Characterization of adenovirus p21 gene transfer, biodistribution, and immune response after local ocular delivery in New Zealand white rabbits

Experimental Eye Research 77 (2003) 355–365 www.elsevier.com/locate/yexer

Characterization of adenovirus p21 gene transfer, biodistribution, and immune response after local ocular delivery in New Zealand white rabbits S.F. Wen, Z. Chen, J. Nery, B. Faha* Canji Inc., 3525 John Hopkins Court, San Diego, CA 92121, USA Received 11 November 2002; accepted in revised form 7 April 2003

Abstract Previous studies suggest that local gene therapy with rAd-p21WAF1/Cip-1 [Perkins et al., 2002. Arch. Ophthalmol. 120, (2002) 941– 949] may provide an effective adjunctive anti-proliferative treatment to prevent glaucoma surgery failure. To further investigate rAd-p21 in this indication, we have characterized several parameters of local gene delivery to conjunctiva including, vector delivery and transgene expression in target tissue, inflammatory response, biodistribution to non-target tissues, and immune response. Quantitative PCR and RTPCR assays were employed to evaluate rAd-p21 dissemination and gene transfer following a single subconjunctival injection. In target tissue, significant levels of rAd-p21 DNA were found in 6/6 animals 1 and 4 days after injection. rAd-p21 DNA and RNA could be detected in the un-injected contralateral eye but at levels that were 10 000–100 000 lower than in the injected eye. Expression of human p21 transgene in conjunctival fibroblasts was confirmed by immunohistochemistry. Biodistribution of rAd-p21 following subconjunctival injection was substantially limited to ocular tissue. In 1/6 rabbits, rAd-p21 DNA was found in whole blood, liver, and spleen at levels that were barely detectable. All non-target organs were negative on day 4. In contrast, in a rabbit injected intravenously as a positive control, all blood samples and tissues samples were positive. rAd-p21 delivery to conjunctiva followed by filtration surgery caused an early acute inflammatory response, which by day 14 was indistinguishable from placebo-treated eyes. Neutralizing anti-adenovirus antibodies were detected following administration of rAd-p21 to conjunctiva, however, vector delivery and transgene expression were unaffected in a subsequent administration to the contralateral eye in the same animal. These results show that local delivery to conjunctiva may be a suitable delivery mode for ocular gene therapy. q 2003 Elsevier Ltd. All rights reserved. Keywords: adenovirus; gene transfer; biodistribution; immune response; conjunctiva; rabbit

1. Introduction Millions of people in the US suffer from chronic blinding diseases such as diabetic retinopathy, macular degeneration, retinal degenerative diseases and glaucoma for which there are no cures. Such diseases may be uniquely amenable to gene therapy for reasons which include: (1) the eye is an easily accessible target suitable for local administration; (2) local gene delivery has high transduction efficiency; (3) local gene delivery to the eye can be incorporated into routine surgical procedures in humans; and (4) local gene * Corresponding author. Dr B. Faha, Canji Inc., 3525 John Hopkins Court, San Diego, CA, USA. E-mail address: [email protected] (B. Faha). 0014-4835/03/$ - see front matter q 2003 Elsevier Ltd. All rights reserved. DOI:10.1016/S0014-4835(03)00122-2

delivery to the eye allows exposure of the target tissue with reduced risk of systemic effects. A number of viral and non-viral vectors have been tested in preclinical models for ocular gene delivery by various routes (Murata et al., 1997; Kaufman et al., 1999; Bennett and Maguire, 2000; Hauswirth and Lewin, 2000). Encouraging results from proof-of-principle experiments in preclinical disease models of proliferative vitreoretinopathy, retinal ganglion cell rescue, retinitis pigmentosa, and Leber congenital amaurosis have recently been reported (Sakamoto et al., 1995; Bennett et al., 1998; Di Polo et al., 1998; Lewin et al., 1998; Acland et al., 2001). Of the gene delivery systems tested to date, recombinant adenoviral vectors (rAd) have been among the most extensively studied and have demonstrated efficient transduction in ocular tissues

356

S.F. Wen et al. / Experimental Eye Research 77 (2003) 355–365

(Bennett et al., 1994; Li et al., 1994; Hoffman et al., 1997; Borras et al., 1998). The conjunctival wound healing response has been a major determinant of surgical failure following glaucoma filtration surgery (for reviews see Skuta and Parrish, 1987; Daniels et al., 1998). The most commonly used agents to inhibit unwanted proliferation and improve surgical success are 5-fluorouracil (5-FU) and mitomycin C (MMC). However, both drugs have problems associated with their use, which have been documented extensively, and include using multiple subconjunctival injections that may cause discomfort to the patient, corneal epithelial defects, chronic hypotony, thin-walled failing filtration blebs and endophthalmitis (Shields et al., 1993; Zacharia et al., 1993; Suner et al., 1997; Akova et al., 1999). Therefore, alternative adjunctive treatments to inhibit wound healing have been investigated and have included, but are not limited to, the use of an anti-TGF-b antibody, cellular photoablation, RGD peptides, etoposide, and paclitaxel (Jampel et al., 1993; Avila et al., 1998; Jampel and Moon, 1998; Cordeiro et al., 1999; Grisanti et al., 1999). We have previously demonstrated efficient gene transfer into rabbit eyes via local delivery of a recombinant adenovirus encoding the p21WAF1/Cip-1 cell cycle inhibitor (for review see Gartel et al., 1996), rAd-p21. rAd-p21 treatment resulted in a gene-specific inhibition of fibroproliferation and wound healing in a rabbit model of glaucoma filtration surgery (Perkins et al., 2002). Selective cell-cycle inhibition induced by rAd-p21 can potentially improve surgical success of glaucoma filtration surgery without the complications associated with current adjunctive agents. We have expanded these studies by characterizing several parameters of local gene delivery to conjunctiva relating to safety for this application, including inflammatory response, biodistribution to non-target tissue, and host immune response.

2. Materials and methods 2.1. Recombinant adenovirus construction and production E1/partial E3-deleted replication-deficient recombinant adenoviruses encoding human p21, green fluorescent protein, and b-galactosidase were used in these experiments (rAd-p21, rAd-GFP, and rAd-b-gal, respectively). Details of the construction of the recombinant adenoviruses are described elsewhere (Wills et al., 1994). Essentially, the coding region of each protein, driven by the human cytomegalovirus (CMV) immediate early promoter, was cloned into the E1 region of adenovirus. Recombinant adenoviruses were grown and propagated in the human embryonic kidney cell line 293 (American Type Culture Collection, Rockville, MD, USA) and purified using standard protocols (Huyghe et al., 1995). Viral particle number of purified rAd was determined using HPLC

quantification and reported as particles/ml (Shabram et al., 1997). 2.2. Gene delivery to conjunctiva and glaucoma surgery All procedures were approved by the Institutional Animal Care and Use Committee and were in accordance with the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research. New Zealand White rabbits weighing between 2 and 3 kg were used for this study. Animals were anesthetized with an intramuscular injection of ketamine (35 –45 mg kg21) and xylazine (5 – 10 mg kg21). Exposure of the eye orbit was achieved with a wire lid speculum. rAdp21 was delivered to the conjunctiva by either a subconjunctival injection at the superolateral quadrant or by topical application via a virus-soaked Wek Cel sponge during filtration surgery. Procedures were either unilateral or bilateral depending on the experiment. For animals undergoing both subconjunctival injection and filtration surgery, the injection (100 ml volume) was performed the day before surgery. Surgical eyes underwent an unguarded sclerectomy using a limbal-based approach, the details of which are described elsewhere (Perkins et al., 2002). Adenovirus particle concentrations used for these studies ranged from 7– 9·5 £ 1011 particles ml21, and, the infectious units/ml ranged from 2·6 £ 109 to 5·5 £ 1010 infectious units ml21. Controls included animals dosed with adenovirus excipient, which is referred to as ‘placebo’ in the text. For the biodistribution study, one animal received 1 ml of rAd-p21 containing 9·5 £ 1011 particles intravenously via the marginal ear vein. For histologic analyses animals were sacrificed 1, 3, 7, or 14 days after surgery. Eyes (n ¼ 3 eyes/timepoint) were enucleated, rinsed in PBS, fixed in 4% paraformaldehyde and processed to slides for either immunohistochemistry or hematoxylin and eosin staining. Hematoxylin and eosin stained slides were evaluated for inflammatory response. Inflammation was graded in a masked manner by a veterinary pathologist (Diplomate, ACVP) using a standard histopathological grading method (minimal , mild , moderate , severe), as previously described (Perkins et al., 2002). The number of inflammatory cells (based on morphology) per 40 £ field were counted from representative sections and averaged: minimal: , 10 inflammatory cells/40 £ field; mild: 11 –40 inflammatory cells/40 £ field; moderate: 41 – 65 inflammatory cells/40 £ field; severe: . 65 inflammatory cells/40 £ field. For rAd-bgal (see above) delivery animals received a subconjunctival injection of , 20 ml virus (7·4 £ 1011 particles ml21) and were sacrificed 3 days later. Eyes were enucleated, rinsed in phosphate-buffered saline (PBS), and fixed for 30 min in 1% paraformaldehyde, 0·2% glutaraldehyde, 0·02% IGEPAL, 0·01% sodium deoxycholate in PBS. Eyes were incubated for 2 hr at room temperature in b-galactosidase substrate solution containing

S.F. Wen et al. / Experimental Eye Research 77 (2003) 355–365

1 mg ml21 X-gal, 1·2 mM MgCl2, 3 mM K4FeCN, 3 mM K3FeCN in PBS at pH 7·4. Eyes were rinsed in PBS and photographs were taken.

357

assumption that one viral particle contains one genome of viral DNA. rAd-p21 RNA was quantified using rAd-p21 cRNA as reported previously (Wen et al., 2000), and was expressed as copies/mg of tissue.

2.3. Sample collection for biodistribution Whole blood was collected from all animals before injection, and at 5 and 30 min, and at 2 hr and 1 day post injection, placed in citrate buffer, and kept frozen until use. Multiple tissues including the injection site and un-injected contralateral eye (full thickness; approximately 1 cm £ 1 cm size), aqueous, liver, ovaries, kidneys, spleen, heart, whole blood, and lungs were sampled 1 and 4 days after unilateral subconjunctival injection. Samples were snap frozen and stored at 2 808C for QPCR and QRT-PCR analyses. 2.4. QPCR/QRT-PCR and preparation of standard curves Quantitative PCR and RT-PCR (QPCR, QRT-PCR) were used to quantify rAd-p21 DNA and transgene expression as previously described (Wen et al., 2000). Briefly, DNA and RNA were co-extracted from approximately 50– 100 mg of tissue using Tri-reagent (Molecular Research Center, Inc., Cincinnati, OH, USA). Purified RNA was DNased and complete removal of DNA was confirmed by PCR. For whole blood, DNA was purified from 0·5 ml of citrated whole blood using the QIAmp DNA mini kit (QIAGEN, Valencia, CA, USA). QPCR/QRT-PCR was performed using the ABI 7700 sequence detector (Applied Biosystems, Inc., Foster City, CA, USA). Rabbit GAPDH DNA and RNA were used as internal controls to assess the quality of purified RNA and DNA and to ensure equal amounts of sample input. Primer and probe sequence used for QPCR and QRT-PCR were as follows: rAd-p21; forward primer 50 -AACGGTACTCCGCCACC-30 , reverse primer 50 -TTCTGACATGGCGCCTACT-30 and probe FAM-TCCGCATCGACCGGATCGG-TAMRA. GAPDH; forward primer, 50 -ACGTGCCGCCTGGAGAA-30 , reverse primer, 50 - CATGAGGTCCACCACCCTGTT-30 and probe FAM-ATGACATCAAGAAGGTGGTGAAGCAGGCTAMRA. To generate standard curves, liver, ovary, spleen, surgical site (full thickness, which includes conjunctiva and retina) heart, kidney, lung and whole blood were collected from naı¨ve animals. Tissues were minced thoroughly and aliquoted into several 100 mg samples. Six to eight 10-fold serially diluted samples of rAd-p21, ranging from 2·5 £ 102 to 2·5 £ 109 particles, were spiked into each aliquot. rAd-p21-spiked tissue aliquots were extracted and processed for QPCR in a manner identical to the test samples, generating standard curves for each type of tissue. Standard curves were used to estimate the number of adenoviral particles in test samples, which were expressed as genome equivalent (GE) per mg of tissue, or per ml of whole blood. A genome equivalent (GE) is based on the

2.5. Immunohistochemistry Immediately following enucleation, eyes were rinsed in PBS, fixed in 4% paraformaldehyde and processed to slides. Slides were immersed in 2 208C ethanol/acetic acid (2:1) for 10 min, followed by antigen retrieval by steaming in high pH buffer (Dako S3307) for 20 min. Endogenous peroxidase was blocked by incubation with 3% (V/V) H2O2. Slides were blocked in 20% (V/V) goat serum for 30 min and then stained for 1 hr with a monoclonal antibody specific for human p21 (1:100, Pharmingen #15091A). After rinsing in PBS (3 £ 5 min), slides were reacted with a biotinylated goat anti-mouse secondary reagent (1:200, Zymed 62-6640) for 30 min. Slides were rinsed in PBS (3 £ 5 min), followed by a 10-minute incubation with streptavidin/HRP conjugate (Dako K0675). Slides were developed with an AEC chromogen/substrate (Dako K3464) and counterstained with hematoxylin. Immunohistochemistry was performed on the Dako autostainer. Serial sections were stained with hematoxylin and eosin using standard protocols. 2.6. Neutralizing antibody assay Blood was drawn at the times indicated in Fig. 7 and sera were prepared using standard procedures. Serum neutralizing capacity of rAd transduction was evaluated using a modified protocol (Rahman et al., 2001) based on a previously described assay (Kozarsky et al., 1994). Duplicate serial dilutions (1:2) of the serum sample of interest (starting at 1:20, 1:40, 1:80, etc.) were plated in a 96 well format (60 ml/well). Appropriately diluted serum samples were incubated with equal volumes of rAd-GFP at a final concentration of 4 £ 108 particles ml21 for 1 hr at 378C. HeLa cells were plated 1 day prior at 1 £ 104 cells/well in flat bottom 96 well format. Subsequently, the media from the HeLa cell cultures were removed and 100 ml of the diluted serum/rAd-GFP mixtures were transferred onto the HeLa cells and were co-incubated overnight. Relative fluorescence was measured the following day on a CytoFluor 4000 Fluorescence Multiwell Plate Reader (PerSeptive Biosystems, Foster City CA, USA) to assay neutralizing capacity. Background fluorescence was assessed in wells containing HeLa cells cultured alone, and GFCB fluorescence was determined in wells with HeLa cells cultured with virus in the absence of immune serum. Sera from placebo-treated animals (two) served as normal rabbit serum controls. The raw fluorescence numbers were then plotted as% transduction using the following equation:

358

%Transduction ¼

S.F. Wen et al. / Experimental Eye Research 77 (2003) 355–365

ðFluorescence value 2 background fluorescence valueÞ £ 100 ðFluorescence value of rAd-GFP control 2 background fluorescence valueÞ

ID-50 titers were subsequently calculated as the titer at which 50% transduction was observed.

evaluate biodistribution incorporates extraction efficiency and matrix effects directly into the QPCR assay.

3. Results

3.2. rAd-p21 delivery and transgene expression after subconjunctival injection: analysis of target tissue

3.1. Quantitative PCR and RT-PCR assay performance To control the nucleic acid extraction and QPCR efficiency across multiple tissue types, standard curves were generated by spiking known amounts of rAd-p21, ranging from 2·5 £ 102 to 2·5 £ 109 particles, into 100 mg of the minced matched tissue collected from naı¨ve animals. The quantification limit for DNA was 2·5 £ 101 GE mg21 of tissue and 1·0 £ 104 GE ml21 of whole blood. As shown in Fig. 1, for all tissues the number of viral particles versus the cycle threshold (the QPCR cycle in which fluorescence separates from baseline) was linear on a logarithmic scale. The slope of each standard curve was used to estimate the QPCR amplification efficiency for DNA extracted for each tissue types. Amplification efficiencies ranged from 102 to 125% in all tissues except whole blood, which was estimated to be 144% (data not shown). This observation indicates that rAd-p21 DNA co-extracted with nucleic acid from the different tissue was amplified with varying efficiency. The use of tissue type matched standards to

Fig. 1. Standard curves of adenovirus spiked tissues for QPCR standards. Known amounts of rAd-p21 (particles as indicated in the figure) were spiked into multiple tissues. Samples were extracted and processed for QPCR as described in Section 2. (B): ovary; (X): kidney; (O): eye; (A): spleen; (W): heart; (K): liver. Threshold cycle: the PCR cycle where fluorescence separates from the baseline. It has an inverse relationship with input DNA quantity.

Rabbits received unilateral subconjunctival injections of 100 ml rAd-p21 (9·5 £ 1010 particles) and were sacrificed 1 and 4 days later (n ¼ 3/timepoint). Vector biodistribution after subconjunctival injection of a small amount of virus was expected to be minimal, therefore, positive control tissues were generated in one rabbit that was injected intravenously with 1 ml rAd-p21 (9·5 £ 1011 particles) and sacrificed 1 day later. Eyes were enucleated and the injection site and analogous tissue from the contralateral un-injected eye were collected and analyzed for rAd-p21 DNA and transgene expression using QPCR and QRT-PCR. Consistent with previous results obtained after intraoperative topical administration using a Wek Cel sponge soaked with rAd-p21 (Perkins et al., 2002), an average of 1·6 £ 107 ^ 3·0 £ 106 GE mg21 tissue of rAd-p21 DNA and 1·8 £ 107 ^ 1·0 £ 107 copies mg21 tissue of p21 RNA were detected at the injection site of each animal (Fig. 2(A) and (B); white bars). These levels decreased slightly by day 4 to 8·4 £ 106 ^ 7·9 £ 106 GE mg21 tissue and 1·1 £ 106 ^ 3·5 £ 105 copies mg21 tissue. Eyes from the intravenously dosed rabbit were also positive for rAd-p21 DNA and p21 RNA (Fig. 2; black bars), but at levels that were , 10 000 fold lower than in subconjunctivally injected eyes. Surprisingly, rAd-p21 DNA was detected in the contralateral un-injected eye in 3/3 rabbits on day 1, but at levels that were , 100 000 lower than in the injected eye (3·0 £ 102 ^ 3·4 £ 102 vs 1·6 £ 107 ^ 3·0 £ 106 GE mg21; Fig. 2(A)). In contrast, p21 transgene expression was detected in only 1/3 contralateral eyes (9·3 £ 101 copies mg21 tissue; Fig. 2(B)). This is consistent with results from other experiments in which we did not detect human p21 or b-galactosidase protein in the contralateral eye (data not shown). By day 4, rAd-p21 DNA was detected in the contralateral eye of only 1/3 rabbits (2·9 £ 101 GE mg21 tissue) and transgene expression was not detected in any rabbit. rAd-p21 DNA or mRNA were not detected in aqueous samples, collected from similarly dosed animals 3 and 14 days after injection (data not shown). To determine whether the cell types relevant to wound healing after filtration surgery were transduced following subconjunctival injection, immunohistochemistry was performed using methods that specifically detect human p21 protein. Rabbits received a subconjunctival injection of 7·6 £ 1010 rAd-p21 particles in a volume of 100 ml,

S.F. Wen et al. / Experimental Eye Research 77 (2003) 355–365

359

Fig. 2. Detection of rAd-p21 DNA and RNA in eye tissue. Eyes (six eyes, six animals) received a subconjunctival injection (n ¼ 3/timepoint) of 100 ml rAdp21 (9·5 £ 1010 total particles). One animal received an intravenous injection of 1 ml of rAd-p21 (9·5 £ 1011 total particles). Animals were sacrificed and injected and contralateral uninjected eyes were processed for QPCR (A) and QRT-PCR (B) analyses to detect rAd-p21 DNA and RNA, respectively. Quantification level QPCR: 2·5 £ 103 GE/100 mg tissue; quantification level QRT-PCR: 10 copies/assay. Black bar: intravenous injection; mean of two eyes from the same animal; white bar: subconjunctival injection; mean ^ SD.

followed by filtration surgery the next day. Animals were sacrificed 1 day later and eyes were enucleated and analyzed for human p21 as described in Section 2. As shown in Fig. 3(A), human p21 was detected in conjunctival fibroblasts, consistent with previous results obtained using intraoperative topical administration of rAd-p21 (Perkins et al., 2002). Expression was confined to the side of the globe that had received the subconjunctival injection. In addition, there was no evidence of p21 expression in scleral fibroblasts, ciliary body epithelial cells, retina, iris, or cornea. Inflammatory cells as determined by morphology were negative (see below). Sections in which the primary antibody was omitted were negative (Fig. 3(B)). Placebo injected eyes were also negative, demonstrating antibody specificity for human p21 (Fig. 3(C)). While these studies have demonstrated the efficiency of rAd-p21 vector delivery and RNA expression and have identified the transduced cell types, it was important to determine the extent of vector delivery in the eye. To do

this, three rabbits received a subconjunctival injection of a recombinant adenovirus encoding bacterial b-galactosidase (, 20 ml; 7·4 £ 1011 particles ml21) and were sacrificed 3 days later. Eyes were enucleated and whole organ X-gal stained as described in Section 2. As seen in Fig. 4 the extent of b-galactosidase expression extends roughly 1/4 –1/3 around the globe, which is well within the size of a filtration surgical site. We also noticed that the bleb caused by the subconjunctival injection is significantly smaller than the b-galactosidase stained area, and, assume that the vector is spread by mechanical means (e.g. blinking) as the vector is absorbed into the tissue. 3.3. Biodistribution of rAd-p21 after subconjunctival injection As shown in Table 1, subconjunctival injection of a single dose of 9·5 £ 1010 particles of rAd-p21 resulted in minimal systemic exposure. In one of three rabbits

Fig. 3. Immunostaining for human p21 after subconjunctival injection of rAd-p21 in rabbits. Animals received a subconjunctival injection of rAd-p21 (7 £ 1010 total particles) or placebo and filtration surgery was performed the next day. One day after surgery animals were sacrificed. Eyes were enucleated and processed for immunohistochemistry using an antibody specific for human p21. Panel (A) shows positive staining for human p21 in conjunctival fibroblasts (examples are indicated by arrows in the figure) and panel (B) shows the antibody negative control. Placebo-treated eyes stained for p21 were negative and are shown in panel (C). Slides were counterstained with hematoxylin; 200 £ .

360

S.F. Wen et al. / Experimental Eye Research 77 (2003) 355–365

Fig. 4. Whole organ X-gal stain after subconjunctival injection of rAd-bgal in rabbits. Eyes received a subconjunctival injection and were enucleated and stained for b-galactosidase expression as described as described in Section 2. Eyes from three animals are shown.

injection (one out of a total of 24 timepoints; 1·1 £ 105 GE ml21 of whole blood), and was the same rabbit in which liver and spleen were positive for rAd-p21 DNA (Fig. 5(C)). In contrast, all positive control tissues from the animal that received an intravenous injection of rAd-p21 (9·5 £ 1011 particles), including whole blood, were positive for rAd-p21 DNA and/or RNA 24 hr post injection (Fig. 5(A)– (C)). All tissues from placebo-injected rabbits were below quantification levels (data not shown). 3.4. Timecourse of inflammatory response

Table 1 rAd-p21 DNA and RNA detection in various tissues after a single subconjunctival injection Tissue

rAd-p21 DNA rAd-p21 RNA (GE/mg of tissue or ml of whole blood) (copies/mg of tissue)

Liver Ovaries Blood Spleen Kidneys Heart Lungs

1/6 (day 1; 2·5 £ 101) 0/6 1/6 (30 min., 1·1 £ 105) 1/6 (day 1, 6·4 £ 101) 0/6 0/6 0/6

0/6 0/6 ND 1/6 (day 1; 1·3 £ 102) 0/6 0/6 0/6

Samples were collected 1 and 4 days after subconjunctival injection, three animal per time point, with a single dose of rAd-p21 containing 9·5 £ 1010 particles; GE: genome equivalents; ND: not done.

quantifiable levels of rAd-p21 DNA were detected in liver (2·5 £ 101 GE/mg of tissue) and spleen (6·4 £ 101 GE/mg of tissue) 24 hr post injection, levels that were at or near the quantification limit. One spleen sample was also positive for p21 RNA (1·3 £ 102 copies/mg of tissue). rAd-p21 DNA and RNA were not detected in any tissue 4 days after subconjunctival injection. rAd-p21 DNA was detected in whole blood in only one rabbit (out of six) 30 min after

We have shown previously that eyes treated with placebo, control adenovirus, or rAd-p21 displayed minimal cellular infiltrate 30 days post intraoperative topical administration (Perkins et al., 2002). In this study we sought to extend these findings by characterizing the inflammatory response within the first 2 weeks after subconjunctival injection of rAd-p21 and filtration surgery. Sections of eyes were evaluated 1, 3, 7, and 14 days after filtration surgery (2, 4, 8, and 15 days after subconjunctival injection). Filtration surgery was also performed on placebo-injected eyes to control for any inflammation related to the surgical procedure. Results are shown in Table 2. Following subconjunctival injection and filtration surgery rAd-p21-treated eyes had an immediate severe inflammatory response, compared to placebo-treated eyes, which was characterized as mild (Fig. 6(A) and (B)). However, 14 days post surgery the response had lessened in rAd-p21-treated eyes, which were indistinguishable from placebo-treated eyes (Fig. 6(C) and (D)). In both groups inflammatory responses were confined to the conjunctiva at the site of injection and surgery and did not extend around the globe. Clinical exams evaluating anterior chamber reactions using slit lamp biomicroscopy showed no

Fig. 5. Distribution of rAd-p21 DNA and RNA in multiple tissues after intravenous injection. Samples were collected 1 day after rAd-p21 administration (intravenous injection; 9·5 £ 1011 total particles; n ¼ 1 rabbit) and processed for QPCR (A) and QRT-PCR (B) analyses. S: spleen; Li: liver; Lu: lung; H: heart; O: ovaries; K: kidneys E: eye (mean of two eyes). rAd-p21 DNA in kidney was below quantification level. (C) rAd-DNA in whole blood collected at 5 min, 30 min, 2 hr., and 24 hr after either subconjunctival (9·5 £ 1010 total particles) or intravenous (9·5 £ 1011 total particles) injection. (Quantification level QPCR: 2·5 £ 103 GE/100 mg tissue or ml whole blood; quantification level QRT-PCR: 10 copies/assay).

S.F. Wen et al. / Experimental Eye Research 77 (2003) 355–365 Table 2 Inflammatory response following subconjunctival injection and filtration surgery Days post surgery

Day 1

Day 3

Day 7

Day 14

Placebo rAd-p21

Mild: 3/3 Severe: 3/3

Mild: 3/3 Severe: 3/3

Mild: 3/3 Moderate: 3/3

Mild: 3/3 Mild: 3/3

Relative scale: minimal , mild , moderate , severe. Numbers represent the number of eyes observed/total number of eye evaluated.

differences between placebo and rAd-p21-treated eyes at any time-point (not shown). 3.5. Generation of neutralizing antibodies and the effect on the ability to dose the contralateral eye Since it has been well established that adenovirus vectors elicit a humoral immune response (Yang et al., 1994, 1995; Michou et al., 1997; Schulick et al., 1997; Otake et al., 1998), we wanted to investigate the humoral response after local administration to conjunctiva and determine the effects on subsequent dosing to the contralateral eye. This is

361

important in the context of glaucoma filtration surgery since glaucoma is a bilateral disease, and patients often require surgery in both eyes. Preliminary studies in rabbits showed that serum neutralizing anti-adenovirus antibodies develop after intraoperative topical administration, peaking at 7 days and decreasing thereafter (data not shown). Therefore, an experiment was performed characterizing how serum neutralizing anti-adenovirus antibodies generated by dosing one eye, affected subsequent rAd-p21 delivery and transgene expression in the contralateral eye. A timeline of rAd-p21 dosing/surgery, blood draws, and termination schedules is shown in Fig. 7(A). All animals were dosed with 100 ml rAd-p21 (7·2 £ 1010 particles) intraoperatively using a virus-soaked ophthalmic sponge as described previously (Perkins et al., 2002), except two, which were treated with placebo soaked sponges. Bilateral procedures were performed in cohorts 2 and 3 per the schedule outlined in Fig. 7. Sacrifices were scheduled such that rAd-p21 DNA and RNA were analyzed 3 days post dosing for all cohorts. At the times indicated, rabbits were sacrificed and eyes processed for QPCR and QRT-PCR to

Fig. 6. Inflammatory response after subconjunctival injection of rAd-p21 in rabbits. Animals received either a subconjunctival injection of placebo (A, C) or rAd-p21 (7 £ 1010 total particles; (B, D) and filtration surgery was performed the next day. Animals were sacrificed and eyes enucleated at the times post surgery indicated in Table 2. Eyes were fixed and processed for histologic analyses. Sections were stained with hematoxylin and eosin and graded according to a relative scale; minimal , mild , moderate , severe. Panels (A) and (B) are representative sections from day 1 and panels (C) and (D) are representative sections from day 14. Sections shown are areas of conjunctiva adjacent to the sclerectomy site; 200 £ .

362

S.F. Wen et al. / Experimental Eye Research 77 (2003) 355–365

Fig. 7. Effect of rAd-p21 pre-exposure to the right eye on subsequent rAd-p21 delivery and RNA expression in the left eye. All animals received one dose of rAd-p21 (7·2 £ 1010 total particles), topically, on day 1, animals in Cohort 2 and 3 received the second dose of rAd-p21 (7·2 £ 1010 total particles), topically, on day 7 or day 31, in the contralateral eye, respectively. Animals in Cohort 1 only received one dose. All eyes were enucleated 3 days after the last dose of rAdp21. (A) Experimental timeline for each cohort: surgical schedule, blood draws (draw) and termination (term). OD, right eye; OS, left eye; ND, not done Levels of rAd-p21 (B) DNA and (C) RNA in injection site; n ¼ 3/cohort. (BQL DNA: 2·5 £ 103 GE/100 mg of tissue; BQL RNA: 10 copies/assay).

determine rAd-p21 delivery and transgene expression. Blood was drawn at several time-points and sera were analyzed to establish the levels of serum neutralizing antiadenovirus antibodies present at the time the second dose was administered. As shown in Fig. 7(B) and (C), there were no differences detected in either rAd-p21 DNA or RNA among any of the cohorts, irrespective of the presence of serum neutralizing anti-adenovirus antibodies (Table 3). Rabbits that received the second dose of rAd-p21 at peak levels of serum neutralizing anti-adenovirus antibodies (cohort 2) had nearly identical levels of rAd-p21 DNA and RNA 3 days post dosing as did naı¨ve rabbits (cohort 1). Similar results were obtained in rabbits given the second dose 31 days after the first dose (cohort 3). QPCR and QRT-PCR analyses were negative for placebo-treated eyes.

4. Discussion Previously, we have shown that rAd-p21 delivered as an adjunct to filtration surgery inhibited wound healing in a rabbit model and was as effective as mitomycin C (Perkins et al., 2002). These data suggested that rAd-p21 might

provide an effective anti-proliferative for the inhibition of wound healing following glaucoma surgery. Therefore to evaluate further the feasibility of using rAd-p21 as a potential therapeutic candidate for preventing glaucoma surgery failure, experiments were initiated which investigated the biodistribution in target and non-target tissues, the development of neutralizing antibodies and inflammatory reactions after local ocular delivery in New Zealand white rabbits. Our data show high levels of rAd-p21 DNA and RNA in conjunctiva (Fig. 2), and, human p21 expression in conjunctival fibroblasts (Fig. 3; the target cells in glaucoma filtration surgery) after subconjunctival injection. In addition, using a reporter gene construct, rAd-bgal, we show that transgene expression covers that portion of the eye that would be expected to be involved in the filtration surgical procedure (Fig. 4). These results are consistent with those previously reported using an intraoperative topical administration with a virus soaked cellulose sponge in rabbits (Perkins et al., 2002). In both studies similar levels of expression were observed 1 day after dosing irrespective of the route of administration (i.e. 1·8 £ 107 ^ 1·0 £ 107 copies mg21 of tissue for subconjunctival injection versus 3·9 £ 107 ^ 6·9 £ 106 copies mg21 for topical). This level of expression following intraoperative topical admin-

S.F. Wen et al. / Experimental Eye Research 77 (2003) 355–365

363

Table 3 Reciprocal titers of individual rabbits

p Sacrificed, animals in cohort 1 were sacrificed on day 3; ND: not done; Bold values: reciprocal titer at time of second dose to contralateral eye; Shaded areas: sample not taken, animal had been sacrificed previously.

istration was shown to inhibit wound healing in a rabbit model of filtration surgery. In addition, human p21 localization to conjunctival fibroblasts after subconjunctival injection is also consistent with previous results obtained after intraoperative topical administration (Perkins et al., 2002). These data suggest that both subconjunctival injection and topical treatment can deliver rAd-p21 to conjunctiva to prevent glaucoma surgery failure. While QPCR and QRT-PCR provide useful tools for quantifying gene delivery efficiency and its tissue dissemination, it is of importance to know whether the cell type involved in the wound healing response is transduced. Proliferation and activation of conjunctival fibroblasts have been shown to play important roles in scarring after filtration surgery (Skuta and Parrish, 1987; Daniels et al., 1998). Controlling fibroblast proliferation by delivery of a cell cycle inhibitor, via rAd-p21, to conjunctival fibroblasts would therefore target the proliferative component of the wound healing response. Others have shown efficient expression of reporter genes (CAT and b-gal) in glaucoma surgery tissues (conjunctival fibroblasts) using either plasmid or adenovirus constructs injected post operatively into the filtering bleb (Angella et al., 2000; Skaf et al., 2001). No p21 expression was observed in scleral fibroblasts, ciliary body epithelial cells, retina, iris, or cornea. Localized gene transduction by subconjunctival injection resulted in high efficiency of gene transfer to the target cells, thereby minimizing the risk of side effects on irrelevant sites within the eye. In contrast both 5-FU and MMC, agents currently in use clinically for this indication, have been shown to distribute to multiple sites within the eye including, cornea, aqueous humor, and vitreous (Fantes et al., 1985; Mietz et al., 1995). We also show that exposure to non-target tissue following a single subconjunctival injection of rAd-p21 was minimal (Table 1 and Figs. 2 and 4). This is not

unforeseen given the relatively small dose of 100 ml (9·5 £ 1010 total particles). In one of three rabbits, rAdp21 DNA was detected in blood (at 30 min), and, in liver and spleen 1 day after injection at levels that were at or near the level of quantification. Exposure to the contralateral eye was unexpectedly observed in all rabbits 1 day after injection, but at levels approximately 10 000 fold lower than in the injected eye, which decreased to below the level of quantification by day 4 in the majority of rabbits. Whether physiological or technical mechanisms explain why rAd-p21 was detected in the contralateral eye is not known and requires further investigation. While the QPCR reaction is extremely sensitive, the ability to detect minute quantities of adenovirus in distant and/or non-target organs begs the question as to the biologic relevance with respect to safety. We do not detect human p21 expression by immunohistochemistry in contralateral eyes from similarly dosed rabbits and histopathological analyses have shown them to be normal (unpublished observations). Similarly, Sewell and colleagues reported no cytopathic or functional effects in liver and kidney, despite the PCR detection of vector, after local dosing of rAd-tk to oral tumors in mice (Sewell et al., 1997). In addition to very low levels of vector, other possible explanations for the lack of functional or histopathologic effects might be that the vector resides in the blood component or interstitial spaces within the tissue. However, the presence of detectable RNA in the spleen sample (Table 1) would tend to negate the latter possibility. Nonetheless, the systemic exposure after local ocular administration is significantly minimal such that any untoward effects would not be expected. It has been well established that cellular and humoral immune responses can limit the effectiveness of adenoviral gene therapy (Yang et al., 1994; Yang et al., 1995; Michou et al., 1997; Schulick et al., 1997; Otake et al., 1998). Local administration to conjunctiva is considered extra-ocular, as

364

S.F. Wen et al. / Experimental Eye Research 77 (2003) 355–365

opposed to intravitreal and subretinal administration. Therefore, ocular immune privilege would be expected not to apply in this indication and all the cellular and humoral responses to adenoviral vectors would be expected to occur. Indeed, histologic analyses showed that local administration to conjunctiva resulted in an immediate severe inflammatory response (Fig. 6 and Table 2), which subsequently lessened to resemble placebo treated eyes within 2 weeks. Although histologic evidence showed an inflammatory response, there were no clinical manifestations of ocular inflammation such as anterior chamber cells and/or flare, tearing, or any evident discomfort of the animals. Cellular immune responses as observed here can result in the loss of transduced cells (Yang et al., 1994; Otake et al., 1998). We have shown previously that p21 transgene expression decreases following local administration to conjunctiva in rabbits (Perkins et al., 2002). However, rAd-p21 RNA expression could be detected out to 60 days. This is well within the timeframe of the most aggressive wound healing response following filtration surgery and the time during which p21 expression would be needed to exert its anti-proliferative effect to inhibit wound healing. Anti-adenovirus neutralizing antibodies were detected to varying degrees in all animals after a single topical administration of rAd-p21 (Table 3). However, their presence did not affect the ability to subsequently dose the contralateral eye either 7 or 31 days after the first dose, as the levels of rAd-p21 DNA and RNA were similar to that of naı¨ve animals that had received only one dose (cohort 1; Fig. 7(B) and (C)). This is especially important for the therapeutic value of rAd-p21 in modulating wound healing after glaucoma surgery since glaucoma is a bilateral disease with patients often requiring a subsequent surgery in the contralateral eye. This finding is not unprecedented, as others have shown little correlation between serum antiadenovirus neutralizing antibodies and gene transfer efficiency in initial and subsequent dosing (Sterman et al., 1998; Nemunaitis et al., 2000; Schuler et al., 2001). We have not addressed whether it is possible to re-dose the same eye. Whether the ability to do so would be needed for this indication given the timing of the proliferative component of the wound healing response has not been investigated. In this regard, multiple subconjunctival injections of 5-FU are often given to patients undergoing filtration surgery both before and after surgery (Goldenfeld et al., 1994; Wong et al., 1994), yet a single dose of MMC is typically used in surgical practice. In conclusion, we have shown that local administration of rAd-p21 (1) efficiently delivered rAd-p21 to the target tissue (conjunctiva) for filtration surgery; (2) resulted in minimal systemic exposure; (3) resulted in a transient inflammatory response and (4) that the generation of serum neutralizing anti-adenovirus antibodies did not affect vector delivery or transgene expression in a subsequent administration to the contralateral eye. Our findings support the use

of rAd-p21 to modulate wound healing following glaucoma surgery.

Acknowledgements We thank Drs Dan Maneval and Monica Zepeda, Canji, Inc., for critical reading of the manuscript; Scott Anderson, Erlinda Quijano, and Mary Hess for technical assistance; Ming Ni for surgical expertise; Van Tsai for help with the neutralizing antibody assay; Melissa Aguirre and Jeff Coleman for histology support; and Drs Suganto Sutjipto, Barry Sugarman and Ken Wills for viral production, characterization and construction.

References Acland, G.M., Aguirre, G.D., Ray, J., Zhang, Q., Aleman, T.S., Cideciyan, A.V., Pearce-Kelling, S.E., Anand, V., Zeng, Y., Maguire, A.M., et al., 2001. Gene therapy restores vision in a canine model of childhood blindness. Nat. Genet. 28, 92 –95. Akova, Y.A., Bulut, S., Dabil, H., Duman, S., 1999. Late bleb-related endophthalmitis after trabeculectomy with mitomycin C. Ophthalmic. Surg. Lasers 30, 146–151. Angella, G.J., Sherwood, M.B., Balasubramanian, L., Doyle, J.W., Smith, M.F., van Setten, G., Goldstein, M., Schultz, G.S., 2000. Enhanced short-term plasmid transfection of filtration surgery tissues. Invest. Ophthalmol. Vis. Sci. 41, 4158–4162. Avila, M., Ortiz, G., Lozano, J.M., Franco, P., de Perez, G., Patarroyo, M.E., 1998. The effects of RGD (Arg-Gly-Asp) peptides on glaucoma filtration surgery in rabbits. Ophthalmic. Surg. Lasers 29, 309 –317. Bennett, J., Maguire, A.M., 2000. Gene therapy for ocular disease. Mol. Ther. 1, 501–505. Bennett, J., Wilson, J., Sun, D., Forbes, B., Maguire, A., 1994. Adenovirus vector-mediated in vivo gene transfer into adult murine retina. Invest. Ophthalmol. Vis. Sci. 35, 2535–2542. Bennett, J., Zeng, Y., Bajwa, R., Klatt, L., Li, Y., Maguire, A.M., 1998. Adenovirus-mediated delivery of rhodopsin-promoted bcl-2 results in a delay in photoreceptor cell death in the rd/rd mouse. Gene Ther. 5, 1156–1164. Borras, T., Matsumoto, Y., Epstein, D.L., Johnson, D.H., 1998. Gene transfer to the human trabecular meshwork by anterior segment perfusion. Invest. Ophthalmol. Vis. Sci. 39, 1503–1507. Cordeiro, M.F., Gay, J.A., Khaw, P.T., 1999. Human anti-transforming growth factor-beta2 antibody: a new glaucoma anti-scarring agent. Invest. Ophthalmol. Vis. Sci. 40, 2225–2234. Daniels, J.T., Occleston, N.L., Crowston, J.G., Cordeiro, M.F., Alexander, R.A., Wilkins, M., Porter, R., Brown, R., Khaw, P.T., 1998. Understanding and controlling the scarring response: the contribution of histology and microscopy. Microsc. Res. Tech. 42, 317 –333. Di Polo, A., Aigner, L.J., Dunn, R.J., Bray, G.M., Aguayo, A.J., 1998. Prolonged delivery of brain-derived neurotrophic factor by adenovirusinfected Muller cells temporarily rescues injured retinal ganglion cells. Proc. Natl Acad. Sci. USA 95, 3978–3983. Fantes, F.E., Heuer, D.K., Parrish 2nd, R.K., Sossi, N., Gressel, M.G., 1985. Topical fluorouracil. Pharmacokinetics in normal rabbit eyes. Arch. Ophthalmol. 103, 953–955. Gartel, A.L., Serfas, M.S., Tyner, A.L., 1996. p21—negative regulator of the cell cycle. Proc. Soc. Exp. Biol. Med. 213, 138– 149. Goldenfeld, M., Krupin, T., Ruderman, J.M., Wong, P.C., Rosenberg, L.F., Ritch, R., Liebmann, J.M., Gieser, D.K., 1994. 5-Fluorouracil in initial

S.F. Wen et al. / Experimental Eye Research 77 (2003) 355–365 trabeculectomy. A prospective, randomized, multicenter study. Ophthalmology 101, 1024–1029. Grisanti, S., Diestelhorst, M., Heimann, K., Krieglstein, G., 1999. Cellular photoablation to control post operative fibrosis in a rabbit model of filtration surgery. Br. J. Ophthalmol. 83, 1353–1359. Hauswirth, W.W., Lewin, A.S., 2000. Ribozyme uses in retinal gene therapy. Prog. Retin. Eye Res. 19, 689– 710. Hoffman, L.M., Maguire, A.M., Bennett, J., 1997. Cell-mediated immune response and stability of intraocular transgene expression after adenovirus-mediated delivery. Invest. Ophthalmol. Vis. Sci. 38, 2224–2233. Huyghe, B.G., Liu, X., Sutjipto, S., Sugarman, B.J., Horn, M.T., Shepard, H.M., Scandella, C.J., Shabram, P., 1995. Purification of a type 5 recombinant adenovirus encoding human p53 by column chromatography. Hum. Gene Ther. 6, 1403–1416. Jampel, H.D., Moon, J.I., 1998. The effect of paclitaxel powder on glaucoma filtration surgery in rabbits. J. Glaucoma 7, 170–177. Jampel, H.D., Thibault, D., Leong, K.W., Uppal, P., Quigley, H.A., 1993. Glaucoma filtration surgery in nonhuman primates using taxol and etoposide in polyanhydride carriers. Invest. Ophthalmol. Vis. Sci. 34, 3076–3083. Kaufman, P.L., Jia, W.W., Tan, J., Chen, Z., Gabelt, B.T., Booth, V., Tufaro, F., Cynader, M., 1999. A perspective of gene therapy in the glaucomas. Surv. Ophthalmol. 43, S91–S97. Kozarsky, K.F., McKinley, D.R., Austin, L.L., Raper, S.E., StratfordPerricaudet, L.D., Wilson, J.M., 1994. In vivo correction of low density lipoprotein receptor deficiency in the Watanabe heritable hyperlipidemic rabbit with recombinant adenoviruses. J. Biol. Chem. 269, 13695– 13702. Lewin, A.S., Drenser, K.A., Hauswirth, W.W., Nishikawa, S., Yasumura, D., Flannery, J.G., LaVail, M.M., 1998. Ribozyme rescue of photoreceptor cells in a transgenic rat model of autosomal dominant retinitis pigmentosa. Nat. Med. 4, 967– 971. Li, T., Adamian, M., Roof, D.J., Berson, E.L., Dryja, T.P., Roessler, B.J., Davidson, B.L., 1994. In vivo transfer of a reporter gene to the retina mediated by an adenoviral vector. Invest. Ophthalmol. Vis. Sci. 35, 2543–2549. Michou, A.I., Santoro, L., Christ, M., Julliard, V., Pavirani, A., Mehtali, M., 1997. Adenovirus-mediated gene transfer: influence of transgene, mouse strain and type of immune response on persistence of transgene expression. Gene Ther. 4, 473– 482. Mietz, H., Rump, A.F., Theisohn, M., Klaus, W., Diestelhorst, M., Krieglstein, G.K., 1995. Ocular concentrations of mitomycin C after extraocular application in rabbits. J. Ocul. Pharmacol. Ther. 11, 49–55. Murata, T., Kimura, H., Sakamoto, T., Osusky, R., Spee, C., Stout, T.J., Hinton, D.R., Ryan, S.J., 1997. Ocular gene therapy: experimental studies and clinical possibilities. Ophthalmic. Res. 29, 242–251. Nemunaitis, J., Swisher, S.G., Timmons, T., Connors, D., Mack, M., Doerksen, L., Weill, D., Wait, J., Lawrence, D.D., Kemp, B.L., et al., 2000. Adenovirus-mediated p53 gene transfer in sequence with cisplatin to tumors of patients with non-small-cell lung cancer. J. Clin. Oncol. 18, 609–622. Otake, K., Ennist, D.L., Harrod, K., Trapnell, B.C., 1998. Nonspecific inflammation inhibits adenovirus-mediated pulmonary gene transfer and expression independent of specific acquired immune responses. Hum. Gene Ther. 9, 2207–2222. Perkins, T.W., Faha, B., Ni, M., Kiland, J.A., Poulsen, G.L., Antelman, D., Atencio, I., Shinoda, J., Sinha, D., Brumback, L., et al., 2002. Adenovirus-mediated gene therapy using human p21WAF-1/Cip-1 to prevent wound healing in a rabbit model of glaucoma filtration surgery. Arch. Ophthalmol. 120, 941–949. Rahman, A., Tsai, V., Goudreau, A., Shinoda, J.Y., Wen, S.F., Ramachandra, M., Ralston, R., Maneval, D., LaFace, D., Shabram, P., 2001. Specific depletion of human anti-adenovirus antibodies facilitates transduction in an in vivo model for systemic gene therapy. Mol. Ther. 3, 768–778.

365

Sakamoto, T., Kimura, H., Scuric, Z., Spee, C., Gordon, E.M., Hinton, D.R., Anderson, W.F., Ryan, S.J., 1995. Inhibition of experimental proliferative vitreoretinopathy by retroviral vector-mediated transfer of suicide gene. Can proliferative vitreoretinopathy be a target of gene therapy? Ophthalmology 102, 1417–1424. Schuler, M., Herrmann, R., De Greve, J.L., Stewart, A.K., Gatzemeier, U., Stewart, D.J., Laufman, L., Gralla, R., Kuball, J., Buhl, R., et al., 2001. Adenovirus-mediated wild-type p53 gene transfer in patients receiving chemotherapy for advanced non-small-cell lung cancer: results of a multicenter phase II study. J. Clin. Oncol. 19, 1750– 1758. Schulick, A.H., Vassalli, G., Dunn, P.F., Dong, G., Rade, J.J., Zamarron, C., Dichek, D.A., 1997. Established immunity precludes adenovirusmediated gene transfer in rat carotid arteries. Potential for immunosuppression and vector engineering to overcome barriers of immunity. J. Clin. Invest. 99, 209–219. Sewell, D.A., Li, D., Duan, L., Westra, W.H., O’Malley Jr., B.W., 1997. Safety of in vivo adenovirus-mediated thymidine kinase treatment of oral cancer. Arch. Otolaryngol. Head Neck Surg. 123, 1298–1302. Shabram, P.W., Giroux, D.D., Goudreau, A.M., Gregory, R.J., Horn, M.T., Huyghe, B.G., Liu, X., Nunnally, M.H., Sugarman, B.J., Sutjipto, S., 1997. Analytical anion-exchange HPLC of recombinant type-5 adenoviral particles. Hum. Gene Ther. 8, 453–465. Shields, M.B., Scroggs, M.W., Sloop, C.M., Simmons, R.B., 1993. Clinical and histopathologic observations concerning hypotony after trabeculectomy with adjunctive mitomycin C. Am. J. Ophthalmol. 116, 673– 683. Skaf, M., di Martino, D.S., de Arruda Mello, P.A., Varma, R., 2001. Adenoviral-mediated gene transfer to the filtering bleb in rabbits. J. Glaucoma 10, 470 –476. Skuta, G.L., Parrish 2nd, R.K., 1987. Wound healing in glaucoma filtering surgery. Surv. Ophthalmol. 32, 149–170. Sterman, D.H., Treat, J., Litzky, L.A., Amin, K.M., Coonrod, L., MolnarKimber, K., Recio, A., Knox, L., Wilson, J.M., Albelda, S.M., Kaiser, L.R., 1998. Adenovirus-mediated herpes simplex virus thymidine kinase/ganciclovir gene therapy in patients with localized malignancy: results of a phase I clinical trial in malignant mesothelioma. Hum. Gene Ther. 9, 1083–1092. Suner, I.J., Greenfield, D.S., Miller, M.P., Nicolela, M.T., Palmberg, P.F., 1997. Hypotony maculopathy after filtering surgery with mitomycin C. Incidence and treatment. Ophthalmology 104, 207–214. Discussion 214– 215. Wen, S.F., Xie, L., McDonald, M., DiGiacomo, R., Chang, A., Gurnani, M., Shi, B., Liu, S., Indelicato, S.R., Hutchins, B., Nielsen, L.L., 2000. Development and validation of sensitive assays to quantitate gene expression after p53 gene therapy and paclitaxel chemotherapy using in vivo dosing in tumor xenograft models. Cancer Gene Ther. 7, 1469–1480. Wills, K.N., Maneval, D.C., Menzel, P., Harris, M.P., Sutjipto, S., Vaillancourt, M.T., Huang, W.M., Johnson, D.E., Anderson, S.C., Wen, S.F., et al., 1994. Development and characterization of recombinant adenoviruses encoding human p53 for gene therapy of cancer. Hum. Gene Ther. 5, 1079–1088. Wong, P.C., Ruderman, J.M., Krupin, T., Goldenfeld, M., Rosenberg, L.F., Shields, M.B., Ritch, R., Liebmann, J.M., Gieser, D.K., 1994. 5Fluorouracil after primary combined filtration surgery. Am. J. Ophthalmol. 117, 149–154. Yang, Y., Li, Q., Ertl, H.C., Wilson, J.M., 1995. Cellular and humoral immune responses to viral antigens create barriers to lung-directed gene therapy with recombinant adenoviruses. J. Virol. 69, 2004–2015. Yang, Y., Nunes, F.A., Berencsi, K., Furth, E.E., Gonczol, E., Wilson, J.M., 1994. Cellular immunity to viral antigens limits E1-deleted adenoviruses for gene therapy. Proc. Nat. Acad. Sci. USA 91, 4407–4411. Zacharia, P.T., Deppermann, S.R., Schuman, J.S., 1993. Ocular hypotony after trabeculectomy with mitomycin C. Am. J. Ophthalmol. 116, 314– 326.