Repeated Administration of Adenovector in the Eye Results in Efficient Gene Delivery

Abstract

purpose. To determine whether repeat administration of an adenovector (Ad) into the eye results in efficient gene delivery and to test whether transgenes can be expressed from an adenovector expression system in the presence of preexisting, neutralizing anti-Ad antibodies.

methods. To assess the efficiency of repeated gene delivery of an adenovector expression system, C57Bl/6 mice received one, two, or three injections (intravitreal [IVT] or periocular [PO]) of AdNull.11D (empty cassette) at 2-week intervals, followed by a single AdLuciferase (AdL.11D) IVT or PO injection. Mice were killed approximately 24 hours after AdL.11D injection and the eyes were enucleated and stored until assayed. Serum samples were also analyzed to determine whether repeated IVT or PO injections lead to induction of neutralizing antibodies directed against an adenovector delivery system. To determine whether preexisting neutralizing anti-Ad antibodies would block transgene expression, mice were preimmunized with one, two, or three intramuscular (IM) injection(s) of AdNull.11D (1 × 10⁹ particle units [pu]). Fourteen days later, when systemic anti-Ad antibody titers were expected to exist, mice were given a single AdL.11D injection (IVT or PO) and killed, and the eyes and serum collected.

results. These studies show that multiple injections at 2-week intervals with adenovectors (IM, IVT, or PO) did not prevent transgene expression in the eye. Moreover, measurement of neutralizing anti-Ad antibody titers revealed that measurable anti-Ad antibody titers in mice did not ablate transgene expression.

conclusions. These studies suggest that transgene expression after repeated adenovector administration into the eye is feasible and repeated injections, whether given IVT or PO, do not lead to an immediate increase in neutralizing anti-Ad antibody titers. Moreover, preimmunization of mice by systemic exposure to adenovector, does not block transgene expression in the eye. These studies indicate that repeat administration of adenovectors (IVT and PO) into the eye can be considered in designing future clinical trials and that the pre-existence of neutralizing anti-Ad antibodies probably does not mitigate activity.

Age-related macular degeneration (AMD), a deterioration of the central portion of the retina, is the chief cause of severe and irreversible vision loss in aging populations. ¹ AMD can be categorized into an atrophic (dry) form and a neovascular (wet, exudative) form. The atrophic form involves alterations of pigment distribution, loss of retinal pigment epithelial (RPE) cells and photoreceptors, and diminished retinal function due to an overall atrophy of the cells. The neovascular AMD involves proliferation of abnormal choroidal vessels, which penetrate the Bruch’s membrane and RPE layer into the subretinal space, and thereby may form extensive clots and/or scars. This eventually leads to the formation of choroidal neovascular (CNV) membranes, from which blood and serum may leak. ² The fluid released from the CNV can damage the structure and function of the overlying retina, including the central macular area, leading to the loss of central vision. The exact cause of CNV is not clear; however, it is often associated with a buildup of abnormal extracellular deposits in the form of soft drusen between the aging RPE and Bruch’s membrane, resulting in localized areas of ischemia and triggering angiogenesis. ³  

Adenovectors are a useful protein expression system for the treatment of ocular diseases, such as AMD. ⁴ They efficiently transduce a variety of ocular cells after intravitreal (IVT) ⁵ and periocular (PO) ⁶ administration; however, transgene expression is transient and is known to decrease from initial levels. ^{5 7 8} In the case of IVT delivery, cells specifically transduced are cells of the iris, ciliary body, trabecular meshwork, and corneal endothelium. ⁵ After PO delivery, cells transduced include the sclera, episclera, and surrounding orbital connective tissue. ⁶ Important aspects of successful adenovirus gene transfer or therapeutic purposes include the amount and persistence of gene expression, the ability to readminister vector, and the localization of vector-directed gene expression to target organs. ⁹ Because repeat administration of adenovector may then be necessary to maintain therapeutic expression, we undertook a series of studies to determine whether repeat administration of an adenovector in the eye is feasible, and to evaluate transgene expression after single and multiple IVT or PO injections. In other tissues, it has been suggested that immunologic limitation to prolonged transgene expression is the development of adenovector-specific neutralizing antibodies after primary administration, which may hinder repeat administration of adenovector. ^{10 11} Because the eye is an immune privileged organ, we also tested whether preexisting neutralizing anti-Ad antibodies would block transgene expression in the eye by inducing anti-Ad antibodies with single and multiple intramuscular (IM) injections of AdNull.11D. 

Materials and Methods

Animals

Female, C57Bl/6 mice were obtained from Harlan Laboratories (Chicago, IL) at 6 to 8 weeks of age. Before treatment, mice were anesthetized with a 100-μL intraperitoneal injection of ketamine hydrochloride (40-mg/kg) and xylazine (12-mg/kg; both from The Butler Company, Columbus, OH). Once the mice were anesthetized, a drop of 0.5% proparacaine hydrochloride (Bausch & Lomb, Rochester, NY) was administered as a topical anesthetic. All animals were treated according to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Adenoviral Vector Expressing Luciferase

Production and quantification of a type-5 adenoviral vector that expresses luciferase from a cytomegalovirus (CMV) immediate early promoter expression cassette has been described. ¹² The vector (AdL.11D) is devoid of the E1A, E1B, E3, and E4 adenovirus replication regions. A similar vector lacking an expression cassette (AdNull.11D) was also used. 

Adenoviral Vector Injections

IVT injections of vectors were performed using a pump microinjection apparatus (Harvard Apparatus, Holliston, MA) and pulled-glass micropipettes, as previously described. ¹³ Each micropipette was calibrated to deliver approximately 2 μL of vehicle containing the appropriate number of viral particles, on depression of a foot switch. The mice were anesthetized, pupils were dilated, and under a dissecting microscope (Nikon, Melville, NY), the sharpened tip of the micropipette was passed through the sclera, just behind the limbus into the vitreous cavity, and the foot switch was depressed. Depression of the foot switch caused a jet injection of fluid to penetrate the vitreous space. 

Periocular injections were performed with a cannula–syringe system, a 100-μL glass syringe (Hamilton Co., Reno, NV) fitted with a 33-gauge cannula, placed into a syringe pump (Harvard Apparatus). The mice were anesthetized as previously described. Under a stereofluorescence microscope with automated focus capabilities (Leica Microsystems, Deerfield, IL), the cannula was inserted under the conjunctiva. A foot pedal was depressed to signal the syringe pump to inject 4 μL of vehicle containing the appropriate number of viral particles. 

Intramuscular injections of vector were delivered in a total volume of 100 μL and administered in three equal injections to the right hind limb. 

Assessment of Luciferase Expression

Immediately after enucleation, eyes were snap frozen and stored at −80°C. Eyes were ground with a cold mortar and pestle on dry ice and then lysed (Reporter Lysis Buffer; Promega, Madison, WI). Lysates were analyzed with a luciferase assay system according to the manufacturer’s protocol. The total protein concentration was determined, to normalize the measurement of luciferase expression based on a Bradford dye-binding procedure with a protein assay (Bio-Rad Laboratories, Hercules, CA). 

Neutralizing Anti-Ad Antibody Assay

Neutralizing antibody titers were determined as previously described. ⁷ Titers were determined by analyzing the ability of serum antibody to inhibit infection of an Ad-green fluorescent protein (GFP; E1-, E3-deleted) serotype 5 adenovector expressing GFP from the CMV promoter) on AE25 cells, which complement for E1 functions and support the growth of E1-deleted adenovectors. AE25 cells were inoculated at 1 × 10⁴ per well on flat-bottomed 96-well plates and grown for 18 to 24 hours at 37°C. A series of two- or fourfold dilutions of the serum samples were incubated with AdGFP at a multiplicity of infection of 5 focal-forming units per cell for 1 hour at 37°C in minimal DMEM. This mixture was incubated with AE25 cells for 1 hour at 37°C, 100 μL of complete medium was added, and the cells were cultured overnight. The percentage of infected cells was determined with an imaging system (Typhoon 9200; GE Healthcare, Sunnyvale, CA), and confirmed with a fluorescence microscope (Nikon). The neutralizing antibody titer was scored as the reciprocal of the last dilution where a 50% reduction in green cells (infected cells) was observed. Because the neutralizing antibody is directed at the exterior protein coat of the adenovector, this assay is applicable to all serotype 5 adenovectors, independent of whether the vector is E1- and E3-deleted or E1-, E3-, and E4-deleted. All samples were assayed in duplicate. 

Statistical Analysis

Results are expressed as the mean ± SEM. An overall test for treatment effect was first performed with one-way analysis of variance (ANOVA). If the overall test indicated a significant treatment effect, individual treatments were compared with the naïve treatment group, using Bonferroni analyses, which adjusted for multiple comparisons. The level established for statistical significance was P < 0.05 (two-tailed test). All analyses were performed on computer (Origin 7.5; Origin Laboratories, Northampton, MA). 

Results

High-Level Gene Expression after Repeated Ocular Doses with the Adenovector

To assess repeated doses of the eye with an adenovector, we used a scheme that exposed mice to a vector without a marker gene (AdNull.11D) and then measured efficiency of delivery using the marker gene luciferase in AdL.11D. Mice were given IVT or PO AdNull.11D (1 × 10⁹ pu) one, two, or three times at 2-week intervals. Twenty-four hours before death, mice were given a single IVT or PO injection of AdL.11D (1 × 10⁹ pu), and luciferase activity was determined. Naïve mice served as the negative (no injection) control; whereas, mice that received a single IVT or PO injection of AdL.11D (1 × 10⁹ pu) served as the positive control, because there are no prior interfering immune responses that might suppress marker (luciferase) gene expression. The experimental design is shown in Figure 1 . We found that single or repeated IVT or PO injections with adenovector (AdNull.11D) did not ablate luciferase expression after one, two, and up to three prior adenovector administrations (Fig. 2) . Although luciferase expression declined modestly after IVT delivery, luciferase activity remained approximately 1000-fold above the noninjected control, even after four closely spaced IVT (Fig. 2A)or PO (Fig. 2B)administrations. This result was surprising, because repeated administration in other tissues (i.e., liver, lung, and pancreas) at 2 weeks after administration of adenovectors leads to ablated responses attributable to increasing titers of circulating neutralizing adenovector antibodies. ^{8 10 14 15 16 17 18} Of interest, repeated PO administration resulted in minimal decreases in gene expression, even less than repeated IVT administration. This was somewhat surprising, as it was anticipated that the periocular space would be a more systemically exposed space than the intraocular space. Also, of interest is that the expression profile after a second administration (e.g., IVT) mimicked the initial administration with respect to maximum transgene levels and expression duration (McVey D, et al., manuscript in preparation). Thus, these data demonstrate that repeated IVT or PO adenovector delivery to the eye (up to four times) does not ablate transgene expression, but rather gene expression is still possible. 

Neutralizing Anti-Ad Antibody Activity after Repeated Doses to the Eye

To assess whether IVT or PO injections of adenovector induce neutralizing antibodies, we collected serum from the same animals after single or multiple injections of AdNull.11D in a delivery scheme similar to that described for experiments measuring gene delivery (Fig 2) . One, two, and three repeat IVT or PO injections of adenovector stimulated neutralizing anti-Ad antibody serum titers (Fig. 3A 3B , respectively). This indicates that the gene delivery shown in Figure 2occurred even when an immune response to the vector existed. These data then indicate the feasibility of repeated IVT or PO administration, even in the presence of circulating neutralizing adenovector antibody titers. That robust neutralizing (>1:10,000) anti-Ad antibody titers were not generated after single or repeated ocular administrations of vector (up to four times) confirms that the eye is immune-privileged. Titers of neutralizing anti-Ad antibody did not increase immediately, but gradually increased (with subsequent IVT or PO administrations). Furthermore, four IVT or PO administrations were necessary to achieve the same level of neutralizing anti-Ad antibody response observed after a single IM injection of AdNull.11D (1 × 10⁹ pu). 

Gene Expression in Preimmunized Animals after Ocular Administration

To further substantiate whether transgene expression in the eye is possible in the presence of preexisting circulating neutralizing anti-Ad antibodies, we purposely induced a systemic immune response by using repeated IM injections of AdNull.11D. The experimental design is shown in Figure 4 . Mice were injected with AdNull.11D (1 × 10⁹ pu) intramuscularly (one, two, or three times) at 2-week intervals. AdNull.11D (1 × 10¹⁰ pu) resulted in similar levels of neutralizing antibodies (data not shown). To measure gene transfer in the eyes, approximately 24 hours before death, mice received a single IVT or PO injection of AdL.11D (1 × 10⁹ pu). Naïve mice were used as the negative control, and mice receiving a single IVT or PO injection of AdL.11D (1 × 10⁹ pu) without preimmunization were used as the positive control for maximum gene expression. Results shown in Figure 5indicate that transgene expression after IVT or PO administration occurred, even in the presence of circulating anti-Ad antibody titers. Although gene expression declined after IVT delivery, as evident by reduction in luciferase expression, gene transfer and expression clearly was able to continue within the eye, despite previous systemic exposure of the animal to adenovector and in the presence of elevated neutralizing adenovector antibody titers compared with naïve animals (discussed later). Luciferase activity remained approximately 100- to 1000-fold above the noninjected control after three preimmunization IM injections in both the IVT (Fig. 5A)and PO (Fig. 5B)groups. 

Neutralizing Anti-Ad Antibody Activity in Preimmunized Animals

To confirm whether single or multiple IM injections would result in increased neutralizing anti-Ad antibodies, levels were measured in serum samples processed from whole blood collected at the time the animals were killed (Fig. 6) . We selected 2-week intervals between adenovector exposures because it has been previously shown that immune responses (i.e., cell-mediated and humoral) are induced 14 days after adenovector exposure. ^{10 11}  

Discussion

Adenovectors are currently undergoing clinical testing for treatment of exudative macular degeneration and retinoblastoma (Chevez-Barrios P, et al. IOVS 2004;45:ARVO E-Abstract 3360). ¹⁹ Adenovectors are of interest to ocular applications, because they have been shown to transduce both the anterior and posterior segments of the eye, ⁵ and their efficiency in gene transfer is high. ²⁰ However, gene expression with adenovectors typically are reported to be transient. Repeated administration of adenovector may be desirable for some disease indications to obtain optimal therapeutic benefit. ²¹ Results in this article indicate that repeat IVT or PO administration of adenovector to the eye can result in repeated transgene expression, even when circulating immune responses to the vector are induced or are elevated compared with nontreated animals. 

Previous studies have indicated that in the practice of gene delivery, readministration of adenovectors may not result in additional gene transfer, particularly after intravenous, intraperitoneal, or intratracheal administration. ^{11 17 21} However, because the eye is an immune-privileged organ, we reasoned that IVT administration to the eye might respond differently. An IVT introduction of adenovectors may lead to a different induction profile of neutralizing anti-Ad antibodies than in other organs injected with adenovectors. ^{11 18 21} Thus, we tested the efficacy of gene transfer and expression after repeated IVT or PO adenovector administration to the eye and determined the neutralizing anti-Ad antibody titers. 

Our results are consistent with the concept that the eye is immune privileged, in that ocular injections produced a weak systemic immune response. Intraocularly, IVT delivery of adenovectors (E1, E3, and E4-deleted) can lead to nonsuppurative inflammation within the eye, but this effect is dose dependent. At the same dose, 1 × 10⁹ pu, as used in this study, the inflammation was minimal with some lymphocytes, a few plasma cells, occasional macrophages, and rare neutrophils and eosinophils. ²² Repeated IVT or PO administration of adenovector in the eye, even up to four times, at a brisk interval (2-weeks) was necessary, to induce circulating neutralizing antibodies equivalent to a single IM injection. Gene delivery experiments in the other systems indicate the rapid induction of multiple immune responses against viral and transgene proteins of adenovectors, including neutralizing antibodies and cytotoxic T lymphocyte (CTL) responses. ²³ B-cell activation and a humoral immune response are widely implicated in the inhibition of successful readministration of adenovectors for nonocular tissues. ^{9 10} After the initial exposure to adenovectors, neutralizing antibodies (IgG and IgA) are generated against major viral capsid proteins, which can attenuate infection of cells, even after the second administration in the heart and lung tissues. ^{9 10} The immune-privileged nature of the eye was also noted after subretinal injection, where minimal anti-Ad antibody production, and successful repeated doses were observed. ²⁴ Most interesting, we found that repeat periocular administrations induced very low neutralizing anti-adenovector titers. These results suggest that the periocular space may be immune privileged as well. 

Preimmunization of the animals before ocular injection of adenovectors indicated that the presence of an immune response has little impact on the efficiency of adenovector expression. Preimmunization of mice through intramuscular delivery, produced little change in adenovector delivery both intravitreally and periocularly. This indicates that elevated serum levels of neutralizing anti-Ad antibody levels do not necessarily ablate gene expression from adenovectors introduced into the eye. Our findings have several clinical implications. First, the presence of neutralizing anti-Ad antibody titers in the serum of patients does not necessarily correlate with transgene expression. Approximately 30% to 70% of the United States population has preexisting neutralizing anti-Ad antibody titers, ^{25 26} yet to date a tight correlation between serum levels of anti-Ad antibodies and any clinical parameter has not been shown. This is of particular interest, as previous clinical trials in other tissues indicate that humans with neutralizing anti-Ad antibody levels may respond to adenovector mediated gene delivery systems. ^{27 28 29} A second clinical implication of this article is to support the concept that repeated adenovector administration (IVT or PO) is feasible. Our results are consistent with an ongoing repeated dose trial with a first-generation adenovector (lacking E1 and E3) in children with retinoblastoma (Chevez-Barrios P, et al. IOVS 2004;45:ARVO E-Abstract 3360). To date, results from this trial indicate that repeated administration to the eye (≥1 e10 pu per injection) is well tolerated and does not induce or boost systemic immune responses to adenovirus and that patients show a clinical response to this experimental therapy (Chevez-Barrios P, et al. IOVS 2005;46:ARVO E-Abstract 5218). Taken together, these data indicate that repeated administration of adenovectors to the eye in patients can be considered an option for future clinical trial designs. 

 Supported by National Eye Institute Grant EY014494.

 Submitted for publication June 10, 2005; revised September 1, 2005; accepted November 18, 2005.

 Disclosure: M.M. Hamilton, GenVec, Inc. (E); D.E. Brough, GenVec, Inc. (E); D. McVey, GenVec, Inc. (E); J.T. Bruder, GenVec, Inc. (E); C.R. King, GenVec, Inc. (E); L.L. Wei, GenVec, Inc. (E)

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Figure 1.

 

Multiple adenovector injections in the eye. Experimental design for testing gene delivery after single and multiple administrations of adenovectors. Mice were given AdNull.11D (1 × 10⁹ pu) IVT or PO (one, two, or three times) every 2 weeks. Twenty-four hours before death, mice were given a single IVT or PO injection of AdL.11D (1 × 10⁹ pu), and luciferase activity was assessed. Naïve mice and mice given a single IVT or PO injection of AdL.11D (1 × 10⁹ pu) served as the control.

Figure 2.

 

Luciferase activity after multiple injections in the eye. Luciferase activity in whole eyes after one, two, or three IVT (A) or PO (B) injections of AdNull.11D, followed by a single IVT (A) or PO (B) injection of AdL.11D. Data are expressed as the mean ± SEM (∼20–30 mice/treatment group) for both IVT (A) and PO (B) administrations. There is a statistically significant difference between results in the treatment and naïve groups (P < 0.05, one-way ANOVA and two-tailed t-test).

Figure 3.

 

Neutralizing anti-Ad antibody activity after multiple injections in the eye. The neutralizing anti-Ad antibody titer for each sample is described as the highest dilution of serum that resulted in a 50% reduction in GFP expression in AE25 cells. Data are the mean ± SEM (error bars) of results of one experiment (10–20 mice per treatment group per experiment) for both IVT (A) and PO (B) administrations. There is a statistically significant difference between results in the treatment and naïve groups (P < 0.05, one-way ANOVA). Furthermore, difference in the results between the treatment groups (except for the AdL only control, P = 0.08 and 0.26; IVT and PO, respectively) and the naïve group is greater than would be expected by chance.

Figure 4.

 

Adenovector administration to the eye in preimmunized animals. Experimental design for testing transgene expression in the eye in the presence of preexisting circulating neutralizing anti-Ad antibodies. Mice were injected with AdNull.11D (1 × 10⁹ pu) intramuscularly (one, two, or three times) at 2-week intervals. Twenty-four hours before death, mice received a single IVT or PO injection of AdL.11D (1 × 10⁹ pu). Naïve mice and mice with a single IVT or PO injection of AdL.11D (1 × 10⁹ pu) alone served as control.

Figure 5.

 

Luciferase activity after ocular delivery in preimmunized animals. Luciferase activity in whole eyes after one, two, or three IM injections of AdNull.11D followed by a single IVT (A) or PO (B) injection of AdL.11D. Data represent the mean ± SEM (error bars) of results in three separate experiments (five mice per treatment group per experiment) for both IVT (A) and PO (B) administrations. The decrease in mean luciferase activity after two IM injections was attributed to 3 of 15 eyes, resulting in low luciferase expression. There is a statistically significant difference between results in the treatment and naïve groups (P < 0.05, one-way ANOVA and two-tailed t-test).

Figure 6.

 

Neutralizing anti-Ad antibody activity in preimmunized animals after ocular injections of adenovector. High systemic neutralizing anti-Ad antibody titers are produced after IVT (A) or PO (B) delivery of adenovectors. The neutralizing anti-Ad antibody titer for each sample is described as the highest dilution of serum that resulted in a 50% reduction in GFP expression in AE25 cells. Data are expressed as the mean ± SEM (error bars) of results of one experiment (five mice per treatment group per experiment) for both IVT (A) and PO (B) administrations. There is a statistically significant difference between results in the treatment and naïve groups (P < 0.05, one-way ANOVA). Furthermore, differences in the results in the treatment groups (except for the AdL only control, P = 0.08 and 0.21; IVT and PO, respectively) compared with the naïve is greater than would be expected by chance.