Hypoxia-targeted 131I therapy of hepatocellular cancer after systemic mesenchymal stem cell-mediated sodium iodide symporter gene delivery

Abstract

Adoptively transferred mesenchymal stem cells (MSCs) home to solid tumors. Biologic features within the tumor environment can be used to selectively activate transgenes in engineered MSCs after tumor invasion. One of the characteristic features of solid tumors is hypoxia. We evaluated a hypoxia-based imaging and therapy strategy to target expression of the sodium iodide symporter (NIS) gene to experimental hepatocellular carcinoma (HCC) delivered by MSCs.MSCs engineered to express transgenes driven by a hypoxia-responsive promoter showed robust transgene induction under hypoxia as demonstrated by mCherry expression in tumor cell spheroid models, or radioiodide uptake using NIS. Subcutaneous and orthotopic HCC xenograft mouse models revealed significant levels of perchlorate-sensitive NIS-mediated tumoral radioiodide accumulation by tumor-recruited MSCs using 123I-scintigraphy or 124I-positron emission tomography. Functional NIS expression was further confirmed by ex vivo 123I-biodistribution analysis. Administration of a therapeutic dose of 131I in mice treated with NIS-transfected MSCs resulted in delayed tumor growth and reduced tumor perfusion, as shown by contrast-enhanced sonography, and significantly prolonged survival of mice bearing orthotopic HCC tumors. Interestingly, radioiodide uptake into subcutaneous tumors was not sufficient to induce therapeutic effects. Our results demonstrate the potential of using tumor hypoxia-based approaches to drive radioiodide therapy in non-thyroidal tumors.

Keywords: gene therapy; hepatocellular carcinoma; hypoxia-targeting; mesenchymal stem cells; sodium iodide symporter.

Figure 1. Induction of hypoxia-responsive promoter-driven transgene expression under hypoxic conditions in vitro

(A, B) For invasion assays, MSCs were stably transfected with mCherry controlled by a hypoxia-responsive promoter (HIF-Cherry-MSCs) and labeled with a green fluorescent cell tracker dye. HIF-Cherry-MSCs located at the surface and thus in non-hypoxic areas of the HuH7 cell spheroid appear green (A), while after invasion into hypoxic areas of spheroids, HIF-Cherry-MSCs appear orange due to an overlay of hypoxia-induced mCherry expression and the green fluorescent cell tracker dye (B). Scale bar = 100 μm. (C) Western blot analysis of MSCs stably transfected with NIS driven by a hypoxia-responsive promoter (HIF-NIS-MSCs) showed increased NIS protein expression with a major band of ~90 kDa after stimulation with the hypoxia mimicking agent cobalt chloride as compared to unstimulated HIF-NIS-MSCs. Cropped blots are shown. MW, molecular weight; Std., standard. (D) ¹²⁵I uptake studies revealed a 31-fold increased perchlorate-sensitive NIS-mediated radioiodide uptake in HIF-NIS-MSCs after stimulation with cobalt chloride as compared to unstimulated HIF-NIS-MSCs. WT-MSCs and unstimulated HIF-NIS-MSCs showed no perchlorate-dependent ¹²⁵I accumulation above background level. Results are reported as mean ± SEM (n = 3; **p < 0.01).

Figure 2. Enhanced tumoral radioiodide accumulation after systemic HIF-NIS-MSC application in subcutaneous and intrahepatic HuH7 xenograft mouse models

¹²³I-scintigraphy (A–C) or ¹²⁴I-PET (D–F) imaging demonstrated enhanced radioiodide accumulation in subcutaneous (A; n = 12) and orthotopic (D; n = 6) HCC tumors after systemic HIF-NIS-MSC application 3 h after radioiodide injection, which was blocked by the competitive NIS inhibitor sodium perchlorate (B, E; subcutaneous model: n = 4; orthotopic model: n = 3), while mice injected with WT-MSC showed no tumoral radioiodide uptake above background level (C, F; subcutaneous model: n = 9; orthotopic model: n = 2). One representative image is shown per group. (G) Time course of radioiodide accumulation in HuH7 tumors as determined by serial scanning. Subcutaneous HCC xenografts showed a maximum ¹²³I uptake of 3.9 ± 0.4% ID/g with a biological half-life of 3.8 h, whereas orthotopic HCC tumors accumulated up to 6.9 ± 0.9% ID/g ¹²⁴I with a biological half-life of 4.0 h. Results are reported as mean ± SEM. Ex vivo biodistribution analysis confirmed perchlorate-sensitive radioiodide accumulation in subcutaneous ((H); 1.6 ± 0.5% ID/g; n = 12; perchlorate: n = 5) and intrahepatic ((I); 3.5 ± 0.6% ID/g; n = 6; perchlorate: n = 3) tumors. No significant radioiodide accumulation was measured in non-target organs. Results are reported as mean ± SEM.

Figure 3. MSC recruitment and hypoxia-induced NIS expression were higher in intrahepatic compared to subcutaneous HCC tumors

Compared to subcutaneous (s.c.) tumors (A), higher NIS-specific immunoreactivity was detected in intrahepatic (i.h.) HuH7 tumors (C). This correlated well with tumoral HIF-NIS-MSC recruitment (B, D). In mice injected with WT-MSCs, no NIS expression (E, G) was detected, though MSCs were recruited (F, H). Non-target organs showed neither MSC recruitment nor NIS expression (I–N), except for the spleen where no NIS staining (O) but positive MSC staining (P) were observed. One representative image is shown each. Scale bar = 100 μm.

Figure 4. Growth inhibition of orthotopic HuH7 tumors after application of a therapeutic dose of radioiodide in HIF-NIS-MSC-treated mice was associated with a prolonged survival

Two groups of mice were established that received 55.5 MBq ¹³¹I 48 h after the final of three HIF-NIS-MSC (subcutaneous model: n = 11; orthotopic model: n = 7) or WT-MSC (subcutaneous model: n = 9; orthotopic model: n = 5) applications in 2-day-intervals. This cycle was repeated once 24 h after the last radioiodide application. 24 h after these treatment cycles, one additional MSC injection was administered followed by a third ¹³¹I injection 48 h later. A further control group received HIF-NIS-MSCs and NaCl (subcutaneous model: n = 13; orthotopic model: n = 5). In animals harboring subcutaneous HCC xenografts, no significant difference in tumor growth ((A); mean ± SEM) or animal survival ((B); percent survival) was observed comparing therapy to control animals (WT-MSC + ¹³¹I and HIF-NIS-MSC + NaCl). In mice harboring orthotopic HuH7 tumors, HIF-NIS-MSC/¹³¹I application resulted in reduced tumor growth ((C–E); mean ± SEM) and increased mouse survival ((F); percent survival) as compared to control groups (WT-MSC + ¹³¹I; *p < 0.05 and HIF-NIS-MSC + NaCl; *p < 0.05).

Figure 5. Reduced perfusion of intrahepatic HCC tumors after application of a therapeutic dose of radioiodide in HIF-NIS-MSC-treated mice

Orthotopic HuH7 xenografts of HIF-NIS-MSC-treated mice showed an overall reduced contrast agent signal (A), PE (3652.6 ± 1364.0; (B)), WiAUC (32089.9 ± 14842.2; (C)), WiR (600.4 ± 172.2; (D)) and WiPI (2556.7 ± 960.5; (E)) after radioiodide therapy as compared to animals treated with HIF-NIS-MSC and NaCl (PE: 9281.9 ± 674.0; WiAUC: 86506.3 ± 7779.2; WiR: 1269.1 ± 106.0; WiPI: 6315.4 ± 342.3; (B–E)). Results are expressed as mean ± SEM in arbitrary units (a.u.; *p < 0.05).

Figure 6. Reduced cell proliferation and blood vessel density in intrahepatic HCC tumors after application of a therapeutic dose of radioiodide in HIF-NIS-MSC-treated mice

Orthotopic HuH7 tumors of HIF-NIS-MSC-treated mice demonstrated significantly reduced tumor cell proliferation (green in (A); 47.6 ± 5.0%; (B)) and blood vessel density (red in (A); 3.0 ± 0.5%; (C)) after radioiodide therapy as compared to animals treated with HIF-NIS-MSCs and NaCl ((A–C); Ki67: 68.7 ± 2.5%; CD31: 6.6 ± 0.5%) or WT-MSCs and ¹³¹I ((A–C); Ki67: 65.2 ± 9.0%; CD31: 6.9 ± 1.6%). Nuclei were counterstained with Hoechst dye. Results are expressed as mean ± SEM (**p < 0.01). Scale bar = 100 μm.