lecture 19 + 20 - therapeutic approaches to reversing sensory loss:

Why is the eye a good target for cellular therapy, and what is the role of the RPE?

• The eye is accessible and visible, allowing cells to be delivered precisely to the retina and monitored.

• It is immune privileged because of the blood-retina barrier.

• It is enclosed, limiting the spread of complications, and one eye can potentially act as a control for the other.

• The macula provides central, high-acuity vision; its centre, the fovea, has a very high density of cones.

• Retinal pigment epithelial (RPE) cells form a monolayer beneath photoreceptors and are essential for their survival.

• RPE functions include nutrient/ion/water transport, absorbing stray light, recycling retinal, phagocytosing photoreceptor debris, and secreting signalling factors.

• RPE also contributes to the blood-retina barrier.

What is age-related macular degeneration (AMD), and what happens in dry vs wet AMD?

• AMD causes progressive loss of central vision, affecting activities such as reading, recognising faces and driving; distorted/wavy lines on an Amsler grid can be an early sign.

• Dry AMD (~90%): accumulation of drusen (fat/protein deposits) → RPE degeneration → photoreceptor death → gradual geographic atrophy and vision loss. No established treatment was discussed for this form.

• Wet AMD (~10%): abnormal blood vessels grow from the choriocapillaries through the RPE (neovascularisation) → leakage/bleeding → scarring and rapid vision loss. Anti-VEGF injections prevent further abnormal vessel growth but do not replace the damaged RPE.

• In both forms, RPE degeneration ultimately causes photoreceptor loss.

What evidence showed that replacing RPE cells could treat AMD?

• Experimental surgeries showed that RPE replacement can restore vision.

• Macular translocation: the retina is physically moved so the macula sits on healthier RPE → about a 5-line improvement in vision in some patients.

• Autologous RPE transplantation: healthy RPE from peripheral retina is transplanted beneath the macula → also produced about a 5-line visual improvement, with one patient regaining their driving licence.

• However, these surgeries are long, complex and high-risk, and mainly suitable for late-stage disease.

How were functional RPE cells produced from human embryonic stem cells, and how was their function tested?

• Human embryonic stem cells are self-renewing and pluripotent, meaning they can be expanded and differentiated into different cell types.

• Cells from the inner cell mass of a blastocyst were expanded in culture; removing factors that maintain pluripotency allowed them to differentiate.

• RPE cells were identified by their pigmentation, isolated and grown into a cobblestone-like monolayer resembling normal human RPE.

• Their identity was confirmed using gene expression (PCR), protein markers and morphology.

• Function was tested by showing that the RPE could phagocytose photoreceptor debris.

How was the RPE therapy delivered as a transplantable patch, and what did animal studies show?

• Injecting RPE cells alone is problematic because they form a disorganised, immature and non-polarised mass rather than the normal monolayer.

• A porous polyester track etched membrane was therefore used as a scaffold, allowing RPE cells to form a mature, polarised monolayer while permitting transport of nutrients, ions and water.

• The surgeon creates an opening, lifts the retina with saline, and inserts the RPE-coated patch beneath the macula, where it is held in place after the saline is removed.

• RPE on the patch maintained the photoreceptor layer, whereas the patch alone caused photoreceptor loss because it separated the photoreceptors from their RPE.

• Thus, the experiments showed that the RPE cells—not simply the patch—are essential for maintaining photoreceptors.

What steps were needed before the RPE patch could be tested in humans?

• Safety testing: confirmed the differentiated RPE cells did not form teratomas; undifferentiated embryonic stem cells did.

• GMP manufacturing: cells were differentiated, expanded and grown on the membrane under Good Manufacturing Practice (GMP) conditions.

• The final patch was only 3 × 6 mm, enough to cover the macula/foveal region.

• A specialised surgical delivery tool was developed to roll the patch up, place it under the retina and safely release it.

• MHRA and ethics approval were required before clinical trials; approval took about 4 years.

Why were patients with wet AMD chosen for the first clinical trial?

Wet AMD causes a rapid loss of vision, but photoreceptors can still be alive shortly after a macular haemorrhage.

The patch could therefore replace damaged RPE and preserve the surviving photoreceptors.

Surgery needed to occur within about 6 weeks for the best chance of saving vision.

What were the clinical outcomes of the RPE patch?

• Expected clinically significant improvement = 3-line increase in visual acuity.

• Both initial patients achieved a 6-line improvement.

• Reading improved:

  • Patient 1: 5 → 51 → 84 → 69 words/min at 0, 6 months, 1 year and 2 years.

  • Patient 2: 0 → 26 → 47 → 41 words/min.

• OCT/autofluorescence/microperimetry showed the RPE survived, functioned and supported photoreceptors.

• No serious adverse reactions or teratomas; complications were mainly surgical.

• Vision improved/stabilised over 6 months–2 years, possibly as cells integrated and matured.

What are the limitations and future of the therapy?

• Some deterioration occurred after ~5 years, possibly due to ongoing AMD, immune response or RPE migration.

• The eye is immune privileged, but long-term immune effects involving microglia remain uncertain.

• One patch costs about £4,000 in materials, although overall treatment costs would be higher.

• The therapy targets the underlying RPE loss/dysfunction rather than just AMD symptoms.

• The therapy has been developed further by Tenpoint Therapeutics.

• Next step: larger multicentre clinical trials.

• Overall: RPE patches were shown to survive, integrate, support photoreceptors and improve vision.

How are inherited retinal diseases (IRDs) clinically classified?

• Progressive: worsen over time;

• stationary: do not progress (e.g. congenital stationary night blindness).

• Progressive IRDs can be non-syndromic (retina only) or syndromic (retinal disease + other body systems, e.g. Bardet-Biedl, Usher syndrome).

• Non-syndromic types include:

  • Macular dystrophy: affects the macula.

  • Rod-cone dystrophy: rods affected first → night blindness and tunnel vision.

  • Cone-rod dystrophy: cones affected first → central visual blurring, later spreading.

  • Choroideremia: affects the retina and other structures such as the choroid.

• IRDs show genetic heterogeneity (different genes can cause the same disease) and phenotypic heterogeneity(different diseases can result from variants in the same gene).

• They can be autosomal recessive, autosomal dominant or X-linked.

How do different inheritance patterns affect the mechanism of IRDs?

• Autosomal recessive: two faulty copies of the gene→ usually loss of function.

• Autosomal dominant: one faulty copy of the gene; mechanisms include haploinsufficiency, toxic gain-of-function and dominant-negative effects.

• X-linked: can be recessive or dominant; inheritance differs between males and females because males have only one X chromosome.

• Understanding the specific mutation and inheritance is important because it determines which therapy can be used.

How does phototransduction convert light into an electrical signal?

• Light causes 11-cis retinal → all-trans retinal, activating rhodopsin.

• Activated rhodopsin activates the G-protein transducin → activates cGMP phosphodiesterase.

• Phosphodiesterase converts cGMP → GMP, causing cGMP-gated cation channels to close.

• Ca²⁺/other cation influx decreases → the photoreceptor hyperpolarises.

• This converts light into an electrical signal transmitted to bipolar cells through ribbon synapses.

• Photoreceptors continuously release neurotransmitter in darkness; light reduces neurotransmitter release.

• Their response is graded rather than an action potential, allowing them to encode small changes in light intensity.

Why is the eye a good target for gene therapy, and what therapies can be used?

• The eye is anatomically accessible and compartmentalised, allowing direct delivery and relatively low therapeutic doses.

• The blood-retinal barrier makes the eye relatively immune privileged, reducing systemic immune responses.

• Non-invasive imaging such as fundus imaging and OCT allows treatment and disease progression to be monitored.

• Therapeutic approaches include:

  • Pharmacotherapy: drugs/small molecules; can be gene- or mutation-specific/agnostic.

  • Neuroprotection: protects surviving cells through stress management, proteostasis and microglial maintenance.

  • Optogenetics: introduces light-sensitive genes/optogenes into surviving retinal cells; limited by sensitivity, wavelength and response speed.

  • Retinal prostheses: electrically stimulate surviving ganglion cells; currently produce visual sensations rather than clear vision.

  • Stem-cell therapy: replaces lost retinal cells.

  • Gene therapy: replaces, supplements or modifies defective genes. (RNA or DNA targeted therapies)

How does AAV gene replacement therapy work, and why is it mainly used for recessive IRDs? +advantages and disadvantages?

• Autosomal recessive diseases usually involve loss of function, so supplying a functional gene can restore the missing function.

• AAV (adeno-associated virus) is an inactivated viral vector carrying the therapeutic transgene.

• AAV enters retinal cells → reaches the nucleus → forms a long-lasting DNA structure that expresses the therapeutic gene.

• Self-complementary AAV (scAAV) allows faster gene expression by bypassing the initial conversion to double-stranded DNA.

• AAV capsids can be engineered to target RPE cells or photoreceptors and provide long-lasting expression in non-dividing cells.

• Example: Luxturna replaces defective RPE65 in LCA2; RPE65 is an enzyme in the visual cycle and is expressed in the RPE. It is delivered by subretinal injection.

What is IPL1-associated LCA and why is early treatment important?

• Leber congenital amaurosis (LCA) caused by autosomal recessive loss-of-function variants in IPL1, a photoreceptor-specific gene.

• Congenital/very early onset: severe visual impairment or complete blindness from birth.

• Children may show roving eye movements and inability to fixate.

• There is no electrophysiological retinal response from the earliest measurable age.

• Therapeutic window is approximately 1–5 years; after ~5 years, photoreceptor degeneration is too advanced for gene therapy because sufficient surviving cells are needed.

What does IPL1 do, and how does its loss cause photoreceptor degeneration?

• IPL1 is located from the connecting cilium to the photoreceptor synapse, including rod/cone inner segments and outer nuclear layer.

• It is not present in rod outer segments, where phototransduction occurs.

• IPL1 is required for proper assembly of the phosphodiesterase (PDE) complex.

• Loss of IPL1 → PDE complex cannot assemble → PDE subunits are degraded.

• Without PDE → cyclic GMP cannot be hydrolysed → cyclic GMP-gated channels remain open → continuous Ca²⁺ influx.

• This resembles constant exposure to light, causing toxic, rapid degeneration of rods and cones.

How are patient-derived retinal organoids (ROs) made, and what are their features?

• Renal epithelial cells from urine → reprogrammed into iPSCs(induced pluripotent stem cells) → differentiated into 3D retinal organoids using different media.

• iPSCs form neuro-retinal vesicles, which are isolated and cultured to develop into organoids.

• ROs follow the spatial and temporal development of the human neuro-sensory retina.

• They form:

  • Outer nuclear layer: rods + cones

  • Inner nuclear layer: bipolar, horizontal + amacrine cells

  • Ganglion cell layer

  • Photoreceptor inner/outer segments + connecting cilium

• Limitations: develop inside-out, lack a normal RPE monolayer, causing less organised photoreceptor outer segments; ganglion cells also become sparse with maturity.

• Despite this, ROs provide a human, patient-specific model for studying disease and testing therapies.

How was CRISPR used in IPL1 disease modelling and gene correction?

• CRISPR-Cas9 uses guide RNA to target a specific DNA sequence and Cas9 cuts the DNA.

• NHEJ repairs the break but can create indels → useful for knocking out IPL1 and modelling the disease.

• IPL1 knockout reproduced the patient phenotype: no IPL1 → loss of PDE6 → increased cGMP.

• HDR can introduce or correct specific mutations, but is inefficient and does not work well in post-mitotic photoreceptors.

• CRISPR can also create isogenic controls: genetically identical cells differing only in the mutation, making it easier to prove that the mutation causes the disease.

• Correcting the IPL1 mutation restored IPL1/PDE6 and reduced cGMP.

How was IPL1 gene therapy developed and what did it achieve?

• Because IPL1 is <4.7 kb, it can fit inside an AAV vector.

• AAV delivered the IPL1 coding sequence under a rhodopsin kinase promoter, so IPL1 was produced specifically in photoreceptors.

• In patient organoids, treatment restored IPL1 and PDE6 and reduced the abnormal cGMP accumulation.

• The same approach was then given to children 1–2.8 years old in one eye.

• After ~3.5 years:

  • Treated-eye logMAR improved from ~2.7 → ~0.9 = major visual improvement.

  • Visually evoked brain responses also improved.

  • Untreated eyes showed no improvement and deteriorated.

• Other potential treatments include antisense oligonucleotides (destroy mutant mRNA or correct abnormal splicing) and base/prime editing.

• These can be highly precise, but challenges include delivery, low efficiency, technical complexity, off-target effects, ethical concerns and very high cost.

How can antisense oligonucleotides (ASOs) treat genetic diseases?

• RNA-targeted therapies can treat both autosomal recessive and dominant diseases.

• Some ASOs bind mRNA → form a DNA–RNA hybrid → RNase H1 degrades the mRNA.

  • Useful for dominant diseases where the harmful allele needs to be removed.

• Other ASOs bind a cryptic splice site and act as a steric block, preventing abnormal splicing.

  • Can restore normal splicing when an intronic variant causes an exon to be incorrectly included/excluded.

What are genome-editing therapies and what are the advantages and limitations of precise editing?

• Non-homologous end joining (NHEJ): creates double-strand breaks to remove problematic regions of DNA.

• Base editing: makes precise single-base changes without a double-strand break.

  • Can convert A → G or C → T.

• Prime editing: allows precise correction of a wider range of variants.

• Advantages: highly precise, versatile, potentially treats many variants and has fewer indels/off-target effects than double-strand-break approaches

• Limitations: difficult delivery, low efficiency, technically complex, incomplete editing, off-target effects, ethical concerns and very high cost/accessibility.