Untitled

Optimal viral strategies for bypassing RNA silencing
Guillermo Rodrigo, Javier Carrera, Alfonso Jaramillo and Santiago F. Elena J. R. Soc. Interface , 257-268 first published online 23 June 2010 Email alerting service
Receive free email alerts when new articles cite this article - sign up in the box at the topright-hand corner of the article or click J. R. Soc. Interface J. R. Soc. Interface (2011) 8, 257–268 Published online 23 June 2010 Optimal viral strategies for bypassing Guillermo Rodrigo1, Javier Carrera1,2, Alfonso Jaramillo3,4 and Santiago F. Elena1,5,* 1Instituto de Biologı´a Molecular y Celular de Plantas, Consejo Superior de Investigaciones Cientı´ficas-Universidad Polite´cnica de Valencia, Campus UPV CPI 8E, Ingeniero Fausto Elio s/n, 46022 Valencia, Spain 2ITACA, Universidad Polite´cnica de Valencia, Camino de Vera s/n, 46022 Valencia, Spain 3E´cole Polytechnique, Route de Saclay, 91128 Palaiseau Cedex, France 4Epigenomics Project, Genopole-Universite´ d'E´vry Val d'Essonne-CNRS UPS3201, Batiment Geneavenir 6, 5 Rue Henri Desbrue res, 91030 E ´ vry Cedex, France 5Santa Fe Institute, 1399 Hyde Park Road, Santa Fe, NM 87501, USA The RNA silencing pathway constitutes a defence mechanism highly conserved in eukaryotes,especially in plants, where the underlying working principle relies on the repressive actiontriggered by the intracellular presence of double-stranded RNAs. This immune system per-forms a post-transcriptional suppression of aberrant mRNAs or viral RNAs by smallinterfering RNAs (siRNAs) that are directed towards their target in a sequence-specificmanner. However, viruses have evolved strategies to escape from silencing surveillancewhile promoting their own replication. Several viruses encode suppressor proteins that inter-act with different elements of the RNA silencing pathway and block it. The differentsuppressors are not phylogenetically nor structurally related and also differ in their mechan-ism of action. Here, we adopt a model-driven forward-engineering approach to understand theevolution of suppressor proteins and, in particular, why viral suppressors preferentially targetsome components of the silencing pathway. We analysed three strategies characterized bydifferent design principles: replication in the absence of a suppressor, suppressors targetingthe first protein component of the pathway and suppressors targeting the siRNAs. Our resultsshed light on the question of whether a virus must opt for devoting more time into transcrip-tion or into translation and on which would be the optimal step of the silencing pathway to betargeted by suppressors. In addition, we discussed the evolutionary implications of suchdesigning principles.
Keywords: RNA silencing; silencing suppression; systems and synthetic biology; transcription – translation tradeoff; virus evolution; virus – host interaction The underlying working principle of RNA silencing relies on the repressive action triggered by the intra- RNA viruses are difficult to control and eliminate cellular presence of double-stranded RNAs (dsRNAs) because of their rapid evolution. This high evolvabil- []. In the case of single-stranded RNA (ssRNA) ity is a consequence of their high mutation rates, viruses, dsRNAs are by-products of genome replication large population size and short generation times mediated by virus-encoded RNA-dependent RNA ] that confer them an astonishing ability to polymerases (RdRps). During viral genome replication, explore genotypic space. Indeed, RNA viruses typi- the dsRNA intermediates become the target of the first component of the silencing pathway, DICER, a higher than their DNA hosts []. Eukaryotic organ- type-III RNase that degrades these dsRNAs into isms have developed a sequence-specific mechanism units of 21 – 24 nucleotides called small interfering to modulate gene expression based on RNA interfer- RNAs (siRNAs; ]). Subsequently, the cellular RNA-induced silencing complex (RISC), which con- Caenorhabditis elegans ] and later on in many tains the argonaute (AGO) endonuclease loads other eukaryotes, including plants and mammals the antisense siRNAs, resulting in an active form.
Likewise, this molecular mechanism is able to Using the antisense siRNA as a guide, AGO cleaves silence viral or aberrant genes.
the target viral ssRNA []. Furthermore, in a second-ary cycle of amplification, the host's RNA-dependent *Author for correspondence ).
RNA polymerase VI (RDR6) uses siRNAs as primers, Received 19 May 2010Accepted 3 June 2010 This journal is q 2010 The Royal Society Bypassing RNA silencing G. Rodrigo et al.
together with partially degraded ssRNAs, to produce on the outcome of the interaction. On the other hand, although many kinetic models of intracellular DICER, a process known as transitivity ]. siRNAs growth have been proposed for different viruses, systemically move from cell-to-cell, immunizing new none of them specifically incorporates the silencing cells against infection ]. Given the properties of response (e.g. [– ]). In this work, we present the the RNA silencing pathway (specificity and amplifica- first model that incorporates the interaction of differ- tion), it represents a sort of innate immune system for ent suppressor proteins with components of the silencing pathway. We perform a dynamical analysis Not surprisingly, viruses have evolved strategies to and show the time course of viral RNA accumulation actively evade the RNA silencing surveillance while under a wide set of parameter states. We also show promoting their own replication ]. Many viruses phase diagrams for different combinations of par- encode a suppressor protein (viral suppressor of ameters and focus our discussion on the behaviour RNA silencing or VSR) that interacts with elements of the system for different viral replication and trans- of the silencing pathway blocking it [– ]. The lation rates in the presence/absence of different targets of these VSRs within the RNA silencing suppressor strategies. These analyses allow us to rationalize why different viruses may opt for different siRNA, RISC or the systemic signal [For strategies in their investment into producing new gen- example, the helper component-protease (HC-Pro) omes (i.e. transcription via antigenomic strains) or encoded by the Potyvirus works as suppressor by into producing large amounts of protein from a few sequestering siRNAs [– ]. This binding prevents initial sense genomes (i.e. translation). Such models the incorporation of siRNAs into the RISC. Further- more, by also binding plant endogenous micro-RNAs design principles of viral systems.
and controlling the expression of other genes, HC-Pro may interfere the expression of DICER proteins[reducing the degradation of dsRNAs and,thus, favouring potyvirus replication. Similarly, the Nodavirus B2 suppressor also sequesters siRNAs[The Tombusviridae P19 and Cucumovirus 2b We have constructed a mathematical model based on suppressors interfere with the systemic spread of the 24 nucleotide siRNAs produced by DCL3 [ Some suppressors act on the RISC, either avoiding positive-sense RNA virus that encodes for a single the upload of siRNAs into AGO, like the Clostero- polyprotein that is processed into mature peptides, virus P21 [by binding to AGO1 and avoiding as is the case for picorna-like viruses (e.g. poliovirus, its interaction with other proteins required to assem- hepatitis C virus, foot-and-mouth disease virus and the potyviruses, which are the largest and more impor- Tombusvirus [by inhibiting the RISC activity tant family of plant viruses). The model involves the after its maturation, like the Begomovirus AC4 following molecular species: genomic and antigenomic [or by targeting AGO for degradation, as it is ssRNA (Sþ and S2, respectively), dsRNA (D), anti- the case for Polerovirus P0 protein ] press).
sense siRNA (I ), viral proteins ( p), virions (V ), It has also been recently shown that the V2 suppres- primed ssRNA (S*) and secondary dsRNA (D*).
sor of Geminivirus competes with SGS3, a key Three different viral proteins are considered, the non- component of the secondary cycle of siRNA amplifi- structural replicase and VSR and the structural CP.
cation, in binding dsRNAs and thus interferes with Their corresponding relative abundances are p, q and transitivity [Finally, the CP of some carmo- 1 2 p 2 q, respectively. This constraint is biologically viruses ] and the P14 of Aureusvirus ] can relevant for picornaviruses as all proteins are self-pro- also bind long dsRNAs, resulting in the protection cessed from a single polyprotein and, thus, their of the intermediaries of replication from DICER relative abundances remain constant during infection.
activity. Accordingly, VSRs have been divided into In addition, the model accounts for several cellular three families [(i) those enhancing within-cell components: the ribosomes (Z ), the RDR6 polymerase virus accumulation, (ii) those essential for cell-to- involved in transitivity (Y ), DICER-like proteins (C ) cell movement but dispensable on virus accumulation and the inactivated and activated RISC (R and R*, in single cells, and (iii) those that facilitate virus respectively). We assume that at the beginning of infection, a single viral ssRNA genome is present, symptoms but are not essential for viral replication which in our particular model must be genomic.
and cell-to-cell movement.
Notice that genomic strands are those that encode The first mathematical models of the RNA silen- for proteins, whereas antigenomic strands are comp- cing pathway focused on aberrant cellular mRNA as lementary and, for simplification, we will assume are triggers of the silencing response – More not coding. To accommodate negative-sense RNA recent models consider viral RNAs as triggers of the viruses into the model, the equations can be straight- response and focused on the spread of viruses in plants ]. However, on the one hand, these encapsidated and cleaved by RISC) changing the studies did not analyse in detail the possible effect initial conditions. For retroviruses or DNA viruses, that different viral suppressor strategies may have the model must be conveniently modified.
J. R. Soc. Interface (2011)

Bypassing RNA silencing G. Rodrigo et al.
Table 1. Values for the kinetic parameters used in the model. Other non-kinetic model parameters are p ¼ q ¼ 0.4, v ¼ 0.1,n ¼ 2n* ¼ 10, f ¼ 0.01, s ¼ 0.1 and k ¼ 10k ¼ 30. The amounts of cellular resources are Z ¼ 105, Y ¼ 105, C0 104 molecules. In the case of a virus encoding a VSR, the corresponding binding constant (GC, GI or GR) takes the value of G.
The cell volume is assumed  10213 l, then 1 nM  100 molecules.
value in the literature 10 h21 for HCV [] 228 h21 in vitro for E. coli [] 25 h21 in vitro for Drosophila melanogaster 100 M21 h21 for nucleation [] 105 M21 h21 for elongation [] 0.06 h21 for HCV 225 nM in vitro for TBSV 8 nM in vitro for D. melanogaster 260 nM in vitro [] 335 nM in vitro for E. coli [] 1000 molecules [,] 10 – 1000 nM in vitro p19, p21 and HC-Pro [] The model is constructed following a generalized enzyme kinetics scheme where both substrates and enzymes are limited in the medium [and there are competitions between different enzymes for the same substrate and different substrates for the same enzyme [This gives a highly coupled formulation.
In , we show the scheme of the RNA silencing pathway, and the kinetic parameters are shown in, with parameter values taken from differentsources.
Viral replication is a process involving multiple reactions aiming to bypass the defence systems ofthe cell. The RNA replication rates (J ), for both Figure 1. Schematic of the RNA silencing pathway and itsinteraction with viral replication. RNA viruses encode for replicase, suppressors of silencing (VSR) and coat proteins.
Three types of suppressors are considered in the scheme: sup- pressors of DICER (I), sequesters of siRNA (II) and suppressors of RISC (III).
where a is the maximum replication rate per mol- ð1  p  qÞP k0 ecule of ssRNA, KP, KR, KZ and KC are the binding constants for the replicase, the activated RISC, the ribosomes and the CP, respectively. The affinity of the replicase for the antigenomic strands is incorporated into the model by the parameter v.
If v ¼ 1, then the RdRp has the same affinity for both strains, whereas v . 1 would imply a larger affinity for the antigenomic strain. By doing so, we geometric (v ¼ 1) to the stamping machine one J. R. Soc. Interface (2011) Bypassing RNA silencing G. Rodrigo et al.
A molecule of dsRNA can be separated into immature virions at a rate given by two ssRNA molecules of complementary polarityat a first-order rate with a constant parameter b lSþ½ð1  p  qÞP=KCk0 Jencapsidation ¼ dissociation ¼ bD In addition, genomic ssRNAs are translated into where l is the maximum assembly rate and k0 , k is viral proteins with rate the number of CP monomers associated to the imma-ture virions, Vimmature. Then, virions are produced at immature½ð1  p  qÞP =KCk f1 þ Vimmature=KC þ ½ðð1  p  qÞP=KCk0gk=k0 ð1  p  qÞP k0 where g is the maximum rate to produce virions, andk is the number of CPs necessary to complete amature virion. All species are thermodynamically where m is the maximum translation rate per molecule degraded at rates kS (ssRNAs), kD (dsRNAs), kI of genomic ssRNA.
(siRNAs) and kP (the rest of proteins or protein The process of RNA silencing is initiated when DICER cleaves dsRNA into siRNAs. The rates describ- The effect exerted by different VSRs on DICER, ing this process are given by the following set of RISC and RDR6 can be conveniently modelled by the following three equations, respectively: 0ð1 þ fqP =GRÞ 1 þ ðD þ D þ C Þ=K 0ð1 þ fqP =GYÞ > where d and KD are the catalytic and binding con- stants of DICER, respectively. Afterwards, the RISCis activated by uploading the antisense siRNAs pro- where C0, R0 and Y0 are the corresponding amounts of each protein in the cell, which are assumed to be in large excess, and GC, GR and GY are the bindingcoefficients of the corresponding VSR to their sub-strate respectively. The parameter f determines the effi- 1 þ ðI þ RÞ=KI ciency at which the suppressor precludes the activityof its target. For example, in the equation for where r and KI are the catalytic and binding DICER, an f ¼ 0.01 means that even at saturating the activation of the RISC, it is now capable DICER molecules will still be active. To account for of directing the cleavage of the viral ssRNA with the suppression on siRNA, we modify JRISC and introduce a new equation to model the sequestrationof siRNAs.
1 þ ðI þ RÞ=K ð1  p  qÞP k0 1 þ ðI þ qPÞ=KI þ R=KI where r and c are, respectively, the rates at which the RISC and the suppressor attach to the siRNA and GI, the binding affinity of the suppressor forthe siRNAs.
where y is the catalytic constant of RNA cleavage.
After defining all the relevant rate equations, it CPs are pre-assembled with ssRNA to produce J. R. Soc. Interface (2011) Bypassing RNA silencing G. Rodrigo et al.
differential equations describing the dynamics of the The vector of steady states is given by F(y1) ¼ 0, which serves to calculate the asymptotic behaviour of the system through the eigenvalues of its Jacobian dissociation þ J  1). The behaviour can change significantly by modifying pivotal parameters of the system. Thus, the construction of bifurcation diagrams is a useful encapsidation  kSS þ; tool for evaluating the behaviour regimes under differ- ent conditions, and also to build up a sensitivity dissociation  J  analysis of the parameters of the system.
We show that the trivial solution of the system (i.e.
translation  kPP ; silenced virus) is stable. The Jacobian matrix evaluated 1 ¼ 0 is given by dissociation  kDD; Jsuppression  kII ; d0 ¼ dC0/KD, r0 ¼ rR0/KI, m0 ¼ mZ/KZ, a00 ¼ a0 þ kS, d00 ¼ d0 þ b þ kD and r00 ¼ r0 þ I. This Jacobian has five negative real eigenvalues encapsidation  Jvirion  kPVimmature > (2a00, 2r00, 2d00, 2k S and 2kP) that represent an asymptotically stable solution of the system. Three of them have multiplicity greater than one. The system also has a second non-trivial solution in which the virus beats the silencing response and replicates and where the stoichiometric parameters n, n* and s accumulates in the cell. Although we have verified represent, respectively, the number of siRNAs pro- numerically the existence of this non-trivial solution on the full model, without lost of generality, the stab- siRNAs produced in the secondary cycle of amplifi- ility analysis for this second solution can be done analytically by simplifying the system (2.11) as secondary siRNA amplification to the degradation of dsRNA relative to the primary siRNAs.
The full model in the Matlab format is available in the electronic supplementary material.
3. STABILITY ANALYSIS ¼ ndD  yRS  k The system (2.11) can be rewritten in a vectorial formas dy/dt ¼ F(y) ¼ VJ(y) 2 Jy, where V is the where the non-trivial steady state is the solution of S matrix of stoichiometric coefficients, J(y) the vector (b 2 d)/(b þ d) ¼ kS þ nday S2/(b þ d) (y S þ kP).
of production rates and J a diagonal matrix with the The characteristic polynomial is 2X3 þ tX2 2 hX 2 vector of degradation rates. The initial condition for the molecular species involved in the system (y0) the trace of the Jacobian matrix, h ¼ (2aS þ y R þ depends on the nature of the virus (i.e. the infectious kS)(b þ dþ yS þ kP) þ (y S þ kP) (b þ d) 2 4abS 2 particle containing a genomic or an antigenomic RNA y 2RS is the trace of its adjoint matrix and D ¼ strand). Here, we have considered for our analyses [4abS 2 (2aS þ y R þ kS) (b þ d)] (y S þ kP) 2 2nad viruses encapsidating genomic RNAs and therefore all y S2 þ y 2RS(b þ d) its determinant. By applying the the elements in y0 are zero except for Sþ ¼ 1. In case Routh – Hurwitz stability criterion, the system will be of negative-sense RNA viruses, the initial condition stable when t , 0, D , 0 and ht , D. Henceforth, by would be S2 ¼ 1. Accordingly, we construct an initial taking the appropriate kinetic parameters that meet value problem to obtain the dynamics of the system.
these three conditions, the system is characterized by J. R. Soc. Interface (2011)

Bypassing RNA silencing G. Rodrigo et al.
Figure 2. Dynamics of viral infection for different initial conditions. (a) The starting condition of the simulation is a single viralgenome; this results in the virus being silenced. (b) The starting condition is that 10 viral genomes infect the cell; this high mul-tiplicity of infection results in exponential viral replication after a period of latency of 1 day required to reach a threshold level ofRdRps. This successful infection happens even in the absence of a VSR. The parameters values are those shown in bistability and, therefore, the initial condition is pivotal In fact, this can be rationalized because viral RdRps to determine the outcome of the process.
compete with ribosomes and with the activated RISCfor genomic strands, whereas they do not compete forantigenomic strands. In addition, high replication rates also allow the virus to escape from the silencingmachinery and to minimize the effect of non-specific 4.1. Virus replication in the absence of a VSR thermodynamic degradation (b).
We have studied the viral replication dynamics by using One question that arises here is whether a tradeoff the mathematical model presented in the previous sec- between replication and translation exists. Upon tion. First, we considered the case of RNA viruses uncoating and the strictly necessary first event of trans- that do not encode suppressor proteins. In we lation, a viral genome can be directed either to show several time-course evolutions of the system transcription, and thus increase the concentration of species (Sþ, S2 and P) for two different sets of initial RNA, or to translation, and thus increase the concen- conditions. When the multiplicity of infection is low tration of viral proteins (in this case only replicase (one single viral Sþ genome per cell) and for the typical and coat). In c, we analysed such tradeoff by parameter values shown in , we show that the considering the binding affinities to positive strands of population is extinguished (a), after a transient replicase (KP) and ribosomes (KZ). We showed that in where the concentration of P reaches a maximum. The the absence of a silencing suppressor, silencing is the model predicts that in this situation, the amount of outcome favoured when translation is more frequent antigenomic strains S2 produced is meaningless and than transcription (KP , KZ). Accordingly, the best its dynamics is dominated by the degradation term in strategy for a virus to bypass the RNA silencing the system of equations (2.11).
response in the absence of a suppressor protein would However, the virus can bypass the silencing mechan- be to increase the affinity of its RNA to the replicase ism if the multiplicity of infection just increases to Sþ ¼ rather than to optimize its binding affinity to the ribo- 10 molecules b). In this case, after a latency period of about 1 day, viral proteins reach a critical con- efficiency, a virus will produce more copies of its centration and promote further exponential replication.
genome up to the point in which the cleavage by Analytically, the latency period can be estimated when DICER would no longer control the accumulation of viral genomes. shows, as expected, that the P. In all these simulations, the condition Sþ . S2 holds, in excellent agreement with the obser- higher the catalytic constants for transcription and vation of an excess of sense siRNAs for positive-sense translation, the higher the chances for a successful viral genomes [The effect of further increasing the viral replication.
multiplicity of infection is to reduce the latency period(data not shown).
4.2. Virus replication dynamics in presence of a We performed several sensitivity analyses to study VSR that acts on DICER the regions in parameter space in which viral replicationoccurs (non-trivial solution) or for which viral silencing Many, if not all, viruses encode proteins capable of takes place (trivial solution). We found that the higher interacting with the cell molecular machinery. The sup- the affinity for the negative strand (lower v), the wider pression mechanism is often a protein – protein or is the parameter space for viral replication ).
RNA – protein interaction resulting in a sequestration J. R. Soc. Interface (2011)

Bypassing RNA silencing G. Rodrigo et al.
m = 10 m = 20 m = 50 κS (h–1) KP = 5×10 KP (molec) Figure 3. Phase diagrams identify different viral strategies. (a) The effect of the catalytic constant of DICER cleavage (d) in thereplication rate (a) and the differential affinity of RdRps for positive and negative strands (v). (b) The relationship between aand the ssRNA degradation rate (kS) for different values of the translation rate (m). (c) The sensitivity of the binding constants ofribosomes (KZ) and replicases (KP) to a. (d ) The effect of KP on m and a. Rep means viral replication bypassing silencing, and Silviral extinction by silencing.
or blockage of one of the many molecules involved in necessary for completing a virion as a function of the the silencing pathway that allows the virus to escape cellular amounts of DICER (C0). For low amounts of from silencing surveillance. Our general model can be DICER, TV is insensitive to variation in GC. In used to analyse and study the effect of various suppres- addition, an increase in the number of DICER mol- sors encoded by different viruses. To analyse the effect ecules per cell does not have any effect on TV for of a suppressor, we considered the virus replication suppressors with weak affinity. However, if GC increases speed as a characteristic scoring function. This speed (moving rightwards in the ordinates axis in ), can be easily computed as the inverse of the time then the time to produce virions significantly grows taken to produce mature virions (TV). In , we up and becomes infinity (indicating viral silencing) for plot 1/TV versus KZ and KP for the case of a VSR oper- high amounts of DICER molecules present in the cell ating over DICER. We found that such a suppressor at the time of infection.
enhances the speed of virus accumulation with respectto a virus without encoding a VSR.
4.3. The effect of suppressing downstream steps To further analyse the suppressor strategy of manip- of the silencing pathway ulating DICER, we constructed a phase diagrambetween the catalytic constant of cleavage by DICER Next, we sought the effect of VSRs operating down- stream in the silencing pathway. Surprisingly, we a). We found that the effect of the suppressor found that suppressors affecting at other levels of the is only significant beyond a threshold level of GC (in pathway (e.g. sequestering siRNAs, interfering with this case 7000 molecules). In other words, if the affinity RISC or with RDR6) did not enlarge the parameter of the suppressor is not high enough, it only represents a space in which the virus successfully replicates within cost for the virus because it cannot help in its replica- a single cell (data not shown). This result suggests tion. b shows the effect that the binding that only by suppressing DICER, the first bottleneck affinity of the suppressor for DICER has on the time to replication imposed by the system, viruses could J. R. Soc. Interface (2011)

Bypassing RNA silencing G. Rodrigo et al.
suppressor sensitivity GC (molec) KZ (molec) KP (molec) Figure 4. Virus replication speed (computed as the inverse of the time to form a mature virion, TV) versus the binding con- stants of ribosomes (K C = 8 × 10 Z) and replicases (KP), with GC molecules. (a) Virus without a VSR. (b) Virus encoding a sup- pressor that blocks DICER. The benefit associated with C0 = 5 × 103 carrying such a suppressor is evaluated as the differencebetween both surfaces and is indicated by the dashed lineand the arrow. The other parameters take the values shown widen the parameter region, resulting in successfulreplication. Hence, the question is why other types of GC (molec) VSRs, such as siRNA sequesters, have evolved? Our Figure 5. (a) Phase diagram to analyse the suppressor effect negative result suggests that the RNA silencing mode on DICER between d and G of action cannot be rationalized by only looking into a C, with a ¼ 20 h21 for different values of m. (b) Time to form one virion (TV) versus the sup- single cell but that a more complex situation in which pressor constant of DICER, with a ¼m ¼ 20 h21 for different cell-to-cell effects may contribute should be considered.
values of C0 (in molecules). The other parameters take the This leads us to consider the role of the space to analyse values shown in . Rep means viral replication bypassing such mechanism.
silencing, and Sil viral extinction by silencing.
In a, we plot the relative amount of accumu- lated siRNAs (normalized by the amount or siRNAproduced in the absence of a VSR, I/IG!1), in the pres- ence of two suppression strategies. For illustrativepurposes, we have chosen the successful operation We have presented a deterministic model of the inter- over DICER described in the previous section and one play between viral replication and the RNA silencing based on sequestering siRNAs. By increasing the affi- pathway. For the sake of biological realism, we modelled nity for the corresponding target molecule (moving a particular type of virus, the picorna-like. By doing so, rightwards on the ordinate axis) to the maximum the model pays the cost of reduced generality and the conclusions may not be applicable to viruses with DICER reduces the concentration of siRNA around other genomic architectures such as negative-sense two orders of magnitude. However, the strategy based RNA, retroviruses or DNA viruses. Although our results on sequestering siRNAs is far less efficient since at the have been performed for positive-sense RNA viruses, strongest affinity it only reduces the accumulation of the model can also be used to study negative-sense virus-derived siRNAs by one order of magnitude.
viruses with minor changes in some rates and the initial However, the transfer of siRNAs from infected to conditions. Readers interested in exploring the inter- neighbouring healthy cells, which allows the peripheral play between the silencing pathway and any of these cells to activate the RISC in the absence of viral infec- viruses must necessarily look at this article as the start- tion, has the expected effect b). In the ing point for developing their own models. Nonetheless, absence of triggering siRNAs, infection progresses our approximation has allowed us to study and compare with the time delay already described above. However, different viral suppression strategies. We have shown if the cell has been already activated, the virus is not that the RNA silencing pathway allows a large variety able to overcome the cleavage by the RISC and runs of behaviours, suggesting multiple potential evolution- to extinction.
ary trajectories for RNA viruses. Future models will J. R. Soc. Interface (2011) Bypassing RNA silencing G. Rodrigo et al.
because one may expect more replication to generate more dsRNA and, therefore, to strength the silencingresponse and, likewise, more translation to producemore suppressor protein. It can be argued that, afterthe very initial burst of translation from the infecting genomic sequence resulting in a few viral proteins, the I/I G optimal strategy involves synthesizing antigenomic strands and using them as templates for producing alarge excess of genomic strands (i.e. using a stampingmachine replication strategy) without diverting theminto translation. If replication is fast enough, this repli-cative strategy works even in the absence of a suppressor protein: a positive feedback is establishedsuch that the replication overcomes the capacity of the available DICER molecules to keep virus replication under control. Once a significant amount of genomic strands has been produced, then translation may take place. If translation results in a VSR protein, then a synergistic effect between fast transcription and trans-lation appears, resulting in successful viral replication.
Among many possibilities, we have focused on three viral strategies. The first one, consisting of blockingDICER, turns out to be the most efficient promoting viral replication. This result is somehow logical from an optimal design perspective. By hitting the first bot- tleneck in the pathway, the virus ensures its own R*(0) = 10 replication. Hitting downstream steps would allowDICER to still exert partial control on virus replication.
The other three strategies explored, sequestering siRNA, blocking the RISC and disrupting the second- ary amplification via RDR6, have been less efficient in promoting intracellular virus accumulation, althoughthey may gain some benefit when looking at cell-to- Figure 6. (a) Amount of siRNA (relative to the amount accu- cell movement. This finding is in good agreement with mulated without a viral suppressor of RNA silencing) versus the observation that Cucumovirus 2b and Tombusvirus the suppressor constant (G) on DICER or siRNA. (b) Viral RNA dynamics in a cell which has not been immunized by receiving siRNA from neighbouring cells (R*(0) ¼ 0) and ina cell that has received a small input of siRNA from an accumulation ].
infected neighbour cell (R*(0) ¼ 10 molecules). The par- Although mathematically convenient, it may be a ameters take the values shown in expect a ¼ 50 h21.
biological oversimplification to assume that suppressorsact at a single stage of the silencing pathway. Evidenceexists showing that VSRs may well simultaneously account for different viral genomic organizations and operate at diverse stages of the pathway. For example, for inherent stochastic effects associated with small the potyviral HC-Pro sequesters siRNAs but also affects numbers of molecules ]. The model presented here the expression of plant genes, including the dcl-like differs from other models of the interaction between genes encoding for the different DICER proteins in Ara- virus and the host silencing response [in which bidopsis thaliana [or by reducing the 30 here we have explored the role played by different methylation of siRNAs, making them sensitive to oli- suppressors of RNA silencing. We have demonstrated and shown in that the system has two stable Another example of multiple actions is the Polerovirus steady states (replication and silencing) and, thus, the P0 that interferes with the silencing pathway at least initial condition of the system (i.e. the initial amount at two levels: binding siRNAs and avoiding the for- of ssRNA in the cell) is important to determine its mation of the activated AGO complex and labelling it dynamics. Likewise, the higher the initial amount of for degradation ([press). Also, a virus may viral RNA, the higher the zone for exponential viral carry more than one VSR, as seems to be the case for replication in the parameter space. This suggests that some tombusviruses (P19 and CP).
increasing the multiplicity of infection is a possible We have also found that in certain regions of par- strategy for virus to escape from the control of RNA ameter space, a virus would be capable of replicating even in the absence of a VSR. The plant subviral patho- We have shown that in the presence of an active gens known as viroids do not encode for any protein at silencing response, it is to the benefit of the virus to all and are still capable of replication in susceptible invest into a transcriptional strategy rather than in hosts ], despite the fact that their RNA molecules translation. This may be somehow counterintuitive are targets of DICER [It has been suggested that J. R. Soc. Interface (2011) Bypassing RNA silencing G. Rodrigo et al.
viroids may evade silencing because of their highly com- 4 Fire, A., Xu, S., Mongomery, M. K., Kostas, S. A., Driver, plex and packed secondary structure [Other S. E. & Mello, C. C. 1998 Potent and specific genetic inter- strategies viruses may use for avoiding silencing consist ference by double-stranded RNA in Caenorhabditis in replicating within spherules in the endoplasmic reti- elegans. Nature 391, 806 – 811. () culum membrane ], where they remain inaccessible 5 Vaucheret, H., Beclin, C. & Fagard, M. 2001 Post-tran- scriptional gene silencing in plants. J. Cell Sci. 14, 3083 – 3091.
Although we have modelled the effect of VSRs on 6 Elbashir, S. M., Harborth, J., Lendeckel, W., Yalcin, A., DICER as a direct protein – protein interaction, VSRs Weber, K. & Tuschl, T. 2001 Duplexes of 21-nucleotide can also interfere with DICER activity by protecting RNAs mediate RNA interference in cultured mammalian the dsRNA as it is produced as reported for CP [ cells. Nature 411, 494 – 498. This particular activity would be easily incorporated 7 Hamilton, A. J. & Baulcombe, D. C. 1999 A species of into our mathematical framework in two simple ways.
small antisense RNA in posttranscriptional gene silencing First, by treating the binding affinity of DICER for in plants. Science 286, 950 – 952. ( D) as a decreasing function of CP concen- tration. Second, by defining a new molecular species 8 Vermeulen, A., Behlen, L., Reynolds, A., Wolfson, A., Marshall, W. S., Karpilow, J. & Khvorova, A. 2005 The for the complex dsRNA/CP and writing down the cor- contributions of dsRNA structure to DICER specificity responding rates and adding a new differential equation and efficiency. RNA 11, 674 – 682. in equation (2.11).
In conclusion, we have shown that from a system 9 Bohmert, K., Camus, I., Bellini, C., Bouchez, D., Caboche, design perspective, the best strategy that a virus may M. & Benning, C. 1998 AGO1 defines a novel locus of Ara- take to ensure its replication in the presence of the anti- bidopsis controlling leaf development. EMBO J. 17, 170 – viral response mediated by RNA silencing would be to (i) replicate fast and by producing an excess of genomic 10 Rand, T., Petersen, S., Du, F. & Wang, X. 2005 Argo- strands, (ii) encode for a VSR that interacts with the naute2 cleaves the anti-guide strand of siRNA during DICER protein, and (iii) exert some control on the mul- RISC activation. Cell 123, 621 – 629. ( tiplicity of infection, ensuring that multiple genomes 11 Voinnet, O., Vain, P., Angell, S. & Baulcombe, D. C. 1998 infect each cell. Obviously, evolution is not a perfect Systemic spread of sequence specific transgene RNA designer and viruses have acquired suppressor proteins degradation is initiated by localized introduction of ecto- that target at different steps of the silencing pathway.
pic promoter less DNA. Cell 95, 177 – 187. Understanding the exact mechanisms by which these VSRs operate will allow us to develop better models 12 Himber, C., Dunoyer, P., Moissiard, G., Ritzenthaler, C. & and to increase our ability to predict the outcome of Voinnet, O. 2003 Transitivity-dependent and -independent the virus – host interaction. Furthermore, VSRs have cell-to-cell movement of RNA silencing. EMBO J. 22, clear biotechnological potential as they can be used to maximize the expression of transgenes ]. Designing 13 Lecellier, C. H. & Voinnet, O. 2004 RNA silencing: no optimal suppressors would benefit from the knowledge mercy for viruses? Immunol. Rev. 1998, 285 – 303.
14 Ding, S. & Voinnet, O. 2007 Antiviral immunity directed advanced in this article.
by small RNAs. Cell 130, 413 – 426. ( This work was supported by the Spanish Ministerio de Ciencia e Innovacio´n grants BFU2009-06993 to S.F.E. and 15 Li, F. & Ding, S. W. 2006 Virus counterdefenses: diverse strategies for evading the RNA-silencing immunity.
(BioModularH2), FP7-ICT-043338 (Bactocom), FP7-KBBE- 503 – 531.
212894 (Tarpol), the Structural Funds of the European Regional Development Fund, the ATIGE-Genopole and the 16 Brigneti, G., Voinnet, O., Li, W. X., Ji, L. H., Ding, S. W. & Foundation pour la Recherche Medicale grants (all to A.J.).
Baulcombe, D. C. 1998 Viral pathogenicity determinants are J.C, G.R. and A.J. also acknowledge the HPC-Europa suppressors of transgene silencing in Nicotiana benthamiana.
programme (RII3-CT-2003-506079). G.R. was supported by EMBO J. 17, 6739–6746. ) a graduate fellowship from the Generalitat Valenciana and 17 Baulcombe, D. 2004 RNA silencing in plants. Nature 431, an EMBO Short-term fellowship. S.F.E. also acknowledges support from the Santa Fe Institute.
18 Kasschau, K. D., Xie, Z., Allen, E., Llave, C., Chapman, E. J., Krizan, K. A. & Carrington, J. C. 2003 P1/HC-Pro, a viral suppressor of RNA silencing, interfereswith Arabidopsis development and miRNA function.
Dev.
205 – 217.
1 Domingo, E. & Holland, J. J. 1997 RNA virus mutations 19 Moissiard, G. & Voinnet, O. 2004 Viral suppression of and fitness for survival. Annu. Rev. Microbiol. 51, 151 – RNA silencing in plants. Mol. Plant Pathol. 5, 71 – 82.
2 Elena, S. F. & Sanjua´n, R. 2007 Virus evolution: insights 20 Dı´az-Pendo´n, J. A. & Ding, S. W. 2008 Direct and indirect from an experimental approach. Annu. Rev. Ecol. Evol.
roles of viral suppressors of RNA silencing in pathogenesis.
Syst. 38, 27 – 52. Annu. Rev. Phytopathol. 46, 303 – 326. 3 Drake, J. W. & Holland, J. J. 1999 Mutation rates among 21 Mallory, A. C., Reinhart, B. J., Bartel, D., Vance, V. B. & RNA viruses. Proc. Natl Acad. Sci. USA 96, 13 910 – Bowman, L. H. 2002 A viral suppressor of RNA silencing differentially regulates the accumulation of short interfering J. R. Soc. Interface (2011) Bypassing RNA silencing G. Rodrigo et al.
RNAs and microRNAs in tobacco. Proc. Natl Acad. Sci.
amplification limits accidental self-directed reactions.
USA 99, 15 228 – 15 233. ( Proc. Natl. Acad. Sci. USA 100, 11 511 – 11 516. ( 22 Chapman, E. J., Prokhnevsky, A. I., Gopinath, K., Dolja, V. & Carrington, J. C. 2004 Viral RNA silencing suppres- 37 Raab, R. M. & Stephanopoulos, G. 2004 Dynamics of gene sors inhibit the microRNA pathway at an intermediate silencing by RNA interference. Biotechnol. Bioeng. 88, step. Genes Dev. 18, 1179 – 1186. ( 38 Groenenboom, M. A. C., Mare´e, A. F. M. & Hogeweg, P.
23 Dunoyer, P., Lecellier, C. H., Parizotto, E. A., Himber, C.
2005 The RNA silencing pathway: the bits and pieces & Voinnet, O. 2004 Probing the microRNA and small that matter. PLoS Comp. Biol. 1, e21. interfering RNA pathways with virus-encoded suppressors of RNA silencing. Plant Cell 16, 1235 – 1250. 39 Groenenboom, M. & Hogeweg, P. 2008 The dynamics and efficacy of antiviral RNA silencing: a model study. BMC 24 Lakatos, L., Szittya, G., Silhavy, D. & Burgyan, J. 2004 Syst. Biol. 2, 28. Molecular mechanism of RNA silencing suppression 40 Groenenboom, M. & Hogeweg, P. 2008 RNA silencing can mediated by p19 protein of tombusviruses. EMBO J. 23, explain chlorotic infection patterns on plant leaves. BMC Syst. Biol. 2, 105. () 25 Mlotshwa, S. et al. 2005 Ectopic DICERLIKE1 expression 41 Endy, D., Kong, D. & Yin, J. 1996 Intracellular kinetics of in P1/HC-Pro Arabidopsis rescues phenotypic anomalies a growing virus: a genetically structured simulation for but not defects in microRNA and silencing pathways.
bacteriophage T7. Biotechnol. Bioeng. 55, 375 – 389.
Plant Cell 17, 2873 – 2885. ) 26 Deleris, A., Gallego-Bartolome, J., Bao, J., Kasschau, K. D., Carrington, J. C. & Voinnet, O. 2006 Hierarchical 42 Reddy, B. & Yin, J. 1999 Quantitative intracellular kin- action and inhibition of plant DICER-like proteins in anti- etics of HIV type 1. AIDS Res. Hum. Retrovir. 15, 273 – viral defense. Science 313, 68 – 71. ( 43 Srivastava, R., You, L., Summers, J. & Yinn, J. 2002 Sto- 27 Qi, Y., Zhong, X., Itaya, A. & Ding, B. 2004 Dissecting RNA chastic vs. deterministic modeling of intracellular viral silencing in protoplasts uncovers novel effects of viral sup- kinetics. J. Theor. Biol. 218, 309 – 321. ( pressors on the silencing pathway at the cellular level.
Nucleic Acids Res. 32, e179. 44 Sidorenko, Y. & Reichl, U. 2004 Structured model of influ- 28 Peremyslov, V. V., Hagiwara, Y. & Dolja, V. V. 1999 enza virus replication in MDCK cells. Biotechnol. Bioeng.
of a plant virus. Proc. Natl Acad. Sci. USA 96, 14 771 – 45 Lim, K., Lang, V., Lam, T. & Yin, J. 2006 Model-based design of growth-attenuated viruses. PLoS Comput. Biol.
29 Azevedo, J. et al. 2010 Argonaute quenching and global changes in Dicer homeostasis caused by a pathogen- 46 Dahari, H., Ribeiro, R. M., Rice, C. M. & Perelson, A. S.
encoded GW repeat protein. Genes Develop. 24, 904 – 2007 Mathematical modeling of subgenomic hepatitis C virus replication in Huh-7 cells. J. Virol. 81, 750 – 760.
30 Vanitharani, R., Chellappan, P., Pita, J. S. & Fauquet, C.
M. 2004 Differential roles of AC2 and AC4 of Cassava 47 Sardanye´s, J., Sole´, R. V. & Elena, S. F. 2009 Replication Geminiviruses in mediating synergism and suppression of mode and landscape topology differentially affect RNA posttranscriptional gene silencing. J. Virol. 78, 9487 – virus mutational load and robustness. J. Virol. 83, 12 579 – 12 589. ) 31 Baumberger, N., Tsai, C. H., Lie, M., Havecker, E. & Baul- 48 DeAngelis, D. L., Goldstein, R. A. & O'Neill, R. V. 1975 A combe, D. C. 2007 The polerovirus silencing suppressor P0 model for trophic interaction. Ecology 56, 881 – 892.
targets ARGONAUTE proteins for degradation. Curr.
Biol. 17, 1609–1614. 49 MacRae, I. J., Zhou, K. & Doudna, J. A. 2007 Structural 32 Csorba, T., Lo´zsa, R., Hutva´gner, G. & Burgya´n, J. 2010 determinants of RNA recognition and cleavage by Polerovirus protein P0 prevents the assembly of small DICER. Nat. Struct. Mol. Biol. 14, 934 – 940. ( RNA containing RISC complexes and leads to degradation of ARGONAUTE1. Plant J. 62, 463 – 472. ( 50 Qi, X., Bao, F. S. & Xie, Z. 2010 Small RNA deep sequen- cing reveals role for Arabidopsis thaliana RNA-dependent 33 Fukunaga, R. & Doudna, J. A. 2009 dsRNA with 50 over- RNA polymerases in viral siRNA biogenesis. PLoS ONE 4, hangs contributes to endogenous and antiviral RNA silencing pathways in plants. EMBO J. 28, 545 – 555.
51 Gillespie, D. T. 1977 Exact stochastic simulation of coupled chemical reactions. J. Phys. Chem. 81, 2340 – 34 Meng, C., Chen, J., Peng, J. & Wong, S. M. 2006 Host- induced avirulence of Hibiscus chlorotic ringspot virus 52 Ebhart, H. A., Thi, E. P., Wang, M. B. & Unrau, P. J.
mutant correlates with reduced gene-silencing suppression 2005 Extensive 30 modification of plant small RNAs is activity. J. Gen. Virol. 87, 451 – 459. modulated by helper component-proteinase expression.
Proc. Natl Acad. Sci. USA 102, 13 398 – 13 403.
35 Me´rai, Z., Kere´ncyi, Z., Molna´r, A., Barta, E., Bisztray, 53 Me´rai, Z., Kere´nyi, Z., Kerste´sz, S., Magna, M., Lakatos, G., Havelda, Z., Burgya´n, J. & Silhavy, D. 2005 Aureus- L. & Silhavy, D. 2006 Double-stranded RNA binding virus P14 is an efficient RNA silencing suppressor that may be a general plant RNA viral strategy to suppress binds double-stranded RNAs without size specificity.
RNA silencing. J. Virol. 80, 5747 – 5756. ( J. Virol. 79, 7217 – 7226. 54 Daro s, J. A., Elena, S. F. & Flores, R. 2006 Viroids: an 36 Bergstrom, C. T., McKittrick, E. & Antia, R. 2003 Math- Ariadne's thread into the RNA labyrinth. EMBO Rep.
J. R. Soc. Interface (2011) Bypassing RNA silencing G. Rodrigo et al.
55 Di Serio, F., Gisel, A., Navarro, B., Delgado, S., Martı´nez 60 Haley, B. & Zamore, P. D. 2004 Kinetic analysis of the de Alba, A. E., Donvito, G. & Flores, R. 2009 Deep RNAi enzyme complex. Nat. Struct. Mol. Biol. 11, 599 – sequencing of the small RNAs derived from two sympto- matic variants of a chloroplastic viroid: implications for 61 Endres, D. & Zlotnick, A. 2002 Model-based analysis of their genesis and for pathogenesis. PLoS ONE 4, e7539.
assembly kinetics for virus capsids or other spherical poly- mers. Biophys. J. 83, 1217 – 1230. 56 Wang, M. B. et al. 2004 On the role of RNA silencing in the pathogenicity and evolution of viroids and viral satel- 62 Rana, T. M. 2007 Illuminating the silence: understanding lites. Proc. Natl Acad. Sci. USA 101, 3275 – 3280. ( the structure and function of small RNAs. Mol. Cell. Biol.
8, 23 – 36.
57 Go´mez, G. & Palla´s, V. 2007 Mature monomeric forms of 63 Rajendran, K. S. & Nagy, P. D. 2006 Kinetics and func- Hop stunt viroid resist RNA silencing in transgenic plants.
tional studies on interaction between the replicase Plant J. 51, 1041 – 1049. proteins of Tomato bushy stunt virus: requirement of p33: p92 interaction for replicase assembly. Virology 345, 58 Schwartz, M., Chen, J., Janda, M., Sulivan, M., den Boon, J. & Ahlquist, P. 2002 A positive-strand RNA virus repli- 64 Scheper, G. C., van Kollenburg Dagger, B., Hu, J., Luo, cation complex parallels form and function of retrovirus Y., Goss, D. J. & Proud, C. G. 2002 Phosphorylation of capsids. Mol.
9, 505 – 514.
eukaryotic initiation factor 4E markedly reduces its affi- nity for capped mRNA. J. Biol. Chem. 277, 3303 – 3309.
59 Sun, W., Jun, E. & Nicholson, A. W. 2001 Intrinsic 65 Lakatos, L. et al. 2006 Small RNA binding is a common Escherichia coli ribonuclease III lacking the dsRNA-bind- strategy to suppress RNA silencing by several viral sup- ing domain. Biochemistry 40, 14 976 – 14 984. ( pressors. EMBO J. 25, 2768 – 2780. J. R. Soc. Interface (2011)

Source: http://jaramillolab.issb.genopole.fr/download/attachments/589899/Rodrigo-Optimal+viral+strategies+for+bypassing+RNA+silencing.-2011.pdf