Screening Method for CRISPR/Cas9 Inhibition of a Human DNA Virus: Herpes Simplex Virus

[Abstract] The efficiency of cleavage of individual CRISPR/Cas9-sgRNAs remains difficult to predict based on the CRISPR target sequence alone. Different intracellular environments (dependent on cell type or cell cycle state for example) may affect sgRNA efficiency by altering accessibility of genomic DNA through DNA modifications such as epigenetic marks and DNA-binding proteins ( e.g. , histones) as well as alteration of the chromatin state of genomic DNA within the nucleus. We recently reported a multi-step screening method for the identification of efficient sgRNAs targeting the Herpes simplex virus (HSV-1) genome and reported a differential mechanism for viral inhibition by CRISPR-Cas9 in the latent versus lytic phase. The screening platform detailed in this protocol allows step-by-step testing of the efficiency of cleavage in a cell-free system and in the context of viral target cells such as human foreskin fibroblasts followed by functional testing of the effects of CRISPR/sgRNA on viral protein expression, replication, and reactivation. This strategy could be readily applied to other target cells such as pluripotent stem cell-derived human sensory neurons or other human DNA viruses. GFP-positive viral yields were determined by plaque assays on FO6 and V27 cells. The histogram shows the mean values and standard deviations of biological (B and and All added conditions showed statistical significance compared to (Ratio


Viruses
The KOS strain is a commonly used wild-type HSV-1 strain (Schaffer et al., 1970). HSV-1 d109 is a KOS strain-derived mutant that is deleted for all five immediate early (IE) genes and contains the green fluorescent protein (GFP) gene under the control of the human cytomegalovirus (HCMV) IE promoter in the HSV-1 UL54 gene locus. d109 can be grown and titrated on U2OS ICP4/27 and FO6 cells (Samaniego et al., 1998).    end (SaCas9 scaffold oligo, Figure 1).

MiSeq adapter primers (Table 1)
ii. Anneal sgRNA primers with the tail primer: For each sgRNA assemble the following reaction in PCR tubes:

Total 10
Anneal protocol: iii. Fill in single-stranded DNA (ssDNA) overhangs to generate double-stranded DNA (dsDNA) templates for T7 transcription by assembling the following reaction for each sgRNA in PCR tubes ( Figure 1):

Total 20
Incubate at 72 °C for a total of 1.5 h and add an additional 1 μl Taq after 45 min.  i. Dilute Cas9 protein in assay buffer to 500 nM.
iii. Confirm each sgRNA is at 200 ng/μl or higher. iii. Incubate for 10 min at 37 °C to allow SaCas9-sgRNA complexes to form.
iv. Add 2 μl purified PCR product to each of the 4 wells.
vi. Add 4 μl of reaction stop buffer to each well and incubate at 80 °C for 10 min.
vii. Run all 10 μl of all reactions on a 2% agarose gel and image.
viii. Analysis: Because the assay provides sgRNA in excess, SaCas9 protein is the limiting factor. Cutting efficiency is constrained by the total amount of template provided, but in 10 www.bio-protocol.org/e3748 Bio-protocol 10 (17) general, sgRNAs that work well show robust cutting (larger band at 0-50% intensity of no SaCas9 control) and there will be evidence of graded activity at lower SaCas9 concentrations. Figure 2 shows the results of the in vitro cleavage assay for three sgRNA targeting the essential HSV UL30 gene (encoding the viral polymerase). ii. The next day, replace the transduction medium with fresh medium and incubate cells for 2 days at 37 °C followed by puromycin treatment (1 µg/ml) for 7-10 days. iii. The next day, replace the transduction medium with fresh medium and incubate cells for 2 days at 37 °C followed by puromycin treatment (1 µg/ml) for 7-10 days.  The barcode PCR primers anneal to the adapter sequences and contain a barcode (blue) as well as P5/P7 sequences for binding to the flow cell. The three sequencing reads cover the amplicon/target sequence and read the barcode.
ii. Perform a Phusion touch-down PCR, which increases the specificity of the reaction, according to the following protocol. Use the NEB calculator to determine the annealing temperature (http://tmcalculator.neb.com/#!/). ii. Run a small amount (2 μl) on a 1.5% agarose gel. This allows you to ensure that the PCR worked and to normalize the amount of each amplicon when PCR products are pooled together for MiSeq. Of note, in some cases, one may see a large primer dimer amplicon around 175 bp in addition to the correct band.

Reagent
iii. Pool amplicons and perform gel extraction: 1) Pool (up to 24) amplicons for different sgRNAs-do not pool across amplicons from different targeted viral genes as this will make the gel extraction more difficult.
From the previous gel, estimate equal molarities when pooling (i.e., compensate for weaker looking amplicons by adding more to the pool). This will even out the read counts across different amplicons during MiSeq.
2) Run pooled amplicons on a 1.5% agarose gel and gel extract the DNA using the Promega Gel Extraction Kit.
iv. Submit purified/pooled amplicons for IlluminaMiSeq using the 16S Metagenomic 6. Indel analysis: Determine indel frequencies for each sgRNA by quantifying aligned reads containing insertions or deletions 1bp or larger using (http://www.outknocker.org/). Analyze indel size using the ICE analysis toolbox (https://www.synthego.com/products/bioinformatics/crispranalysis). Figure 4 shows the analysis of indel mutations in the HSV-1 genome during the quiescent infection for sgRNAs targeting the essential HSV-1 genes UL29 and UL30. i. Detect near-infrared fluorescence using Odyssey (LI-COR) and quantify protein expression levels using Image J or ImageStudio V4 (LI-COR) (see Figure 5D).

Inhibition of lytic replication
a. Transduction of HFFs with SaCas9/sgRNA: i. Prepare HFF cells at a density of 2 x 10 5 /well in a T25 flask one day prior to transduction ii. Transduce with 2-5 ml of lentivirus (or at an MOI of 5) and 3-4 µg/ml of polybrene.
iii. The next day, replace the transduction medium with fresh medium and incubate cells for additional 2 days at 37 °C followed by puromycin treatment (1 µg/ml) for 7-10 days. iv. Remove media and fix cells with 100% cold methanol for 10 min.
v. Stain monolayer with GIEMSA stain and count plaques (see Figures 5B and 5C) 3. Inhibition of reactivation of quiescent virus (see Figure 6A) ii. The next day, replace the transduction medium with fresh medium and incubate cells for 2 days at 37 °C followed by puromycin treatment (1 µg/ml) for 7-10 days.
iii. To reactivate quiescent d109 virus, superinfect HFF cells with wild-type HSV-1 at an MOI of 5 in PBS containing 0.1% glucose (wt/vol) and 0.1% BCS (v/v) for 1 h with shaking at 37 °C. iv. Replace virus inoculum to DMEM containing 1% BCS for 24 h. viii. Reactivation analysis: For the reactivation assay, it is important to include a control for recombination between WT HSV-1 and d109 which could transfer the GFP sequence to WT HSV-1 and result in production of false GFP-positive plaques. We quantify d109 viral reactivation using plaque assays by counting GFP-positive plaques on complementing FO6 cells. FO6 is a Vero-derived cell line expressing ICP4, ICP27, and ICP0 upon HSV-1 infection, thereby complementing replication of d109 virus (Samaniego et al., 1998). To measure these recombinant viruses, we count GFPpositive plaques formed on V27 cells, which express ICP27 (but not ICP4 and ICP0) upon infection with HSV. Because ICP27 is replaced with GFP in HSV-1 d109, any recombinant GFP-positive but ICP27-negative HSV mutants that arise (but not HSV-1 d109), can replicate in V27 cells. To calculate the number of plaques originating from reactivated d109 genomes, we subtract the number of GFP-positive plaques on V27 cells from the number of GFP-positive plaques on FO6 cells (see Figures 6B and 6C). 20 www.bio-protocol.org/e3748 Bio-protocol 10 (17)

Data analysis
Controls: We include a control without sgRNA (SaCas9 only) for all experiments.

Statistical analysis:
We use a minimum of three biological replicates (independent experiments that are performed using the same test at different times) to determine statistical significance. We use Prism 6 (Version 6.01) software (GraphPad Software) for statistical analysis. To determine statistical significance, we use Student's t-test, ratio paired t-test, or one-way ANOVA with Dunnett's multiple comparison test with two-sided (95% confidence level) (Oh et al., 2019).