Nanoparticle-mediated gene transfer specific to retinal pigment epithelial cells

Abstract

Previously, we demonstrated that CK30PEG10k-compacted DNA nanoparticles (NPs) efficiently target photoreceptor cells and improve visual function in a retinitis pigmentosa model. Here, we test the ability of these NPs in driving transgene expression in the retinal pigment epithelium (RPE), using an RPE-specific reporter vector (VMD2-eGFP). NPs, uncompacted plasmid, or saline were subretinally delivered to adult BALB/c mice. NP-based expression was specific to RPE cells and caused no deleterious effects on retinal structure and function. eGFP expression levels in NP-injected eyes peaked at post-injection day 2 (PI-2), stabilized at levels ~3-fold higher than in naked DNA-injected eyes, and remained elevated at the latest time-point examined (PI-30). Unlike naked DNA, which only transfected cells at the site of injection, NPs were able to transfect cells throughout the RPE. Subretinal injections of rhodamine labeled NPs and naked DNA showed comparable initial uptake into RPE cells. However, at PI-7 and -30 days significantly more fluorescence was detected inside the RPE of NP-injected eyes compared to naked DNA, suggesting NPs are stable inside the cell which could possibly lead to higher and sustained expression. Overall, our results demonstrate that NPs can efficiently deliver genes to the RPE and hold great potential for the treatment of RPE-associated diseases.

Fig. 1. Vector construction and Nanoparticle EM

(a) Vector map depicting the organization of the expression cassette. VMD2 promoter was cloned into EcoRI and BamHI sites in the pEGFP-1 vector purchased from Invitrogen. (b) EM image of NP formulation at 40,000x. The rod shaped particles were determined to have diameter of ~8–11nm. Scale bar =200nm.

Fig. 2. Relative eGFP expression levels following subretinal injection

Eyes were injected at P30, enucleated at 2, 7 and 30 days PI, processed for total RNA preparation followed by cDNA synthesis and qRT-PCR. Data represent eGFP expression levels relative to HPRT. NP-driven expression of eGFP was significantly higher than naked DNA at 2, 7 and 30 days PI (P<0.01, P<0.05, and P<0.05, respectively). Data are presented as means ± SEM (n=4 for saline, n=7 for others).

Fig. 3. RPE specific eGFP expression

Mice were subretinally injected at P30 with 1μl naked DNA or NP at a concentration of 4.3mg/ml, or saline, and eyes were collected at PI-2 & 30 days. Representative confocal images of native expression of eGFP in the RPE layer are shown. Nuclei were counterstained with DAPI. Images in panel a and b were taken at 2x and 40x magnification, respectively. Images in panel b are taken from regions near the site of injection (indicated by arrows). Scale bar in a =500μm, b =50μm. (OS- outer segment, IS- inner segment, ONL- outer nuclear layer, INL- inner nuclear layer). (DAPI- Blue, eGFP- Green).

Fig. 4. Native eGFP fluorescence in RPE whole mounts

Eyes were dosed with naked DNA, NPs, or saline as presented in Fig. 3 legend. Eyes were enucleated at PI-2 and PI-30 days. The eye cup with the RPE cells layer was flattened by making radial incisions. Shown are representative planes from confocal images of RPE flat mounts (a) or representative regions near the site of injection (b, c) (approximate site of injections indicated by asterisks). Nuclei were counterstained with DAPI. Images in panel a, b and c were taken at 2x, 40x and 100x magnification, respectively. Saline injected and uninjected eyes (data not shown) were used as controls. Scale bars in a =500μm, b =20μm, and c =10μm. (DAPI- Blue, eGFP- Green)

Fig. 5. Quantification of the distribution of eGFP positive RPE cells

(a) RPE whole mount image was taken at 2x (scale bar =400μm). (b–f) Examples of representative 10x images from a NP-treated eye used to count eGFP positive cells (b-superior, c-inferior, d-temporal, e-nasal and f-nasal) (scale bar =40μm). For example, the boxed region in panel a represents the 10x image in panel b. (g) The total percent of RPE cells expressing eGFP in each treatment group. (h) Quantification of the percent of total RPE cells expressing eGFP in each region. Data have been plotted as mean ± SEM (n=3). At PI-2 days, the eGFP distribution in NP-treated eyes was significantly higher in all regions when compared to naked DNA treatment (P<0.0001). Similarly, at PI-30 days, the distribution was significantly higher at the superior and temporal regions (P<0.01, and P<0.01, respectively) but no significant differences were found at the nasal and inferior region (P>0.05). However, compared to saline the distribution in NP treated eyes were significantly higher at both time points at all regions (P<0.001).

Fig. 6. In-vivo Fundus images showing eGFP expression in the RPE cell layer

Images were taken with two filter settings. The upper panels of a (PI-7 days) and b (PI-14 days) were taken using white light filter while the bottom panels were taken using GFP filter (excitation at ~482nm and emission at ~536nm). Injections were performed through the sclera for in vivo imaging purposes. NP injected eyes show a significant area of RPE transfection depicted by the bright green fluorescence over the light green background fluorescence. Approximate injection site has been indicated by arrows on bottom panels of both a and b.

Fig. 7. Cellular uptake and distribution of nanoparticles and naked DNA

For in vivo imaging, DNA was labeled with rhodamine. Tissues were collected from NPs (a), naked DNA (b), and saline (c) treated eyes at PI-2, 6, 24, 48 hours, 7, and 30 days for immunostaining with Lamin B (nuclear envelope marker). At 100x magnification, ~25 sections of 0.5μm thickness were imaged. Left panels show confocal stacked images while 3D surface views are shown on the right from areas boxed in the left panels. Cytoplasmic/nuclear localization of the labeled DNA and NPs was assessed. (c) Saline injections were performed as vehicle controls and did not exhibit any rhodamine fluorescence. Scale bar in c =10μm. (d) Retinal whole mounts from eyes injected with rhodamine labeled naked DNA and NPs were collected at PI-48 hrs and imaged for rhodamine fluorescence to evaluate the uptake by the retina. Images were taken at 60x magnification. Scale bar in d =20μm.(Green- Lamin B, Red- TM-Rhodamine, Blue –DAPI).

Fig. 8. VMD2-eGFP NPs cause no gross retinal toxicity

No significant effects of naked DNA or NPs on scotopic (a) or photopic (b) retinal function are observed at PI-30 days (P>0.05). Data are presented as mean ± SEM (n = 5). (c) qRT-PCR for RPE65 was performed on treated eyes at the indicated time points. No significant differences in RPE65 values (normalized to HPRT, P>0.05) were detected at PI-2, -7, and -30 days in NP injected eyes when compared with naked DNA and saline treated eyes. Data are presented as mean ± SEM (n = 5–7). (d) Number of rows of photoreceptor nuclei counted for each cohort at PI-30 days. No significant differences in number of rows of photoreceptor nuclei were seen between cohorts (P>0.05, n = 4). Data are presented as mean ± SEM.