Formulation and evaluation of polymer-based nanoparticles for intravitreal gene delivery applications

Abstract

The advent of the first-ever retinal gene therapy product, involving subretinal administration of a virus-based gene delivery platform, has garnered hope that this state-of-the-art therapeutic modality may benefit a broad spectrum of patients with diverse retinal disorders. On the other hand, clinical studies have revealed limitations of the applied delivery strategy that may restrict its universal use. To this end, intravitreal administration of synthetic platforms, such as polymer-based gene delivery nanoparticles (PNPs), has emerged as an attractive alternative to the current mainstay. To achieve success, however, it is imperative that synthetic platforms overcome key biological barriers in human eyes encountered following intravitreal administration, including vitreous gel and inner limiting membrane (ILM). Here, we introduce a series of experiments from the fabrication of PNPs to a comprehensive evaluation of relevant experimental models to determine whether PNPs overcome these barriers and efficiently deliver therapeutic gene payloads to retinal cells. We conclude the protocol by providing a few important considerations for successful implementation of the strategy.

BASIC PROTOCOL 1: Preparation and characterization of PNPs

BASIC PROTOCOL 2: Evaluation of in vitro transfection efficacy

BASIC PROTOCOL 3: Evaluation of PNP diffusion in vitreous gel

BASIC PROTOCOL 4: Ex vivo assessment of PNP penetration within vitreoretinal explant culture

BASIC PROTOCOL 5: Assessment of in vivo transgene expression mediated by intravitreally administered PNPs

INTRODUCTION:

Over 2.2 billion people worldwide experience vision impairment or irreversible vision loss (Blindness, Vision Impairment, & Vision Loss Expert Group of the Global Burden of Disease, 2021). The discovery of numerous disease-causing genetic defects and disease-remedying genetic medicines has ushered in an era of gene therapy that can potentially treat previously incurable retinal disorders. Moreover, there has been an enormous effort to develop gene delivery platforms, based on naturally-occurring or synthetic materials, that enable safe and efficient gene transfer to target tissues and cells while protecting the gene payloads from enzymatic degradation during the transit (Hou, Zaks, Langer, & Dong, 2021; Jiang, Abedi, & Shi, 2021; Rai, Alwani, & Badea, 2019; Sung & Kim, 2019; D. Wang, Tai, & Gao, 2019).

In 2017, the first-ever retinal gene therapy product was approved by the FDA for gene augmentation therapy of a rare type of inherited retinal disorder (IRD), Leber’s congenital amaurosis (Ciulla, Hussain, Berrocal, & Nagiel, 2020). The product, named Luxturna^®, utilizes the adeno-associated virus (AAV) which has been the gene delivery vector of choice to date due to its inherent ability to infect retinal cells with relatively minimal pathologic risks (Bulcha, Wang, Ma, Tai, & Gao, 2021). However, AAV faces various hurdles in practical and widespread applications, including the limited cargo capacity (< 4.7 kb), the uncontrollable lifespan of transgene expression, clinically reported pro-inflammatory response, therapy-inactivating immunogenicity, and extremely high cost (Chowdhury et al., 2021; Colella, Ronzitti, & Mingozzi, 2018; “Spark’s gene therapy price tag: $850,000,” 2018). As of 2021, 47 clinical trials have been approved by the FDA to investigate ocular gene therapy strategies, including those currently ongoing and completed (Tawfik, Chen, Goldberg, & Sabel, 2022). As the coverage of retinal gene therapy expands from IRDs to more prevalent acquired retinal diseases (Guimaraes, Georgiou, Bainbridge, & Michaelides, 2021; J. H. Wang, Roberts, & Liu, 2020), the need for safe, efficient, and cost-effective gene delivery strategies is ever-increasing. Non-viral synthetic gene delivery platforms offer an attractive alternative to AAV with a favorable safety profile, negligible immunogenicity, dose flexibility, and affordable cost (H. Yin et al., 2014). Encouragingly, ocular drug delivery systems based on various synthetic materials have already been approved by the FDA, albeit for small molecule or protein drugs rather than genetic medicines (Allyn, Luo, Hellwarth, & Swindle-Reilly, 2022). Most of the synthetic gene delivery platforms are based on or inclusive of positively charged materials, such as cationic polymers, that compact negatively charged nucleic acids (i.e., plasmid DNA) into small nanoparticles. In addition to the above-mentioned merits of synthetic platforms, polymer-based gene delivery nanoparticles (PNPs) possess vast engineering potential to modulate the chemical properties necessary to enhance the gene transfer efficacy and safety profile within specific tissues (Rai et al., 2019; Samal et al., 2012). More detailed information regarding the selection and design of polymers for PNP formulation is provided in the “Background Information” and the “Critical Parameters” sections (vide infra).

Most current clinical gene therapy studies involve the subretinal injection of the product to gain immediate access to photoreceptor or retinal pigmented epithelium (RPE) following administration (Xue, Groppe, Salvetti, & MacLaren, 2017). However, this administration modality is relatively complex and highly invasive, often incurring ocular damage or requiring an additional surgical procedure, and confines the spatial therapeutic coverage to near the injection site (Bordet & Behar-Cohen, 2019; Hartman & Kompella, 2018; Holekamp, Shui, & Beebe, 2005). The intravitreal injection has emerged as an alternative means of administration that circumvents these limitations. The method is less invasive and relatively safe while potentially providing widespread gene transfer throughout the entire retinal surface area (Duncan et al., 2018). One caveat is that a relatively high gene therapy dose would be needed due to the presence of a series of delivery barriers encountered in this administration route, such as the vitreous gel and inner limiting membrane (ILM) (Leclercq, Mejlachowicz, & Behar-Cohen, 2022; Mains & Wilson, 2013; K. Y. Zhang & Johnson, 2021a). This aspect is of particular concern for virus-based gene delivery platforms (e.g., AAV) as the necessity of large viral doses creates significant issues of safety, immunogenicity, and economic burden (Timmers et al., 2020). In contrast, synthetic platforms, such as PNPs, are well positioned to be delivered by intravitreal administration, which is attributed to the dosing flexibility and cost-effectiveness described above. The barrier properties of vitreous gel and ILM are further discussed in the “Background Information” and the “Critical Parameters” sections (vide infra).

Here, we introduce a series of protocols for the formulation, characterization, and evaluation of PNPs for intravitreal gene delivery applications (Figure 1). We first introduce methods for the formulation and comprehensive characterization of PNPs (Basic Protocol 1), followed by step-by-step protocols for evaluating in vitro transfection efficiency (Basic protocol 2), ex vivo diffusion in vitreous gel (Basic Protocol 3), and ex vivo penetration within vitreoretinal explant cultures (Basic Protocol 4). Finally, we detail the methods for assessing in vivo gene transfer efficacy of PNPs following intravitreal administration into mouse eyes (Basic Protocol 5).

Figure 1. Formulation, characterization, and evaluation of polymer-based gene delivery nanoparticles (PNPs) for intravitreal gene transfer.

TEM: transmission electron microscopy.

BASIC PROTOCOL 1: Preparation and characterization of PNPs

This protocol describes the detailed methods of PNP preparation and characterization. We and others have engineered PNPs using a variety of cationic polymers, including but not limited to polyethyleneimine (PEI) (K. Negron et al., 2019; Suk et al., 2014), poly-L-lysine (PLL) (Konstan et al., 2004; Suk et al., 2014), poly(amidoamine) dendrimer (Tarach & Janaszewska, 2021), and poly(β-amino ester) (PBAE) (Figure 2) (Anderson, Akinc, Hossain, & Langer, 2005; Mastorakos et al., 2015; Karina Negron et al., 2020; Patel et al., 2019). For simplicity, we here introduce a method of preparing PNPs based on PEI, which is the most widely utilized carrier material for gene delivery applications that has entered into multiple clinical studies (Buscail et al., 2015; Shi et al., 2017; H. Yin et al., 2014). Nevertheless, we note that more recently introduced polymers have demonstrated superior performances compared to PEI in several in vitro and preclinical studies (Anderson et al., 2005; Anderson, Lynn, & Langer, 2003; Mastorakos et al., 2015; Routkevitch et al., 2020). Importantly, a panel of formulation parameters, including DNA and polymer concentrations, polymer-to-DNA weight and volume ratios, and buffer conditions (e.g., ionic strength, pH, etc.), should be optimized for individual polymers and their modifications. However, the methodology of PNP characterization described here can be generally applied irrespective of polymer types and specific applications.

Figure 2. Cationic polymers widely used for PNP preparation.

Chemical structures of (A) polyethylenimine (PEI), (B) poly-L-lysine (PLL), (C) poly(amidoamine) (PAMAM) dendrimer, and (D) poly(β-amino ester) (PBAE). A generation 2 (G2) PAMAM dendrimer is shown here as an example, but dendrimers with greater generations are widely used. PBAE polymers are synthesized with various diacrylate monomers (R₁; e.g., 1,3-propanediol diacrylate), amine alcohols (R₂; e.g., 5-amino-1-pentanol), and amine-containing end-capping groups (R₃; e.g., propane-1,3-diamine).

Materials

• 1.5 mL Microcentrifuge tubes (Eppendorf Microcentrifuge Tubes or equivalent)

• Polyethyleneimine (PEI; Sigma-Aldrich, cat. no. 408727)

• Plasmid DNA; Luciferase-pcDNA3 (Addgene, cat. no. 18964)

• UltraPure^™ DNase/RNase-Free Distilled Water (Invitrogen, cat. no. 10977015)

• Amicon^® Ultra-4 Centrifugal Filter Unit (MWCO of 100 kDa; Millipore Sigma, cat. no. UFC810024)

• Transmission Electron Microscope (TEM; Hitachi or equivalent)

• Transmission Electron Microscopy Grids (Electron Microscopy Sciences, cat. no. FCF-300-CU)

• UranyLess (Electron Microscopy Sciences, cat. no. 2240920)

• Certified Molecular Biology Agarose (Bio-Rad, cat. no. 1613101)

• 50x TAE (Tris/Acetic Acid/EDTA) Buffer (Bio-Rad, cat. no. 1610743)

• SYBR^™ Green I Nucleic Acid Gel Stain - 10,000X concentrate in DMSO (Invitrogen, cat. no. S7563)

• 6x DNA Loading Dye (Thermo Fisher, cat. no. R0611)

• DNase I (Thermo Fisher, cat. no. EN0521)

• 0.5 M EDTA (Invitrogen, cat. no. AM9260G)

• Heparin sodium salt from porcine intestinal mucosa (Sigma-Aldrich, cat. no. H3149–10KU)

• Vortex machine (Fisherbrand or equivalent)

• Benchtop centrifuge (Thermo Scientific or equivalent)

• NanoDrop^™ 2000 Spectrophotometer (Thermo Scientific, ND2000)

• Dynamic Light Scattering device (DLS; Malvern Panalytical or equivalent)

• Electrophoresis unit (Malvern Panalytical or equivalent)

• ChemiDoc Imagers (Bio-Rad or equivalent)

PNP formulation

1. Prepare aqueous solutions of plasmid DNA and polymers using UltraPure^™ DNase/RNase-free distilled water. Note that most cationic polymers exhibit basic pH when dissolved in a non-buffered aqueous solution, which is generally adjusted to a slightly acidic pH using 1M HCl or other buffer solutions. As an example, DNA and polymer solutions described in the table below are prepared to formulate PEI-based PNPs with a plasmid DNA batch size of 20 μg.

   Material	Amount
   DNA solution (0.2 μg/μL)	100 μL
   PEI solution (0.8 μg/μL, pH 6.5)	20 μL

   Note that an in-house optimized polymer-to-DNA weight ratio for PEI is applied here to introduce an example protocol for PNP formulation, but the ratio can be further adjusted as needed. Importantly, a range of the polymer-to-DNA ratios should be screened for other carrier materials (e.g., newly synthesized cationic polymers) to determine the conditions that enable full compaction of DNA (vide infra “DNA compaction”).

2. Transfer the polymer solution to a 1.5-mL Eppendorf tube.

3. Add the DNA solution to the polymer solution dropwise while vortexing, followed by an additional 10-second vortexing.

4. Incubate the solution for 30 minutes at room temperature.

5. Transfer the PNP solution to an Amicon^® Ultra-4 centrifugal filter unit and add 3.5 mL of water.

6. Centrifuge the Amicon^® Ultra-4 centrifugal filter unit for 10 minutes at 1,000 × g using a benchtop centrifuge and discard the flowthrough.

7. Fill the Amicon^® Ultra-4 centrifugal filter unit with UltraPure^™ DNase/RNase-free distilled water up to 3.5 mL.

   Repeat at least 3 times from step 6 to step 7 to remove free polymers that are not involved in DNA compaction.

8. Collect the unfiltered PNP solution and transfer it to a 1.5-mL tube.

DNA compaction

1. To prepare a 1% agarose gel, dissolve 1 g of agarose in 100 mL 1x TAE buffer in microwavable glassware.

2. Microwave for less than a minute until the agarose is completely dissolved.

   Do not overboil the gel. Intermittent microwaving and swirling of the glassware are recommended.

3. Cool down the temperature approximately to 50°C and add 10 μL of SYBR^™ Green in agarose solution.

4. Pour slowly the agarose solution into a gel tray placed in the electrophoresis unit and place the well comb.

5. Cool the gel at 4°C or room temperature for 30 minutes to completely solidify.

6. During cooling of the gel, dilute the PNP solution or uncompacted DNA solution (as negative control) with UltraPure^™ DNase/RNase-free distilled water to a 2.5-ng/μL DNA concentration.

7. Transfer 10 μL of each sample to a 1.5-mL tube and add 2 μL of 6x DNA loading dye.

8. After confirming that the gel is completely solidified, carefully remove the well comb.

9. Fill the electrophoresis unit with 1x TAE buffer until the gel is completely soaked into the buffer.

10. Carefully load 10 μL of DNA ladder or sample solutions into individual wells.

11. Run the gel at 80–100 V until the visible dye is 75% of the way down the gel, which typically takes about 30 – 40 minutes.

12. Analyze the gel with ChemiDoc Imagers or a functionally equivalent device (Figure 3A).

Figure 3. Formulation and characterization of PNPs.

(A) Gel electrophoretic migration assay demonstrating DNA compaction by polymers at various polymer-to-DNA weight ratios. (B) Hydrodynamic diameters and (C) ζ-potentials of PNPs measured by DLS. (D) Representative transmission electron micrograph of PNPs. Scale bar = 200 nm. (E) Gel electrophoretic migration assay demonstrating the protection of DNA payloads by PNPs against DNase. L: DNA ladder.

Fundamental physicochemical characterization of PNPs

1. Quantify the yield of or the DNA concentration in the final PNP solution using NanoDrop^™ 2000 Spectrophotometer, using UltraPure^™ DNase/RNase-free distilled water as a blank.

   PEI does not interfere with DNA absorbance at 260 nm, but if a carrier polymer exhibits inherent absorbance at this wavelength (e.g., PBAE), use of fluorescently labeled DNA is recommended for a fluorimetric analysis to determine the final yield.

2. Adjust or dilute the PNP solution to a DNA concentration of 0.1 μg/μL by adding UltraPure^™ DNase/RNase-free distilled water.

3. Measure the hydrodynamic diameter, polydispersity index (PDI), and ζ-potential of PNPs using a DLS device (Figures 3B and 3C).

4. Determine the geometric diameter and morphology of PNP using TEM (Figure 3D). Place a drop of a diluted PNP solution on a TEM grid, incubated for 5 minutes, and washed 3 times with water. After carefully removing the water, perform negative staining by adding a droplet of 1% UranyLess aqueous solution to the grid, wash with water 3 times, and completely air-dry the grid prior to TEM.

DNA protection

1. Prepare the gel as described in the steps 1 – 5 of the “DNA compaction” section.

2. During cooling of the gel, dilute the PNP solution or uncompacted DNA solution to a 25-ng/μL DNA concentration

3. Transfer 10 μL of each sample to a 1.5-mL tube and add 2 μL of DNase, followed by a 15-minute incubation at 37°C.

4. Deactivate the DNase by adding 1.2 μL of 0.5 M EDTA solution at 65°C.

5. Induce the release of DNA from the PNPs (i.e., decompaction) by adding 30 μL of 0.1 mg/mL heparin solution for a 10-minute incubation at room temperature.

6. Add 57 μL of nuclease-free ultrapure water to the sample.

7. Transfer 10 μL of sample solution to a 1.5-mL tube and add 2 μL of 6x DNA loading dye.

8. Carefully load 10 μL of DNA ladder or sample solutions into individual wells. Example test conditions are shown in the table below:

   Sample	DNase	Heparin	Description
   DNA	−	−	Untreated DNA
   	+	−	DNase-treated DNA
   PNP	−	−	Untreated PNP
   	−	+	Heparin-treated PNP
   	+	−	DNase-treated PNP
   	+	+	DNase/heparin-treated PNP

9. Run and analyze the gel as described in the steps 11 – 12 of the “DNA compaction” section (Figure 3E).

   DNA protected against enzymatic degradation and subsequently released from PNPs retains the original molecular weight (i.e., location of bands identical to the unperturbed DNA). Ideally, no band for any degraded DNA products should be visible to ensure full protection of the DNA payloads by PNPs.

Colloidal stability in physiological conditions

1. Transfer 20 μL of the PNP solution to a 1.5-mL tube and add 180 μL of PBS or other physiologically relevant solutions (e.g., a solution of hyaluronan, a primary component of vitreous gel).

2. Incubate at 37°C and monitor the changes in hydrodynamic diameter, PDI, and ζ-potential over time up to an hour or more.

   Particles with excellent colloidal stability retain hydrodynamic diameter and PDI in a physiologically relevant condition.

BASIC PROTOCOL 2: Evaluation of in vitro transfection efficacy

In vitro reporter transgene expression is one of the most standard outcome measures that determine the performance of PNPs. In general, PNPs are prepared to carry plasmid DNA expressing a reporter protein of choice and compared with a commercially available transfection agent, such as Lipofectamine^™ 3000. Here, we describe two different methods of evaluating the ability of PNPs to transfect cells under a conventional culture condition. The first approach involves the packaging of plasmid DNA expressing a fluorescence protein (e.g., ZsGreen1) into PNPs, followed by transfection of cells of interest and flow cytometric analysis to quantify the percentage of transfected cells and average fluorescence intensity. An alternative approach entails the transfection of cells with PNPs carrying luciferase-expressing plasmid DNA, followed by cell lysis to recover proteins to be analyzed for reporter luciferase activity using a luminometer. The luciferase bioluminescent activity is directly proportional to the amount of luciferase enzyme produced by the transfected cells and thus represents the overall level of transgene expression. These two methodologies are complementary in that they provide distinct information regarding the ability of PNPs to transfect cells in vitro; fluorescent reporters demonstrate the percentage of a cell transfected whereas bioluminescent reporters demonstrate the overall level of protein expression in transfected cells. Of note, it is now widely appreciated that transfection outcomes in vitro do not necessarily correlate with in vivo performances of PNPs and thus the data should be interpreted with caution and in a context-dependent manner.

Materials

• 6-well plate (Corning, cat. no. 3516, or equivalent)

• 96-well plate (Corning, cat. no. 3585, or equivalent)

• Human retinal pigment epithelial (RPE) ARPE-19 cell line (ATCC, cat. no. CRL-2302)

• Plasmid DNA; ZsGreen1-N1 (Addgene, cat. no. #54702) and Luciferase-pcDNA3 (Addgene, cat. no. 18964)

• DMEM, high glucose (Thermo Fisher, cat. no. 11965092)

• Opti-MEM^™ Reduced Serum Medium (Thermo Fisher, cat. no. 31985062)

• Fetal Bovine Serum (FBS; Thermo Fisher, cat. no. 26140)

• Penicillin-Streptomycin (10,000 U/mL) (Gibco, cat. no. 15140122)

• Phosphate Buffer Saline (PBS, Thermo Fisher, cat. no. 10010023)

• Trypsin-EDTA (0.05%) (Gibco, cat. no. 25300054)

• 5X Reporter Lysis Buffer (Promega, cat. no. E3971)

• Lipofectamine^™ 3000 Transfection Reagent (Invitrogen, cat. no. L3000001)

• Luciferase Activity System (Promega, cat. no. E1500)

• Pierce^™ BCA Protein Assay Kit (Thermo Fisher, cat. no. 23225)

• Cell incubator (Thermo Fisher or equivalent)

• Fluorescence-activated cell sorting (FACS; SONY, SH800 or equivalent)

• Luminometer (Promega, GloMax, or equivalent)

• Microplate reader (Thermo Fisher or equivalent)

In vitro reporter fluorescent protein expression: flow cytometric analysis

1. Seed human RPE cells (ARPE-19), or a target cell type of interest, on a 6-well plate at a cell density of 5 × 10⁵ cells per well and replenish with 2 mL of culture medium (i.e., DMEM supplemented with 10% FBS and 1% penicillin-streptomycin).

   Cell seeding density can be adjusted depending on the cell type of interest.

2. Incubate the cells at 37°C and 5% CO₂ level for 24 hours.

3. Remove the medium and wash the cells in each well with 1 mL of PBS.

4. Replace the PBS with 1 mL of Opti-MEM^™ medium.

5. Incubate the cells at 37°C and 5% CO₂ level for 30 minutes.

6. Prepare an aqueous solution of PNPs carrying ZsGreen1-expressing plasmid DNA.

   Note that the PNP formulation condition should be optimized for a carrier polymer of interest, which applies to all subsequent studies beyond this protocol. Cells treated with saline or Lipofectamine^™ 3000 carrying the identical plasmid DNA serve as a negative or a positive control, respectively.

7. Dilute the PNP solution to a DNA concentration of 5 μg/mL with Opti-MEM^™ medium.

8. Remove the medium and wash the wells with 1 mL of PBS.

9. Treat the cells in each well with 1 mL of PNPs in Opti-MEM^™ medium. The wells replenished with Opti-MEM^™ medium devoid of PNPs serve as an untreated control.

10. Incubate the cells at 37°C and 5% CO₂ level for 48 hours.

11. Wash the wells by replacing the medium with 1 mL of PBS and repeat the washing twice.

12. Treat the cells in each well with 300 μL of trypsin-EDTA (0.05%).

13. Incubate the cells at 37°C for 3 minutes.

14. Add 1.2 mL of fresh culture medium to the wells.

15. Detach the cells from the bottom of the wells by pipetting.

16. Collect the 1.5 mL of cell suspension.

17. Centrifuge the cell suspension at 100 × g for 3 minutes at 4°C.

18. Gently discard the supernatant.

19. Resuspend the cell pellet with 500 μL of PBS containing 1% FBS.

    After the cells are resuspended, the solution needs to be stored at 4°C using an ice box or a cooling rack.

20. Transfer the cell suspension to the flow cytometry tube.

21. Insert the tube into flow cytometer.

22. Run flow cytometry to assay at least 1 × 10⁴ cells with a proper fluorescence setting.

    The reporter fluorescent ZsGreen1 protein has the excitation and emission peak wavelengths of 488 and 509 nm, respectively.

23. Gate the cell population based on the forward scatter (FSC) and side scatter (SSC) signal.

24. Plot the histogram of the cell population based on the fluorescence intensity.

25. Set a background threshold fluorescence intensity to exclude over 99% of untreated cell population.

26. Apply the identical threshold to the other treatment groups. The cell populations beyond the threshold are deemed positive for reporter transgene expression (Figure 4).

    Experiments are conducted at least with a triplicate of each treatment condition (e.g., comparison of multiple PNPs) for quantitative and statistical analysis. PNPs that exhibit a significantly greater percentage of transfected cells compared to Lipofectamine^™ 3000 or other commercial agents are generally considered promising.

Figure 4. In vitro transfection of ARPE-19 cells by PNPs.

Flow cytometric analysis of ARPE-19 cells treated with saline or either Lipofectamine^™ 3000 (LIPO) or PNPs carrying plasmid DNA expressing fluorescent ZsGreen1 protein. The percentages of ARPE-19 cells expressing the ZsGreen1 protein are noted.

In vitro reporter luciferase activity: cell homogenate-based luciferase assay

1. Seed human RPE cells (ARPE-19), or a cell type of interest, on a 96-well plate at a cell density of 1 × 10⁴ cells per well and replenish with 200 μL of culture medium (DMEM containing 10% FBS and 1% penicillin-streptomycin).

   Cell seeding density can be adjusted depending on the cell type of interest.

2. Incubate the cells at 37°C and 5% CO₂ level for 24 hours.

3. Remove the medium and wash the cells in each well with 100 μL of PBS.

4. Replace the cultured media with 100 μL of Opti-MEM^™ media

5. Incubate the cells at 37°C and 5% CO₂ level for 30 minutes.

6. Prepare an aqueous solution of PNPs carrying luciferase-expressing plasmid DNA.

   Cells treated with saline or Lipofectamine^™ 3000 carrying the identical plasmid DNA serve as a negative or a positive control, respectively.

7. Dilute the PNP solution to a DNA concentration of 5 μg/mL with Opti-MEM^™ medium.

8. Remove the medium and wash the wells with 100 μL of PBS.

9. Treat the cells in each well with 100 μL of PNPs in Opti-MEM^™ medium.

10. Incubate the cells at 37°C and 5% CO₂ level for 48 hours.

11. Wash the wells by replacing the medium with 100 μL of PBS and repeat the washing twice.

12. Treat the cells in each well with 100 μL of reporter lysis buffer.

    5X reporter lysis buffer should be diluted to 1X concentration with distilled water to apply an optimal osmotic pressure to the cells.

13. Apply three freeze-and-thaw cycles to the cells to ensure cell lysis. Cells are frozen at −80°C and subsequently thawed at room temperature.

14. Collect 100 μL of cell lysate.

15. Centrifuge the lysate at 12,000 × g at 4°C for 20 minutes.

16. Collect the supernatant gently without disturbing the pellet.

17. Mix 10 μL of the supernatant with 100 μL of luciferin substrate buffer.

18. Measure the relative light unit (RLU) as luciferase activity with a luminometer.

19. Quantify the protein concentration of the supernatant by BCA protein quantification assay.

20. Normalize the RLU value to the protein content.

    Experiments are conducted at least with a triplicate of each treatment condition (e.g., comparison of multiple PNPs) for quantitative and statistical analysis. PNPs that exhibit a significantly greater bioluminescence signal compared to Lipofectamine^™ 3000 or other commercial agents could be considered promising.

BASIC PROTOCOL 3: Evaluation of PNP diffusion in vitreous gel

Intravitreally administered PNPs are delivered directly into the vitreous gel, which serves as the frontline extracellular delivery barrier prior to reaching the retinal tissue. The vitreous gel is a biological hydrogel matrix composed of adhesive macromolecules that can trap particulate matters via multivalent adhesive interactions (Xu et al., 2013). Thus, it is critical that PNPs are designed and experimentally confirmed to traverse this barrier by avoiding the adhesive interactions to gain access to the retinal tissue. Here, we introduce a straightforward means to investigate the diffusion behaviors of PNPs in vitreous gel collected from a rabbit eye using the multiple particle tracking (MPT) technique. We note that the method can be equally applicable to vitreous gels collected from other sources, such as bovine eyes and post-mortem human eyes.

Materials

• Rabbit eye

• Surgical scissors and forceps (Pakistan, cat. no. 40-230-21 and 40-252-23)

• Sterile Rigid Pack 3-mL Syringes, 23 G (Medline, cat. no. 8881513033)

• Label IT^® Nucleic Acid Labeling Kit, Cy5 (Mirus, cat. no. MIR 3725)

• Dynabeads^™ Spot-On^™ Slides (Applied Biosystems, cat. no. 74004)

• Fisherbrand^™ Cover Glasses (Thermo Fisher, cat. no. 12-542-B)

• Fluorescence microscope (Carl Zeiss, Axiovert, or equivalent)

• MetaMorph^® Microscopy Automation & Image Analysis Software (Molecular Devices LLC)

Particle diffusion in vitreous gel

1. Fluorescently label plasmid DNA with Cy5 using Label IT^® Nucleic Acid Labeling kit as per the manufacturer’s protocol.

2. Prepare an aqueous solution of PNPs carrying Cy5-labeled plasmid DNA for microscopic visualization.

   We have previously confirmed that the use of fluorescently labeled DNA does not perturb the particle physicochemical properties irrespective of the type of carrier polymer utilized for PNP formulation.

3. Dissect the eye from a euthanized rabbit by clasping behind the globe using curved forceps and applying brisk outward force, and keep it at 4°C during the sample preparation for the PNP diffusion analysis.

   Animal procedures are performed in accordance with the guidelines and policies of the Institutional Animal Care and Use Committee (IACUC) for humane treatment of vertebrates.

4. Aspirate the vitreous gel through the sclera at the limbus using a 23G needle syringe.

   The volume of the rabbit vitreous is approximately 1.5 mL. The needle tip is inserted into the vitreous cavity approximately to a depth of 1 cm, and the vitreous gel is gently pulled out by the syringe. The harvested vitreous gel is immediately used for particle diffusion analysis.

5. Mix 30 μL of the vitreous gel with 1 μL of Cy5-labeled PNP solution prepared at a DNA concentration of 1 μg/μL.

   A minimal volume of the PNP solution is added to the vitreous gel not to perturb the inherent barrier properties of vitreous gel (i.e., <5% dilution by the addition of PNP solution).

6. Transfer the mixture into the central circular well of a SPOT-ON^™ slide and seal the well with a cover glass.

   The mixture should be sealed without creating air bubbles.

7. Place the SPOT-ON^™ slide on the stage of the fluorescence microscope under a 100X oil-immersion objective.

8. Adjust the objective to focus a plane away from the cover glass using the excitation and emission peak wavelengths of 651 and 670 nm, respectively.

   The plane that shows uniform immobilization of PNPs carrying Cy5-labeled plasmid DNA indicates the interface between the cover glass and the vitreous gel.

9. Record the motions of Cy5-labeled PNPs in the vitreous gel at a frame rate of 15 frames per second (i.e., 67 ms per frame) for 20 seconds using the MetaMorph^® software (Figure 5A).

   Movies are captured at more than 10 randomly selected locations.

10. Calculate the mean square displacement (MSD) using a custom-made automated MPT computational software coded in MATLAB (Figure 5B).

    The MSD is a square of the distance traveled by an individual particle at a given time interval (i.e., time scale), and thus it is directly proportional to the particle diffusion rate. The slope (m) of MSD vs. time scale curve of 1 represents the unhindered Brownian motion whereas m < 1 indicates hindered motions of particulate matters. Figure 4B shows that diffusion of PNPs used in this example study is negligibly hindered in rabbit vitreous gel.

Figure 5. Diffusion of PNPs in vitreous gel.

(A) Representative trajectories of vitreous-impermeable (left) and vitreous-penetrating (right) PNPs in rabbit vitreous gel. Scale bars = 10 μm. (B) The mean square displacement (MSD) over time scale for vitreous-penetrating (solid line) and vitreous-impermeable (dashed line) PNPs. MSD is the square of distance traveled by an individual nanoparticle at a given time interval (i.e., time scale). The inset shows the representative relationships of MSD and time scale for nanoparticles exhibiting unhindered Brownian (m = 1; red) or hindered (m < 1; blue) motions.

BASIC PROTOCOL 4: Ex vivo assessment of PNP penetration within vitreoretinal explant culture

The PNPs that manage to diffuse within the vitreous gel and reach the vicinity of retinal tissue face an additional extracellular barrier, inner limiting membrane (ILM). The ILM is the basement membrane at the ocular vitreoretinal interface, which is now well appreciated as a significant barrier to intravitreally administered ocular therapies, such as gene therapy (K. Y. Zhang & Johnson, 2021a). Apparently, PNPs incapable of traversing the ILM cannot reach the retinal parenchyma to mediate therapeutic transgene expression in retinal cells. The organotypic culture of vitreoretinal explant has been shown to maintain the integrity of a continuous ILM at least for 7 days (K. Y. Zhang & Johnson, 2021a) and thus serves as an excellent testbed to evaluate the ILM-penetrating ability of PNPs. Here, we provide a brief introduction to the methods of preparing the vitreoretinal explant culture using a post-mortem human eye and evaluating the ability of PNPs to traverse the ILM in the model (Figure 5). A more detailed method of preparing the vitreoretinal explant culture can be found elsewhere (K. Y. Zhang & Johnson, 2022).

Materials

• 6-well plate (Corning or equivalent)

• Human cadaver eye (Eye Bank Association of America)

• Plasmid DNA; ZsGreen1-N1 (Addgene, cat. no. #54702)

• Phosphate Buffer Saline (PBS; Thermo Fisher, cat. no. 10010023)

• BD Microlance^™ 3 Needles, 30 G (BD, cat. no. 304000)

• Westcott spring scissors (Fine Science Tools, cat. no. 15000–11)

• Trephine 5 mm (Fine Science Tools, cat. no. 18004–50)

• Organotypic culture inserts (Millipore Sigma, cat. no. PICM0RG50)

• Whatman^® cellulose filter paper (Whatman, cat. no. 1001090)

• Neurobasal^™-A Medium (Gibco, cat. no. 10888022)

• B-27^™ Plus Supplement (50X) (Gibco, cat. no. A3582801)

• N-2 Supplement (100X) (Gibco, cat. no. 17502048)

• GlutaMAX^™ Supplement (Gibco, cat. no. 35050061)

• Penicillin-Streptomycin (10,000 U/mL) (Gibco, cat. no. 15140122)

• Paraformaldehyde Solution, 4% in PBS (PFA; Thermo Fisher, cat. no. J19943.K2)

• ProLong^™ Diamond Antifade Mountant with DAPI (Thermo Fisher, cat. no. P36971)

• Fisherbrand^™ Superfrost^™ Plus Microscope Slides (Thermo Fisher, cat. no. 12-550-15)

• Fisherbrand^™ Cover Glasses (Thermo Fisher, cat. no. 12–542-B)

• Confocal Microscope (Carl Zeiss, LSM 880, or equivalent).

Ex vivo penetration through ILM

1. Obtain a human cadaver eye from the Eye Bank Association of America. A minimal time from death to retinal dissection is preferable and should be less than 18 hours.

   Alternatively, bovine, porcine, leporine, or murine eye can be used to prepare vitreoretinal explant culture following the procedures described below.

2. Prepare an aqueous solution of PNPs carrying either Cy5-labeled plasmid DNA or ZsGreen1-expressing plasmid DNA at a DNA concentration of 1 μg/μL. Refer to the “Basic Protocol 1” (vide supra).

3. Puncture the sclera at 4 mm posterior to the limbus using a 30G needle.

4. Incise the sclera using Westcott spring scissors at the ora serrata circumferentially.

5. Separate the anterior and posterior segments of the globe.

6. Create 4 radial relaxing incisions in the poster segment and flatten the tissue in a Maltese cross configuration.

7. Punch out circles of retinal tissue from the eyeball using a 5-mm trephine. Record the location of each retinal sample, as the composition of retinal cells and the ILM differs according to neuroanatomical location (i.e., fovea, macula, periphery).

8. Gently separate the vitreoretinal tissue from the underlying RPE and choroid.

9. Transfer the free-floating retina in PBS carefully onto an organotypic culture insert using a sterile P1000 pipette with tip-widened by cutting with a straight razor.

   Orient the vitreous and retinal ganglion cell (RGC) layer of the retina inside a culture insert to face up while photoreceptors are against the membrane of the insert.

10. Aspirate excess PBS from the insert and remove residual fluid by applying Whatman^® cellulose filter paper to the undersurface of the organotypic membrane. This will help to gently adhere the tissue to the organotypic membrane and prevent detachment of the tissue in the next step.

11. Place the culture insert with the flat-mounted retina in the 6-well plate with 1.5mL of Retinal Explant Culture Media (Neurobasal^™- A Medium containing 2% B-27^™ Plus Supplement, 1% N-2 Supplement, 1% GlutaMAX^™ Supplement, and 0.5% penicillin-streptomycin) in the well. The organotypic culture is maintained under an air-fluid interface with media perfusing through the organotypic filter feeds the culture.

12. To evaluate the ability of PNPs to traverse the explant culture, particularly through the ILM, apply 1 μL of an aqueous solution of PNPs carrying Cy5-labeled plasmid DNA to the middle of a few vitreoretinal explant cultures in an identical manner.

13. Incubate the inserts at 37°C and 5% CO₂ level for varying amounts of time (e.g., 1, 2, 6 and 12 hours).

14. Remove the medium and wash each insert with 3 mL of PBS.

15. Fix the explant for 1 hour by adding 3 and 5 mL of 4% PFA solution to the 6-well plate and the insert, respectively.

16. Wash each insert by replacing the PFA solution with PBS and repeat the washing 3 times.

17. Collect the explant attached to the membrane by cutting the edge of the membrane with a circumferential incision with a crescent blade.

18. Place the explants on microscope slides.

19. Mount the explants with 2 drops of ProLong^™ DAPI solution.

20. Cover the explants with a cover glass without creating air bubbles.

21. Perform Z-stack confocal imaging of explants incubated with PNPs for different time periods to assess the time-course PNP penetration through the ILM and the retinal layer.

    The Cy5 fluorophore has the excitation and emission peak wavelengths of 651 and 670 nm, respectively

22. For evaluating reporter transgene expression by retinal cells, apply 1 μL of an aqueous solution of PNPs carrying ZsGreen1-expressing plasmid DNA in the middle of the explant culture.

23. Incubate the insert at 37°C and 5% CO₂ level for 48 hours.

24. Follow the previous steps 13 to 19.

25. Perform Z-stack confocal imaging of the explant to assess the reporter transgene expression throughout the retinal layer.

    The reporter ZsGreen1 protein has the excitation and emission peak wavelengths of 488 and 509 nm, respectively. Successful ILM penetration and subsequent retinal cell transduction are determined by the positive fluorescent reporter transgene expression profile over the retinal layer spanning ganglion cell layer, inner nuclear layer, and outer nuclear layer.

BASIC PROTOCOL 5: Assessment of in vivo transgene expression mediated by intravitreally administered PNPs

The development of novel gene delivery platforms for retinal applications is culminated by validating their ability to mediate transgene expression in retinal tissue in vivo. Small animals, such as rodents, are highly relevant to subsequent studies to evaluate therapeutic gene delivery and efficacy given that most of the widely used preclinical ocular disease models are based on rodents. In this final protocol, we describe procedures evaluating the ability of PNPs to deliver reporter plasmid DNA to and mediate transgene expression in retinal cells following intravitreal administration.

Materials

• C57BL/6 mice (Charles River)

• 1.5 mL Microcentrifuge tubes (Eppendorf Microcentrifuge Tubes or equivalent)

• Plasmid DNA; ZsGreen1-N1 (Addgene, cat. no. #54702)

• Ketamine HCl Injectable Solution 100 mg/mL C3N (Covetrus, cat. no. 11695-6840-1)

• Xylazine Injection 100 mg/mL 50 cc (Vet One, cat no. Vet-Rx-MW 510650)

• Ofloxacin Ophthalmic Solution 0.3% (Ofloxacin, cat. no. 146835)

• Neo-Poly-Bac Ophthalmic Ointment (Bausch+Lomb, cat. no. IWM044525)

• Forane (Isoflurane, UPS) Iiquid for Inhalation (Baxter, cat. no. 10019036060)

• Far Infrared Warming Pad Controller with Warming Pad (20.3 cm W × 25.3 cm L) (Kent Scientific, cat. no. RT-0520)

• D-luciferin Potassium Salt (GoldBio, cat. no. LUCK-100)

• Cotton Tipped Applicators – Medical, 6” (Uline, cat no. S-21102)

• Paraformaldehyde Solution, 4% in PBS (PFA; Thermo Fisher, cat. no. J19943.K2)

• Whatman^® cellulose filter paper (Whatman, cat. no. 1001090)

• Phosphate Buffer Saline (PBS; Thermo Fisher, cat. no. 10010023)

• Bovine Serum Albumin (Sigma-Aldrich, cat. no. A7030)

• Triton^™ X-100 (Sigma-Aldrich, cat. no. X100)

• Tissue-Tek^® O.C.T. Compound (Sakura, cat. no. 4583)

• Living Colors^® ZsGreen Monoclonal Antibody (Takara, cat. no. 632598)

• Goat Anti-Rabbit IgG H&L (Alexa Fluor^® 647) (Abcam, cat. no. ab150079)

• ProLong^™ Diamond Antifade Mountant with DAPI (Thermo Fisher, cat. no. P36971)

• Fisherbrand^™ Superfrost^™ Plus Microscope Slides (Thermo Fisher, cat. no. 12-550-15)

• Fisherbrand^™ Cover Glasses (Thermo Fisher, cat. no. 12–542-B)

• 1 mm Spherical Tungsten Carbide Milling Media Balls (Polished) (MSE Supplies, cat. no. BA0622)

• Microinjector (Warner Instruments, cat. no. PLI-100A)

• Ophthalmic Microsope (Carl Zeiss, Stemi 508 or equivalent)

• In Vivo Imaging System (IVIS; PerkinElmer or equivalent)

• Microtome (Leica or equivalent)

• Confocal Microscope (Carl Zeiss, LSM 880 or equivalent)

• TissueLyser LT (Qiagen, 85600 or equivalent)

In vivo reporter luciferase activity: live animal imaging and tissue homogenate-based luciferase assay

1. Prepare an aqueous solution of PNPs carrying luciferase-expressing plasmid DNA at a DNA concentration of 1 μg/μL. Refer to the “Basic Protocol 1” (vide supra).

2. Turn on the microinjector and a nitrogen gas tank

3. Adjust the pressure meter of the microinjector to approximately 40 psi.

4. Insert the capillary needle into the pipette holder.

5. Anesthetize a mouse by intraperitoneally injecting a mixture solution of ketamine (100 mg/kg) and xylazine (10 mg/kg) solution.

   Animal procedures are performed in accordance with the guidelines and policies of the IACUC for humane treatment of vertebrates.

6. Check the depth of anesthesia by monitoring the pinch-toe/tail reflex.

7. After confirming the absence of the pinch-toe/tail reflex, place the mouse under the ophthalmic microscope.

8. Place one drop of an ophthalmic solution on the ocular surface and wait for 1 minute.

9. Gently open the eyelids with two fingers and puncture the sclera at the limbus with the tip of the capillary needle by carefully adjusting the angle to the ocular surface at 45 degrees, thereby avoiding the damage to the lens.

   Lens damage can cause vision disability or cataract and incites an inflammatory reaction within the eye that can affect the experimental results.

10. Inject 1 μL of the PNP solution using the microinjector.

11. Leave the needle tip inserted in the mouse eye for approximately 2 seconds prior to removal to prevent the leakage of the injected content.

12. Apply the Neo-Poly-Bac ophthalmic ointment to the ocular surface using a cotton tipped applicator.

13. Leave the mouse on the heating pad until full recovery from anesthesia.

14. Forty-eight hours after the administration, anesthetize the treated mouse with Isoflurane gas (2.5%, O₂ flow rate 0.5 L/minute).

15. Inject D-luciferin potassium salt solution intraperitoneally at a dose of 150 mg/kg.

    D-luciferin is prepared in PBS at a concentration of 15 mg/mL, and 200 μL of this stock solution is administered to a mouse weighing 20 g.

16. Place the animals in the IVIS chamber.

17. Take luminescence images of the eye at 10 minutes post-injection of D-luciferin (Figure 7A).

    The post-injection incubation time can vary depending on the route of D-luciferin administration.

18. For the complementary tissue homogenate-based luciferase assay, euthanize the mouse with CO₂ 2 hours after the IVIS study.

    The time allows clearance of D-luciferin from the body.

19. Enucleate eyes by clasping behind the globe using curved forceps and applying brisk outward force.

20. Treat the whole eyeball in a 1.5 mL tube with 300 μL of reporter lysis buffer.

    5X reporter lysis buffer should be diluted to 1X concentration with distilled water to apply an optimal osmotic pressure to the cells.

21. Add 10 of 1 mm spherical tungsten carbide milling media balls into the 1.5 ml tube.

22. Homogenize the eyeball with a bead-based TissueLyser LT with 50/s oscillation rate for 5 minutes.

    Tissue lysis is conducted at 4°C to avoid denaturation of proteins during homogenization.

23. Apply three freeze-and-thaw cycles to the eyeball lysate to ensure cell lysis. Cells are frozen at −80°C and subsequently thawed at room temperature.

24. Centrifuge the lysate at 12,000 × g at 4°C for 20 minutes.

25. Collect the supernatant gently without disturbing the pellet.

26. Mix 10 μL of the supernatant with 100 μL of luciferin substrate buffer.

27. Measure the relative light unit (RLU) as luciferase activity with a luminometer.

28. Quantify the protein concentration of the supernatant by BCA protein quantification assay.

29. Normalize the RLU value to the protein content.

Figure 7. In vivo transfection of mouse retina by intravitreally administered PNPs.

(A) Representative bioluminescence images of mouse eyes intravitreally treated with saline (upper) or PNPs carrying plasmid DNA expressing luciferase (lower). (B) Representative confocal images of mouse retina following intravitreal administration of saline (upper) or PNPs carrying plasmid DNA expressing ZsGreen1 protein (lower). Reporter ZsGreen1 transgene expression (green) is observed throughout the retinal layer of an eye treated by PNPs. Blue staining represents cell nuclei. GCL: ganglion cell layer; INL: inner nuclear layer; ONL: outer nuclear layer.

In vivo reporter fluorescent protein expression: immunofluorescence

1. Prepare an aqueous solution of PNPs carrying ZsGreen1-expressing plasmid DNA at a DNA concentration of 1 μg/μL.

2. Follow the steps 2 – 13 of the “in vivo reporter luciferase activity” section.

3. Forty-eight hours after the administration, euthanize the mouse with CO₂ and enucleate eyes by clasping behind the globe using curved forceps and applying brisk outward force.

4. Fix the harvested eyeball with 4% PFA solution at 4°C for 12 hours.

5. Wash the eyeball by replacing the PFA solution with 1 mL of PBS, followed by a 5-minute incubation, and repeat the washing 3 times.

6. Remove the surface fluid from the eyeball using Whatman^® cellulose filter paper.

7. Embed the eyeball into the O.C.T. compound.

8. Store the O.C.T. block at −20°C for 1 hour.

9. Prepare a slide of a 10 μm-thickness eyeball sections using a microtome.

   The section is conducted perpendicular to the retina layer to prepare a slice of cup-shaped retinal layer.

10. Block and permeabilize the retinal tissue on the slide with 100 μL of blocking buffer (PBS containing 1% BSA and 0.1% Triton^™ X-100).

11. Wash the slides by replacing the buffer with 100 μL of PBS and repeat the washing 3 times.

12. Immerse the slides in 100 μL of Living Colors^® ZsGreen monoclonal antibody diluted in blocking buffer (1:200).

    Antibodies against retinal cell markers (RBPMS for RGCs; Chx10 for bipolar cells; GFAP for astrocytes; Sox2 for Muller glia; rhodopsin for rods; cone arrestin for cones) can be used to further confirm the specific cell types transfected by PNPs. The host of an anti-retinal cell primary antibody should be carefully selected not to overlap with that of the anti-ZsGreen1 antibody (i.e., rabbit) to avoid the second antibody cross-reactivity.

13. Incubate the slide at 4°C for 12 hours in a humid chamber to minimize the evaporation of the buffer.

14. Wash the slide by replacing the buffer with 100 μL of PBS for 1 minute and repeat the washing 3 times.

15. Incubate the slides in 100 μL of the secondary antibody, Goat Anti-Rabbit IgG H&L (Alexa Fluor^® 647), diluted in blocking buffer (1:400) at room temperature for 1 hour.

    To stain the retina tissue with multiple primary antibodies, the secondary antibody should be host-specifically matched with the primary antibody to avoid cross-reactivity between the antibodies. In addition, the fluorescence of secondary antibodies should be carefully selected to avoid conflicts in excitation and emission wavelengths.

16. Wash the slide by replacing the buffer with 100 μL of PBS and repeat the washing 3 times.

17. Mount the slide with two drops of ProLong^™ DAPI solution.

18. Cover the slide with a cover glass without creating air bubbles.

19. Perform confocal microscopy to detect the reporter transgene expression in the retinal layer (Figure 7B).

    The reporter ZsGreen1 protein has the excitation and emission peak wavelengths of 488 and 509 nm, respectively.

Background Information

The approval of the first-ever retinal gene therapy product by the FDA has spurred efforts to expand the use of this state-of-the-art therapeutic modality to numerous inherited and acquired ocular diseases. Although the subretinal injection of AAV-based gene therapy constitutes the mainstay delivery strategy at this moment, accumulated evidence highlights the necessity of exploring alternative delivery platforms and administration routes. In particular, synthetic nucleic acid delivery platforms, such as PNPs, have emerged as a cost-effective alternative that can circumvent several limitations inherent to virus-based platforms to potentially accommodate ever-increasing needs. Albeit not for ocular applications, PNPs based on PEI (Buscail et al., 2015; H. Yin et al., 2014) and PLL (Konstan et al., 2004) have shown tolerability in humans, and lipid nanoparticles have been approved for clinical use over the past recent years (Hou et al., 2021; Kulkarni, Cullis, & van der Meel, 2018). As such, an increasing number of preclinical studies evaluating synthetic nucleic acid delivery platforms for ocular applications are being reported (Ryals et al., 2020; Shen et al., 2020; Y. H. Wang, Rajala, & Rajala, 2018; Zulliger et al., 2018).

The majority of systemically administered gene delivery vectors are readily accumulated in the mononuclear phagocyte system, such as liver and spleen, prior to reaching a target organ (Y. N. Zhang, Poon, Tavares, McGilvray, & Chan, 2016). Further, tight junctions of retinal capillary endothelium and retinal pigmental epithelium form the inner and the outer blood-retinal barriers, respectively, which restrict an access of systemically administered gene delivery vectors to the eye (Campbell & Humphries, 2012; Del Amo et al., 2017). Thus, preclinical and clinical ocular gene delivery and therapy studies primarily employ localized administration modalities, particularly, the subretinal route that enables direct access to the retinal layer. Recently, recognition of drawbacks to subretinal injection in clinical studies has broadened interest in an alternative administration method, intravitreal injection, which is being increasingly investigated in preclinical and clinical settings (Duncan et al., 2018; Ochakovski, Bartz-Schmidt, & Fischer, 2017; Ross & Ofri, 2021). However, we note that there are distinct challenges to PNPs and other gene delivery platforms for the successful implementation of intravitreal retinal gene transfer (Peeters et al., 2005; Peynshaert, Devoldere, Minnaert, De Smedt, & Remaut, 2019). First, intravitreally administered PNPs must penetrate the vitreous gel to distribute radially away from the site of administration and reach the retinal surface. The vitreous gel is a biological hydrogel matrix previously shown to permit unhindered diffusion of non-adhesive nanoparticles as large as 500 nm in diameter (Xu et al., 2013). Thus, the vitreous gel unlikely poses a significant steric/physical gene delivery barrier given that PNPs are generally far smaller (<100 nm). However, the primary building blocks of the vitreous gel are adhesive macromolecules, such as negatively charged glycosaminoglycans (Kasdorf, Arends, & Lieleg, 2015) and hydrophobic collagens (Le Goff & Bishop, 2008), which can immobilize PNPs irrespective of size via electrostatic and/or hydrophobic interactions. The PNPs that manage to traverse the vitreous gel and reach the retinal surface encounter the ILM, an interfacial barrier that constitutes a boundary between vitreous gel and retina (Halfter, Dong, Dong, Eller, & Nischt, 2008). The ILM is a supramolecular network composed of various basement membrane constituents, including collagen, laminin, nidogen, and heparan sulfate proteoglycans (Halfter et al., 2008). Recent studies demonstrated that removal of ILM significantly increased the coverage and level of transgene expression mediated intravitreally administered AAV in the eyes of non-human primates (Takahashi et al., 2017; Teo et al., 2018), which underscores the ILM as a critical delivery barrier. Adhesivity of the ILM is well expected from its macromolecular constituents. Although information regarding the size cut-off for ILM penetration is very limited, it has been shown that negatively charged polystyrene nanoparticles possessing the diameters of 40 nm, but not 100 and 200 nm, are capable of penetrating the ILM in a bovine vitreoretinal explant model (Peynshaert et al., 2017; Teo et al., 2018). The finding suggests that the mesh spacing of ILM is markedly tighter than that of the vitreous gel. To this end, penetration of PNPs through the ILM can be restricted by both adhesive interactions and steric hindrance. Overall, it is critical that PNPs are designed to traverse both vitreous gel and ILM while possessing the ability to transfect retinal cells.

Critical Parameters

Polymer selection and/or design

PNPs are generally formed by multivalent interactions of positively charged nitrogen species (e.g., amines) of cationic polymers with negatively charged phosphates of the nucleic acid backbone. An excess of polymers with high positive charge content are used to ensure full and stable compaction of DNA, yielding positively charged nanoparticles. However, the high charge content in carrier polymers can cause cytotoxic effects on cells by disrupting plasma and lysosomal membranes (Lv, Zhang, Wang, Cui, & Yan, 2006; Monnery et al., 2017; Nel et al., 2009) and/or compromising mitochondrial metabolic activity (Bhattacharjee et al., 2010). Thus, environmentally sensitive polymers (e.g., biodegradable and bioreducible polymers) that break into small pieces in vivo, have been developed and validated for enhanced safety profiles (Chen et al., 2020). Of note, bioreducible polymers are preferentially degraded in reducing environments, such as inside the cell, thereby providing an additional benefit of facilitating the intracellular release of nucleic acid payloads (Kim, Ou, Bull, & Kim, 2010; Lee & Kim, 2014). As described above, both vitreous gel and ILM are rich in negatively charged macromolecules and thus restrict the diffusion of positively charged PNPs through these barriers (Figure 5) and (Xu et al., 2013)). To address this limitation, PNPs can be prepared to possess surface coatings of polyethylene glycol (PEG). PEG is a hydrophilic and neutrally charged polymer that minimizes nonspecific adhesive interactions of nanoparticles with an array of biological entities (Suk, Xu, Kim, Hanes, & Ensign, 2016). For example, PEGylation has been shown to facilitate the penetration of otherwise positively charged PNPs through various biological barriers, including airway mucus (Mastorakos et al., 2015; Osman et al., 2018; Suk et al., 2014), brain extracellular matrix (K. Negron et al., 2019; Karina Negron et al., 2020) and tumor tissue (Mastorakos et al., 2017; K. Negron et al., 2019; Karina Negron et al., 2020). Likewise, we demonstrate here that PEGylated PNPs are capable of efficiently penetrating rabbit vitreous gel (Figure 5).

Selection of experimental models

Most of our knowledge of ocular gene transfer is derived from rodent studies due to several practical aspects, including a short life cycle, easy maintenance, and the availability of various disease models, but the ultimate therapeutic target for clinical translation is human patients. Differences in the anatomical and molecular compositions between vitreous gel and ILM of human and rodent eyes can undermine the translation of results from rodent models into human subjects (K. Y. Zhang & Johnson, 2021b). To this end, rodent model studies should be complemented by studies in models that closely emulate the characteristics of these delivery barriers in human eyes. Encouragingly, human and rabbit vitreous gels are comparable at anatomical, histological, and biomechanical levels (Los, 2008; Luo et al., 2022), and thus rabbit vitreous gel can serve as a reliable human surrogate for investigating the ability of PNPs to overcome this barrier. The data necessary to answer which animals exhibit morphological and compositional similarities of the ILM with humans are very limited. Many laboratory animals, including rodents and rabbits, resemble in ILM thickness and structure to the fetal human ILM; the thickness of human fetal ILM has been estimated by TEM to be < 100 nm (Peynshaert et al., 2019). However, human ILM thickens gradually and substantially from the equatorial toward posterior zone as well as over age, up to a few microns (Candiello, Cole, & Halfter, 2010; Heegaard, 1994; Henrich et al., 2012; K. Y. Zhang & Johnson, 2021a). Likewise, non-human primate (NHP) exhibits the regional and age-related variation in the ILM thickness (Peynshaert et al., 2019), with the thickness falling in the range of human ILM thickness (Matsumoto, Blanks, & Ryan, 1984; Peynshaert et al., 2019). Of note, while intravitreally administered AAV yields widespread transduction of retinal neurons in rodents, similar applications in NHPs (i.e., rhesus macaque) result in the transduction of only a narrow region surrounding the fovea (Ramachandran et al., 2017; L. Yin et al., 2011). To this end, NHPs are arguably the most relevant preclinical model when it comes to intravitreal retinal gene transfer applications, but experimentation with NHPs entails a large amount of time and cost. The human vitreoretinal explant model (Figure 6) may serve as an attractive alternative that enables prompt and cost-effective assessment of the ability of PNPs to traverse the human ILM. Overall, it is critical to carefully select and use complementary experimental models that mimic the barriers to clinical efficacy that will be encountered in human patients with high fidelity.

Figure 6. Schematic for assessing the PNP penetration and transfection in human vitreoretinal explant culture.

The explant culture is prepared from a post-mortem human eye and is treated by PNPs on the vitreous side to evaluate the penetration into and transfection in the retinal tissue. ILM: inner limiting membrane.

Troubleshooting

Table 1 lists potential problems that may arise while implementing the protocols along with their possible causes and solutions.

Table 1.

Troubleshooting guide

Problem	Possible Cause	Solution
Inconsistent DLS measurement	The low particle count rate of PNPs	Increase the particle concentration
	Suboptimal DNA compaction by carrier polymers	Optimize the polymer-to-DNA weight ratio
High in vitro cytotoxicity	High dose of PNPs	Conduct cell viability/cytotoxicity test with an optimal dose of PNPs
	Presence of high-level residual free carrier polymers	Wash the PNP solution with Amicon® Ultra-4 centrifugal filter unit to remove the free polymers
High well-to-well variability in in vitro transfection efficacy within experimental groups	Inconsistent cell seeding density	Frequently and gently pipet the cell medium to avoid settling of cells
	Use of the outer wells of the multi-well plate	Use the inner wells exhibiting similar and mild medium evaporation
Poor signal-to-noise ratio in MPT studies	Low fluorescent labeling density of DNA	Enhance labeling density by increasing the incubation time and/or the concentration of fluorophores
	Photobleaching of fluorescence	Reduce the exposure time in each frame Consider the use of a fluorophore with enhanced quantum yield
Difficulty in focusing PNPs in the vitreous gel	Failure to locate the middle of the vitreous gel sample	Focus on the interface of cover glass and vitreous gel and gradually move up
	Low concentration of PNPs	Add PNPs with a higher concentration
Degeneration of cultured retinal tissue	Prolonged the period from death to retinal explanation	Utilize eyes that can be harvested with a minimal time from death to dissection
	Inadequate dissection technique	Perform retinal dissections quickly Limit direct contact with retinal tissue Maintain retinal tissue in media and transfer within fluid by pipette
Damage to an eyeball during the enucleation	Reduced mechanical property of the punctured/injected eyeball	Use curved scissors and forceps to harvest the eyeball at the deep bottom
Infection or inflammation after ocular injection	Contamination of materials or tools	Check the endotoxin levels in PNPs and ingredients and sterilize all tools
	Accidental penetration or contact with the crystalline lens	Perform intravitreal injections carefully to avoid the lens, under direct visualization of the retina through the pupil where feasible
Poor preservation of histological tissue	Insufficient fixing	Incubate at room temperature for another 30 minutes
Low fluorescent intensity on confocal imaging	Low transgene expression in the tissue	Increase the dosage of PNPs for transfection
	Low concentrations of primary and/or secondary antibodies	Increase/optimize the concentration

Understanding Results

This protocol introduces a series of methods to engineer PNPs, characterize physicochemical and other essential properties as a gene delivery platform, and evaluate their delivery performances using in vitro, ex vivo, and in vivo assays highly relevant to intravitreal gene transfer to the retina.

The formulation of PNPs is achieved by a dropwise mixing of a diluted plasmid DNA solution to a concentrated polymer solution. Screening at a range of polymer-to-DNA weight ratios is the first step to establish a condition that allows complete compaction of DNA by a carrier polymer of interest (Figure 3A). PNPs are then characterized for physicochemical properties, including morphology, size, and surface charge (Figures 3B–D), which collectively impact on the performance of PNPs. To confirm that nucleic acid payloads are protected against enzymatic activity expected in vivo, PNPs are incubated with DNase and then payloads are released by heparin following the deactivation of DNase with EDTA. The DNA protected by PNPs appears at the location of the agarose gel expected for the intact unprocessed DNA, as revealed by the gel electrophoretic migration assay (Figure 3E). The ability of PNPs to transfect human retinal cells (e.g., ARPE-19) in vitro can be evaluated by two complementary methods as described above. The first method involves the transfection of cells with PNPs carrying plasmid DNA expressing a reporter fluorescent protein (e.g., ZsGreen1), and the percentage of the transfected cells and the average level of transgene expression (e.g., fluorescent intensity) are determined by flow cytometry (Figure 4). The second method entails the transfection of cells with PNPs carrying luciferase-expressing plasmid DNA and measuring the luciferase activity from the cell lysate using a luminometer. The subsequent steps evaluate the ability of PNPs to overcome key extracellular barriers present en route to the retina following intravitreal administration, vitreous gel and ILM, using ex vivo models. The ability of PNPs to penetrate the former barrier is assessed by quantifying the diffusion rates of fluorescently labeled PNPs in vitreous gel collected from an animal or a human cadaver eye using the MPT technique (Figure 5). In parallel, the culture of human vitreoretinal explants treated with fluorescently labeled PNPs is subjected to confocal microscopy to visualize PNPs manage to penetrate the ILM and transfect retinal cells (Figure 6). Finally, PNPs carrying plasmid DNA expressing either luciferase or a fluorescent protein (e.g., ZsGreen1) are intravitreally administered into the eyes of inbred mice (e.g., C57BL/6) and in vivo reporter transgene expression is assessed by live animal imaging, tissue homogenate-based luciferase assay and/or confocal microscopy (Figure 7).

Time Considerations

Formulation and characterization of PNPs take 3 hours, including DLS and TEM, while the time could vary depending on the number of PNP types being processed simultaneously. The PNP protection assay takes 2 hours including DNase incubation and deactivation, heparin-mediated DNA decompaction, and agarose gel electrophoresis. The evaluation of In vitro transfection requires at least 3 days, including overnight cell incubation to reach desired confluency, 2-day transfection period, and evaluation of transgene expression. The quantification of diffusion rates of PNPs in vitreous gel necessitates at least one day to accommodate sample preparation, movie recording, and MPT analysis. Ex vivo ILM penetration study takes tightly up to 2 – 3 days to prepare the vitreoretinal explant culture, treat the culture with PNPs, and evaluate the reporter transgene expression but penetration of fluorescently labeled PNPs across the ILM can be determined on the same day of the culture preparation. The ability of intravitreally administered PNPs to transfect retinal cells in vivo can be determined 2 – 3 days, from intravitreal injection to the assessment of transfection, if the transgene expression is driven by the cytomegalovirus promoter which generally peaks at 48-hour post-administration. However, the study period can be prolonged if plasmid DNA controlled by a eukaryotic promoter is used or the goal of the study is to monitor transgene expression kinetics over time.

DATA AVAILABILITY STATEMENT:

The data, tools, and materials (or their source) that support the protocol are available from the corresponding author upon reasonable request.