Detection and Differentiation of Multiple Viral RNAs Using Branched DNA FISH Coupled to Confocal Microscopy and Flow Cytometry

Due to the exceptionally high mutation rates of RNA-dependent RNA polymerases, infectious RNA viruses generate extensive sequence diversity, leading to some of the lowest barriers to the development of antiviral drug resistance in the microbial world. We have previously discovered that higher barriers to the development of drug resistance can be achieved through dominant suppression of drug-resistant viruses by their drug-susceptible parents. We have explored the existence of dominant drug targets in poliovirus, dengue virus and hepatitis C virus (HCV). The low replication capacity of HCV required the development of novel strategies for identifying cells co-infected with drug-susceptible and drug-resistant strains. To monitor co-infected cell populations, we generated codon-altered versions of the JFH1 strain of HCV. Then, we could differentiate the codon-altered and wild-type strains using a novel type of RNA fluorescent in situ hybridization (FISH) coupled with flow cytometry or confocal microscopy. Both of these techniques can be used in conjunction with standard antibody-protein detection methods. Here, we describe a detailed protocol for both RNA FISH flow cytometry and confocal microscopy.

viruses that are present in the same cell as newly synthesized drug-resistant variants. Drugresistant viral RNA must first be amplified and translated in its cell of origin, making newly synthesized drug-resistant viruses susceptible to dominant suppression by their drugsusceptible parents and cousins. We have coined the term "dominant drug targets" to describe viral targets with higher barriers to the development of antiviral drug resistance due to genetic dominance of drug-susceptible viruses. Study of genetic interactions and physical location of distinct viral genomes in the same cell required the development of the new technology described here.
To identify dominant drug targets for which tool antiviral compounds are available, we first generated drug-resistant viruses and built the mutations into an infectious cDNA clone. To test whether drug-resistant or drug-susceptible viruses were genetically dominant, we generated cells co-infected with drug-susceptible and drug-resistant viruses and then monitored selection from within them. In studies using poliovirus (Crowder and Kirkegaard, 2005;Tanner et al., 2014) and Dengue virus (Mateo et al., 2015) we were able to generate sufficiently high-titer virus stocks to perform coinfections at high multiplicities of infection and thus ensure that all cells in our cultures were coinfected. Recently, we expanded tests for dominance to hepatitis C virus (HCV) (van Buuren et al., 2018), for which high-titer stocks are often difficult to obtain, especially for drug-resistant variants that have reduced fitness. Therefore, when we co-infected Huh7.5.1 cells with two strains of HCV at multiplicities of infection of less than 1 PFU/cell, we generated four cell populations: co-infected cells, two types of singly infected cells and a significant population of uninfected cells. We needed to differentiate co-infected cells from the two types of singly infected cells and learn about genetic selection while doing so. To accomplish this, we were early adopters of the branched DNA (bDNA) technology originally developed by Affymetrix (now Thermo Fisher Scientific). This technology uses tiered DNA oligos to build a network of up to 8,000 fluorophores on each target RNA. This unique type of RNA fluorescent in situ hybridization (FISH) can be coupled with protein detection using standard antibody conjugation and detected using confocal microscopy (ViewRNA® Cell Plus Assay) and flow cytometry (PrimeFlow™ RNA Assay).
These bDNA FISH techniques first generate a series of target probes that bind the RNA of interest at adjacent sequences, but leave 3' extensions of unique sequence to bind the preamplifier DNA that is complementary to two different probes. Cooperative binding of the pre-amplifier DNA to two probes increases the signal-to-noise ratio because any individual mistargeted probe cannot be amplified. Typically, twenty pairs of target probes are designed to bind the RNA of interest; this requires roughly 1,000 nucleotides of sequence space. Each of the twenty pre-amplifier DNAs is then bound by a series of amplifier DNAs, and then subsequently by a series of oligonucleotide-conjugated fluorophores. This process leads to the labeling of each individual target RNA with up to 8,000 fluorophores, sufficient to visualize individual RNAs by confocal microscopy. To apply this technology to dominant drug targeting in HCV, we needed to generate a strain of HCV with sufficient dissimilarity in its RNA sequence that we could differentiate it from wild-type viral RNA. To accomplish this, we generated three codon-altered versions of the JFH1 strain of HCV. Codon optimization algorithms available through GeneArt (Thermo Fisher Scientific) were used to design three approximately 1,000-nucleotide regions of the JFH1 genome that had altered codon usage but retained the same protein sequence. These codon-altered JFH1 strains all contained 200-250 synonymous mutations over the 1,000nucleotide regions. Of these three strains, two demonstrated decreased fitness, likely due to disruption of RNA secondary structures required for viral replication (Pirakitikulr et al., 2016). The third strain, however, displayed growth kinetics that mimicked wild-type virus and could be used in co-infection experiments and differentiated from wild-type JFH1 using both RNA FISH and flow cytometry.

1.
Pipettes (with or without filter tips)

Procedure
A. Construction of codon altered sequences

1.
Roughly 1,000 nucleotides of RNA sequence are required to support the hybridization of twenty bDNA trees and 8,000 fluorophores. Targets that contain less than the full complement of bDNAs can still be detected by flow cytometry but require higher copy numbers to achieve the same resolution.

2.
For viral RNAs, when possible, scan the literature for any structural information available to determine which areas of the genome are the least likely to contain essential RNA secondary structures. If possible, also choose a region that has convenient cut sites for insertion of your codon-altered sequence. We chose to clone three codon-altered regions of the JFH1 genome because we anticipated decreased viability from some of the codon-altered strains.

3.
GeneArt is a product offered through Thermo Fisher Scientific and can be used to synthesize genes up to 9,000 bp in length (https:// www.thermofisher.com/us/en/home/life-science/cloning/gene-synthesis/geneartgene-synthesis.html). The GeneArt homepage offers several tools, including the gene optimizer tool. Use the gene optimizer algorithms to design codon-altered sequences with wild-type viral RNA sequence as your template. We submitted three regions of JFH1 that were all roughly 1,000 nucleotides in length and flanked by convenient cut sites. The optimizer tool was able to alter nearly 25% of nucleotides in all three cases.

4.
The newly synthesized sequence will arrive incorporated into a plasmid with a defined antibiotic-resistance marker. At this time, your codon-altered gene fragment can be subcloned into a plasmid that encodes the viral genome using restriction digestion and ligation.

5.
Target Probes that differentiated viral RNA sequences were designed and manufactured by Affymetrix (now Thermo Fisher Scientific) for use with both the ViewRNA and PrimeFlow platforms ( Figure 1).

1.
The pJFH1 plasmid encodes the full-length genome of the JFH1 strain of HCV. The wild-type plasmid and all codon-altered versions contain an Xbal cut site at the 3' end of the genome. Digest 5 μg of plasmid DNA with 20 U of Xbal in the CutSmart Buffer provided in a final reaction volume of 25 μl. Incubate digestions at 37 °C for 2 h.

2.
Purify linearized DNA using the QIAquick PCR Purification Kit, as per manufacturer's protocol.

3.
Using 1 μg of linearized plasmid as your template, perform in vitro transcription with the MEGAscriptT7 kit to make full-length genomic viral RNA. Incubate in vitro transcription reaction at 37 °C for 6 h. The temperature and duration of this incubation can be altered for optimal yield of individual transcripts.

4.
Isolate synthesized viral RNA using Trizol as per the manufacturer's protocol. Resuspend vRNA pellet in 50 μl of RNase-free water.

b.
Add 2 ml of Trypsin and incubate at 37 °C for 5 min.

c.
Add 5 ml of 10% FBS media and harvest cell suspension into a 15 ml conical tube.

d.
Centrifuge cells at 400 × g for 4 min.

e.
Resuspend cell pellet in 5 ml of PBS.

f.
Centrifuge cells at 400 × g for 4 min.

g.
Resuspend cell pellet with 5 ml of CytoMix. Cytomix recipe can be found below under "Recipes".

h.
Centrifuge cells at 400 × g for 4 min.

i.
Resuspend cell pellet in 400 μl of CytoMix and transfer to a 4 mm GenePulser cuvette.

j.
Mix 10 μg of viral RNA into cell suspension inside cuvette and gently pipet up and down to mix.

l.
Allow cells to rest at room temperature for 10 min.

m.
Transfer electroporated cells to a fresh 10 cm culture dish with 10 ml of 10% FBS media (see Recipes).

7.
Culture electroporated cells for up to two weeks in 10% FBS media, passaging every 3-4 days as required. As you passage, expand the culture. Typically, cultures of 10 7 electroporated cells are expanded into either five or ten T150 flasks. This gives HCV time to spread and generates a culture with a higher percentage of cells that are infected and productively synthesizing progeny virus. Further expansion of cells to larger capacity can be done if needed.

8.
Convert JFH1 cultures to Human Serum Media (Steenbergen et al., 2013). Growth of HCV in human serum has two benefits. First, Huh7.5.1 cells differentiate and cease cell division, therefore trypsinization and biweekly passage are no longer required. Instead, virus containing cell supernatants can simply be collected biweekly and directly replaced with fresh medium. Second, growth in Human Serum Media increases viral yield by 10 to 100-fold.

C. Simultaneous infection with two HCV strains and detection of co-infected cells with
Prime-Flow.

1.
The description of this protocol has been adapted from the PrimeFlow Assay user's manual.

2.
Huh7.5.1 cells are seeded into 12-well plates at a density of 10 5 cells per well using 1 ml of 10% FBS media.

3.
In our hands, JFH1 cultured in human serum media can produce viral titers of 10 5 -10 6 focus forming units (FFU) per ml. Infect Huh7.5.1 cells at a multiplicity of infection of one virus particle per cell with both wild-type and codon-altered JFH1. This often equates to roughly 1-2 ml of each virus preparation. A total volume of 4 ml can be used carefully in 12-well plates.

4.
Incubate infected cells in a CO 2 incubator at 37 °C for 4-6 h. Following initial incubation, remove virus-containing media by aspiration. Replace media with fresh 10% FBS media and incubate infected cells for 72 h.

5.
Replace 10% FBS media with fresh 10% FBS media that either contains antiviral drugs or vehicle and incubate infected cells for 24-36 h.

6.
Aspirate off media containing antivirals or vehicle and wash cells with 1 ml PBS.

7.
Harvest infected cells by treating cells with 0.5 ml trypsin and incubating at 37 °C with CO 2 for 5 min.

8.
Inhibit trypsin by adding 1 ml of 10% FBS media to each well. Harvest all cells and transfer to one of the 1.5 ml microfuge tubes supplied in the PrimeFlow Assay kit.

10.
Aspirate off media and trypsin, being careful not to lose any cells. This is achieved by only aspirating down to the 100 μl marker on the side of the Eppendorf tube. Wash cells with 1 ml of Flow Cytometry Staining Buffer. Vortex and spin at 400 × g for 5 min.

11.
Aspirate Flow Cytometry Staining Buffer and fix cells using 1 ml of Fixation Buffer 1 at 4 °C for 30 min.

13.
Resuspend cells in 1 ml of Permeabilization Buffer. Spin cells at 800 × g for 5 min. Repeat wash with Permeabilization Buffer 3 ×.

14.
Aspirate final Permeabilization Buffer wash and resuspend cells in 1 ml of Fixation Buffer 2. Incubate cells in the dark at room temperature for 60 min.

15.
Spin cells at 800 × g for 5 min and resuspend in 1 ml of Wash Buffer.

18.
Resuspend cells in 100 μl of the Target Probe mixture. Incubate at 40 ± 1 °C for 2 h. We use a heat block in our 40 °C incubator to increase heat conduction to the tubes and protect from large fluctuation in heat ( Figure 2). This incubation can be extended from 2 h to overnight. Longer incubations periods allowed for all amplification steps, flow cytometry and data analysis to be completed the following day.

19.
Wash cells by adding 1 ml of Wash Buffer, vortex, and spin at 800 × g for 5 min.

22.
Wash cells by adding 1 ml of Wash Buffer, vortex, and spin at 800 × g for 5 min.

25.
Wash cells by adding 1 ml of Wash Buffer, vortex, and spin at 800 × g for 5 min.

27.
Prepare Label Probe mix by diluting Label Probes into the Label Probe Diluent at 1:100.

29.
Wash cells by adding 1 ml of Wash Buffer, vortex, and spin at 800 × g for 5 min.

31.
Aspirate Wash Buffer leaving 100 μl of residual liquid to resuspend stained cells.
Resuspend cells by pipetting up and down and transfer to a labeled BD FACS tube containing 250 μl of PBS.

32.
Analyze cells using a flow cytometer and FlowJo software (details below).
D. Quantitation of RNA-protein colocalization using confocal microscopy.

1.
Huh7.5.1 cells are plated on Micro Cover Glass inside 12-well tissue culture plates at a density of 10 5 cells per well one day prior to infection ( Figure 3).

2.
Co-infect cells with wild type JFH1 and codon altered JFH1 at a multiplicity of infection equal to one virus per cell.

3.
At 6 h post infection, aspirate inoculum and replace with 1 ml of 10% FBS media.

4.
At 24 h post infection, wash cells 2 × with 1 ml of PBS.

5.
Add 400 μl of Fixation/Permeabilization Buffer to each well and incubate for 30 min at room temperature.

7.
Overlay cells with 400 μl of Blocking/Antibody Diluent and incubate at room temperature for 20 min.
Overlay cells with antibody mixture and incubate at room temperature for 1 h.

9.
Wash cells three times with PBS.

10.
Dilute secondary antibody in 400 μl Blocking/Antibody Diluent as required. We use anti-mouse AlexaFluor-647 diluted at 1:200 for our experiment with HCV.
Overlay cells with antibody mixture and incubate at room temperature for 1 h.

12.
Add 400 μl of Fixation Solution to each well and incubate in the dark at room temperature for 1 h.

15.
After the final wash, overlay cells with 400 μl of Target Probe mixture. Incubate at 40 ± 1 °C for 2 h.

18.
After the final wash, overlay cells with 400 μl of Pre-Amplifier mixture and incubate at 40 ± 1 °C for 1 h.

21.
After the final wash, overlay cells with 400 μl of Amplifier mixture and incubate at 40 ± 1 °C for 1 h.

24.
After the final wash, overlay cells with 400 μl of Label Probe mixture and incubate at 40 ± 1 °C for 1 h.

28.
Using forceps, carefully remove stained Micro Cover Glass from the 12-well dish, dab on a Kimwipe to remove excess Wash Buffer and place "cells down" onto the drop of Permafluor/DAPI. Allow to harden for at least 4 h.

29.
Visualize cells using a confocal microscope. We use a Leica SP8 Confocal Microscope fitted with a White Light Laser. Green channel. The result is plotted as the total number of puncta that share mutual space between channels versus the total number of puncta in each channel.

5.
Determining colocalization between RNA and protein requires a separate algorithm as the proteins often not localize into discrete countable puncta. We therefore ask Volocity to determine how many of the viral RNA puncta "Touch" anywhere within the protein signal. We then graph the number of RNA puncta that touched protein versus the total number of RNA puncta in each cell.  Huh7.5.1 cells are plated onto Micro Cover Glass within a 12-well tissue culture plate. These cells are infected following a 24 h incubation to allow cell adherence to the glass. Following infection, the cells are fixed, stained for protein and RNA using the ViewRNA Cell Plus Assay kit, all within the 12-well plate. The Micro Cover Glass is then carefully transferred to a Microslide spotted with PermaFluor/DAPI using fine forceps.  Huh7.5.1 cells were co-infected with JFH1-WT and JFH1-CA on Micro Cover Glass for 72 h. Cells were stained for HCV core protein and both viral RNAs using the ViewRNA Cell Plus Assay. Quantification of colocalization was performed using Volocity software. Scale bars are 2.5 μm in length.