Cancer screening by systemic administration of a gene delivery vector encoding tumor-selective secretable biomarker expression

Abstract

Cancer biomarkers facilitate screening and early detection but are known for only a few cancer types. We demonstrated the principle of inducing tumors to secrete a serum biomarker using a systemically administered gene delivery vector that targets tumors for selective expression of an engineered cassette. We exploited tumor-selective replication of a conditionally replicative Herpes simplex virus (HSV) combined with a replication-dependent late viral promoter to achieve tumor-selective biomarker expression as an example gene delivery vector. Virus replication, cytotoxicity and biomarker production were low in quiescent normal human foreskin keratinocytes and high in cancer cells in vitro. Following intravenous injection of virus >90% of tumor-bearing mice exhibited higher levels of biomarker than non-tumor-bearing mice and upon necropsy, we detected virus exclusively in tumors. Our strategy of forcing tumors to secrete a serum biomarker could be useful for cancer screening in high-risk patients, and possibly for monitoring response to therapy. In addition, because oncolytic vectors for tumor specific gene delivery are cytotoxic, they may supplement our screening strategy as a "theragnostic" agent. The cancer screening approach presented in this work introduces a paradigm shift in the utility of gene delivery which we foresee being improved by alternative vectors targeting gene delivery and expression to tumors. Refining this approach will usher a new era for clinical cancer screening that may be implemented in the developed and undeveloped world.

Figure 1. Overall strategy for exogenous cancer biomarkers.

(a) Steps for cancer screening strategy are as follows: 1) Cancer-targeting Herpes Simplex Virus (HSV) is injected systemically. 2) Engineered HSV selectively replicates in tumors while being cleared from healthy non-cancerous tissues. 3) Biomarker is selectively produced in tumors. 4) Blood samples are collected and analyzed for biomarker levels. 5) Serum levels of exogenously delivered biomarker are higher in tumor bearing mice than healthy tumor-free mice. (b) Gene maps for wild-type HSV and novel recombinant rQ-M38G.

Figure 2. Differential sensitivity of cell lines to virus infection.

(a) Gaussia luciferase (GLuc) transgene expression, (b) virus replication, and (c) cytotoxicity following infection of Vero cells and a panel of human tumor cell lines with rQ-M38G (MOI = 0.001).

Figure 3. Detection of tumor presence by serum biomarker levels.

(a) Time course of GLuc expression in renal subcapsule (r.s.c) SK-NEP_Luc-bearing and tumor-free mice injected systemically with 1.2×10⁷ pfu or rQ-M38G. Numbers in legend represent individual mice. (b) GLuc expression and in vivo imaging of SK-NEP_Luc-bearing mice four days following systemic injection of 1.2×10⁷ pfu of rQ-M38G. Serum GLuc levels for mice injected with tumor and control that did not receive tumor injections (bar graph), and in vivo luciferase imaging of mice that received tumor cell injections.

Figure 4. Virus biodistribution.

(a) GFP immunofluoresence in SK-NEP_Luc-bearing mice with and without systemic virus administration. Mice receiving virus (#1 and #2) showed selective GFP expression in tumor, while mice receiving no virus (No Virus Control) showed global GFP-negativity. Scale bar = 15 microns. (b) Biodistribution of virus in mice with SK-NEP_Luc tumors following systemic administration as determined by qPCR for viral genomes.

Figure 5. Sensitivity limits of detection.

a) in vivo imaging of two mice (i and ii) with vastly different SKNEP-Luc tumor burdens. b) Serum GLuc concentration in mouse i, ii and four tumor-free control mice 4 days following systemic infection with 1×10⁷ pfu of rQ-M38G. c) Hematoxylin and eosin macroscopic images of tumor bearing kidneys (T = tumor, scale bars = 1 mm) and insets of microscopic tumor foci in the renal parenchyma of mouse ii (scale bar = 200 µm).