Analysis of IAV Replication and Co-infection Dynamics by a Versatile RNA Viral Genome Labeling Method

Resource

Analysis of IAV Replication and Co-infection

Dynamics by a Versatile RNA Viral Genome Labeling Method

Graphical Abstract

Highlights

d

PLP labeling is a versatile method for analyzing RNA viral genomes in situ

d

Genome labeling can resolve cell infections with single- nucleotide specificity

d

Viral entry and replication can be monitored in large cell populations

d

Productive influenza A virus cell co-infections are time restricted

Authors

Dan Dou, Iva´n Herna´ndez-Neuta, Hao Wang, ..., G€ulsah Gabriel, Mats Nilsson, Robert Daniels

Correspondence

mats.nilsson@scilifelab.se (M.N.), robertd@dbb.su.se (R.D.)

In Brief

Dou et al. demonstrate how padlock probe (PLP) RNA labeling can be

harnessed to analyze RNA viruses in situ.

The localization and labeling specificity are combined to visualize the entry and replication dynamics of the influenza A virus gene segments and define the time window for productive same-cell co- infections.

Dou et al., 2017, Cell Reports 20, 251–263 July 5, 2017 ª 2017 The Authors.

http://dx.doi.org/10.1016/j.celrep.2017.06.021

Cell Reports

Resource

Analysis of IAV Replication and Co-infection Dynamics by a Versatile RNA Viral

Genome Labeling Method

Dan Dou,

Iva´n Herna´ndez-Neuta,

Hao Wang,

Henrik O ¨ stbye,

Xiaoyan Qian,

Swantje Thiele,

Patricia Resa-Infante,

Nancy Mounogou Kouassi,

Vicky Sender,

Karina Hentrich,

Peter Mellroth,

Birgitta Henriques-Normark,

G€ulsah Gabriel,

Mats Nilsson,

* and Robert Daniels

*

1Department of Biochemistry and Biophysics, Stockholm University, 10691 Stockholm, Sweden

2Science for Life Laboratory, 17121 Stockholm, Sweden

3Heinrich Pette Institute, Leibniz Institute for Experimental Virology, 20251 Hamburg, Germany

4Department of Microbiology, Tumor and Cell Biology, Karolinska Institutet, 17177 Stockholm, Sweden

5Clinical Microbiology, Karolinska University Hospital, 17176 Stockholm, Sweden

6SCELSE and LKC, Nanyang Technological University (NTU), Singapore 639798, Singapore

7These authors contributed equally

8Present address: European Molecular Biology Laboratory, Grenoble Outstation, Grenoble Cedex 9, 38042 Grenoble, France

9Lead Contact

*Correspondence:mats.nilsson@scilifelab.se(M.N.),robertd@dbb.su.se(R.D.) http://dx.doi.org/10.1016/j.celrep.2017.06.021

SUMMARY

Genome delivery to the proper cellular compartment for transcription and replication is a primary goal of viruses. However, methods for analyzing viral genome localization and differentiating genomes with high identity are lacking, making it difficult to investigate entry-related processes and co-examine heterogeneous RNA viral populations. Here, we present an RNA labeling approach for single-cell analysis of RNA viral replication and co-infection dy- namics in situ, which uses the versatility of padlock probes. We applied this method to identify influenza A virus (IAV) infections in cells and lung tissue with single-nucleotide specificity and to classify entry and replication stages by gene segment localization.

Extending the classification strategy to co-infections of IAVs with single-nucleotide variations, we found that the dependence on intracellular trafficking places a time restriction on secondary co-infections necessary for genome reassortment. Altogether, these data demonstrate how RNA viral genome la- beling can help dissect entry and co-infections.

INTRODUCTION

Advancements in sequencing technology have increased the available information about viruses and made host responses to infection more accessible (Brister et al., 2015; Capobianchi et al., 2013; Law et al., 2013; Westermann et al., 2017). These improvements have been particularly useful for examining how heterogeneous RNA viral populations or quasispecies, generated by error-prone replication, contribute to viral fitness

(Borderı´a et al., 2016; Varble et al., 2014; Vignuzzi et al., 2006;

Xue et al., 2016). They have also made it easier to determine the origin of new RNA viruses and mutations responsible for changes in pathogenicity (Gire et al., 2014; Herfst et al., 2012;

Imai et al., 2012). However, sequencing techniques generally provide limited spatial information about RNA viral genomes, which is necessary to examine processes related to genome trafficking, co-infections, and reassortment.

Due to technical limitations, cells infected by RNA viruses are typically identified with antibodies, which do not account for population heterogeneity. Approaches using single-molecule fluorescence in situ hybridization (smFISH) have shown promise for detecting the localization of RNA viral genomes in cells (Chou et al., 2012, 2013; Lakdawala et al., 2014). Although these tech- niques are not easily modified to distinguish betweeen RNA viruses that are highly similar, using the genome for detection provides several advantages. First, viral genomes are a defining feature of an infection and can be used to monitor viral entry.

Second, genome replication is an accurate infection reporter, because it occurs early and in defined cellular regions. Third, sequence-based methods have the potential to achieve the specificity needed to co-analyze cell infections by viruses that differ by as few as one nucleotide.

Since single-molecule RNA localization techniques are rela-

tively new (Crosetto et al., 2015; Gaspar and Ephrussi, 2015),

much of the information about RNA viral genomes and trafficking

has come from biochemical assays and electron and immunoflu-

orescence microscopy. This especially applies to influenza A vi-

ruses (IAVs), for which these approaches helped define the

genome as eight negative-sense, single-stranded, viral RNA

(vRNA) gene segments (Arranz et al., 2012; McGeoch et al.,

1976; Palese and Schulman, 1976). They also were used to

establish the general entry pathway to the nucleus (Herz et al.,

1981; Jackson et al., 1982; Martin and Helenius, 1991; Matlin

et al., 1981) and to identify cellular proteins involved in gene

segment nuclear export (Elton et al., 2001; Ma et al., 2001).

Recent studies have analyzed certain aspects of IAV gene segment trafficking using vRNA-associated proteins as re- porters (reviewed in Breen et al., 2016; Hutchinson et al., 2010) and smFISH labeling (Chou et al., 2012, 2013; Huang et al., 2012; Lakdawala et al., 2014). However, these methods have not been extensively applied and many of the remaining ques- tions about IAV genome trafficking require alternative ap- proaches with high sequence specificity and multiplex capacity.

A rarely used technique for analyzing RNA viruses involves padlock probes (PLPs), which were initially designed to detect DNA sequences and later developed for labeling RNA se- quences in cells and tissue (Andersson et al., 2012; Gyarmati et al., 2008; Ke et al., 2013; Larsson et al., 2010; Lizardi et al., 1998; Nilsson et al., 1994). PLPs function in a variety of biological samples, have high specificity and a short nucleotide recognition requirement, and can generate a localized amplified signal alone or in multiplex reactions (Bane´r et al., 1998; Hardenbol et al., 2003; McGinn et al., 2016; Nilsson et al., 1997; Zieba et al., 2012). In this study, we used the versatility of PLPs to develop an in situ RNA labeling method for analyzing IAV infections.

The approach is capable of detecting all eight IAV vRNA gene segments, can monitor the stochastic aspects of IAV entry, and can be used to define the infection stages across cell pop- ulations. By using this methodology to analyze co-infections of IAVs with single-nucleotide variations, we obtained evidence that productive IAV cell co-infections occur in a limited time window, defined by the replication stage of the primary infection.

The specificity and direct nature of this multifunctional PLP label- ing approach extends how RNA viral infections can be analyzed and detected in vitro and in vivo.

RESULTS

Approach for IAV Gene Segment vRNA Labeling and Single-Cell Analysis

The IAV genome consists of eight gene segments, or ribonucleo- protein complexes, that contain a single vRNA with numerous copies of the nucleoprotein (NP) and one copy of the viral poly- merase (Figure 1A). Due to the segmentation, productive infec- tions only occur when all eight gene segments reach the nucleus for replication (Eisfeld et al., 2015; Herz et al., 1981; Te Velthuis and Fodor, 2016). Based on this requirement, we sought to analyze IAV infections by tracking the intracellular localization of the vRNAs in single cells. PLPs were chosen as a basis for the approach because of their ability to detect specific RNA sequences in situ and their multiplex labeling capacity (Ke et al., 2013; Larsson et al., 2010; Lizardi et al., 1998). Each PLP was designed for a specific vRNA from the H1N1 strain A/WSN/33 (WSN) by incorporating 5

and 3

arms that hybridize to adjacent 20-nucleotide sequences, as well as two identifica- tion barcodes (Figure 1B) (Supplemental Information).

The labeling and single-cell analysis were established using infected Madin Darby Canine Kidney (MDCK) cells (Figure 1C).

Following cell fixation, cDNAs were synthesized from each IAV gene segment using locked nucleic acid (LNA) modified primers (Supplemental Information). The template was then digested with RNase H, and PLPs were hybridized and ligated to enable

synthesis of a rolling-circle product (RCP), which locally am- plifies the signal by creating a concatemer with multiple comple- mentary barcode copies. After RCP synthesis, cell nuclei were stained, fluorescently labeled barcodes were added (Supple- mental Information), and the cells were imaged by fluorescence microscopy (Figure 1D). To accommodate all eight gene seg- ments, the first barcode labels were removed and the cells were reimaged after labeling with the second set of barcodes.

Images were then aligned and compiled, and nuclei staining combined with cell autofluorescence was used to define the cell segments used to assign the RCP-labeled gene segments.

Padlock Probe Labeling Efficiently Tracks IAV Gene Segment LocalizationIn Situ

The general IAV cell entry pathway is well defined (Figure 2A) (de Graaf and Fouchier, 2014; Hutchinson et al., 2010; Skehel and Wiley, 2000). IAVs bind and traffic to the endosome where the low pH promotes hemagglutinin (HA)-mediated membrane fusion releasing the gene segments into the cytosol (Rust et al., 2004; White et al., 1982; Yoshimura and Ohnishi, 1984).

From there, importins transport them to the nucleus for replica- tion by the viral polymerase through a positive-sense cRNA inter- mediate (Gabriel et al., 2011; Resa-Infante and Gabriel, 2013).

The new gene segments then use the chromosomal region main- tenance 1 (CRM-1) nuclear export pathway and Rab proteins to reach the plasma membrane for viral packaging in a gene segment-dependent manner (Amorim et al., 2011; Eisfeld et al., 2011; Elton et al., 2001; Gavazzi et al., 2013; York and Fodor, 2013). To establish time parameters for this complex pro- cess in our system, vRNA levels in the cells and culture medium were measured to determine when replication ( 3 hr post-infec- tion [p.i.]) and viral release ( 7 hr p.i.) initiate ( Figure 2B).

Due to the limited IAV gene segment entry data, we asked whether our approach can label vRNAs during entry by trapping the virus in endosomes using the pH-dependent viral fusion in- hibitor bafilomycin A1 (BFLA1) (Figure 2A) (Ochiai et al., 1995).

In both untreated and BFLA1-treated cells, gene segment RCPs were visualized 2 hr p.i. (Figure 2C). However, the RCPs in BFLA1-treated cells mainly resolved as single clusters, indic- ative of endosomes with a nuclear localization frequency match- ing the average image occupancy of the nucleus ( 40%), whereas the RCPs in untreated cells showed high nuclear local- ization. Next, we assessed the labeling of newly synthesized vRNAs by treating cells with leptomycin B (LMB), which inhibits their nuclear export (Figure 2A) (Elton et al., 2001). The gene seg- ments were resolved in both samples, and similar to the distribu- tion of the viral protein controls NP and PA (Figure S1), LMB treatment increased the RCP nuclear localization to 80% ( Fig- ure 2D). These experiments confirmed that the developed in situ PLP labeling approach is applicable to the entire IAV replication cycle, because it can detect vRNA localization during entry and after replication.

IAV Gene Segment Labeling Accurately Identifies Productive Cell Infections

In tissue culture systems, defective interfering viral particles often exceed the number of infectious particles (Brooke, 2014;

Dimmock and Easton, 2014; Fonville et al., 2015). Therefore,

we used an MOI corresponding to 0.3, or an infection rate of

25%, to further assess the labeling specificity ( Figure 2E).

Although the recognition sequences caused some label variation (Figure S2), gating by the number of distinct gene segment RCPs revealed that 20% of the cells had productive infections (all eight vRNAs), whereas most of the population was uninfected or subject to potential nonproductive infections (Figure 2F). We then measured the number of NP gene segment RCPs per cell over time and found a reasonable linear correlation with NP vRNA levels obtained by qPCR (Figure 2G), indicating the label- ing also can provide semiquantitative data for the vRNAs.

Design for Labeling IAV Gene Segments with Single-Nucleotide Substitutions

Similar to many RNA viruses, IAVs frequently acquire genome mutations, and these can significantly alter viral growth and pathogenicity (Boni, 2008; Drake, 1993; Schrauwen and Fouch- ier, 2014). However, most classical assays for monitoring viral properties are indirect and do not accommodate heterogeneity in viral populations (Dulbecco and Vogt, 1953; Reed and Muench, 1938). Consequently, IAVs with high identity are often analyzed by sequencing or independent comparisons. PLPs offer a potential solution to this problem, because the recognition is based on sequence rather than epitopes or cell death. To test the ability to resolve IAV infections with single-nucleotide

specificity, we rescued an isogenic virus (WSN

), which con- tained a silent (synonymous) point mutation in each WSN gene segment (Figure 3A), by reverse genetics (da Silva et al., 2015;

Hoffmann et al., 2000). During the rescue, both viruses showed similar in vitro growth (Figure 3B), and no changes occurred in the sequences or the distributions of HA, NA, M1, and NP in the viral particles after passaging (Figure 3C).

A pair of PLPs was then designed for each gene segment such that the terminal nucleotide of the 3

arm hybridized to the nucle- otide substitution site (Figure 3D). Using this approach, only a properly hybridized probe can ligate and produce a RCP (Fig- ure S3A). As an initial test, the labeling was performed on vRNA isolated from WSN and WSN

particles in solution and the RCPs were quantified using automated amplified single-mole- cule detection (Figure S3B) (Go¨ransson et al., 2012; Jarvius et al., 2006). In solution, the PLP pairs detected each gene segment from the two IAVs with similar efficiency (Figure 3E), and several pairs showed low cross-reactivity, including the HA gene segment pair (Figure S3C). We then analyzed RNA isolated from MDCK cells and mouse lung tissue infected with WSN, infected with WSN

, or co-infected with WSN and WSN

. Using the HA PLP pair, the viruses responsible for the in vitro and in vivo infections were appropriately identified, and similar RCP ratios were observed for samples that received equivalent numbers of infectious WSN and WSN

particles (Figure 3F).

A B

D

C

Figure 1. Single-Cell IAV Gene Segment vRNA Labeling Approach (A) Diagram of an IAV and the eight single-stranded vRNA gene segments.

(B) PLP design for detecting the IAV gene segments. Each PLP contains20-nucleotide 5⁰and 3⁰arms that hybridize to adjacent sequences in a gene segment and barcodes for identifying the gene segment (colored) and strain (brown).

(C) In situ labeling of IAV gene segments. cDNAs are synthesized from each vRNA in fixed cells. Templates are digested using RNase H. PLPs are hybridized and ligated to produce a template for synthesizing a rolling-circle product (RCP), which creates a concatemer of complementary barcode copies that are detected using complementary fluorescently labeled barcode probes.

(D) Single-cell RCP visualization and quantification. RCPs are detected by cycles of barcode probe hybridization (up to five probes at a time), fluorescence microscopy imaging, and stripping. RCP images from different cycles are compiled and aligned, and the RCPs are assigned to the nuclei and cell segments defined by nuclei staining and cell autofluorescence.

Single-Nucleotide Resolution of IAV Infections in Cells and Lung Tissue

Although deep sequencing is more suited to identify unknown RNA sequence variants, our labeling approach is inexpensive

and uses a straightforward analysis that offers the advantage of spatial information in situ. This unique property has the poten- tial to enhance the cellular analysis of viral genome trafficking, heterogeneity and selection, gene segment reassortment, and

F C

A B

E D

G

Figure 2. PLPs Detect IAV Gene Segments during Entry and Replication

(A) The IAV genome (vRNA) replication cycle. After endocytosis, IAVs fuse with the endosomal membrane, releasing the vRNAs into the cytosol. The vRNAs are replicated in the nucleus by the vRNA-dependent RNA polymerase (vRdRP), assembled, and exported for packaging at the plasma membrane. BFLA1 inhibits endosomal acidification and fusion. LMB inhibits the nuclear export.

(B) qPCR analysis of cellular and extracellular (media) NP and NA vRNA levels. vRNA levels were normalized to 0 hr p.i. (cellular) or 4 hr p.i. (extracellular). Mean values with the SD from two independent experiments are displayed.

(C) The in situ labeling of IAV gene segments in untreated and BFLA1-treated (100 nM) MDCK cells 2 hr p.i. The percentage of gene segment RCPs that co-localize with the nucleus and the SD were determined from two independent experiments analyzing1,000 cells. The box-and-whisker plot (5%–95% CI) shows the nucleus image occupancy from the corresponding image fields.

(D) In situ labeling of IAV gene segments in untreated or LMB-treated (10 nM) MDCK cells at 8 hr p.i. LMB was added 5 hr p.i. The RCP nuclear co-localization and SD were determined from three independent experiments.

(E) Representative image of the eight IAV gene segment RCPs generated in MDCK cells 8 hr p.i. with an MOI of0.3. Cell nuclei were stained with Hoechst. Inset shows the higher magnification of a cell.

(F) Graph showing the percentage of1,000 cells that contained the indicated number of distinct gene segment RCPs in the nucleus at 8 hr p.i.

(G) Correlation plot showing the relative increases in the NP vRNA levels versus the number of NP gene segment RCPs in500 infected cells at the indicated times post-infection. Values normalized to 4 hr p.i. and are shown as± SD.

other aspects in which sequencing is not ideal. Therefore, we examined the single-nucleotide resolution in situ by labeling WSN- and WSN

-infected MDCK cells with the gene segment PLP pairs individually (Figures 4A and S4) and together (Fig- ure 4B). In both cases, RCPs were primarily generated to the appropriate virus. The specificity was confirmed using MDCK cells that were independently infected with WSN or WSN

and combined before the labeling (Figure 4C). Finally, we tested the ability to analyze in vivo samples by labeling lung tissue sec- tions isolated from mice infected with WSN, infected with WSN

, or co-infected with WSN and WSN

. While the labeling was less efficient, the virus responsible for the infection was accurately identified (Figure 4D), which opens up a number of potential in vivo applications for this technique.

Analysis of IAV Entry Kinetics by Gene Segment Labeling

Despite the central role of a viral genome in defining an infection, a comprehensive analysis of IAV gene segment trafficking during entry has not been performed due to technical limitations. There- fore, we used the labeling approach to directly monitor IAV gene segment trafficking to the nucleus using two MOIs. Because IAVs are endocytosed within minutes (Matlin et al., 1981), cell binding was synchronized at 4

C, and the gene segments were labeled before and after initiating entry by shifting the cells to 37

C (Figure 5A). With low and high MOIs, few RCPs were observed from bound IAVs (0 and 5 min) (Figures 5C–5E), likely because the cell-bound viral particles are inefficiently fixed.

However, the fixation requirement enabled the measurement of IAV entry kinetics as the labeling increased upon internalization.

With a low MOI, the percentage of cells associated with a gene segment RCP increased until 30 min post-binding (Figure 5B),

reaching a higher level than predicted by the MOI, presumably due to defective particles. When a higher MOI was used, RCPs were assigned to most cells earlier, with a half-time of 5 min (Figure 5B), consistent with entry measurements by single-parti- cle live imaging (Lakadamyali et al., 2003; Rust et al., 2004).

The single-cell low MOI data also showed that the number of RCPs per cell plateaued from 30 min until replication began at 180 min (Figures 5C, 5D, and S5). When the RCPs from the high MOI were analyzed together or individually, the number per cell continuously increased even though replication was also initiated at 180 min (Figures 5C, 5E, and S5B). The combi- nation of the results from the in situ labeling and qPCR indi- cated that two IAV entry steps are stochastic: (1) cell entry, when more viral particles are bound, the entry half-time for a single virus decreases but it takes longer for all particles to enter the cell, and (2) replication; although the replication initia- tion time is fixed because of the lengthy gene segment traf- ficking process, the replication rate is determined by the num- ber of gene segment templates and bound polymerases that reach the nucleus.

Classification of IAV Infection Stages by Gene Segment Labeling

We next asked whether single-cell labeling can be used to clas- sify IAV infection stages in cells. First, we established a RCP per cell cutoff ( 20) that distinguishes early replication stages from later ones (Figure S6). We then combined this parameter with the nuclear co-localization values from the entry inhibitor BFLA1 ( 40%) and nuclear export inhibitor LMB (80%) to assign each cell to one of five stages, as shown in Figure 5F.

As expected, most infected cells were initially assigned to stage 1 (cell entry), and over time, they shifted to stage 2 (nuclear

D E

A B

F

C

with Single-Nucleotide Variations

(A) Diagram depicting the single-nucleotide dif- ferences in each WSN and WSN^Isogene segment.

Amino acid codons for the indicated IAV gene are given as a position reference.

(B) Viral titers obtained during the rescue of WSN and WSN^Isoby reverse genetics in co-cultured 293T and MDCK cells.

(C) Representative immunoblots showing the relative NA, HA, NP, and M1 protein levels in iso- lated WSN and WSN^Isoparticles.

(D) PLP pair design to differentiate IAV gene seg- ments that differ by one nucleotide. In each gene segment PLP pair, the terminal 3⁰ nucleotide (strain-specific nucleotide) is designed to hybrid- ize with the respective vRNA nucleotide substitu- tion in the WSN^Isoor WSN sequence.

(E) The recognition for each gene segment PLP pair was determined in solution (Figure S3) by quantifying the number of RCPs generated from the same amount of vRNA.

(F) RNA was extracted from MDCK cells (16 hr p.i.) or mouse lung tissue harvested 12 hr p.i. with the indicated viruses and RCPs were generated with the HA PLP pair. The graph shows the percentage of WSN (green) and WSN^Iso(red) RCPs from each sample.

import) and stage 3 (replication), the latter of which combines cells in which the vRNAs reached the nucleus with cells actively replicating vRNAs (Figure 5G). At the low MOI, the stage 1 and 2 populations did not fully disappear, likely because they represent defective infections that are masked at higher MOIs. The higher MOI also caused a sharper stage 1, earlier stage 2 initiation, and more pronounced stage 3 population that dropped at 180 min when vRNA-replicating cells entered stage 4 (nuclear export).

An extended analysis at the low MOI showed a later shift from stage 3 to stages 4 and 5 (exported) that correlated with viral release (Figures S7A–S7C). In agreement with previous work (Chou et al., 2013; Lakdawala et al., 2014), the gene segments did not show any bias in nuclear export (Figures S7D and S7E).

These results demonstrate how IAV gene segment labeling can be used to analyze the entry process and to define cell infection stages within large populations.

The IAV Replication Cycle Defines the Time Window for Cell Co-infections

Reassortment can occur when two IAVs infect and replicate within the same cell (Steel and Lowen, 2014). However, the complex trafficking of the gene segments to the nucleus is likely to limit the time window for productive secondary IAV infections. To investigate this concept, we designed a co-infec- tion experiment in which cells receive a primary infection (WSN) followed by a secondary infection (WSN

) at increasing time intervals and all gene segments were labeled 480 min (8 hr) af- ter the primary infection (Figure 6A). This endpoint was chosen because the primary infection caused increasing cell loss at later times. To control for the decreasing duration of the sec- ondary infection, a set of WSN

-infected cells was generated without a primary infection (Figure 6B). As expected in the con- trol samples, the decreasing infection times corresponded with

A B C

D

Figure 4. In Situ IAV Genome Labeling Can Identify Infections with Single-Nucleotide Precision

(A) MDCK cells infected with the indicated virus were labeled 6 hr p.i. using the individual gene segment PLP pairs (Figure S4), and the resulting number of RCPs for each PLP are displayed as a box-and-whisker plot.

(B) MDCK cells infected with WSN or WSN^Isowere labeled 6 hr p.i. with the eight PLP pairs. Strain-specific barcodes were used for visualization, and the RCP per cell numbers for each PLP set are shown for25,000 cells.

(C) MDCK cells infected with WSN or WSN^Isowere trypsinized 2 hr p.i., the cells were combined, reseeded on the same slide, and labeled 4 hr later with the eight WSN and WSN^IsoPLP pairs. A representative image is shown with arrows indicating WSN-infected (green) or WSN^Iso-infected (red) cells. The an inset shows the nuclei staining.

(D) Representative images of lung cryosections from mice 2 days post-inoculation with WSN or WSN^Iso, or co-inoculation with WSN and WSN^Iso, which were labeled with the eight WSN and WSN^IsoPLP pairs. The insets display the mosaics of the lung sections with the respective image region. The number of RCPs and the SD from the indicated PLP set was obtained from analyzing two entire lung cryosection mosaics. RCP signals are displayed as enlarged circles for better visualization.

a lower number of RCPs per cell (4 hr) (Figures 6B and S8A) and a shift to a more pronounced stage 3 (replication) popula- tion (Figure 6D).

In the co-infection samples, the number of WSN

RCPs per cell decreased when the virus was added more than 45 min after WSN (Figures 6C and S8B), whereas the number of WSN RCPs A

F

C

E D

B

G

Figure 5. Single-Cell Analysis of IAV Entry byIn Situ Gene Segment Labeling

(A) Infection procedure for IAV gene segment labeling during viral entry. WSN at either a low (0.3) or a high (3) MOI was bound to MDCK cells at 4^C, unbound virus was removed, and entry was initiated by shifting the cells to 37^C. Labeling was performed at the indicated times post-binding.

(B) Percentage of cells possessing one or more RCPs plotted with respect to the time post-binding for the low and high MOI infections. For each time point,

1,000 cells were examined.

(C) The number of gene segment RCPs per infected cell is shown as a box-and-whisker plot for1,000 cells at each time point following infection with the low (left panel) or high (right panel) MOI.

(D and E) Representative images of the IAV gene segment labeling at the indicated times post-binding with the low (D) and high (E) MOI. The RCPs were visualized with the strain-specific barcode (low and high MOI) and the gene segment barcodes (high MOI).

(F) Cell diagrams of the parameters for classifying the five infection stages based on the number of labeled gene segments and the percentage of nuclear co- localization.

(G) Distribution of the infection stages in the cell population at the indicated times following infection with the low (left panel) or high (right panel) MOI. More than 12,000 cells were analyzed.

A

D

E

B C

Figure 6. Co-labeling of IAV Genomes Reveals the Time Window for Productive Cell Co-infections

(A) Diagram of the co-infection experiment. Cells were bound with the primary IAV (WSN) at 4^C using an MOI of3, unbound virus was removed, and the secondary IAV (WSN^Iso) was added at an equal MOI at the indicated times (red arrows) after the 37^C temperature shift. Cells were labeled 8 hr p.i. with WSN, resulting in increasingly shorter durations of the WSN^Isoinfections.

(B) Representative control images of the WSN^Isogene segment labeling to account for the shorter durations of the infection. Cells were only infected with WSN^Iso. (C) Representative images of the WSN (green) and WSN^Iso(red) gene segment labeling in co-infected cells. The secondary WSN^Isoinfections were initiated at the indicated times after WSN binding (listed on the left). The durations of the WSN^Isoinfections are listed on the right.

(D) Distribution of the WSN^Isoinfection stages in the control samples (no WSN primary infection) as the duration of the infection becomes shorter.

(E) Left: distribution of the WSN^Isoinfection stages as the virus was added at increasing time intervals (listed in parentheses) after the primary WSN infection. For each time point,2,000 cells were examined. Right: combined distribution of the WSN infection stages in the cell populations from all samples.

remained largely unaffected (Figures 6C and S8C). The decrease in the WSN

RCPs was accompanied by a small decrease in the percentage of WSN

-infected cells, which was distinct from the one associated with shorter infection times (Figure 6E). As the time interval between the two infections increased, the WSN

-infected cell population also showed a pronounced shift to stage 1 (cell entry), instead of the expected shift to stage 3 (replication) that is attributed to the shorter infection duration.

Because the WSN

gene segments were still visible in the cells, it implies the loss in productive WSN

infections (stages 3–5) occurred at these times because the primary infection impairs the nuclear import of the WSN

gene segments, but not their entry into the cell. By considering the temporal aspect of the replication cycle, these results show that productive IAV cell co-infections are restricted once an IAV genome has reached the nucleus and initiated the replication process (Figure 7).

DISCUSSION

All viruses use particular cell compartments for genome replica- tion. Despite the advantages of monitoring cell infections through the viral genome, protein-based reporters are more commonly used due to technical limitations. This is especially true for RNA viruses. Therefore, we sought to develop an approach to detect the location of RNA viral genomes in situ and based the methodology on the versatile PLP labeling prop- erties (Ke et al., 2013; Larsson et al., 2010). We used IAVs as a test case for the technique due to the challenges associated with labeling eight relatively short RNA gene segments and the limited data about the IAV life cycle before replication initiation.

By combining the PLP labeling with a single-cell analysis, the approach provided semiquantitative data about the IAV gene segments in individual cells and could distinguish cell infections by IAVs that vary by single nucleotides. The addition of a tempo- ral parameter made it possible to follow gene segment dynamics

Figure 7. The Primary IAV Replication Stage Defines the Cell Susceptibility to Secondary Infections

The model displays the distribution of the five IAV replication stages that were observed by gene segment PLP labeling over the course of an infection. Each stage is diagramed with respect to the cell population and the corresponding infection time that was observed using a single (primary) IAV infection. The time window and the stages of a primary IAV infection in which the cell is suscepti- ble to co-infection by a secondary IAV are shown above the graph.

from entry through replication (Figure 7), monitor the stochastic aspects of IAV en- try, and define the time window for pro- ductive cell co-infections.

The PLPs, which facilitate the labeling, also provide in situ resolution of IAV sequences that is unattainable with con- ventional methods. We used the single- nucleotide specificity to alleviate bias in the co-infection experiments that showed the replication of a second IAV is restricted once the replication of the primary IAV initiates (Figure 7). While the time frame may vary among strains and cell types, it is reasonable to assume that reassortment events happen less frequently when replication of a secondary IAV genome is restricted. In our system, the restriction was observed when the time interval between infections reached

2 hr, which supports previous results examining the temporal contributions to reassortment in IAVs (Marshall et al., 2013).

However, the labeling approach revealed that secondary IAV gene segments still enter the cell but seem unable to reach the nucleus for replication. From a mechanistic aspect, impairing nu- clear import once replication begins would benefit IAVs, because it would prevent new gene segments from recycling into the nu- cleus when they are needed for viral packaging at the plasma membrane.

In addition to being ideal for investigating the cell machinery that facilitates viral genome import and trafficking, this approach can potentially make several aspects of viral infections in vivo more accessible and be developed further for diagnostic appli- cations. For instance, a sequence analysis revealed that only seven PLPs are needed to identify and differentiate among almost all known IAVs and influenza type B viruses (Figure S9).

By including additional PLP permutations, HA and NA subtypes can be identified (Gyarmati et al., 2008), as well as the species of origin for each gene segment. However, this strategy would require further development for clinical specimens.

Viral protein labeling is a sufficient infection reporter in many

cases, but genome labeling is more applicable for differentiating

highly similar viruses, especially those with changes in inacces-

sible epitopes, such as transmembrane domains (da Silva et al.,

2015; Nordholm et al., 2013), or epitopes that are removed (e.g.,

signal sequences) or shielded by glycans (Krammer and Palese,

2015; Stewart-Jones et al., 2016). Genome labeling is also more

appropriate for analyzing different aspects of entry, which are

crucial parts of an infection. Live cell imaging of the genome during entry is ideal, but technical constraints like retaining in- fectivity and the number of gene segments are limiting. To over- come these issues, previous investigations have incorporated fluorescent dyes into viral membranes to visualize viral entry and membrane fusion, in which the association of the signal with the viral genome is lost (Lakadamyali et al., 2003; Rust et al., 2004). Although our technique labels the viral gene seg- ments, it has shortcomings. Similar to all FISH-based methods, the resolution is not sufficient to elucidate gene segment co- localization, and the target sequences must be known. The PLPs also have limited sensitivity and the recognition efficiencies vary, but modifications such as increasing the number of PLPs per target can aid in solving these issues.

In summary, the versatility of this approach for in vitro and

of specificity and spatial resolution. The technique is relatively inexpensive, and by altering the PLP design, it can be applied to any RNA virus. Our results showing that IAVs must satisfy both spatial (infect the same cell) and temporal parameters to establish productive cell co-infections is likely a contributing factor to the low reassortment frequencies reported for human infections (Sobel Leonard et al., 2017). They also raise the ques- tion of how IAVs limit the nuclear import of gene segments once replication has initiated, because nuclear import is a dominant process. With this method, addressing these types of questions for any RNA virus will become more accessible, and the unique ability to label viral genomes in tissue sections makes it possible to start examining RNA viral replication processes in vivo.

EXPERIMENTAL PROCEDURES

Reagents and Plasmids

MDCK and HEK293T cells were purchased from LGC Standards. DMEM, Opti-MEM (OMEM), trypsin, fetal bovine serum (FBS), penicillin and strepto- mycin (P/S), Hoechst, cell mounting media, dinucleotide triphosphates (dNTPs), and RiboLock RNase inhibitor were all obtained from Thermo Fisher Scientific. Leptomycin (LMB), BFLA1, Tween 20, formamide, diethyl pyrocar- bonate (DEPC), paraformaldehyde (PFA), and tosyl phenylalanyl chloromethyl ketone (TPCK)-trypsin were from Sigma-Aldrich. LT-1 transfection reagent was purchased from Mirus, phi29 DNA polymerase was purchased from O-Link Bioscience, and BSA was purchased from New England Biolabs.

PLPs and barcode detection probes were obtained from Integrated DNA Technologies, and the LNA-modified cDNA primers were obtained from Exiqon (seeSupplemental Informationfor sequences). TransciptMe reverse transcriptase (tRT) and RNase H were purchased from DNA-Gdansk, the RNeasy mini kit was purchased from QIAGEN, Ampligase was purchased from Epicenter, and HybriWell secure seals were purchased from Grace Bio-Lab. Monoclonal PA antiserum (NR-19225) and polyclonal HA antisera (NR-15696) were obtained through BEI Resources, National Institute of Allergy and Infectious Diseases (NIAID), NIH. The influenza WSN reverse genetic plas- mids were kindly provided by Robert Webster (St. Jude Children’s Research Hospital). WSN^Isowas created by site-directed mutagenesis using the WSN plasmids as templates.

Cell Culture, IAV Reverse Genetics, and Viral Titer Determination MDCK and HEK293T cells were cultured in DMEM containing 10% FBS and 100 U/mL P/S in a 5% CO2humidified atmosphere at 37^C. WSN and WSN^Iso viruses were rescued by a previously described reverse transfection approach (da Silva et al., 2015). Briefly, 1.5 mL of trypsinized MDCK cells, at a density of

10⁶cells/mL in OMEM 10% FBS, was added to each T-25 cm²flask. WSN or WSN^Isoplasmids (1mg of each) were mixed in 1.5 mL OMEM with 20 mL LT-1

for 20 min at room temperature, and 1.5 mL trypsinized 293T cells in OMEM at a density of10⁶cells/mL were added to each mixture. The mixture was incubated 15 min and transferred to the MDCK cell flask at 37^C. At 24 hr post-transfection, media was replaced with infection media (DMEM with 0.1% FBS, 0.3% BSA, 100 U/mL P/S, and 4mg/mL TPCK-trypsin). Viral titers were determined by calculating the median tissue culture infectious dose (TCID50/mL) on MDCK cells as described (da Silva et al., 2015; Reed and Muench, 1938). WSN and WSN^Isosequencing (Eurofins MWG Operon) was performed using PCR-amplified cDNA copies of the vRNAs extracted from isolated particles.

Viral Particle Isolation and Immunoblotting

WSN or WSN^Isoviral particles were passaged in MDCK cells, the culture media was clarified by centrifugation (1,0003 g; 5 min), and the viral titers were measured. Equal infectious particle numbers were then sedimented (115,0003 g; 60 min) through a 0.2 M sucrose cushion as previously described (da Silva et al., 2015). The isolated particles were resuspended in equivalent volumes of Laemmli sample buffer, resolved by SDS-PAGE, transferred to a polyvinylidene fluoride (PVDF) membrane, and processed (da Silva et al., 2013) using the HA antisera and antisera raised against purified recombinant NA, M1, and NP expressed in E. coli.

MDCK Cell Infections, Co-infections, BFLA1 and LMB Treatment, and Fixation

MDCK cells grown on 3.5 cm dishes (qPCR analysis), or chamber slides (NP, PA, and gene segment labeling), were incubated at 4^C with infection media containing WSN at a low (0.3) or high (3) MOI for 30 min to facilitate binding.

Unbound virus was removed, cells were washed twice with 4^C PBS, new 37^C infection media was added, and the cells were placed at 37^C. For co-infection experiments, WSN was bound to cells at 4^C using an MOI of

3, unbound virus was removed, and WSN^Isowas added at an MOI of3 to the cells before the 37^C temperature shift to initiate entry or at the indicated time after the shift. To prevent IAV fusion, cells were treated with 100 nM of the vacuolar ATPase inhibitor BFLA1 from 1 hr before binding until labeling 2 hr p.i.

To inhibit CRM-1-mediated nuclear export, cells were treated with 10 nM LMB 5 hr p.i. At the indicated times, cells were fixed with cold 4% (w/v) PFA in PBS (pH 7.4) treated with DEPC (PBS-D) for 20 min, followed by two PBS-D washes and three 5 min fixations with increasing ethanol concentrations (70%, 85%, and 99%). Slides were stored at80^C.

In Situ PLP Ligation, Rolling-Circle Amplification, and Barcode Labeling

The following procedure was carried out in 50mL hybridization chambers created by attaching secure seals to the slide. Cells were washed with PBS-D containing 0.05% Tween 20 (PBST-D), treated with 0.1 M HCl-DEPC for 10 min, and washed. cDNA synthesis was then performed for 3 hr at 37^C using 20 U/mL tRT in 13 tRT buffer with 0.5 mM dNTPs, 0.2 mg/mL BSA, 1mM LNA-modified cDNA gene segment primers, and 0.8 U/mL Ribo- Lock RNase inhibitor. After washing (PBST-D), cDNAs were fixed with 4%

(w/v) PFA for 10 min and washed (PBST-D). PLPs (0.1mM of each) were annealed and ligated using 0.5 U/mL Ampligase in 13 Ampligase reaction buffer containing 20% formamide, 0.2mg/mL BSA, 50 mM KCl, and 0.4 U/mL RNase H for 30 min at 37^C followed by 45 min at 45^C. Cells were washed (PBST-D) and RCPs were synthesized using 1 U/mL phi29 polymerase in 13 phi29 buffer with 5% glycerol, 0.25 mM dNTPs, and 0.2 mg/mL BSA for 16 hr at 25^C. Cells were washed and RCPs were labeled after 20 min at 37^C using 0.1mM of each barcode probe in 30 mM saline-sodium citrate buffer (pH 7.0) containing 300 mM NaCl and 20% formamide. After a final wash, cell nuclei were stained with 0.5mg/mL Hoechst for 10 min, secure seals were removed, and the cells were dehydrated for 1 min in an ethanol series (70%, 85%, and 99%) and mounted. To exchange barcodes, coverslips were removed, new secure seals were placed, and three 5 min incubations were performed with 65% formamide at 65^C before washing (PBST-D).

Mouse Infections, Lung Sectioning, and IAV vRNA Labeling

Experiments were performed in accordance with the local ethical committee (Stockholms Norra djurfo¨rso¨ksetiska na¨mnd). Male 6- to 8-week-old C57BL/6

mice (Janvier Labs) were sedated by intraperitoneal injection of 80 mg/kg ke- tamine and 5 mg/kg xylazine (Rompun, Bayer) and intranasally inoculated with

5 3 10⁴plaque-forming units (p.f.u.) of WSN or WSN^Isoor co-infected with WSN and WSN^Iso(2.5 3 10⁴p.f.u. of each). Mice were sacrificed 2 days post-inoculation and lungs were inflated with a 3:1 mixture of 4% PFA to FSC22 cryosectioning media (Leica), tied off at the trachea, extracted, and fixed in 4% PFA for 4 hr. Lungs were placed in base molds (Leica) covered with FSC22 and frozen at80^C, and 20mm airway sections were cut with a Biosystems CM3050 cryostat (Leica). Sections were attached to super-frost slides, washed twice (PBST-D), fixed with 4% PFA for 45 min, washed, and treated with 2 mg/mL pepsin in 0.1 M HCl-DEPC for 1.5 min. The sections were dehydrated with a 2 min ethanol series (70%, 85%, and 99%) and secure seals were mounted. PLP labeling was performed in 100mL reactions as described for cells with the following changes: cDNAs were synthesized for 16 hr at 45^C and fixed with 4% (w/v) PFA for 30 min, and RCPs were synthe- sized at 37^C and labeled for 30 min.

Image Acquisition and Processing

To capture RCP signals, multiple focal plane images were acquired using an AxioplanII epifluorescence microscope (Zeiss) and merged using maximum- intensity projections (Zen software). Single-cell analysis was performed using CellProfiler software (Kamentsky et al., 2011). To increase the nucleus seg- mentation accuracy, a Gaussian filter correction was applied to the Hoechst staining and refined using shape descriptors. Individual cells were defined by automated thresholding using the segmented nuclei as a seed, followed by watershed segmentation using the cytoplasmic autofluorescence signal.

For RCP counting, top-hat filtering was used to enhance the RCP resolution and each RCP signal, assigned by localization to a particular cell or nucleus, was tabulated.

qRT-PCR

cDNAs of segment 5 (NP) and segment 6 (NA) were generated from equivalent amounts of cellular RNA using segment-specific primers (0.25mM) with a vRNA tag (5⁰-GGCCGTCATGGTGGCGAAT-3⁰) at the 5⁰ end as previously described (Kawakami et al., 2011). qPCR was performed with a CFX96 Real-Time System (Bio-Rad) and iTaq Universal SYBR Green Supermix ac- cording to the manufacturer’s instructions (Bio-Rad). Relative vRNA quantities were calculated using CFX manager software v.3.1 (Bio-Rad) and a standard curve for each vRNA. Cycling parameters were 95^C for 30 s, followed by 40 cycles of 95^C for 5 s and 60^C for 30 s.

In-Solution Labeling of vRNAs from IAVs, MDCK Cells, or Mouse Lung Tissue

RNA templates were extracted using an RNeasy mini kit from sedimented WSN and WSN^Isoparticles, or1 3 10⁵MDCK cells or30 mg of mouse lung tissue. MDCK cells were infected with WSN or WSN^Iso(MOI of0.5), or co-infected with equal infectious units of WSN and WSN^Isofor 16 hr. Lung tissue was isolated from 8-week-old female BALB/c mice, intranasally infected under isoflurane sedation with 5.83 10⁴p.f.u. of WSN or WSN^Iso, or co- infected with 2.93 10⁴p.f.u. of both WSN and WSN^Isodiluted in 50mL PBS.

Following cervical dislocation 12 hr p.i., lungs were harvested and transferred to RNAlater (QIAGEN) for 12 hr at 4^C and stored at80^C. The animal exper- iments were performed in strict accordance with the guidelines of the German animal protection law. All animal protocols were approved by the relevant German authority (Beho¨rde f€ur Stadtentwicklung und Umwelt Hamburg).

cDNAs were synthesized using equal RNA amounts, 0.1mM of the gene segment primers, and tRT according to the manufacturer’s instructions.

Each cDNA (10mL) was combined with a 10 mL ligation mix containing 200 pM PLPs, 0.4mg/mL BSA, 23 Ampligase buffer, and 500 mU/mL Ampligase enzyme for 40 min at 55^C. The ligated PLPs were mixed with 200 mU/mL phi29 polymerase in 13 phi29 buffer, 125 mM dNTPs, and 0.2 mg/mL BSA for 2 hr at 37^C. The reaction was terminated at 65^C for 5 min and the RCPs were labeled with 5 nM barcode probes in hybridization buffer (50 mM Tris- HCl [pH 7.5], 5 mM EDTA, 1 M NaCl, and 0.01% Tween 20) for 2 min at 75^C and 15 min at 55^C. RCPs from fixed volumes (20mL) were quantified by automated amplified single-molecule detection (ASMD) with an Aquila 400 (Q-Linea).

Statistical Analysis

Means from technical replicates of qPCR experiments were used for normal- ization, and the SD was calculated from the means of independent replicates.

Box-and-whisker plots show the 5%–95% confidence interval (CI) for the indi- cated single-cell samples, and the percentages of nuclear localizations are presented as population means with the SD. Infection stage analysis was per- formed using all raw single-cell data collected for each condition. All statistical analyses were performed with GraphPad Prism 6.0.

SUPPLEMENTAL INFORMATION

Supplemental Information includes Supplemental Experimental Procedures and nine figures and can be found with this article online athttp://dx.doi.org/

10.1016/j.celrep.2017.06.021.

AUTHOR CONTRIBUTIONS

Conceptualization, R.D., M.N., D.D., and I.H.-N.; Primary Methodology and Investigation, D.D. and I.H.-N.; Supporting Methodology and Investigation, H.W., H.O¨ ., and X.Q.; Animal Methodology and Investigation, P.R.-I., N.M.K., V.S., K.H., P.M., B.H.-N., G.G., and I.H.-N. Writing – Original Draft, R.D., M.N., D.D., and I.H.-N.; Writing – Review & Editing, all co-authors.

ACKNOWLEDGMENTS

We thank Sara Badreh for help with procedure optimization, Alex Sountoulidis and Christos Samakovilis for technical assistance with lung tissue preparation and sectioning, and Johan Nordholm for critical reading of the manuscript.

This work was supported by grants from the Swedish Research Council (R.D., M.N., and B.H.-N.), Swedish Foundation for Strategic Research (R.D., B.H.-N., and Flu-ID to M.N.), FORMAS (Biobridges Project to M.N.), Knut and Alice Wallenberg Foundation (B.H.-N.), Carl Trygger Foundation (R.D.), and Harald Jeanssons Stiftelse (R.D.)

Received: March 1, 2017 Revised: April 20, 2017 Accepted: June 6, 2017 Published: July 5, 2017

REFERENCES

Amorim, M.J., Bruce, E.A., Read, E.K., Foeglein, A., Mahen, R., Stuart, A.D., and Digard, P. (2011). A Rab11- and microtubule-dependent mechanism for cytoplasmic transport of influenza A virus viral RNA. J. Virol. 85, 4143–4156.

Andersson, C., Henriksson, S., Magnusson, K.E., Nilsson, M., and Mirazimi, A.

(2012). In situ rolling circle amplification detection of Crimean Congo hemor- rhagic fever virus (CCHFV) complementary and viral RNA. Virology 426, 87–92.

Arranz, R., Coloma, R., Chicho´n, F.J., Conesa, J.J., Carrascosa, J.L., Val- puesta, J.M., Ortı´n, J., and Martı´n-Benito, J. (2012). The structure of native influenza virion ribonucleoproteins. Science 338, 1634–1637.

Bane´r, J., Nilsson, M., Mendel-Hartvig, M., and Landegren, U. (1998). Signal amplification of padlock probes by rolling circle replication. Nucleic Acids Res. 26, 5073–5078.

Boni, M.F. (2008). Vaccination and antigenic drift in influenza. Vaccine 26 (Suppl 3), C8–C14.

Borderı´a, A.V., Rozen-Gagnon, K., and Vignuzzi, M. (2016). Fidelity variants and RNA quasispecies. Curr. Top. Microbiol. Immunol. 392, 303–322.

Breen, M., Nogales, A., Baker, S.F., and Martı´nez-Sobrido, L. (2016). Replica- tion-competent influenza A viruses expressing reporter genes. Viruses 8, E179.

Brister, J.R., Ako-Adjei, D., Bao, Y., and Blinkova, O. (2015). NCBI viral ge- nomes resource. Nucleic Acids Res. 43, D571–D577.

Brooke, C.B. (2014). Biological activities of ‘noninfectious’ influenza A virus particles. Future Virol. 9, 41–51.

Capobianchi, M.R., Giombini, E., and Rozera, G. (2013). Next-generation sequencing technology in clinical virology. Clin. Microbiol. Infect. 19, 15–22.

Chou, Y.Y., Vafabakhsh, R., Doganay, S., Gao, Q., Ha, T., and Palese, P.

(2012). One influenza virus particle packages eight unique viral RNAs as shown by FISH analysis. Proc. Natl. Acad. Sci. USA 109, 9101–9106.

Chou, Y.Y., Heaton, N.S., Gao, Q., Palese, P., Singer, R.H., and Lionnet, T.

(2013). Colocalization of different influenza viral RNA segments in the cytoplasm before viral budding as shown by single-molecule sensitivity FISH analysis.

PLoS Pathog. 9, e1003358.

Crosetto, N., Bienko, M., and van Oudenaarden, A. (2015). Spatially resolved transcriptomics and beyond. Nat. Rev. Genet. 16, 57–66.

da Silva, D.V., Nordholm, J., Madjo, U., Pfeiffer, A., and Daniels, R. (2013).

Assembly of subtype 1 influenza neuraminidase is driven by both the trans- membrane and head domains. J. Biol. Chem. 288, 644–653.

da Silva, D.V., Nordholm, J., Dou, D., Wang, H., Rossman, J.S., and Daniels, R.

(2015). The influenza virus neuraminidase protein transmembrane and head domains have coevolved. J. Virol. 89, 1094–1104.

de Graaf, M., and Fouchier, R.A. (2014). Role of receptor binding specificity in influenza A virus transmission and pathogenesis. EMBO J. 33, 823–841.

Dimmock, N.J., and Easton, A.J. (2014). Defective interfering influenza virus RNAs: time to reevaluate their clinical potential as broad-spectrum antivirals?

J. Virol. 88, 5217–5227.

Drake, J.W. (1993). Rates of spontaneous mutation among RNA viruses. Proc.

Natl. Acad. Sci. USA 90, 4171–4175.

Dulbecco, R., and Vogt, M. (1953). Some problems of animal virology as stud- ied by the plaque technique. Cold Spring Harb. Symp. Quant. Biol. 18, 273–279.

Eisfeld, A.J., Kawakami, E., Watanabe, T., Neumann, G., and Kawaoka, Y.

(2011). RAB11A is essential for transport of the influenza virus genome to the plasma membrane. J. Virol. 85, 6117–6126.

Eisfeld, A.J., Neumann, G., and Kawaoka, Y. (2015). At the centre: influenza A virus ribonucleoproteins. Nat. Rev. Microbiol. 13, 28–41.

Elton, D., Simpson-Holley, M., Archer, K., Medcalf, L., Hallam, R., McCauley, J., and Digard, P. (2001). Interaction of the influenza virus nucleoprotein with the cellular CRM1-mediated nuclear export pathway. J. Virol. 75, 408–419.

Fonville, J.M., Marshall, N., Tao, H., Steel, J., and Lowen, A.C. (2015). Influenza virus reassortment is enhanced by semi-infectious particles but can be sup- pressed by defective interfering particles. PLoS Pathog. 11, e1005204.

Gabriel, G., Klingel, K., Otte, A., Thiele, S., Hudjetz, B., Arman-Kalcek, G., Sauter, M., Shmidt, T., Rother, F., Baumgarte, S., et al. (2011). Differential use of importin-a isoforms governs cell tropism and host adaptation of influ- enza virus. Nat. Commun. 2, 156.

Gaspar, I., and Ephrussi, A. (2015). Strength in numbers: quantitative single- molecule RNA detection assays. Wiley Interdiscip. Rev. Dev. Biol. 4, 135–150.

Gavazzi, C., Yver, M., Isel, C., Smyth, R.P., Rosa-Calatrava, M., Lina, B., Moule`s, V., and Marquet, R. (2013). A functional sequence-specific interaction between influenza A virus genomic RNA segments. Proc. Natl. Acad. Sci. USA 110, 16604–16609.

Gire, S.K., Goba, A., Andersen, K.G., Sealfon, R.S., Park, D.J., Kanneh, L., Jal- loh, S., Momoh, M., Fullah, M., Dudas, G., et al. (2014). Genomic surveillance elucidates Ebola virus origin and transmission during the 2014 outbreak. Sci- ence 345, 1369–1372.

Go¨ransson, J., Ke, R., Nong, R.Y., Howell, W.M., Karman, A., Grawe´, J., Sten- berg, J., Granberg, M., Elgh, M., Herthnek, D., et al. (2012). Rapid identification of bio-molecules applied for detection of biosecurity agents using rolling circle amplification. PLoS ONE 7, e31068.

Gyarmati, P., Conze, T., Zohari, S., LeBlanc, N., Nilsson, M., Landegren, U., Bane´r, J., and Bela´k, S. (2008). Simultaneous genotyping of all hemagglutinin and neuraminidase subtypes of avian influenza viruses by use of padlock probes. J. Clin. Microbiol. 46, 1747–1751.

Hardenbol, P., Bane´r, J., Jain, M., Nilsson, M., Namsaraev, E.A., Karlin-Neu- mann, G.A., Fakhrai-Rad, H., Ronaghi, M., Willis, T.D., Landegren, U., and Da- vis, R.W. (2003). Multiplexed genotyping with sequence-tagged molecular inversion probes. Nat. Biotechnol. 21, 673–678.

Herfst, S., Schrauwen, E.J., Linster, M., Chutinimitkul, S., de Wit, E., Munster, V.J., Sorrell, E.M., Bestebroer, T.M., Burke, D.F., Smith, D.J., et al. (2012).

Airborne transmission of influenza A/H5N1 virus between ferrets. Science 336, 1534–1541.

Herz, C., Stavnezer, E., Krug, R., and Gurney, T., Jr. (1981). Influenza virus, an RNA virus, synthesizes its messenger RNA in the nucleus of infected cells. Cell 26, 391–400.

Hoffmann, E., Neumann, G., Kawaoka, Y., Hobom, G., and Webster, R.G.

(2000). A DNA transfection system for generation of influenza A virus from eight plasmids. Proc. Natl. Acad. Sci. USA 97, 6108–6113.

Huang, S., Yu, C., Cheng, G., and Chen, Y. (2012). Detection of single influenza viral RNA in cells using a polymeric sequence probe. Anal. Chem. 84, 8118–

8121.

Hutchinson, E.C., von Kirchbach, J.C., Gog, J.R., and Digard, P. (2010).

Genome packaging in influenza A virus. J. Gen. Virol. 91, 313–328.

Imai, M., Watanabe, T., Hatta, M., Das, S.C., Ozawa, M., Shinya, K., Zhong, G., Hanson, A., Katsura, H., Watanabe, S., et al. (2012). Experimental adaptation of an influenza H5 HA confers respiratory droplet transmission to a reassortant H5 HA/H1N1 virus in ferrets. Nature 486, 420–428.

Jackson, D.A., Caton, A.J., McCready, S.J., and Cook, P.R. (1982). Influenza virus RNA is synthesized at fixed sites in the nucleus. Nature 296, 366–368.

Jarvius, J., Melin, J., Go¨ransson, J., Stenberg, J., Fredriksson, S., Gonzalez- Rey, C., Bertilsson, S., and Nilsson, M. (2006). Digital quantification using amplified single-molecule detection. Nat. Methods 3, 725–727.

Kamentsky, L., Jones, T.R., Fraser, A., Bray, M.A., Logan, D.J., Madden, K.L., Ljosa, V., Rueden, C., Eliceiri, K.W., and Carpenter, A.E. (2011). Improved structure, function and compatibility for CellProfiler: modular high-throughput image analysis software. Bioinformatics 27, 1179–1180.

Kawakami, E., Watanabe, T., Fujii, K., Goto, H., Watanabe, S., Noda, T., and Kawaoka, Y. (2011). Strand-specific real-time RT-PCR for distinguishing influ- enza vRNA, cRNA, and mRNA. J. Virol. Methods 173, 1–6.

Ke, R., Mignardi, M., Pacureanu, A., Svedlund, J., Botling, J., Wa¨hlby, C., and Nilsson, M. (2013). In situ sequencing for RNA analysis in preserved tissue and cells. Nat. Methods 10, 857–860.

Krammer, F., and Palese, P. (2015). Advances in the development of influenza virus vaccines. Nat. Rev. Drug Discov. 14, 167–182.

Lakadamyali, M., Rust, M.J., Babcock, H.P., and Zhuang, X. (2003). Visualizing infection of individual influenza viruses. Proc. Natl. Acad. Sci. USA 100, 9280–

9285.

Lakdawala, S.S., Wu, Y., Wawrzusin, P., Kabat, J., Broadbent, A.J., Lamir- ande, E.W., Fodor, E., Altan-Bonnet, N., Shroff, H., and Subbarao, K. (2014).

Influenza A virus assembly intermediates fuse in the cytoplasm. PLoS Pathog.

10, e1003971.

Larsson, C., Grundberg, I., So¨derberg, O., and Nilsson, M. (2010). In situ detection and genotyping of individual mRNA molecules. Nat. Methods 7, 395–397.

Law, G.L., Korth, M.J., Benecke, A.G., and Katze, M.G. (2013). Systems virology: host-directed approaches to viral pathogenesis and drug targeting.

Nat. Rev. Microbiol. 11, 455–466.

Lizardi, P.M., Huang, X., Zhu, Z., Bray-Ward, P., Thomas, D.C., and Ward, D.C.

(1998). Mutation detection and single-molecule counting using isothermal roll- ing-circle amplification. Nat. Genet. 19, 225–232.

Ma, K., Roy, A.M., and Whittaker, G.R. (2001). Nuclear export of influenza virus ribonucleoproteins: identification of an export intermediate at the nuclear pe- riphery. Virology 282, 215–220.

Marshall, N., Priyamvada, L., Ende, Z., Steel, J., and Lowen, A.C. (2013). Influ- enza virus reassortment occurs with high frequency in the absence of segment mismatch. PLoS Pathog. 9, e1003421.

Martin, K., and Helenius, A. (1991). Transport of incoming influenza virus nucle- ocapsids into the nucleus. J. Virol. 65, 232–244.

Matlin, K.S., Reggio, H., Helenius, A., and Simons, K. (1981). Infectious entry pathway of influenza virus in a canine kidney cell line. J. Cell Biol. 91, 601–613.

McGeoch, D., Fellner, P., and Newton, C. (1976). Influenza virus genome con- sists of eight distinct RNA species. Proc. Natl. Acad. Sci. USA 73, 3045–3049.

McGinn, S., Bauer, D., Brefort, T., Dong, L., El-Sagheer, A., Elsharawy, A., Evans, G., Falk-So¨rqvist, E., Forster, M., Fredriksson, S., et al. (2016). New technologies for DNA analysis—a review of the READNA Project.

N. Biotechnol. 33, 311–330.

Nilsson, M., Malmgren, H., Samiotaki, M., Kwiatkowski, M., Chowdhary, B.P., and Landegren, U. (1994). Padlock probes: circularizing oligonucleotides for localized DNA detection. Science 265, 2085–2088.

Nilsson, M., Krejci, K., Koch, J., Kwiatkowski, M., Gustavsson, P., and Land- egren, U. (1997). Padlock probes reveal single-nucleotide differences, parent of origin and in situ distribution of centromeric sequences in human chromo- somes 13 and 21. Nat. Genet. 16, 252–255.

Nordholm, J., da Silva, D.V., Damjanovic, J., Dou, D., and Daniels, R. (2013).

Polar residues and their positional context dictate the transmembrane domain interactions of influenza A neuraminidases. J. Biol. Chem. 288, 10652–10660.

Ochiai, H., Sakai, S., Hirabayashi, T., Shimizu, Y., and Terasawa, K. (1995).

Inhibitory effect of bafilomycin A1, a specific inhibitor of vacuolar-type proton pump, on the growth of influenza A and B viruses in MDCK cells. Antiviral Res.

27, 425–430.

Palese, P., and Schulman, J.L. (1976). Mapping of the influenza virus genome:

identification of the hemagglutinin and the neuraminidase genes. Proc. Natl.

Acad. Sci. USA 73, 2142–2146.

Reed, L.J., and Muench, H. (1938). A simple method of estimating fifty percent end points. Am. J. Epidemiol. 27, 493–497.

Resa-Infante, P., and Gabriel, G. (2013). The nuclear import machinery is a determinant of influenza virus host adaptation. BioEssays 35, 23–27.

Rust, M.J., Lakadamyali, M., Zhang, F., and Zhuang, X. (2004). Assembly of endocytic machinery around individual influenza viruses during viral entry.

Nat. Struct. Mol. Biol. 11, 567–573.

Schrauwen, E.J., and Fouchier, R.A. (2014). Host adaptation and transmission of influenza A viruses in mammals. Emerg. Microbes Infect. 3, e9.

Skehel, J.J., and Wiley, D.C. (2000). Receptor binding and membrane fusion in virus entry: the influenza hemagglutinin. Annu. Rev. Biochem. 69, 531–569.

Sobel Leonard, A., McClain, M.T., Smith, G.J., Wentworth, D.E., Halpin, R.A., Lin, X., Ransier, A., Stockwell, T.B., Das, S.R., Gilbert, A.S., et al. (2017). The

effective rate of influenza reassortment is limited during human infection.

PLoS Pathog. 13, e1006203.

Steel, J., and Lowen, A.C. (2014). Influenza A virus reassortment. Curr. Top.

Microbiol. Immunol. 385, 377–401.

Stewart-Jones, G.B., Soto, C., Lemmin, T., Chuang, G.Y., Druz, A., Kong, R., Thomas, P.V., Wagh, K., Zhou, T., Behrens, A.J., et al. (2016). Trimeric HIV-1- Env structures define glycan shields from clades A, B, and G. Cell 165, 813–826.

Te Velthuis, A.J., and Fodor, E. (2016). Influenza virus RNA polymerase: in- sights into the mechanisms of viral RNA synthesis. Nat. Rev. Microbiol. 14, 479–493.

Varble, A., Albrecht, R.A., Backes, S., Crumiller, M., Bouvier, N.M., Sachs, D., Garcı´a-Sastre, A., and tenOever, B.R. (2014). Influenza A virus transmission bottlenecks are defined by infection route and recipient host. Cell Host Microbe 16, 691–700.

Vignuzzi, M., Stone, J.K., Arnold, J.J., Cameron, C.E., and Andino, R. (2006).

Quasispecies diversity determines pathogenesis through cooperative interac- tions in a viral population. Nature 439, 344–348.

Westermann, A.J., Barquist, L., and Vogel, J. (2017). Resolving host-pathogen interactions by dual RNA-seq. PLoS Pathog.13, e1006033.

White, J., Helenius, A., and Gething, M.J. (1982). Haemagglutinin of influenza virus expressed from a cloned gene promotes membrane fusion. Nature 300, 658–659.

Xue, K.S., Hooper, K.A., Ollodart, A.R., Dingens, A.S., and Bloom, J.D. (2016).

Cooperation between distinct viral variants promotes growth of H3N2 influ- enza in cell culture. eLife 5, e13974.

York, A., and Fodor, E. (2013). Biogenesis, assembly, and export of viral messenger ribonucleoproteins in the influenza A virus infected cell. RNA Biol. 10, 1274–1282.

Yoshimura, A., and Ohnishi, S. (1984). Uncoating of influenza virus in endo- somes. J. Virol. 51, 497–504.

Zieba, A., Grannas, K., So¨derberg, O., Gullberg, M., Nilsson, M., and Landeg- ren, U. (2012). Molecular tools for companion diagnostics. N. Biotechnol. 29, 634–640.