Modulation of GSK-3β activity in Venezuelan equine encephalitis virus infection

Abstract

Alphaviruses, including Venezuelan Equine Encephalitis Virus (VEEV), cause disease in both equine and humans that exhibit overt encephalitis in a significant percentage of cases. Features of the host immune response and tissue-specific responses may contribute to fatal outcomes as well as the development of encephalitis. It has previously been shown that VEEV infection of mice induces transcription of pro-inflammatory cytokines genes (e.g., IFN-γ, IL-6, IL-12, iNOS and TNF-α) within 6 h. GSK-3β is a host protein that is known to modulate pro-inflammatory gene expression and has been a therapeutic target in neurodegenerative disorders such as Alzheimer's. Hence inhibition of GSK-3β in the context of encephalitic viral infections has been useful in a neuroprotective capacity. Small molecule GSK-3β inhibitors and GSK-3β siRNA experiments indicated that GSK-3β was important for VEEV replication. Thirty-eight second generation BIO derivatives were tested and BIOder was found to be the most potent inhibitor, with an IC(50) of ∼0.5 µM and a CC(50) of >100 µM. BIOder was a more potent inhibitor of GSK-3β than BIO, as demonstrated through in vitro kinase assays from uninfected and infected cells. Size exclusion chromatography experiments demonstrated that GSK-3β is found in three distinct complexes in VEEV infected cells, whereas GSK-3β is only present in one complex in uninfected cells. Cells treated with BIOder demonstrated an increase in the anti-apoptotic gene, survivin, and a decrease in the pro-apoptotic gene, BID, suggesting that modulation of pro- and anti-apoptotic genes contributes to the protective effect of BIOder treatment. Finally, BIOder partially protected mice from VEEV induced mortality. Our studies demonstrate the utility of GSK-3β inhibitors for modulating VEEV infection.

Figure 1. Small molecule inhibitors of GSK-3β impede VEEV replication.

A) U87MG astrocytes were treated with BIO (10, 5, 2.5, or 1.25 µM) and cell viability assayed 24 hours later by CellTiter Glo Assay (Promega). DMSO treated cells are displayed at 100% viability and all treatments compared to those values. B) U87MG astrocytes were pretreated for 2 hours with DMSO, or BIO (5, 2.5, or 1.25 µM), infected with VEEV TC-83 at MOI 0.1, and post-treated with compounds. Twenty-four hours post infection viral supernatants were collected and assayed for viral replication by plaque assays. C) U87MG astrocytes were treated with DMSO, SB 415286 (50, 10, or 2 µM), or SB 216763 (50, 10, or 2 µM) and cell viability assayed 24 hours later by CellTiter Glo Assay (Promega). DMSO treated cells are displayed at 100% viability and all treatments compared to those values. D) U87MG astrocytes were pretreated for 2 hours with DMSO, SB 415286 (50, or 10 µM), or SB 216763 (50 or 10 µM), infected with VEEV TC-83 at MOI 0.1, and post-treated with compounds. Twenty-four hours post infection viral supernatants were collected and assayed for viral replication by plaque assays. E) U87MG astrocytes were transfected with negative control siRNA (siNEG) or GSK-3β siRNA (siGSK3) at 50 nM. Forty-eight hours post-transfection, cells were infected with VEEV TC-83 at MOI 0.1. Viral supernatants were collected 24 hours post-infection and assayed for viral replication by plaque assays. Western blot results are displayed below the graph, indicating that GSK-3β expression was decreased following siRNA treatment.

Figure 2. Identification of a BIO derivative that inhibits VEEV replication and CPE.

A) U87MG astrocytes were pretreated for 2 hours with DMSO or various BIO derivatives at 1 µM, infected with VEEV TC-83 at MOI 0.1, and post-treated with compounds. Twenty-four hours post infection viral supernatants were collected and assayed for viral replication by q-RT-PCR. Compounds with an asterisk displayed greater than 1 log inhibition of viral replication. The horizontal bar indicates the level of viral replication displayed by the DMSO control. B) U87MG astrocytes were treated with DMSO or various BIO derivatives at 1 µM. Forty-eight hours post infection cell viability was measured by CellTiter Glo luminescence cell viability assay. The horizontal bar indicates the cell viability displayed by the DMSO control (set to 100%). C) U87MG astrocytes were pretreated for 2 hours with DMSO, BIO, #6, #8, #10, or #19 at 1 µM, infected with VEEV TC-83 at MOI 0.1, and post-treated with compounds. Twenty-four hours post infection viral supernatants were collected and assayed for viral replication by plaque assays. D) BIO derivatives were assayed for their ability to inhibit VEEV induced CPE. U87MG astrocytes were pretreated for 2 hours with DMSO or various BIO derivatives at 1 µM, infected with VEEV TC-83 at MOI 0.1, and post-treated with compounds. Forty-eight hours post infection CPE was measured by MTT assay. Mock infected cells are displayed at 100% viability. The horizontal bar indicates the cell viability displayed by the VEEV infected and DMSO control. Compound #6 has an asterisk as cells treated with it displayed the greatest cell viability following VEEV infection.

Figure 3. Characterization of BIOder.

A) Structure of BIO: 6-bromoindirubin-3′-oxime and BIOder (#6): 6-bromo-5-methyl-1H-indole-2,3-dione 3. B) U87MG astrocytes were pretreated for 2 hours with DMSO or BIOder (10, 1.0, and 0.1 µM), infected with VEEV TC-83 at MOI 0.1, and post-treated. Twenty-four hours post-infection, supernatants were collected and analyzed by plaque assay to determine the amount of infectious virus released. C) U87MG astrocytes were pretreated for 2 hours with DMSO or BIOder (10, 1.0, and 0.1 µM), infected with VEEV TC-83 at MOI 0.1, and post-treated. Seventy hours post infection, CPE was measured by MTT assay. Mock infected cells are displayed at 100% viability. D) Uninfected U87MG astrocytes were treated with BIOder (10, 1.0, and 0.1 µM), and cell viability determined by MTT assay. Untreated (UT) cells are displayed at 100% viability.

Figure 4. BIO and BIOder inhibit GSK-3β in VEEV infected cells.

A) U87MG astrocytes and B) Vero cells were pretreated for 2 hours with DMSO, BIO (0.1 and 1.0 uM), or BIOder (0.1 and 1.0 uM), infected with VEEV TC-83 at MOI 0.1, and post-treated with DMSO, BIO, or BIOder. Mock infected cells were treated with DMSO, 1 uM BIO, or 1 uM BIOder. Twenty-four hours post infection cells were collected and protein extracts prepared. One mg of extract was IPed at 4°C overnight with GSK-3β antibody. The next day complexes were precipitated with A/G beads for two hours at 4°C. IPs were washed twice with TNE buffer and kinase buffer. Phosphorylation reactions were performed with IP material and 200 ng of glycogen synthase peptide 2 (Millipore) as substrate. Following incubation, samples were separated by SDS-PAGE, dried and subjected to analysis using Molecular Dynamics Phosphor Imager software. C) Immunoprecipitation of GSK-3β and IgG control from titration of extracts followed by staining. Constant amount of anti-GSK-3β or IgG Rabbit antibodies (10 µg each) were used for overnight precipitation using 10,100, and 1000 µg of VEEV infected U87MG extracts. Samples were IPed overnight and next day Protein A and G were added, washed in TNE₁₅₀ and 0.1% NP-40 and then ran on a 4–20% followed by staining with coomassie blue. A dominant band of GSK-3β along with both heavy and light chains are stained.

Figure 5. Presence of novel GSK-3β complexes in VEEV infected cells.

A) Samples from both infected and uninfected cells were loaded on a sizing column and separated in presence of a 500 mM salt buffer. No detergents were used during the chromatography steps. Each fraction (over a range of 60 fractions) contained 500 µL and only 250 µL samples were precipitated, resuspend in low volume of TNE₅₀ and 0.1% NP-40 (10 µL) and ran on a gel for western blotting. Samples were western blotted for presence of GSK-3β and actin. B) GSK-3β IPs from fractions 31, 41, 51 (100 µL each) were mixed with GSK-3β Ab (5 µg) overnight and washed next day first with RIPA buffer (1×) and then TNE₅₀ and 0.1% NP-40 (2×) followed by kinase buffer (3×) prior to addition of substrate and ³²P-ATP. Samples were ran on a gel, stained, destained (over 4 hrs), dried and then exposed to phosphoImager cassette. Control immunoprecipitation from these fraction (50 uL each) were mixed and used for IP/Kinase using IgG (5 µg). C) IP/kinase reactions were treated with either BIO or BIOder (0.1 uM or 1 µM) in vitro during the kinase reaction. Samples were ran on a gel and exposed to phosphoImager cassette.

Figure 6. BIOder treatment alters expression of apoptotic genes to promote survival of U87MGs.

A) U87MGs were treated with 1 µM BIO, 1 µM BIOder, or DMSO and infected with VEEV TC-83. RNA was harvested from infected cells 24 hours post infection and analyzed by RT-PCR for expression of the indicated anti- and pro-apoptotic genes. B) Band intensities corresponding to triplicate samples were quantified and represented as fold change in gene expression over the DMSO control, with the DMSO control being set as a fold change of 1.0.

Figure 7. BIOder inhibits VEEV cell death in vivo.

A) Female C3H/HeN mice were treated subcutaneously with either DMSO or various concentration of BIOder (10 mg/kg, 20 mg/kg, 40 mg/kg) every day for 5 days. Mice were weighed daily and the average % of the mouse weight is plotted in panel A. B) and C) Female C3H/HeN mice were infected intranasally with 5×LD50 (2×10⁷ pfu) of VEEV TC-83. Groups of 10 mice were treated SQ with vehicle, BIO (50 mg/kg) or BIOder (20 mg/kg) on days −1, 1, 3, and 5 and were monitored for survival for 14 days. Kaplan-Meier curves for survival between DMSO and BIO (panel B). Kaplan-Meier curves for survival between DMSO and BIOder (panel C). Significance was determined using Mantel-Cox Log-rank test. P-value of 0.057 between control and BIOder. This experiment was performed one time.