Nuclear import of the influenza A virus transcriptional machinery

Abstract

Unusually for an RNA virus, influenza A viruses transcribe and replicate their genomes in the nuclei of infected cells. As a result the viral ribonucleoprotein complexes (RNPs), and their newly synthesised protein subunits, must interact with the host nuclear import machinery. In this review we discuss how the virus exploits nuclear import pathways to allow regulated and chaperoned assembly of RNPs in the nucleus, and describe how the import machinery itself can be a determinant of host tropism.

Introduction

Influenza A virus, a cause of widespread disease in humans and livestock, is the prototypic member of the orthomyxovirus family. The majority of RNA viruses that infect vertebrates replicate their genomes in the cytoplasm [1] but, in common with other orthomyxoviruses, influenza A virus transcribes and replicates its genome in the nucleus of an infected cell [2], [3], [4]. The viral genome comprises eight segments of single-stranded, negative-sense RNA, each of which is encapsidated in a ribonucleoprotein complex (RNP). The viral lifecycle requires nuclear import, first of RNPs introduced by the infecting virus, and then of newly synthesised RNP protein subunits–the acidic subunit and the two basic subunits of the trimeric viral polymerase (PA, PB1 and PB2), and the viral nucleoprotein (NP) [5], [6], [7]. This review describes how influenza A virus utilises the host's nuclear import machinery to allow nuclear transcription of its genes and replication of its genome, and the implications of these interactions for the virus.

Nuclear genome replication and transcription allows influenza A virus to exploit various cellular functions that are compartmentalised to the nucleus. By accessing the cellular splicing machinery, influenza is able to expand its coding capacity in a manner that would not be possible for a cytoplasmically replicating virus. Spliced transcripts are produced from the mRNAs of two of the virus’ eight genomic segments (encoding the ion channel M2, as well as several spliced RNAs of unknown function, from segment 7, and the nuclear export protein (NEP) from segment 8) [8], [9]. Access to the nucleus also facilitates ‘cap-snatching,’ the process by which the viral polymerase obtains 5′ cap structures for its mRNA by cleaving them from cellular transcripts [10]. Cap-snatching, which removes a structure required for translation from host mRNAs and transfers it to viral transcripts, is carried out in the cytoplasm by some cytoplasmically replicating viral families such as the bunyaviruses and the arenaviruses [11]. By contrast, the influenza A virus polymerase is active in the nucleus, where it has been shown to bind to cellular DNA-dependent RNA polymerase II (Pol II), the transcriptional machinery for cellular mRNA [12], [13], [14], [15]. It has been suggested that this may increase the efficiency of the cap-snatching process, as well as linking viral transcription to cellular pathways of mRNA processing and nuclear export [16], [17], [18]. It has also been shown that this interaction increases the efficiency of host shut-off as the viral polymerase not only inhibits Pol II elongation but causes it to be degraded [15], [19], [20], a mechanism that appears to contribute to virulence [21]. As viral transcription, but not replication, is dependent on Pol II-generated capped transcripts, the suppression of Pol II synthesis as infection proceeds may contribute to the decline in viral mRNA synthesis late in infection [16].

Cap-snatching of host mRNAs and degradation of Pol II are examples of mechanisms where nuclear localisation of the transcriptional machinery allows the virus to maintain control of an infected cell and prevent the activation of antiviral responses. In another example of this, one of the major mechanisms by which infections with influenza virus are detected is through the cytoplasmic receptor retinoic acid inducible gene 1 (RIG-I). By encapsidating newly synthesised RNA into RNPs within the nucleus, influenza viruses may at least partially mask a molecular pattern capable of triggering an antiviral response [22].

Replicating in the nucleus therefore provides influenza A virus with a wide range of potential benefits, including access to host splicing machinery, efficient cap-snatching, and increased opportunities for host shut-off and evasion of antiviral responses. Entry to the nucleus is, however, a tightly gated process, and the viral transcriptional machinery must interact with a range of cellular factors in order to bring this about.

The nucleus of a cell is enclosed by a double membrane. Spanning this barrier are channels formed by nuclear pore complexes (NPCs), large macromolecular structures that permit traffic between the cytoplasm and the nucleoplasm [23]. While small molecules can pass through NPCs by passive diffusion, structures larger than approximately 9 nm [24] or 20–30 kDa [23] can only be translocated across the NPC by forming specific interactions with transport receptors, notably members of the importin-β (karyopherin-β) superfamily, which shuttle between the nucleoplasmic and the cytoplasmic sides of the nuclear membrane. Most of these transport receptors bind cargo directly, although importin β (karyopherin-β1) uses isoforms of the importin-α to mediate binding to specific cargoes [23], [25]. Directionality of transport is maintained through association with the small protein Ran, which binds to GTP in the nucleus (due to the nuclear localisation of the Ran guanosine exchange factor, RanGEF, also known as regulator of chromosome condensation 1, RCC1) and hydrolyses it to GDP in the cytoplasm (facilitated by the cytoplasmic Ran GTPase activating protein, RanGAP). Importins bind specific cargo in the cytoplasm, and release it on binding RanGTP in the nucleus. Conversely, exportins bind cargo along with RanGTP in the nucleus, and release them when Ran hydrolyses GTP to GDP in the cytoplasm.

Nuclear transport receptors interact with specific cargo proteins by binding to short motifs – nuclear localisation signals (NLSs) or nuclear export signals (NESs). NLSs typically consist of basic residues, forming either a single cluster or a bipartite motif in which two basic regions are separated by a short linker. In ‘classical’ nuclear import the NLS binds to an isoform of importin α, which binds in turn to importin β1, whereas in ‘non classical’ nuclear import the NLS of the cargo binds directly to an importin-β family member. While classical NLSs typically conform to a consensus sequence, importin-β binding cargoes are much more diverse and the requirements for a functional NLS often remain unclear [25], [26].