Cotreatment of T1012G and propranolol exerted a synergistic killing effect in gastric cancer
The IC50 of T1012G were determined to be 0.15 MOI, 0.04 MOI, and 0.21 MOI in HGC-27, AGS and MFC cell lines, respectively and the IC50 of propranolol were 70µM, 67µM and 86µM in the three cell lines (Fig. 1A and 1B). The combined therapy of these two agents exhibited enhanced inhibition on cell viability in a concentration dependent manner (Fig. 1C-1E). The synergistic effect was measured by combination index (CI) using Chou-Talalay algorithm. The lowest CI values (0.523, 0.607, and 0.657) were observed in co-treatment group 60µM + 1MOI, 80µM + 0.05MOI and 60µM + 0.01MOI in HGC-27, AGS, MFC cell lines, respectively (Table 1 and supplementary Table 1). In addition, pre-treated propranolol exhibited stronger synergistic effect than co-treatment model in HGC-27 (Fig. 1F). Under low dose virus infection (0.01, 0.05 and 0.1 MOI), the CI value of the pre-treatment group was significantly lower than that of the co-treatment group (0.549 vs. 1.023 P < 0.05, 0.624 vs. 0.944 P < 0.05, 0.540 vs. 0.829 P < 0.05) (Table 1).
Table 1
Combination index (CI) values for propranolol and T1012G combination for HGC-27 gastric cancer cell lines. The CI value of combination models were measured by Chou-Talalay method where CI value quantitatively defines synergism (CI < 1), additive effect (CI = 1) and antagonism (CI > 1). Data are presented as mean ± SEM, P < 0.05 (unpaired t-test).
|
Propranolol (40 µM) + T1012G (0.01MOI)
|
Propranolol + T1012G
|
Propranolol → T1012G
|
CI1
|
1.023 ± 0.035
|
0.549 ± 0.017
|
Combination effect
|
Nearly additive ±
|
Synergism +++
|
|
Propranolol (40 µM) + T1012G (0.05MOI)
|
Propranolol + T1012G
|
Propranolol → T1012G
|
CI1
|
0.944 ± 0.008
|
0.624 ± 0.042
|
Combination effect
|
Nearly additive ±
|
Synergism +++
|
|
Propranolol (40 µM) + T1012G (0.1MOI)
|
Propranolol + T1012G
|
Propranolol → T1012G
|
CI1
|
0.829 ± 0.130
|
0.540 ± 0.049
|
Combination effect
|
Moderate synergism ++
|
Synergism +++
|
1Combination index ( ±: Nearly additive (CI 0.90–1.10); +: slight synergism (CI 0.85–0.90); ++: moderate synergism (CI 0.70–0.85); +++: synergism (CI 0.30–0.70)) (55)
|
Propranolol pretreatment enhanced the propagation of T1012G in vitro
Propranolol (40 µmol/l) pre-treatment for 12 hours also did not affect T1012G (0.1 MOI) replication in HGC-27 cells (Fig. 2A, P>0.05). However, when the load of T1012G was decreased to 0.01 MOI and the pre-treatment was extended to 12, 24 and 48 hours, propranolol (40 µmol/l) yield a 3, 7, 6 folds increase of T1012G titers respectively when compared with T1012G only group(Fig. 2B, P < 0.05, P < 0.001, P < 0.001 ). This data suggested that a low dose of OVs is needed when it is used in combination with propranolol. On the other hand, the co-treatment and T1012G (0.1 MOI) pre-treatment could significantly shut down T1012G replications in HGC-27 cells (Fig. 2A), which suggested the importance of the sequential administration of these two drugs.
Propranolol pretreatment inhibited tumor growth and enhanced T1012G propagation in vivo
The synergistic effect of propranolol and T1012G was assessed in HGC-27 tumor engrafted in BALB/C nude mice. The mean tumor sizes of propranolol pre-treated mice were smaller than propranolol co-treated group (335.3 ± 36.92 mm3 vs. 659.3 ± 49.26 mm3, P < 0.01; Fig. 3B) and PBS group (335.3 ± 36.92 mm3 vs. 2100 ± 275.4 mm3, P < 0.01; Fig. 3B) on day 22 in HGC-27 tumor models. There was no statistical difference between T1012G only group and propranolol co-treated group (1118 ± 210.0 mm3 vs. 659.3 ± 49.26 mm3, P > 0.05; Fig. 3B). The differences in body weight of the mice were not observed among groups (Fig. 3D). The titers of T1012G were measured via standard plaque assay. Propranolol pre-treatment significantly increased T1012G replication in tumors when compared with T1012G only group (1.15×106 ± 2.5×105 pfu/ml vs. 2.84×105 ± 3.5×104 pfu/ml, P < 0.01; Fig. 3E), while there was no statistical difference between propranolol co-treated group and T1012G only group (3.84×105 ± 3.5×104 pfu/ml vs. 2.84×105 ± 1.1×105 pfu/ml, P >0.05; Fig. 3E). Overall, these results indicated that pretreatment with propranolol can enhance the efficacy of T1012G in gastric cancer models, possibly because of the elevated T1012G titers in tumors.
Propranolol pretreatment enhanced the expression of viral proteins and affected T1012G induced antiviral immune response
We next examined the expression of viral proteins at different stage of infection including immediate early viral proteins (ICP4, ICP0, ICP27), early viral protein (ICP8), and late viral proteins (VP16, US11). After 20 hours infection, ICP4, ICP0, and ICP27 was increased 10.3-fold, 3.2-fold, and 5.0-fold compared with T1012G only group, respectively (Fig. 4A-B, P<0.001, P<0.01, P<0.01, ). The ICP8 was significantly upregulated 9.4 fold compared with T1012G only group (Fig. 4A-B, P<0.01). A 3.2-fold and 3.4-fold increase of VP16 and US11 was observed between these groups (Fig. 4A-C, P<0.01, P<0.01). These data indicated that propranolol pre-treatment enhanced propagation of T1012 by facilitating the synthesis of vital viral proteins at different stages.
In order to detect specific mechanism, we found that the phospho-STAT3 (p-STAT3) and total STAT3 was inhibited after a 48 hours treatment of propranolol (40 µM) compared with untreated group (Fig. 4D-E, P < 0.05, P<0.0001). After 9 and 20 hours infection of T1012G, p-STAT3 in the propranolol pretreatment group was significantly up-regulated 2.9 and 2.0-fold compared with T1012G only group (Fig. 4D and 4F, P < 0.05, P < 0.01) and total STAT3 was increased 2.3 and 1.9-fold (Fig. 4D and 4F, P<0.05,P<0.05), while the expression and activation of other IFN-I responsive genes-STAT1, STAT2 were not affected significantly (Supplementary Fig. 1, P>0.05). Meantime, phospho-PKR (p-PKR) increased 2.2-fold in T1012G only group when compared with control group (Fig. 4G and 4H, P<0.05), suggesting that the activation of PKR may play an important role in mediating antiviral response. However, propranolol pretreatment induced a 62.46% ± 5.06%/65.94% ± 10.11% decrease of total-PKR/p-PKR compared with T1012G only group (Fig. 4G and 4H, P<0.05)indicating a potent suppression on interferon induced antiviral response. These data suggested that propranolol pretreatment could inhibit the virus induced expression of antiviral genes PKR through enhancing the activation of STAT3. However, cotreatment or virus pretreatment could inhibit the expression of p-STAT3 and total STAT3 compared with virus only group (Fig. 4I-4L, P < 0.05, P<0.05 or P < 0.05, P<0.05). Meantime, p-PKR and total PKR increased 1.5-fold and 3.0-fold or 1.7-fold and 2.4-fold in cotreatment or virus pretreatment group when compared with virus only group (Fig. 4I-4L, P<0.05, P<0.01 or P<0.05, P<0.01).
STAT3 altered HSV-1 propagation and cytotoxicity in gastric cancer cells by inhibiting IFN-I antiviral pathway in response to T1012G infection
Human HGC-27 gastric cancer cells were transfected with siRNA target to STAT3 (si-1, si-2) and si-NC (negative control). The cytotoxicity of HSV-1 against cells was measured with STAT3 knockdown by si-STAT3 (Fig. 5A) or with STAT3 overexpression (Fig. 5B). Cells with STAT3 knockdown exhibited virus cytotoxicity at an IC50 of 6.2 MOI and 5.5 MOI, when compared to control cells (IC50 of 0.18 MOI). Conversely, overexpressing cells exhibited cytotoxicity that was measured with an IC50 of 0.009 MOI compared with control cells (IC50 of 0.195 MOI). Genetic manipulation of STAT3 altered T1012G-mediated cytotoxicity of gastric cancer cells. Figure 5C showed that propranolol pretreatment enhanced the cytotoxicity of T1012G, meantime, this effect could be reversed by knocking down STAT3 or be enhanced by STAT3 overexpression.
In contrast to parental cells that expressed basal levels of STAT3, there was 94.97% ± 0.30% and 96.89% ± 0.28% reduction in viral production in HGC-27 with STAT3 knocked down (Fig. 5E and 5G, P<0.01, P<0.001). Meantime, a 6-fold increase of viral titers was observed in STAT3 overexpressed HGC-27 cell lines compared with control group (Fig. 5H and 5I, P<0.001). Collectively, this data showed that STAT3 expression positively correlates with HSV-1 replication.
Figure 5 Cytotoxicity of T1012G against gastric cancer cells, virus replication and expression of IFN-I response genes after T1012G infection with altered STAT3 gene expressions. (A-B) Cell viability (measured by cck8) of HGC-27 gastric cancer cells transfected with siRNA or plasmid was assayed 2 days after infection of T1012G at different MOI. (C) After 48 hours of incubation with or without propranolol (40 µmol/l), cells transfected with control treatment, siRNA or plasmid were infected with different dose of virus (0.01, 0.05, 0.1, 0.5, 1 MOI). The number of surviving cells in each well was determined 2 days after infection. (D, F, H) Western blot analysis of STAT3 in the transfected siRNA (si-1 and si-2) and plasmid. (E, G, I) Cells were infected (MOI 0.01) with HSV-1 (T1012G) and virus yields were determined on Vero cells after 48-hour infection. (J, L, N) Expression of IFN-I responsive genes, PKR, 9 or 20 hours following T1012G infection (0.01 MOI) in (J,L): HGC-27 cells treating with STAT3 siRNA or vehicle and (N): HGC-27 cells treating with plasmid-STAT3 or vehicle. (K, M, O) Quantification of J, L, N. Results are presented as mean ± SEM, significant differences were evaluated using one-way ANOVA. * P < 0.05, ** P < 0.01, *** P < 0.01 and **** P < 0.0001 (Tukey test for multiple comparisons). ** P < 0.01 and *** P < 0.01 (Dunnett's test for multiple comparisons). OE: overexpression
Propranolol Pretreatment Counteracted Ifn-α/β-mediated Inhibition Of Viral Propagation
IFN-α/β treatment inhibited the replication of T1012G by > 70% in HGC-27 cells at 6h (Fig. 6A-B, *** P < 0.01). With the prolongation of interferon treatment time, the degree of virus replication inhibition was greater than 6h treatment (~ 90% reduction). Propranolol treatment counteracted IFN-α/β in a dose-dependent manner in cells infected with T1012G (Fig. 6A-B). The 40 µM propranolol treatment cotreated with IFN-α/β increased the viral yields (compared with IFNα: 3.9-fold, 4.2-fold, 5.1-fold; IFNβ: 4.6-fold, 4.6-fold, 3.0-fold) (Fig. 6A-B). The expression of PKR or p-PKR was significantly upregulated 2.0-fold/2.7-fold or 2.8-fold/2.5-fold in response to IFN-α/β (Fig. 6C-F, * P < 0.05, *** P < 0.001, * P < 0.05, *** P < 0.001), and the IFN-mediated upregulation was significantly attenuated (91.57% ± 4.40%/92.57% ± 1.77% or 99.63% ± 0.16%/99.34%± 0.13% decrease) by propranolol treatment ((Fig. 6C-F, ** P < 0.01, *** P < 0.001, ** P < 0.01, *** P < 0.001), while the cotreatment with propranolol and IFN-α/β induced 4.2-fold/2.2-fold or 5.5-fold/4.7-fold increase in total STAT3 or p-STA3 compared to IFN-α/β treatment ((Fig. 6C-F, ** P < 0.01,* P < 0.05, ** P < 0.01, ** P < 0.01). These results indicated that propranolol pretreatment could counteract the actions of IFNs and rescue viral yields by preventing IFN-mediated upregulation of PKR.
Virus infection enhanced the secretion of IFNα, and propranolol pretreatment could further enhance its secretion
With the prolongation of virus infected time, the secretion of IFNα increased continuously (compared with oh: 9h: * P < 0.05; 20h: ** P < 0.01; 36h: *** P < 0.001; Fig. 7A), while there was no obvious effect on the secretion of IFNβ(P>0.05;Figure 7B). Meantime, we also found that propranolol treatment could slightly enhance IFNα secretion compared with untreated group, and propranolol pretreatment could further enhance IFNα secretion compared with single treatment group after 9h or 20h virus infection (9h: compared with propranolol group: #P < 0.05; compared with T1012G group: &&P < 0.01; 20h: compared with propranolol group: ##P < 0.01; compared with T1012G group: &&P < 0.01;Figure 7C), while there was no significant effect on the secretion of IFNβ (P>0.05; Figure 7D). These results indicated that propranolol pretreatment may further enhance virus replication by enhancing the secretion of IFNα and then promoting the activation of its downstream product-STAT3.