A Deviant Immune Response to Viral Proteins and ... - Cell Press

2 downloads 0 Views 252KB Size Report
A Deviant Immune Response to Viral Proteins and. Transgene Product Is Generated on Subretinal. Administration of Adenovirus and Adeno-associated Virus.
doi:10.1006/mthe.2002.0525, available online at http://www.idealibrary.com on IDEAL

ARTICLE

A Deviant Immune Response to Viral Proteins and Transgene Product Is Generated on Subretinal Administration of Adenovirus and Adeno-associated Virus Vibha Anand, Bethany Duffy, Zaixin Yang, Nadine S. Dejneka, Albert M. Maguire, and Jean Bennett* University of Pennsylvania, F. M. Kirby Center for Molecular Ophthalmology, 310 Stellar Chance Labs, Scheie Eye Institute, 422 Curie Blvd, Philadelphia, Pennsylvania 19104-6069, USA *To whom correspondence and reprint requests should be addressed. Fax: (215) 573-7155. E-mail: [email protected].

The immune response after ocular exposure to foreign antigens varies substantially from that of a typical systemic response. Anterior chamber associated immune deviation (ACAID) has been well documented. The immune response of the subretinal space has not been studied in as much detail. Here, we characterized the immune response of the subretinal space when it encounters the antigens AdV-GFP and AAV-GFP (recombinant adenovirus or adeno-associated virus, respectively), each delivering the reporter gene encoding green fluorescent protein (GFP). Results indicate that the subretinal space possesses an immune-deviant property similar to ACAID. AdVelicited immune responses following subretinal injections are significantly reduced compared with systemic responses elicited by intradermal injections of the same virus. Furthermore, subretinal AdV administration results in transduction of retinal pigment epithelial cells (RPE), which are the potential antigen presenting cells of the retina. This subsequently generates a population of immunosuppressive Th2-type, cytokine-secreting, splenic T cells. This response may be advantageous to the development of ocular gene therapy. Key Words: adenovirus, adeno-associated virus, subretinal space, immune privilege, immune deviation

INTRODUCTION Induction of an immune-deviant response in the eye on encounter with antigen has been well documented-especially in the context of the anterior chamber. The immune response in the posterior part of the eye has not been characterized in as much detail. The studies so far suggest that this area may also have immune-deviant features. Of note, Streilein et al. [1–3] reported an immune-deviant response against soluble and cell-bound antigens in the subretinal space. This space is able to accept retinal pigment epithelium (RPE) allografts and neonatal ocular tissue and sustain them without induction of a significant delayed type hypersensitivity (DTH) response [1]. The subretinal space is the area between the RPE and the photoreceptors. It may be accessed during a retinal detachment, a process that causes the separation of photoreceptors and RPE. A physical barrier, created on the outer side by the RPE and on the inner side by the intracellular junctions between the Muller cells, forms the geographical border and separates the subretinal space from

MOLECULAR THERAPY Vol. 5, No. 2, February 2002 Copyright © The American Society of Gene Therapy 1525-0016/02 $35.00

the blood supply. Apart from the physical barrier, it is postulated that the subretinal space possesses molecules that contribute to the “immune privilege” of this site. One such molecule may be transforming growth factor- (TGF-), which has been postulated to have a role in inducing different phases of anterior chamber associated immune deviation (ACAID) [4]. Viral vectors such as adenovirus (AdV) and adeno-associated virus (AAV) have been used in several ocular gene therapy studies and may target photoreceptors and RPE after subretinal injection; consequently, it is important to characterize the immune responses to these viruses in the subretinal space. Induction of a cell-mediated response can limit gene expression following AdV delivery in the eye [5,6]. Various immunomodulation strategies, such as immunosuppression [7] and modulation of antigen presenting function by delivery of CTLA4-Ig [5,8], have been evaluated for their ability to prolong transgene expression following virus delivery into the retina.

125

ARTICLE

doi:10.1006/mthe.2002.0525, available online at http://www.idealibrary.com on IDEAL

TABLE 1: Systemic humoral responses following subretinal or intradermal injections of AAV-GFP and AdV-GFP Subretinal injections AAV-GFP

AdV-GFP

Day

Total IgG

NAb

Total IgG

NAb

0

0.12 ± 0.01

20

0.27 ± 0.04

20

21

0.38 ± 0.22

20

1.62 ± 0.79*

20

Intradermal injections AAV-GFP

AdV-GFP

Day

Total IgG

NAb

Total IgG

NAb

0

0.15 ± 0.06

20

0.09 ± 0.02

20

21

0.51 ± 0.06

640

2.39 ± 0.09*

1280

Serum antibody titers (total and neutralizing (NAb)) against AAV and AdV proteins following subretinal and intradermal injections as measured by ELISA. Total antibody levels (as reflected by absorbance at 405 nm) are presented as average values in 10 mice. The NAb levels are presented as reciprocal dilutions. Serum Ab levels were measured at day 0 (before injection) and day 21 for both subretinal and intradermal groups. *P < 0.0001: day 21 AAV versus AdV subretinal injection; day 21 AdV subretinal versus intradermal injection.

but the AdV produced significantly higher antibody levels (P < 0.001). These antibodies were non-neutralizing in nature. For the neutralizing antibody assay, the intensity of GFP was captured digitally in the wells and was inversely proportional to the concentration of neutralizing antibodies. An increase in the amount of GFP, as assessed with fluorescence microscopy, correlated with a higher neutralizing antibody (Nab) titer detected with the Fluoroimager. Intradermal injection (Table 1) resulted in a significant increase in both anti-AAV and anti-AdV antibodies. AdV-injected animals had higher antibody levels than AAV-injected animals and intradermal injection of either vector resulted in neutralizing antibodies. There was a higher neutralizing antibody response for AdV than for AAV (1280 versus 640 reciprocal dilutions, respectively). Isotyping of the responses (Table 2) after subretinal administration revealed predominant induction of IgG2b levels for rAAV (P < 0.0001), and IgG2b and IgG3 levels for AdV (P < 0.0001). In contrast, IgG2a levels increased significantly in the intradermal AdV-injected animals. Humoral Response to GFP Antibody induction against the transgene protein GFP was also evaluated following both subretinal and intradermal injection of AAV and AdV. There was no significant antibody response to GFP after delivery of AAV. However, AdV administration (both intradermal and subretinal) induced significant anti-GFP antibodies (Table 3).

Ocular administration of antigen affects the cytokine milieu and the overall immune status of the eye. Administration of antigen to the anterior chamber can result in an antigen-specific downregulation of Th1 responses, IL-12 production by the antigen presenting cells (APCs), and induction of a cross-regulatory Th2 response. Particulate or cell-associated antigens may also induce IL- Cell-Mediated Responses to Viral Proteins and GFP 12 and Th1 responses after anterior chamber injection [9]. following Subretinal and Intradermal Administration As shown previously [10], herpes simplex virus 1 (HSV1) Bright-field microscopic examination of sections (from was able to induce ACAID following anterior chamber injections. We examined whether antigen-speTABLE 2: Isotyping of systemic humoral response following cific immune deviation exists in the subretinal space. injections of AAV-GVP or Adv-GVP We also evaluated the possibility that AdV and AAV Subretinal injections serve as cell-associated antigens and induce similar Post-treatment (day 21) immune responses in the subretinal space. Here we Antibody isotype Pretreatment demonstrate that recombinant viruses can be used to AAV AdV deliver genes to the subretinal space without induc- IgG2a 0.07 ± 0.01 0.17 ± 0.05* 0.98 ± 0.18 ing unwanted deleterious responses. The nature of IgG2b 0.120 ± 0.01 0.56 ± 0.33* 1.35 ± 0.40 the response differs, however, after delivery of AdV 0.08 ± 0.01 0.27 ± 0.06* 2.39 ± 0.37 and AAV. The immunological reactions to delivery of IgG3 viral vectors in the eye will have an impact on the success of ocular gene therapy. Intradermal injections Antibody isotype

RESULTS

Pretreatment

Post-treatment (day 21) AAV

AdV

Subretinal and Intradermal Injections of Both IgG2a 0.087 ± 0.02 0.68 ± 0.15* 2.97 ± 0.25 AAV and AdV Induce a Systemic Humoral IgG2b 0.12 ± 0.01 0.42 ± 0.09* 2.35 ± 0.57 Response to Viral Proteins 0.08 ± 0.00 0.20 ± 0.04* 2.09 ± 0.73 We injected animals subretinally and intradermally IgG3 with AdVGFP or AAV-GFP after obtaining pretreat- Isotype antibody responses to AAV and AdV in serum after subretinal or intradermal injections. ment serum samples. At 3 weeks following subretinal Mean + SD of absorbance (405 nm) values in sera of cohorts of 10 mice, before treatment and 3 weeks after subretinal/intradermal injections of 13 109 IU of AAV-GFP or AdGFP. injection, ELISA analysis revealed that both AAV and *P < 0.0001; IgG2b, AAV versus AdV after subretinal injection; IgG2a, IgG3, AAV versus Adv after AdV induced a systemic humoral response (Table 1), subretinal injection; IgG2a, IgG2b, IgG3, AAV versus AdV after intradermal injection.

126

MOLECULAR THERAPY Vol. 5, No. 2, February 2002 Copyright © The American Society of Gene Therapy

doi:10.1006/mthe.2002.0525, available online at http://www.idealibrary.com on IDEAL

ARTICLE

FIG. 1. DTH measurements following injection of AdV-GFP or AAVGFP. Animals were injected subretinally or intradermally with AdVGFP or AAV-GFP (50 l of 1  109p/ml). Animal footpads were measured 7 days following injection. Animals were subsequently challenged with either AdV-GFP or AAV-GFP in their right footpads. 24 hours post-challenge, footpad measurements were obtained; a swelling or increase in thickness would be indicative of a DTH response. Animals treated with an initial intradermal injection of AdV-GFP exhibited a significant increase in footpad thickness (*), indicative of a DTH response. This response was absent from all other groups.

cells was confirmed by assaying for the release of immunosuppressive cytokines such as IL-4, IL10, and TGF-1 using Elispot. In contrast, a proinflammatory response developed in animals receiving naive T cells following subconjunctival both AdV- and AAV-injected eyes) adjacent to transduced injection. Naive T cells were not immunosuppressive and hence failed to halt proinflammatory responses. Animals cells revealed no inflammatory cells, and immunohistochemical analysis confirmed the absence of CD4-, CD8-, receiving naive T cells thus developed a DTH response. Cell-mediated responses were analyzed in vitro using Tand CD16-positive cells (data not shown). Mice injected subretinally with rAdV-GFP or rAAV-GFP cell proliferation assays. Elispot assay was used to detect were evaluated for DTH responses by footpad challenge. the induction of cytokines specific to adenoviral and GFP Figure 1 indicates that there was no significant DTH antigens. In the adenovirus-specific cytokine assay, all response in animals primed with AAV or AdV (subreti- cytokine levels were elevated 24 hours after subretinal nally) following challenge with AAV, adenoviral proteins AV-GFP injection compared with unstimulated controls. or GFP. In contrast, animals primed with AdV via intra- At 7 days, spleen-derived T cells primarily produced the immunosuppressive cytokines IL-4, IL-10, and TGF-, dermal injection developed an intense DTH following whereas the pro-inflammatory cytokines IL-2, IFN-, and challenge by either the adenoviral proteins, or GFP. Adoptive transfer of T cells can be used to character- TNF- were maintained close to baseline (unstimulated) ize cell-mediated responses. As the results from the DTH levels (Fig. 3A). The 7-day group had a greater number of spot-forming T cells (SFTs) for all groups than the 24studies showed that AAV was unable to elicit substantial cell-mediated responses, animals injected with rAdV alone hour group. Similar cytokine profiles were observed in the GFP-specific cytokine assay (Fig. 3B). served as donors for spleen-derived T cells in adoptive Control animals receiving intradermal injection of transfer experiments. The cells were administered to naive AV-GFP were also analyzed for cytokine induction at the mice via tail vein injection and their ability to suppress DTH was monitored. Control animals received spleno- corresponding time points by T-cell proliferation assay. cytes and lymph-node-derived T cells from naive animals. In both the adenovirus and GFP-specific cytokine assays, there was a significant induction of IL-2, IFN-, and TNFA significant suppression of DTH was observed in animals that had received adoptive transfer of T cells from the  at 24 hours following intradermal injection of AV-GFP subretinally injected group following intradermal or sub- (Figs. 3C and 3D), but levels of IL-2, IFN-, IL-4, IL-10, and TGF- were similar to unstimulated controls. At 7 conjunctival challenge. Because an intradermal or subconjunctival challenge stimulates a pro-inflammatory T- days, levels of TNF- remained elevated and there was a significant induction of IL-2 and IFN-, whereas levels of cell population, both of these procedures serve to detect IL-4, Il-10, and TGF- were close to baseline (Figs. 3C DTH suppression (Fig. 2). Function of the transferred T and 3D). TABLE 3: Total IgG response to GFP following subretinal or RPE Cell Cultures intradermal injection of AAV-GFP or AdV-GFP Cultures of RPE cells displayed features typical of Pre-immune levels Subretinal Intradermal these cells [20], including hexagonal packing and AdV-GFP 0.11 ± .01 2.47 ± 0.25* 3.40 ± 0.08 acquisition of a flat, polygonal morphology after AAV-GFP 0.11 ± 0.03 0.16 ± 0.10* 0.13 ± 0.08 dissociation with trypsin. RPE cell specificity was confirmed using pan-cytokeratin staining; 70–80% *P < 0.001; AdV, subretinal versus intradermal; AdV versus AAV, subretinal. of cells stained positive (data not shown).

MOLECULAR THERAPY Vol. 5, No. 2, February 2002 Copyright © The American Society of Gene Therapy

127

ARTICLE

doi:10.1006/mthe.2002.0525, available online at http://www.idealibrary.com on IDEAL

Class II Induction on in Vitro Cultures and in Vivo AdV-GFP Transduced Mouse Retinas In cultures of RPE cells transduced with AdV-GFP, most of the infected cells stained positive for class II protein (Ia) confirming the antigen-presenting function of these cells. The positive controls included cells induced with 100 U/ml IFN- before staining with PE anti-mouse Ia (Fig. 4). An induction of I-Ab protein was also observed in mouse retinas transduced earlier with AdV-GFP. I-Ab protein was not detected in control saline injected retinas or AAV-GFP injected retinas (Fig. 4).

DISCUSSION Our study aimed to characterize the immune responses mounted to viral vectors (AdV or AAV) delivered to the subretinal space. While both viruses induced a systemic humoral response, neither virus induced neutralizing antibodies after delivery to the subretinal space. Intradermal injection of AdV, in contrast, induced high levels of neutralizing antibodies. The antibody titers were significantly higher against adenoviral proteins. Also, both viruses induced antibodies directed against GFP. Significant differences in the immune responses to these vectors, however, were noticed for cell-mediated immune response. AdV induced a cell-mediated response after systemic application. However, cellular infiltrate was not observed in eyes of animals that had been injected with AdV and then challenged systemically with the same virus. Subretinal injection of AdV resulted in suppressed humoral and cell mediated responses towards both viral antigens and GFP. AAV also induced suppressive humoral responses, similar to results from previous studies [11]. The immune response following subretinal injection of AdV (and AAV) is similar to that described in ACAID.

In ACAID, exposure of antigen to the anterior chamber environment results in a predominant Th2-like response [12]. There is suppression of DTH, but an induction of IgG1 and IgG2b (Th2-type) antibodies [13]. We found that AAV failed to invoke a DTH response regardless of the site of delivery. Suppression of DTH after subretinal delivery of AdV was quite remarkable, however. Like rAAV, subretinal injection of AdV elicited a predominant Th2 response (characterized by IgG2b and IgG3). Subretinal adenovirus injections also induced weaker levels of IgG2a. The responses were reversed after intradermal injection of AdV, where a predominant Th1 response (with increased IgG2a levels) was found. The contrast in immune response following subretinal (Th2) and systemic (Th1) exposure of AdV lends support to the notion that the subretinal space induces immune deviation. Additional support for immune deviation of the subretinal space comes from experiments evaluating suppression of DTH. We found that DTH towards viral proteins and GFP was suppressed after subretinal injection with AdV. This was in marked contrast to the results of intradermal injection. The reversal of DTH after subretinal delivery suggests that a suppressor T-cell population is generated after injection of AdV in this location and that this T-cell population downregulates the cytotoxic T-cell reaction mediating the DTH response. Finally, additional evidence of immune deviance of the subretinal space was furnished by the observation that adoptive transfer of T cells in mice that had been subretinally injected with AdV conferred immunosuppressive properties on naive mice. The transferred T cells protected these animals from proinflammatory responses and allowed prolonged transgene expression. What are the factors that result in immune deviation in the subretinal space? The physical/anatomical barriers [14] of the subretinal space may play a role in this phenomenon. We are presently testing this possibility by performing additional studies in animals with retinal degeneration, where the anatomical barriers have been disrupted by the disease process. Immunosuppressive cytokines and immunosuppressive regulatory T cells may also be involved. TGF- and IL-10 have been implicated in altering the immune response from a Th1 to a Th2 response in the anterior chamber [15]. TGF-2 [16] and other

FIG. 2. Adoptive transfer of immunosuppressive T cells induced by AdV in C57Bl/6 mice. Animals were injected subretinally with AdV-GFP (1 l/eye of 5  1012p/ml stock virus). T cells were harvested 7 days post-injection. T cells were also obtained from untreated controls. The cells (5 105 cells/ml) were then injected via tail vein into additional C57BL6 mice. In these new mice, one group of animals received naive T cells, whereas the other received the subretinally primed cells. At 7 days post administration, animals were administered Ad-GFP via intradermal or sub-conjunctival injection. One week postinjection, footpads were measured. Animals were subsequently challenged with AdV-GFP. After 24 hours, footpad measurements were obtained. Following adoptive transfer of subretinally primed T cells, a significant reduction in DTH response was observed in the right footpad (*). In comparison, no DTH suppression was observed in control animals that had received adoptive transfer of naive T cells.

128

MOLECULAR THERAPY Vol. 5, No. 2, February 2002 Copyright © The American Society of Gene Therapy

doi:10.1006/mthe.2002.0525, available online at http://www.idealibrary.com on IDEAL

A

B

C

D

ARTICLE

FIG. 3. In vitro analysis of cell mediated responses. Spleen-derived T cells were collected from mice following injection with AdV-GFP. ELISPOT assays were used to characterize virus-specific (A) and GFP-specific (B) cytokine profiles 24 hours and 7 days following subretinal injection of AdV-GFP. The same technique was performed to evaluate the AdV-specific (C) and GFP-specific (D) cytokine profiles following intradermal injection of AdV-GFP. Individual cytokines are plotted on the x axis, whereas the y axis indicates the number of spots/well in the assay. This latter number correlates with the number of T cells elaborating a particular cytokine. Hatched bars indicate control assays: “-”, unstimulated spleen-derived T-cells (- control); +, spleen-derived T-cells treated with the T-cell stimulant phorbol 12-myristate 13-acetate (+ control). *P = 0.001; **P < 0.0001.

secreted mediators [17] have also been proposed as factors which endow T cells with the ability to suppress the induction and expression of DTH. Results from the in vitro cytokine quantitation assay, Elispot, confirmed the immunosuppressive nature of these T cells. There was a predominant IL-10 and TGF- secreting T-cell population in both draining lymph nodes and spleens of subretinally injected animals directed against adenoviral proteins and GFP. Thus, immune-suppressive mechanisms similar to those present in the anterior chamber occur in the subretinal space. Other viruses may have this effect, as recently we have observed immune deviation after lentivirus delivery to the subretinal space (Petros Karakousis et al., manuscript submitted). Furthermore, our preliminary data suggest that the variable immune responsiveness of the retina might be due to differences in transduction of antigen-presenting cells by Adv and AAV. The cell-mediated response associated with AdV may be due to the fact that AdV infects APCs more efficiently than AAV. What are the APCs in the

MOLECULAR THERAPY Vol. 5, No. 2, February 2002 Copyright © The American Society of Gene Therapy

retina that are infected by AdV? Potential candidates include RPE cells and dendritic cells in the choroid or retinal microglia. In our preliminary findings, we show an activation of class II MHC proteins on RPE cells in retina following subretinal injections in mouse eyes as well as in vitro, following AdV infection in RPE cell primary cultures. AAV transduces RPE cells much less efficiently that AdV following subretinal injection or in vitro infection. This may contribute to the diminished cell-mediated responses of this virus. Such a finding has been reported [18]. Identification of APCs in the retina could give a clue as to the mechanisms by which the cell-mediated response after subretinal injection differs from that typically induced by systemic application of AdV. Immune deviation in the subretinal space may be advantageous for ocular gene therapy. This may allow viral vectors which may be toxic at systemic sites to be safely delivered to this region. Further delineation of the limits of this “immune privilege” should be made in order to assure that such vectors will do no harm. The elicitation

129

ARTICLE

doi:10.1006/mthe.2002.0525, available online at http://www.idealibrary.com on IDEAL

FIG. 4. RPE cells infected with AdV-GFP are positive for class II protein (I) indicating that these cells have an antigen-presenting function. (A) RPE cells possess green fluorescence following infection with AdV-GFP. (B) Ia was not detected in control saline-injected retinas. (C) Control RPE cells were induced to stain positive for Ia with 100 U/ml IFN-. (D) GFP and Ia colocalize in an RPE cell infected with AdV-GFP.

of immune deviation by subretinal injection may also have application for systemic gene therapy. It may be possible to take advantage of this phenomenon to suppress the immune response towards a vector/transgene being tested for treatment of a systemic disease.

MATERIALS

AND

METHODS

Virus preparation: AAV-EGFP and AdV-EGFP. All manipulations of recombinant viruses were carried out in accordance with institutional and national biosafety guidelines. The AAV-GFP virus was prepared by packaging the recombinant DNA into AAV particles by complementation as described [19,20]. The transgene cassette contained enhanced cDNA coding green fluorescent protein (EGFP; Clontech, Palo Alto, CA) under control of a cytomegalovirus (CMV) promoter. The recombinant virus was purified through three successive centrifugations in CsCl gradients. Titer of the purified virus was determined by infectious unit (i.u.) assay on 84-31 cells, an E1/E4-complementing cell line derived from 293 cells, and defined as i.u./ml [19]. The titer of the purified virus was 1.6  1011 i.u./ml. The number of virion DNA particles was also determined by slotblot analysis.

130

Purified rAAV-GFP was tested for adenovirus contamination through histochemical assay of the lacZ product, -galactosidase [20]. No -galactosidase activity was detected. This assay is capable of detecting one adenovirus particle in 106 cells. AAV-GFP was also tested for contamination with wild-type AAV. This was performed by infecting 293 cells with rAAVGFP in the presence of adenovirus and analyzing DNA harvested from the resulting lysate by Southern and western analyses using a Rep probe. Rep, reflecting the presence of wild-type AAV, was present at a level of approximately 0.1%. Ad.CMVGFP (provided by James Wilson, University of Pennsylvania) was prepared, purified, and titered as described [21]. The virus lacks the E1A and E3 sequences required for replication and instead contains the EGFPencoding cDNA and a CMV promoter. Adenovirus preparations were tested for and found to lack contamination with replication-competent virus by PCR as described [22]. Intraocular administration of recombinant virus. All animals were cared for in accordance with federal, state, and local regulations. Injections in cohorts of 10 adult (6- to 8-week-old immunocompetent C57Bl/6) mice were performed through a transscleral transchoroidal approach as described [22–24]. Briefly, after animals were anesthetized with avertin, a 30-gauge cannula was inserted into the subretinal space of the peripheral retina. The cannula was advanced and secured in the subretinal space. One l of virus-

MOLECULAR THERAPY Vol. 5, No. 2, February 2002 Copyright © The American Society of Gene Therapy

doi:10.1006/mthe.2002.0525, available online at http://www.idealibrary.com on IDEAL

containing solution (1  1010 particles) of AAV-GFP or Ad-GFP was injected, thereby raising a dome-shaped retinal detachment. Contralateral eyes were uninjected or injected with vehicle alone (Hepes-buffered saline, pH 7.8, containing 10% glycerol by volume). Injections were monitored by indirect ophthalmoscopy and by direct visualization through the operating microscope. Visualization of the detachment confirmed that the solution was injected into the targeted location. Exposure of RPE and photoreceptor cells to the subretinal injection solution was confirmed histologically at termination of the experiment by observation of GFP in those cells [19,25]. All injections in which blebs were not successfully raised were noted and analyzed separately. Experiments were repeated three times. For subconjunctival injections, virus was delivered using a 30-gauge needle in the fornix of the conjunctiva. Evaluation of immune responses. Humoral responses were evaluated by assay of serum antibody levels following both subretinal and intradermal injections using enzyme linked immunosorbent assay (ELISA) [11]. Cellmediated immune (CMI) responses were followed by a functional assay for memory T cells. These assays included in vitro assays of lymphoproliferative responses and in vivo assays of DTH and adoptive transfer. Responses were compared with those obtained after systemic administration of virus. Systemic administration was achieved through intradermal injections. These were performed in cohorts of 10 mice. Injections of 1  109 i.u. of rAAV-GFP or AdV-GFP contained in a total volume of 0.1 ml were made using a 30-gauge needle. The date of injection was designated as day 1. The animals were bled from the tail vein at different time points (0, 7, 14, and 21 days) after virus administration, and sera were isolated for assay of antibodies against viral proteins and the transgenic protein GFP. The antibodies were isotyped using ELISA and characterized for their neutralizing nature. DTH assay. Cell-mediated responses were evaluated via DTH assays. Briefly, 7 or 14 days following subretinal administration of rAdGFP, animals were divided into two groups. Baseline footpad thicknesses were measured with a micrometer (Warren-Knight Instrument Co, PA). Animals were then challenged with AdV or GFP (1  109 i.u. of rAAV-GFP/rAdGFP contained in a total volume of 0.1 ml or 50 g recombinant GFP protein (Clontech Lab Inc, Palo Alto, CA) via intradermal injections into the right footpad. After 24 hours, footpad measurements were repeated. Left footpad measurements served as a negative control. Similar assays were performed on animals injected subretinally with AAV-GFP. Two groups of animals were monitored. The first was challenged with AAV 1 week following subretinal administration. The second group was challenged with GFP 3 weeks following subretinal administration. Measurements of the footpads were repeated 24 hours after injections. The measurements were performed in triplicate and averaged. The extent of footpad swelling was recorded as the results of the 24-hour measurements minus the 0-hour measurements. Adoptive transfer assay. AdV-GFP was administered to C57Bl/6 mice (n = 10) via subretinal injection. Seven days following administration, both virus-exposed and saline-injected (naive) animals were sacrificed and spleen and lymph node derived T cells were isolated as described [26]. Thus, there were four different populations of cells: virus-exposed spleen, virus-exposed lymph node, naive spleen, and naive lymph node. These cells were then administered into four groups of naive animals (n = 12) via tail vein injection. One hour post injection, subconjunctival injections were carried out to deliver AdV-GFP. This mode of delivery ensures a positive pro-inflammatory response mediated through the mucosa and hence serves as a positive control for a DTH response. Seven days post-subconjunctival injections, footpad thicknesses were measured. Animals were then challenged again by footpad injection and, 24 hours later, the footpads were measured once again to assess the DTH response. To assess GFP expression in these animals, eyes were enucleated upon termination of the experiment and processed for histology as described [11]. ELISA. Sera were obtained via tail vein puncture and stored at -80°C. Samples were analyzed for antibodies to viral capsid proteins and the transgenic protein GFP. Enhanced protein binding ELISA plates (Costar, Corning, NY) were coated overnight at 4°C with antigen using 109 particles/well of rAAV or rAdV-GFP and 100 ng/well GFP in bicarbonate buffer,

MOLECULAR THERAPY Vol. 5, No. 2, February 2002 Copyright © The American Society of Gene Therapy

ARTICLE

pH 9.6. Plates were then washed, blocked, and incubated with diluted (1:100) serum. Saline and serum from an uninjected group of mice were used as negative controls. Human serum containing high levels of anti-AAV antibodies (CF 493 #31, provided by Narin Chirmule) served as a positive control. Samples were then incubated with a 1:1000 dilution of alkaline phosphatase-conjugated sheep anti-mouse IgG (Sigma, St. Louis, MO; 100 l/well) and washed. The isotyping of the antibody response was performed by incubating with biotinylated anti-mouse IgG1, IgG2a, IgG2b, and IgG3 (Pharmingen, San Diego, CA, and Becton Dickinson Co., Franklin Lakes, NJ; 100 l/well) for 2 hours. The wells were washed with phosphate buffered saline containing 0.5% Tween and then incubated with 1:1000 diluted alkaline phosphatase-conjugated avidin (Sigma, 100 ml/well) for 2 hours. The wells were washed again and the color was developed using the Sigma Fast paranitrophenyl phosphate substrate (Sigma). The plates were read at an optical density of 405 nm. Measurements were repeated three times. Identification of neutralizing antibodies. To identify neutralizing antibodies, serum samples (10 l) were plated in 96-well, round-bottom tissue culture plates. Eight serial dilutions of each sample were made in serumfree DMEM. rAAV-GFP virus (2  108 IU) was added to each well. The plates were incubated for 1 hour at 37°C and then the samples were transferred to 96-well, flat-bottom tissue culture plates that had been seeded with 84-31 cells (293 cells which express the E3 and E4 genes). The cells were 60% confluent. Infections were performed overnight. The following day, plates were inspected visually using a fluorescence microscope and analyzed using a Fluoroimager (Molecular Dynamics, Sunnyvale, CA). The neutralizing antibody titer was defined as the highest dilution that allowed GFP to be produced to levels 50% of control samples where no serum was added before virus infections. Antigen-specific cytokine release assay (Elispot). The induction of viral and GFP-specific cytokines following subretinal AV-GFP injection was analyzed at various time points (24 hours, 48 hours, 72 hours, 7 days, 14 days, 21 days) using a T-cell proliferation assay, Elispot [27]. Elispot plates (Cellular Technology Ltd., Cleveland, OH) were coated with primary antibodies directed against the cytokines IL-2, IFN-, TNF-, IL-4, IL-10, and TGF- (PharMingen, San Diego, CA; 100 ml/well) and incubated overnight at 4°C. Primary antibodies were removed and the Elispot plates were blocked for 1 hour with 3% bovine serum albumin in 1 phosphate buffer (3% BSA). Animals were sacrificed and spleen-derived T cells were isolated [26] and plated on the primary antibody-coated Elispot plates at a concentration of 1  105 or 5  105 cells/well (100 l total volume/well, diluted in Dulbecco’s modified Eagle’s medium (DMEM) + 10% fetal bovine serum (FBS)). The plates were then incubated for 48 hours at 37°C under the following conditions: 1) in cell medium alone (unstimulated control group); 2) in the presence of the T-cell stimulants phorbol 12-myristate 13-acetate (PMA) and ionomycin (positive control group; each stimulant diluted 1:100 in DMEM; Sigma); 3) in the presence of adenovirus (AdV-stimulated group; 1  109 IU of AdV-GFP diluted in DMEM); and 4) in the presence of GFP antigens (GFP-stimulated group; 0.1 mg GFP/well, diluted in DMEM). After 48 hours, cells were removed and the plates were washed three times with filtered 1 PBS containing 0.1% Tween and then twice with filtered 1 phosphate buffer. Secondary (biotinylated) antibodies against the cytokines listed above were added to the appropriate wells and the plates were incubated overnight at 4°C. Primary and secondary antibodies for all cytokines were diluted 1:500 in 3% BSA. The following day, secondary antibodies were removed and the Elispot plates were washed again. Extravidin alkaline phosphatase (Sigma; diluted 1:1000 in 3% BSA; 100 l/well) was added to each well and the Elispot plates were again incubated overnight at 4°C. BCIP/NBT phosphatase substrate (Kirkegaard & Perry Laboratories, Gaithersburg, MD) was added (100 l/well), and the plates were left in the dark at room temperature for 5–7 minutes. During this time the plates were monitored by magnifying glass for the appearance of color-reactive spots. The plates were read by Cellular Technologies Ltd. (Cleveland, OH) and analyzed by Immunospot software provided by the same company. Isolation and culture of primary mouse RPE cells. C57BL/6 mice (6–8 weeks old) were used for isolation of RPE cells using modifications of described methods [28,29]. Mice were sacrificed and eyes were enucleated and rinsed three times with PBS containing 50 g/ml of gentamicin

131

ARTICLE

doi:10.1006/mthe.2002.0525, available online at http://www.idealibrary.com on IDEAL

(Sigma), 2.5 g/ml fungizone (Sigma), and 20 mM Hepes (Gibco). We added 2% dispase (Boehringer Mannheim) and incubated the samples at 37°C for 30–45 minutes. A circumferential incision was made just below the ora serrata and the anterior segment and vitreous were discarded. The retina was gently detached from the eyecup and RPE cell layer removed from Bruch’s membrane. The isolated RPE was placed in fresh medium and incubated in 1 ml of 0.1% trypsin in calcium and magnesium-free PBS. RPE tissue was mechanically dissociated into a single-cell suspension and trypsin activity was quenched with addition of excessive serumcontaining medium. RPE cells were pelleted by low speed centrifugation and resuspended in growth medium. Cell viability was assessed using trypan blue exclusion dye. Cells were cultured at 37°C, 95% air + 5% CO2 atmosphere. Evaluation of antigen presenting marker (Class II protein, Ia) in primary mouse RPE cultures and in vivo AdV-GFP injected retinas. Purity of the RPE cell cultures was evaluated by morphological and immunohistochemical examination as described before [29]. The cells were immunostained for cytokeratin staining using a pan-cytokeratin antibody (1:100, DAKO-AE1/AE3; M3515) to ascertain the cell purity. The cultured mouse RPE cells as well as AdV-GFP transduced retina were infected with AdV-GFP (5  109p/ml) for 72 hours, following which the cells were fixed with 4% paraformaldehyde, and stained for Ia (class II protein) induction using a phycoerythrin-conjugated anti-mouse Ia (Pharmingen, USA). Control experiments included cells stimulated with mouse IFN- (100 U/ml). Histological studies. Tissue sections were stained with hematoxylin and eosin and observed under bright field to assess the presence of inflammatory cells. Additional adjacent sections were subjected to immunohistochemical analyses to evaluate the presence of CD4-, CD8-, and CD16-positive cells. These sections were incubated with monoclonal rat anti-mouse antibodies (Boehringer Mannheim, Indianapolis, IN) specific for L3T4 (CD4) at a dilution of 1:50 or for Ly-2 (CD8a) at a dilution of 1:100 for 30 minutes at room temperature. The sections were washed with PBS and incubated for 30 minutes at room temperature with biotin-SP-conjugated goat anti-rat IgG (Jackson Immunoresearch Laboratories, West Grove, PA). The sections were then incubated with Streptavidin-phycoerythrin (Sigma; 1:1000) for 45 minutes and mounted with vectashield (Vector Labs, CA). All studies were performed at 24-hour, 48-hour, 72-hour, 7-day, and 14day time points following subretinal injections.

ACKNOWLEDGMENTS Supported by NIH RO1 grants EY12156 and 10820, Foundation Fighting Blindness, the Lois Pope Life Foundation, the Milton and Ruth Steinbach Foundation, the Paul and Evanina Mackall Foundation Trust, the William and Mary Greve International Research Scholar Award (Research to Prevent Blindness), and the F. M. Kirby Foundation. RECEIVED FOR PUBLICATION MARCH 27; ACCEPTED DECEMBER 6, 2001.

REFERENCES 1. Jiang, L. Q., Jorquera, M., Streilein, J. W., and Ishioka, M. (1995). Unconventional rejection of neural retinal allografts implanted into the immunologically privileged site of the eye. Transplantation 59: 1201–1207. 2. Streilein, J. (1987). Immune regulation and the eye: a dangerous compromise. FASEB J. 1: 199–208. 3. Wenkel, H., and Streilein, J. W. (1998). Analysis of immune deviation elicited by antigens injected into the subretinal space. Invest. Ophthalmol. Vis. Sci. 39: 1823-1834. 4. Kosiewicz, M., Alard, P., and Streilein, J. (1998). Alterations in cytokine production fol-

132

lowing intraocular injection of soluble protein antigen: Impairment in IFN- and induction of TGF- and IL-4 production. J. Immunol. 161: 5382–5390. 5. Hoffman, L. M., Maguire, A. M., and Bennett, J. (1997). Cell-mediated immune response and stability of intraocular transgene expression after adenovirus-mediated delivery. Invest. Ophthalmol. Vis. Sci. 38: 2224–2233. 6. Reichel, M., et al. (1998). Immune responses limit adenovirally mediated gene expression in the adult mouse eye. Gene Ther. 5: 1038-1046. 7. Di Polo, A., Aigner, L. J., Dunn, R. J., Bray, G. M., and Aguayo, A. J. (1998). Prolonged delivery of brain-derived neurotrophic factor by adenovirus-infected Muller cells temporarily rescues injured retinal ganglion cells. Proc. Natl. Acad. Sci. USA 95: 3978–3983. 8. Ali, R. R., et al. (1998). Co-injection of adenovirus expressing CTLA4-Ig prolongs adenovirally mediated lacZ reporter gene expression in the mouse retina. Gene Ther. 5: 1561–1565. 9. D’Orazio, T., and Niederkorn, J. (1998). The nature of antigen in the eye has a profound effect on the cytokine milieu and resultant immune response. Eur. J. Immunol. 28: 1544–1553. 10. Whittum, J., Niederkorn, J., McCulley, J., and Streilein, J. (1984). Role of suppressor T cells in HSV induced immune deviation. J. Virol. 51: 556-558. 11. Anand, V., Chirmule, N., Fersh, M., Maguire, A. M., and Bennett, J. (2000). Additional transduction events after subretinal readministration of recombinant adeno-associated virus. Hum. Gene Ther. 11: 449-457. 12. Li, X., D’Orazio, L., and Niederkorn, J. (1996). Role of Th1 and Th2 cells in anterior chamber-associated immune deviation. Immunol. Today 16: 220–223. 13. Garside, P., and Mowat, A. (1995). Polarization of Th-cell responses: a phylogenetic consequence of nonspecific immune defence? Immunol. Today 16: 220–223. 14. Niederkorn, J. (1990). Immune privilege and immune regulation in the eye. Adv. Immunol. 48: 191–226. 15. D’Orazio, T., and Niederkorn, J. (1998). A novel role for TGF-{b} and IL-10 in the induction of immune privilege. J. Immunol. 160: 2089–2098. 16. Kezuka, T., and Streilein, J. (1999). Analysis of in vivo regulatory properties of T cells activated in vitro by TGFb2-treated antigen presenting cells. Invest. Ophthal. Vis. Sci. 41: 1410–1421. 17. Griffith, T., Herndon, J., Lima, J., Kahn, M., and Ferguson, T. (1995). The immune response and the eye. TCR -chain related molecules regulate the systemic immunity to antigen presented in the eye. Int. Immunol. 7: 1617-1625. 18. Jooss, K., Yang, Y., Fisher, K. J., and Wilson, J. M. (1998). Transduction of dendritic cells by DNA viral vectors directs the immune response to transgene products in muscle fibers. J. Virol. 72: 4212-4223. 19. Bennett, J., Duan, D., Engelhardt, J., and Maguire, A. M. (1997). Real-time non-invasive in vivo assessment of adeno-associated virus-mediated retinal transduction. Invest. Ophthalmol. Vis. Sci. 38: 2857-2863. 20. Dudus, L., et al. (1999). Persistent transgene product in retina, optic nerve and brain after intraocular injection of rAAV. Vision Res. 39: 2545–2554. 21. Gao, G.-P., Yang, Y., and Wilson, J. M. (1996). Biology of adenovirus vectors with E1 and E4 deletions for liver-directed gene therapy. J. Virol. 70: 8934–8943. 22. Bennett, J., et al. (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. 23. Bennett, J., et al. (1996). Photoreceptor cell rescue in retinal degeneration (rd) mice by in vivo gene therapy. Nat. Med. 2: 649–654. 24. Bennett, J., Wilson, J., Sun, D., Forbes, B., and Maguire, A. (1994). Adenovirus vectormediated in vivo gene transfer into adult murine retina. Invest. Ophthalmol. Vis. Sci. 35: 2535–2542. 25. Liang, F.-Q., Anand, V., Maguire, A. M., and Bennett, J. (2000). Intraocular delivery of recombinant virus. In Methods in Molecular Medicine: Ocular Molecular Biology Protocols (P. E. Rakoczy et al., Eds.), pp. 125–139. Humana Press, Totowa, NJ. 26. Plebanski, M. (2000). Preparation of lymphocytes and identification of lymphocyte subpopulations. In Lymphocytes-A Practical Approach (S. Rowland-Jones and A. McMichael, Eds.), pp. 1–26. Oxford University Press, NY. 27. Favre, N., Bordmann, G., and Rudin, W. (1997). Comparison of cytokine measurements using ELISA, ELISPOT and semi-quantitative RT-PCR. J. Immunol. Meth. 204: 57–66. 28. Castillo, B., Little, C., Cerro, C., and Cerro, M. (1995). An improved method of isolating fetal human retinal pigment epithelium. Curr. Eye Res. 14: 677–683. 29. Percopo, C., Hooks, J., Shinohara, T., Caspi, R., and Detrick, B. (1990). Cytokine-mediated activation of a neuronal retinal resident cell provokes antigen presentation. J. Immunol. Methods 145: 4101–4107.

MOLECULAR THERAPY Vol. 5, No. 2, February 2002 Copyright © The American Society of Gene Therapy