A Novel Listeria monocytogenes -Based DNA Delivery System for Cancer Gene Therapy

Abstract

Bacteria-mediated transfer of plasmid DNA to mammalian cells (bactofection) has been shown to have significant potential as an approach to express heterologous proteins in various cell types. This is achieved through entry of the entire bacterium into cells, followed by release of plasmid DNA. In a murine model, we show that Listeria monocytogenes can invade and spread in tumors, and establish the use of Listeria to deliver genes to tumors in vivo. A novel approach to vector lysis and release of plasmid DNA through antibiotic administration was developed. Ampicillin administration facilitated both plasmid transfer and safety control of vector. To further improve on the gene delivery system, we selected a Listeria monocytogenes derivative that is more sensitive to ampicillin, and less pathogenic than the wild-type strain. Incorporation of a eukaryotic-transcribed lysin cassette in the plasmid further increased bacterial lysis. Successful gene delivery of firefly luciferase to growing tumors in murine models and to patient breast tumor samples ex vivo was achieved. The model described encompasses a three-phase treatment regimen, involving (1) intratumoral administration of vector followed by a period of vector spread, (2) systemic ampicillin administration to induce vector lysis and plasmid transfer, and (3) systemic administration of combined moxifloxacin and ampicillin to eliminate systemic vector. For the first time, our results reveal the potential of Listeria monocytogenes for in vivo gene delivery.

Introduction

G ene therapy is a realistic prospect for the treatment of cancer and other diseases, and involves the delivery of genetic information to a cell to facilitate the production of a therapeutic protein. However, there are still significant difficulties to be overcome before achieving an efficient and safe gene medicine, most significantly overcoming biological barriers to cell delivery. Current gene delivery vectors are broadly grouped as “viral” or “nonviral” (plasmid based). Viral vector clinical trials have highlighted safety fears concerning vector integration (which has resulted in cases of leukemia) and fatal immune responses and have led to increased interest in nonviral systems (Thomas et al., 2003; Yamamoto and Curiel, 2005). Nonviral systems are less immunogenic and toxicity is generally low. Plasmid DNA has greater potential for delivery of larger genetic units, and large-scale production is relatively easy. However, the potential for efficiency of current nonviral approaches is limited. The transfer of naked DNA is typically an inefficient process, with cell and tissue damage caused by administration of physical and chemical modalities (Boulaiz et al., 2005). It is feasible that the optimal delivery modality for various diseases and anatomical locations and/or tissue types will differ.

Bacteria-mediated transfer of plasmid DNA into mammalian cells (bactofection) has been shown to have potential as a potent approach to express heterologous proteins in various cell types (Vassaux et al., 2006). Delivery of genetic material is achieved through entry of the entire bacterium into target cells. Various bacterial species including Salmonella, Clostridium, and Escherichia coli have been examined (Dietrich et al., 1998; Hense et al., 2001; Grillot-Courvalin et al., 2002; Pilgrim et al., 2003; Schoen et al., 2005; Loeffler et al., 2006; Vassaux et al., 2006). The majority of this work has focused on the use of bacteria in vaccine settings, to deliver antigen-expressing plasmid to antigen-presenting cells. Bacterial gene delivery presents several potential advantages over current vectors. For example, viral vector particle manufacture is an extremely cumbersome, time-consuming, and expensive process. Traditional Good Manufacturing Practice (GMP)-grade naked plasmid DNA isolation is significantly less expensive, but often requires combination with expensive chemical vector (liposomes, cationic polymers, etc.) or delivery equipment (electroporation, etc.). In contrast, large-scale GMP production of bacterial vectors is a well-established industrial process.

Listeria monocytogenes (LMO) is particularly efficient in mediating internalization into host cells through expression of internalin proteins that bind the host cell surface and trigger entry by a zipper-like mechanism. Once inside cells, the bacterium produces specific virulence factors that lyse the vacuolar membrane and allow escape into the cytoplasm (Portnoy et al., 1992). LMO is capable of actin-based motility through production of the virulence factor ActA and cell-to-cell spread without an extracellular phase (Pilgrim et al., 2003). One of the advantages of employing a gram-positive bacterium such as LMO is that genomes of various LMO strains have been sequenced and many well-defined virulence-attenuated mutants have been constructed (Glaser et al., 2001; Loeffler et al., 2006). Taking these advantages into account, it is therefore not surprising that LMO has been employed in several studies for vaccine delivery (Dietrich et al., 1999, 2001; Schoen et al., 2008), including ongoing human clinical trials (Maciag et al., 2009). Thus, LMO is a promising platform for development as a vector for gene therapy, and in vitro proof of concept employing LMO as a gene delivery vector has been established (Dietrich et al., 1998, 2000; Grillot-Courvalin et al., 2002).

However, Listeria-mediated plasmid delivery to tumors in vivo has not been reported. Furthermore, plasmid release from bacterial cells frequently goes unaddressed in described bactofection strategies and reliance on intracellular spontaneous or host-mediated lysis and plasmid release results in suboptimal plasmid transfer. We aimed to address this failing by development of a novel antibiotic-mediated lysis strategy for in vivo gene delivery to tumors.

Materials and Methods

Bacterial strains, media, and growth conditions

Listeria monocytogenes EGDe was cultured at 37°C under aerobic conditions in brain heart infusion (BHI) broth (Oxoid, Basingstoke, UK). Chemically competent E. coli TOP10 (Invitrogen, Carlsbad, CA) was used as a host for cloning. Escherichia coli was propagated in Luria–Bertani (LB) broth (Sambrook et al., 1989) at 37°C under aerobic conditions. If required, cultures were supplemented with a 5-μg/ml concentration of erythromycin or chloramphenicol (Listeria), or with erythromycin (300 μg/ml; E. coli).

DNA manipulations

Oligonucleotides were purchased from MWG Biotech (Ebersberg, Germany) (Table 1). For cloning purposes KOD hot start DNA polymerase (Merck, Nottingham, UK) was used, whereas for screening purposes, Taq polymerase (Bioline, London, UK) was used. Restriction endonucleases and T4 DNA ligase were purchased from Roche (Mannheim, Germany). Plasmid purification kits, PCR purification kits, and gel extraction kits were purchased from Qiagen (Crawley, UK).

Table 1.

Primers Used in Study

Oligonucleotide	Sequence (5′→3′) ^a	Target
JP504	AGCAGCGGATCC TGCGGCCGCACGCGTGGAGCTAG	CMV-luciferase
JP505	AGCAGCGGTACC TTCCTGCGGCCGCTCGGTCCGCACGTGG	CMV-luciferase
JP514	AGCAGCGGATCC TGACCGACGCCGACCAAC	bGH-poly-A
JP522	AGCAGCGAATTC CCTCCTATAGTGAGTCGTATTATACTATGC	IRES
JP523	AGCAGCGAATTC ATGACAAGTTATTATTATAGTAGAAGTTTGG	lysA
JP524	AGCAGCGGATCC CTAAATCTTTTTAACAAACTTCGTGTTAGC	lysA
JP528	CTCGAGGCAAGCTTCTGCAGG	pFXluc
JP529	AGCTTTTGTTCCCTTTAGTGAGGG	pFXluc
JP530	AGCAGCCCCGGG TGAGTGTCCCGTCCCTGGGGCTCC	IRES
JP531	AGCAGCCCCGGG TTCGATAGAGAAATGTTCTGGCACC	bGH-poly-A
JP533	TTTACACGAAATTGCTTCTGGTGGC	pFXluc
JP580	GAATTCAATCGATGTTCGAATCCCAATTC	pFXluc-IRES-lysA
JP581	GTGTCCCGTCCCTGGGGCTC	pFXluc-IRES-lysA

Italic sequence represents clamp; underlined sequence indicates restriction site.

Electrotransformation of Listeria monocytogenes

Electrocompetent Listeria monocytogenes EGDe was prepared as described previously (Monk et al., 2008). As the ΔyycH derivative of Listeria monocytogenes lyses in the presence of ampicillin and/or high sucrose concentrations, we applied the following protocol. BHI containing 20 mM d,l-threonine (Sigma-Aldrich, St. Louis, MO) was inoculated (1%, v/v) from a freshly grown overnight culture of ΔyycH in BHI containing 5 mM d,l-threonine. Cells were harvested at an optical density at 600 nm of 1.5 by centrifugation at 4°C (20 min, 3000 × g). The pellet was washed twice in ice-cold water containing 10% (v/v) glycerol (pH 7), and resuspended in 1/100th the culture volume in the same solution. For transformations, DNA was precipitated with Pellet Paint (Merck) and dissolved in 5 μl of sterile distilled water. DNA–cell suspensions were electroporated at 10 kV/cm, 400 Ω, and 25 μF in a GenePulser (Bio-Rad, Hercules, CA). On transformation, bacteria were recovered in 1 ml of BHI after incubation for 3 hr at 37°C. Bacteria were plated on selective BHI agar plates. Using the replicating vector pFX3EM, a maximum of 10² colony-forming units (CFU)/μg could be obtained.

Construction of reporter plasmids

Reporter plasmids were constructed with a derivative of pFX3 (Xu et al., 1991) in which the choramphenicol marker was replaced with the erythromycin marker (I. Monk, unpublished data). The plasmid is a derivative of a lactococcal plasmid, and can replicate in both gram-positive and gram-negative bacteria. To generate pFXluc-IRES-lysA, first, the internal ribosomal entry site (IRES) region and the bovine growth hormone (bGH) poly(A) region were amplified (primer pairs JP530–JP522 and JP514–JP531, respectively), using pGT62LacZ (Invivogen, Toulouse, France) as the template. The gene encoding LysA was amplified (primer pair JP523–JP524) with genomic DNA of L. monocytogenes EGDe as the template. The resultant replicons corresponding to IRES and bGH poly(A) were subjected to a single digest with EcoRI and BamHI, respectively, whereas the PCR product corresponding to lysA was digested with both enzymes. The digested PCR products were purified, and ligated for 16 hr at 4°C at a 1:1:1 molar ratio. Subsequently, the ligation mix was run on a 1% agarose gel, and a band of 2.3 kb, corresponding to the size of IRES-lysA-(poly-A), was excised and purified. The gel-purified fragment was diluted to 20 ng/μl, and served as a template for PCR using primer pair JP530–JP531. Each oligonucleotide had a phosphate incorporated at the 5′ end.

The backbone of pFXluc was amplified with primer pair JP528–JP529. Oligonucleotides JP528 and JP529 start upstream and downstream, respectively, of the poly(A) region in pFXluc. After amplification, the PCR product was treated with DpnI, purified, and ligated for 16 hr at 4°C with the IRES-lysA-(poly-A) cassette at a molar ratio of 1:3 [pFXluc:IRES-lysA-(poly-A)]. The orientation of IRES-lysA-(poly-A) was determined by PCR with oligonucleotides JP533 and JP522.

Assessment of ampicillin sensitivity of ΔyycH

Overnight cultures of L. monocytogenes EGDe and its ΔyycH derivative were washed once in phosphate-buffered saline (PBS) and subcultured in a 1/100th volume of BHI broth and, when applicable, supplemented with ampicillin at 50, 100, or 150 ng/ml. Growth characteristics were determined in a 96-well plate in a SpectraMax M2 plate reader (Molecular Devices, Sunnyvale, CA) for 23 hr. Experiments were performed in triplicate, in three independent assays.

Invasion assays

The human colonic cell line Caco-2 (HTB-37; American Type Culture Collection [ATCC]) and the human mammary tumor cell line MCF-7 (HTB-22; ATCC) were maintained according to the recommendations of the ATCC. Invasion assays were performed as described previously (Monk et al., 2008), using a multiplicity of infection (MOI) of 50:1, with the following modifications. Tissue culture cells were not supplemented with antibiotics during propagation; invasion and gentamicin treatment were performed for 1.5 hr instead of 1 hr. Host cells were lysed by resuspension in ice-cold water, and bacteria were subsequently enumerated by plate count. Percentage invasion was expressed as follows: (CFU internalized/CFU added) × 100%. Experiments were perfor-med in triplicate on three independent days.

In vitro bactofection assays

Invasion assays were performed as described previously, using an MOI of 50:1. For gene delivery assays, in 24-well plates, MCF-7 cells were seeded at 1 × 10⁵ cells per well and grown until confluence. Invasion assays were performed, and after gentamicin treatment, the medium was replaced with Dulbecco's modified Eagle's medium (DMEM; 10% [v/v] bovine serum) only, or with ampicillin (20, 100, or 500 μg/ml). Ampicillin treatment was for 24 or 48 hr, and was refreshed after 12 hr. To determine the internalized CFU, infected monolayers were washed three times with PBS, and cells were lysed in 1 ml of ice-cold water. Internalized CFU was determined by plate counts. To assess firefly luciferase expression, monolayers were washed three times with PBS, and resuspended in 250 μl of reporter lysis buffer (Promega, Madison, WI). The samples were subjected to one freeze–thaw cycle to obtain complete lysis. Lysis solutions were centrifuged (1 min at 13,000 × g), and the supernatants were transferred to a 24-well plate. Levels of firefly luciferase protein expression were determined by adding an equal volume of the substrate luciferin. Luminescence was measured in relative light units (RLU) (in photons sec^–1) in a Xenogen IVIS 100 (Xenogen/Caliper Life Sciences, Alameda, CA).

Expression analysis of pFXluc-IRES-lysA

To ensure that the genes flanking the IRES region were expressed, we transfected MCF-7 cells with appropriate plasmid DNA. MCF-7 cells were seeded at 1 × 10⁵ cells per well in a six-well plate. Transfections were performed at 70% confluence with N-[1-(2,3-dioleoyloxy)propyl]-N,N,N-trimethylammonium methyl sulfate (DOTAP) reagent (Roche) according to the manufacturer's recommendations. After transfection, cells were incubated for 24 hr to allow for gene expression. Total RNA was extracted as described previously (Hegde et al., 2000), and subjected to DNase treatment (Ambion, Cambridgeshire, UK). An Omniscript RT kit (Qiagen) was used to prepare cDNA, using gene-specific primers, according to the manufacturer's recommendations. The resultant cDNA was subjected to PCR to assess the expression of the firefly luciferase gene and lysA (Table 2).

Table 2.

Strains and Plasmids Used in Study

Strain or plasmid	Characteristic(s)	Source
Strain
Escherichia coli TOP10	Cloning host	Invitrogen
Listeria monocytogenes EGDe	Wild type of serotype 1/2a, with completed genome sequence	Glaser et al. (2001)
LMOlux	EGDe with pPL2lux-P_HELP integrated in chromosome, Cm^r	Riedel et al. (2007)
LMOluc	EGDe harboring pFXluc, Em^r	This study
LMOlys	EGDe harboring pFXluc-IRES-lysA, Em^r	This study
ΔLMO	EGDe derivative in which yycH gene is deleted	Monk (2009)
ΔLMOluc	ΔLMO harboring pFXluc, Em^r	This study
ΔLMOlys	ΔLMO harboring pFXluc-IRES-lysA, Em^r	This study
Plasmid
pFX3EM	Derivative of pFX3 in which Cm^r marker was replaced by Em^r marker	Monk (2009)
pFXluc	pFX3EM harboring firefly luciferase gene under control of CMV promotor, Em^r	This study
pFXluc-IRES-lysA	Derivative of pFXluc in which IRES region and lysA gene are cloned downstream of firefly luciferase gene, Em^r	This study

Abbreviations: Cm^r, chloramphenicol resistance; CMV, cytomegalovirus; Em^r, erythromycin resistance; IRES, internal ribosomal entry site.

Animal studies

Tumor induction

All in vivo experiments were approved by the ethics committee of University College Cork (Cork, Ireland). Athymic MF1 nu/nu and BALB/c mice were obtained from Harlan Laboratories (Oxfordshire, UK). Before experiments, the mice were afforded an adaptation period of at least 14 days. Healthy female mice, weighing 16–22 g and 6–8 weeks of age, were included in experiments. For routine tumor induction, 2 × 10⁶ MCF-7 cells suspended in 200 μl of serum-free DMEM were injected subcutaneously into the flank of athymic MF1 nu/nu mice.

In vivo gene delivery

Mice were randomly divided into experimental groups and subjected to specific experimental protocols. For tumor experiments, mice were treated when tumors reached approximately 100 mm³ in volume (major diameter, 5–7 mm). When applicable, tumors were injected with LMO as follows: 100 μl of LMO suspension (10⁶ CFU) was administered intratumorally by injection into the center of the tumor tissue, using a 29-gauge needle. For LMO-mediated transfection, after 6 days, if applicable, a single dose of 100 μl of ampicillin (100 mg/ml) was administered for three consecutive days. After euthanasia, assessment of CFU in liver and spleen was performed as described previously (Rea et al., 2005). In a similar manner CFU in tumor tissue was determined.

For plasmid delivery by electroporation, a custom-designed applicator with two needles 4 mm apart was used, with both needles placed through the skin central to the tissue. Tissue was injected between electrode needles with 12.5 μg of plasmid DNA in sterile injectable saline in an injection volume of 50 μl. After 80 sec, square-wave pulses (1200 V/cm, 100 μsec × 1 and 120 V/cm, 20 msec; eight pulses) were administered in sequence with a custom-designed pulse generator (Cliniporator; IGEA, Carpi, Italy).

Whole body imaging

In vivo luciferase activity from tissues was analyzed at set time points posttransfection as follows: 80 μl of 30-mg/ml firefly luciferin (Biosynth, Basil, Switzerland) was injected intraperitoneally or intratumorally. Mice were anesthetized by intraperitoneal administration of 200 μg of xylazine and 2 mg of ketamine. Ten min after luciferin injection, live anesthetized mice were imaged for 3 min at high sensitivity with an intensified charge-coupled device (CCD) camera (IVIS imaging system; Xenogen/Caliper Life Sciences). All data analysis was carried out with the Living Image 2.5 software package (Xenogen/Caliper Life Sciences).

In vivo infection study

In vivo bacterial survival was determined by inoculating two groups of five (8- to 12-week-old) BALB/c mice intraperitoneally with either wild-type L. monocytogenes EDGe or the mutant ΔyycH strain. To prepare cultures for inoculation both strains were grown to log phase, determined as an OD₆₀₀ nm reading of between 0.8 and 1.0. The cultures were centrifuged, washed once with PBS, resuspended, and subsequently diluted in PBS to a final concentration of 3 × 10⁵ CFU in 200 μl of PBS. The mice were killed on day 3 postinfection, and bacteria were harvested from the spleen or liver by homogenization of organs in PBS. Total bacterial counts were determined by serial dilution of organ homogenates on BHI agar followed by overnight incubation at 37°C.

Histological analyses

The Gram stain protocol is a modified version of the Hucker Conn stain (Hucker and Conn, 1928). Briefly, paraffin-embedded sections were hydrated and flooded for 3 min with crystal violet and Gram's iodine, respectively. Slides were thoroughly rinsed between wash steps. Tissue sections were differentiated by submersion of the slide in a solution consisting of equal parts absolute alcohol and acetone before rinsing with water and counterstaining for 2 min in 1% (v/v) neutral red. After drying, dehydrated samples were examined for the presence of differentially stained bacterial cells.

Flow cytometric analysis

Tumor tissue was excised and finely minced with a scalpel. Tissue was chemically dissociated in DMEM containing collagenase I (300 U/ml; Sigma-Aldrich) and Dispase II (1 mg/ml; Roche) for 40 min at 37°C and then applied to a cell strainer (70-μm mesh size; BD Biosciences, Oxford, UK). Cells were washed in PBS, fixed with 70% ethanol, and rinsed with permeabilization buffer (PBS, FBS, 0.1% Triton X-100; Sigma-Aldrich). Fluorescence-activated cell sorting (FACS) analysis for luciferase was performed with a luciferase-specific monoclonal antibody (Luci17; Abcam, Cambridge, UK) and a relevant secondary antibody (cyanine-5; Abcam). For identification of cell types, fluorescein isothiocyanate (FITC)-conjugated anti-mouse CD68 (clone FA-11; AbD Serotec), phycoerythrin (PE)-conjugated anti-pan-cytokeratin (clone C-11; Abcam), and relevant isotype controls as per the manufacturer were used. Analysis was preformed on a FACSDiva (BD Biosciences) and analyzed with BD FACSDiva software 6.0 (BD Biosciences).

Ex vivo bactofection

Freshly resected breast tumor samples were obtained from consenting patients of the South Infirmary Victoria University Hospital (Cork, Ireland). Tissue samples were maintained in 6-well plates containing DMEM supplemented with erythromycin (10 μg/ml) and 10% (v/v) fetal calf serum. Tissue was divided into three portions of approximately 2 mm in thickness. When appropriate, L. monocytogenes ΔyycH harboring pFXluc-IRES-lysA (ΔLMOlys) was injected in a total of 100 μl into three locations of the tissue (approximately 33 μl per site) corresponding to a total administration of 10⁶ CFU per tissue sample. When applicable, 100 μl of ampicillin (200 mg/ml) was administered by injection 24 hr postinfection. Forty-eight hours after ampicillin treatment, 100 μl of luciferin (3 mg/ml) was administered per sample. Luminescence was determined as described previously.

Statistical analysis

The Student t test was employed to investigate statistical differences. Samples with p values less than 0.05 were considered statistically significant.

Results

Invasion, replication, and mobility of LMO in MCF-7 cells

We examined the ability of LMO to invade MCF-7 human breast adenocarcinoma cells in vitro. The Caco-2 human epithelial colorectal adenocarcinoma cell line is commonly used to study listerial infection, and was used for comparison (Braun et al., 1998; Van Langendonck et al., 1998). In an in vitro invasion model we showed that LMO invades MCF-7 cells at levels comparable (p = 0.28) to Caco-2 (Fig. 1), indicating that breast tumor cells may be a target for Listeria-mediated bactofection.

FIG. 1.

Cell invasion and replication and intratumoral mobility of Listeria monocytogenes (LMO) in MCF-7 cells. (A) Invasion of LMO in the breast adenocarcinoma cell line MCF-7 was comparable to that of the colonic adenocarcinomic cell line Caco-2 (p = 0.28). The results shown are the averages of three independent experiments wherein each experiment was performed in triplicate. Error bars represent the standard error of the mean (SEM). (B) LMO replicates intratumorally. Athymic MF1 nu/nu mice harboring subcutaneous MCF-7 tumors were injected intratumorally with 10⁶ CFU of LMOlux. Over a period of 8 days a 50% increase in intratumoral luminescence was observed, confirming intratumoral replication of LMO. (C) Mobility of LMO in tumor tissue is ActA dependent. LMO wild type and LMOΔactA were injected intratumorally as described previously. On day 8 postinfection tissue sections were Gram-stained, and examined microscopically at a total original magnification of × 400. LMO stains blue, and bacteria are indicated with arrows.

Although it is well established that LMO can replicate and mobilize itself in human colonic cells in vitro (Mounier et al., 1990; Dramsi et al., 1993; Braun et al., 1998), this has not been reported in tumor tissue. We therefore first addressed whether LMO can replicate intratumorally. To this end, a lux-tagged strain of LMO (LMOlux) previously constructed in our laboratory was used (Riedel et al., 2007). Athymic MF1 nu/nu mice harboring MCF-7 tumors were injected intratumorally with LMOlux, and luminescence was measured on days 1 and 8 postinjection. A 50% increase in luminescence was observed on day 8, confirming replication of LMO (Fig. 1). We also confirmed intratumoral growth of LMO by CFU counts. On day 1 we recovered 8.9 × 10⁵ CFU, and on day 8 we recovered 1.5 × 10⁶ CFU, corresponding to an increase of 68%.

We hypothesized that the LMO-specific trait of cell-to-cell spread would provide enhanced bactofection in tumors as it may broaden the tissue area of DNA delivery postadministration. It is known that the protein ActA mediates cell-to-cell spread of LMO (Robbins et al., 1999; Portnoy et al., 2002), but the ability to (and mechanism of ) spread within tumors has not been investigated. Whereas intratumoral growth can be determined with lux-tagged bacteria, cell-to-cell spread can likely not. Observed luminescence adjacent to the site of injection can represent signal spread from the area of infection rather than cell-to-cell spread. Therefore, we used a microscopic approach to determine bacterial spread. LMO wild type and LMOΔactA (in the EGDe background; a gift from P. Cossart, Unité des Interactions Bactéries Cellules, Institut Pasteur, Paris, France) were intratumorally administered to MCF-7 tumors in vivo and, 8 days later, Gram-stained tissue sections (n = 3) were examined microscopically. Whereas LMO wild-type cells were observed throughout broad areas of tumor tissue, LMOΔactA was confined to the region of injection (Fig. 1), confirming the ability of LMO to spread intratumorally in an ActA-dependent manner.

Ampicillin induces plasmid release from intracellular LMO

To assess DNA delivery in MCF-7 cells, the eukaryotic cytomegalovirus (CMV) early promoter and firefly luciferase gene sequences were cloned in pFX3EM, yielding the reporter plasmid pFXluc (Table 2). Luminescence detected in subsequent studies with this construct was taken to be mammalian cell specific, and was confirmed with no luminescence detected from bacterial cells harboring pFXluc (data not shown). We examined antibiotic-mediated bacterial lysis in the context of intracellular plasmid release. The β-lactam antibiotic ampicillin is known to induce lysis in gram-positive bacteria (Hopkins, 2003). A standard in vitro invasion assay was performed with LMO harboring the control plasmid pFX3EM (LMOctrl) or pFXluc (LMOluc). On invasion, we treated half the infected MCF-7 monolayers for 24 hr with ampicillin to induce plasmid release from LMO, and to allow for expression of firefly luciferase by the eukaryotic cells. Treatment with ampicillin resulted in a 4-log reduction of viable LMO (of both LMOctrl and LMOluc) (Fig. 2). To prove that ampicillin treatment results in plasmid delivery, lysates of the infected monolayers were prepared and levels of luminescence were assessed. A 6-fold increase in luminescence was observed when LMOluc was incubated with ampicillin (Fig. 2). Taken together, these data demonstrate that ampicillin can facilitate reporter gene delivery in vitro.

FIG. 2.

In vitro bactofection mediated by ampicillin. Results of bactofection assays with LMO harboring pFX3EM (LMOctrl) or pFXluc (LMOluc) are shown. Solid columns represent recovered bacteria (primary y axis). Red data points represent bioluminescence intensity (secondary y axis). Representative images of luminescence readings of the corresponding group are displayed below. Data shown are the averages of three independent experiments, each performed in duplicate. Color images available online at www.liebertonline.com/hum.

To examine this strategy in vivo, subcutaneous MCF-7 tumors were injected with LMOluc, and 6 days later a 3-day course of ampicillin was commenced before IVIS imaging. No tumor-derived luminescence was detected in vivo at levels greater than background. When tumors were excised and imaged ex vivo, low-level luminescence was detected in 40% of tumors (data not shown). These data suggested that improvement of plasmid delivery was required for in vivo DNA delivery.

Improved bacterial lysis via incorporation of CMV-lysin cassette

PLY118 is a bacteriophage-derived lysin that has been shown to lyse LMO efficiently (Loessner et al., 1995; Gaeng et al., 2000). We searched the genome sequence of LMO EGDe for a functional homolog of PLY118, and identified a gene annotated as l-alanyl-d-glutamate peptidase (lysA). The predicted protein sequence shares 96 and 98% identity and similarity, respectively, with PLY118. A derivative of pFXluc was constructed in which an internal ribosomal entry site (IRES) and the lysA gene of LMO EGDe were cloned downstream of the firefly luciferase gene, generating pFXluc-IRES-lysA (Fig. 3). The IRES region allows for internal initiation of translation of the mRNA (Pelletier et al., 1988), thereby placing both the firefly luciferase and lysA genes under the control of the CMV promoter. In doing so, the expression of lysin is confined to mammalian cells, which is hypothesized to amplify further lysis of intracellular bacteria, subsequently resulting in further release of plasmid encoding lysA. Expression of both the firefly luciferase and lysA genes was confirmed by reverse transcriptase-PCR on RNA extracted from pFXluc-IRES-lysA Lipofectamine-transfected MCF-7 cells (data not shown).

FIG. 3.

Overview of construction of the P_CMV-luc-IRES-lysA-(poly-A) cassette. Using genomic DNA of LMO as template, lysA was amplified whereby an EcoRI site and a BamHI site were incorporated at the 5′ and 3′ ends, respectively. Subsequently, an internal ribosomal entry site (IRES) and bovine growth hormone (bGH) poly(A) were amplified whereby restriction sites were incorporated at the 3′ end of the IRES (EcoRI) and the 5′ end of bGH poly(A) (BamHI). The three amplicons were digested with the corresponding enzymes, and subsequently joined by ligation. A gel-purified fragment, corresponding to IRES-lysA-bGH poly(A) served as template for amplification of the cassette whereby phosphate groups were incorporated by the oligonucleotides at both extremities. The cassette was subsequently ligated into the PCR-amplified backbone of pFXluc, generating pFXluc-IRES-lysA. HSV1 tk, herpes simplex virus type 1 thymidine kinase gene; P_CMV, cytomegalovirus promoter.

pFXluc-IRES-lysA was introduced into LMO, yielding LMOlys, and a comparative bactofection assay was performed with LMOluc. Relative to LMOluc, 20% of CFU were recovered with a 40% increase in luminescence detected from LMOlys-infected monolayers (plus ampicillin) (Fig. 4A), demonstrating improved lysis due to expression of bacterial lysin by the eukaryotic cells, and with corresponding improved efficacy of bactofection.

FIG. 4.

Improvement of bactofection efficacy. (A) ΔLMOlys yielded the highest level of bactofection in vitro. In vitro bactofection efficacy was determined for L. monocytogenes harboring pFXluc (LMOluc) or pFXluc-IRES-lys (LMOlys), and L. monocytogenes ΔyycH harboring pFXluc (ΔLMOluc) or pFXluc-IRES-lys (ΔLMOlys). The recovered CFU on ampicillin treatment (shaded columns; primary y axis) and the levels of luminescence (data points; secondary y axis) are expressed as percentage CFU recovered and percentage luminescence, respectively, relative to the levels of LMOluc. Data shown are the averages of three independent experiments, with each experiment performed in duplicate. (B) ΔLMOlys significantly improved bactofection in vivo compared with LMOluc. Color images available online at www.liebertonline.com/hum.

Selection of LMO strain with enhanced ampicillin sensitivity

One of the limiting factors for the level of lysin production by eukaryotic cells is initial plasmid release induced by ampicillin treatment. To address this, we selected a derivative of LMO that has increased sensitivity to ampicillin. yycH is part of a two-component system that is conserved in low guanine–cytosine content (low GC%) bacteria. In B. subtilis yycH regulates the activity of the two-component system YycFG, and a yycH mutant in B. subtilis revealed a cell wall defect (Szurmant et al., 2005, 2007). It has been shown that a yycH mutant in Listeria is sensitive to ampicillin (Monk, 2009). To evaluate the sensitivity of ΔyycH to ampicillin at concentrations that are more relevant to intracellular levels (Chanteux et al., 2005) we assessed the sensitivity at concentrations up to 150 ng/ml. Clearly, at concentrations as low as 50 ng/ml ΔyycH has increased sensitivity to ampicillin relative to the wild type (Fig. 5) whereas in the absence of ampicillin the growth characteristics are similar to that of the wild-type strain.

FIG. 5.

Listeria monocytogenes ΔyycH has increased ampicillin sensitivity. Listeria monocytogenes EGDe wild type (LMO) and L. monocytogenes ΔyycH (ΔLMO) were grown in BHI, or in BHI supplemented with ampicillin (50, 100, or 150 ng/ml). Growth kinetics were monitored for 23 hr. Data shown are the averages of three independent experiments, each performed in triplicate. Error bars represent the standard error of the mean (SEM).

We introduced pFXluc and pFXluc-IRES-lysA into ΔLMO, yielding ΔLMOluc and ΔLMOlys, respectively. ΔLMOluc displayed 3.2-fold increased efficiency in bactofection compared with LMOluc, whereas ΔLMOlys-mediated bactofection efficiency was 3.8-fold higher than that of ΔLMOluc as evidenced by luciferase assays in vitro (Fig. 4A). In contrast, the introduction of lysA into LMO increased bactofection only by 1.4-fold. We therefore concluded that ΔLMO was more sensitive to lysis on exposure to ampicillin, and/or the lysin protein, than LMO.

Comparative in vivo bactofection with LMOluc versus ΔLMOlys

LMOluc or ΔLMOlys (10⁶ CFU) was intratumorally administered to MCF-7 tumors and allowed to replicate and spread intratumorally for 6 days, after which systemic ampicillin treatment was initiated, followed by assessment of luminescence. High relative levels of luminescence were observed specifically in tumors injected with ΔLMOlys (Fig. 4B).

Safety profile of ΔLMOlys

Given the requirements for safety of any gene delivery system, we investigated the safety profile of ΔLMOlys. Groups of subcutaneous MCF-7-bearing athymic mice (n = 6) were intratumorally administered ΔLMOlys with or without subsequent ampicillin treatment. Viable bacteria were recovered from liver and spleen and enumerated (Fig. 6A and B). Results demonstrated almost complete eradication of systemic vector, with ampicillin treatment eliminating ΔLMOlys from the liver, and reducing intrasplenic bacteria 3.5-log fold. When combined moxifloxacin and ampicillin treatment was examined, systemic infection was completely eliminated (data not shown). Reporter gene detection was not increased by the dual antibiotic treatment (data not shown), which may be explained by the fact that moxifloxacin does not induce lysis, being a member of the fluoroquinolone group of antibiotics that inhibit bacterial DNA replication (Hopkins, 2003).

FIG. 6.

Safety profile L. monocytogenes ΔyycH. ΔLMO was injected intratumorally (day 0). After 6 days, half the animals in each group were treated intraperitoneally with ampicillin for 3 days. Recovery of ΔLMO was assessed on days 1, 7, and 9 in (A) liver and (B) spleen. (C) ΔLMO has attenuated virulence. Animals were injected intraperitoneally with LMO and ΔLMO (both, 3 × 10⁵ CFU in 200 μl). After 6 days, CFU were assessed in liver and spleen. *p < 0.05.

To further evaluate the ability of ΔLMO to cause systemic infection, we compared LMO and ΔLMO in a BALB/c infection assay. ΔLMO was found to be significantly less virulent than the wild-type strain because significantly less CFU was recovered from liver and spleen (p = 0.026 and p = 0.024, respectively) (Fig. 6C).

Prolonged gene expression by bactofection-mediated delivery

Plasmid-based gene delivery systems are known to provide only short-term gene expression in tumors, because of plasmid loss and/or promoter silencing by methylation of microbial DNA sequences shortly posttransfection (Brooks et al., 2004). We hypothesized that in the LMO system described here, host cell-mediated expression of lysin gradually lyses more bacteria, thereby creating a continuous flow of newly released plasmid DNA. To test this, we compared reporter gene expression over time in MCF-7 tumors administered pFXluc-IRES-lysA by ΔLMOlys or by electroporation. Figure 7 shows that ΔLMOlys displayed higher relative levels of luminescence on days 5 and 8 (64 and 50%, respectively) compared with levels derived from electroporation (34 and 14%, respectively), suggesting sustained plasmid release and prolonged expression by listerial delivery.

FIG. 7.

Prolonged gene expression mediated by ΔLMOlys plasmid delivery. MCF-7 tumors were injected with 10⁶ CFU of ΔLMOlys. Six days postinfection, ampicillin treatment commenced. On the second day of ampicillin treatment, a parallel group of tumor-bearing mice was electroporated intratumorally with 10 μg of pFXluc-IRES-lysA. Luminescence levels were assessed 1 day postelectroporation (equivalent to the third and final day of ampicillin treatment in the bactofection group), 5 days postelectroporation, and 8 days postelectroporation. Results are expressed as percentage luminescence relative to the levels obtained on day 1 for each group. Results shown are the averages of four tumors per group.

Flow cytometric analysis of ΔLMOlys-transfected tumors

Data from flow cytometric analyses of the nature of plasmid-expressing cell types demonstrated that 72 hr after initiation of ampicillin treatment, 69 (±8)% of tumor-associated luminescence originated from epithelial cells (pan-cytokeratin positive) (Fig. 8) and 0% from macrophages (CD68 marker), indicating active LMO-mediated gene delivery to tumor cells. Although luciferase-expressing cells will reflect the cell types taking up LMO cells subsequent to administration, it must be noted that by the time of sampling (8 days after LMO administration), it is likely that LMO-containing macrophages will have cleared the tumor environment. Of total cells isolated from tumors 31.4 (±21)% expressed luciferase, demonstrating the transfection efficiency of this vector strategy to be comparable with that of other nonviral techniques (Fig. 8).

FIG. 8.

Flow cytometric analysis of ΔLMOlys-transfected tumors Subcutaneous MCF-7 tumors were subjected to the ΔLMOlys–ampicillin schedule as described previously. Tumors were excised 72 hr after ampicillin administration, and single-cell suspensions were prepared and stained with anti-luciferase–Cy5, anti-pan-cytokeratin–PE, and anti-CD68–FITC. Cells were analyzed by flow cytometry. (A) Rat isotype control IgG. (B) 52.6% of total cells isolated from tumors were positive for luciferase (average, 31.4 [ ± 21], n = 3). Within luciferase-positive cells (C), 69 (±8)% were pan-cytokeratin positive (epithelial/tumor cells). None (0%) of the luciferase-positive cells were CD68⁺ (macrophage cells) (data not shown).

LMO-mediated gene delivery to patient tumor tissue ex vivo

Freshly resected malignant breast tumor tissues were divided into three equal portions and placed in culture for a maximum of 5 days, the limit to keep the tissue viable. On day 1, two samples received 10⁶ ΔLMOlys and one sample was untreated. One of the transfected samples and the untransfected sample were administered ampicillin 24 hr postinjection. At 48 hr after ampicillin treatment, all tissue sections were imaged. Successful bactofection was observed in ampicillin-treated ΔLMOlys tissue, which showed luminescence throughout the sample (Fig. 9). No luminescence was observed in untransfected samples, whereas low levels of luminescence were observed in tissue injected with ΔLMOlys (without ampicillin), suggestive of natural lysis of ΔLMOlys. Similar results were observed for three independent tissue samples.

FIG. 9.

Gene delivery in patient ex vivo breast tissue. The ability of LMO to mediate gene delivery to spontaneous human tumors was demonstrated. Ampicillin treatment-induced bactofection of resected patient breast tumor tissue ex vivo is shown.

Discussion

The ability of LMO to initiate cell invasion, coupled with cell-to-cell spread, makes this bacterium a powerful candidate bactofection vector for tumor gene delivery. Using an attenuated derivative of LMO with improved sensitivity to ampicillin, we demonstrate for the first time plasmid delivery in tumors by means of antibiotic treatment. Several features make LMO a superior vehicle for DNA delivery than nonspreading gram-negative Salmonella, for which tumor bactofection has been reported (Fu et al., 2008). LMO has previously been investigated as a vehicle for gene delivery using in vitro (cell culture) models (Dietrich et al., 1998; Pilgrim et al., 2003) and use of a plasmid that expressed a phage lysin under the control of the listerial actA promoter was found to mediate bacterial lysis once internalized to mammalian cells (Dietrich et al., 1998). However, this approach does not allow for cell-to-cell spread. In addition, we have noted that the actA promoter is not completely switched off in vitro (data not shown), thereby potentially inducing bacterial lysis before reaching the target tissue, unlike the eukaryotic cell-specific CMV system described here.

We aimed to develop a novel approach employing LMO, maintaining the ability to spread, yet attenuated in virulence and where eradication of the vector in systemic infection can be achieved. To this end, we have demonstrated that the combination of ampicillin with a super-ampicillin-sensitive strain achieves simultaneous plasmid release and destruction of bacterial vector. We demonstrated LMO to spread intratumorally in an ActA-dependent manner. The observation of a moderate 68% increase in bacterial numbers over 8 days might be explained in part by death of transfected tumor cells over time in the rapidly growing murine tumor model. It is well established that central regions of tumor masses become necrotic during development (Bertout et al., 2008), which as a consequence cut off LMO from host-derived nutrients such as lipoic acid, essential for intracellular growth (O'Riordan et al., 2003). Furthermore, it has previously been shown that tumor growth is severely impacted by LMO infection, mediated by natural killer cell cytotoxicity (Shen et al., 2008). For tumor gene therapy, this facet may be a beneficial adjunct in treatment, and Fig. 7 indicates that the LMO system is capable of mediating prolonged expression despite any death of originally transfected cells.

To assess bactofection, we used the eukaryotic cell-specific CMV promoter to express a firefly luciferase reporter gene. Although low-level transcription from this promoter has been reported in E. coli (Mishin et al., 2001), we did not observe luciferase expression from the CMV promoter in LMO, and concluded that observed luminescence is specific to mammalian cell expression. Given the heterogeneity of tumor-associated cells, there is potential for passive and active uptake of LMO by many intratumoral cell types (Whiteside, 2008). This notwithstanding, L. monocytogenes is known to target epithelial cells through the interaction of the bacterial protein InlA with host cell E-cadherin (Hamon et al., 2006). Because most tumors are of epithelial origin (carcinomas) (Garcia et al., 2000), of which the majority (including MCF-7; Wang et al., 2009) retain expression of E-cadherin, LMO possesses an ability to directly target carcinoma cells, thereby further increasing the attractiveness of this bactofection vector for cancer gene therapy. Indeed, in the current study FACS characterization of the cell types transfected by LMO demonstrated that tumor cells, rather than tumor-associated macrophages, are the major cell type bactofected in vivo after intratumoral injection.

Previously we have shown that Listeria can migrate to 4T1 tumor tissue on systemic delivery (our unpublished data), and on the basis of the findings of Yu and colleagues (2008) we anticipate that systemic delivery to MCF-7 tumor tissue can also be obtained. Moreover, in the current study we have shown that intraperitoneal ampicillin administration also results in intratumoral plasmid release (data not shown). Work is ongoing in our laboratory to optimize routes of administration of LMO vectors in tumor-bearing mice.

Using wild-type Listeria we refined our system to optimize plasmid delivery in vitro by means of ampicillin treatment. In our system, host cell-specific expression of both the firefly luciferase and listerial lysin is achieved. Although low-level spontaneous lysis was observed, an increase in luciferase expression was obtained after ampicillin treatment as a result of plasmid release followed by host-specific expression of lysin that will subsequently cause further bacterial lysis. Incorporation of this construct in a strain that is more sensitive to ampicillin (ΔLMOlys) than the wild type increased bactofection levels further. Despite the attenuated nature of ΔLMO, reflected by significantly reduced levels in spleen and liver in a murine infection model, the highest level of bactofection was obtained with this strain, both in vitro and in vivo, highlighting the sensitivity of this strain to both ampicillin and the lysin.

Using ex vivo human mammary tumor tissue we were also able to obtain successful bactofection with ΔLMOlys. Even though ex vivo culture conditions did not permit application of the full bactofection regimen, we clearly demonstrated gene delivery to patient tumors. Such an ability to transfect tumors in a “human setting” further establishes the potential for this strategy given that many gene delivery protocols fail in the clinic following extrapolation from murine data, for example, due to absence of viral receptors in patient tumors (Summerford and Samulski, 1998).

The combination of antibiotic treatment with an antibiotic-sensitive strain is a novel and simple approach to establish bactofection. There is potential for the efficacy of both delivery and safety to be further refined. The ΔprfA LMO vector assessed in human trials demonstrates greater attenuation than the mutant used in the current study (Maciag et al., 2009). We are currently investigating the optimal combination of mutations that will give rise to further attenuation while retaining the efficacy of gene delivery and cell-to-cell spread in LMO. We did not observe a linear correlation between vector lysis and luminescence levels. The poor accessibility of plasmid DNA from the cytoplasm to the nucleus may be a factor, and possibilities to improve the efficiency of expression include incorporation of nuclear localization sequences (Diamond and Greenbaum, 2008), particularly for the nondividing cellular component of solid tumors. Delivery of RNA instead of DNA is a promising alternative (Schoen et al., 2005; Loeffler et al., 2006), that is, in addition to protein delivery strategies, under current investigation in our laboratory.

The system described here displays many advantages over current plasmid delivery methods. Traditional nonviral systems mediate gene expression only in cells originally transfected, with no spread posttransfection; suffer from transient gene expression; and deliver plasmid only. Our system has the potential to permit self-delivery of DNA, RNA, or protein, with long-term gene expression mediated by sustained plasmid release, which can be induced/silenced by antibiotic administration. Taken together, Listeria-mediated gene delivery for the treatment of tumors is a realistic concept.

Footnotes

Acknowledgments

This work was supported through a grant from Science Foundation Ireland 07/RFP/BIMF542.

Author Disclosure Statement

Potential competing financial interests declared as patent applications relating to this work have been filed (U.S. Patent Application No. 61/099,449 and I.E. Patent Application No. 2008/0724).

References

Bertout

J.A.

, Patel

S.A.

, Simon

M.C.

2008. The impact of O₂ availability on human cancer. Nat. Rev. Cancer, 8:967–975.

Boulaiz

, Marchal

J.A.

, Prados

, Melguizo

, Aranega

2005. Non-viral and viral vectors for gene therapy. Cell. Mol. Biol. (Noisy-le-Grand), 51:3–22.

Braun

, Ohayon

, Cossart

1998. The InIB protein of Listeria monocytogenes is sufficient to promote entry into mammalian cells. Mol. Microbiol., 27:1077–1087.

Brooks

A.R.

, Harkins

R.N.

, Wang

, Qian

H.S.

, Liu

, Rubanyi

G.M.

2004. Transcriptional silencing is associated with extensive methylation of the CMV promoter following adenoviral gene delivery to muscle. J. Gene Med., 6:395–404.

Chanteux

, Van Bambeke

, Mingeot-Leclercq

M.P.

, Tulkens

P.M.

2005. Accumulation and oriented transport of ampicillin in Caco-2 cells from its pivaloyloxymethylester prodrug, pivampicillin. Antimicrob. Agents Chemother., 49:1279–1288.

Diamond

S.L.

, Greenbaum

2008. Proteases' prime targets revealed. Nat. Biotechnol., 26:652–653.

Dietrich

, Bubert

, Gentschev

, Sokolovic

, Simm

, Catic

, Kaufmann

S.H.

, Hess

, Szalay

A.A.

, Goebel

1998. Delivery of antigen-encoding plasmid DNA into the cytosol of macrophages by attenuated suicide Listeria monocytogenes. Nat. Biotechnol., 16:181–185.

Dietrich

, Gentschev

, Hess

, Ulmer

J.B.

, Kaufmann

S.H.

, Goebel

1999. Delivery of DNA vaccines by attenuated intracellular bacteria. Immunol. Today, 20:251–253.

Dietrich

, Spreng

, Gentschev

, Goebel

2000. Bacterial systems for the delivery of eukaryotic antigen expression vectors. Antisense Nucleic Acid Drug Dev., 10:391–399.

10.

Dietrich

, Kolb-Maurer

, Spreng

, Schartl

, Goebel

, Gentschev

2001. Gram-positive and gram-negative bacteria as carrier systems for DNA vaccines. Vaccine, 19:2506–2512.

11.

Dramsi

, Kocks

, Forestier

, Cossart

1993. Internalin-mediated invasion of epithelial cells by Listeria monocytogenes is regulated by the bacterial growth state, temperature and the pleiotropic activator prfA. Mol. Microbiol., 9:931–941.

12.

, Lan

, Li

, Han

, Gao

, Ren

2008. Synergistic antitumor efficacy of suicide/ePNP gene and 6-methylpurine 2′-deoxyriboside via Salmonella against murine tumors. Cancer Gene Ther., 15:474–484.

13.

Gaeng

, Scherer

, Neve

, Loessner

M.J.

2000. Gene cloning and expression and secretion of Listeria monocytogenes bacteriophage-lytic enzymes in Lactococcus lactis. Appl. Environ. Microbiol., 66:2951–2958.

14.

Garcia

S.B.

, Novelli

, Wright

N.A.

2000. The clonal origin and clonal evolution of epithelial tumours. Int. J. Exp. Pathol., 81:89–116.

15.

Glaser

, Frangeul

, Buchrieser

, Rusniok

, Amend

, Baquero

, Berche

, Bloecker

, Brandt

, Chakraborty

, Charbit

, Chetouani

, Couve

, De Daruvar

, Dehoux

, Domann

, Dominguez-Bernal

, Duchaud

, Durant

, Dussurget

, Entian

K.D.

, Fsihi

, Garcia-Del Portillo

, Garrido

, Gautier

, Goebel

, Gomez-Lopez

, Hain

, Hauf

, Jackson

, Jones

L.M.

, Kaerst

, Kreft

, Kuhn

, Kunst

, Kurapkat

, Madueno

, Maitournam

, Vicente

J.M.

, Ng

, Nedjari

, Nordsiek

, Novella

, De Pablos

, Perez-Diaz

J.C.

, Purcell

, Remmel

, Rose

, Schlueter

, Simoes

, Tierrez

, Vazquez-Boland

J.A.

, Voss

, Wehland

, Cossart

2001. Comparative genomics of Listeria species. Science, 294:849–852.

16.

Grillot-Courvalin

, Goussard

, Courvalin

2002. Wild-type intracellular bacteria deliver DNA into mammalian cells. Cell Microbiol., 4:177–186.

17.

Hamon

, Bierne

, Cossart

2006. Listeria monocytogenes: A multifaceted model. Nat. Rev. Microbiol., 4:423–434.

18.

Hegde

, Qi

, Abernathy

, Gay

, Dharap

, Gaspard

, Hughes

J.E.

, Snesrud

, Lee

, Quackenbush

2000. A concise guide to cDNA microarray analysis. Biotechniques, 29:548–550552–544, 556.

19.

Hense

, Domann

, Krusch

, Wachholz

, Dittmar

K.E.

, Rohde

, Wehland

, Chakraborty

, Weiss

2001. Eukaryotic expression plasmid transfer from the intracellular bacterium Listeria monocytogenes to host cells. Cell Microbiol., 3:599–609.

20.

Hopkins

S.J.

2003. Drugs and Pharmacology for Nurses. Churchill Livingstone/Elsevier: New York.

21.

Hucker

G.J.

, Conn

H.J.

1928. Further studies on the methods of gram staining. N.Y. Agric. Exp. Stat. Tech. Bull., 128:1.

22.

Loeffler

D.I.

, Schoen

C.U.

, Goebel

, Pilgrim

2006. Comparison of different live vaccine strategies in vivo for delivery of protein antigen or antigen-encoding DNA and mRNA by virulence-attenuated Listeria monocytogenes. Infect. Immun., 74:3946–3957.

23.

Loessner

M.J.

, Wendlinger

, Scherer

1995. Heterogeneous endolysins in Listeria monocytogenes bacteriophages: A new class of enzymes and evidence for conserved holin genes within the siphoviral lysis cassettes. Mol. Microbiol., 16:1231–1241.

24.

Maciag

P.C.

, Radulovic

, Rothman

2009. The first clinical use of a live-attenuated Listeria monocytogenes vaccine: A phase I safety study of Lm-LLO-E7 in patients with advanced carcinoma of the cervix. Vaccine, 27:3975–3983.

25.

Mishin

V.P.

, Cominelli

, Yamshchikov

V.F.

2001. A “minimal” approach in design of flavivirus infectious DNA. Virus Res., 81:113–123.

26.

Monk

I.R.

, Gahan

C.G.

, Hill

2008. Tools for functional postgenomic analysis of Listeria monocytogenes. Appl. Environ. Microbiol., 74:3921–3934.

27.

Monk

I.R.

2009. Development and application of genetic tools for the functional genomics of Listeria monocytogenes. PhD Thesis. National University of Ireland.

28.

Mounier

, Ryter

, Coquis-Rondon

, Sansonetti

P.J.

1990. Intracellular and cell-to-cell spread of Listeria monocytogenes involves interaction with F-actin in the enterocytelike cell line Caco-2. Infect. Immun., 58:1048–1058.

29.

O'Riordan

, Moors

M.A.

, Portnoy

D.A.

2003. Listeria intracellular growth and virulence require host-derived lipoic acid. Science, 302:462–464.

30.

Pelletier

, Kaplan

, Racaniello

V.R.

, Sonenberg

1988. Cap-independent translation of poliovirus mRNA is conferred by sequence elements within the 5′ noncoding region. Mol. Cell. Biol., 8:1103–1112.

31.

Pilgrim

, Stritzker

, Schoen

, Kolb-Maurer

, Geginat

, Loessner

M.J.

, Gentschev

, Goebel

2003. Bactofection of mammalian cells by Listeria monocytogenes: Improvement and mechanism of DNA delivery. Gene Ther., 10:2036–2045.

32.

Portnoy

D.A.

, Chakraborty

, Goebel

, Cossart

1992. Molecular determinants of Listeria monocytogenes pathogenesis. Infect. Immun., 60:1263–1267.

33.

Portnoy

D.A.

, Auerbuch

, Glomski

I.J.

2002. The cell biology of Listeria monocytogenes infection: The intersection of bacterial pathogenesis and cell-mediated immunity. J. Cell Biol., 158:409–414.

34.

Rea

, Hill

, Gahan

C.G.

2005. Listeria monocytogenes PerR mutants display a small-colony phenotype, increased sensitivity to hydrogen peroxide, and significantly reduced murine virulence. Appl. Environ. Microbiol., 71:8314–8322.

35.

Riedel

C.U.

, Monk

I.R.

, Casey

P.G.

, Morrissey

, O'Sullivan

G.C.

, Tangney

, Hill

, Gahan

C.G.

2007. Improved luciferase tagging system for Listeria monocytogenes allows real-time monitoring in vivo and in vitro. Appl. Environ. Microbiol., 73:3091–3094.

36.

Robbins

J.R.

, Barth

A.I.

, Marquis

, De Hostos

E.L.

, Nelson

W.J.

, Theriot

J.A.

1999. Listeria monocytogenes exploits normal host cell processes to spread from cell to cell. J. Cell Biol., 146:1333–1350.

37.

Sambrook

, Fritsch

E.F.

, Maniatis

1989. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press: Cold Spring Harbor, NY.

38.

Schoen

, Kolb-Maurer

, Geginat

, Loffler

, Bergmann

, Stritzker

, Szalay

A.A.

, Pilgrim

, Goebel

2005. Bacterial delivery of functional messenger RNA to mammalian cells. Cell Microbiol., 7:709–724.

39.

Schoen

, Loeffler

D.I.

, Frentzen

, Pilgrim

, Goebel

, Stritzker

2008. Listeria monocytogenes as novel carrier system for the development of live vaccines. Int. J. Med. Microbiol., 298:45–58.

40.

Shen

, Kanoh

, Maruyama

, Matsumoto

, Zhang

, Asano

2008. Attenuated Listeria infection activates natural killer cell cytotoxicity to regress melanoma growth in vivo. Microbiol. Immunol., 52:107–117.

41.

Summerford

, Samulski

R.J.

1998. Membrane-associated heparan sulfate proteoglycan is a receptor for adeno-associated virus type 2 virions. J. Virol., 72:1438–1445.

42.

Szurmant

, Nelson

, Kim

E.J.

, Perego

, Hoch

J.A.

2005. YycH regulates the activity of the essential YycFG two-component system in Bacillus subtilis. J. Bacteriol., 187:5419–5426.

43.

Szurmant

, Fukushima

, Hoch

J.A.

2007. The essential YycFG two-component system of Bacillus subtilis. Methods Enzymol., 422:396–417.

44.

Thomas

C.E.

, Ehrhardt

, Kay

M.A.

2003. Progress and problems with the use of viral vectors for gene therapy. Nat. Rev. Genet., 4:346–358.

45.

Van Langendonck

, Bottreau

, Bailly

, Tabouret

, Marly

, Pardon

, Velge

1998. Tissue culture assays using Caco-2 cell line differentiate virulent from non-virulent Listeria monocytogenes strains. J. Appl. Microbiol., 85:337–346.

46.

Vassaux

, Nitcheu

, Jezzard

, Lemoine

N.R.

2006. Bacterial gene therapy strategies. J. Pathol., 208:290–298.

47.

Wang

, Li

, Wang

, Yang

, Sun

, Yao

, Cai

, Wu

, Zhou

, Zha

2009. E-cadherin decreased human breast cancer cells sensitivity to staurosporine by up-regulating Bcl-2 expression. Arch. Biochem. Biophys., 481:116–122.

48.

Whiteside

T.L.

2008. The tumor microenvironment and its role in promoting tumor growth. Oncogene, 27:5904–5912.

49.

F.F.

, Pearce

L.E.

, Yu

P.L.

1991. Construction of a family of lactococcal vectors for gene cloning and translational fusion. FEMS Microbiol. Lett., 61:55–59.

50.

Yamamoto

, Curiel

D.T.

2005. Cancer gene therapy. Technol. Cancer Res. Treat., 4:315–330.

51.

Y.A.

, Zhang

, Szalay

A.A.

2008. Establishment and characterization of conditions required for tumor colonization by intravenously delivered bacteria. Biotechnol. Bioeng., 100:567–578.