A Genome-Wide CRISPR/Cas9 Screen Reveals the Requirement of Host Sphingomyelin Synthase 1 for Infection with Pseudorabies Virus Mutant gD-Pass

Abstract

Herpesviruses are large DNA viruses, which encode up to 300 different proteins including enzymes enabling efficient replication. Nevertheless, they depend on a multitude of host cell proteins for successful propagation. To uncover cellular host factors important for replication of pseudorabies virus (PrV), an alphaherpesvirus of swine, we performed an unbiased genome-wide CRISPR/Cas9 forward screen. To this end, a porcine CRISPR-knockout sgRNA library (SsCRISPRko.v1) targeting 20,598 genes was generated and used to transduce porcine kidney cells. Cells were then infected with either wildtype PrV (PrV-Ka) or a PrV mutant (PrV-gD^-Pass) lacking the receptor-binding protein gD, which regained infectivity after serial passaging in cell culture. While no cells survived infection with PrV-Ka, resistant cell colonies were observed after infection with PrV-gD^-Pass. In these cells, sphingomyelin synthase 1 (SMS1) was identified as the top hit candidate. Infection efficiency was reduced by up to 90% for PrV-gD^-Pass in rabbit RK13-sgms1_KO cells compared to wildtype cells accompanied by lower viral progeny titers. Exogenous expression of SMS1 partly reverted the entry defect of PrV-gD^-Pass. In contrast, infectivity of PrV-Ka was reduced by 50% on the knockout cells, which could not be restored by exogenous expression of SMS1. These data suggest that SMS1 plays a pivotal role for PrV infection, when the gD-mediated entry pathway is blocked.

Keywords: CRISPR/Cas9 gene editing; PrV; SMS1; gD–Pass; herpesvirus; pseudorabies virus; sgms1; sphingomyelin synthase.

Figure 1

Identification of sphingomyelin synthase 1 (SMS1, sgms1) as an entry factor for PrV-gD^–Pass. (A) Schematic overview of the genome-wide CRISPR/Cas9 screening protocol. PK15 cells were transduced with lentiviruses encoding the porcine CRISPR knockout library (SsCRISPRko.v1) at a multiplicity of transduction (MOT) of 0.3. Two days post transduction, cells were selected using puromycin for 10 to 14 days and subsequently infected with PrV at a multiplicity of infection (MOI) of 0.5. Surviving cells were split in half and were either reinfected or harvested to isolate genomic DNA for subsequent analysis by sequencing. (B) Genes enriched in the surviving cell pool after infection with PrV-gD^–Pass were ranked by the robust rank aggregation (RRA) scores, displaying a very strong selection of sgRNAs targeting the sgms1 gene in all three independent assays.

Figure 2

Characterization of RK13-sgms1_KO cells. (A) Shown is the targeted sequence of exon 3 of the rabbit sgms1 gene region. The sequence derived from RK13-sgms1_KO (lower line) was compared to the parental rabbit sequence (ENSOCUG00000010965.4, upper line). Due to the large deletion (Δ) identical regions between wildtype and knockout are indicated by black bars and the deleted sequence is represented by a thin line. A detailed analysis on base pair level is shown in (B). Locations of sgRNAs are marked in blue (A,B), while the corresponding PAM sequence is marked in green. Deletions are shown by hyphen. (C) Lysates of wildtype RK13 and RK13-sgms1_KO, as well as RK13-sgms1_KO cells transduced either with an empty vector (RK13-sgms1_KO-X) or with a vector expressing the rabbit SMS1 (RK13-sgms1_KO/sgms1⁺) were probed with a SMS1-specific antiserum. The specific ca. 38 kDa band is marked by an arrow. The α-tubulin monoclonal antibody was used as loading control. Molecular masses of marker proteins (in kDa) are indicated on the left.

Figure 3

Characterization of RK13-sgms1_KO cells. SMS1 knockout cells were compared to parental RK13 cells. Cell viability and propagation was measured at 24, 48 and 72 h post seeding using the PrestoBlue^TM reagent (A). Shown is the mean of three independent experiments. Differences in the lipid metabolism were detected by Oil Red O staining (red) showing an accumulation of lipids in RK13-sgms1_KO cells compared to parental RK13 cells (B). Nuclei were counterstained with Hematoxylin and scale bars indicate 50 µm. (C) Cytoplasmic membranes were stained with CellBrite^TM for 24 h. Nuclei were counterstained with DAPI, and fluorescence was detected with a confocal laser-scanning microscope with constant settings. Scale bars indicate 10 µm. (D) Fluorescence was quantified by calculation of the corrected total cell fluorescence. Significant differences were calculated by two-way ANOVA followed by Sidak’s multiple comparison test (A, ns: not significant) or by an unpaired t-test (D, **** ≤0.0001).

Figure 4

Effect of sgms1 gene knockout on PrV replication in RK13 cells. (A) RK13 and RK13-sgms1_KO cell lines were infected with PrV-1112 or PrV-gD^–Pass at an MOI of 5 and harvested at different times p.i. Progeny virus titers were determined on RK13 cells. Given are mean values as log₁₀ plaque forming units (pfu) per ml of three independent experiments with corresponding standard deviations. Asterisks indicate statistically significant differences calculated by two-way ANOVA followed by Sidak’s multiple comparison test compared to the parental RK13 in the same symbol as the corresponding graph (* ≤0.05, ** ≤0.01, **** ≤0.0001). (B) For determination of the efficiency of plating (EOP) cells were infected with approx. 200 pfu per 12-well. The EOP was determined by counting plaques two days post infection and the numbers were normalized (in %) to parental RK13. Given are mean values of six independent experiments with corresponding standard deviations.