A simplified cloning strategy for the generation of an endothelial cell selective recombinant adenovirus vector

Journal of Virological Methods 135 (2006) 127–135

Short communication

A simplified cloning strategy for the generation of an endothelial cell selective recombinant adenovirus vector Chitladda Mahanivong, J¨org A. Kr¨uger, Dafang Bian, Ralph A. Reisfeld, Shuang Huang ∗ Department of Immunology, The Scripps Research Institute, 10550 North Torrey Pines Rd., La Jolla, CA 92037, USA Received 17 November 2005; received in revised form 15 February 2006; accepted 22 February 2006 Available online 3 April 2006

Abstract Specifically targeting adenoviral vectors to particular cell/tissue types can be achieved by genetically modifying the adenovirus fiber protein. Two common strategies are: (1) directly modifying the fiber gene in the adenovirus genome and (2) in trans supply of the modified fiber. The former however, suffers from difficulties in directly manipulating large adenoviral genomic DNA. Although the latter allows easy manipulation of the small fiber gene, our studies show that the in trans supplement of the modified fiber causes incomplete fiber assimilation in the virus. Thus an alternate cloning strategy was devised to facilitate the insertion of cell-targeting sequences into the HI loop of a CAR binding-ablated fiber gene in the Ad5 genomic backbone. Our approach retains the advantage of easily modifying the fiber with the additional benefit of genetic re-insertion into the Ad genomic backbone to ensure complete modified fiber incorporation. Using this strategy, an endothelial cell binding peptide sequence (Asn-Gly-Arg) was introduced into the Ad fiber and showed that the generated Ad vector displayed selective transduction of endothelial cells both in vitro and in vivo compared to the conventional vector. Furthermore, this Ad vector cloning strategy can be adapted to introduce other peptide sequences to target other cell types. © 2006 Elsevier B.V. All rights reserved. Keywords: Adenovirus vector; Fiber; Re-targeting; Recombination; Endothelium

Recombinant adenoviruses (Ad) have a number of important characteristics that make them useful as gene transfer tools for applications in gene therapy (McConnell and Imperiale, 2004; Mizuguchi and Hayakawa, 2004). They are easy to propagate and can transduce a wide range of target cells with high efficiency. Their genomes can accommodate large fragments of DNA (Bett et al., 1993), the viral genome is maintained epichromosomally through successive rounds of replication and does not affect host genes (Kovesdi et al., 1997). These advantageous features have led to the rapid development of recombinant Ad from the laboratory to the clinic. But the efficacies observed in the laboratory have not translated as successfully in clinical applications. One of the major reasons is the general dissemination of Ad gene transfer to non-target tissue regions thereby reducing therapeutic effects and causing vector-induced immunogenic toxicities (Liu and Muruve, 2003; Russell, 2000). In order to address this problem, there has been increasing

∗

Corresponding author. Tel.: +1 858 784 9211; fax: +1 858 784 8472. E-mail address: [email protected] (S. Huang).

0166-0934/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.jviromet.2006.02.010

studies directed at manipulating the viral structure to alter viral tropism to achieve selective transduction (Barnett et al., 2002; Mizuguchi and Hayakawa, 2004; Everts and Curiel, 2004; Einfeld and Roelvink, 2002). Ad infection is initiated through the interaction of the viral fiber protein and its principal cell surface coxsackie and adenovirus receptor (CAR) (Bergelson et al., 1997; Bewley et al., 1999; Roelvink et al., 1999). As such, the fiber structure is a primary candidate for genetic modification to redirect its interaction to alternate cell surface receptors. Initial modifications of the fiber involved the assembly of chimeric Ad, by swapping the fiber proteins between Ad serotypes (Stevenson et al., 1997; Gall et al., 1996; Shayakhmetov et al., 2000; Havenga et al., 2001). However the resultant Ad tropism is dictated and ultimately limited by the chimeric fiber. Subsequent generation of fiber modifications involved the introduction of targeting ligands into the C-terminal peripheral end of the fiber structure (Wickham et al., 1996; Michael et al., 1995). Elucidation of the crystal structure of the fiber C-terminal knob domain (Xia et al., 1994) revealed a flexible HI loop (connects ␤-strands H and I) segment that is exposed away from the knob, suggesting a possible position for ligand

128

C. Mahanivong et al. / Journal of Virological Methods 135 (2006) 127–135

addition. Krasnykh et al. (1998) introduced a FLAG octapeptide into the HI loop and demonstrated that it did not affect fiber trimerization or function and the FLAG ligand was accessible to interact with an anti-FLAG monoclonal antibody. Ruoslahti and co-workers (Ruoslahti, 1996; Koivunen et al., 1995) showed peptides containing an RGD motif binds cell surface integrins with high affinity and RGD modified Ad significantly improve transduction of many cancers including cervical (Rein et al., 2004), ovarian (Dmitriev et al., 1998), glioma (Koizumi et al., 2001), breast carcinoma and melanoma (Pasqualini et al., 1997) that express low levels of CAR and are generally refractory to wild type Ad infection. Similarly Nicklin et al. (2000, 2001, 2004) incorporated two peptide sequences (SIGYPLP and MSLTTPPVARP) into the fiber HI loop and modified Ad vectors showed selective targeting to quiescent human saphenous vein endothelial cell and human saphenous vein vascular smooth muscle cells. Thus peptide introduction into the HI loop to alter viral tropism presents a promising approach to generate stable fiber modifications. Angiogenesis, the formation of new vasculature plays a key role in the growth and metastasis of solid tumors (Folkman, 1995; Bergers and Benjamin, 2003). These blood vessels are generally comprised of endothelial cells and the angiogenic endothelium overexpresses markers that are present in low abundance in resting quiescent endothelial cells (Carmeliet, 2003; Cleaver, 2003). Researchers have capitalized on this information to develop treatments that can be targeted to endothelial cells to inhibit angiogenesis and ultimately reduce disease progression. A number of peptide sequences have been identified to interact with tumor-associated vascular endothelial cells. They include: (1) CDCRGCCFC (RGD) (Arap et al., 1998), (2) CNGRCVSGCAGRC (NGR) (Arap et al., 1998), (3) CGSLVRC (GSL) (Arap et al., 1998), (4) TCDLDNDKYIALEEWAGCFG (SPARC) (Kupprion et al., 1998) and (5) CSCKNTDSRCKARQLELNERTCRC (HBDt) (El-Sheikh et al., 2002). RGD, GSL and NGR were identified using phage display peptide library to select for sequences that targeted tumor angiogenic blood vessels. SPARC is a 20 amino acid sequence derived from a matricellular secreted protein, acidic and rich in cysteine (SPARC) known to bind endothelial cells. HBDt derives from the heparinbinding domain of the vascular endothelial growth factor that also localizes to the vascular endothelium. It was also of interest to examine if these peptides could facilitate Ad targeting to human endothelial cells. Given the increasing reports of success with modifying the Ad fiber protein to alter viral tropism, this strategy was chosen to develop an endothelial cell-targeted Ad vector. A number of protocols have been described in literature for the construction of fiber modified Ad (Belousova et al., 2002; Krasnykh et al., 1998; Stevenson et al., 1997). Most of these methods involve complicated and extensive cloning and PCR steps involving numerous plasmid constructs. One method involved inserting oligonucleotide sequences directly into the fiber gene of the virus genomic backbone (Mizuguchi et al., 2001), however genetic manipulation of large viral DNA is technically challenging. Preliminary attempts to generate cloning sites in the Ad genomic fiber gene resulted in difficulties in ligating large DNA frag-

ments containing the 34 kbp viral genomic DNA. As such, the trans-supply strategy originally developed by Von Seggern et al. (1998) was used. In this approach, a fiber-deleted Ad genome is constructed and the fiber protein is supplied in trans by a plasmid containing the modified fiber gene. The fiber-encoded plasmid is transiently transfected into mammalian cells, followed by a subsequent infection with a fiberless Ad vector (Jakubczak et al., 2001; Nicklin et al., 2001). During the packaging process, the modified fiber proteins are incorporated into the progeny Ad. The key advantage to this strategy is the relative ease of mutating the fiber sequence (small) rather than manipulating the Ad genome, which is technically difficult. The protocol involves the initial synthesis of oligonucleotides encoding the peptide sequences into the HI loop of the fiber gene in the plasmid pDV137. The pDV137 vector previously developed by Nicklin et al. (2001) encoded a modified version of the Ad5 fiber that contained a 48 bp linker (with a BspEI unique cloning site) in the HI loop and contained mutations (S408E and P409A) in the CAR recognition sites. The plasmids were transiently transfected into HEK293 cells and expression of the modified fibers confirmed (data not shown). The cells were then infected with ␤-galactosidase containing fiber-deleted Ad vector (Nicklin et al., 2001) (Ad
F) and the modified fibers incorporated into the progeny virus during packaging. The infectivities of these packaged virus stocks were compared in HeLa (nonendothelial cell) and HUVEC cells by incubating the cells with increasing virus concentrations (105 –106 VP/cell) for 16 h. Cells were then washed, incubated a further 24 h and then assayed for the ␤-galactosidase expression to determine transduction efficiencies of these Ads (Fig. 1A). When 106 VP/cell concentration of Ad
F.NGR was added, a two-fold greater transduction of HUVEC than HeLa cells was observed. SPARC and HBDt containing viruses showed moderate enhancements in Ad transduction, whereas RGD and GSL containing viruses showed only minor differences in their transduction between HUVEC and HeLa cells. These results demonstrate that incorporation of the NGR peptide could selectively target the Ad vector to endothelial cells. However, the high viral concentrations (106 VP/cell) of Ad
F.NGR required in these experiments to achieve an efficient level of transduction was a concern, particularly as a two-order of magnitude lower concentration of a wild-type virus was sufficient to attain significant transduction in HUVEC (Fig. 1B). To explain the poor infectivity of Ad
F.NGR, the viral capsid fiber and penton proteins (two proteins involved in the Ad entry into host cells) were analysed by immunoblotting (Fig. 1C). Samples of control Ad.LacZ (genomically encoded fiber) and Ad
F.NGR was electrophoresced on a 12% SDS-PAGE gel, then immunoblotted and analysed with the respective polyclonal antibodies (Fig. 1C). The results showed that Ad
F.NGR contained significantly less fiber content compared to Ad.LacZ. However the genomically encoded penton base protein showed identical levels of expression between both viruses. This suggested that the in trans addition of the mutant fibers was inefficient and the majority of the Ad
F.NGR preparation contained viral particles with either incomplete or no fiber proteins. It thus explained why very high viral concentrations were necessary for the cell transduction experiments (Fig. 1A). In a further effort

C. Mahanivong et al. / Journal of Virological Methods 135 (2006) 127–135

129

Fig. 1. Transduction assays using recombinant viruses containing modified fibers. (A) Cervical carcinoma HeLa (closed squares) and primary human vascular endothelial HUVEC (open squares) cells were seeded (1 × 104 cells/well) into 24-well plates with varying doses of CsCl purified (Green and Pina, 1964) recombinant viruses (that carried the fiber produced in trans) modified with NGR, RGD, SPARC, GSL or HBDt for 16 h. The cells were washed and fresh complete media added and incubation continued for another 24 h. The cells were washed with PBS, fixed with 4% paraformaldehyde for 30 min and then treated with X-gal staining solution in PBS (0.5 mg/ml X-gal (DMSO), 3 mM potassium ferricyanide, 3 mM potassium ferrocyanide, 40 mM HEPES pH 7.4, 15 mM NaCl, 1.3 mM MgCl2 ) for 16 h. Cells that were transduced, expressed ␤-galactosidase activity and developed a blue coloring. Transduction efficiencies were estimated by the percentage of blue cells counted from three microscopic views. All experiments were performed in duplicates and repeated on three separate occasions and averages from the data are represented. (B) Comparison of transduction efficiencies of Ad.LacZ (solid circles) and Ad
F.NGR (solid triangles) viruses for HUVEC. Transduction assays were conducted and percentage transduction calculated as described above. (C) Western analysis comparing fiber and penton base expression of purified control Ad.LacZ and Ad
F.NGR viruses. CsCl purified virus (1 ␮g) was subjected to SDS-PAGE (12% gel) and immunoblot analysis. The fiber and penton base protein expression were detected with polyclonal antibodies against the respective structural proteins as described previously (Nicklin et al., 2001).

to solve this problem, stable cell lines were generated to uniformly express the modified fibers, but this did not alleviate the problem of deficient fiber incorporation (data not shown). And so, although this protocol was successful for the incorporation of some peptides (Nicklin et al., 2004; Jakubczak et al., 2001), comparable fiber incorporation with the NGR peptide was not achieved and as a result the infectivity of these mutant viruses were significantly reduced. Although some success was obtained with retargeting using the method of peptide incorporation into the HI loop of the fiber, the in trans fiber incorporation strategy was inefficient and hampered the production of viral particles with correct fiber content. So a new strategy was developed that would facilitate the incorporation of a peptide sequence directly into the genomic backbone of the Ad fiber and would eliminate the need to sup-

ply the fiber in trans so that complete fiber incorporation into the new viral particles can be ensured. Genetic manipulation of the fiber is conducted with a small plasmid construct prior to a double recombination event to incorporate the mutant fiber into the viral genome. The efficacy of this approach was demonstrated by generating an Ad vector with NGR expressing fiber proteins. The cloning strategy involves two simple cloning steps followed by a recombination step to incorporate the NGR peptide into the CAR binding-ablated fiber gene of the pAdEasy genomic backbone and can be adapted to easily enable the insertion of any targeting peptide directly into the HI loop of the Ad fiber. By using the pAdEasy vector it was possible to take advantage of the pAdEasy cloning system (Stratagene, La Jolla, CA) involving the recombination with pShuttle.CMV to facilitate foreign gene expression. In this

130

C. Mahanivong et al. / Journal of Virological Methods 135 (2006) 127–135

Fig. 2. Schematic of the cloning strategy for the insertion of the NGR sequence into pAdEasy. (A) The generation of pDV137, containing the Ad fiber defective in CAR binding and modified to enable a peptide insertion into the HI loop has been previously described (Nicklin et al., 2001). Briefly, site-directed mutagenesis was conducted to introduce a unique BspEI site (red) into the HI loop (between nucleotide base positions 1445 and 1446) of the 1581 bp fiber gene. The complementary NGR5 and NGR3 oligonucleotide sequences with the BspEI overhangs (red) are indicated along with the encoded the 13 amino acid NGR peptide sequence. (B) I, The NGR5/3 oligonucleotides were annealed and subcloned into the BspEI-linearised pDV137. (II) This clone (pDV137.NGR) was then digested with BsrDI (generates variable single-stranded overhangs). (III) The 1.3 kbp fragment containing the fiber/NGR modified sequence was subcloned into the corresponding region in pCM11 to generate pCM12. pCM11 contains a 6.2 kbp SpeI/PacI [position 27234–33437 bp (based on the Stratagene sequence)] fragment from the wild type pAdEasy plasmid that encompasses the fiber gene cloned into pCR-Blunt II-TOPO (Invitrogen, Carlsbad, CA). (IV) Next, we devised a strategy that relies on a double recombination event between a linearised pAdEasy and a DNA fragment containing the fiber/NGR in order to incorporate the NGR modified fiber gene into the pAdEasy backbone. pAdEasy was linearised with a unique cutter SrfI (position 12679 bp) (Stratagene) and gel purified. The second DNA fragment was a gel purified 6.3 kbp SpeI/PacI from pCM12 containing fiber/NGR. A mixture of 750 ng of each linearised DNA was co-electroporated into recombination proficient E. coli strain BJ5183; V. A double recombination event is necessary to give rise to ampicillin resistant pAdEasy clones. The position of the second recombination event 3 or 5 to the NGR sequence would result in NGR incorporation or lack thereof respectively. PCR and DNA sequence analyses were used to identify correct pAdEasy clones containing the NGR insertion.

C. Mahanivong et al. / Journal of Virological Methods 135 (2006) 127–135

study, the ␤-galactosidase gene was incorporated, but a similar cloning strategy could be used to incorporate a therapeutic agent and thereby generating a targeted recombinant Ad for therapeutic applications. Details of this new approach are outlined in the schematics in Fig. 2. The initial basis involves a plasmid construct pDV137, which contains the Ad fiber mutated to include a unique BspEI restriction site in the HI loop to facilitate cloning of a sequence to enable retargeting of the Ad vector (Nicklin et al., 2001). A schematic depicting the introduced BspEI linker sequence in the fiber is shown in Fig. 2A. An oligonucleotide sequence of the NGR peptide (Fig. 2A) was initially cloned into the fiber gene in plasmid pDV137 (Fig. 2A and B-I) and then subcloned into a second plasmid (pCM11) that encoded a larger flanking sequence encompassing a 6.2 kbp SpeI/PacI sequence of pAdEasy (Fig. 2B-II). In this step, pDV137.NGR was digested with BsrDI and the 1.3 kbp fragment containing NGR was religated (3 fragment ligation) into the identical position of pCM11 (thus replacing the wild type HI loop sequence) to generate pCM12 (Fig. 2B-III). The BsrDI restriction sequence is nonpalindromic (CGTTAC↑NN) and digestion of pCM11 generated three fragments with unique two base overhangs, which ensured the three fragments, ligate correctly. pAdEasy contains unique SpeI and PacI restriction sites and initially we attempted to clone the 6.3 kbp SpeI/PacI (containing NGR) fragment from pCM12 directly into a 27.5 kbp linearised SpeI/PacI digested pAdEasy (containing the remaining genome). Over 1000 colonies were screened (data not shown), but did not yield a complete clone of pAdEasy.NGR. So instead an alternative method was devised utilizing the E. coli bacterial recombination system (Chartier et al., 1996; Bubeck et al., 1993). As depicted in Fig. 2B-IV, pAdEasy was linearised with SrfI and co-electroporated with the 6.3 kbp SpeI/PacI DNA fragment (from pCM12) containing the NGR sequence into the recombination proficient E. coli strain BJ5183. The flanking DNA (3 kbp upstream and 3.2 kbp downstream) surrounding the NGR sequence in the 6.3 kbp fragment contained homologous sequence with pAdEasy and a double recombination event was necessary to give rise to transformants containing complete pAdEasy genomes (Fig. 2B-V). The initial recombination event had to occur between the SpeI and SrfI sites and the position of the second recombination event would give rise to the wild type fiber gene (if the event occurred upstream of the NGR sequence) or the NGR modified fiber (if the event occurred downstream of the NGR sequence). The transformation in BJ5183 yielded 43 colonies and restriction digestion was conducted to analyze the plasmid DNA from all the colonies (data not shown). Electroporation of the SrfI-digested pAdEasy DNA alone yielded no colonies. PCR and DNA sequence analyses identified 12 correct pAdEasy clones containing the NGR insertion. Thus approximately 28% of the clones contained the desired NGR peptide insert and the remaining clones contained unmodified fiber sequence. One of these correct clones, labeled pAdEasy.NGR and used as a basis for the construction of an adenovirus expressing the ␤-galactosidase reporter protein according to the pAdEasy vector system protocol (Stratagene, CA). Briefly, a NotI fragment containing the ␤-galactosidase gene

131

Fig. 3. Virus profiles comparing Ad.NGR and Ad.LacZ. (A) CsCl purified virus (2 ␮g) was boiled in SDS loading buffer prior to electrophoresis in a 12% SDS-PAGE gel and immunoblotted to detect the fiber and penton base capsid proteins with their respective polyclonal antibodies (Nicklin et al., 2001). (B) CsCl purified virus (50 ␮g) was precipitated by centrifugation (30,000 × g) in a Beckman airfuge for 30 min and boiled in SDS loading buffer containing 2% ␤-mercaptoethanol, prior to electrophoresis in a 12% SDS-PAGE gel under reducing conditions. The gel was stained with Coomassie blue and destained overnight in 15% acetic acid. The major viral proteins are indicated on the right hand side. The Benchmark Pre-stained (Invitrogen) protein ladder (kDa) is shown on the left hand side.

was subcloned from pRSV␤-gal (Promega) into the NotI site of the transfer vector pShuttle-CMV (Stratagene). The orientation and sequence of the positive clones (named pShuttle.␤-gal) were confirmed by DNA sequencing and purified DNA was isolated and digested with PmeI and co-electroporated with pAdEasy.NGR into BJ5183. Recombinants are selected with kanamycin and analysed by restriction digestion. The correct recombinant adenoviral clone, labeled Ad.NGR was then transfected into HEK293 cells for the generation of infectious viral particles (Ad.NGR) that: (1) encoded the NGR modified fiber, (2) was deleted of CAR recognition signals and (3) expressed the assayable ␤-galactosidase reporter gene. To determine the viral fiber content, samples of the purified Ad.NGR virus was analysed by immunoblot (Fig. 3A) and Coomassie staining (Fig. 3B). The immunoblot results show that the Ad.NGR contained a similar fiber and penton base protein content compared to the wild type Ad.LacZ virus (Fig. 3A). The Coomassie stain of the viruses indicated identical protein profiles between the wild type virus and the fiber modified virus (Fig. 3B). These results confirm that using our cloning strategy, the NGR sequence could be conveniently and successfully incorporated into the Ad genomic fiber

132

C. Mahanivong et al. / Journal of Virological Methods 135 (2006) 127–135

Fig. 4. Viral transduction in vitro. Lung carcinoma A549, hepatic carcinoma Hep2G, cervical carcinoma HeLa, primary human endothelial HUVEC and HMVEC (Cascade Biologics, Inc., Portland, Oregon) cells (1 × 104 cells/well) were seeded in 24 well plates overnight. Various concentrations of Ad.LacZ and Ad.NGR virus were added to the cells for 16 h. The cells were then washed and fresh media added and incubation continued for a further 24 h prior to fixing with paraformaldehyde and X-gal staining conducted as described previously. (A) Representative microscopic (×100) images of Ad.LacZ (top panels) and Ad.NGR (bottom panels) [3 × 103 virus particles/cell] transduction of the five cell lines. (B) The calculated percentage transduction of Ad.LacZ and Ad.NGR (104 VP/cell) in A549, HeLa, Hep2G, HUVEC and HMVEC cells. (C) HUVEC or HMVEC were treated with increasing doses of Ad.NGR and the percentage transduction efficiencies calculated as described earlier.

sequence and functional viral particles expressing peptide modified fiber proteins could be generated. In subsequent experiments, the transducibility of Ad.NGR for various cell lines was examined to determine whether the introduced NGR peptide conferred selectivity for endothelial cells. The Ad.NGR and Ad.LacZ viral (104 viral particles/cell) transduction was tested on five cell types, A549, HeLa, Hep2G cells and endothelial HUVECs and HMVECs (Fig. 4A and B). Cell lines A549 and HeLa express high levels of CAR and are known to be efficiently transduced by wild type Ad and this was evident by the high transduction efficiency with Ad.LacZ. The Ad.NGR virus however demonstrated significantly reduced transduction efficiency of 1, 9 and 18% in A549, HeLa and Hep2G cells, respectively, indicating the CAR binding-ablated Ad.NGR lacked the ability to use CAR and so poorly transduced these three cell lines. Unlike the Ad.LacZ control virus which transduced all the cell lines test, the Ad.NGR showed selectivity for transducing endothelial (approximately 80%) HUVEC and HMVEC cells (Fig. 4A and B). The addition of increasing concentrations of Ad.NGR to HUVEC or HMVEC cells confirmed the modified virus demonstrated similar levels of transduction in both endothelial cell types (Fig. 4C). These results demonstrate that the NGR modified virus exhibited significant selectivity for endothelial cells compared to the conventional Ad.LacZ virus that transduced all cell types equally well.

Next investigations of whether the NGR modified Ad also conferred selectivity to endothelial cells in vivo were conducted. A matrigel plug model system in BALB/c mice (Niethammer et al., 2002) was used to induce neovasculature and then assayed the plugs to ascertain transduction efficacy after systemic application with control or Ad.NGR viruses. A solution containing growth factor reduced matrigel (BD Biosciences) and murine fibroblast growth factor 2 was injected subcutaneously into the mice (n = 4) underbelly and allowed 10 days for the recruitment of endothelial cells to form new blood vessels in the matrigel substrate. Subsequently, Ad.LacZ or Ad.NGR (2 × 1010 or 1 × 1011 VP/0.1 ml each) virus or 0.1 ml PBS was injected by intravenous (i.v.) injection into the lateral tail vein. Three days later, lectin-FITC (0.15 ml) was i.v. injected and 30 min later the mice were sacrificed and matrigel plugs harvested. Samples of the virus infected matrigel plugs were analysed by activitybased assay and immunohistochemistry (Fig. 5). The addition of 2 × 1010 or 1 × 1011 viral particles resulted in more than 3.5-fold greater levels of Ad.NGR transduction in the matrigel endothelium compared to the Ad.LacZ control (Fig. 5A). In these assays, ␤-galactosidase activity was normalized against the lectin-FITC fluorimetry reading (at 490 nm), which was a representative measure of endothelial cell content and more likely corresponds to a more accurate measurement than the total protein content. The results of the ␤-galactosidase activity assays thus indicated

C. Mahanivong et al. / Journal of Virological Methods 135 (2006) 127–135

133

Fig. 5. Viral transduction in vivo. Five groups of BALB/c mice (n = 4) were implanted subcutaneously in the lower abdominal region with 0.4 ml growth factor reduced matrigel (BD Biosciences) containing 0.4 ␮g/ml murine fibroblast growth factor 2 (PeproTech, Rocky Hill, NJ). Seven days later, Ad.LacZ or Ad.NGR (2 × 1010 or 1 × 1011 VP/0.1 ml each) or 0.1 ml PBS was injected by intravenous (i.v.) injection into the lateral tail vein. Three days later, 0.15 ml of fluorescent Bandeiraea Simplicifolia lectin I, Isolectin B4 at 0.1 mg/ml (Vector Laboratories) was i.v. injected to stain endothelial tissue (Xiang et al., 2005). The animals were sacrificed 30 min later and matrigel plugs excised and dissected in half. One portion was assayed for ␤-galactosidase activity and the remaining portion was placed in Tissue-Tek OCT compound and snap frozen for histology. (A) ␤-Galactosidase activity assay of matrigel extracts from mice infected with (2 × 1010 or 1 × 1011 viral particles of) Ad.LacZ or Ad.NGR. The matrigel samples were homogenized in 0.4 ml RIPA lysis buffer (0.15 mM NaCl/0.05 mM Tris pH 7.2/1% Triton-X 100/1% sodium deoxycholate/0.1% SDS). The lysate was assayed for lectin-FITC by fluorimetry at 490 nm (Tecan GENios Pro, Switzerland) as a measure of endothelial tissue and the ␤-galactosidase activity assayed for viral transduction using the Galacto-light Plus chemiluminesence kit (Applied Biosystems). Viral transduction was expressed as ␤-galactosidase activity/lectin fluorescence. Data are presented as the mean ± S.E. (n = 4 in each group). * p < 0.05 vs. Ad.LacZ. (B) Cryostat sections of the OCT frozen matrigel samples prepared on microscope slides were fixed (2% formaldehyde/0.2% gluteraldehyde in PBS) for 10 min, flushed with PBS and then incubated with 2.5% BSA in PBS for 10 min to block non-specific reactions. After washing with PBS, the matrigel cryo-sections were incubated with 1:300 dilution of rat anti-mouse biotin-CD31 antibody (BD Biosciences) for 2 h at room temperature. The sections were then rinsed thoroughly with PBS and incubated with 1:300 dilution of HRP–Streptavidin (BD Biosciences) for a further 2 h at room temperature. This was followed by a wash with PBS, then the X-gal staining solution (described earlier) was applied and the slides incubated on a moistened paper towel in a covered Petri-dish overnight at room temperature. Finally, the matrigel samples were rinsed with water prior to 5 min incubation with a peroxidase substrate chromogen solution (DAKO AEC Substrate System). A light microscope was used to visualize brown staining indicative of EC vascularization and blue coloring which was indicative of viral transduction. Magnification ×100.

that the Ad.NGR virus more effectively transduced the matrigel vasculature than the conventional virus. In order to confirm this data, sections of the matrigel plug were analysed by immunohistochemistry using the CD31 antibody to stain endothelial cells (reddish-brown) and X-gal staining to detect viral transduction (blue). Representative slides are shown in Fig. 5B. In the samples treated with Ad.NGR, the blue staining indicative of viral transduction co-localized with the reddish-brown stained endothelial cells (Fig. 5B, top panels). This demonstrated that

the Ad.NGR indeed delivered ␤-galactosidase gene expression in the endothelium of the matrigel substrate. In contrast, little or no specific blue staining was observed in slide samples from animals treated with Ad.LacZ (Fig. 5B, bottom panels). The absence of Ad.LacZ in the matrigel endothelium may be a result of a dilution of the control Ad by the surrounding tissues en route to the endothelium due to is ability to bind a wide spectrum of cell types and so fewer particles reach the matrigel substrate. An alternative explanation for the observed poor transduction

134

C. Mahanivong et al. / Journal of Virological Methods 135 (2006) 127–135

of the control virus for the matrigel vasculature may be that the murine derived endothelium lacks sufficient/or appropriate surface receptors for conventional Ad vectors. The development of targeted delivery of therapeutic gene expression is necessary to improve the efficacy of genetic treatments. This study aimed to capitalize on the advantages of Ad vectors as gene transfer tools, by modifying the vector in order to generate targeted transduction of gene expression. In order to achieve this, a simplified and convenient cloning strategy was developed for the genetic manipulation of the Ad genome to facilitate incorporation of a targeting peptide into the fiber gene in order to modify viral tropism for selective cellular transduction. The efficacy of this approach was demonstrated by generating an Ad vector with NGR expressing fiber proteins for the goal of creating an angiogenic endothelial cell-targeting Ad vector. Such vectors are highly desired for their potential in targeting the endothelium and to deliver gene therapies to treat diseases that cause excessive angiogenesis, such as cancer, arthritis and atherosclerosis. Acknowledgments We would like to thank Dan Von Seggern for his consultation and providing us with plasmid constructs and Joan Gausepohl for her administrative help with the preparation of this manuscript. This is manuscript #17259 from The Scripps Research Institute. References Arap, W., Pasqualini, R., Ruoslahti, E., 1998. Cancer treatment by targeted drug delivery to tumor vasculature in a mouse model. Science 279, 377–380. Barnett, B.G., Crews, C.J., Douglas, J.T., 2002. Targeted adenoviral vectors. Biochim. Biophys. Acta 1575, 1–14. Belousova, N., Krendelchtchikova, V., Curiel, D.T., Krasnykh, V., 2002. Modulation of adenovirus vector tropism via incorporation of polypeptide ligands into the fiber protein. J. Virol. 76, 8621–8631. Bergelson, J.M., Cunningham, J.A., Droguett, G., Kurt-Jones, E.A., Krithivas, A., Hong, J.S., Horwitz, M.S., Crowell, R.L., Finberg, R.W., 1997. Isolation of a common receptor for coxsackie B viruses and adenoviruses 2 and 5. Science 275, 1320–1323. Bergers, G., Benjamin, L.E., 2003. Tumorigenesis and the angiogenic switch. Nat. Rev. Cancer 9, 661–668. Bett, A.J., Prevec, L., Graham, F.L., 1993. Packaging capacity and stability of human adenovirus type 5 vectors. J. Virol. 67, 5911–5921. Bewley, M.C., Springer, K., Zhang, Y.B., Freimuth, P., Flanagan, J.M., 1999. Structural analysis of the mechanism of adenovirus binding to its human cellular receptor. Science 286, 1579–1583. Bubeck, P., Winkler, M., Bautsch, W., 1993. Rapid cloning by homologous recombination in vivo. Nucleic Acids Res. 21, 3601–3602. Carmeliet, P., 2003. Angiogenesis in health and disease. Nat. Med. 9, 653–660. Chartier, C., Degryse, E., Gantzer, M., Dieterl´e, A., Pavirani, A., Mehtali, M., 1996. Efficient generation of recombinant adenovirus vectors by homologous recombination in Escherichia coli. J. Virol. 70, 4805–4810. Cleaver, O.M.D.A., 2003. Endothelial signaling during development. Nat. Med. 9, 661–668. Dmitriev, I., Krasnykh, V., Miller, C.R., Wang, M., Kashentseva, E., Mikheeva, G., Belousova, N., Curiel, D.T., 1998. An adenovirus vector with genetically modified fibers demonstrates expanded tropism via utilization of a coxsackievirus and adenovirus receptor-independent cell entry mechanism. J. Virol. 72, 9706–9713.

Einfeld, D.A., Roelvink, P.W., 2002. Advances towards targetable adenovirus vectors for gene therapy. Curr. Opin. Mol. Ther. 4, 444–451. El-Sheikh, A., Liu, C., Huang, H., Edgington, T.S., 2002. A novel vascular endothelial growth factor heparin-binding domain substructure binds to glycosaminoglycans in vivo and localizes to tumor microvascular endothelium. Cancer Res. 62, 7118–7123. Everts, M., Curiel, D.T., 2004. Transductional targeting of adenoviral cancer gene therapy. Curr. Gene Ther. 4, 337–346. Folkman, J., 1995. Angiogenesis in cancer, vascular, rheumatoid and other disease. Nat. Med. 1, 27–31. Gall, J., Kass-Eisler, A., Leinwand, L., Falck-Pedersen, E., 1996. Adenovirus type 5 and 7 capsid chimera: fiber replacement alters receptor tropism without affecting primary immune neutralization epitopes. J. Virol. 70, 2116–2123. Green, M., Pina, M., 1964. Biochemical studies on adenovirus multiplication. VI. Properties of highly purified tumorigenic human adenoviruses and their DNA’s. Proc. Natl. Acad. Sci. U.S.A. 51, 1251–1259. Havenga, M.J., Lemckert, A.A., Grimbergen, J.M., Vogels, R., Huisman, L.G., Valerio, D., Bout, A., Quax, P.H., 2001. Improved adenovirus vectors for infection of cardiovascular tissues. J. Virol. 75, 3335–3342. Jakubczak, J.L., Rollence, M.L., Stewart, D.A., Jafari, J.D., Von Seggern, D.J., Nemerow, G.R., Stevenson, S.C., Hallenbeck, P.L., 2001. Adenovirus Type 5 viral particles pseudotyped with mutagenized fiber proteins show diminished infectivity of coxsackie B-adenovirus receptor-bearing cells. J. Virol. 75, 2972–2981. Koivunen, E., Wang, B., Ruoslahti, E., 1995. Phage libraries displaying cyclic peptides with different ring sizes: ligand specificities of the RGD-directed integrins. Nat. Biotechnol. 13, 265–270. Koizumi, N., Mizuguchi, H., Hosono, T., Ishii-Watabe, A., Uchida, E., Utoguchi, N., Watanabe, Y., Hayakawa, T., 2001. Efficient gene transfer by fiber-mutant adenoviral vectors containing RGD peptide. Biochim. Biophys. Acta 1568, 13–20. Kovesdi, I., Brough, D.E., Bruder, J.T., Wickham, T.J., 1997. Adenoviral vectors for gene transfer. Curr. Opin. Biotechnol. 8, 583–589. Krasnykh, V., Dmitriev, I., Mikheeva, G., Miller, C.R., Belousova, N., Curiel, D.T., 1998. Characterization of an adenovirus vector containing a heterologous peptide epitope in the HI loop of the fiber knob. J. Virol. 72, 1844–1852. Kupprion, C., Motamed, K., Sage, E.H., 1998. SPARC (BM-40, Osteonectin) inhibits the mitogenic effect of vascular endothelial growth factor on microvascular endothelial cells. J. Biol. Chem. 273, 29635–29640. Liu, Q., Muruve, D.A., 2003. Molecular basis of the inflammatory response to adenovirus vectors. Gene Ther. 10, 935–940. McConnell, M.J., Imperiale, M.J., 2004. Biology of Adenovirus and its use as a vector for gene therapy. Hum. Gene Ther. 15, 1022–1033. Michael, S.I., Hong, J.S., Curiel, D.T., Engler, J.A., 1995. Addition of a short peptide ligand to the adenovirus fiber protein. Gene Ther. 2, 660–668. Mizuguchi, H., Hayakawa, T., 2004. Targeted adenovirus vectors. Hum. Gene Ther. 15, 1034–1044. Mizuguchi, H., Koizumi, N., Hosono, T., Utoguchi, N., Watanabe, Y., Kay, M.A., Hayakawa, T., 2001. A simplified system for constructing recombinant adenoviral vectors containing heterologous peptides in the HI loop of their fiber knob. Gene Ther. 8, 730–735. Nicklin, S.A., Von Seggern, D.J., Work, L.M., Pek, D.C.K., Dominiczak, A.F., Nemerow, G.R., Baker, A.H., 2001. Ablating adenovirus type 5 fiber-CAR binding and HI loop insertion of the SIGYPLP peptide generates an endothelial cell-selective adenovirus. Mol. Ther. 4, 534– 542. Nicklin, S.A., White, S.J., Nicol, C.G., Von Seggern, D.J., Baker, A.H., 2004. In vitro and in vivo characterization of endothelial cell selective adenoviral vectors. J. Gene Med. 6, 300–308. Nicklin, S.A., White, S.J., Watkins, S.J., Hawkins, R.E., Baker, A.H., 2000. Selective targeting of gene transfer to vascular endothelial cells by use of peptides isolated by phage display. Circulation 102, 231–237. Niethammer, A.G., Xiang, R., Becker, J.C., Wodrich, H., Pertl, U., Karsten, G., Eliceiri, B.P., Reisfeld, R.A., 2002. A DNA vaccine against VEGF receptor 2 prevents effective angiogenesis and inhibits tumor growth. Nat. Med. 8, 1369–1375.

C. Mahanivong et al. / Journal of Virological Methods 135 (2006) 127–135 Pasqualini, R., Koivunen, E., Ruoslahti, E., 1997. ␣v integrins as receptors for tumor targeting by circulating ligands. Nat. Biotechnol. 15, 542–546. Rein, D.T., Breidenbach, M., Wu, H., Han, T., Haviv, Y.S., Wang, M., Kirby, T.O., Kawakami, Y., Dall, P., Alvarez, R.D., Curiel, D.T., 2004. Gene transfer to cervical cancer with fiber-modified adenoviruses. Int. J. Cancer 111, 698–704. Roelvink, P.W., Lee, G.M., Einfeld, D.A., Kovesdi, I., Wickham, T.J., 1999. Identification of a conserved receptor-binding site on the fiber proteins of CAR-recognizing adenoviridae. Science 286, 1568–1571. Ruoslahti, E., 1996. RGD and other recognition sequences for integrins. Annu. Rev. Cell Dev. Biol. 12, 697–715. Russell, W.C., 2000. Update on adenovirus and its vectors. J. Gen. Virol. 81, 2573–2604. Shayakhmetov, D.M., Papayannopoulou, T., Stamatoyannopoulos, G., Lieber, A., 2000. Efficient gene transfer into human CD34+ cells by a retargeted adenovirus vector. J. Virol. 74, 2567–2584.

135

Stevenson, S.C., Rollence, M., Marshall-Neff, J., McClelland, A., 1997. Selective targeting of human cells by a chimeric adenovirus vector containing a modified fiber protein. J. Virol. 71, 4782–4790. Von Seggern, D.J., Kehler, J., Endo, R., Nemerow, G.R., 1998. Complementation of a fiber mutant adenovirus by packaging cell lines stably expressing the Ad5 fiber protein. J. Gen. Virol. 79, 1461–1468. Wickham, T.J., Roelvink, P.W., Brough, D.E., Kovesdi, I., 1996. Adenovirus targeted to heparin-containing receptors increases its gene delivery efficiency to multiple cell types. Nat. Biotechnol. 14, 1570–1573. Xia, D., Henry, L.J., Gerard, R.D., Deisenhofer, J., 1994. Crystal structure of ˚ the receptor-binding domain of adenovirus type 5 fiber protein at 1.7 A resolution. Structure 2, 1259–1270. Xiang, R., Mizutani, N., Luo, Y., Chiodoni, C., Zhou, H., Mizutani, M., Ba, Y., Becker, J.C., Reisfeld, R.A., 2005. A DNA vaccine targeting survivin combines apoptosis with suppression of angiogenesis in lung tumor eradication. Cancer Res. 65, 553–561.