Lipid nanoparticles enable mRNA delivery to diverse cell types of the inner Retina.

Scientists have found that a common way to deliver genetic medicine—lipid nanoparticles (LNPs)—can reach more types of eye cells than previously thought. This includes neurons and blood vessel cells. This discovery is critical. It suggests that existing delivery technologies might be capable of treating a much wider array of eye diseases than currently assumed.

The limitations of viral delivery

For years, the gold standard for retinal gene therapy has been adeno-associated viruses (AAVs). These viral vectors are highly efficient at entering cells. However, they act like small envelopes with strict weight limits. Their limited packaging capacity makes them unsuitable for delivering large genes. They are also unideal for complex systems like CRISPR gene editors. Furthermore, AAVs cause persistent transgene expression. This is not always desirable when a temporary pulse of genetic instruction is needed.

Lipid nanoparticles (LNPs) have emerged as a powerful non-viral alternative. Much like the mRNA vaccines used against SARS-CoV-2, LNPs encapsulate messenger RNA (mRNA) to provide transient instructions to cells. While LNPs offer a much higher payload capacity than AAVs, their utility in the eye has been constrained by a lack of clarity regarding their "tropism." Tropism refers to the specific tendency of a delivery vehicle to target certain cell types. Historically, researchers believed that in the inner retina, LNPs primarily targeted Müller glia (support cells) or phagocytes (cells that engulf debris). This narrow perceived scope suggested that LNPs might struggle to reach the neurons or vascular cells essential for many retinal disorders.

Mapping the LNP landscape

To move beyond these assumptions, the authors systematically assessed how conventional LNPs distribute genetic material across various retinal environments. They utilized a dual-pronged approach. First, they examined dissociated retinal cells to remove all physical barriers. This allowed them to see where the particles could go. Second, they used murine retinal explants (slices of tissue maintained in a lab) to see where they actually go in a structured environment.

The researchers used a specialized setup involving semi-permeable PTFE membrane inserts. This allowed them to mimic two distinct clinical administration routes: 1. Quasi-intravitreal delivery: Applying LNPs to the side of the retina containing the ganglion cells. This simulates an injection into the vitreous humor (the clear gel filling the eye). 2. Quasi-subretinal delivery: Applying LNPs to the outer side of the retina. This simulates an injection into the space between the photoreceptors and the retinal pigment epithelium.

By using chemically modified mRNA encoding a fluorescent protein called mCherry, the team could visually track successful "transfection." Transfection is the process by which a cell takes up and expresses the delivered genetic material.

Broader tropism and delivery hurdles

The study reveals that LNP capability is significantly more expansive than previously documented. In the absence of anatomical barriers, the authors report that LNPs successfully transfected not only Müller glia and macrophages but also astrocytes and various neurons .

Even in intact tissue, the authors found that LNPs reached vascular structures, such as pericytes or endothelial cells .

They also reached neurons, including cells with morphologies characteristic of horizontal cells .

However, the effectiveness of this delivery is heavily dictated by the physical architecture of the eye. The authors find that subretinal delivery is consistently more efficient than intravitreal administration .

A primary culprit for this disparity is the inner limiting membrane (ILM). The ILM is a basement membrane that acts as a structural barrier to incoming particles. The paper demonstrates that in healthy retinas, the ILM significantly hinders intravitreal transfection. Notably, when the researchers mechanically removed the ILM (a process called "peeling"), the transfection efficiency following intravitreal delivery increased significantly .

Furthermore, the study notes that the state of the retina itself changes the math. The authors report that transfection efficiency was higher in degenerate (diseased) retinas compared to healthy ones . This might be due to the breakdown of structural barriers like the outer limiting membrane during disease progression.

Assessing the cost of delivery

Every powerful intervention carries a biological cost. A sensitive indicator of retinal stress is the activation of Müller glia. This manifests as an increase in the expression of Glial Fibrillary Acidic Protein (GFAP). The authors report that LNP administration does trigger this stress response. This is particularly true in degenerate retinas and following subretinal delivery .

While the researchers observed that intravitreal delivery to healthy retinas caused minimal stress, removing the ILM changed the outcome. The act of peeling the ILM to improve delivery efficiency led to a measurable increase in GFAP signaling. This highlights a fundamental engineering tradeoff in ocular therapeutics. The very barriers that protect the retina from external stressors are the same obstacles that prevent efficient drug delivery.

The verdict on LNP versatility

Is LNP technology ready to replace AAVs for inner retinal disease? The answer is: probably, but with refinement.

The study provides strong evidence that conventional LNPs are not restricted to a single cell type. This makes them viable candidates for treating disorders like Congenital Stationary Night Blindness (CSNB). In CSNB, targeting specific neurons is required. Because LNPs can carry much larger payloads and offer transient expression, they solve several "hardware" limitations inherent to viral vectors.

However, practitioners should not expect immediate success. The authors note that while neurons can be transfected, certain critical types—like bipolar cells—remained rare in their observations. Achieving high-precision, cell-type-specific targeting will require further tuning of LNP formulations. For now, the work establishes that the bottleneck for LNP therapy is not necessarily the chemistry of the particles. Instead, it is the physical navigation of the eye's complex anatomical barriers.

Figures from the paper