Potent anti-tumor effects of receptor-retargeted syncytial oncolytic herpes simplex virus

Abstract

Most oncolytic virotherapy has thus far employed viruses deficient in genes essential for replication in normal cells but not in cancer cells. Intra-tumoral injection of such viruses has resulted in clinically significant anti-tumor effects on the lesions in the vicinity of the injection sites but not on distant visceral metastases. To overcome this limitation, we have developed a receptor-retargeted oncolytic herpes simplex virus employing a single-chain antibody for targeting tumor-associated antigens (RR-oHSV) and its modified version with additional mutations conferring syncytium formation (RRsyn-oHSV). We previously showed that RRsyn-oHSV exhibits preserved antigen specificity and an ∼20-fold higher tumoricidal potency in vitro relative to RR-oHSV. Here, we investigated the in vivo anti-tumor effects of RRsyn-oHSV using human cancer xenografts in immunodeficient mice. With only a single intra-tumoral injection of RRsyn-oHSV at very low doses, all treated tumors regressed completely. Furthermore, intra-venous administration of RRsyn-oHSV resulted in robust anti-tumor effects even against large tumors. We found that these potent anti-tumor effects of RRsyn-oHSV may be associated with the formation of long-lasting tumor cell syncytia not containing non-cancerous cells that appear to trigger death of the syncytia. These results strongly suggest that cancer patients with distant metastases could be effectively treated with our RRsyn-oHSV.

Keywords: HSV; cancer; cell death; fusogenic; gene therapy; herpes simplex virus; intravenous; oncolytic virotherapy; systemic delivery; targeting.

Graphical abstract

Figure 1

Schematic representation of the genomic structures of the recombinant HSVs used in this study U_L, unique long segment; U_S, unique short segment; CMVp, human cytomegalovirus major immediate-early (HCMV-IE) promoter; EGFP, expression cassette for EGFP inserted in the intergenic region between the UL3 and UL4 genes; ΔICP34.5, deletion of ICP34.5 gene; Bh, R858H mutation in gB; Kt, A40T mutation in gK; NT, D285N/A549T mutations in gB that increase the rate of virus entry; scEGFR, anti-hEGFR scFv-fused gD; scEpCAM, anti-hEpCAM scFv-fused gD; Closed boxes, terminal and internal inverted repeats.

Figure 2

Efficiency of anti-tumor activity of a single intra-tumoral injection of hEGFR-specific RR-oHSV and ICP34.5-deleted CR-oHSV U87 cells were injected subcutaneously into SCID-beige mice. When tumor volumes reached ∼310 mm³ (arrow), PBS (black circles; n = 6), 10⁴ (blue diamonds; n = 6) or 10⁷ (blue triangles; n = 6) pfu of KGΔ, or 10⁴ (red diamonds; n = 6) or 10⁷ (red triangles; n = 6) pfu of KGNE was injected into the tumors. Tumor volumes were measured, and mean volumes ± standard errors (SEs) are shown. Two-way repeated-measures analysis of variance (ANOVA) was used for comparative analysis. ∗p < 0.05; ns, not significant. Experiments were repeated twice, and similar results were observed.

Figure 3

In vitro plaque formation and cell killing by the syncytial derivatives of hEGFR-specific RR-oHSV and ICP34.5-deleted CR-oHSV (A) U87 cells were infected with the viruses shown above the panels and overlaid with methylcellulose-containing medium. EGFP signals and phase-contrast images were recorded 2 days post-infection. Photographs of representative plaques are shown. Scale bars, 500 μm. (B) U87 cells were infected with the viruses shown below the graph and overlaid with methylcellulose-containing medium. Areas of plaques 3 days post-infection are shown (n = 15). Means ± standard deviations (SDs) are also shown. Dunn’s test was used for comparative analysis. ∗p < 0.05; ns, not significant. (C) U87 cells were infected at MOIs indicated below the graph for 3 days, and cell viability relative to uninfected cells was assessed by MTT assay. Means ± SDs are shown (n = 6). Sidak test was used for comparative analysis. ∗p < 0.05; ns, not significant. White bars, KGΔ; black bars, KGNE; vertical-striped bars, KGΔ-BhKt; horizontal-striped bars, KGNE-BhKt. Experiments were repeated twice, and similar results were observed.

Figure 4

Syncytium formation in U87 xenografts caused by the syncytial derivatives of hEGFR-specific RR-oHSV and ICP34.5-deleted CR-oHSV U87 cells were injected subcutaneously into SCID-beige mice, and, when tumor volumes reached >780 mm³, PBS or 10⁵ pfu of KGΔ, KGNE, KGΔ-BhKt, or KGNE-BhKt was injected into the tumors. Consecutive tumor sections were prepared 3 days post-injection and stained with hematoxylin and eosin (HE; upper) or immunostained for EGFP expression (lower). Arrowheads, ground-glass nuclei; Arrows, multinucleated giant cells. Scale bars, 50 μm.

Figure 5

Effects of the introduction of syncytial mutations into hEGFR-specific RR-oHSV and ICP34.5-deleted CR-oHSV on the efficiency of anti-tumor activity when injected intra-tumorally (A) U87 cells were injected subcutaneously into SCID-beige mice, and, when tumor volumes reached ∼300 mm³ (arrow), PBS (black circles; n = 6), 10² (blue diamonds; n = 6) or 10³ (blue triangles; n = 6) pfu of KGNE, or 10¹ (red squares; n = 6), 10² (red diamonds; n = 6), or 10³ (red triangles; n = 6) pfu of KGNE-BhKt was injected into the tumors. Tumor volumes were measured, and mean volumes ± SEs are shown. Two-way repeated-measures ANOVA was used for comparative analysis. ∗p < 0.05; ns, not significant. (B) PBS (left; n = 6), 10⁷ pfu of KGNE (center; n = 6), or 10² pfu of KGNE-BhKt (right; n = 6) were intra-tumorally injected into SCID-beige mice bearing subcutaneous U87 xenografts (∼780 mm³, 23 days after inoculation; arrows). Tumor volumes were measured and plotted. (C) Survival of the mice in (B) was monitored and is shown in a Kaplan-Meier plot. Log-rank testing was used for comparative analysis. Black line, PBS; blue line, 10⁷ pfu of KGNE; red line, 10² pfu of KGNE-BhKt. (D) PBS (black circles; n = 6), 10⁷ pfu of KGΔ (blue circles; n = 6), 10¹ (red squares; n = 6), 10³ (red diamonds; n = 6), 10⁵ (red triangles; n = 6), or 10⁷ (red circles; n = 6) pfu of KGΔ-BhKt, or 10¹ pfu of KGNE-BhKt (green circles; n = 6) was intra-tumorally injected into SCID-beige mice bearing subcutaneous U87 xenografts (∼320 mm³; arrow). Tumor volumes were measured, and mean volumes ± SEs are shown. Two-way repeated-measures ANOVA was used for comparative analysis. ∗p < 0.05; ns, not significant. Experiments were repeated twice, and similar results were observed.

Figure 6

Efficiency of anti-tumor activity of a single intra-venous injection of the syncytial derivatives of the hEGFR- or hEpCAM-specific RR-oHSVs (A) PBS (left; n = 5) or 10⁵ (center; n = 5) or 10⁷ (right; n = 5) pfu of KGNE-BhKt was injected into SCID-beige mice bearing subcutaneous U87 xenografts (∼700 mm³, 24 days after inoculation; arrows) via the tail vein. Tumor volumes were measured and plotted. (B) KGNE-BhKt (10⁷ pfu) was injected into SCID-beige mice bearing subcutaneous U87 xenografts (∼400 mm³) via the tail vein. Tumor sections were prepared 1 or 3 days post-injection and immunostained for EGFP expression. Scale bars, 2 mm. See also Figure S1. (C) PBS (left; n = 5) or 10⁵ pfu of KGNE-BhKt (center; n = 5) or KGNEp-BhKt (right; n = 5) was injected into SCID-beige mice bearing subcutaneous HepG2 xenografts (∼580 mm³, 26 days after inoculation; arrows) via the tail vein. Tumor volumes were measured and plotted. Two-way repeated-measures ANOVA was used for comparative analysis. (D) PBS (black circles; n = 6) or 10⁶ pfu of KGNE-BhKt (blue circles; n = 6) or KGNEp-BhKt (red circles; n = 6) was injected into SCID-beige mice bearing subcutaneous U87 xenografts (∼280 mm³, 9 days after inoculation; arrow) via the tail vein. Tumor volumes were measured and mean volumes ± SEs are shown. Two-way repeated-measures ANOVA was used for comparative analysis. ns, not significant. Experiments were repeated twice, and similar results were observed.

Figure 7

In vitro expansion of syncytia generated by the syncytial derivatives of the ICP34.5-deleted CR-oHSV and hEGFR- or hEpCAM-specific RR-oHSVs (A) U87 cells were infected with viruses shown on the left at an MOI of 0.00015 for 2 h, seeded in combination with the same number of cells indicated below the panels, and overlaid with methylcellulose-containing media. Mixed cultures were observed for EGFP and mCherry signals in a time-lapse manner. Photographs at 29 or 48 hpi are shown (see time-lapse movies in Videos S1 and S2). (B) U87 cells were infected with viruses shown on the left at an MOI of 0.000004 for 2 h, seeded in combination with the same number of cells indicated below the graphs, and overlaid with media containing methylcellulose and SYTOX Orange. Eight syncytia were monitored during the period of 24–78 hpi. Areas of syncytia were measured and plotted. Two-tailed Mann-Whitney U test was used for comparative analysis. ×, last measurement before becoming positive for SYTOX Orange staining. (C–F) U87 (C), HepG2 (D and F), or Hepa1-6-hEpCAM (E) cells were infected with viruses shown on the left at MOIs of 0.000006–0.0003 for 2 h, seeded in combination with the same (C–E) or double (F) number of cells indicated below the panels, and overlaid with media containing methylcellulose and SYTOX Orange. Mixed cultures were observed for EGFP and SYTOX signals 5 (C), 4 (D and F), or 2 (E) days post-infection, respectively. Experiments were repeated twice, and similar results were observed. Scale bars, 500 μm (A, D, and E), 1 mm (C), or 2 mm (F).