Vegf or EphA2 Antisense Polyamide-nucleic acids; Vascular Localization and Suppression of Retinal Neovascularization

original article

© The American Society of Gene Therapy

Vegf or EphA2 Antisense Polyamide-nucleic acids; Vascular Localization and Suppression of Retinal Neovascularization Jikui Shen1,2, Bing Xie1,2, Christina M Hatara1,2, Sean F Hackett1,2 and Peter A Campochiaro1,2 The Department of Ophthalmology, The Johns Hopkins University School of Medicine, Baltimore, Maryland, USA; 2The Department of Neuroscience, The Johns Hopkins University School of Medicine, Baltimore, Maryland, USA 1

Localized gene knockdown is a valuable tool for investigating the function of gene products in tissues. It may also be a good therapeutic strategy for selective targeting of a gene product implicated in disease pathogenesis. While small interfering RNAs (siRNAs) are useful for localized gene knockdown and have achieved well-deserved attention, other strategies may also have applications. Polyamide nucleic acids (PNAs) are DNA–protein chimeric molecules that can be designed with modifications so as to allow good cell entry and high affinity binding to complementary RNA. After intraocular injection of fluorescein isothiocyanate (FITC)-labeled antisense PNAs directed against Vegf or EphA2 (genes that are highly expressed in retinal vessels), labeling was observed to persist in retinal blood vessels even after staining elsewhere in the retina had faded. This did not occur after injection of FITC-labeled antisense human cAMP responsive element binding protein 1 (hCreb) PNA. Subretinal injection of antisense EphA2 PNA was seen to label choroidal blood vessels. Intraocular injection of antisense Vegf PNA or antisense EphA2 PNA significantly reduced their respective target messenger RNAs (mRNAs) in ischemic retinas and suppressed retinal neovascularization (NV). These data suggest that signaling through EphA2 contributes to retinal NV, and that antisense PNAs may be an advantageous way to target EphA2 and other endothelial cell receptors that contribute to ocular NV. Received 13 October 2006; accepted 4 July 2007; published online 7 August 2007; doi:10.1038/sj.mt.6300276

Introduction Targeted gene disruption in mice can provide a powerful tool for investigating the role of gene products in physiologic or patho-­ logic processes. However, elimination of a gene product that is critical for development results in embryonic death and makes it impossible to use this approach to test gene function in adult ­animals. Conditional knockouts may provide a solution, but are difficult and expensive to produce. Antisense oligonucleotides and small interfering RNA (siRNA) are reagents that can be

injected into adults and have the potential for specific knock-­ down of gene products. In addition, their injection into isolated tissue compartments such as the eye can provide local knock-­ down and prevent remote effects involving other tissues. These reagents work well in some situations and provide unequivocal answers, but in other situations they are inefficient and the results are either equivocal or uninformative. Polyamide nucleic acids (PNAs) are DNA–protein chi-­ meric molecules in which the deoxyribose backbone of DNA is replaced with a polyamine backbone that maintains the structure of the nucleotide side chain.1 PNAs with backbones consisting of 2-­aminoethylglycine units attached to nucleo-­ tides through a methylenecarbonyl group were demonstrated to hybridize to complementary DNA (cDNA) or RNA with higher melting temperatures than ­ corresponding DNA–DNA or DNA–RNA duplexes.2 However, the advantage of their supe-­ rior binding affinity was negated by poor water solubility and limited entry into cells. Replacement of amido groups with phosphonate groups resulted in phosphono-PNAs (pPNAs) with a negative charge; while this improved solubility and cell entry, it reduced binding affinity.3 A reasonable solution was provided by construction of antisense PNA–pPNA hybrids that are water-soluble, enter cells efficiently, and have strong bind-­ ing affinity to cDNA or RNA.4 Substitution of PNA-like trans4-hydroxy-l-proline monomers (HypNA) for PNA monomers yields HypNA–pPNA hybrids in which binding affinity is increased to levels seen with pure PNAs without compromis-­ ing solubility and cell entry.5 One would predict that these qualities would make antisense HypNA–pPNA hybrids effi-­ cient tools for achievement of gene knockdowns. In this study, therefore, we sought to test their usefulness for in vivo gene knockdown in the retina. In the course of the study we found that HypNA–pPNA hybrids, that target genes highly and/or dif-­ ferentially expressed in retinal blood vessels, show prolonged localization in retinal vascular cells. They are therefore ideal for testing the ­contribution of gene products to vascular processes and pathologies. They may be particularly ­useful for evaluation of receptors on vascular cells. In order to test this application, we utilized antisense EphA2 PNA to investigate the role of the receptor EphA2 in retinal neovascularization (NV).

Correspondence: Peter A. Campochiaro, Maumenee 719, The Johns Hopkins University School of Medicine, 600 N. Wolfe Street, Baltimore, Maryland 21287-9277, USA. E-mail: [email protected]

1924

www.moleculartherapy.org vol. 15 no. 11, 1924–1930 nov. 2007

© The American Society of Gene Therapy

Results Antisense Vegf PNA stains retinal blood vessels and is detectable for more than 2 weeks Six hours after injection of 2 µg of fluorescein isothiocyanate (FITC)-labeled antisense Vegf PNA, there was a strong fluores-­ cent signal in cells along the surface of the retina (Figure 1a). Twenty-four hours after injection, there was increased fluo-­ rescence on the retinal surface and also in the inner plexiform layer and outer plexiform layers, and some fluorescent structures resembled penetrating blood vessels (Figure 1b). Forty-eight hours after injection (Figure 1c) the background fluorescence had faded somewhat, and there were three distinct regions of strong fluorescence: the retinal surface, the outer border of the inner plexiform layer, and the outer plexiform layer. The superficial, intermediate, and deep capillary beds of the retina are located in these three regions. The horizontal tube-like structures with occasional branches have the typical appearance of blood vessels (arrows). Five days after injection (Figure 1d), there were tubelike structures, similar to those seen at 48 hours, but with less background. At 15 days after injection (Figure 1e) there was still detectable focal staining at the inner surface of the retina with some structures clearly recognizable as blood vessels (arrow) and some diffuse staining at the outer border of the retina. An ocular section from an eye harvested 24 hours after injection of FITClabeled antisense Vegf PNA that was stained for platelet endo-­ thelial cell adhesion molecule (PECAM), a marker of vascular endothelial cells, showed co-localization of the PECAM labeling in the superficial capillary bed with FITC-labeled antisense Vegf PNA (Figure 1f), although extravascular FITC-labeled antisense Vegf PNA was also present. At that time point (24 hours), the PNA had not penetrated very far into the retina, and the deep vessels showed staining only for PECAM (small arrow). As a control for purposes of comparison, FITC-labeled dex-­ tran was injected into the vitreous. Twenty-four hours after injection (Figure 1g), there was strong fluorescence in the pho-­ toreceptor inner segments, retinal pigmented epithelial cells, and the choroid. By 48 hours after injection (Figure 1h), the fluores-­ cence in the retina had faded and there was a stronger signal in the choroid.

Intravitreous injection of antisense Vegf PNA reduces Vegf messenger RNA and VEGF164 protein At postnatal day 12 (P12), the onset of the ischemic period, mice with ischemic retinopathy were given an intraocular injection of 2 µg of antisense Vegf PNA in one eye and phosphate-buffered saline (PBS) in the other eye. At P15 the level of Vegf messenger RNA (mRNA) in the retina was measured by real time reverse transcriptase-polymerase chain reaction (PCR) and the level of VEGF164 protein was measured by enzyme-linked immunosor-­ bent assay. The level of Vegf mRNA was significantly reduced in eyes injected with antisense Vegf PNA compared to those injected with vehicle (Figure 2a). Injection of antisense Vegf PNA also caused a significant reduction of 26.4% in VEGF164 compared to the eyes injected with PBS (Figure 2b). When compared against the baseline level of vascular endothelial growth factor (VEGF) in non-ischemic P15 retina, this represents a 40% reduction in ischemia-induced increase in VEGF in the retina. Molecular Therapy vol. 15 no. 11 nov. 2007

EphA2 Knockdown Suppresses Retinal NV

a

b INL ONL

c

d INL ONL

e

f INL ONL

g

h INL ONL 100 µm

Figure 1 Distribution of fluorescein isothiocyanate (FITC)-labeled antisense Vegf polyamide nucleic acid (PNA) in the retina after intraocular injection. (a) Six hours after intraocular injection of 2 µg of FITC-labeled antisense Vegf PNA, fluorescence was limited to the surface of the retina (counterstained with Hoechst). (b) Twenty-four hours after injection, there was strong fluorescence at the retinal surface, extending into the inner plexiform layer (IPL) and outer plexiform layer (OPL). A tube-like structure in the IPL has the appearance of a penetrating retinal vessel (arrow). (c) Forty-eight hours after injection, background fluorescence had faded and numerous tube-like branching structures were seen (arrows) in the location of the intermediate and deep capillary beds of the retina. (d) Five days after injection, fluorescence at the retinal surface was still strong and faint outlines of tube-like structures could be seen in the IPL and OPL. (e) Fifteen days after injection, there was still some fluorescence at the inner and outer surfaces of the retina with an occasional vessel still seen (arrow). (f) Twenty four hours after injection of 2 µg of FITC-labeled antisense Vegf PNA, there was co-localization, and ­ platelet-endothelial cell adhesion molecule-1, a specific marker for vascular endothelial cells, confirmed that the vascularshaped structures containing PNA were retinal vessels. (g) Twenty-four hours after injection of 2 µg of FITC-labeled dextran, there was strong fluorescence in photoreceptor inner segments and retinal pigmented epithelial cells, with mild staining in the choroid (CH). There was no significant staining in the inner nuclear layer, outer nuclear layer, or the photoreceptor outer segments. (h) At 48 hours after injection, most of the staining in the retina had cleared, but there was still some staining in the CH.

Intravitreous injection of antisense Vegf PNA suppresses retinal NV In P17 mice with ischemic retinopathy, those that had received an intraocular injection of 1 µg of antisense Vegf PNA at P12, appeared to have substantially less NV on the ­surface of the ret-­ ina (Figure 3a and c) than those that had received an ­injection 1925

© The American Society of Gene Therapy

EphA2 Knockdown Suppresses Retinal NV

b

6,000 5,000

PECAM 250

P = 0.00845

VEGF pg/10 µg of retinal protein

Vegf transcripts/100,000 Cyclophilin A transcripts

a

4,000 3,000 2,000 1,000

200

a

P = 0.0346

c INL

150

ONL

Intraocular EphA2-PNA

100 50

d

0

0 Antisense VEGF PNA

Merge

b

Control

Antisense Control VEGF PNA

e

f

Room air INL

Figure 2 Significant reduction of vascular endothelial growth factor (VEGF) in the retinas of mice with ischemic retinopathy after intravitreous injection of antisense Vegf polyamide nucleic acid (PNA). (a) Bars represent the mean (± SEM) Vegf messenger RNA (mRNA) copy number per 105 copies of Cyclophilin A mRNA, as determined by real time reverse transcriptase-polymerase chain reaction. There was a significant reduction in Vegf mRNA in the eyes of postnatal day 15 mice with ischemic retinopathy injected with antisense Vegf PNA, when compared with eyes injected with vehicle (n = 16 for both). (b) Enzyme-linked immunosorbent assay showed that, compared to fellow eyes injected with vehicle (n = 9), eyes injected with 2 µg of antisense Vegf PNA (n = 9) had a 26.4% reduction in VEGF164 protein. Taking into consideration the baseline amount of VEGF present in control retinas placed in room air, this represents a 40% reduction in the excess VEGF164 in ischemic retinas. Statistical comparisons were made by unpaired t-test.

ONL

Intraocular hCREB-PNA

g

h

i INL ONL

Periocular EphA2-PNA

j

k

l INL ONL

Periocular hCREB

m

n

o INL ONL

Subretinal EphA2-PNA

a

b

p

q

r INL

Subretinal hCREB-PNA

500 µm

c

d

100 µm

e Total area of retinal NV

0.16 0.12

P = 0.02935

0.08 0.04 0 PBS control

Antisense VEGF-PN

Figure 3  Antisense Vegf polyamide nucleic acid (PNA) reduces retinal neovascularization (NV) in mice with oxygen-induced ischemic retinopathy. Eyes of postnatal day mice with ischemic retinopathy, injected with antisense Vegf PNA (a and c), showed little NV on the surface of the retina (stained with Griffonia simplicifolia lectin), while those injected with vehicle (b and d) showed large areas of NV. (e) Measurement of the area of NV on the surface of the retina by image analysis, with the observer masked with regard to treatment group, showed a significant reduction in mean (± SEM) area of NV on the surface of the retina in eyes injected with antisense Vegf PNA as compared to corresponding phosphate-buffered saline (PBS) controls. Statistical comparison was made by unpaired t-test.

1926

ONL 56.677 µm

Figure 4 Distribution of antisense EphA2 polyamide nucleic acid (PNA) after intraocular, periocular or subretinal injection. Seventytwo hours after intraocular injection of 2 µg, periocular injection of 50 µg, or subretinal injection of 2 µg of fluorescein isothiocyanate (FITC)-labeled antisense EphA2 PNA, ocular sections were stained with platelet endothelial cell adhesion molecule-1 (PECAM-1) coupled to phycoerythrin. Visualization of fluorescence using the red channel showed retinal and choroidal vessels (first column), and visualization with the green channel showed localization of the antisense EphA2 PNA (second column). The merged images are shown in the third column. (a–c) After intravitreous injection, antisense EphA2 PNA was located predominantly within cells of retinal vessels. (d–f) In contrast, 72 hours after intravitreous injection of 2 µg of FITC-labeled antisense human cAMP responsive element binding protein 1 (hCreb) PNA, there was mild diffuse fluorescence throughout the retina. (g–i) After periocular injection of 50 µg of FITC-labeled antisense EphA2 PNA, there was strong fluorescence in the retinal pigmented epithelium and choroid with no staining in the retina. (j–l) After periocular injection of 50 µg of FITC-labeled antisense hCreb PNA, there was weak florescence in photoreceptor outer segments. (m–o) After subretinal injection of 2 µg of FITC-labeled antisense EphA2 PNA, there was strong fluorescence in the choroid that co-­localized with PECAM-1, thereby indicating concentration of FITC-labeled antisense EphA2 PNA in choroidal vessels; however, (p–r) after injection of 2 µg of FITC-labeled antisense hCreb PNA, there was only faint diffuse fluorescence.

of vehicle (Figure 3b and d). Measurement of the area of NV by image analysis, with the observer masked with respect to treatment group, confirmed that intraocular injection of 1 µg of antisense Vegf PNA caused significant suppression of retinal NV (Figure 3e). www.moleculartherapy.org vol. 15 no. 11 nov. 2007

© The American Society of Gene Therapy

EphA2 Knockdown Suppresses Retinal NV

Distribution of antisense EphA2 PNA after intraocular, periocular, or subretinal injection Seventy-two hours after intravitreous injection of 2 µg of FITClabeled antisense EphA2 PNA, there was strong fluorescence that co-localized with staining for PECAM-1, thereby indicat-­ ing that the antisense EphA2 PNA was specifically localized in retinal blood vessels (Figure 4a–c). In contrast, 72 hours after intravitreous injection of 2 µg of FITC-labeled antisense human cAMP responsive element binding protein 1 (hCreb) PNA, there was mild diffuse fluorescence throughout the retina (Figure 4d–f). Seventy-two hours after periocular injection of 10 µl containing 50 µg of FITC-labeled antisense EphA2 PNA, there was strong fluorescence in the retinal pigmented epithe-­ lium and choroid, with no staining in the retina (Figure 4g–i). Periocular injection of 10 µl containing 50 µg of FITC-labeled antisense hCreb PNA showed only weak fluorescence signals in the photoreceptor outer segments (Figure 4j–l). Seventy-two hours after subretinal injection of 5 µg of FITC-labeled antisense EphA2 PNA, there was strong fluorescence in the choroid that co-localized with PECAM-1, thereby indicating concentration of FITC-labeled antisense EphA2 PNA in choroidal blood vessels (Figure 4m–o); however, after injection of 5 µg of FITC-labeled antisense hCreb PNA, there was only faint, diffuse fluorescence (Figure 4p–r).

a

Antisense EphA2 PNA knocks down EphA2 mRNA and inhibits retinal NV The preferential, prolonged localization of antisense EphA2 PNA in retinal vascular cells by antisense PNA suggests that it may be a useful tool to explore Eph receptor participation in reti-­ nal vascular processes. The Ephrin/Eph signaling pathway has been demonstrated to play a role in some types of NV. A recent report from a study that utilized a soluble Eph receptor has sug-­ gested its possible involvement in retinal NV, but it is unclear which Eph receptor is involved. We found that EphA2 mRNA was present in retina and was significantly increased in ischemic retina (data not shown). Intraocular injection of 2 µg of anti-­ sense EphA2 PNA at P12, the onset of ischemia, caused signifi-­ cant reduction of EphA2 mRNA in the retina at P15 (Figure 5a) when compared with injection of 2 µg of antisense hCreb PNA. Immunoblots of retinal homogenates from eyes of mice injected with antisense EphA2 PNA showed a substantial reduction in EphA2 protein as compared to retinal homogenates from eyes

a

b

c

d

P = 0.0244

1,600 1,400 1,200 1,000 800 600 400 0

b

400 µm

200 Control (hCREB-PNA)

hCREB

EphA2

EphA2-PNA hCREB

EphA2

EphA2

e

0.12 Total area of retinal NV

EphA2 transcripts/100,000 cyclophilin A transcripts

1,800

0.1 P = 0.02025

0.08 0.06 0.04 0.02 0 hCREB-PNA

Actin

Figure 5  Intravitreous injection of antisense EphA2 polyamide nucleic acid (PNA) reduces messenger RNA (mRNA) and protein for EphA2 in ischemic retina. At postnatal day 15, real time polymerase chain reaction was carried out on retinal RNA from mice with ischemic retinopathy that had received an intraocular injection of 2 µg of antisense EphA2 PNA in one eye and 2 µg of antisense hCreb PNA in the other eye. (a) Bars represent the mean (± SEM) copy number of EphA2 mRNA per 105 copies of Cyclophilin A mRNA. When compared against retinas from eyes injected with 2 µg of antisense hCreb PNA (n = 9), retinas from eyes injected with antisense EphA2 PNA (n = 9) showed a significant reduction in EphA2 mRNA (P = 0.0224, paired t-test). (b) Western blots of retinal homogenates from two eyes injected with antisense EphA2 PNA showed a substantially lower level of EphA2 protein than retinal homogenates from two eyes with antisense hCreb PNA.

Molecular Therapy vol. 15 no. 11 nov. 2007

EphA2-PNA

Figure 6  Antisense EphA2 polyamide nucleic acid (PNA) reduces retinal neovascularization (NV) in mice having oxygen-induced ischemic retinopathy. At postnatal day 17, mice with ischemic retinopathy (n = 6) that had received an intraocular injection of 2 µg of antisense EphA2 PNA in one eye and antisense human cAMP responsive element binding protein 1 (hCreb) PNA in the other eye were given an intraocular injection of rat-anti mouse platelet endothelial cell adhesion molecule-1 antibody. After 12 hours, the mice were killed and the retinas were dissected intact and incubated in goat-anti rat IgG conjugated with Alexa488, and whole mounted. (a and b) Two representative retinas from eyes injected with antisense hCreb PNA show extensive NV (arrows) on the surface of the retina with lightly stained normal retinal vessels providing a diffuse green background. (c and d) In contrast, two representative retinas from eyes injected with antisense EphA2 PNA show much less NV (arrows). (e) Measurement of the area of NV on the surface of the retina by image analysis, with the investigator masked with respect to treatment group, showed a significant reduction in NV in eyes injected with antisense EphA2 PNA as compared to eyes injected with antisense hCreb PNA. Statistical comparison was made by paired t-test.

1927

EphA2 Knockdown Suppresses Retinal NV

injected with antisense hCreb PNA (Figure 5b). When compared with injections of antisense hCreb PNA, injections of antisense EphA2 PNA did not cause a significant reduction in Sonic hedgehog or EphA1 mRNA. At P17, 5 days after the onset of ischemia, retinal flat mounts from eyes injected with antisense hCreb PNA showed extensive NV on the surface of the retina (Figure 6a and b), whereas those from eyes injected with antisense EphA2 PNA showed only a few small areas of NV. (Figure 6c and d). Image analysis confirmed that antisense EphA2 PNA caused a statistically significant reduc-­ tion in ischemia-induced retinal NV (Figure 6e).

Discussion Gene knockdown is a valuable technique for exploring the func-­ tion of gene products. It is not surprising that siRNA has gener-­ ated a great deal of excitement, because it provides a reliable gene knockdown approach, provided care is taken to avoid any off-­target effects. We recently utilized siRNA technology to demonstrate that (i) VEGF receptor 1 plays an important pro-angiogenic role in several types of ocular NV;6 (ii) that vasohibin functions in a negative feedback loop to inhibit VEGF-induced retinal NV;7 and (iii) that p66Shc promotes oxidative stress in the retina.8 Labeled siRNA injected into the vitreous cavity was visualized in the retina for about 1 week, and appeared to enter most retinal cells with no particular pattern.6 In this study, we performed intraocular injections with labeled antisense PNAs and found that antisense Vegf PNA and antisense EphA2 PNA showed selective, prolonged localization in retinal cells, while antisense hCreb PNA did not. Antisense Vegf PNA was detectable in retinal vascular cells for at least 15 days (the longest time point examined) after a single injection. This suggests that PNAs may be particularly useful for investigating retinal vascular disease. VEGF plays a central role in retinal vascular diseases, and knockdown of VEGF with siRNA was seen to suppress ocular NV.9 Intraocular injection of 1 µg of antisense Vegf PNA reduced VEGF in ischemic retina and suppressed retinal NV. Confirmation of the well-established role of VEGF provides validation for using anti-­ sense PNA to investigate the role of gene products in the retina. The preferential localization in vascular endothelial cells is particularly advantageous for exploring the role of genes that are differentially expressed in vascular endothelial cells, such as receptors. The Eprhrin/Eph signaling system plays a role in axon guidance, in developmental angiogenesis, and in angiogenesis in some adult tissues.10–15 Intraocular injection of a soluble Eph receptor resulted in suppression of ischemia-induced retinal NV, thereby suggesting that the Ephrin/Eph signaling system may be involved. However, there was no indication as to which Eph recep-­ tor was involved.16 We theorized that EphA2 was a good candi-­ date and this was supported by our finding that EphA2 mRNA is significantly increased in ischemic retina. The ischemia-induced increase in EphA2 mRNA and retinal NV were significantly blocked by intraocular injection of antisense EphA2 PNA. This indicates that EphA2 plays an important role in retinal NV. These data help to further elucidate the molecular pathogen-­ esis of ischemia-induced retinal NV. The VEGF signaling sys-­ tem, which is very complex, plays a central role.17–19 Four VEGF receptors, VEGF receptor 1, VEGF receptor 2, neuropilin 1, and 1928

© The American Society of Gene Therapy

­neuropilin 2 collaborate, and each seems to play an independently important role, because perturbation of any of them substantially reduces retinal NV.6,20–22 This study identifies EphA2 as a recep-­ tor target outside the VEGF signaling system, that contributes to retinal NV and thereby provides a parallel track for therapeutic intervention. The Tie 2 receptor is also involved, but its blockage is not beneficial in all situations; blockage of Tie 2 stimulates retinal NV when VEGF levels are high and promotes its regression when VEGF levels are low, or when signaling through VEGF recep-­ tors is blocked.23 Stimulation of Tie 2 by over-expression of Ang1 inhibits retinal NV.24,25 It is therefore clear that multiple receptors on retinal vascular endothelial cells participate in the growth and regression of retinal NV, and there may be others that have roles yet to be defined. The preferential localization of intraocularly injected antisense PNAs directed against genes preferentially and/or highly expressed in retinal vascular cells indicates that PNAs provide a valuable tool for carrying out those investigations. In addition, the favorable pharmacokinetics after intraocular injection suggest that PNAs may serve not only as valuable research tools, but may also have therapeutic potential.

Materials and Methods Mice. Pathogen-free female C57BL/6 mice were purchased from Charles

Rivers Laboratories (Wilmington, MA) and were treated in accordance with the requirements of the Committee on Animal Welfare of the Johns Hopkins University School of Medicine and the guidelines of the Association for Research in Vision and Ophthalmology for the use of ani-­ mals in Research. Modified antisense PNAs targeting VEGF or EphA2. gripNAs are com-­

mercially available modified antisense PNAs consisting of a backbone of alternating HypNA and pPNA monomers attached to bases by methylene carbonyl linkages (Active Motif, Carlsbad, CA). A gripNA targeting the 5′-end of Vegf mRNA adjacent to the transcription start site with sequence AGA GAG CAG AAA GTT CAT (antisense Vegf PNA) was generated and labeled with FITC. A gripNA targeting the 5′-end of EphA2 mRNA adja-­ cent to the transcription start site with sequence GAC TGC CCG GAG CTC CAT (antisense EphA2 PNA) was also generated. National Center for Biotechnology Information blast search showed no matches for the corre-­ sponding sense sequences other than the target mRNAs. Antisense hCreb PNA was used as a control; National Center for Biotechnology Information blast search (http://www.ncbi.nlm.nih.gov/BLAST/) showed no overlaps between any mouse mRNA and its target sequence. Lyophilized antisense PNAs were reconstituted in PBS and stored at –80 °C. Determination of distribution of FITC-labeled antisense Vegf PNA in the retina after intravitreous injections. Mice were anesthetized and given

intravitreous injections of 1 µl of PBS containing 2 µg of FITC-labeled antisense Vegf PNA, FITC-labeled EphA2 PNA, FITC-labeled antisense hCreb PNA or 2 µg of FITC-labeled dextran. The mice were killed at vari-­ ous time points after injection and the eyes were frozen in optimal cutting temperature embedding compound (Miles Diagnostics, Elkhart, IN). In some experiments, mice were given subretinal injections of 2 µg or peri-­ ocular injections of 50 µg of FITC-labeled EphA2 PNA. Frozen sections (10 µm) were fixed in 4% paraformaldehyde, washed with PBS, counter-­ stained with Hoechst (1:1,000; Sigma, St. Louis, MO), and viewed with a Nikon fluorescence microscope. For PECAM staining, sections were blocked with 10% normal goat serum or normal mouse serum for 30 min-­ utes and incubated overnight at 4 °C with rat anti-mouse PECAM (1:100; Pharmingen, San Diego, CA). After three washes with PBS/Tris, sections www.moleculartherapy.org vol. 15 no. 11 nov. 2007

© The American Society of Gene Therapy

were incubated for 1 hour at room temperature with 1:800 Cy3-labeled goat anti-rat IgG (Jackson ImmunoResearch Lab, West Grove, PA). Mice with oxygen-induced ischemic retinopathy. Ischemic retinopathy

was induced by placing P7 mice in 75 ± 3% oxygen for 5 days and then returning them to room air.26 Treatments were administered at P12, con-­ sisting of intravitreous injection of antisense Vegf PNA or antisense EphA2 PNA in one eye and antisense hCreb PNA or vehicle in the other eye. Some of the mice were killed at P15 for measurement of Vegf or EphA2 mRNA levels or VEGF164 protein level. Others were killed at P17 to measure the amount of retinal NV. Measurement of the amount of retinal NV. Ten micrometer serial sections

were cut through the entire eye starting with sections that included the iris root on one side of the eye and proceeding to the iris root on the other side. Every tenth section, roughly 100 µm apart, was stained with biotinylated Griffonia simplicifolia lectin B4 (GSA, Vector Laboratories, Burlingame, CA), which selectively binds to vascular cells. Slides were incubated in methanol/H2O2 for 10 minutes at 4 °C, washed with 0.05 mol/l Tris­buffered saline (TBS), pH 7.6, and incubated for 30 ­minutes in 10% nor-­ mal porcine serum. Slides were incubated for 2 hours at room temperature with biotinylated GSA and, after being rinsed with 0.05 mol/l TBS, they were incubated with avidin coupled to peroxidase (Vector Laboratories, Burlingame, CA) for 45 minutes at room temperature. After being washed for 10 minutes with 0.05 mol/l TBS, the slides were incubated with diami-­ nobenzidine (Invitrogen, Carlsbad, CA) to give a brown reaction product, and mounted with Cytoseal (Stephens Scientific, Riverdale, NJ). The slides were examined with an Axioskop microscope (Zeiss, Thornwood, NY) and images were digitized using a three charge-coupled device color video camera and a frame grabber. Image-Pro Plus software was used for delin-­ eating GSA-stained cells on the surface of the retina and the area occupied by them was measured. The mean of the measurements from each eye was used as a single experimental value. For quantification of retinal NV after injection of antisense EphA2 PNA in one eye and vehicle in the other eye, P17 mice were given intraocular injections of rat anti-mouse PECAM. The mice were killed 12 hours after the injection, the eyes were fixed in 10% formalin for 4 hours, and the retinas were dissected. After washing twice with PBS, the retinas were stained with 1:500 goat anti-rat IgG conjugated with Alexa488 (Invitrogen, Carlsbad, CA) for 40 minutes, washed, and whole-mounted. The slides were examined with a Nikon microscope, images were acquired by Spot software, and the area of NV was measured using ImagePro Plus. Real-time reverse transcriptase-PCR. Retinas were dissected and total

RNA was isolated using RNeasy kits (Qiagen, Chatsworth, CA). After quantification of RNA concentration using Gene SpecIII (Hitachi, Tokyo, Japan), 1 µg was treated with DNase I (Ambion, Austin, TX) to remove any contaminating genomic DNA, and cDNA was synthesized with reverse transcriptase (SuperScript III; Invitrogen, Carlsbad, CA) and 5.0 µmol/l random hexamer. Samples of cDNA were aliquoted and stored at –80 °C. Real-time PCR using the SYBR Green I format on a Chromo4 thermal cycler system (BioRad Laboratories, Hercules, CA) was used for quantify-­ ing levels of mRNA for VEGF, and EphA2. Reactions were performed in a 20 µl volume using the SYBR Green Master mix (Qiagen) with 0.5 mmol/l primers. Cyclophilin A mRNA was measured for normalization. The sequences of the PCR primer pairs used were: Vegf, forward-GCC AGC ACA TAG AGA GAA TGA GC and reverse-CAA GGC TCA CAG TGA TTT TCT GG; EphA2, forward-CAC CAA GAT TGA CAC CAT CG and reverse-TGT ATC GGA GAC AGC GAC AG; and Cyclophilin A, forwardCAG ACG CCA CTG TCG CTT T and reverse-TGT CTT TGG AAC TTT GTC TGC AA. The mRNA fragments were amplified with Pfu Taq polymerase (Stratagene, La Jolla, CA) and purified with a gel extraction kit (Qiagen). mRNA copy numbers were calculated according to the Roche Molecular Therapy vol. 15 no. 11 nov. 2007

EphA2 Knockdown Suppresses Retinal NV

quantification technique manual. Standard curves for each gene were plotted with quantified cDNA template during each real-time PCR. Target gene mRNA copy number was normalized to 105 copies of Cyclophilin A. Enzyme-linked immunosorbent assays for mouse VEGF. Retinas were

removed and sonicated in 200 µl of lysis buffer containing 50 µl of 1 mol/l Tris–HCl (pH 7.4), 50 µl of 10% sodium dodecyl sulfate, 5 µl of 100 nmol/l phenylmethaneculfonyl, and 5 µl of sterile de-ionized water. Retinal homogenates were centrifuged at 13,000g for 5 minutes at 4 °C and the pro-­ tein concentration in supernatants was measured by micro-bicinchoninic acid assay (Pierce, Rockford, IL). Enzyme-linked immunosorbent assay was performed using the murine Quantikine VEGF assay kit in accordance with the manufacturer’s instructions (R & D Systems, Minneapolis, MN). Serial dilutions of recombinant VEGF were assayed to generate a standard curve with a 10/pg/ml limit of detection. Immunoblots for EphA2. Retinas were dissected from adult C57BL/6 mice 72 hours after intravitreous injection of 2 µg of antisense EphA2 PNA in one eye and 2 µg of antisense hCreb PNA in the other eye. The retinas were immediately placed in lysis buffer containing 50 mmol/l Tris–HCl (pH 7.4), 150 mmol/l NaCl, 1% deoxycholic acid, 0.1% sodium dodecyl sulfate, 0.5% NP-40, and complete protease inhibitor cocktail (Roche, Mannheim, Germany). The retinas were sonicated for 4 seconds at 4 °C. Protein con-­ centrations were determined by micro-bicinchoninic acid assay (Pierce, Rockford, IL) in accordance with the manufacturer’s instructions. Twenty micrograms of protein were loaded in wells of a 10% sodium dodecyl sul-­ fate gel and, after electrophoresis, separated proteins were transferred to a nitrocellulose membrane. The membrane was incubated in 0.05 mol/l TBS, pH 7.6, containing 5% skim milk to block nonspecific binding sites, and then incubated in TBS containing 0.05% Tween-20, 2.5% skim milk, and 0.5 µg/ml rabbit anti-mouse EphA2 polyclonal antibody (Novus Biologicals, Littleton, CO). After three washes with TBS–Tween-20, the membrane was incubated in horseradish peroxidase-conjugated goat anti-rabbit polyclonal antibody (1:5,000; Amersham-Pharmacia Biotech, Piscataway, NJ). For signal development, membranes were incubated in Enhanced Chemoluminescence-Plus Western blotting detection reagent (Amersham-Pharmacia Biotech, Piscataway, NJ) and exposed to X-ray film (Kodak, Rochester, NY). Statistical analyses. Data are presented as mean ± SEM. For within-­animal

comparisons between the antisense PNA-treated eye and the ­vehicle-treated eye, data were analyzed using a two-way analysis of variance. Paired t-tests were used for comparisons between vehicle-treated eyes.

ACKNOWLEDGMENTs This study was supported by EY12609 from the National Eye Institute and a senior scientist award from Research to Prevent Blindness, New York, NY. P.A.C. is the George S. and Dolores Dore Eccles Professor of Ophthalmology and Neuroscience.

References

1. Nielsen, PE, Egholm, M, Berg, RH and Buchardt, O (1991). Sequence-selective recognition of DNA by strand displacement with a thymine-substituted polyamide. Science 254: 1497–1500. 2. Egholm, M, Buchardt, O, Christensen, L, Behrens, C, Freler, SM, Driver, DA et al. (1993). PNA hybridizes to complementary oligonucleotides obeying the Watson–Crick hydrogen-bonding rules. Nature 365: 566–568. 3. van der Laan, AC, Stromberg, R, van Boom, JH and Kuyi-Yeheskiely, E (1996). An approach towards the synthesis of oligomers containing a N-2-hydroxyethyl-aminomethylphosphonate backbone: a novel PNA analogue. Tetrahedron Lett 43: 7857–7860. 4. Efimov, VA, Choob, MV, Buryakova, AA, Kalinkina, AL and Chakhmakhcheva, OG (1998). Synthesis and evaluation of some properties of chimeric oligomers containing PNA and phosphono-PNA residues. Nucleic Acids Res 26: 566–575. 5. Efimov, V, Choob, M, Buryakova, A, Phelan, D and Chakhmakhcheva, OG (2001). PNA-related oligonucleotide mimics and their evaluation for nucleic acid hybridization studies and analysis. Nucleosides Nucleotides Nucleic Acids 20: 419–428.

1929

EphA2 Knockdown Suppresses Retinal NV

6. Shen, J, Samul, R, Silva, RL, Akiyama, H, Liu, H, Saishin, Y et al. (2005). Suppression of ocular neovascularization with siRNA targeting VEGF receptor 1. Gene Ther 13: 225–234. 7. Shen, J, Yang, X, Xiao, WH, Hackett, SF, Sato, Y and Campochiaro, PA (2006). Vasohibin is up-regulated by VEGF in the retina and suppresses VEGF receptor 2 and retinal neovascularization. FASEB J 20: 723–725. 8. Wu, Z, Rogers, B, Kachi, S, Hackett, SF, Sick, A and Campochiaro, PA (2006). Reduction of p66Shc suppresses oxidative damage in retinal pigmented epithelial cells and retina. J Cell Physiol 209: 996–1005. 9. Reich, SJ, Fosnot, J, Kuroki, A, Tang, W, Yang, X, Maguire, AM et al. (2003). Small interfering RNA (siRNA) targeting VEGF effectively inhibits ocular neovascularization in a mouse model. Mol Vis 9: 210–216. 10. Drescher, U, Kremoser, C, Handwerker, C, Löschinger, J, Noda, M and Bonhoeffer, F (1995). In vitro guidance of retinal ganglion cell axons by RAGS, a 25 kDa tectal protein related to ligands for Eph receptor tyrosine kinases. Cell 82: 359–370. 11. Drescher, U, Bonhoeffer, F and Müller, BK (1997). The Eph family in retinal axon guidance. Curr Opin Neurobiol 7: 75–80. 12. McBride, JL and Ruiz, JC (1998). Ephrin-A1 is expressed at sites of vascular development in the mouse. Mech Dev 77: 201–204. 13. Adams, RH, Wilkinson, GA, Weiss, C, Diella, F, Gale, NW, Deutsch, U et al. (1999). Roles of ephrinB ligands and EphB receptors in cardiovascular development: demarcation of arterial/venous domains, vascular morphogenesis, and spouting angiogenesis. Genes Dev 13: 295–306. 14. Pandey, A, Shao, H, Marks, RM, Polverini, PJ and Dixit, VM (1995). Role of B61, the ligand for the Eck receptor tyrosine kinase, in TNF-alpha-induced angiogenesis. Science 268: 567–569. 15. Ogawa, K, Pasqualini, R, Lindberg, RA, Kain, R, Freeman, AL and Pasquale, EB (2000). The ephrin-A1 ligand and its receptor, EphA2, are expressed during tumor neovascularization. Oncogene 19: 6043–6052. 16. Chen, J, Hicks, D, Brantley-Sieders, D, Cheng, N, McCollum, GW, Qi-Werdich, X et al. (2006). Inhibition of retinal neovascularization by soluble EphA2 receptor. Exp Eye Res 82: 664–673.

1930

© The American Society of Gene Therapy

17. Ozaki, H, Hayashi, H, Vinores, SA, Moromizato, Y, Campochiaro, PA and Oshima, K (1997). Intravitreal sustained release of VEGF causes retinal neovascularization in rabbits and breakdown of the blood-retinal barrier in rabbits and primates. Exp Eye Res 64: 505–517. 18. Okamoto, N, Tobe, T, Hackett, SF, Ozaki, H, Vinores, MA, LaRochelle, W et al. (1997). Transgenic mice with increased expression of vascular endothelial growth factor in the retina: a new model of intraretinal and subretinal neovascularization. Am J Pathol 151: 281–291. 19. Ozaki, H, Seo, MS, Ozaki, K, Yamada, H, Yamada, E, Okamoto, N et al. (2000). Blockade of vascular endothelial cell growth factor receptor signaling is sufficient to completely prevent retinal neovascularization. Am J Pathol 156: 679–707. 20. Oh, H, Takagi, H, Otani, A, Koyama, S, Kemmonchi, S, Uemura, A et al. (2002). Selective induction of neuropilin-1 by vascular endothelial growth factor (VEGF): a mechanism contributing to VEGF-induced angiogenesis. Proc Natl Acad Sci USA 99: 383–388. 21. McLeod, DS, Taomoto, M, Cao, J, Zhu, Z, Witte, L and Lutty, GA (2002). Localization of VEGF receptor-2 (KDR/Flk-1) and effects of blocking it in oxygen-induced retinopathy. Invest Ophthalmol Vis Sci 43: 474–482. 22. Shen, J, Samul, R, Zimmer, J, Liu, H, Liang, X, Hackett, SF et al. (2004). Deficiency of neuropilin 2 suppresses VEGF-induced retinal neovascularization. Mol Med 10: 12–18. 23. Oshima, Y, Oshima, S, Nambu, H, Kachi, S, Takahashi, K, Umeda, N et al. (2005). Different effects of angiopoietin 2 in different vascular beds in the eye; new vessels are most sensitive. FASEB J 19: 963–965. 24. Nambu, H, Nambu, R, Oshima, Y, Hackett, SF, Okoye, G, Wiegand, S, et al. (2004). Angiopoietin 1 inhibits ocular neovascularization and breakdown of the blood-retinal barrier. Gene Ther 11: 865–873. 25. Nambu, H, Umeda, N, Kachi, S, Oshima, Y, Akiyama, H, Nambu, R et al. (2005). Angiopoietin 1 prevents retinal detachment in an aggressive model of proliferative retinopathy, but has no effect on established neovascularization. J Cell Physiol 204: 227–235. 26. Smith, LE, Wesolowski, E, McLellan, A, Kostyk, SK, D’Amato, R, Sullivan, R et al. (1994). Oxygen-induced retinopathy in the mouse. Invest Ophthalmol Vis Sci 35: 101–111.

www.moleculartherapy.org vol. 15 no. 11 nov. 2007