Molecular imaging with bioluminescence and PET reveals viral oncolysis kinetics and tumor viability

Abstract

Viral oncolysis, the destruction of cancer cells by replicating virus, is an experimental cancer therapy that continues to be explored. The treatment paradigm for this therapy involves successive waves of lytic replication in cancer cells. At present, monitoring viral titer at sites of replication requires biopsy. However, repeat serial biopsies are not practically feasible for temporal monitoring of viral replication and tumor response in patients. Molecular imaging provides a noninvasive method to identify intracellular viral gene expression in real time. We imaged viral oncolysis and tumor response to oncolysis sequentially with bioluminescence and positron emission tomography (PET), revealing the kinetics of both processes in tumor xenografts. We demonstrate that virus replication cycles can be identified as successive waves of reporter expression that occur ∼2 days after the initial viral tumor infection peak. These waves correspond to virions that are released following a replication cycle. The viral and cellular kinetics were imaged with Fluc and Rluc bioluminescence reporters plus two 18F-labeled PET reporters FHBG [9-(4-18F-fluoro-3-[hydroxymethyl] butyl) guanine] and FLT (18F-3'-deoxy-3-'fluorothymidine), respectively. Correlative immunohistochemistry on tumor xenograft sections confirmed in vivo results. Our findings show how PET can be used to identify virus replication cycles and for real-time measurements of intratumoral replicating virus levels. This noninvasive imaging approach has potential utility for monitoring viral oncolysis therapy in patients.

Fig 1. Schematic characterization of molecular reporters, cells and virus

A: The dual bioluminescence detection system is based on the enzymatic reaction of Fluc and Rluc where luciferin and coelenterazine serves as substrates. The PET ligand [¹⁸F]FHBG, serves as the substrate for HSV-TK, which phosphorylates and traps it intracellularly. B: Stable cancer cells lines that express Rluc and mCherry can be identified with Rluc bioluminescence, and mCherry fluorescence. We show transformed MC26 cells imaged for Rluc in vitro. C: HSV-Luc expresses the Fluc and TK genes. The replicating virus was identified in MC26 cells with Fluc bioluminescence in vitro (MC26+HSV-Luc), and the signal was comparable to MC26 cells that transiently express Fluc (MC26-Fluc). The virus replicating in MC26 cells was detected with the PET tracer [¹⁸F]FHBG in vitro (MC26+hrR3). The [¹⁸F]FHBG signal was comparable to that of MC26 cells which express the TK gene transiently (MC26sr39TK).

Fig 2. Kinetics of virus and cells in vitro determined with bioluminescence and PET substrates

A: The replication kinetics of HSV-Luc in MC26 cells studied with Fluc bioluminescence. Two peaks (arrows) were observed in the Fluc signal in the high virus titers (1×10⁸, 1×10⁷, and 1×10⁶ pfu) whereas the signal was undetectable in the lower virus titers. Fluc signal is expressed as the mean net intensity/area ± SD. B: Viability of MC26 cells infected the virus. The cell counts decreased when infected with high viral titers (1×10⁸, 1×10⁷, and 1×10⁶ pfu). Cell count is expressed as the mean ± SD. C: [¹⁸F]FHBG uptake in MC26 and Vero cells 24 hrs after HSV-Luc (1×10⁸ pfu) infection. Data are the mean % of ID of [¹⁸F]FHBG uptake/cell ± SD. D: Changes in MC26 and Vero cell concentration once infected with virus. The MC26 cell concentration was markedly reduced compared to cells not infected with virus. Cell concentration is expressed as the mean ± SD.

Fig 3. Dynamics of virus replication in vivo imaged with bioluminescence and PET

A: Kinetics of virus replication in flank tumors studied with Fluc bioluminescence. The Fluc signal expressed 3 peaks (arrows) that decreased in intensity over time. Fluc signal is presented as the mean net intensity/area ± SD. B: Kinetics of virus replication imaged with [¹⁸F]FHBG- PET. The [¹⁸F]FHBG uptake in tumors peaked (arrows) at 6 hrs (0.3 days) in the high virus titers. It decreased in highest virus titer (1×10⁹ pfu), while it peaked again at 3 days in the lower titer (1×10⁵ pfu). The lowest titer (1×10³ pfu) had no signal. The radiolabel uptake is presented as the mean ± SD of the % injected dose/cc.

Fig 4. Tumor growth imaged with Rluc bioluminescence correlates with volumetric measurements obtained with caliper readings

The growth of human tumor xenografts imaged with Rluc bioluminescence. The Rluc signal areas in scans of bilateral MDA-MB-231-Rluc flank tumors increased over time (A). The Rluc signal intensity increased and corresponded with caliper-measured tumor volume (B). A similar expression was observed where the Rluc signal increase correlated with caliper-measured tumor volume in A2058-Rluc (C, D), and HT29-Rluc (E, F) flank tumors. Data are expressed as the mean ± SEM.

Fig 5. Tumor growth and viral replication kinetics imaged with dual bioluminescence

A: Growth of MDA-MB-231-Rluc flank tumors imaged with Rluc. B: Growth inhibition of MDA-MB-231-Rluc tumors undergoing viral oncolysis (1×10⁸ pfu) imaged with Rluc. C: Rluc signal intensities in the control and virus-treated groups. Rluc signal is shown as the mean net intensity/area ± SD. Asterisks indicate P < 0.004. D: Tumor volume measured with caliper readings in the control and virus treated groups. Tumor volume is shown as the mean ± SD. Asterisks indicate P < 0.002. E: Sequential imaging of virus replication with Fluc. F: The Fluc signal intensities peaked (arrows) several times as it decreased over time. Fluc signal is expressed as the mean net intensity/area ± SD.

Fig 6. Dynamics of viral replication and cell proliferation imaged with microPET

A: Localization of sites of virus replication in tumors with [¹⁸F]FHBG-PET. Virus was injected into the left of bilateral flank tumors. [¹⁸F]FHBG uptake was seen in the virus infected tumor but not in the control tumor. B: Time activity curves obtained from dynamic PET scans after [¹⁸F]FHBG injection. [¹⁸F]FHBG accumulated in the virus infected tumor, while it was washed off from the heart, liver and control tumor. C: Identification of proliferating tumor with [¹⁸F]FLT-PET. Photopenia identifies the site of virus replication in the tumor core.

Fig 7. Correlation of bioluminescence, fluorescence and microPET imaging data in vitro

A: Tumor characteristics before virus administration. The tumor is identified with Rluc bioluminescence, mCherry fluorescence, and [¹⁸F]FLT-PET (small arrows). H&E and PCNA are show highly proliferative tumors in vitro. B: Changes in tumor characteristics following lytic virus replication. A destructed tumor core is identified with Rluc, mCherry and [¹⁸F]FLT-PET (arrows), 48 hrs after intra tumor virus (1×10⁸ pfu) injection. The changes are seen on H&E and PCNA where the destructed tumor core (arrow) is surrounded by rim of viable tumor (asterisk). C: Replicating virus is identified with Fluc bioluminescence and [¹⁸F]FHBG-PET. IHC staining for LacZ and HSV-TK (arrow heads) in excised tumors confirm the presence of virus. Apoptosis is seen in the destructed tumor core.