Background
Cancer is a complex genetic disease that results in malignancy of native tissue [
1]. Gene therapy was initially conceived as an approach for treating genetic diseases; however, its scope has expanded to include treatments for cancer [
2]. Currently, over 60 % of ongoing clinical gene therapy trials are designed to treat cancer [
3]. Most cancer gene therapy vectors are invasively administered via intratumoral injections, and the development of non-invasive systemic vectors are greatly warranted. In addition to potentially causing less harm to the patient, it may target both localized and metastatic tumors. Successful cancer gene therapy depends on the development of vectors able to deliver therapeutic genes to tumors specifically and efficiently, while sparing healthy tissues. Animal viruses have the potential to be developed for targeted gene transfer, but require elimination of their native tropism for mammalian cells [
4]. This approach however, is challenging because engineering viruses to target non-natural receptors greatly reduce their efficiency [
5].
We have previously reported the generation of a hybrid bacteriophage vector for targeted systemic cancer gene therapy and molecular imaging [
6]. This construct, termed (adeno-associated virus/phage) AAVP, is a combination of a mammalian transgene cassette flanked by inverted terminal repeats (ITRs) from adeno-associated virus 2 (AAV2) and a fUSE-5 (peptide display) vector derived from the fd bacteriophage genome. Phages have evolved to infect bacteria only and thus, unlike eukaryotic viruses, have no strategies to deliver genes to mammalian cells. The AAVP displays the cyclic RGD4C (CDCRGDCFC) fusion peptide on its pIII capsid protein, allowing homing to and entry via αv integrin receptors such as α
vβ
3 and α
vβ
5, which are expressed by cancer cells or cancer-associated endothelial cells, but not on normal tissues or vasculature [
7,
8]. Moreover, the RGD4C.AAVP construct has allowed for drastic improvements in gene delivery rates to cancer cells over conventional bacteriophage vectors, as shown in numerous
in vitro and
in vivo studies, including a large-scale cancer trial involving pet dogs with natural cancers [
9]. Even though the targeting and efficiency of the RGD4C.AAVP has improved with the modifications applied thus far, there still exists a large room for improvement.
An important consideration is not all limitations are attributable to the vector. Cancer cells in particular, possess macro- and microanatomical barriers that impede gene delivery. Specifically, desmoplastic reactions result in substantial extracellular matrix (ECM) formation around tumors, cancer-associated fibroblasts and infiltrating immune cells [
10]. The resultant high interstitial fluid pressure (IFP), spatial hindrance and inhibition of cell-surface receptors decrease uptake of therapeutics [
11]. As such, depletion of the ECM before administration of therapeutics constitutes a mechanism for tumor priming [
12]. ECM clearance should allow increased transport and binding of RGD4C.AAVP to αv integrin receptors on the tumor cell surface. This principle of transduction has already been demonstrated in multiple studies through the use of ECM-depleting enzymes [
13‐
15].
We sought to test the hypothesis that ECM depletion can increase the tumor transduction efficacy of RGD4C.AAVP vectors by evaluating the effects of co-administering AAVPs after treatment of cancer cells with collagenase, hyaluronidase or a combination of both. Our results show that ECM degradation is a powerful adjuvant in raising transduction rates for phage-guided cancer therapy. These findings were further verified through RGD4C.AAVP-mediated cancer killing by delivering the conditionally toxic Herpes simplex virus-thymine kinase (HSVtk) suicide gene in conjunction with ganciclovir treatment. We have also validated this strategy through a multicellular tumor spheroid model (MCTS). Our results demonstrate how modulating the tumor microenvironment may enhance the efficacy of RGD4C.AAVP and other gene delivery vectors as powerful implements against cancer.
Discussion
In this study, we demonstrate the effectiveness of ECM depletion as a strategy to enhance bacteriophage-guided gene transfer to cancer cells. By removing various ECM constituents using collagenase, hyaluronidase or a combination of both, we were able to substantially increase RGD4C.AAVP internalization and RGD4C.AAVP-mediated gene expression, leading to enhanced therapeutic efficacy. The most dramatic effect was observed when various tumor cell lines were treated with a combination of both enzymes in both 2D and 3D MCTS settings.
The data provide evidence that a combination of spatial and biochemical changes are responsible for the increase in gene transfer efficacy following enzyme-induced ECM depletion. Because the tumor ECM is a primary obstacle that impedes therapy from reaching target cells, when ECM is depleted, the spatial distribution and adsorption of RGD4C.AAVP vectors on the tumor cell surface are improved. Furthermore, ECM constituents may also biochemically impede vector access to tumor cells. Two interesting ECM constituents that had the highest impact in terms of competitive inhibition are collagen and fibronectin; both of these constituents are ligands for α
v integrins which are also the receptor targets of the RGD4C ligand displayed on RGD4C.AAVP and substantially decreased following enzymatic treatment [
26,
27]. Our data also show that hyaluronic acid was also released by enzyme treatment; however, it has not been shown that it may compete or interact with α
v integrin receptors. These data suggest that enhanced gene expression by RGD4C.AAVP vectors after enzyme-mediated ECM depletion is a function of both physical and biochemical changes in the ECM.
When judging the effect of ECM depletion on RGD4C.AAVP gene therapy, it is crucial to consider gene transduction at all stages. Unlike drug molecules, RGD4C.AAVP cannot easily diffuse through the ECM; it is a relatively narrow (6.5 nm width) elongated cylinder reaching up to 1400 nm length [
28] making cellular access difficult in the presence of ECM. Additionally, the ECM is hydrophilic, subsequently increasing the relative volume of the microenvironment surrounding tumor cells.
Our experiments provide the first proof of concept evidence that ECM clearance can be used in phage-guided gene transfer to improve targeted cancer cell killing. As we have shown, collagenase and hyaluronidase are able to substantially increase RGD4C.AAVP-mediated gene transfer efficacy in cancer cells and exert additive effects when both are applied. These effects can be explained, as collagen and hyaluronic acid are separate ECM constituents that may inhibit vector transduction in various ways. However, high levels of enzymes, especially collagenase, seem to be counterproductive to transduction. We suspect this is likely due to physical loss of detached cells during washing, rather than a direct alteration of cellular function due to enzyme-treatment.
Administration of collagenase into tumor-bearing animal models has been proven to decrease the interstitial fluid pressure and improve gene therapy vector accessibility as well as transduction [
29]. Despite FDA-approved use as a locally injected treatment in man [
30], intravenous administration of collagenase remains a particular concern since they do not discriminate between healthy and diseased tissues. As a result, injections may cause tissue damage, as demonstrated by lung necrosis and hemorrhages in murine models [
31]. Moreover, the ECM has been implicated in many cancer pathways including metastasis [
32]. High levels of interstitial matrix metalloproteinases (MMPs), specially collagenase, have been found to correlate with poorer prognosis and increased metastatic potential for some cancers [
33]. Therefore, the clinical translatability of ECM depletion as a strategy to enhance cancer therapy warrants further investigation.
One potential way of circumventing systemic ECM degradation is to use a pharmacological agent to inhibit its production. Losartan is a candidate angiotensin II type I receptor antagonist for hypertension, which has shown anti-fibrotic activity mediated by the Tumor Growth Factor Beta 1 pathway through thrombospondin-1 (TSP-1), leading to inhibition of collagen type I synthesis [
34,
35]. Its combination with oncolytic HSV has been efficacious in murine xenograft models of human cancers, and has shown limited and manageable side effects in patients [
14,
36], offering a better safety profile than intravenous collagenase. Losartan’s mechanism of action inhibits new synthesis of collagen rather than degrading collagen, meaning it may preferably affect more metabolically active cancer cells. Furthermore, losartan has potentially multiple anti-cancer properties such as metastatic suppression via TGF-β1 signaling [
37] that make it especially attractive for use in this field. Thus, drug modulation of collagen may provide an alternative strategy for enhancing gene transfer in cancer. In our study we showed that losartan produced a significant gene expression increase by RGD4C.AAVP vector without affecting its tumor cell specificity.
While we observed changes in multiple ECM constituents, fibronectin is clearly affected after treatment with collagenase. Multiple studies have shown that fibronectin synthesis is positively related to progression and metastases of cancer [
38]. Suppression of growth factor signaling pathways can down-regulate fibronectin synthesis, and is a potential strategy for enhancing penetration of vector/cell-based therapies to solid tumors [
39‐
41]. Direct or indirect inhibition of fibronectin will significantly contribute to enhanced therapy, as it is a physical barrier to tumor cells, in addition to being competitive inhibitor of receptors that RGD4C.AAVP vectors use for tumor cell entry [
26]. Our data suggest that down-regulation of fibronectin, combined with collagen, may significantly account for increased gene expression from the RGD4C.AAVP vector; indeed, this is supported by previously reported
in vivo data for other gene therapy vectors [
42]. Strategies to inhibit fibronectin synthesis or degrade existing tumor-associated fibronectin are crucial in the context of ECM depletion in gene therapy vectors, including RGD4C.AAVP.
Though we have focused on collagen and fibronectin as the main inhibitory ECM constituents, therapies involving hyaluronidase are also used in clinical settings. Intravenous hyaluronidase has already been used as a safe drug for myocardial infarction in man [
43]. Different levels of hyaluronidase have been associated with both cancer growth and suppression with some studies showing tumor-associated production of the enzyme, whereas others have shown that higher concentrations inhibit tumorigenesis. This suggests the potential role of hyaluronidase as an adjuvant for breast cancer chemotherapy [
44]. Though efficacious, it has not found widespread use due to its immunogenicity, as it is of bovine origin [
45]. However, a recently FDA-approved recombinant human hyaluronidase (PEGPH20, Halozyme Therapeutics) is being evaluated in an ongoing trial for patients with advanced solid tumors [
46,
47], suggesting that hyaluronidase can be a clinically feasible anti-cancer adjuvant.
Alternative strategies for decreasing hyaluronic acid production, such as enzyme inhibition have also been considered. Inhibition of hyaluronic acid synthase (
HAS) using antisense oligonucleotides has been used in murine
in vivo models to halt cancer progression [
48]. Another small molecule HAS inhibitor, 4-methylumbelliferone has even shown anticancer effects independent of ECM depletion effects, consequently reducing proliferation and cellular metastatic potential [
49].
We have also demonstrated the feasibility of ECM depletion in 3D MCTS, which are thought to be superior models for
in vivo tumors [
50]. The MCTS model reflects possible resistance to RGD4C.AAVP’s penetration to reach the target cells, due to their metabolic activities, physical characteristics (hypoxia and lack of vascularization) and differential regulation when compared to 2D models [
23]. Because cell monolayers lack the complexities of 3D tumors, they are more susceptible to therapy in vitro. By demonstrating the effectiveness of ECM modulation in enhancing RGD4C.AAVP-meditaed gene transfer in an MCTS setting, we show that our strategy of ECM depletion is relevant for
in vivo gene based therapies.
Our findings also establish the applicability of combining ECM depletion with
HSVtk-GCV therapy. A prominent problem with cancer gene therapy is incomplete eradication of tumors, which frequently translates into cancer recurrence [
51,
52]. ECM depletion allowed significant eradication of tumor cells by RGD4C.AAVP gene therapy, which demonstrates its potential as a therapeutic strategy for future studies.
Material and methods
Cell culture
Rat 9L gliosarcoma cells were a gift from Dr Hrvoje Miletic (University of Bergen, Norway) and human M21 Melanoma cells were provided by Dr David Cheresh (University of California, La Jolla). Human LN229 and SNB19 glioblastoma cells were provided by Dr Nelofer Syed (Imperial College London, UK), LNCaP prostate cancer cell line was a gift from Dr Paul Mintz (Imperial College London) and the mouse C2C12 myoblast cell line was provided by Dr Francesco Muntoni (University College London, UK). The human breast cancer MCF-7 and glioblastoma U87 cell lines were from Cancer Research UK. Cells were sustained in Dulbecco’s Modified Eagle’s Medium (DMEM, Sigma) supplemented with 10 % Fetal Bovine Serum (FBS, Sigma), L-Glutamine (2 mM, Sigma), Penicillin (100 units/ml, Sigma) and Streptomycin (100 μg/ml, Sigma). The C2C12 cells were grown in 20 % FBS. Cells were maintained at 37 °C in a humidified atmosphere supplemented with 5 % CO2. Collagenase from Clostridium histolyticum (Type I, ≥125 CDU/mg, Sigma) and hyaluronidase from bovine testes (Type I-S, 400-1000 units/mg, Sigma) were solubilized in phosphate buffered saline (1x PBS, Sigma).
Immunostaining
Immunostaining experiments were performed as previously described [
28]. Brief, cells were seeded on 18 mm
2 coverslips in 12-well plates and allowed to proliferate until 70-80 % confluence. Cultures were then treated with enzymes as indicated, washed with PBS, and fixed in 4 % PFA in PBS for 10 min at room temperature. Next, cells were quenched with 50 mM Ammonium Chloride (NH
4Cl), washed with PBS and blocked with PBS containing 2 % bovine serum albumin (BSA) for 30 min. Cells were incubated with mouse anti-fibronectin or anti-laminin antibodies (BD Biosciences) at 1:200 dilution overnight at 4 °C, or 1 h at room temperature with a rabbit anti-α
v integrin (diluted 1:50). Then, cells were incubated with secondary AlexaFluor-conjugated antibodies (diluted 1:750) and with 4',6-diamidino-2-phenylindole (DAPI) (diluted 1:2000 in 1 % BSA-PBS) for 1 h at room temperature. Finally, cells were mounted in Prolong Gold antifade mounting medium (Invitrogen) and images were obtained using a fluorescent microscope (Nikon Eclipse TE2000U) and analysed by Openlab imaging software.
Hyaluronic acid (HA) ELISA
HA ELISA was performed as previously described [
53]. Briefly, samples or standard umbilical cord HA (Sigma) at various concentrations (19–10,000 ng/ml) in PBS pH 7.4, were added to 1.5 ml plastic tubes containing biotinylated hyaluronic acid binding protein (HABP, 1:200 in 0.05 M Tris–HCl buffer, pH 8.6). The tubes were incubated at room temperature for 1 h, then samples were added to the microplate, which was pre-coated with umbilical cord HA (100 μl/well of 10 mg/ml), and blocked with 1 % BSA (150 μl/well). The plate was then incubated at room temperature for 1 h. The wells were subsequently washed and 100 μl of peroxidase-conjugated anti-biotin antibody (Sigma) were added. The plate was incubated at room temperature for another hr. The reaction was stopped with 50 μl/well of 4 M sulfuric acid and the absorbance was determined using a microplate reader (Molecular Devices) at 492/690 nm. The concentration of HA in samples was calculated by reference to a standard curve.
Collagen depletion assay (Sirius Red)
Collagenase from Clostridium histolyticum (Type I, ≥125 CDU/mg, Sigma) was solubilized in 1x PBS. Tumor cells were seeded in 12-well plates at a density of 120,000 cells/well and cultured for 72 h until confluence. Cells were then washed with PBS and 500 μl of 0.2 mg/ml collagenase in serum free media was added to the cells and left to incubate at 37 °C for 1 h. Next, both cells and supernatant were processed as different samples in a collagen I detection assay. The assay was conducted based on the use of the Sirius Red dye and by following the manufacturer’s instructions of a Sirius Red Total Collagen Detection Kit (Chondrex, Inc.), with final measurements obtained through a Promega plate reader.
AAVP diffusion assay
200 μl of ECM Gel from Engelbreth-Holm-Swarm murine sarcoma (Sigma) at 2.5 mg/ml or 5.0 mg/ml were added to a 48-well plate and allowed to set at 37 °C. In the meantime fluorescently-tagged RGD4C.AAVP was prepared at a concentration of 5 μg/ml. 50 μl of the RGD4C.AAVP solution were taken up in a pipette tip, which was inserted at a fixed position into the ECM-gel matrix and left to diffuse through the material for 1 h. Measurements were obtained at 1 h post diffusion and at 30 min intervals thereafter.
AAVP production and purification
AAVP bacteriophages were generated as previously reported [
54], by inserting an AAV mammalian viral gene transfer cassette into the fUSE5 plasmid derived from the fd-tet bacteriophage. Reporter or therapeutic genes were carried by the cassette, driven by a Cytomegalovirus (CMV) promoter. Targeting and specificity were achieved through genetic manipulation and display of the RGD4C peptide on the pIII minor coat protein conferring tumor-targeting properties [
54]. AAVPs were produced and purified from the culture supernatant of
Escherichia Coli K91 host bacteria as previously reported [
54]. Isolated AAVPs were sterile-filtered through 0.45 μm filters and titration levels were quantified using a photospectrophotometer (NanoDrop, Thermo Scientific).
Generation of fluorescently-labelled RGD4C.AAVP
RGD4C.AAVP phage particles (1x10
13) were conjugated to fluorescein isothiocyanate (FITC) for 1 h at room temperature in the dark [
55]. Then the phage was purified by three consecutive polyethyleneglycol precipitations, titrated as described above. Phage labelling was confirmed by double-staining using a rabbit anti-phage antibody (Sigma). We have previously confirmed that the fluorescently-labelled phage retains target specificity and cancer cell transduction (unpublished data).
AAVP internalization
Internalization assay was carried out as previously described [
28]. After cell treatment with optimized concentrations of the enzymes, cells were treated with vectors for 4 h at 37 °C, then placed on ice to stop endocytosis and washed 3 times with PBS to remove unbound vectors. Surface bound vectors were removed by trypsinization after which cells were pelleted by centrifugation at 2000 rpm for 5 min and fixed in 4 % paraformaldehyde (PFA) for 10 min at room temperature. Untreated cells were used as negative controls. To detect internalised phage-derived vectors, cells were blocked with 0.1 % saponin in 2 % bovine serum albumin in PBS (BSA-PBS) for 30 min followed by staining with rabbit anti-phage antibody (diluted 1:1000) in 0.1 % saponin in 1 % BSA-PBS for 1 h at room temperature. Cells were pelleted and re-suspended three times in 0.1 % saponin in 1 % BSA-PBS, then incubated with goat anti-rabbit AlexaFluor-647 (diluted 1:500) for 1 h at room temperature. Finally, cells were washed twice with 0.1 % saponin-PBS and re-suspended in PBS before analysis.
Fluorescence-activated cell sorting (FACS) analysis was carried out using a FACscalibur Flow cytometer (BD Biosciences) equipped with an argon-ion laser (488 nm) and red-diode laser (635 nm). The mean fluorescence intensity was measured for at least 10,000 gated cells per triplicate well. Results were analysed using Flojo (TreeStar) software.
AAVP transduction
Transduction of cells with vectors were carried out as we previously described [
54]. Brief, 10,000–20,000 cells/well were seeded in 48-well plates and cultured for 48-72 h to achieve 70–80 % confluence. The cells were washed, then collagenase or hyaluronidase or combination of both enzymes in serum free media were added (as indicated) to the cells for 1 h at 37 °C before the enzyme-containing medium was removed. 110 μl of serum-free media with AAVP (as indicated) was then applied followed by incubation at 37 °C for 4 h, 400 μl of complete medium were subsequently added to the cells and left for 72 h at 37 °C. For transduction using RGD4C.AAVP/
HSVtk, DEAE.DEXTRAN (60 ng/ 1 μg AAVP) was used to enhance gene transduction in line with our previously established protocol and findings [
16]. The cationic polymer does not affect cell viability or specificity of the RGD4C.AAVP vector [
16].
Determination of tumor cell killing in vitro
Cells were seeded in 48-well plates and incubated for 48 h to reach 80 % confluence. Then, cultures were treated with collagenase or hyaluronidase enzymes or combination of both, as indicated. Enzyme solutions were removed from cultures and washed with PBS before treatment with AAVP vectors carrying the Herpes Simplex Virus-thymidine kinase (HSVtk) gene. Ganciclovir (GCV, Sigma) was added to the cells (40 μM) at day 3 post vector transduction and renewed daily. Viable cells were monitored under microscope and cell viability was measured at day three post GCV treatment by using the trypan blue exclusion method (Sigma). When treating the M21-tk cell population stably expressing the HSVtk gene, GCV was added to cells the next day following enzymatic treatments and renewed daily for 3 days.
Reporter gene assays
Quantification of luciferase expression from cells transduced with AAVP-
Luc, was performed as described [
16]. Medium was aspirated before the addition of 110 μl of Glo Lysis buffer, 1x (Promega). 50 μl of the cell lysate was then transferred to a 96-well white opaque microplate (BD Falcon) and mixed with an equal volume of Steady-Glo® luciferase substrate (Promega) for luciferase and read with a Promega Glomax plate reader. Luciferase expression was normalized to 1 μg protein levels determined by the Bradford assay (Sigma) and data presented as relative luminescence units (RLU). GFP expression was visualized using a Nikon Eclipse TE2000-U fluorescence microscope, as were morphological observations of cell monolayers.
Cell viability assays (CellTiter Glo)
Cells were seeded, treated and transduced in line with the cell viability assay for trypan blue exclusion method. However, a bioluminescence assay for cell viability (CellTiter Glo, Promega) was used according to the manufacturer’s instruction. The CellTiter Glo assay is a luciferase-based assay to measure cell viability based on available adenosine triphosphate in the cell lysate.
Multicellular tumor spheroids (MCTS) culture and treatment
MCTS were seeded in 96-well ultra-low attachment surface plates (Corning) from trypsinized monolayer cultures at an optimized cell suspension density [
56] of 0.5 x 10
4 cells/ml in 200 μl. They were then incubated for 48 h at 37 °C prior to any treatment. Media was completely removed before the application of 110 μl of enzyme/serum-free medium mixture and left to incubate at 37 °C for 1 h. The enzyme-supplemented medium was then removed and the spheroids were washed with PBS before transduction with AAVP/
Luc vectors (in 110 μl of serum-free medium) for 4 h at 37 °C. Following transduction, 100 μl of complete media were added to make up a total volume of 200 μl and the cells were left to develop gene expression over a 72 h incubation period at 37 °C.
For MCTS luciferase assays, media was removed and 15 μl of Glo Lysis buffer was added to one spheroid per well. The MCTS were then incubated for 30 min. 10 μl of cell lysate/well from each spheroid per experimental condition were taken from 4 wells and combined to make up 40 μl of total volume. This was mixed in a 1:1 ratio together with the Steady-Glo® luciferase substrate and left 10 min before measurement in the Promega plate reader. The luminescence signal was then normalized to 1 μg protein levels determined by the Bradford assay.
Losartan treatment
Various losartan concentrations of 0 μM, 20 μM, 50 μM, 100 μM, 150 μM and 200 μM were applied in serum-free media to cells and left overnight in an incubator at 37 °C. The following day, losartan-containing media was removed and cells were transduced with vectors as described above. Luciferase assays were carried out at 72 h post-transduction.
Statistical analyses
Data are expressed as mean ± standard error of the mean (s.e.m.) and analysed using student’s t test when comparing two groups. To compare more than two groups, we used one-way ANOVA and post hoc Turkey tests. P values were considered significant when <0.05 and denoted as follows: *p < 0.05, **p < 0.01, ***p < 0.001. Statistical analyses were performed using the GraphPad Prism software (version 5.0).
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
TY carried out the experiments, performed analysis of data, contributed to draft the figures and text of the manuscript, and helped in the design of the study. EL performed experiments, analysed the data, wrote the manuscript and contributed in the design of the project. KS performed experiments, analysed and discussed the data. NS supported the project, discussed the results and helped revise the manuscript. PA wrote and revised the manuscript and figures, and discussed and analysed the data. AH conceived and designed the study, analysed results, supervised the project, edited the manuscript that was read and approved by all authors.
TY: Researcher in the National Nanotechnology Center (NANOTEC), Thailand. EL: BSc student in the Phage Therapy Group.
KS: Senior Postdoctoral Researcher in the Phage Therapy Group.
NS: Research Lecturer in Neuro-oncology and Head of the John Fulcher Molecular Neuro-oncology Research Group at Imperial College London.
PA: Senior PhD student in the Phage Therapy Group and Postgraduate Representative for the Faculty of Medicine at Imperial College London.
AH: Senior Lecturer in Medicine and Head of the Phage Therapy Group at Imperial College London.