Subretinal injection in mice to study retinal physiology and disease

Abstract

Subretinal injection (SRI) is a widely used technique in retinal research and can be used to deliver nucleic acids, small molecules, macromolecules, viruses, cells or biomaterials such as nanobeads. Here we describe how to undertake SRI of mice. This protocol was adapted from a technique initially described for larger animals. Although SRI is a common procedure in eye research laboratories, there is no published guidance on the best practices for determining what constitutes a 'successful' SRI. Optimal injections are required for reproducibility of the procedure and, when carried out suboptimally, can lead to erroneous conclusions. To address this issue, we propose a standardized protocol for SRI with 'procedure success' defined by follow-up examination of the retina and the retinal pigmented epithelium rather than solely via intraoperative endpoints. This protocol takes 7-14 d to complete, depending on the reagent delivered. We have found, by instituting a standardized training program, that trained ophthalmologists achieve reliable proficiency in this technique after ~350 practice injections. This technique can be used to gain insights into retinal physiology and disease pathogenesis and to test the efficacy of experimental compounds in the retina or retinal pigmented epithelium.

Extended Data Fig. 1 ∣. Custom needle parameters.

Detailed illustration of the needles designed for subretinal injection in rat (top) and mice (bottom).

Fig. 1 ∣. Applications of SRI in rodent models.

SRI has been used in various applications, e.g., to deliver reagents to create disease models, for gene and cell replacement therapies. A2E, N-retinylidene-N-retinylethanolamine; DR, diabetic retinopathy; LCA, Leber Congenital Amaurosis; PEC, macrophage-rich peritoneal exudate cell.

Fig. 2 ∣. Depictions of three SRI approaches.

a, The corneal approach. b, The transscleral posterior approach. c, The pars plana approach.

Fig. 3 ∣. Representative images of a proper and improper SRI procedure to study the RPE phenotype.

a, Clear view of the fundus shown as seen through the surgical microscope after mounting the ocular ring on the eye and adding the lubricant gel. After adjusting the focus under the surgical microscope, we can clearly see the optic nerve located in the center of the field. Scale bar, 1 mm. b, Ideal postinjection fundus as visualized under the surgical microscope after SRI of 0.6 μl PBS. The blue triangle indicates the injection site. Scale bar, 1 mm. c, Extensive bleb (retinal detachment) as visualized under the surgical microscope induced by improper SRI of 0.6 μl PBS. Scale bar, 1 mm. d–f, The same eye as that observed in b; fundus photography (d) and OCT (e) showed no RPE degeneration (scale bar, 100 μm); confocal images of RPE flat mounts stained for ZO-1 also showed normal cellular architecture (scale bar, 25 μm) (f). g–i, The same eye as that observed in c; fundus photography (g) and OCT (h) showed hypopigmentation and subretinal vacuolization on day 7 (D7) (scale bar, 100 μm); confocal images of RPE flat mounts staining for ZO-1 showed disrupted cellular architecture (scale bar, 25 μm) (i).

Fig. 4 ∣. Overview of the procedure.

The general protocol for SRI involves reagent and animal preparation (Steps 1–7) followed by SRI injection (Steps 8–20), additional experimental treatment (Step 21) or use of transgenic mice, fundus photo examination (Steps 22–23), euthanasia and enucleation (Steps 24–25), sample preparation (Steps 26–34) and confocal imaging (Steps 35–36). IVT, intravitreous; I.p., intraperitoneal; I.v., intravenous.

Fig. 5 ∣. Representative images of a successful SRI.

a,b, Representative fundus photo (left) and flat mounts stained for ZO-1 (red, right). The pictures depict the classical phenotypic characteristics of mice administered with SRI of PBS (a, no degeneration) or Alu RNA (b, note focal hypopigmentation). Blue triangle, injection site. Scale bars, 20 μm. c,d, Representative fundus photo (left) and fluorescent micrographs of 10-μm-thick retinal sections labeled with DAPI and anti-Cre in C57BL/6J mice after injection of AAV2-hRPE(0.8)-iCre-WPRE (c) or AAV2-CMV-Null (d). Note the Cre expression 10 d after injection shown by white arrowheads (c). Scale bars, 100 μm. WPRE, woodchuck hepatitis posttranscriptional regulatory element.

Fig. 6 ∣. Identification of the injection site in mouse RPE flat mounts.

a,b, Fundus photographs (left) and RPE flat mounts stained for ZO-1 (red) at lower (10× and 20×) and higher magnifications (60×) after SRI of PBS (a) or AβOs (b). The injection site is identified by characteristic stellate pattern in the RPE (white arrow), which corresponds to the region where the needle touches the RPE. The surrounding area is examined, and higher-magnification images are acquired and analyzed. Scale bars, 10× (200 μm), 20× (100 μm) and 60× (50 μm). AβO, amyloid-beta oligomers. Figure reproduced with permission from ref. .

Fig. 7 ∣. Follow-up of successful and improper SRI of PBS.

a, After successful SRI of PBS, the retina reattaches as early as day 1 as visualized on OCT (left). On day 7, fundus photography showed no RPE degeneration, and OCT and confocal images of ZO-1-immunostained RPE flat mounts (red, right) at lower (10× and 20×) and higher magnification (60×) of the same eye showed normal RPE cellular architecture. b, After improper SRI, a bullous retinal detachment was observed intraoperatively. OCT showed RPE and outer retinal changes as early as day 1. On day 7, fundus photos showed areas of hypopigmentation, and OCT and confocal images of ZO-1-immunostained RPE flat mounts showed distorted RPE cellular architecture. c, An improper injection technique could also cause injection-related RPE degeneration, even flattening of the bleb. In this case, the injection needle has entered the subretinal space. Blue triangles indicate the site of the injection. Scale bars, 10× and 20× (100 μm), 60× (50 μm) for ZO-1 and 100 μm for OCT scans.