The technique, called nano-enhanced optical delivery (NOD), sidesteps the viral vectors that dominate current ophthalmic gene therapy. Adeno-associated viruses (AAVs), the workhorses of retinal gene transfer, carry a payload limit of roughly 4.7 kilobases, which restricts the delivery of large therapeutic genes and gene-editing tools such as CRISPR/Cas9. Viral vectors can also provoke inflammatory and immunological reactions, carry theoretical risks of genomic integration, and offer essentially no spatial control over where the therapeutic gene is expressed once administered. For diseases characterized by localized degeneration, such as geographic atrophy in advanced AMD, the inability to confine gene expression to the affected retinal regions is a significant drawback.

NOD addresses these limitations with a physical mechanism. The researchers functionalized gold nanorods with 5-kilodalton carboxyl-terminated polyethylene glycol to minimize nonspecific binding and prevent aggregation in physiological solutions. Transmission electron microscopy confirmed rod-shaped particles averaging approximately 120 nanometers in length and 25 nanometers in width, with a zeta potential of roughly −28 millivolts at pH 7. Ultraviolet–visible–near-infrared spectroscopy revealed a strong longitudinal plasmonic absorption band in the near-infrared region overlapping the 980-nanometer wavelength of the delivery laser. When irradiated at this wavelength, the nanorods convert optical energy into localized thermal energy, generating hotspots at the cell surface that transiently permeabilize the plasma membrane and allow extracellular plasmid DNA to enter the cell.

The therapeutic payload in this study was the MCO-020 plasmid, which encodes a multi-characteristic opsin designed to confer ambient-light sensitivity to retinal neurons. The expression cassette uses a CAG promoter driving the MCO-020 coding sequence, followed by a human growth hormone polyadenylation signal. The researchers first optimized delivery parameters in HEK293 cells, where immunostaining for the mCherry reporter protein revealed that transgene expression occurred only in cells exposed to laser irradiation. Cells incubated with plasmid and nanorods in the absence of laser exposure showed no detectable reporter expression, confirming that laser activation is required for plasmid entry. Increasing laser power from 10 to 20 milliwatts produced substantially stronger reporter fluorescence, and longer irradiation durations of 400 seconds yielded higher expression than 200 seconds. Adjacent non-irradiated areas within the same culture showed negligible signal, demonstrating that gene delivery was confined to the laser-targeted area.

To integrate this delivery mechanism with real-time anatomical guidance, the team combined a spectral-domain OCT subsystem centered at 840 nanometers with the 980-nanometer delivery laser through a fiber coupler, so that both beams share a common optical path toward the retina. The combined beam is collimated and directed toward the eye using micro-electromechanical system scanning mirrors and telescopic lenses. Fourier transformation of the detected interference spectrum produces depth-resolved structural images of the retina. Using en-face OCT imaging, the targeted region of interest is identified and marked on the imaging interface, and retinal features such as blood vessels serve as reference landmarks for tracking the position of the targeted region throughout the procedure. This feedback system enabled dynamic adjustment of the delivery beam to maintain precise alignment with the selected retinal area.

In vivo experiments were conducted in rd1 mice, which carry a mutation in the Pde6b gene encoding the β-subunit of rod photoreceptor phosphodiesterase. This mutation causes rapid degeneration of rod photoreceptors followed by secondary cone loss, producing severe visual impairment while preserving inner retinal neurons. The researchers injected 1 microliter of MCO-020 plasmid mixed with gold nanorods into the vitreous cavity of one eye, then positioned the animal under the OCT-guided delivery system approximately 2 hours later. A 980-nanometer continuous-wave near-infrared laser with an approximately 15-micrometer Gaussian focal spot was raster scanned over an approximately 1 × 1 millimeter retinal area, with laser powers of 10 to 30 milliwatts and total irradiation durations of 200 and 400 seconds evaluated.

OCT imaging performed immediately after laser irradiation revealed no detectable disruption of retinal structure or evidence of additional tissue damage within the targeted region. Retinal thickness measurements obtained from B-scan images showed no statistically significant differences between baseline and post-treatment time points. Confocal microscopy demonstrated robust MCO-020 expression, visualized through mCherry immunostaining, in retinal cells located within the laser-targeted region. Quantitative analysis showed that the number of MCO-expressing cells increased with increasing laser power, indicating that higher power produced greater transfection efficiency. Longitudinal quantification revealed persistent transgene expression for several months following delivery, with detectable expression remaining evident at the 12-month partial dataset point.

Electrophysiological recordings confirmed that the delivered opsin was functionally active. Visually evoked potential (VEP) recordings from the visual cortex showed dose-dependent increases following MCO-020 delivery, with treated mice exhibiting mean cortical responses of approximately 15 microvolts compared with roughly 6 microvolts in untreated controls, a statistically significant difference. In animals with strong retinal transgene expression, VEP responses were clearly detectable in the visual cortex, confirming successful transmission of visual signals through the retinal circuitry. These cortical responses persisted for up to 52 weeks following treatment, indicating long-term functional activity of the optogenetically modified retinal neurons. Intraocular pressure measurements remained comparable to control eyes throughout the study period, and retinal cell viability assays showed no significant differences between untreated controls, laser-only, nanorod-only, and combined treatment groups.

Biodistribution analysis performed at 12 months showed MCO-020 vector DNA below the limit of detection of 50 copies per nanogram in all evaluated non-target tissues, including brain, gonads, kidney, mesentery, spleen, heart, lungs, and liver. This finding supports limited systemic distribution following ocular delivery. In a separate focal photoreceptor injury model generated in wild-type mice using a 532-nanometer green laser, OCT en-face imaging confirmed localized retinal damage, and microfocal ERG recordings obtained from the damaged area showed markedly reduced responses prior to treatment. Following OCT-guided NOD delivery of MCO-020, significant recovery of focal ERG responses was observed within 1 week, demonstrating restoration of localized retinal activity in an experimentally defined region of injury.

The authors note several limitations. Direct retinal temperature measurements and nanorod-specific thermal modeling were not performed, so the reported exposure calculations characterize the delivered optical dose but do not establish the local temperature rise at the nanorod–cell interface. The full therapeutic window was not defined by escalation from ineffective exposure to the threshold for tissue injury, and published retinal laser-damage thresholds may not directly predict the NOD threshold because plasmonic nanorods locally enhance optical absorption and heat generation. The rd1 model represents inherited retinal degeneration rather than age-related macular degeneration and does not reproduce the chronic RPE dysfunction, inflammatory processes, or macular anatomy associated with human geographic atrophy. Nevertheless, the absence of detectable changes in retinal morphology, thickness, cell viability, intraocular pressure, and electrophysiological function supports tolerability within the tested operating range. The researchers suggest that the combination of OCT guidance and nano-enhanced optical delivery could serve as a platform for localized delivery of therapeutic genes, including gene-editing tools that exceed viral packaging limits, to retinal regions affected by degeneration.

Subject of Research: Optogenetic gene delivery

Article Title: Optical coherence tomography guided targeted non-viral optical delivery of opsin-encoding genes into the retina

Article References: Kim, S., Carlson, M., Batabyal, S., Dibas, A., & Mohanty, S. (2026). Optical coherence tomography guided targeted non-viral optical delivery of opsin-encoding genes into the retina. Frontiers in Pharmacology, 17, Article 1831339. https://doi.org/10.3389/fphar.2026.1831339

Image Credits: AI Generated

DOI: 10.3389/fphar.2026.1831339

Keywords: gene therapy, optogenetics, retinal degeneration, non-viral delivery, optical coherence tomography, gold nanorods, MCO-020, Optical, coherence, tomography, guided, targeted