Selective gene transfer in vitro to tumor cells via recombinant Newcastle disease virus

Cancer Gene Therapy (2005) 12, 295–303 All rights reserved 0929-1903/05 $30.00

r 2005 Nature Publishing Group

www.nature.com/cgt

Selective gene transfer in vitro to tumor cells via recombinant Newcastle disease virus Huijie Bian,1,3 Philippe Fournier,1 Rob Moormann,2 Ben Peeters,2 and Volker Schirrmacher1 1

Division of Cellular Immunology, German Cancer Research Center, D010, Im Neuenheimer Feld 280, 69120 Heidelberg, Germany; 2Animal Sciences Group, Wageningen UR, Division of Infectious Diseases, PO Box 65, 8200 AB Lelystad, The Netherlands; and 3Cell Engineering Research Center, Fourth Military Medical University, Xi’an 710032, China. We developed a novel strategy to target recombinant Newcastle disease virus (NDV) to tumor cells for gene therapy. Modifying the virus with a bispecific fusion protein allowed virus receptor-independent tumor cell binding and gene transfer. The targeting molecule aHN-IL-2 contains an scFv antibody cloned from a neutralizing hemagglutinin-neuraminidase (HN)-specific hybridoma linked to the human cytokine IL-2. A recombinant NDV expressing the enhanced green fluorescent protein (NDFL-EGFP) was applied to show the expression of foreign genes in virus-infected tumor cells. At 24 hours after infection with the modified virus (NDFL-EGFP/aHN-IL-2), FACS analysis and fluorescence microscopy revealed neutralization of natural infection in IL-2 receptornegative Jurkat leukemia cells, but targeted expression of EGFP in IL-2 receptor-positive human leukemia-derived MT-2 cells. The targeted gene delivery of NDFL-EGFP/aHN-IL-2 in MT-2 cells was blocked by the target ligand human IL-2. Selective virus entry to IL-2 receptor bearing tumor cells was also observed in a mixture of Jurkat and MT-2 cell lines. These results demonstrate that a recombinant NDV carrying a foreign gene can be successfully targeted to a specific tumor through a bispecific protein, which thereby increases the selectivity of gene transfer. Cancer Gene Therapy (2005) 12, 295–303. doi:10.1038/sj.cgt.7700774 Published online 17 December 2004 Keywords: newcastle disease virus; bispecific; targeted; enhanced green fluorescent protein

esistance of malignant diseases to conventional therapies has inspired the search for novel strategies R such as immunotherapy or gene therapy. Clinical antitumor vaccination studies employing Newcastle disease virus (NDV)-modified autologous tumor cell vaccine (ATV-NDV) indicated improvement of survival.1–6 These and other clinical data7 suggest that NDV is an interesting antineoplastic agent for treatment of human cancer. Recent studies showed that a reporter gene encoding human secreted alkaline phosphatase can be inserted at different positions in the RNA genome of NDV without severely affecting replication efficiency or virus yield.8 A recombinant NDV expressing a reporter gene, the enhanced green fluorescent protein (EGFP) was also generated independently by two groups by applying reverse genetics techniques.9,10 The results suggest a novel potential usefulness of NDV as a viral vector for gene therapy. The molecular biology of NDV,11 of oncolytic strains12 and of recombinant NDV as a vaccine vector13 have been Received May 20, 2004.

Address correspondence and reprint requests to: Professor Volker Schirrmacher, PhD, German Cancer Research Center (DKFZ), Division of Cellular Immunology (D010), Im Neuenheimer Feld 280, D-69120 Heidelberg, Germany. E-mail: [email protected]

described. This avian paramyxovirus has a nonsegmented negative-stranded RNA genome encoding, among others, two membrane glycoproteins, hemagglutinin-neuraminidase (HN) and F, which are expressed as spike proteins in its envelope. The HN protein is involved in cell attachment and virus release, and the F protein is mediating fusion of the viral envelope with cellular membranes. NDV binds to both normal cells and tumor cells via the interaction between HN and sialic acidcontaining molecules on cell surfaces. It replicates, however, selectively in tumor cells because these make a weaker interferon response than normal cells.14,15 Binding to normal cells might compromise the therapeutic effect of NDV when administered systemically in vivo. In our previous animal studies, antitumor activity was observed only when NDV was injected locally but not systemically (i.v. or i.p.).16 One of the reasons is probably that NDV binds to normal cells, and therefore does not reach the tumor cells. In this study, we developed a model for tumor-targeted gene delivery by using the recombinant EGFP expressing NDV (named NDFL-EGFP), which is modified with a bispecific fusion protein. Bispecific molecules can be used for targeting of viral vectors to localize gene transfer to specific cell types, which may improve gene delivery and reduce immunogenicity and side effects in gene

Selective gene transfer by recombinant NDV H Bian et al

296

therapy.17–19 We designed a bispecific fusion protein that simultaneously blocks native cell binding of NDV and creates a new binding specificity to a defined target. As new target we chose the interleukin-2 receptor (IL-2R). Adult T-cell leukemia (ATL) cell lines, which are infected with human T-cell leukemia virus-I (HTLV-I), uniformly express IL-2Ra chains.20 The observation that IL-2Ra is not expressed by resting normal cells, but is expressed by a proportion of the abnormal cells in certain forms of lymphoid neoplasia, provides the rationale for the use of the IL-2Ra as a target.21,22 For this purpose, a singlechain antibody (scFv) with virus neutralizing activity specific for HN of NDV was fused with a cDNA encoding the human cytokine IL-2 (named aHN-IL-2). Using IL2R
and IL-2R þ tumor cell lines, we show that the modification of the recombinant NDFL-EGFP with the bispecific fusion protein aHN-IL-2 blocks native receptor binding to IL-2R
cells but redirects NDV to IL-2R þ cell lines in which the viral replication leads to the expression of the transgene EGFP in a very selective way. The approach will be further followed for targeting of NDV vectors to the site of metastases after systemic injection. This study first proves the principle of changing the specificity of NDV with regard to tumor cell binding and gene transfer.

Materials and methods

Cells Dihydrofolate reductase (dhfr)-deficient CHO cells (ATCC, CRL-9096) were used for the production of the bispecific fusion protein aHN-IL-2. These cells are grown in alpha MEM containing 5% dialysed fetal calf serum (FCS), 2 mM L-glutamine, 100 U/ml penicillin, 100 mg/ml streptomycin and 10 mM methothrexate. The human Jurkat CD3 cell line that was sorted from Jurkat cells for CD3-positivity was grown in RPMI-1640 medium supplemented with 5% inactivated FCS, 2 mM L-glutamine, 2% HEPES and 100 U/ml penicillin–100 mg/ml streptomycin. MT-2, an HTLV-1 transformed T cell line, was kindly provided by Dr. Masahiko Makino (Department of Microbiology, National Institute of Infectious Disease, Tokyo, Japan). MT-2 cells were propagated in the same RPMI-1640 medium supplemented with 10% FCS. All reagents were purchased from Gibco Life technologies (Karlsruhe, Germany) except for the dialyzed FCS, which was purchased from Biochrom (Krefeld, Germany) and for the methotrexate, which was obtained from Calbiochem–Novabiochem (Schwalbach, Germany). All cell lines were maintained at 371C in a humidified atmosphere of 5% CO2.

Antibodies and cytokines Mouse anti-HN monoclonal antibody (HN.B mAb, IgG 2a) and mouse anti-F mAb lcii (IgG1) were kindly provided by Dr Iorio (Department of Molecular Genetics and Microbiology, University of Massachusetts, Medical School, MA). Anti-F scFv with an E Tag epitope

Cancer Gene Therapy

sequence was derived via recombinant scFv technology from immune spleen cells of the mice which had been immunized with NDV-Ulster. All three molecules were used in flow cytometry to detect NDV viral antigens on host cell surfaces. Anti Histidin-Tag mAb and its FITC conjugated form (both from Dianova GmbH, Hamburg, Germany) was used for the characterization of bispecific fusion protein aHN-IL-2. Mouse anti-E Tag antibody was purchased from Amersham Biosciences (Freiburg, Germany). Goat F(ab0 )2 anti-mouse Ig-RPE was obtained from Southern Biotechnology Associates, Inc. (Birmingham, AL). Anti-human IL-2Ra mAb was obtained from R&D Systems (Wiesbaden, Germany). rhIL-2 was purchased from Promokine (PromoCell GmbH, Heidelberg, Germany).

Construction of a recombinant NDFL-EGFP A DNA fragment containing the gene encoding the EGFP was obtained by means of PCR using plasmid pEGFP (Clontech BD Biosciences, Alphen aan den Rijn, The Netherlands) as a template. The DNA fragment containing the EGFP gene was inserted into plasmid pNDFL þ 23 between the transcription-start box of the NP gene and the NP open reading frame. In order to allow transcription of downstream genes, a synthetic nucleotide sequence corresponding to the consensus transcription-end and transcription-start box (ATTAAGAAAAAATACGGG TAGAAG) of NDV was inserted behind the EGFP gene. The resulting plasmid was used to rescue recombinant NDV as described previously.23 Rescued virus (designated NDFL-EGFP) was propagated in the allantoic cavity of 10–11-day-old specific pathogen-free embryonated eggs, which were incubated at 37–381C. Allantoic fluid was harvested 48 hours after inoculation. The virus titer (TCID50/ml) was 107.9. Single-step growth curves indicated that replication of the virus that carried the EGFP gene did not differ significantly from the parent strain NDFL þ (data not shown).

Other NDV strains Nonvirulent recombinant NDFL þ virus was generated from a cDNA clone of NDV strain LaSota.23 NDVUlster 2C was obtained in 1984 from Dr PH Russel (University London, England). All viruses were propagated in embryonated chicken eggs, harvested from the allantoic fluid, purified by ultracentrifugation as described previously,4 and cryopreserved in aliquots at
701C. The virus was quantified by a hemagglutination assay. One hemagglutination unit (HU) is defined as the smallest virus concentration leading to visible sheep erythrocyte agglutination.

Construction and expression of the bispecific fusion protein aHN-IL-2 The HN-specific scFv was elaborated from a hybridoma (HN.B) producing an HN-specific mAb using scFv recombinant technologies. The cDNA encoding the human IL-2 was derived from the plasmid pFC54.t

Selective gene transfer by recombinant NDV H Bian et al

(ATCC, Rockeville, MD). In order to assemble these two sequences into a mammalian expression vector (kindly provided by Dr P Ba¨uerle, Micromet, Martinsried, Germany), appropriate restriction sites were inserted by PCR at both ends of these sequences. The obtained fragments were cloned in this plasmid upstream of a dhfr gene in such a way that the expression of both genes in the cistron remains under the control of the same elongation factor promotor (hEF-A). The resulting plasmid was transfected into dhfrdeficient CHO cells by electroporation. Clones that were expressing the desired protein in an efficient way were selected. The production was performed in high density cell culture systems (Integra Biosciences, Fernwald, Germany). The produced protein containing a His tag was purified by IMAC. It was then characterized for its binding to NDV and IL-2R by flow cytometry. The purity was assessed by Coomassie staining after migration of the purified proteins on SDS-PAGE gels. The size of the molecule was checked by Western blot. For that, after SDS-PAGE (12.5% acrylamide) and electroelution onto polyvinylidene difluoride membranes, proteins were detected with anti-Flag mAb (Sigma, Taufkirchen, Germany) and goat anti-mouse IgG horseradish peroxidase (HRP)-conjugated Ab (Jackson Immunoresearch, distributed by Dianova GmbH, Hamburg, Germany). Blots were developed with the Lumilight Plus substrate (Roche Molecular Biochemicals, Mannheim, Germany).

Modification of NDV with bispecific fusion protein aHN-IL-2 Modification of NDV was performed by incubation of NDV with appropriate amounts of aHN-IL-2 for 1 hour on ice.

NDV binding and infection of tumor cells. Tumor cell suspensions were washed twice with FCS-free RPMI-1640 medium and 1  107 cells were incubated with 100 HU (or 10 HU) of NDV or the same doses of modified NDV/ aHN-IL-2 in a final volume of 1 ml for 1 hour at 371C in a CO2 incubator. During the incubation, cells were shaken every 15 minutes. The cells were then washed twice and either stained with antibodies and analyzed by FACS to measure the bound virus or the cells were further cultured for 24 hours to allow for viral replication. After staining with NDV-specific antibodies, viral replication could be monitored by flow cytometry as described below. EGFP fluorescence was observed either via flow cytometry or fluorescence microscopy. Flow cytometry. A total of 5  105 cells/sample were used for analysis by a FACScan flow cytometer (Becton Dickinson, Heidelberg, Germany). All antibodies were diluted in FACS buffer (PBS containing 5% FCS and 0.1% NaN3). Cells were washed twice with FACS buffer and then incubated with the first antibody. Subsequently the cells were washed and incubated with PE-conjugated second antibody (goat F(ab0 )2 anti-mouse Ig-RPE) for 20–30 minutes on ice in the dark. All FACS data were

analyzed with CELLQuest software (Becton Dickinson, Heidelberg, Germany).

Cytospin and fluorescence microscopy. Cells were harvested after infection with NDFL-EGFP or modified NDFL-EGFP/aHN-IL-2 for 24 hours. Single cell suspensions were dispersed onto slides using a Cytospin Centrifuge (Cytospin 2, Shandon, UK) at 700 rpm for 4 minutes with low acceleration. EGFP fluorescence was examined using a fluorescence microscope with FITC filter set (Carl Zeiss, Oberkochen, Germany). CFSE staining. To differentiate cells from a mixture of two different cell populations, one of them was first stained with CFSE (Molecular Probes, Leiden, The Netherlands). Briefly, before mixing the same number of Jurkat and MT-2 cells, either Jurkat or MT-2 cells were first washed twice with PBS at room temperature, and then incubated for 10 minutes at room temperature in the dark with CFSE (made from a 0.5 mM stock of CFSE in tissue culture grade DMSO). The optimized final concentration in PBS was 2 mM for Jurkat cells and 8 mM for MT-2 cells. The reaction was quenched with PBS containing 10% FCS.

Results

Characterization of the bispecific fusion protein aHN-IL-2 In order to block the native cell binding site of NDV and confine its tropism to IL-2R þ human ATL cells, the bispecific protein aHN-IL-2 was constructed (Fig 1a). In this fusion protein, the anti-HN scFv was derived from the hybridoma HN.B which can inhibit virus-cell binding. As a consequence, the HN.B mAb was able to block viral infection by over 90% (Fig 2a), while anti-F mAb neutralized maximally 50% of NDV infection in human leukemia-derived T cells MT-2. These results prompted us to use HN.B and to construct the bispecific protein from it. The new bispecific protein was produced in CHO cells and purified by immobilized metal affinity chromatography (IMAC) (Fig 1b and c). We first investigated its virus neutralization capacity and its binding properties towards HN. FACS analysis showed that IL-2Ra is not expressed at the surface of the Jurkat CD3 cell line. This observation was correlated with the absence of binding of aHN-IL-2 to these cells. When NDFL-EGFP was preincubated with different amounts of aHN-IL-2 for 1 hour on ice before being added to the Jurkat CD3 cells and the cells were then incubated for 24 hours, expression of EGFP by NDFL-EGFP was inhibited (Fig 2b). The 50% inhibitory concentration of aHN-IL-2 was 6.7, 5.9 and 2.5 mg/ml when using NDFL-EGFP at 100 HU, 10 HU or 1 HU per 107 cells, respectively. When Jurkat cells were first incubated with NDV-Ulster for 1 hour at 371C to expose viral HN molecules, aHN-IL-2 was found to bind to these Jurkat-NDV cells in a dose-dependent manner (Fig 3a).

Cancer Gene Therapy

297

Selective gene transfer by recombinant NDV H Bian et al

298

a

a SS

Flag

VH

VL

L

L

hIL-2

His

anti-HN scFv

b M (kDa)

c

% inhibition of infection

C-terminus

N-terminus

100 anti-HN anti-F

80 60 40 20 0 0.01

97

0.1

1

66 45

10

100

1000

mAb (µg/ml)

b

30 Figure 1 Characteristics of the bispecific fusion protein aHN-IL-2. (a) Design of the fusion protein sequence. The molecule is composed of an scFv specific for HN (made of the variable region of the heavy chain VH linked to the variable region of the light chain VL) joined by a linker (L; Gly4Ser) to the human IL-2 sequence. An eukaryotic secretory sequence (SS) including a 50 -terminal Kozak site and a Flag sequence (Flag) was introduced at the N-terminus. The Cterminal His tail permits purification of the protein via IMAC columns. Purified proteins were analyzed by 12.5% SDS-PAGE under reducing conditions and stained with Coomassie (b) or with an anti-Flag antibody after immunoblotting (c).

Next we investigated the binding properties of aHN-IL2 towards the IL-2 receptor (IL-2R). For this purpose, we used the HTLV-1-transformed T-cell line MT-2, which constitutively expresses high levels of human IL-2Ra. Figure 3a shows the dose-dependent binding of aHN-IL-2 to MT-2 cells. This binding could be blocked by the addition of anti-human IL-2Ra antibody, demonstrating the true binding of the aHN-IL-2 protein to the IL-2Ra (Fig 3b).

Alteration of the binding specificity of NDV by aHN-IL-2 We next tested the effect of aHN-IL-2 on the binding specificity of NDV. For that, NDFL-EGFP was preincubated with different amounts of aHN-IL-2 for 1 hour on ice. Then this modified virus (NDFL-EGFP/aHN-IL2) was used to bind to IL-2R
(Jurkat) and IL-2R þ (MT2) cells for 1 hour at 371C. Figure 3c shows that only 0.6– 0.7% of Jurkat cells bound the modified virus at concentrations of aHN-IL-2 ranging from 36.8 to 18.4 mg/ml, while 76.2% of Jurkat cells bound the native NDFL-EGFP virus. Thus, native binding of NDV was neutralized to a level higher than 99%. Modified virus NDFL-EGFP/aHN-IL-2, however, showed strong binding towards MT-2 cells. This IL-2R þ cell line bound the modified viral particles very well (96.2% positive cells). The native NDFL-EGFP was also bound (44.8% positive cells), but less well. These results demonstrate alteration

Cancer Gene Therapy

% inhibition of infection

100 80 60 40 20 0 0.1

1

10

100

α HN-IL-2 (µg/ml) Figure 2 Virus neutralization. (a) Neutralization of NDV infection in MT-2 cells by HN.B mAb and anti-F mAb. In all, 10 HU of NDFLEGFP was pre-incubated with different amounts of HN.B mAb (’) or anti-F mAb (&) for 1 hour on ice, then incubated with 106 MT-2 cells for 1 hour at 371C. After washing of unbound virus, the cells were further cultured for 24 hours. The signal of EGFP fluorescence was quantified by FACS analysis. (b) Neutralization of NDV infection in Jurkat cells by aHN-IL-2. Conditions were similar as in (a) except that NDFL-EGFP was pre-incubated with aHN-IL-2 and then further incubated with Jurkat CD3 leukemia cells. The percentage of inhibition was calculated by comparing the numbers of positive cells after infection with modified NDV with the number of positive cells after infection with native NDV.

of the binding specificity of NDV by aHN-IL-2. Native virus receptor-dependent binding to Jurkat cells could be completely neutralized and a new and selective binding to IL-2R-positive cells was observed.

Selectivity of NDV entry by aHN-IL-2 in a mixture of target positive and negative tumor cell lines To demonstrate the selectivity of retargeted NDV entry to IL-2R þ tumor cells, we designed a 24-hours virus infection assay with two target tumor cell populations, which had been differentiated by labeling with carboxyfluorescein diacetate succinimidyl ester (CFSE) before mixing. Virus entry and replication, which is associated with amplification of cell surface expression of HN and F molecules,14 was evaluated quantitatively by FACS analysis using anti-F antibodies. Without aHN-IL-2, the parental strain NDFL þ infected both cell types (labeled and unlabeled). In the presence of aHN-IL-2, however,

Selective gene transfer by recombinant NDV H Bian et al

NDFL þ infected selectively the IL-2R-positive MT-2 cells, either unlabeled (39.7%) or labeled (44.8%). Thus, similar results of selective entrance of NDV/aHN-IL-2 to MT-2 cells were obtained in this criss-cross experiment irrespective of whether Jurkat cells or MT-2 cells were labeled with CFSE (Fig 4).

Selective gene transfer via NDFL-EGFP by aHN-IL-2 Optimal infection in Jurkat and MT-2 cells by native NDFL-EGFP was observed at 24 hours. At this time

% positive cells

a 100 75

50

Specific target inhibition Jurkat-NDV (HN) MT-2 (IL-2R)

25

0 0.01

0.1

1 10 αHN-IL-2 (µg/ml)

100

1000

b

If the retargeted virus entry was truly dependent on specific recognition of the IL-2 receptor, then the receptor ligand IL-2 should be capable of competitive inhibition. This was indeed the case. Retargeted expression of EGFP in MT-2 cells 24 hours after infection by NDFL-EGFP/ aHN-IL-2 could be blocked by recombinant human IL-2 (rhIL-2). FACS data revealed that 86% of retargeted delivery of the transgene was inhibited with 5 mg of rhIL-2 (Fig 6).

counts

Discussion

Tumor-selective replication competent RNA viruses represent promising new vectors for gene therapy and oncolytic virotherapy.24,25 Replication of such viruses takes place in the cytoplasm of tumor cells. Since the viral genomes do not integrate into host cellular DNA there is

anti-His-FITC c 100 Jurkat % positive cells

point, we saw maximal cytoplasmic EGFP expression and maximal density of cell surface expressed HN and F molecules (data not shown). To test for maximal retargeted infection, we followed EGFP expression in MT-2 cells at 24 hours post-infection. An optimum was seen when using 12 mg/ml aHN-IL-2 and 10 HU NDFLEGFP in a volume of 50 ml. Under the optimized conditions, both flow cytometry and fluorescence microscopy revealed the difference between native and retargeted gene transfer of NDFL-EGFP in IL-2R
Jurkat and IL-2R þ MT-2 cells (Fig 5). Expression of EGFP by native NDFL-EGFP was stronger in Jurkat cells than in MT-2 cells (Fig 5a and b, thin black line). In contrast, expression of EGFP by modified NDFL-EGFP/aHN-IL2 was stronger in MT-2 cells (85.1%) than in Jurkat cells (5.4%) (Fig 5a and b, bold black line). The retargeted delivery of the reporter gene EGFP was confirmed by fluorescence microscopy (Fig 5c).

MT-2

75

50

25

0 36.8

27.6

18.4

9.2

4.6

HN-IL-2 ( g/ml)

2.3

0

Figure 3 Binding characteristics of aHN-IL-2 with or without NDV. (a) Bispecific binding of aHN-IL-2 to HN and IL-2R. Human IL-2R
(Jurkat) leukemia cells were modified with NDV by incubation with 400 HU of NDV-Ulster per 107 cells for 1 hour at 371C and then washed. Different amounts of aHN-IL-2 were then incubated with the NDV modified Jurkat cells (’) or with the human IL-2R þ (MT-2) cells (&) for 1 hour on ice. FACS analysis was performed after staining of the cells with anti-His mAb and goat anti-mouse F(ab0 )2PE. (b) Blocking of the binding of aHN-IL-2 to MT-2 cells with antihuman IL-2Ra antibodies. A total of 5  105 MT-2 cells were preincubated with 1.25 mg of anti-human IL-2Ra antibody for 1 hour on ice before another 1 hour incubation with aHN-IL-2 (bold black line). For control, cells were incubated only with PBS (gray peak) or only with aHN-IL-2 (thin black line). FACS analysis was performed after staining of the cells with anti-His-FITC. (c) Neutralization of normal virus receptor dependent binding and retargeted binding of NDFL-EGFP (10 HU per 107 cells) modified with aHN-IL-2. 1 HU of NDFL-EGFP was preincubated with different amounts of aHN-IL-2 for 1 hour on ice and then further incubated with 106 Jurkat or MT-2 cells for 1 hour at 371C. After unbound virus was washed away, the cells were stained with anti-F mAb and goat anti-mouse F(ab0 )2-PE before FACS analysis.

Cancer Gene Therapy

299

Selective gene transfer by recombinant NDV H Bian et al

300

NDFL+/ HN-IL-2

NDFL+ 52.2 %

39.7 %

44.6 %

4.0 %

anti-F scFv

Jurkat*

0.2%

1.5%

3.0%

30.0 %

63.9 %

1.7 %

54.8%

44.8 %

MT-2*

1.4 %

3.1 %

4.6%

50.4 %

CFSE Figure 4 Selective targeted virus entry in a mixture of Jurkat and MT-2 cells. Either Jurkat or MT-2 cells were labeled with CFSE before the two cell lines were mixed. Then the mixture of Jurkat and MT-2 cells was infected with native NDFL þ (both left histograms) or with modified NDFL þ / aHN-IL-2 (both right histograms) for 24 hours. Cells were stained with anti-F scFv, followed by anti-E Tag antibody and goat anti-mouse F(ab0 )2PE for FACS analysis. *indicates the cells labeled with CFSE.

a high safety profile. Measles virus (MV), a human negative-strand RNA virus of the family Paramyxoviridae, which binds to its receptor CD46 was genetically modified to target some novel cell surface molecules on tumor cells (e.g. epidermal growth factor,26 CD20,27 CD38,28 and human carcinoembryonic antigen29). By fusing scFv anti-CD20 antibody to the C-terminus of the hemagglutinin of an attenuated MV, Bucheit et al27 showed that the growth of CD20 þ tumor cells was retarded by the modified MV as compared with the nonmodified virus. Recent studies showed that NDV, an avian paramyxovirus can be genetically modified by insertion of foreign sequences by reverse genetics. This technology provides a promising novel RNA vector to deliver and express therapeutic genes for cancer gene therapy.8–10,23,30–32 We here describe a new procedure to alter the binding specificity of recombinant NDV to make it more tumorspecific for systemic application. This was achieved by using an aHN-IL-2 bispecific fusion protein that neutralizes the receptor binding site of the HN-protein and which introduces a new binding moiety, in this case the cytokine IL-2. Attachment of the fusion protein aHN-IL2 allowed the virus to bind to IL-2 receptor positive but not to IL-2 receptor negative tumor target cells. Selective binding to the IL-2 receptor enabled the virus to infect tumor cells expressing this receptor and to replicate within them efficiently, thereby leading to strong expression of the transgene EGFP.

Cancer Gene Therapy

Although NDV selectively replicates in tumor cells, its broad binding properties, including normal cells and tumor cells, is a disadvantage for systemic application. It is therefore necessary to limit the binding of NDV to tumor cells when NDV is administered systemically. In this regard, we developed this new strategy. Bispecific reagents have been used in different virus vectors, for example adenovirus,33 and adeno-associated virus.34 By employing a bispecific antibody consisting of an anti-Ad knob monoclonal Fab fragment conjugated with an anti-EGFR Ab, Miller et al35 achieved EGFR specific gene transfer and significantly enhanced adenoviral gene delivery in 7/12 established glioma cell lines. Others used a retroviral vector preloaded with a viral receptor–ligand bridge protein to target cells expressing EGFR.36 In this study, we report the first attempt to target NDV to IL-2R-positive cells via the bispecific fusion protein aHN-IL-2. IL-2Ra/p55 (Tac, CD25) is overexpressed on leukemia cells of almost all ATL patients and HTLV-1-infected T-cell lines.20,37 By complexing NDV with aHN-IL-2, we have blocked the natural cell-binding site of the HN protein while simultaneously supplying a binding alternative with IL2. This modified virus showed 2.2-fold stronger binding to IL-2R þ MT-2 cells compared to its natural binding behavior. Targeted virus entry through IL-2R could be blocked by preincubation of MT-2 cells with IL-2, confirming that the redirected NDV binding was specific for the IL-2R. Further assays, based on the addition of

Selective gene transfer by recombinant NDV H Bian et al

301

a

b Jurkat

MT-2

HN-IL-2

HN-IL-2

+

-

+

EGFP

c

I

-

-

EGFP HN-IL-2

+

II

Jurkat

III

IV

MT-2

Figure 5 Targeted gene delivery. Gene delivery of EGFP by NDFL-EGFP or NDFL-EGFP/aHN-IL-2 virus (100 HU per 107 cells) in Jurkat and MT-2 cells. For targeting purpose, 10 HU of NDFL-EGFP were preincubated with 12 mg/ml of aHN-IL-2 for 1 hour on ice, then further incubated with 106 Jurkat or MT-2 cells for 1 hour at 371C. After unbound virus was washed away, the cells were further cultured for 24 hours. The signal of EGFP fluorescence was measured by FACS. Cells incubated in serum-free RPMI-1640 medium (gray peak) served as negative control. Cells incubated with native NDFL-EGFP (thin black line) were used as positive control (a and b). In (a), 5.4% of IL-2R
Jurkat cells (bold black line) were infected by NDFL-EGFP/aHN-IL-2. In (b), 85.1% of IL-2R þ MT-2 cells (bold black line) were infected by NDFL-EGFP/aHN-IL-2. The percentage of cells with targeted virus infection was calculated as the ratio of the percentage of positive cells after infection with the modified virus to the percentage of positive cells after infection with the native virus. The results were confirmed by fluorescence microscopy (c). I: Jurkat þ NDFL-EGFP, II: Jurkat þ NDFL-EGFP/aHN-IL-2, III: MT-2 þ NDFL-EGFP, IV: MT-2 þ NDFL-EGFP/aHN-IL-2. Each picture was a merge of two images that were taken from the same field under normal light and under excitation of green fluorescence from the same field (  250 magnification).

NDV/aHN-IL-2 to a mixture of Jurkat and MT-2 cells, corroborated this approach of selective gene transfer to targeted tumor cells. The high efficiency of expression of EGFP in the infected tumor cells indicates that NDV could become an interesting new vector for delivering transgenes to tumors for the purpose gene therapy. Our previous studies demonstrated that human tumor cell modification by NDV infection was efficient and safe.1,38 It led to

upregulation of HLA and cell adhesion molecules, induction of IFN-a, -b and chemokines and, finally to apoptosis.14 In addition, NDV infection of tumor cells led to paracrine stimulation of high amounts of IFN-a production in human peripheral blood mononuclear cells via the cell surface-expressed viral HN protein15 and via the cytoplasmic viral dsRNA,39 which activates Toll-like receptor 3 (TL-R3).40 By targeting NDV with a therapeutic gene to the site of a tumor or of metastases,

Cancer Gene Therapy

Selective gene transfer by recombinant NDV H Bian et al

302

rhIL-2 + -

EGFP Figure 6 Specific entry of targeted NDV in MT-2 cells. A total of 106 MT-2 cells were preincubated with 5 mg of rhIL-2 for 1 hour on ice before the modified NDFL-EGFP/aHN-IL-2 virus was added. After unbound virus was washed away, the cells were cultured for 24 hours. The signal of EGFP fluorescence was measured by FACS. Gray peak: cells þ PBS, bold black line: cells þ rhIL-2 þ NDFLEGFP/aHN-IL-2, thin black line: cells þ NDFL-EGFP/aHN-IL-2.

synergistic therapeutic effects can be expected from the therapeutic gene product and from the immunostimulatory properties of this virus.

References 1. Schirrmacher V. Clinical trials of antitumor vaccination with an autologous tumor cell vaccine modified by virus infection: improvement of patient survival based on improved anti-tumor immune memory. Cancer Immunol Immunother. 2004 Oct 30. 2. Liang W, Wang H, Sun TM, et al. Application of autologous tumor cell vaccine and NDV vaccine in treatment of tumors of digestive tract. World J Gastroenterol. 2003;9:495–498. 3. Batliwalla FM, Bateman BA, Serrano D, et al. A 15-year follow-up of AJCC stage III malignant melanoma patients treated postsurgically with Newcastle disease virus (NDV) oncolysate and determination of alterations in the CD8T cell repertoire. Mol Med. 1998;4:783–794. 4. Schirrmacher V, Ahlert T, Probstle T, et al. Immunization with virus-modified tumor cells. Semin Oncol. 1998;25:677– 696. 5. Ahlert T, Sauerbrei W, Bastert G, et al. Tumor-cell number and viability as quality and efficacy parameters of autologous virus-modified cancer vaccines in patients with breast or ovarian cancer. J Clin Oncol. 1997;15:1354–1366. 6. Ockert D, Schirrmacher V, Beck N, et al. Newcastle disease virus-infected intact autologous tumor cell vaccine for adjuvant active specific immunotherapy of resected colorectal carcinoma. Clin Cancer Res. 1996;2:21–28. 7. Lorence RM, Pecora AL, Major PP, et al. Overview of phase I studies of intravenous administration of PV701, an oncolytic virus. Curr Opin Mol Ther. 2003;5:618–624.

Cancer Gene Therapy

8. Zhao H, Peeters BP. Recombinant Newcastle disease virus as a viral vector: effect of genomic location of foreign gene on gene expression and virus replication. J Gen Virol. 2003;84(Part 4):781–788. 9. Al-Garib SO, Gielkens AL, Gruys E, Peeters BP, Koch G. Tissue tropism in the chicken embryo of non-virulent and virulent Newcastle diseases strains that express green fluorescence protein. Avian Pathol. 2003;32:591–596. 10. Engel-Herbert I, Werner O, Teifke JP, Mebatsion T, Mettenleiter TC, Romer-Oberdorfer A. Characterization of a recombinant Newcastle disease virus expressing the green fluorescent protein. J Virol Methods. 2003;108:19–28. 11. Nagai Y, Hamaguchi M, Toyoda T. Molecular biology of Newcastle disease virus. Prog Vet Microbio Immunol. 1989;5:16–64. 12. Sinkovics JG, Horvath JC. Newcastle disease virus (NDV): brief history of its oncolytic strains. J Clin Virol. 2000; 16:1–15. 13. Huang Z, Elankumaran S, Panda A, Samal SK. Recombinant Newcastle disease virus as a vaccine vector. Poult Sci. 2003;82:899–906. 14. Washburn B, Schirrmacher V. Human tumor cell infection by Newcastle disease virus leads to upregulation of HLA and cell adhesion molecules and to induction of interferons, chemokines and finally apoptosis. Int J Oncol. 2002;21: 85–93. 15. Zeng J, Fournier P, Schirrmacher V. Induction of interferon-alpha and tumor necrosis factor-related apoptosisinducing ligand in human blood mononuclear cells by hemagglutinin-neuraminidase but not F protein of Newcastle disease virus. Virology. 2002;97:19–30. 16. Schirrmacher V, Griesbach A, Ahlert T. Antitumor effects of Newcastle disease virus in vivo: local versus systemic effects. Int J Oncol. 2001;18:945–952. 17. Barker SD, Dmitriev IP, Nettelbeck DM, et al. Combined transcriptional and transductional targeting improves the specificity and efficacy of adenoviral gene delivery to ovarian carcinoma. Gene Therapy. 2003;10:1198–1204. 18. Jongmans W, van den Oudenalder K, Tiemessen DM, et al. Targeting of adenovirus to human renal cell carcinoma cells. Urology. 2003;62:559–565. 19. Nicklin SA, Baker AH. Tropism-modified adenoviral and adeno-associated viral vectors for gene therapy. Curr Gene Ther. 2002;2:273–293. 20. Horiuchi S, Koyanagi Y, Tanaka Y, et al. Altered interleukin-2 receptor alpha-chain is expressed in human T-cell leukaemia virus type-I-infected T-cell lines and human peripheral blood mononuclear cells of adult T-cell leukaemia patients through an alternative splicing mechanism. Immunology. 1997;91:28–34. 21. Waldmann TA. The IL-2/IL-15 receptor systems: targets for immunotherapy. J Clin Immunol. 2002;22:51–56. 22. Zhang M, Zhang Z, Garmestani K, et al. Pretarget radiotherapy with an anti-CD25 antibody-streptavidin fusion protein was effective in therapy of leukemia/ lymphoma xenografts. Proc Natl Acad Sci USA. 2003; 100:1891–1895. 23. Peeters BP, de Leeuw OS, Koch G, Gielkens AL. Rescue of Newcastle disease virus from cloned cDNA: evidence that cleavability of the fusion protein is a major determinant for virulence. J Virol. 1999;73:5001–5009. 24. Russell SJ. RNA viruses as virotherapy agents. Cancer Gene Ther. 2002;9:961–966. 25. Palese P, Zheng H, Engelhardt OG, Pleschka S, GarciaSastre A. Negative-strand RNA viruses: genetic engineering

Selective gene transfer by recombinant NDV H Bian et al

26.

27.

28.

29.

30.

31.

32.

33.

and applications. Proc Natl Acad Sci USA. 1996;93: 11354–11358. Schneider U, Bullough F, Vongpunsawad S, Russell SJ, Cattaneo R. Recombinant measles viruses efficiently entering cells through targeted receptors. J Virol. 2000;74: 928–936. Bucheit AD, Kumar S, Grote DM, et al. An oncolytic measles virus engineered to enter cells through the CD20 antigen. Mol Ther. 2003;7:62–72. Peng KW, Donovan KA, Schneider U, Cattaneo R, Lust JA, Russell SJ. Oncolytic measles viruses displaying a singlechain antibody against CD38, a myeloma cell marker. Blood. 2003;101:2557–2562. Hammond AL, Plemper RK, Zhang J, Schneider U, Russell SJ, Cattaneo R. Single-chain antibody displayed on a recombinant measles virus confers entry through the tumor-associated carcinoembryonic antigen. J Virol. 2001; 75:2087–2096. Krishnamurthy S, Huang Z, Samal SK. Recovery of a virulent strain of newcastle disease virus from cloned cDNA: expression of a foreign gene results in growth retardation and attenuation. Virology. 2000;278:168–182. Romer-Oberdorfer A, Mundt E, Mebatsion T, Buchholz UJ, Mettenleiter TC. Generation of recombinant lentogenic Newcastle disease virus from cDNA. J Gen Virol. 1999; 80(Part 11):2987–2995. Huang Z, Krishnamurthy S, Panda A, Samal SK. High-level expression of a foreign gene from the most 30 -proximal locus of a recombinant Newcastle disease virus. J Gen Virol. 2001;82(Part 7):1729–1736. Nettelbeck DM, Miller DW, Jerome V, et al. Targeting of adenovirus to endothelial cells by a bispecific single-chain

34.

35.

36.

37.

38.

39.

40.

diabody directed against the adenovirus fiber knob domain and human endoglin (CD105). Mol Ther. 2001;3: 882–891. Bartlett JS, Kleinschmidt J, Boucher RC, Samulski RJ. Targeted adeno-associated virus vector transduction of nonpermissive cells mediated by a bispecific F(ab0 gamma)2 antibody. Nat Biotechnol. 1999;17:181–186. Miller CR, Buchsbaum DJ, Reynolds PN, et al. Differential susceptibility of primary and established human glioma cells to adenovirus infection: targeting via the epidermal growth factor receptor achieves fiber receptor-independent gene transfer. Cancer Res. 1998;58: 5738–5748. Boerger AL, Snitkovsky S, Young JA. Retroviral vectors preloaded with a viral receptor-ligand bridge protein are targeted to specific cell types. Proc Natl Acad Sci USA. 1999;96:9867–9872. Uchiyama T, Hori T, Tsudo M, et al. Interleukin-2 receptor (Tac antigen) expressed on adult T cell leukemia cells. J Clin Invest. 1985;76:446–453. Schirrmacher V, Feuerer M, Fournier P, Ahlert T, Umansky V, Beckhove P. T-cell priming in bone marrow: the potential for long-lasting protective anti-tumor immunity. Trends Mol Med. 2003;9:526–534. Fournier P, Zeng J, Schirrmacher V. Two ways to induce innate immune responses in human PBMCs: paracrine stimulation of IFN-alpha responses by viral protein or dsRNA. Int J Oncol. 2003;23:673–680. Alexopoulou L, Holt AC, Medzhitov R, Flavell RA. Recognition of double-stranded RNA and activation of NF-kappaB by Toll-like receptor 3. Nature. 2001;413: 732–738.

Cancer Gene Therapy

303