EGFR-targeted nonviral NIS gene transfer for bioimaging and therapy of disseminated colon cancer metastases

Abstract

Liver metastases present a serious problem in the therapy of advanced colorectal cancer (CRC), as more than 20% of patients have distant metastases at the time of diagnosis with less than 5% being cured. Consequently, new therapeutic approaches are of major need together with high-resolution imaging methods that allow highly specific detection of small metastases. The unique combination of reporter and therapy gene function of the sodium iodide symporter (NIS) may represent a promising theranostic strategy for CRC liver metastases allowing non-invasive imaging of functional NIS expression and therapeutic application of ¹³¹I. For targeted NIS gene transfer polymers containing linear polyethylenimine (LPEI), polyethylene glycol (PEG) and the epidermal growth factor receptor (EGFR)-specific ligand GE11 were complexed with human NIS DNA (LPEI-PEG-GE11/NIS). Tumor specificity and transduction efficiency were examined in high EGFR-expressing LS174T metastases by non-invasive imaging using ¹⁸F-tetrafluoroborate (¹⁸F-TFB) as novel NIS PET tracer. Mice that were injected with LPEI-PEG-GE11/NIS 48 h before ¹⁸F-TFB application showed high tumoral levels (4.8±0.6% of injected dose) of NIS-mediated radionuclide uptake in comparison to low levels detected in mice that received untargeted control polyplexes. Three cycles of intravenous injection of EGFR-targeted NIS polyplexes followed by therapeutic application of 55.5 MBq ¹³¹I resulted in marked delay in metastases spread, which was associated with improved animal survival. In conclusion, these preclinical data confirm the enormous potential of EGFR-targeted synthetic polymers for systemic NIS gene delivery in an advanced multifocal CRC liver metastases model and open the exciting prospect of NIS-mediated radionuclide therapy in metastatic disease.

Keywords: colon cancer metastases; gene therapy; nonviral EGFR-targeted gene transfer; sodium iodide symporter; theranostic application.

Figure 1. NIS-mediated ¹²⁵I uptake studies in vitro

Incubation of EGFR expressing (A) LS174T cells with polyplexes at a N/P ratio of 6 resulted in high transduction efficiency and EGFR-specificity of NIS-encoding LPEI-PEG-GE11/NIS (n=4) polyplexes compared to polyplexes without ligand (LPEI-PEG-Cys/NIS; n=4) and empty polymers (LPEI-PEG-GE11/HBG; n=4) (B). Pre-treatment with the NIS-specific inhibitor perchlorate resulted in reduced iodide uptake confirming NIS-specificity. Cell viability was not affected by polyplex-mediated NIS gene transfer (C) (^*p≤0.05; ^**p≤0.01; ^*** p≤0.001; n/s = not significant). Results are reported as mean ± SEM.

Figure 2. PET imaging studies after systemic NIS gene transfer in vivo

Hepatic colon cancer metastases (A, B) showed high EGFR expression (C) in contrast to normal liver tissue. Significant tumor-specific accumulation of the NIS PET tracer ¹⁸F-TFB was detected in metastases of animals that received LPEI-PEG-GE11/NIS (D; n=5) in contrast to mice injected with control vectors (E; n=3). To verify NIS-mediated tracer uptake, LPEI-PEG-GE11/NIS-injected mice received the competitive NIS inhibitor perchlorate (NaClO4) 30 min before tracer administration (F; n=2), which blocked NIS-mediated tracer accumulation in metastases and tissues that exhibit endogenous NIS expression. Serial scanning revealed an accumulated dose of 4.8±0.6% of injected dose (G) (^*p≤0.05). Results are reported as mean ± SEM. (M=metastases; S=nasal secretion; T=thyroid; MG=mammary gland; St=stomach; B=bladder).

Figure 3. NIS expression in metastatic tumor sections

qPCR revealed high NIS mRNA expression in metastatic areas of mice injected with LPEI-PEG-GE11/NIS (n=5) compared to untreated tumors (n=2) and low NIS mRNA expression levels were observed in tumors of mice that were treated with the control vector LPEI-PEG-Cys/NIS (n=3) (A). Results are reported as mean ± SEM. Immunohistochemical staining confirmed NIS expression specifically in metastatic tissue of LPEI-PEG-GE11/NIS treated mice. NIS-specific immunostaining was visible only in metastatic tissue and occurred in clusters. No NIS expressing cells were found in normal liver tissue (B).

Figure 4. Radioiodide therapy studies

Metastases-bearing mice were treated with three cycles of either LPEI-PEG-GE11/NIS followed by ¹³¹I (n=9) or saline (n=9) 48 h later or received saline only (n=9). Hepatic tumor load (A) and animal survival (Kaplan-Meier-Plot) (B) of the three different treatment groups were compared (Student's t-test: ^*p≤0.05; ^**p≤0.01; ^*** p≤0.001; Mann-Whitney U test: ^#p≤0.05; ^##p≤0.01; ^###p≤0.001). Results are either reported as mean ± SEM for tumor volumes or in percent for survival plots.

Figure 5. Hepatic contrast enhanced ultrasound (CEUS) measurement

CEUS imaging of the entire liver on day 5 after therapy start revealed strong differences in contrast agent uptake and perfusion between therapy (A; LPEI-PEG-GE11/NIS + ¹³¹I) and control groups (B; LPEI-PEG-GE11/NIS + NaCl and C; NaCl + NaCl). Due to hypovascularization of metastases and necrotic areas metastases (M) appear dark. Animals treated with ¹³¹I showed an overall enhanced contrast agent signal and a higher maximum uptake (D), in comparison to both control groups, which exhibited higher hepatic metastases load, reduced amount of healthy hepatic tissue and thus overall decreased perfusion levels.

Figure 6. Immunofluorescence staining for tumor cell proliferation and blood vessel density

Frozen metastatic tissue sections of the three different treatment groups exhibited reduced cell proliferation (green; Ki67) and blood vessel density (red; CD31) in animals treated with LPEI-PEG-GE11/NIS + ¹³¹I (n=9; A) compared to control groups that received LPEI-PEG-GE11/NIS + NaCl (n=9; B) or NaCl + NaCl (n=9; C). Differences in Ki67 and CD31 staining between metastatic tissue and normal liver tissue can be seen in (D). Higher cell proliferation and increased pattern of vascularization of tissue surrounding metastases along with low vascularization inside metastases was found in both control groups (E, F) (^*p≤0.05). Results are reported as mean ± SEM.