Biochemical and genetic analysis of viral replication has identified roles for many host-encoded RNA helicases. Specifically, several helicases have been identified by analyzing global changes in host mRNA expression following viral infection28,29
and small interfering RNA (siRNA) screens to identify genes required for viral replication.30
For example, following HIV-1 infection, changes in gene expression are observed for dhx9, ddx10, ddx18, ddx21, ddx23, ddx39
The expression levels of cytosolic sensors ddx58/RIG-I, mda-5
, are increased by infection with Dengue virus, yellow fever virus, vaccinia virus, Hantaan virus and Sin nombre virus, consistent with induction of the innate antiviral response.22
The ubiquitous involvement of cellular RNA helicases in virus replication is attributable to their ability to as act a scaffold for alternative protein-protein interactions. While some host-encoded RNA helicases, such as RIG-I and MDA-5, induce the cellular antiviral response and limit viral replication, eight host-encoded helicases have been shown to promote virus replication. The following sections describe the structure of these helicases and their role in the virus life cycle.
DDX3 is a 73 kD nucleo-cytoplasmic DEAD helicase that is ubiquitously expressed in a wide range of tissues.31–33
DDX3 has multiple activities in the cell, which involve ATPase activity and ability to act as a scaffold for alternative protein-protein interactions. The central helicase and DEAD domains are flanked at the N-terminus by a protein-protein interaction domain and a putative nuclear export signal and a glycine-rich C-terminus (). DDX3 is highly conserved in metazoans (). Ded1 is the yeast homolog and is essential for general translation initiation.34
Ded1 was identified in a yeast genetic screen and the ded1-deficiency can be rescued by human DDX3, implicating DDX3 in translation of metazoan RNAs.34,35
Since then, metazoan DDX3 has been observed to modulate gene expression at the transcriptional and post-transcriptional levels, both of which affect the broader processes of cell growth regulation, apoptosis and the innate response to virus infection.32,36,37
Members of four virus families use DDX3 to promote their life cycle ().
Significant conservation is observed amongst RNA helicases involved in virus biology
Overview of the role of cellular helicases in replication cycle of indicated virus
Jeang and colleagues identified DDX3 affects HIV-1 posttranscriptional gene expression.33
The exogenous expression of DDX3 increases production of HIV-1 Gag virion structural protein (5 to 12-fold), while suppression of endogenous DDX3 by antisense RNA decreases Gag production (factor of 13).33
Northern analysis determined that the overexpression of DDX3 did not change the abundance of the unspliced gag transcript or the ratio of unspliced to spliced viral transcripts. The results indicated that DDX3 did not affect HIV-1 transcription or pre-mRNA processing and is likely to enhance Gag production by increasing the Rev-dependent nuclear export of the unspliced gag transcripts. Consistent with this activity, DDX3 is a binding partner of the CRM1 nuclear export receptor and the HIV-1 viral posttranscriptional trans-activator, Rev.33
DDX3 expression is enhanced in a HeLa cell line that expresses HIV-1 Tat.33
Two possible explanations are that ddx3
expression is upregulated by the transcriptional trans-activation activity of Tat or the post-transcriptional RNA silencing suppressor activity of Tat, which may attenuate miRNA activity against ddx3.38,39
Yedavalli et al. explored DDX3 as a target to inhibit HIV-1 replication.40
Two-ring expanded nucleoside analogs were identified that inhibit DDX3 activity, suppress HIV-1 replication and exhibit insignificant toxicity in cell culture or in mice. These data document that cellular helicases are a feasible target for the development of novel antiretroviral drugs.
DDX3 promotes the innate response to virus infection by its ability to act as a scaffold for protein-protein interactions.41–44
DDX3 augments the activity of TBK1/IKKε, which activates IRF-3, IRF-7 and the NFκB transcription factor to stimulate interferon and interferon response genes that induce the antiviral state.41,42
While the helicase activity of DDX3 is not required, the N-terminal domain of DDX3 is required and associates with IKKε. Notably the same domain is targeted by vaccinia virus K7 protein.41
The crystal structure of K7 in complex with the N-terminal DDX3 peptide reveals that a hydrophobic cleft of K7 residues engage a thumb-like projection of tandem phenylalanine residues.45
Presumably this protein-protein interaction prevents DDX3 from engaging IKKe to promote interferon induction. The findings suggest that DDX3 is a prime target for viral manipulation of interferon induction.41,42
Yeast two-hybrid studies initially identified an interaction between DDX3 and the HCV core structural protein.31,35,46
Subsequently RNA interference verified that DDX3 is important for viral RNA replication in HuH-7 liver cells.47
The N-terminal residues of DDX3 are important for interaction with the HCV core and DDX3 interaction is significantly reduced by a single alanine substitution in HCV core.48
However it remains to be determined whether or not the HCV core blocks interaction between DDX3 and IKKε, which would attenuate IRF signalling. Alternatively, HCV core may block interaction between DDX3 and another cellular protein that bolsters replication by another mechanism.
Whereas DDX3 enhances replication of HIV-1 and HCV, DDX3 restricts hepatitis B virus (HBV) replication in both hepatoma and non-hepatoma cells. DDX3 binds to HBV polymerase (Pol), which is responsible for reverse transcription of viral RNA to DNA49
and is incorporated into nucleocapsid with the pre-genome RNA.50
DDX3 ATPase activity, but not helicase activity, is necessary for the inhibitory effect on HBV reverse transcription, which posits that inhibition of viral DNA synthesis occurs at a step following ATP hydrolysis, but prior to RNA unwinding.50
In addition to catalyzing reverse transcription, HBV Pol exhibits an immune modulatory role that is attributed to disrupted IRF signaling.50
Similar to vaccinia virus K7, HBV Pol disrupts the interaction between DDX3 and IKKε. The results suggest that DDX3 is manipulated to evade IFN induction at the expense of viral reverse transcription. The identification of DDX3 residues essential for the interaction with HBV Pol may provide another selective target for small molecule inhibitors to impair virus replication.
RNA helicase A (RHA)/DHX9.
Human RHA/DHX9 is an SF2 DEIH helicase that is the homolog of bovine nuclear DNA helicase II and Drosophila melanogaster
RHA is the largest and most complex of the helicases involved in viral replication (140 kDa and six domains) (). RHA executes protein-protein interactions that impact the continuum of gene expression, beginning at transcription and culminating in mRNA translation and influence cell growth and apoptosis, and the innate response to virus infection.3,54–59
RHA contains two double-stranded RNA binding domains (dsRBDs) of conserved α-β-β-β-α topology; the DEIH helicase core; the domain of unknown function (DUF), which is conserved in Ago2; the helicase-associated domain 2 (HA2) of a diversity of RNA helicases; a nuclear transport domain that is recognized by importin-α; and the carboxyl-terminal arginine and glycine-rich residues that are recognized by PRMT1 methylase.60,61
RHA is prominently nuclear62
and interacts with RNA polymerase II and transcription factors and coactivators that modulate gene transcription.63–65
In the cytoplasm, RHA interacts with structural features of the 5′ untranslated region (UTR) of the junD mRNA to facilitate efficient cap-dependent translation of this protooncogene.57,66,67
In sum, virus interaction with RHA may be advantageous for efficient expression of viral genes. Furthermore, the impact of RHA on cellular gene expression may produce cell growth conditions that are advantageous for the viral life cycle. Members of three virus families use RHA to promote their replicative cycle.
RHA overexpression has been suggested to increase retrovirus gene transcription, the ratio of spliced to unspliced viral transcripts68,69
and nuclear export of unspliced RNA.62,70
However downregulation experiments with siRNAs and quantitative RNA analysis determined RHA is not required for retrovirus transcription, RNA splicing or nuclear export, but that RHA is necessary for retrovirus translation.57,71–73
RHA selectively interacts with structural features of the 5′ UTR of many retroviruses and overexpression of RHA increases their translation.57,71–73
RHA interaction with the 5′ UTR increases polysome association and requires the ATP-binding activity of the DEIH helicase domain.57,73
The molecular basis of RHA translation activity is catalytic rearrangement of the RNP to facilitate ribosome scanning and translation initiation, and possibly and protein-protein interaction that secures a circular polysome for efficient translation reinitiation.57
In sum, RHA is necessary for efficient translation of many retroviruses including HIV-1 ().
Figure 3 Cellular helicases involved in HIV-1 replication. DExH/D helicases with a known function in HIV-1 replication are indicated beside the corresponding step in the replication cycle of HIV-1: (1) Receptor-mediated entry; (2) Reverse transcription of viral (more ...)
In the case of the simian Mason-Pfizer monkey virus (MPMV), RHA interacts with the 5′ UTR to facilitate translation.74
Additionally, RHA interacts with the 3′ UTR of MPMV in a region necessary for nuclear export of the unspliced viral transcript.70
This region is designated the constitutive transport element (CTE)62
and facilitates nuclear export by recruiting the TAP nuclear export receptor.75
Conceivably RHA interacts with MPMV RNA in the nucleus and facilitates consecutive remodeling of the viral RNP necessary for nuclear export and subsequent translation in the cytoplasm.76
Studies on HIV-1 determined that RHA promotes the infectivity of progeny virions77
(). Molecular analysis of the defect determined that RHA in the virus producer cells fosters the reverse transcription activity of the progeny virions77
(). Kleiman and colleagues determined the defect is attributable to reduced annealing of tRNA to viral RNA (40% of control) during the virion assembly process.133
Furthermore, RHA is incorporated into the progeny virions, possibly through interaction with the nucleocapsid domain of Gag,77
the HIV-1 RNA or an adapter protein.73
RHA-deficient virions are less infectious to Hela-based TZM-bl reporter cells77
and primary T lymphocytes.73
The defect is attributed, in part, to reduced reverse transcription77
but may also involve replication steps that are downstream of reverse transcription. In sum, RHA appears to be involved in multiple steps in the retroviral life cycle.
Behrens and colleagues identified that RHA binds to the 5′ and 3′ UTRs of bovine viral diarrhea virus RNA and affects viral replication.78
Two other proteins that contain the conserved dsRBD (NF-45 and NF-90/NFAR-1) facilitate viral replication. These findings suggest the dsRBD may be acting as a scaffold to secure the viral RNA in a configuration helpful to replication of the bovine diarrheal virus.
To investigate possible importance of RHA in a human flavivirus, He et al. studied HCV replicon cells using RHA downregulation strategies. A lentivector-based inducible RNA interference system administered partial RHA downregulation. Partial RHA downregulation produced a gradual reduction of HCV RNA and protein over 21 days, as measured by RT-PCR and immunoblot, respectively.79
The results suggest RHA is a necessary host factor for HCV in replicon cells.
Recent study of foot and mouth disease virus (FMDV) determined that RHA downregulation decreases viral replication.80
In infected cells, RHA co-precipitates FMDV replication proteins, 2C and 3A.80
Furthermore, FMDV infection changes the subcellular localization of RHA from prominently nuclear and a uniform distribution, to prominently cytoplasmic and a punctuate distribution. Similar to results in retroviruses and bovine diarrheal virus,57,73,78
RHA interacts with terminal regions of FMDV RNA. Together the results posit the model that FMDV uses RHA to promote viral RNA replication and to facilitate viral protein synthesis. The authors speculate that FMDV uses RHA to connect the 5′ and 3′ UTRs to switch between viral RNA replication to viral protein synthesis, which may occur on a circular polysome.80
In summary, RHA interacts with the terminal regions of many viral RNAs. Future analysis is warranted to determine the domain(s) of RHA required for interaction with viral RNA and test the possibility that similar residues are required across several virus families. This fundamental investigation will be informative to development of antiviral therapy that selectively discriminates between the cellular and viral RNA targets of RHA.
To date a single virus, HIV-1, has been shown to interact with DHX30 (isoform 2).81
DHX30 is a 131 kDa protein with the conserved ATP-dependent helicase domain (DEVH variant) followed by the DUF domain and HA2 domain and one dsRBD at the amino-terminus (). The DHX30 domain structure is highly similar to RHA. Similar to results with RHA, overexpression of DHX30 increases expression of the HIV-1 provirus (2 to 3-fold) and an HIV-1 LTR luciferase reporter (1.5–4 fold), and this activity is eliminated by a DEVH mutation.81
In distinction from RHA, DHX30 overexpression decreased viral RNA packaging into progeny virions (factor of ~10 as measured by RT-real time PCR on equivalent particle-associated Gag), suggesting proper trafficking of HIV-1 RNA between the cytoplasm and virions was disrupted (). The authors speculated that DHX30 might cause a modification of the packaging signal by binding to the 5′ UTR. As expected, the reduction in viral RNA packaging correlated with reduced infectivity of progeny virions (by a factor of 6). Downregulation of DHX30 did not produce an effect on HIV-1 gene expression,81
which is likely attributable to redundant activity by RHA. RHA is abundantly expressed in many types of animal cells and displays high sequence identity with DHX30 in basic residues that direct RHA interaction with retrovirus RNA.134
Domain exchange experiments between DHX30 and RHA will be instrumental to understand the specificity and potential functional redundancy of these RNA helicases.
Members of three virus families exhibit interaction with DDX1 (). DDX1 is an 82 kDa ATP-dependent DEAD box family member that contains a SPRY domain (). DDX1 exhibits affinity for poly(A) RNA, heterogenous nuclear ribonucleoprotein K and cleavage stimulatory CSTF-64, which suggests DDX1 contributes to proper 3′ end maturation of mRNAs.82,83
In common with DDX3, RHA and DHX30, DDX1 activity affects HIV-1 replication (). Yeast and mammalian two hybrid systems identified DDX1 interacts with HIV-1 Rev.84
DDX1 downregulation reduces the activity of an HIV-1 Rev/RRE-dependent luciferase reporter gene, and RT-PCR determined that nuclear export of unspliced reporter mRNA was impaired. The reduced Rev/RRE activity correlated with disordered Rev subcellular distribution in nucleus and nucleolus,85
consistent with an important role for this host factor in HIV-1 biology. While DDX1 is upregulated in neuroblastoma and retinoblastoma cell lines and tumors,86
expression in astrocytes is low. Astrocytes are nonpermissive for HIV-1 replication and ectopic expression of DDX1 is sufficient to complement the defect in HIV-1 replication.85
The findings that DDX1 facilitates Rev/RRE activity and is a rate-limiting factor for HIV-1 replication in astrocytes indicates that DDX1 is a required host factor for HIV-1 replication.87
DDX1 also modulates cell specificity of the human polyoma virus, JC virus.88
DDX1 expression is high in cells susceptible to JC virus infection and low in nonpermissive cells; exogenous DDX1 expression renders nonpermissive cells susceptible to JC virus infection. DDX1 and CSTF64 bind the JC virus transcriptional control region and the outcome is productive expression and accumulation of early and late JC virus transcripts. By contrast, downregulation of DDX1 impairs JC virus gene expression. The results are likely attributable to impaired 3′ end processing that culminates in the decay of viral transcripts.
Yeast 2-hybrid screens determined DDX1 interacts with coronavirus nonstructural protein 14.89
Vero cells infected with infectious bronchititis virus exhibited transit of DDX1 from the nucleus to the cytoplasm and colocalization with the coronaviral RNA.89
Mutation of DDX1 DEAD motif to AAAA eliminated the colocalization and the replication of coronavirus, indicating a necessary role for the ATPase-dependent helicase activity of DDX1.89
The results suggest that DDX1 is important for efficient coronavirus replication.
DDX24 is a 96 kDa DEAD box helicase that lacks other conserved motifs () and exhibits amino acid content that is dissimilar to typical DEAD box family members.91
Proteomic analysis identified DDX24 coprecipitates with HIV-1 Gag. This interaction is mediated by the Gag nucleocapsid domain and is disrupted by RNase, indicating that the interaction is RNA-dependent.77
DDX24 also coprecipitates with HIV-1 Rev.25
The downregulation of DDX24 changes the subcellular localization of Rev from nucleolar to nuclear and moderately increases Rev/RRE-dependent nuclear export of unspliced HIV-1 mRNA (2–3 fold).25
Despite the increase of viral transcript in the cytoplasm, the infectivity of HIV-1 progeny virions was reduced on Hela-based TZM-bl reporter cells. Real time PCR determined reduced abundance of virion-associated RNA, which is sufficient to account for the reduction in titer.25
In sum, DDX24 appears to be important for proper trafficking of the viral RNA for assembly into virions ( and
p68 is a DEAD box helicase that contains a p68HR domain that is characteristic of p68-like RNA helicases (). Named for its molecular weight of 68 kDa, p68 cross reacts with monoclonal antibody to SV40 large-T antigen (pAB204 epitope), which suggests these proteins display a similar epitope.92–94
In common with SV40 large-T antigen, p68 affects cellular proliferation and tumor development, and mechanistically p68 intersects the processes of transcriptional regulation, RNA splicing and the processing of pre-ribosomal RNA.95–100
Using results from a yeast two-hybrid screen, Goh et. al. determined that p68 interacts with HCV NS5B RNA-dependent RNA polymerase.101
Expression of NS5B of HCV changes the subcellular localization of endogenous p68 from the nucleus to the cytoplasm,101
similar to the effect of FMDV on RHA.80
Downregulation of p68 reduces the transcription of negative strand HCV RNA. Taken together, p68 is an important host factor for HCV at the level of HCV gene expression ().
Moloney leukemia virus 10 homolog (Mov10) contains a putative SF1 helicase (DEAG variant) and lack other defined domains (). Mov10 appears to be heavily post-translationally modified because of substantial difference between the predicted and observed molecular weights (114 kDa and 130 kDa, respectively). Mov10 is required for RNA interference-mediated gene silencing by RNA-induced silencing complex (RISC);102,103
however RNA interference is successfully employed to test the effect of Mov10 on virus replication (below). Mov10 is expressed in a variety of cell types including lymphocytes and embryonic stem cells.104,105
Mov10 in HeLa cells coprecipitates with RISC components Ago1 and Ago2 and colocalizes with Ago proteins in cytoplasmic processing bodies.102
Virus interaction with Mov10 has the potential to interface with antiviral RNA silencing and influence the viral replication efficiency.
Hepatitis delta virus (HDV, unassigned family).
The downregulation of Mov10 or the Mov10 binding partner, Ago2, is sufficient to reduce HDV replication.106
HDV replication requires RNA-directed transcription and host RNA-dependent RNA polymerase catalytic activity. In principle, these processes are similar to RNA-directed RISC loading to miRNA target RNA and RNA-dependent catalytic activity of RISC. We posit a corollary model in which Mov10 facilitates remodeling of the incoming viral RNA to produce a transcription initiation competent RNP.
The downregulation of endogenous Mov10 reduces HIV-1 infectivity. However, exogenous expression of Mov10 in primary or transformed cells severely reduces infectivity of progenyvirions.103,104,107
These non-reciprocal results suggest an appropriate balance of Mov10 and cofactors is responsible for control of virus replication. Recently published studies determined Mov10 is a restriction factor against several retroviruses: simian influenza virus, murine leukemia virus, feline immunodeficiency virus, equine infectious anemia virus and HIV-1.103,104,107
Mov10 of human, simian and murine origin exhibit antiretroviral activity, consistent with their significant conservation among various species103
(). HIV-1 infection of human cell lines (A3.01, CEM-T4, CEM-SS) reduces endogenous Mov10,103
which is reminiscent of virus-associated degradation of the APOBEC3G restriction factor by Vif.108
Mov10 interacts with APOBEC3G in primary CD4+
T cells and 293T cells, although Vif does not downregulate Mov10 activity.104
Similar to DHX30, Mov10 interacts with the nucleocapsid domain of HIV-1 Gag in an RNA-dependent manner and is packaged in virions (). Restriction occurs at an early, post-entry step in viral replication that occurs before reverse transcription is initiated (). Results from producer cells indicate that proteolytic processing of nascent virions is reduced,107
which may indicate defective viral core uncoating or translocation to the nucleus. The issue remains uncertain of whether or not the anti-HIV activity requires ATPase-dependent helicase activity of Mov10. Mutation of the DEAG helicase domain eliminated anti-HIV-1 activity in one study,103
but not in another investigation.104
Both studies used 293T cells to produce virus, but they employed different cells to measure virus titer: osteosarcoma HOS-derived HIV GFP GHOST reporter cells and HeLa, respectively. The use of different target cells may contribute to the conflicting outcomes. The restriction activity of Mov10 in species from mouse to man and against many retroviruses raises the possibility that Mov10 restricts replication of other viral families. Either way, deciphering the fundamental mechanisms of Mov10 activity may provide a target for development of selective anti-viral molecules.