Simple and Rapid High-Throughput Assay to Identify HSV-1 ICP0 Transactivation Inhibitors

Abstract

Herpes simplex virus 1 (HSV-1) is a ubiquitous virus that results in lifelong infections due to it’s ability to cycle between lytic replication and latency. As an obligate intracellular pathogen, HSV-1 exploits host cellular factors to replicate and aid in its life cycle. HSV-1 expresses infected cell protein 0 (ICP0), an immediate-early regulator, to stimulate the transcription of all classes of viral genes via its E3 ubiquitin ligase activity. Here we report an automated, inexpensive, and rapid high-throughput approach to examine the effects of small molecule compounds on ICP0 transactivator function in cells. Two HSV-1 reporter viruses, KOS6β (wt) and dlx3.1-6β (ICP0-null mutant), were used to monitor ICP0 transactivation activity through the HSV-1 ICP6 promoter::lacz expression cassette. A ≥10-fold difference in β-galactosidase activity was observed in cells infected with KOS6β compared to dlx3.1-6β, demonstrating that ICP0 potently transactivates the ICP6 promoter. We established the robustness and reproducibility with a Z′-factor score of ≥0.69, an important criterium for high-throughput analyses. Approximately 19,000 structurally diverse compounds were screened and 76 potential inhibitors of the HSV-1 transactivator ICP0 were identified. We expect this assay will aid in the discovery of novel inhibitors and tools against HSV-1 ICP0. Using well-annotated compounds could identify potential novel factors and pathways that interact with ICP0 to promote HSV-1 gene expression.

1. Introduction

Herpes simplex virus 1 (HSV-1) infects ~80% of the world’s population. HSV-1 is the major cause of recurrent oral-facial sores and can give rise to severe diseases such as herpes stromal keratitis and encephalitis (Roizman et al., 2007). During lytic infection, the virus productively replicates in the epithelial and fibroblast cells at the periphery. Latency occurs in the innervating sensory neurons, characterized by no infectious virus but presence of the HSV-1 genome. Various stressors trigger the latent virus to lytically reactivate leading to recurrent symptoms and persisting as a lifelong infection (Bloom, 2016). Since HSV-1 is an obligate intracellular pathogen, cellular factors play an important role in replication and reactivation (Grinde, 2013).

Current treatments remain limited to targeting HSV-1 lytic infection and viral DNA replication. First line therapeutics include acyclic guanosine analogues such as acyclovir, which upon phosphorylation by HSV thymidine kinases selectively inhibits viral DNA polymerase (Vadlapudi et al., 2013; Wilson et al., 2009). The lifelong use of these drugs has led to viral resistance (Piret and Guy, 2011). Second line therapeutics, cidofovir and foscarnet, are limited in their use due to nephrotoxicity (Wilson et al., 2009). Therefore, it is essential to identify inhibitors of novel targets that block HSV-1 lytic infection and reactivation.

In this study we focused on infected cell protein 0 (ICP0), an immediate-early viral protein of HSV-1. ICP0 transactivates all three classes of HSV-1 genes, in part, through the destabilization and/or inhibition of host factors. ICP0 utilizes its RING-finger domain for E3 ubiquitin ligase activity targeting specific cellular proteins by conjugating them with ubiquitin, a post-translational modification (Everett, 2000; Boutell et al., 2002). Ubiquitin-mediated degradation of cellular proteins by ICP0 leads to the disruption of nuclear domain 10 (ND10), associated with cellular proliferation and differentiation, senescence, apoptosis, and virus restriction (Cai et al., 1993; Ching et al., 2005; Zhong S et al., 2000). Two ND10 constituent proteins, promyelocytic leukemia protein (PML) and Sp100, are degraded by ICP0, inactivating the antiviral properties of ND10s (Everett et al., 1998; Muller and Dejean, 1999, Lanfranca et al., 2014).

Genetic studies have shown that ICP0-null mutants are reduced for viral replication compared to wild type HSV-1 strains, demonstrating that ICP0 promotes efficient viral replication in cell culture and animal models of HSV-1 infection (Sacks and Schaffer, 1987; Leib et al., 1989, Halford and Schaffer, 2000; Everett, 1989; Everett et al., 2009; Stow and Stow, 1986). Animal studies have demonstrated that ICP0 enhances the establishment of viral latency and significantly stimulates viral reactivation (Halford and Schaffer, 2001; Halford et al., 2006; Cai et al., 1993). Given this pivotal role of ICP0 in the HSV-1 life cycle, mechanisms by which ICP0 functions and the cellular pathways have not been fully explored (Smith et al., 2011; Hagglund and Roizman, 2004; Boutell and Everett, 2013). Genetic and cell-based assays have led to the discovery of ICP0-host interactions, but chemical biological approaches to examine these interactions have been limited.

We developed a novel approach to examine and identify potential inhibitors of HSV-1’s ICP0 transactivator function. Our approach utilizes two reporter viruses, KOS6β (Davido et al., 2002) and dlx3.1-6β (Davido et al., 2003). KOS6β (wt) and dlx3.1-6β (an ICP0-null mutant) have an ICP6 promoter::lacz cassette inserted between UL49 and UL50 genes. Notably, ICP0 is observed to be a potent and specific inducer of the early ICP6 gene, which encodes the large subunit of ribonucleotide reductase (Davido and Leib, 1996; Davido et al., 2002; Sze and Herman, 1992; Goldstein and Weller, 1998). ICP0 transactivation activity can be monitored using a simple colorimetric-based β-galactosidase activity assay. This assay provides an inexpensive and automated high-throughput screening method. We conducted an initial screen with roscovitine, a broad inhibitor of cyclin-dependent kinases (cdks) and HSV-1 transcription, to validate the sensitivity, robustness, and reproducibility of our assay.

This assay was used in a pilot study that screened ~19,000 compounds, and we identified 76 hits as potential ICP0 transactivator inhibitors, which included trichothecenes, lipopeptides, and cdk inhibitors. Some of the compounds have been previously shown to impair HSV-1, confirming the utility of our screen. Implications of our system are discussed.

2. Materials and Methods

2.1. Cell culture, viruses, and compounds

HepaRG cell line is derived from a liver tumor patient (Gripon et al., 2002). HepaRG cells (a gift from Roger Everett) were grown in William's E Medium containing 10% fetal bovine serum (FBS), 2mM L-glutamine, 10 U/mL penicillin, 10 U/mL Streptomycin, 50 μg/mL Insulin, and 50 μM Hydrocortisone. HepaRG cells were maintained by incubation at 37°C in 5% CO₂. Reporter viruses, HSV-1 KOS6β (wt) and dlx3.1-6β (ICP0-null mutant), were used in our assays. For viral yield assays, 100,000 HepaRG cells were infected at an MOI of 5 with KOS6β and dlx3.1-6β. After 24 hpi cells were harvested and viral titers were determined by plaque assays (Davido et al., 2002; Davido et al., 2003). Roscovitine, was prepared in dimethyl sulfoxide (DMSO) at a stock concentration of 50 mM (Schang et al., 1998). The final concentrations of roscovitine tested were 50 μM and 100 μM.

2.2. Optimization of high-throughput assay

To optimize our assay, we examined the variables of fetal bovine serum (FBS) percentage, multiplicity of infection (MOIs), infection period, β-galactosidase assay kinetics and stability. HepaRG cells were seeded in 384-wells-plates, 25 μL of 6,750 cells per well, in phenol red-free William’s E Medium containing either 1% or 2% FBS, and incubated for 24 hours at 37°C in 5% CO₂. Then, 10 μl of KOS-6β or dlx3.1-6β were added to wells at MOIs equivalent to 0, 0.2, 1, and 5. Infections proceeded for 6, 12, or 24 hours. At each time point, 10 μl of 1X lysis buffer (1% Triton X-100; 20 mM Tris-HCl [pH 8.0]; 150 mM NaCl; 1 mM dithiothreitol) was added to each well and incubated at 37°C for 20 minutes. β-Gal Assay buffer/CPRG solution was made with Chlorophenol Red-β-D-galactopyranoside (CPRG) (Calbiochem) and β-Gal Assay buffer (2.475 mL 1M KCl; 19.8 mL 1M Phosphate buffer (pH 7.3); 225 μL 1M MgCl₂; 1.984 mL of 14.4 M BME; H2O to 50 mL). The β-Gal Assay buffer/CPRG solution was added to each well (10 μl/well) with a final concentration of 0.2 mg/mL. Absorbance was measured at 595 nm at 5, 35, 80, 120, 1080 minutes post-addition of the β-Gal/CPRG solution with a PerkinElmer EnVision reader.

2.3. Primary screen with roscovitine

A primary screen was conducted using roscovitine with optimized conditions (Fig. 3). HepaRG cells were prepared as described in section 2.2. Roscovitine was then transferred into each well echo 555 acoustically (Labcyte Inc.) and preincubated for 40 minutes at 37°C in 5% CO₂. KOS-6β and dlx3.1-6β were dispensed at MOI 5 and processed as described in section 2.2. For studying compound effects on cell viability, HepaRG cells were treated with roscovitine for 12 hours. An ATP-based cell viability assay was performed using Promega Cell-Titer Glo reagent according to manufacturer’s instructions. GraphPad Prism 8 was used to determine IC₅₀ and CC₅₀.

Figure 3. Optimization results.

Histograms show optimization results of: (A) KOS6β infection and (B) dlx3.1-6β infection. For each reporter virus the data compares: 1% or 2% FBS, MOIs: 0, 0.2, 1, and 5, and β-galactosidase levels at 6 hpi, 12 hpi, and 24 hpi using CPRG, and assay stability from 5 to 1080 minutes.

2.4. Screen with KU-HTSL libraries and cell cytotoxicity

HepaRG cells were seeded as previously described in 2.2. Each compound from the KU-HTSL was transferred by echo 555 acoustically into each well for a final concentration of 10 μM. The libraries included Selleck Bioactives, Natural Products (GreenPharma), CMLD Diversity, Analyticon Natural Products, Life Natural Products. Each compound was preincubated for 40 minutes at 37°C in 5% CO₂. KOS-6β and dlx3.1-6β, were dispensed at an MOI 5 per well, and processed as described in section 2.2. To eliminate potential false positive hits based on cytotoxicity, Promega Cell-Titer Glo assay was performed, and each compound was incubated at 10 μM for 12 hours before analysis.

2.5. Cycloheximide block and release

HepaRG cells were seeded as previously described in 2.2. Cycloheximide (CHX), protein synthesis inhibitor, was transferred into each well for a final concentration of 50 μg/mL and incubated for 1 hour. KOS-6β and dlx3.1-6β were then dispensed at an MOI 5 per well. After 4 hours post infection, virus and CHX was washed off with phosphate-buffer saline twice. Complete Williams E media was added to each well and the compounds were transferred by echo 555 acoustically into each well for a final concentration of 10 μM. After 20 hours of incubation, β-galactosidase assays were performed.

2.6. Chemo-informatics screen

The hits from the initial screen were clustered using Canvas by Schrodinger (Schrodinger Release, 2017; Duan et al., 2010; Sastry et al., 2010). MACCS fingerprints were calculated for each compound. Compounds were clustered using hierarchical clustering and leader-follower clustering, with various merge distances/cluster radii. Leader-follower clustering of MACCS hits with a cluster radius of 0.3 yielded visually intuitive clusters.

3. Results

3.1. Assay optimization

We first examined the replication of our reporter viruses, KOS6β and dlx3.1-6β (Figure 1A), in HepaRG cells (Figure 1B). The KOS6β titer is 2.95 x 10⁶ PFU/mL, and dlx3.1-6β titer is 1.13 x 10⁴ PFU/mL, a 261-fold difference. This difference is comparable to a previous study examining KOS and dlx3.1 replication in Vero cells (Sacks and Schaffer, 1987). The β-galactosidase viral reporter system allowed us to develop an automated colorimetric screen in 384-well plate format to rapidly process large numbers of compounds. To optimize our screen we examined several variables: serum levels, MOI, length of infectivity, and longevity of colorimetric signal (Figure 2). HepaRG cells, human hepatocytes, was selected for our cell-based assay because these cells are easy to culture and are readily infected by HSV-1 (Everett et al., 2008). To maintain viable growth conditions for HepaRG cells and reduce possible non-specific binding of viral particles to serum, we tested 2 concentrations (1% or 2%) of fetal bovine serum (FBS). As shown in Figure 3, absorbance signals for β-galactosidase activity at 1% and 2% FBS remained <0.16 in absence of virus, demonstrating no effect on β-galactosidase activity. Our assay utilizes reporter viruses, KOS6β (wt) and dlx3.1-6β (ICP0-null mutant), that have an ICP6 promoter::lacz expression cassette. ICP0 is a specific and potent inducer of ICP6, allowing us to utilize a β-galactosidase reporter system to monitor ICP0 transactivation activity. In presence of KOS6β or dlx3.1-6β, the absorbance remained consistent for either virus irrespective of MOIs or time of infectivity. As 2% FBS did not appear to impact cell viability, we decided to use media containing 2% FBS in subsequent experiments.

Figure 1. Reporter viruses.

(A) The background strain used in this approach is KOS, an HSV-1 wt strain (not drawn to scale), with unique long (U_L) and unique short (U_S) regions of the genome flanked by an inverted repeat sequences (ab U_L b’a’ and a’c’ U_s ca). Two arrows represent ICP0 which contains two copies in the HSV-1 genome, and ICP0 is a specific inducer of ICP6. The reporter viruses KOS6β and dlx3.1-6β both contain a ICP6 promoter fused with a lacz reporter gene cassette inserted between the UL49 and UL50 genes of HSV-1. dlx3.1-6β, an ICP0-null mutant, has a 3.1 kb deletion in both copies of ICP0 gene. This figure is adapted from Davido, et al., 2003. (B) HepaRG cells were infected at an MOI of 5 with KOS6β or dlx3.1-6β. At 24 hpi, cells were harvested, and viral yields were determined by standard plaque assays. Error bars represent the standard deviations from 3 independent experiments.

Figure 2. Schematic of optimization assay.

Serum percentage, multiplicity of infection (MOI), infection period, and β-galactosidase assay stability. Scheme of optimization process: 6,750 HepaRG cells were seeded in each well of 384-well plates with 1% or 2% FBS, incubated at 37°C in 5% CO₂ for 24 hours. Cells were then infected with KOS6β or dlx3.1-6β at MOIs of 0, 0.2, 1, and 5. β-galactosidase levels were analyzed at 6 hpi, 12 hpi, and 24 hpi using CPRG. β-galactosidase stability was examined at 5, 35, 80, 120, and 1080 minutes.

We then examined MOIs at 0, 0.2, 1, and 5 PFU/cell to determine the optimal MOI for β-galactosidase activity signal. KOS6β, absorbance signals ranged from 0.5 to 1.5, with a MOI of 5 reaching optimal signal by 24 hours post-infection (hpi). The MOIs of 0.2, 1, and 5 were clearly differentiated at 12 hours post-infection, irrespective of FBS concentration. By 24 hpi β-galactosidase levels of cells infected at MOIs of 0.2 and 1 began to approach those samples with an MOI of 5. For dlx3.1-6β infections at the lower MOIs (i.e., 0.2 and 1.0), the absorbance was at background levels, regardless of the infection time point. A reproducible increase in absorbance (2-3-fold) at an MOI of 5 by 24 hpi was observed compared to mock-infected cells. These data indicate that ICP0 strongly transactivates the ICP6 promoter of HSV-1, mirroring results from other published reports. We ultimately used an MOI of 5, as it gave the highest signal to the background control for both viruses.

The kinetics of β-galactosidase activity for all groups were analyzed at 6, 12, and 24 hpi. For KOS6β, β-galactosidase expression was noticeably detected at 6 hpi for all MOIs, with absorbance values showing marginal to substantial increases by 12 and 24 hpi. 24 hpi was selected as optimal infection time point, as maximal β-galactosidase activities were observed with the reporter viruses at an MOI of 5.

Lastly, we analyzed the stability of β-galactosidase assay after stop solution was added and absorbance read 5, 35, 80, 120, and 1080 minutes later. Given the stability of β-galactosidase activities for MOI or time of infection, all subsequent assays were read 1080 minutes after stop solution was added. In summary, we selected the optimized conditions of 2% FBS, MOI of 5, 24 hpi, and 1080 minutes plate reads for our reporter assays.

3.2. Assay validation: primary screen with roscovitine

After establishing our final conditions described in Figure 4, this assay was initially validated in a screen setting. Roscovitine, a broad cdk inhibitor, blocks the expression of many HSV-1 genes (Schang et al., 1998, Schang et al., 1999; Schang et al., 2000; Davido et al., 2002). Roscovitine was used as a positive control to validate the inhibition of HSV-1 gene expression in our reporter assay. Cells were pre-treated for 40 minutes with roscovitine over a range of concentrations. The OD signals of β-galactosidase for both KOS6β and dlx3.1-6β decrease in dose-dependent response to roscovitine (Figure 5A). A sigmoidal-dose response was obtained for inhibition of roscovitine against the two HSV-1 reporter viruses. Roscovitine showed an inhibitory concentration of 50% (IC50) at 17.39 μM for KOS6β and 8.18 μM for dlx3.1-6β (Figure 5B). Approximately 50% loss of cell viability was observed with 100 μM roscovitine (Figure 5C).

Figure 4. Schematic of primary screen.

The assay used 2% of FBS, 5 PFU/cell for both reporter viruses, and 24 hpi. The approach was tested using a known small molecule inhibitor of HSV gene expression, roscovitine, as a positive control and the negative control, DMSO (0.6%). These conditions were also used to screen multiple compound libraries.

Figure 5. Inhibition of β-galactosidase activity from reporter viruses by roscovitine.

(A) At an MOI of 5, the relative OD values of dlx3.1-6β as compared to KOS6β at various concentrations of roscovitine are shown. To confirm the activity of roscovitine in our assay, dose-dependent responses to (B) β-galactosidase expression and (C) cytotoxicity were measured. Error bars represent the standard deviations from 4 independent trials.

3.3. Assay robustness and reproducibility

This assay was tested for its robustness and reproducibility to be used in screening libraries of small molecule inhibitors (Figure 6). With KOS6β, roscovitine at 50 μM and 100 μM had a 3.75- and 9.3-fold reduction in β-galactosidase activity, respectively, compared with untreated controls. This result validates the sensitivity of our assay with the use of a known small molecule inhibitor of HSV-1 gene expression. Furthermore, wells only containing the reporter virus, KOS6β, exhibited a low absorbance signal (0.16 ± 0.014), whereas wells infected with KOS6β and treated with roscovitine reached an average Abs of 0.8 ± 0.048 and 0.32 ± 0.042 at 50 μM and 100 μM concentrations, respectively. The Z′ score is a statistical indicator of assay quality, measuring assay signal dynamic range, data variation associated with sample measurement, and data variation associated with reference controls. A score between 0.5 and 1 indicates suitability of assay for high throughput screening (Zhang et al., 1999). The Z′-factor for all samples were above ≥0.69, indicative of good separation of the positive and negative controls in our assay.

Figure 6. Well plate uniformity.

Scatterplot of 16 wells examining in samples containing KOS6β, cells, and KOS6β and DMSO (0.6%), KOS6β-infected cells plus roscovitine (50 μM), or KOS6β-infected cells plus roscovitine (100 μM). The average and standard deviation from each sample were measured to calculate the Z′ score.

3.4. Screen with KU-HTSL: pilot screen and secondary screen

We employed our high-throughput assay and screened ~19,000 compounds from the High Throughput Screening Laboratory at the University of Kansas (KU-HTSL). The KU-HTSL has a collection of diverse small molecules with unique scaffolds from several commercial vendors. These included bioactive FDA approved inhibitors, natural product scaffolds amenable to chemical synthesis, purified drug-like compounds, and purified secondary metabolites. The screen of ~19,000 compounds at 10 μM resulted in a hit rate of 4.6%, (840 compounds), focusing on compounds above 3 standard deviations from the plate median. To eliminate false positives due to cytotoxicity of the compounds, ATP levels were measured for cell viability. The cytotoxicity assay reduced the number of hits to 349 compounds, a hit rate of 1.9%. To help assess if the compounds are directly or indirectly blocking ICP0 transactivation activity, we employed a secondary assay, a cycloheximide (CHX) block and release. CHX blocks protein synthesis and allows ICP0 transcripts to accumulate. At 4 hpi, the CHX block is released and each compound is added when ICP0 protein is expressed. Inclusion of the secondary assay resulted in 76 final hits, a 0.4% hit rate, which helped focus our efforts on specific compounds. We then utilized a chemo-informatic approach (Canvas by Schrodinger) to cluster compounds based on chemical structure and filter-out compounds that are promiscuous or reactive. This resulted in 42 clusters, including singletons, and eliminated 6 hits flagged as pan-assay interference compounds (PAINS).

3.4.1. Clusters & singletons: trichothecenes, lipopeptides, and cyclin-dependent kinases

One cluster of 10 compounds was identified to be a family of trichothecenes, secondary metabolites, produced by fungi. Previous studies have shown the inhibitory effect of trichothecenes on HSV are likely due to the binding of the compound to the polyribosomes, inhibiting viral protein synthesis (Tani et al., 1995; Okazaki et al., 1992; Okazaki et al., 1988). Another cluster contained in two compounds of cyclic lipopeptides, biosurfactants produced by Bacillus subtilis. A previous study showed treatment with one of these lipopeptides reduced HSV-1 titers by >25,000-fold, which was due to the disintegration of the HSV-1 viral envelope and capsid (Vollenbroich et al., 1997).

Several singletons, which have unique chemical structures, are pan-cdk inhibitors. The activities of cdks have been shown to be required for HSV-1 replication and transcription and regulate ICP0 function (Schang et al., 1998; Davido et al., 2003; Davido et al., 2002). One hit compound we identified, a known cdk-7 and −9 inhibitor, was shown to inhibit expression of all immediate-early genes including ICP0 (Hou et al., 2017). Overall, a subgroup of the hits (Table 1) provide a proof of concept that our high-throughput assay is capable of identifying inhibitors of HSV-1.

Table 1.

Select hits from KU-HTSL screen.

Compoundsa	Percent Inhibition in Primary Screenb	Percent Inhibition in Secondary Screenc
Trichothecenes
KU0188522	100	99
KU0280950	93	95
KU0281790	100	99
KU0281792	99	99
KU0282448	90	91
KU0283070	100	99
KU0283235	98	99
KU0283795	97	97
KU0283847	101	100
Lipopeptides
KU0283335	100	99
KU0283824	100	94
Cyclin-Dependent Kinase Inhibitors
KU0191030	97	96
KU0190175	90	95
KU0190205	94	96
KU0189256	99	97
KU0190358	99	95
KU0190362	98	96

^aSubgroup of final hits from KU-HTSL screen listed as University of Kansas compound identifiers

^bPercent inhibition of KOS6β from primary screen as described in section 2.4

^cPercent inhibition of KOS6β from secondary screen as described in section 2.5

4. Discussion

ICP0 is a crucial viral regulatory protein that can dictate lytic infection or latency during an HSV-1 infection. We and others have demonstrated that ICP0 is a potent transactivator to all classes of HSV-1 genes, stimulating viral infection and reactivation. ICP0 is an attractive target for the development of novel antiherpetics, and such antivirals would be expected to limit HSV-1 lytic infection from reactivation. To date, identification of compounds that specifically inhibit ICP0 are limited.

To achieve this goal, we established a chemical-biological assay utilizing HSV-1 reporter viruses that allowed us to monitor ICP0 transactivation function in tandem with small molecule compounds. Our optimization experiments led us to select the conditions of 2% FBS, MOI of 5 for the reporter viruses, 24 hpi, and measuring β-galactosidase 1080 minutes after the addition of stop solution. The assay achieved a robust and reproducible Z′-factor of ≥0.69 for various controls, which meets high throughput screening criteria. The sensitivity and effectiveness of the reporter system was validated using roscovitine, which displayed a dose-dependent response with the reporter viruses. Overall, this method was optimized to produce consistent and reproducible measurements for monitoring ICP0 transactivating activity in a high-throughput approach.

It is notable that our approach confers several advantages over other assays. One is the simplicity of the assay, where all components are added directly to the 384-well plate, requiring minimal handling of the samples through automation. The reporter viruses provide another advantage, as the lacZ gene enables for a simple, inexpensive, and direct colorimetric screen for β-galactosidase activity. Previous assays utilized radioactive substrates or expensive equipment in fluorescent-based assays (Sekulovich et al., 1988); our approach eliminates those potential issues. Additionally, this assay provides a rapid and feasible method to screen multiple libraries of small molecule compounds in a 384-well plate format, requiring small amounts of compounds for testing that are in micromolar range. Lastly, our high-throughput assay has the potential to be used in combination with other genetic screens (e.g., CRISPR/Cas9) to identify novel cellular factors involved in HSV-1 replication. A potential limitation of our assay is that false positive hits may be identified in our high-throughput assay. Such compounds could target ICP0 protein levels, ICP0-dependent transcriptional coactivators, and ICP0 antagonists. As previously discussed, we employed a secondary assay using a cycloheximide block and release experiment to minimize such false positive hits. Future studies will be focused on determining how several of the 76 hits we identified in our screen inhibit HSV-1 and ICP0.

Highlights.

• Rapid high-throughput approach to screen multiple libraries of compounds to identify HSV-1 ICP0 transactivation inhibitors.

• ICP0 is a specific transactivator of the HSV-1 ICP6 promoter::lacz expression cassette in the reporter virus, KOS6β.

• The assay conditions were optimized for serum concentrations, MOIs, hours post infections, and time of plate reads.

• Previous reported compounds were identified as hits, indicating proof of concept for our high-throughput approach.

Glossary

CHX

cycloheximide

DMSO

dimethylsulfoxide

FBS

fetal bovine serum

hpi

hours post infection

HTS

high-throughput screening

HSV-1

herpes simplex virus 1

ICP0

infected cell protein 0

ICP6

infected cell protein 6

KU HTSL

KU High-Throughput Screening Laboratory

MOI

multiplicity of infection

wt

wild type