New Frontier in the Management of Corneal Dystrophies: Basics, Development, and Challenges in Corneal Gene Therapy and Gene Editing

Abstract

Corneal dystrophies represent a group of heterogeneous hereditary disorders causing progressive corneal opacification and blindness. Current corneal transplant management for corneal dystrophies faces the challenges of repeated treatments, complex surgical procedures, shortage of appropriate donor cornea, and, more importantly, graft rejection. Genetic medicine could be an alternative treatment regime to overcome such challenges. Cornea carries promising scope for a gene-based therapy involving gene supplementation, gene silencing, and gene editing in both ex vivo and in vivo platforms. In the cornea, ex vivo gene therapeutic strategies were attempted for corneal graft survival, and in vivo gene augmentation therapies aimed to prevent herpes stromal keratitis, neovascularization, corneal clouding, and wound healing. However, none of these studies followed a clinical trial–based successful outcome. CRISPR/Cas system offers a broad scope of gene editing and engineering to correct underlying genetic causes in corneal dystrophies. Corneal tissue--specific gene correction in vitro with minimal off-target effects and optimal gene correction efficiency followed by their successful surgical implantation, or in vivo CRISPR administration targeting pathogenic genes finds a way to explore therapeutic intervention for corneal dystrophies. However, there are many limitations associated with such CRISPR-based corneal treatment management. This review will look into the development of corneal gene therapy and CRISPR-based study in corneal dystrophies, associated challenges, potential approaches, and future directions.

The cornea’s multilayer structure could undergo slow and progressive degeneration, which are primarily bilateral, noninflammatory, nonsystemic, hereditary, and causes corneal structural abnormalities and opacification leading to blindness. Such a group of disorders are known as corneal dystrophies. There are 22/23 such dystrophies, each involving different corneal layers and underlying genetic causes.¹ Each dystrophy entails deposits within stromal lamellae, clouding due to fluid retention out of endothelial dysfunction, opacities, and morphological changes. In Figure 1, corneal conditions in various types of corneal dystrophies have been marked. Most corneal dystrophies are monogenic and show the autosomal dominant form of inheritance with high penetrance, and very few reflect autosomal recessive conditions. The International Committee for Classification of Corneal Dystrophies (IC3D) was organized to categorize and define the corneal dystrophies keeping the well-defined corneal dystrophies to category-1 with identified causative genes and known mutations, while the least defined belongs to category-4, in which the clinical and genetic evidence is still unclear (Table 1).² Management of corneal dystrophies depends upon their type, stage, and severity. Mild cases can be managed by gel/ointments, therapeutic contact lens, or conventional corneal scraping; and for severe cases, corneal transplant using different surgical interventions, such as Descemet stripping automated endothelial keratoplasty (DSAEK), Descemet membrane endothelial keratoplasty (DMEK), full-thickness penetrating keratoplasty (PK), and endothelial keratoplasty (EK), needs to be employed.³ The cornea is highly compatible and relatively successful tissue for allograft transplantation.⁶ However, there are several challenges with current corneal transplant management, including complex surgical procedures, the requirement of repeated treatments, extreme shortage of appropriate donor cornea, postsurgical complications such as wound leak, glaucoma, endophthalmitis, and more importantly, graft rejections.^4,5 Moreover, ready access to a supply and facilities of viable donor corneas is simply not available in many countries, specially low-income African-Asian countries, making the current treatment regime unavailable for the large population. Another serious concern is the transmission of donor-associated infectious diseases in the eye.⁷ To overcome these hurdles, genetic medicine–based approaches, including gene therapy and gene editing, have been getting attention in the field of corneal ophthalmology.

FIGURE 1: Digital collage images of distinct types of corneal dystrophies (derived from host institute hospital). A-I, Examples of corneal dystrophies. The corneal images (left) were obtained under low magnification with broad-beam illumination on a slitlamp camera. The corresponding anterior segment optical coherence tomography (ASOCT) (right) were taken under standard procedure. A, Epithelial basement corneal dystrophy: gelatinous drop--like corneal dystrophy in a patient with severe disease associated with the characteristic mulberry-like deposits. ASOCT of the same patient with gelatinous drop–like keratopathy (GDLK) demonstrates markedly thickened cornea, irregular anterior stromal architecture, and diffuse anterior stromal deposition. B, Stromal granular corneal dystrophy: granular corneal dystrophy type 1 (GCD1) with breadcrumb-like lesions amidst clear intervening spaces. ASOCT shows refractile deposits throughout the stroma with an intervening clear stroma. C, Stromal lattice corneal dystrophy: lattice corneal dystrophy type 1, with refractile branching lines within the corneal stroma, visualized direct illumination. ASOCT shows Bowman scarring with a hypertrophied epithelium. D, Stromal granular lattice corneal dystrophy: Avellino dystrophy with refractile branching lines with breadcrumb-like lesions amidst clear intervening spaces. ASOCT shows diffuse anterior stromal scarring with refractile deposits with clear intervening stroma. E, Stromal macular corneal dystrophy: macular corneal dystrophy visualized with broad-beam illumination and slit-beam illumination, demonstrating extensive stromal deposits extending to the corneal limbus. ASOCT shows Bowman scarring with refractile deposits throughout stroma with an opaque intervening stroma. F, Stromal Schyder corneal dystrophy: Schnyder corneal dystrophy, central corneal opacity sparing the corneal periphery. ASOCT shows dense arcus with pre-Descemet deposits. G, Corneal hereditary endothelial dystrophy: typical greyish white discoloration due to pan corneal edema. ASOCT shows markedly thickened epithelium, stroma, and Descemet membrane. Bowman membrane breaks were noted. H, Fuchs endothelial corneal dystrophy: with corneal endothelial pigments and guttae visible. ASOCT shows confluent central guttate with Descemet thickening. I, Posterior polymorphous corneal dystrophy: visualized with direct illumination, including a subtle presentation with a faint tram track–like lesion on the corneal endothelium running obliquely across the cornea near the central visual axis. ASOCT shows markedly thickened Descemet endothelial complex with abnormal deposits in the predescemetic plane.

TABLE 1 - List of Corneal Dystrophies Based on Latest IC3D Classification, 2015²

Corneal Dystrophies Type	Name	Onset	Pattern of Inheritance	Causative Gene[s]	Genetic Locus	IC3D Category	Current Treatment Managements
Corneal Epithelial Dystrophies	EBMD	Late onset, or early onset in some cases	Sporadic mostly	TGFβl in some cases.	5q31	Cl in some case	Eye patch, ointments and eye drops, Gas-permeable contact lens, surgery
	EREDs	First decade of life	AD	Unknown, COL17Al	Unknown 10q25.1	C3ĆC1 in some cases	Lubrication, patching & cycloplegia, antibiotics and pain relievers, Inhibitors of matrix metalloproteinase-9 (MMP-9) & topical corticosteroids
	SMCD	First decade of life	AD, some cases of X-linked inheritance	Unknown	Unknown	C4	PTK
	MECD	Early childhood	AD	KRT3ĆKRT12	12ql3 [KRT3], Ć17ql2 [KRT12]	Cl	Superficial mechanical debridement, PTK
	LECD	Childhood	X-linked dominant	Unknown	Xp22.3.	C2	PTK
	GDLD	First to second decade	AR	TACSTD2, MISI	Ip32	Cl	PTK
Bowman Layer Dystrophies	RBCD	Childhood	AD	TGFβl	5q31	Cl	PK
	TBCD	Early childhood	AD	TGFβl	5q31	Cl	PK
Stromal Dystrophies	GCD1	Early Childhood	AD	TGFβl	5q31	Cl	PTK, LK, PK
	GCD2	Childhood	AD	TGFβl	5q31	Cl	PTK, LK, PK
	LCD1	First to second decade	AD	TGFβl	5q31	Cl	Patching and soft contact lenses, PK, PTK
	MCD	Childhood	AR	CHST6	16q22	Cl	PTK, LK, PK
	SCD	First to third decade	AD	UBIAD1	lp36	Cl	PTK, LK, PK
	CSCD	Congenital	AD	DCN	12q21.33	Cl	PTK, LK, PK
	FCD	Congenital or first years of life	AD	PIKFYVE	2q34	Cl	PTK, LK, PK
	PACD	Childhood	AD	EPYC, KERA, DCNLUM	12q21.33	Cl	PTK, LK, PK
	CCDF	Childhood	Unknown	Unknown	Unknown	C4	PK
	PDCD	Usually after third decade	Not clear, some cases reported with AD	Unknown	Unknown	C4	PK
Descemet Membrane and Endothelial Dystrophies	FECD	Early-onset Childhood	Unknown	CO8A2	Ip34.3-p32	Cl	PK, DSAEK, DMEK
	Late-onset Usually 4th to 5th decade	AD	TCF4, SLC4A11 ZEB1, AGBL1	13pter-ql2.13, 20pl3-pl2, 18q21.2-q21.3, 9p24.1-p22.1, 5q33.1-q35.2, 15q25 and 1 Op 11.2.	C2 & C3 in some cases	PK, DSAEK, DMEK
	PPCD	PPCD1 Early childhood	AD	unknown	20pll.2-qll.2	C2	B-blockers, alpha-adrenergic agonists, and carbonic anhydrase inhibitors, PK, DSEK, DMEK
	PPCD2 Early childhood	AD	COL8A2,	Ip34.3-p32.3	Cl	PK, DSAEK, DMEK
	PPCD3 Early childhood	AD	ZEB1	10pll.22	Cl	PK, DSAEK, DMEK
	CHED	Congenital	AR	SLC4A11	20pl3	Cl [C3 in some cases]	PK, DSAEK, DMEK
	XECD	Congenital	X-linked Dominant	unknown	Xq25	C2	PK, DSAEK, DMEK

AD indicates autosomal dominant; AR, autosomal recessive; CCDF, central cloudy dystrophy of Francois; CHED, congenital hereditary endothelial dystrophy; CSCD, congenital stromal corneal dystrophy; DMEK, Descemet membrane endothelial keratoplasty; DSAEK, Descemet stripping automated endothelial keratoplasty; EBMD, epithelial basement membrane dystrophy; EREDs, epithelial recurrent erosion dystrophies; FCD, Fleck corneal dystrophy; FECD, Fuchs endothelial corneal dystrophy; GCD1, granular corneal dystrophy, type 1; GCD2, granular corneal dystrophy, type 2; GDLD, gelatinous drop--like corneal dystrophy; LCD1, lattice corneal dystrophy, type 1; LECD, Lisch epithelial corneal dystrophy; LK, lamellar keratoplasty; MCD, macular corneal dystrophy; MECD, Meesmann corneal dystrophy; PACD, posterior amorphous corneal dystrophy; PDCD, pre-Descemet corneal dystrophy; PK, penetrating keratoplasty; PPCD, posterior polymorphous corneal dystrophy; PTK, phototherapeutic keratectomy; RBCD, Reis-Bucklers corneal dystrophy; SCD, Schnyder corneal dystrophy; SMCD, subepithelial mucinous corneal dystrophy; TBCD, Thiel-Behnke corneal dystrophy; XECD, X-linked endothelial corneal dystrophy.

Gene Therapy Strategies

Gene therapy (GT) consists of 3 possible strategies: gene supplementation, gene silencing, and gene editing with 2 modes of delivery, ex vivo and in vivo in target cells. Gene supplementation involves the delivery of functional wild-type gene; gene silencing requires the delivery of molecule which can inactivate the functionality of a gene involved in disease pathogenicity; and gene editing includes the delivery of a molecular system that enable targeted alterations in the host gene sequences.⁸ The gene supplementation utilizes vectors carrying gene-specific cDNA, (sometimes codon-optimized to enhance the gene expression); gene silencing includes antisense RNA, si-RNA, micro-RNA, ribozymes; and for genome editing, tools including TALEN (transcription activator-like effector nuclease), ZFN (zinc finger nucleases), and recently evolved CRISPR system (clustered regularly interspaced short palindromic repeats) can be used.⁹ For all these, the delivery of genes can be mediated using viral or nonviral methods (reviewed in many publications).¹⁰ In the current decade, there are few remarkable achievements in ophthalmic genetic medicine, including the landmark success of retinal adeno-associated virus (AAV)-based GT for Leber Congenital Amaurosis (LCA) in the form of commercial GT drug LUXTURNA.¹¹ On the other hand, gene therapy in the cornea is progressive to pave its pathway for a breakthrough.

Gene Therapy in the Cornea

Cornea carries several advantages, making it suitable for GT. Being immune-privileged, avascular, transparent, located externally, and having advanced imaging and well-established surgical techniques with dedicated eye banks for its storage, makes it suitable for both in vivo and ex vivo. However, it also possesses certain barriers such as tear flow, blinking, and strong epithelial tight junctions, which provide strong resistance against external gene delivery.¹² The future prospect of corneal dystrophies will require advancement in approaches and overcoming associated challenges both at ex vivo and in vivo platform.

Ex Vivo Corneal Gene Therapy

In ex vivo corneal GT, corneal cells/tissue to be transplanted is cultured for weeks or months in a lab setup and administered with a recombinant viral or nonviral vector carrying the essential therapeutic gene. The positive selected healthy cells are then transplanted to the host using various corneal surgical procedures. Although several developments for corneal graft procedure have been made, significant improvement in long-term corneal graft survival rates has not been achieved. Administration of corticosteroids in the form of drops, tablets, or injectables and systemic immune suppression is currently the way to suppress allograft rejection. Still, it is associated with a lot of side effects and remains ineffective in the “high-risk patients,” which includes patients related to neovascularization, inflammation, glaucoma, and a previous history of graft rejection.¹⁴ Therefore, GT approaches by delivering factors aiming to modulate the immune response, inflammation, and cellular proliferation and differentiation have been tested for corneal graft survival (Table 2).¹⁵

TABLE 2 - Corneal Gene Therapy Studies for Graft Survival

GT Vector	Gene	Graft Donor	Graft Recipient	Study Outcome	PMID Ref
rAd	CTLA4-Ig	C57BL/6 mice	BALB/c mice	Prolonged corneal allograft survival on systemic administration blocking CD28/B7 T-cell signals.	11923251
rAd	IL-10	Ovine	Sheep	Corneal allograft survival increased from 20 days to 55 days.	11397952
rAd	IL-10	Wistar-Furth rats	Lewis rats	Prolonged graft survival with systemic adeno viral mediated transduction while no effect with local adeno viral or liposome-mediated method.	17093506
rAd	NGF	Dark Agouti rat	Lewis Rat	Prolonged mean survival time of donor graft by 16 days.	17325145
rAAV	TGFβl	Rat	Rat	Overexpression of TGFβl prolonged high-risk allograft survival.	23624044
rLV	PD-L1	Rat	Rat	Significantly prolongs corneal allograft survival with reduced expression of proinflammatory cytokine.	22300371
rLV	Bcl-xL	C57BL/6 mice	BALB/c mice	Improved corneal allograft survival by preventing apoptosis of the endothelium.	17614980
rLV	E-K-5	New Zealand white rabbits	New Zealand white rabbits	Showed inhibition of neovascularization and graft rejection.	12714613
rEIAV	IDO	C3H mice	BALB/c mice	IDO overexpression in donor cornea significantly increased graft survival by inhibiting T cell and inflammatory response.	16482510
rAd	p40 IL-12 and IL-4	Ovine	Sheep	Local administration of p40 IL-12 prolonged corneal graft survival significantly, while IL-4 induced rejection.	16081789
rAd	TNFR-Ig	Dutch Belted rabbits	New-Zealand white rabbits	Marginally increased graft survival in vivo compared to control.	11381062
rAd	IL-4	Wistar-Furth rats	Lewis rats	No prolongation of corneal allograft survival	10943680
rAd Entranster™ vector	Monomerie anti-CD4 antibody ĆCD25 siRNA	Wistar-Furth rats ĆSprague-Dawley (SD) rat	Fisher-344 ratsĆSD rat	No prolongation of graft survival. ĆCD25-siRNA prolonged graft survival via upregulating anti-inflammatory molecule (IL10) expression	1585156826024991
Cationic polymer jetPEI	IL-lra	Wistar-SD rat model	Wistar-SD rat model	Inhibited graft failure via downregulating immune mediators.	23723965
MIDGE	IL-4 + CTLA4CTLA4	C3H/mice	Balb/C mice	Prolongation of corneal allograft survival when treated to the recipient cornea only using gene gun technique.	12397435

Bcl-xl indicates B-cell lymphoma-extra-large; CD, cluster of differentiation; CTLA4-Ig, cytotoxic T lymphocyte antigen 4-immunoglobulin; E-K5, endostatin-kringle 5 of plasminogen fusion gene; IDO, indoleamine 2,3-dioxygenase; IL, interleukin; IL-lra, IL1 receptor antagonist; MIDGE, minimalistic immunologically defined gene expression; NGF, nerve growth factor; PDL1 programmed death-ligand 1; rAAV, recombinant adeno-associated virus; rAAV, recombinant adeno-associated virus; rAD, recombinant adenovirus; rEIAV, recombinant equine infectious anemia virus; rEIAV, recombinant equine infectious anemia virus; rLV, recombinant lenti virus; rLV, recombinant lentivirus; siRNA, small interfering ribonucleic acid; TGFβl, transforming growth factor-beta 1; TNFR-Ig, tumor necrosis factor receptor-immunoglobulin; TNFR-Ig, tumor necrosis factor receptor-immunoglobulin.

In Vivo Corneal Gene Therapy

In vivo corneal GT applies direct administration of therapeutic genes using viral or non-viral methods. To deliver therapeutic genes to the cornea, direct intra-stromal injection, topical administration, intracameral injection, or subconjunctival injection route can be applied.¹⁶ However, few studies have employed non-ocular gene route of gene administration (intramuscular, intraperitoneal).¹⁷ The viral/non-viral mediated therapeutic gene administration through microinjection may also be accompanied by other physical methods such as electroporation or ultrasound or involve surgeries such as excimer laser ablation to enhance the gene delivery.¹⁸

In Vivo Corneal Gene Therapy Using Reporter Gene Delivery

Early attempts of in vivo GT started in 1994–95, mainly with reporter genes (LacZ, GFP, Luciferase) to access the corneal gene delivery efficacy and efficiency. Viral and nonviral methods were tested in various animal models to analyze the reporter gene expression, optimum dose, cytotoxicity, and duration (Table 3).

TABLE 3 - Corneal In Vivo Gene Therapy

GT Vector and Delivery Method	Gene	Animal Model	Outcome	PMID Ref
Reporter genes AV vector [Microinjection to anterior chamber]	lacZ	Rabbits	Detected endothelial expression after 48h of administration with the significant sign of inflammation	8641831
AV vector [Microinjection to vitreous body, anterior chamber or peribulbar space]	lacZ	Mice	Continuous 50 days expression detected in 20-30% of the endothelial cells without resulting detectable cytopathic effect	7584067
AV vector Intravitreal/ Intracameral injection	lacZ	Mice	Corneal endothelial and trabecular meshwork expression for 14 days was achieved without toxic effect.	7558714
HIV-1-based lentiviral vector [Intracameral injection]	GFP	Mice	Led endothelial GFP expression for 12 weeks	11895005
HIVl-derived lentiviral vector [Intrastromal microinjection assisted with femtosecond LASER]	GFP	Pig	Constant transgene expression was observed for 21 days.	19387484
Baculovirus vector [Intravitreal injection]	GFP	Mice	Detected continuous endothelial expression for 14 days after administration.	11726636
AAV5 [Topical administration]	GFP	Rabbit	Detected expression at day 3, maximum expression at day 7 and maintained up to 16 weeks.	21533273
CAV-2 vector [Microinjection]	GFP	Mice Gray	Widespread expression in cornea diffused throughout stromal region for a week	24607662
rAV-5 vector [Topical administration]	LacZ	mouse Lemurs Cotton rats	b-Gal expression was observed in conjunctival epithelium only at 24 and 48 hours.	9878215
rAAV [Anterior chamber injection along with LPS induced inflammation]	LacZ	New Zealand
White rabbits			Endothelial LacZ expression lasted for 15 days with induction of inflammation	11867594
rAAV vector (serotypes) [Intrastromal injection or topical administration]	EGFP	Mouse	Successful transfection in mouse cornea showed expression in week 1, maximum expression at week 2 and lasted up to week 4 without any morphological changes	27001051
rAV and rAAV [Topical administration]	GFP	Rabbits	Significant GFP keratocyte expression was detected using AAV while higher expression IN cornea was detected using rAV	19023450
Plasmid [Gene gun]	GFP	Rabbits	A steady expression for 7 days in corneal epithelium using gene gun without apparent ocular damage	9298220
Plasmid DNA [Electric pulse]	LacZ	Brown Norway rat	Transgene expression was observed in corneal stroma for 15 days and maximum between day 4-6 without apparent ocular damage	11950229
Cationic lipoplexes vector [Intravitreal injection]	Luciferase	Rabbits	Led continuous expression for 1 –7 days with peak expression at day3 with maximum expression in aqueous humor	15196630
NOVAFECT chitosans oligomer [Intrastromal injection]	GFP	Rat	GFP expression was observed in keratocytes after transfection.	19879644
Linear PEI-DNA nanoconstruct [Topical administration]	GFP	Mice	Post topical single dose administration showed significant GFP uptake at day 3 with no significant ocular side effects.	30699061
AAV6, AAV8 and AAV9 vector [Topical administration]	AP	Mouse	A continued transgene expression for 30 days was observed without significant apoptotic effect with maximum transduction efficiency using AAV9	20599959
Cationic lipoplexes, and AAV vector [lamellar flap-technique]	CAT, lacZ	Rabbits	Lamellar flap allowed access to the stromal delivery. Cationic lipoplex showed expression in cornea for shorter duration [4–72 h] while AAV led to expression for longer duration [3–l0 days]	12573666
Plasmid (KRT1– 12, Promoter) [Gene gun] Transgenes	lacZ	Rabbit	Corneal specific expression for 2 days	9856765
naked plasmid DNA [Topical administration]	IL-10	Animal model	Topical administration of DNA coding IL-10 was found to be effective in resolving HSK induced lesions	9257860
Plasmids [Electroporation]	IL-10	Mice	Transgene expression was observed in day 1 –3 but had adverse effect when used DNA nuclear targeting sequence in vector.	17327469
rAV vector [Anterior chamber injection]	sFlt-1	Rat	Efficient transduction to corneal endothelial cells and trabecular meshwork cells were obtained and maintained for 10 days post injection.	11440623
AAV vector [Intrastromal injection]	HLA-G	Rabbit	Prevented corneal vascularization, lymphocyte infiltration and reduced myofibroblast formation significantly.	29259248
AAV vector [Topical administration to abraded cornea]	Lyir-targeting hammerhead ribozyme	Rabbits	AAV vectors expressing LAT-targeting ribozymes prevented reactivation of HSV latent infection	29875240
Plasmid mixed with lipofectamine [Microinjection to anterior chamber, vitreous cavity, and subretinal space]	HO-1	Rabbits	Intracameral administration led both endothelial and epithelial expression while only endothelial expression observed when administered by intravitreal injection.	7558713
Lipofectamine 2000, Entranster[TM]-in vivo, polyethyleneimine [PEI] [Topical administration]	siRNA targeting HSV-1 LCP4 gene	Mouse	Reduced corneal damage reduction and HSK VP16 gene expression.	25336327
rAAV vector [Microinjection]	Antisense VEGF	SD rats	Significantly reduced corneal neovascularization compared to control group	16762237
Direct injection [corneal micropocket assay]	HGF/NK4 against VEGF	Rabbit	The co-injection of HGF/NK4 inhibits phosphorylation of ERK and ets-1 expression which results in angiogenesis inhibition caused by VEGF.	12637990
AV vector [intramuscular injection]	TGF-βLL receptor	Mice	Inhibition of TGF-beta by delivering AV expressing soluble TGF-βII receptor fused with IgG (AdTbeta-ExR) in muscle reduced corneal edema, opacification, and angiogenesis.	11127579
p-PEDF-SAINT [Subconjunctival injection]	PEDF	Rat	Sustained PEDF expression inhibited corneal neo-vascularization up to 3 months with minimal immune reaction	19596319
HSV-1 [Intraperitoneal injection]	IFN-y, IL-2, LL-4	Mice	Immunization with rHSV-1 carrying IL4 showed more host immune responses and cleared the virus compared to IL2 and the effect was least for IFN- g	12719570
AAV5 [Topical administration]	Smad7	Rabbit	Significantly reduced corneal haze and corneal fibrosis with single administration	28339457
[PEI2-GNPs] [Topical administration]	BMP-7	Rabbit	Modulated wound healing and prevented corneal fibrosis without cytotoxicity.	23799103
Retroviral vector [Topical administration after post phototherapeutic keratectomy (PTK)	dnGl	Rabbits	Administration of dominant negative mutant cyclin Gl [dnGl] led corneal haze prevention	11923236
AAV-5 vector [Topical administration after epithelium removal]	DCN	Rabbit	Decorin GT showed significant decrease in CNV with no major side effects.	22039486

AP indicates alkaline phosphatase gene; IL-10, interleukin 10; BMP7, bone morphogenetic protein 7; CAT, catalase; CAV-2, canine adenovirus type-2; DCN, decorin; dnGl, dictyostelium homolog of the mammalian tumor suppressor ING gene; EGFP, enhanced green fluorescent protein; GFP, green fluorescent protein; HGF-NK4, N-terminal hairpin and subsequent four-kringle domains of hepatocyte growth factor [HGF] ; HLAG, human leukocyte antigen G; HO, heme oxygenase ICP4, infected-cell polypeptide 4; IFN-g, interferon gamma; IL-2, interleukin 2; IL-4, Interleukin 4; LacZ, p-galactosidase; LAT, latency-associated transcript; MMP9, matrix metallopeptidase 9; PEDF, pigment epithelium-derived factor; PEI2-GNP2, polyethylimine-conjugated gold nanoparticles; sFlt-1, soluble variant of the VEGF receptor; Smad7, SMAD family member 7; SOX-2, SRY-Box transcription factor 2; TNF-a, tumor necrosis factor; VEGF-A, vascular endothelial growth factor A.

In Vivo Corneal Gene Therapy for “Herpes Stromal Keratitis”

In vivo corneal GT using therapeutic transgene initially focused on treating herpes stromal keratitis (HSK). HSK accounts for 10% of the corneal transplantation burden. Local administration of IL10 had shown prevention against HSK.¹⁹ Accounting short half-life of IL10, its gene therapy--based administration overcomes the requirement of repeated doses. Therefore, IL10 as a therapeutic transgene was administered using different delivery methods, including the electroporation method and adeno and lentiviral-based delivery in murine, ovine, or human cornea.^20,21 In another study, single adeno- associated virus (AAV) mediated intrastromal injection of the immunomodulatory and anti-inflammatory molecule HLA-G (human leukocyte antigen-G), prevented corneal neovascularization (CNV), reduced trauma-induced T-lymphocyte infiltration, and condensed myo-fibroblasts formation in rabbit cornea.²² Herpes simplex virus 1 (HSV1) infection can enter in sensory nerve endings, and upon trigger such as ultraviolet (UV) exposure, stress, ocular surgery, and hormonal factors, it can further reactivate its latent infection.²³ To prevent this, adenoviral-mediated delivery of ribozymes targeting HSV1 late gene UL20 and AAV-mediated ribozymes targeting HSV latency-associated transcript (LAT) region were used, and it could block the viral reactivation in more than 60% of infected eyes in animal studies, thus showing the relevance of the LAT targeting GT approach for the treatment of HSV latent infection.²⁴ Besides IL-10, other anti-inflammatory mediators such as IL-2, IL-4, IL-18, IL-21, IL12p35, IL-12p40, IFN-b, IFN-a-1, IFN-γ, colony-stimulating factor (GM-CSF) and TNF-a, have been investigated as other possible gene therapy targets for the treatment of corneal inflammation (Table 3).²⁵

In Vivo Corneal Gene Therapy for ”Corneal Neovascularization”

Different viral systems were investigated for efficient and safe delivery of genes in the cornea to protect against corneal neovascularization (CNV), caused due to corneal injury, inflammation, or infection. The pathogenic corneal vascularization originates from the perilimbal plexus and invades toward the pannus (between epithelium and Bowman layer) or stroma.²⁶ Both pannus or stromal neovascularization is regulated by the balance between vascular endothelial growth factors such as VEGF, VEGF-C, VEGF-D, sFlt, VEGFR3, endostatin, and thrombospondin-1 and -2 within the cornea and therefore are the prime targets for their modulation using GT (Table 3).²⁷

In Vivo Corneal Gene Therapy for Corneal Haze, Clouding, Wound Healing

Further attempts using in vivo corneal GT were made for the therapeutic purpose of corneal haze, opacity, corneal clouding management of mucopolysaccharidosis VII, wound healing (Table 3).^28,29 Retro viral-mediated delivery of dominant-negative mutant cyclin G1 (dnG1) prevented the development of corneal haze after phototherapeutic keratectomy (PTK) in rabbits.³⁰ Using “decorin” an important leucine-rich proteoglycan in corneal stromal connective tissue, and a natural antagonist of transforming growth factor-beta (TGF-β), it was found that its gene transfer in human donor corneas derived fibroblast (HSF) could inhibit the formation of TGF-β driven myofibroblast and haze, indicating the potential use of decorin-GT to treat corneal haze.³¹ Following this study, in vivo decorin-GT in the CNV model of rabbit stroma was performed, delivering a single dose of AAV titer carrying the decorin gene, resulting in a significant reduction in neovascularization with no major side effects, thus suggesting a decorin-GT clinical trials.³² Topical bone morphogenic protein 7 (BMP7) gene delivery using polyethyleneimine-conjugated gold nano-particles (PEI2-GNPs) in laser-induced corneal wound healing rabbit model showed modulation in wound healing and inhibition in fibrosis by regulating TGFβl-mediated profibrotic Smad signaling.³³ In another study, single dose of AAV5 based gene delivery of Smad7 to the corneal stroma repressed corneal haze and fibrosis post photorefractive keratectomy (PRK) in a rabbit corneal injury model without major immune response. Combination GT of BMP7 and hepatocyte growth factor (HGF) using PEI2-GNP in alkali injury rabbit in vivo and human in vitro models had a protective effect against corneal fibrosis and corneal transparency.³⁵ In mucopolysaccharidosis type VII (MPS VII), canine adenovirus type 2 (CAV-2) vectors mediated administration of deficient enzyme – β-glucuronidase (β-glu or GUSB) in corneal keratocyte of mice and nonhuman primates (in vivo) and in the corneal graft of dog and human (ex vivo), led to stable release and diffusion of β-glu throughout the collagen-dense stroma and improved corneal clouding compared to control, suggesting CAV-2 vectors–based GT for diseases affecting corneal keratocytes.³⁶ Recently, using adenoviral mediated intracameral or intrastromal injection of GUSB gene in mice model of mucopolysaccharidoses (MPS) a widespread transgene expression was observed in cornea which was effective in clearing corneal cloudiness without generating significant neutralizing antibodies against adeno virus.³⁷ Using AAV8/9 chimeric capsid serotypes (8G9) in a study, delivering a codon-optimized wildtype gene iduronidase in donor human cornea, tenfold increase in its enzymatic (alpha-L-iduronidase, IDUA) activity was demonstrated to potentially treat corneal clouding and progressive corneal blindness.³⁸

Gene Editing and Crispr Technology

Recent advancement in gene editing tools has opened a new therapeutic potential other than conventional to correct genetic alterations in ocular disorders and restore normal visual function. The early generation genome editing tools such as meganucleases, zinc finger nucleases (ZFNs), and transcription activator-like effector nucleases (TALENs), could not gain scientific momentum because of their complexity in designing and practical application. With the advent of “CRISPR/Cas system” --a technological derivation of bacterial immune defense mechanism against the virus, the targeted alteration in the cellular genome for research and therapeutic purposes has become radially doable, accessible, designable, and affordable. This RNA-guided genome editing tool is now being widely used for genomic alterations, transcriptional regulation, and epigenetic regulation. The CRISPR-Cas technology has become a universal practical tool for genome editing and engineering with a wide range of applications.³⁹

CRISPR Composition and Mechanism

The bacterial antiviral adaptive immune system comprises 3 to 4 classes of CRISPR systems in which the CRISPR-Cas9 belongs to class-2, comprising a single effector protein molecule Cas9. The fundamental component of CRISPR- Cas9 technology involves i) Cas9 protein, which creates double-strand breaks (DSB) to the target region near PAM [protospacer adjacent motif (NGG) sequence (N- any nucleotide base]; ii) single-guide RNA (sgRNA) which consist 2 parts: (a) around 20 base length CRISPR RNA (crRNA) complimentary to the target region, (b) tracer RNA which is a scaffold sequence facilitating guide RNA binding to Cas9 protein.⁴⁰ The 2 components, Cas9 protein and single-guide RNA, now can be available in different forms or formats, such as Cas9 protein, Cas9 plasmid constructs, Cas9 mRNA, Cas9 stable cell line, Cas9 recombinant viruses, and similarly for guide RNA, crRNA/tracer RNA complex linked with linker, crRNA/tracer heteroduplex, cloned sgRNA in a vector, or in vitro synthesized sgRNA using T7 polymerase and a dsDNA template.⁴¹ Once administered inside target cells, the tracer RNA component of encoded sgRNA helps form a complex with bilobed Cas9 protein within nucleus and this complex then binds to the dsDNA target complimentary to the crRNA and undergo a conformational change which allows it to create DSB (2 to 4 bases) before PAM sequence motif.⁴² The DSB induces an endogenous repair system in the cell which may be of 2 types: i) nonhomologous end-joining (NHEJ) and ii) homologous end-joining (HEJ) discussed below.⁴³

CRISPR Nonhomologous End-Joining Mechanism and Applications

The NHEJ, an error-prone repair system within the cell, can be exploited to cause random errors, including insertions or deletions (indels) of various sizes. It was also seen that these indels might be of distinctive pattern based on the type of gRNA used.⁴⁴ The NHEJ directed genome editing has been widely used for gene deletion and insertion for creating knock-out cell lines and animal models in a time-effective manner. Using 2 gRNA simultaneously, large genomic deletions from kilobase to megabase sizes can be achieved between 2 DSBs of 2 gRNA targets, and the frequency of large deletions in such scenario decreases on increasing the size of deletions.⁴⁶ Allele-specific gene editing can also employ the NHEJ pathway and has been developed as one of the potential approaches to treat autosomal dominant disorders with dominant-negative effects.⁴⁷

CRISPR Homology End-Joining Mechanism and Applications

The HEJ-directed editing, on the other hand, uses homologous sequence--based correction. Therefore, it requires an additional “donor element” with desired alterations and homology arms on either side of the target and acts as a reference template for DNA repair, like that cell uses the homologous chromosome in a normal cellular DNA repair system. The donor element can be in the form of a vector or single-stranded or double-stranded sequences. In the presence of excess delivered donor elements, homologous recombination between target and donor across their homology arm occurs at the site of Cas9 meditated cut, and the HDR mediated repair gets executed.⁴⁸ Since the donor element sequence is identical to the target, to avoid Cas9 mediated cleavage on donor itself, it needs to be introduced with silent mutation, preferably at its PAM site.⁴⁹ The HEJ system, which repairs in an error-free manner, has been utilized to create knockin, gene alteration, or gene correction in the target region.⁵⁰ A large number of genetic diseases results from a single-base change and thus prime targets for HDR-mediated CRISPR corrections. In a remarkable clinical study using CRISPR-based HDR, COL7A1 gene mutation correction was performed in patient-derived induced pluripotent stem cells (iPSCs) with recessive dystrophic epidermolysis bullosa (RDEB)—a severe inherited skin blistering disorder. Followed by this, from gene-corrected iPSCs, 3-dimensional human skin equivalents (HSEs) were generated and differentiated into keratinocytes (KCs) and fibroblasts (FBs), and grafted into immunodeficient mice, which showed wild-type expression of COL7A1 and restored anchoring fibrils with no evident off-targets activity or other safety concern.⁵¹ Similar approaches can be implemented in corneal dystrophies as well, where patient-derived iPSC could be generated and corrected for its defective mutation using CRISPR tool followed by its differentiation into endothelial cells, which can be surgically grafted in patients cornea for the rescue of disease phenotype. Interestingly, in 1 study, it was shown that dental pulp contains a population of adult stem cells known as DPSc, that can be harvested from third molars with the capability to differentiate into corneal keratocytes.⁵² Such stem cells from the patient can be an alternative source for corneal keratocytes for gene correction.

Advancement in Crispr Technology

There are many modifications done on Cas9 protein to improvise further action of the CRISPR/Cas9 system. These advancements have been aimed to reduce off-target effects, minimizing Cas9 size for efficient delivery, and mediate CRISPR-based gene regulations.

Cas9 Modifications

The CRISPR/Cas9 protein used was originally derived from Streptococcus pyogenes known as (SpCas9). Many mutant SpCas9 variants have been developed using directed evaluation which has different PAM sequence specificity other than NGG, with reduced off-target effects such as SpCas9-HF1 (high fidelity SpCas9), eSpCas9 (enhanced-specificity SpCas9), HypaCas9 (hyper-accurate SpCas9), SpCas9-NG (Streptococcus pyogenes Cas9), and xCas9.⁵³ Recent development in RNA targeting Cas9 such as modified SpyCas9pPAMer and type II-A and II-C Cas9 has enabled PAM less binding and thus, broadened the target specificities and scope for CRISPR editing.^54,55 Modifications in Cas9 domains have facilitated CRISPR/Cas9 efficiency and accuracy. Nuclease domains, either HNH or RuvC of the Cas9 nuclease, have been mutated independently (H840A in HNH or D10A in RuvC) to create Cas9 DNA “nickases” (nCas9) capable of introducing a single-strand cut with the same specificity as a regular CRISPR/Cas9 nuclease. Using 2 different nCas9 in close proximity dsDNA double breaks can be generated, with reduced off-target cleavage.⁵⁶ Mutation-based modification in Rec3 domain led to HypaCas9 that showed genome-wide high specificity.⁵⁷ dCas9 (dead Cas9) is a mutant form of Cas9 lacking endonuclease activity due to silent mutations D10A and H841A in the RuvC1 and HNH nuclease domains, respectively. However, it retains its gRNA directed PAM dsDNA binding. Such dCas9 can be fused with other proteins at its N or C terminus to get desirable CRISPR effects for genetic and epigenetic regulation without a double-strand break in the target templates.⁵⁸ For example, by fusing DNA methyltransferase (DNMT) (an enzyme responsible for methylating cytosine residue in DNA) with the dCas9, it was possible to induce DNA methylation at that site and repress unwanted gene expression.⁵⁹ Similarly, fusing Ten-eleven translocation methyl-cytosine dioxygenase l(Tetl)—a demethylating enzyme catalyzing the conversion of the methylcytosine to 5-hydroxymethylcytosine, it was possible to induce promoter demethylation and activate gene expression for silent beta-galactoside alpha-2,6-sialyltransferase 1 (ST6GAL1) gene while promoter methylation fusing DNMT repressed the expression.⁶⁰ Another enzyme, histone acetyl-transferase, which can regulate gene expression by acetylation of histones and the induction of chromatin remodeling, could be coupled with dCas9 to control gene expression at target regions, including promoter, enhancer, or repressor regions. CRISPR interference (CRISPRi) by fusing dCas9 with Krupple-associated box (KRAB) enzyme—a transcriptional repressor has been widely used and have the advantage over other CRISPR silencing approach that it avoids DNA gene editing cut, is reversible, and can be utilized in different cell types in a multiplexed manner.⁶² On the other hand, upregulation of transcription was achieved in plant system by coupling dCas9 with a transcriptional activator VP64 (VP64 activation system) and using modified guide RNA scaffold, it could be further enhanced.⁶³ Thus, dCas9 has provided versatility in the CRISPR genome editing approaches.

CRISPR Base Editors

The natural property of Cytidine deaminase, which can convert cytidine into adenine by irreversible hydrolytic deamination of cytidine and deoxycytidine, was utilized to generate CRISPR base editors. Linking cytidine deaminase with the inactive dCas9 resulted in a CRISPR system that could convert cytosine (C) to uracil (U) without DSB.⁶⁴ Uracil is then subsequently converted to thymine by DNA replication or repair. However, cellular uracil DNA glycosylase (UGI) can reverse U back to C. To avoid such reversal, third-generation base editors known as “BE3” were developed by linking DNA glycosylase (UGI) inhibitor along with fused cytidine deaminase and dCas9. Such next-generation base editors could reach stable targeted conversion of C to T with efficiency ranges from 15 – 75%.⁶⁵ The A to G base editors were a roadblock in base editing because of the unavailability of natural adenine deaminase. However, using a transfer RNA adenosine deaminase and fusing it to nickase Cas9, another novel base editor was developed to direct A-G conversion at the target sites.⁶⁶ More recently, a novel “prime editing” approach has been developed by fusion of catalytically impaired Cas9 H840A nickase to an engineered reverse transcriptase (RT) enzyme. It allows all possible base corrections, by using a long guide RNA (with desired edits) which is reverse transcribed by RT. The newly synthesized edited strand gets incorporated, forms heteroduplex, and snipped off the unedited strand by nickase mechanism and generate edits in both strands.⁶⁷ The continued developments of such next-generation base editors have significantly expanded the scope of gene corrections.

Crispr Gene Editing in the Cornea

The cornea is an ideal tissue for the targeted gene editing owing to its tissue availability and accessibility during corneal transplantation. Since the underlying causes of corneal dystrophies are genetic, in situ or in vivo corrections of these mutations through the CRISPR mechanism are the prime target. Table 4 enlists the various CRISPR-based studies for corneal dystrophies. Shin et al showed that endogenous overexpression of transcriptional factor known as sex-determining region Y-box 2 (SOX2) using CRISPR activation system in corneal endothelial cells promoted wound healing and regeneration both in vitro and in cryoinjury rat model.⁶⁸

TABLE 4 - Targeted Gene Therapy using CRISPR/Cas9 in Corneal Diseases

Disease	Targeted Gene	Cell Line/Animal	CRISPR Editing	Result	PMID Ref
Granular corneal dystrophy (GCD)	TGFßl	GCD2 patient derived primary keratocytes	R124H mutation correction	Homology directed gene correction with efficiency 20.6% in heterozygous cells & 41.3%, in homozygous cells with no apparent off target	29196743
Meesmann’s epithelial corneal dystrophy (MECD)	KRT12	Humanized MECD mouse model.	L132P mutation-based knock down	Demonstrated NHEJ
directed gene deletion in and reduction of mutant allele m-RNA and protein					26289666
Corneal endothelial dystrophies	SOX2	SD rats	SOX2 overexpression	CRISPR activation of SOX2 promoted wound healing and regeneration in CECs.	30270540
Corneal epithelial dystrophy	TGFßl	Human limbal epithelial stem cells	TGFbl knockout	TGFβl knockout limbal epithelial stem cells were generated successfully.	30753226
Lattice corneal dystrophy typel (LCD1)	TGFßl	Mice	Specific- Point mutation [R124C] insertion	Generated mouse model of TGFb1-R124C corneal dystrophy which mimics features of the human corneal dystrophy.	32029872
Fuchs endothelial corneal dystrophy (FECD)	COL8A2	Adult COL8a2 mutant mice	COL8a2 start codon disruption	Rescued structural and functional basis of the disease phenotype in a FECD mouse model.	34100716
Posterior polymorphous corneal dystrophy (PPCD)	ZEB1	CEnC cell line	ZEB1 Knockout	Cell-based model of PPCD
generated					31194824

CEnC indicates corneal endothelial Cells; COL8A2, collagen type VIII alpha 2 chain; KRT12, keratin 12; SOX2, SRY-box transcription factor 2; TGFβl transforming growth factor-beta-induced protein ig-h3; ZEB1, zinc finger E-box binding homeobox 1.

In a similar study from the same group demonstrated that endogenous overexpression of antiaging protein Sirtuinl (SIRT1) had similar effects.⁶⁹ By introducing patient-derived heterozygous nonsense mutation using CRISPR in telomerase immortalized limbal epithelial stem cells (T-LSCs), an aniridia-related keratopathy invitro model was generated which on the addition of the recombinant PAX6 protein via cell-penetrating peptide could rescue the disease phenotype suggesting PAX6 gene supplementation for such disorders.⁷⁰ In a study by Fuller et al, gene deletion and its replacement with hygromycin resistance cassette by targeting ura3 virulence gene were demonstrated in corneal clinical isolates of fungal keratitis providing the scope of CRISPR in corneal infectious disease.^7l

CRISPR Approaches as a Molecular Therapy

Toward the development of CRISPR as a molecular therapy, many preclinical and few clinical studies are in progress. Different strategies have been adopted based on the nature of the genetic problem. Deletion of pathogenic deep intronic mutations using dual AAV CRISPR has shown a gain of gene function in preclinical and clinical studies in LCAl0. (Clinical trial.gov NCT03872479). CRISPR editing in isolated T cells led to the development of engineered T cells with chimeric antigen receptor (CAR) on them, known as “CAR-T cells”. These CAR-T cells are specific to the antigen present on cancer cells and act as a “living drug” against the cancerous cell. Many types of cancers are being targeted under clinical trials using CAR-T cells. (ClinicalTrial.gov NCT04035434, NCT04502446, NCT04438083). CRISPR modified stem cell is another approach. Recently, the US Food and Drug Administration (FDA) has granted CTX001-a CRISPR-Cas9 gene-editing hematopoietic stem cells, (isolated from patient blood) which can produce high levels of fetal hemoglobin in red blood cells for the treatment of sickle cell disease (SCD) (ClinicalTrials.gov NCT03655678).

CRISPR Editing in Corneal Dystrophies

The corneal CRISPR gene editing approaches started in the recent past. Corneal dystrophies caused due to TGF-beta-1 (TGFβl) mutations affect several layers in the cornea and collectively are known as the epithelial-stromal TGFβl corneal dystrophies. They exhibit strong allelic and phenotypic heterogeneity.⁷² Broadly they can be classified as lattice corneal dystrophy (LCD) or granular corneal dystrophy (GCD). CRISPR/Cas9 mediated HDR gene repair has radical potential to correct mutations in TGFβl-related corneal dystrophy. In this direction, a successful example of ex vivo CRISPR/Cas9--mediated gene editing was demonstrated in primary corneal keratocytes derived from GCD patient’s corneal tissue. Using Cas9/gRNA (px 458) plasmid and a single-stranded HDR donor template, R124H mutation was corrected with a higher gene correction efficiency of 20.6% in heterozygous cells and 41.3% in homozygous cells with no detectable off-target effects.⁷³ Generating R124C mutation in TGFβl using CRISPR/Cas9-mediated homology-directed repair mechanism in zygotes of C57BL/6Ncr mice a novel transgenic mice model was generated mimicking human LCD.⁷⁴

Allele CRISPR Editing in Corneal Dystrophies

Allele-specific targeting involves designing gRNA specific to mutant allele to cause disruption, keeping the wild-type allele intact and phenotypically active. First in vivo allele-specific genome editing using CRISPR/Cas9 tool in corneal dystrophies was studied in a knock-in humanized mouse model of MECD (Meesmann epithelial corneal dystrophy) carrying pathogenic Leu132Pro mutation in the keratin-12 (KRT12) gene. This mutation generated a novel PAM site in one allele, which could be specifically targeted. Upon intrastromal injection of CRISPR-gRNA complex in humanized MECD mice, the mutant allele underwent NHEJ with an efficiency of 38.5% and caused deletions of up to 53 nucleotides, causing a frameshift and thus cleavage of the dominant-negative mutant gene.⁷⁵ In FECD, it is found that approximately 70% of cases are caused by a TCF4 gene trinucleotide (CTG) repeats, (mostly in the range of 30–50) for which different disease-causing mechanisms have been studied including (1) dysregulated expression of TCF4 isoforms, (2) accumulation of toxic repetitive RNA transcripts, (3) non-AUG dependent (RAN) transcription and translation of repetitive RNA transcripts, and (4) age and tissue-dependent genomic instability of the repeat element.⁷⁶ To remove such expended repeats, dCas9 CRISPR-guideRNA plasmids targeting TCF4 repeats in patient-derived endothelial cells were delivered using lipofection or lentiviral transduction. Post transfection, the percentage of cells with repeated foci decreased from 59% to 5.6,% with a significant decrease in the number of foci per 100 cells.⁷⁷ In a recent study, inhibition of transcriptional factor TCF4 using siRNA (si-TCF4) and its activation using (CRISPR)/dCas9 activation systems (pl-TCF4) in the endothelium of cryo-injury model Sprague-Dawley (SD) rats was evaluated. si-TCF4 showed a reduced percentage of cells in the S phase while it was increased with pl-TCF4, with induction of corneal endothelium regeneration. These result further supports, TCF4 may be a potential target for gene editing--based treatment for FECD and other disorders requiring endothelial regeneration.⁷⁸ In a recent study by Kathleen et al, targeting natural variants containing a PAM sequence cis to mutated allele a mutation independent allele-specific gene editing could be achieved in the TGFβl gene.⁷⁹ Another CRISPR-based target found in early FECD is Col8a2. Single knockout of COL8a2 does not develop FECD phenotype and may not be indispensable to corneal function, yet causes FECD in some cases due to its mutant dominant gain-of-function activity.⁸⁰ In a recent study, it was shown that gene knocking down of mutant COL8a2 by abolishing its first initiation codon using adenovi-ral-mediated delivery of CRISPR cas9 in mice model of FECD rescued the disease pheno- type.⁸¹ Another strong potential candidate for CRISPR-based editing is SLC4A11 (sodium bicarbonate transporter-like protein 11), which acts as an ion transporter of Na+ coupled OH− ion and facilitates transmembrane water movement in corneal endothelium. SLC4All is found to be a causative gene for pediatric-onset congenital hereditary endothelial dystrophy (CHED) and influences some cases of late-onset Fuchs endothelial corneal dystrophy (FECD).⁸² The SLC4A11-point mutations are common in CHED and can be targeted for HDR-directed repair or using base editors. The CHED patients undergo DSAEK/DMEK surgeries where endothelial tissue is excised. These excised tissue/cells can be cultured in vitro. CRISPR editing of these endothelial cells and clonal amplification of CRISPR corrected cells followed by their surgical implant could be a potential treatment regime that needs to be studied. The autologous gene corrected endothelial cells implantation will avoid graft rejections. Gene corrected endothelial cells could be preserved and reserved for future use for dystrophy-related complications in same or another eye. Another potential approach could be a CRISPR-based drug administration in vivo targeting the pathogenic gene for gene correction or gene modulation. Schematic in Figure 2 shows the design of the CRISPR-based gene correction approach in corneal endothelial dystrophies. Such approaches in corneal dystrophies are needed to be studied in large for an effective clinical outcome.

FIGURE 2: CRISPR-based potential treatment process in cornea. The picture depicts a possible CRISPR-based treatment procedure for endothelial dystrophies. Patient endothelial cells/tissue sourced from DSAEK/DMEK/PK surgeries can be CRISPR-edited and gene-corrected cells can be selected, amplified, stored and eventually administered into cornea; or CRISPR components or its drug form can be delivered via viral or nonviral methods. CRISPR indicates clustered regularly interspaced short palindromic repeats; DMEK, Descemet membrane endothelial keratoplasty; DSAEK, Descemet stripping automated endothelial keratoplasty; PK, penetrating keratoplasty.

CRISPR Limitations Specific to Corneal Diseases

CRISPR technology and its applications in the cornea are in its explorative phase and possess many limitations which need to be solved. In addition to common challenges associated with CRISPR, such as off-targets, indel generations, toxicity, lower gene editing efficiency, and large size Cas9 delivery, there are other limitations for CRISPR application specific to corneal disease. Lower cellular density of monolayer endothelium due to aging provides limited cell for CRISPR gene editing studies. In a study by Charlesworth et al, the presence of pre-existing humoral and cell-mediated adaptive immune responses to Cas9 was observed in the majority of his human subjects, hindering the prospect of CRISPR-based GT application.⁸³ The preclinical studies have revealed the outcome of CRISPR-based editing, but the long-term effect of CRISPR therapeutics are still unknown due to a lack of clinical studies. Last but not least, like any new technology the CRISPR-based corneal genome editing medicine carries a potential risk of its misuse or bioterrorism.

Discussions

Corneal blindness is the fourth major cause of global blindness and involves a huge demand for corneal transplantation.⁸⁴ Owing to a large number of corneal graft failures, the management of corneal dystrophies has moved its focus from conventional surgical corneal transplant to correct the underlying genetic cause of the disease. GT and gene editing are the prime approaches in this direction. The earlier attempts of conventional in vivo and ex vivo corneal GT were mainly focused on corneal graft rejection, neovascularization, and HSK. However, these studies were limited in preclinical stages and prolonged sustained expression of transgene could not be achieved to support clinical trials. Though, such studies provided the proof of principle for the possibility of gene delivery in the cornea. As most of the corneal dystrophies are autosomal dominant, the conventional gene supplementation strategy might not be effective due to dominant negative effects of the mutated gene.

In such a scenario, genome editing is a better approach towards corneal dystrophies treatment. The new evolving approaches in CRISPR genome editing have opened more opportunities for successful GT outcomes. CRISPR gene corrections in the cornea are possible both ex vivo and in vivo. Using target-specific CRISPR gRNA corneal pathogenic gene can be altered or pathogenic region can be spliced out. Targeting allele-specific mutations by CRISPR, autosomal dominant corneal dystrophies can be managed by deletion of pathogenic repetitive repeats. Most of the genetic alterations in corneal dystrophies are single-base changes. The HEJ-directed CRISPR editing or application of next generations base editors can be instrumental in correcting these single base mutations. There are many metabolic disorders or syndromes which manifest their phenotype on the cornea. CRISPR can be useful tools for correcting their corneal manifestations. One way of broadening the CRISPR approach in corneal dystrophy is to manage corneal manifestations of these disorders which do not have any treatment regime. GT and gene editing in corneal dystrophies have promising potentials but have limitations of their substantial need against existing modalities. Since corneal dystrophies are not extreme sight-threatening at their initial stages, and slowly progressive and surgical interventions are available, GT or editing, which requires long-term extensive research, ethically bound rules, regulations, and multiple follow-ups may not be an essential therapeutic choice. In addition, for half of the corneal dystrophies which fall in ICD3/4 categories, the involved genetic factor is largely unknown. Therefore, unlike retinal GT, where there is no treatment option available, the corneal gene augmentation therapy clinical trials are very limited, and no FDA-approved corneal GT/editing is achieved so far.⁸⁵ However, the scope and hope for genetic medicine in the form of GT and editing is immense.⁸⁵ The corneal tissue availability with available tools and techniques for corneal imaging and surgery makes the wide scope of gene editing studies. With the optimization of higher gene editing efficiency and eliminating off-targets and possible cytotoxic effects in preclinical studies, a successful therapeutic application can be achieved. The success of CRISPR genome engineering will enable the progression of personalized medicine for corneal dystrophies. As genome editing has now become so approachable as never before in the history of genomic science, the regulatory steps to avoid misuse of CRISPR-based techniques against patients or in the form of bioterrorism should always be in vigilance.