Packaging of influenza virus genome: Robustness of selection

Influenza virus is unique in that its genome is fragmented just like human chromosomes. This genome segmentation confers evolutionary advantages such as genetic reassortment, which contributes to the emergence of pandemic strains (1). Also like humans, the influenza virus must pass its genome accurately onto its offspring for species survival. This process is complicated but biologically interesting because the viral genome must be selected from a large pool of host genetic materials in the virus-infected cell and each infectious virus particle must contain all of the genome fragments. However, despite exhaustive research, the mechanisms by which the influenza virus particle ensures correct packaging of its fragmented genome have remained largely unclear. In PNAS, Chou et al. (2) develop an experimental system to address this long-time enigma and, in so doing, make important contributions to our understanding of the genome packaging process of the influenza virus.

The influenza virus genome is composed of eight single-stranded, negative-sense RNA segments. Historically, two models have been proposed to explain the mechanisms by which the viral RNA segments are packaged into progeny virus particles: the random packaging model and the selective packaging model (3). The former model assumes that the various viral RNA segments are arbitrarily incorporated into virus particles, and infectious particles are produced only when at least one copy of each viral RNA segment is incorporated into the particle by chance (Fig. 1A). Infectious bursal disease virus, which possesses a two-segmented dsRNA genome, employs this system (4). The latter model proposes that each viral RNA segment has a distinct packaging signal sequence that differentiates it from the other viral RNA segments, ensuring the packaging of eight unique viral RNA segments into each virus particle (Fig. 1A). The dsRNA bacteriophages of the Cystoviridae family are good examples of organisms that use this model (5). With respect to the influenza virus, however, conclusive evidence regarding its genome packaging mechanism is still lacking, and controversy remains.

Fig. 1.

(A) Conflicting hypothetical models of the genome packaging mechanism. Replication of the viral genome takes place in the nucleus of the virus-infected cell. The viral RNA segments, in the form of RNPs, are transported to the plasma membrane, where the segmented genome is packaged, and progeny virions are produced. a, The selective packaging model, highlighted by Chou et al. (2), proposes that eight unique viral RNA segments are consistently packaged into every progeny virus particle. b. In the random packaging model, viral RNA segments are arbitrarily packaged into virus particles, and only a small proportion of the progeny virus particles are infectious. (B) EM analysis shows that the eight RNPs are arranged in a distinct pattern when packaged in a virus particle (11). (C) A 3D model of the eight RNPs within a virus particle reconstructed by using ET (13). The RNPs differ in length and are labeled with different colors.

Early evidence in support of the selective packaging model for influenza virus came from analyses of defective-interfering (DI) viral RNAs (6, 7). The DI segment, which is derived from a viral RNA segment as a result of an internal deletion in the coding region, competitively inhibits the packaging of its parental viral RNA segment but not that of other viral RNA segments, and is preferentially incorporated into progeny virus particles. Such segment-specific competition implies that each viral RNA segment is distinct during the genome packaging process and that each DI segment possesses a so-called genome packaging signal. Later, reverse genetics studies provided compelling evidence that all eight viral RNA segments possess segment-specific packaging signal sequences for their efficient incorporation into progeny virus particles (3, 8), as predicted in the selective packaging model. These packaging signals include bipartite sequences at both ends of the viral RNA segment, which house not only the conserved promoter region that is common to all eight viral RNA segments, but also protein-coding and segment-specific noncoding regions adjacent to the promoter region. Thus, the signal sequences are unique to each viral RNA segment, which may be important during the selection of the eight different viral RNA segments.

EM analyses also strongly support the selective packaging model (9). Eight ribonucleoprotein complexes (RNPs), which are composed of the viral RNA segments, nucleoproteins, and viral RNA polymerase complexes, are arranged within a progeny virus particle in a distinct pattern, in which seven RNPs surround a central RNP (10, 11) (Fig. 1B). Importantly, no progeny virus particles contain more than eight RNPs (11). Recent 3D analyses using electron tomography (ET) showed that the eight RNPs arranged in the aforementioned “seven plus one” configuration differ in length (12, 13) (Fig. 1C). They are composed of three long RNPs and five shorter RNPs of significantly different lengths, suggesting that the eight RNPs contain at least six different kinds of viral RNA segments (13, 14). Interestingly, ET also suggests that the eight RNPs within a virus particle are connected to each other, forming a supramacromolecular complex (12, 13). This observation is consistent with reverse genetics studies demonstrating that mutations in a packaging signal of a viral RNA segment affect the packaging efficiency of the other viral RNA segments in the progeny virus particles (15). Thus, viral RNA segments do not appear to be autonomous within virus particles; rather, their packaging into virions is coordinated via interactions among the different viral RNA segments.

Thus, growing evidence from reverse genetics studies and EM analyses has favored the selective packaging model. However, conclusive evidence to directly prove that a single virus particle contains eight unique viral RNA segments has been missing, mainly because no one had found the right experimental approach. In PNAS, Chou et al. (2) describe an experimental system based on FISH, which enables them to examine the composition of viral RNA segments within virus particles quantitatively at single-virus resolution. By using two probe sets targeting different viral RNA segments, the authors initially examined the copacking efficiency of two different viral RNA segments in each virus particle and showed that a certain viral RNA segment is copackaged with all the other viral RNA segments in a large proportion of virus particles (2). Then, they quantitatively examined the copy number of each viral RNA segment being packaged in each virus particle by counting the number of hybridized probes in each particle by using a photobleaching technique. The authors demonstrated that most of the virus particles packaged a single copy of each of the eight different viral RNA segments. This report by Chou et al. (2) thus offers indisputable evidence to support the selective packaging model

Growing evidence from reverse genetics studies and EM analyses has favored the selective packaging model.

and moves us one step closer to understanding the genome packaging mechanism of influenza virus.

Nevertheless, our knowledge about the genome packaging mechanism is still limited. As Chou et al. point out (2), the precise mechanisms for the selection of the eight viral RNA segments, as well as the interactions among the different viral RNA segments at the molecular level, remain largely unknown. Although the packaging signal sequences, which are unique to each viral RNA segment, should be involved in the genome selection and the multisegmental interactions (3, 8, 15), their roles in the selective genome packaging process have not yet been revealed. The approach established by Chou et al. (2), in combination with reverse genetics, will allow us to examine the effects of mutations or deletions in a packaging signal on the copackaging of viral RNA segments at the single-virus level. Such studies, if performed with various combinations of two different viral RNA segments, would help us enormously to identify specific interactions among viral RNA segments and to understand the role of each packaging signal in the selective genome packaging process.

Finally, the controversy that has existed for half a century has drawn to an end (16). Now we know that an influenza virus particle typically incorporates eight different viral RNA segments through yet-unidentified selective packaging machineries. This genome packaging process, which is an essential step for influenza virus replication and is likely conserved among all influenza A viruses (11), is strictly regulated to select eight unique viral RNA segments. Therefore, it could be an excellent drug target for pan-influenza A viruses. With this great advancement in the basic science of virology, we can look to the development of novel strategies to combat seasonal and pandemic influenza infections.