Sustained Vision Recovery by OSK Gene Therapy in a Mouse Model of Glaucoma

Abstract

Glaucoma, a chronic neurodegenerative disease, is a leading cause of age-related blindness worldwide and characterized by the progressive loss of retinal ganglion cells (RGCs) and their axons. Previously, we developed a novel epigenetic rejuvenation therapy, based on the expression of the three transcription factors Oct4, Sox2, and Klf4 (OSK), which safely rejuvenates RGCs without altering cell identity in glaucomatous and old mice after 1 month of treatment. In the current year-long study, mice with continuous or cyclic OSK expression induced after glaucoma-induced vision damage had occurred were tracked for efficacy, duration, and safety. Surprisingly, only 2 months of OSK fully restored impaired vision, with a restoration of vision for 11 months with prolonged expression. In RGCs, transcription from the doxycycline (DOX)-inducible Tet-On AAV system, returned to baseline 4 weeks after DOX withdrawal. Significant vision improvements remained for 1 month post switching off OSK, after which the vision benefit gradually diminished but remained better than baseline. Notably, no adverse effects on retinal structure or body weight were observed in glaucomatous mice with OSK continuously expressed for 21 months providing compelling evidence of efficacy and safety. This work highlights the tremendous therapeutic potential of rejuvenating gene therapies using OSK, not only for glaucoma but also for other ocular and systemic injuries and age-related diseases.

Introduction

Glaucoma is the leading cause of blindness worldwide and is a chronic condition that targets retinal ganglion cells (RGCs) that reside within the innermost layer of the retina (Resnikoff et al., 2004; Sun et al., 2022). RGCs are responsible for transmitting visual signals to the brain through long axons that form the nerve fiber layer, which gathers in the posterior retina to form the optic nerve. Importantly, RGC axons lack myelin sheaths until they pass through the lamina cribrosa. Therefore, axon bundles are unprotected in the optic nerve head, which is believed to be the location where glaucomatous injury initially occurs (Quigley et al., 1981; Syc-Mazurek and Libby, 2019), which progresses to the cell body and ultimately leads to nerve cell death through apoptosis, resulting in an irreversible loss of vision.

An important trigger for developing glaucoma is increased intraocular pressure (IOP), which is believed to cause mechanically induced stress injury of the axons in the optic nerve head (Peters et al., 2014; Quigley et al., 1983). Lowering the IOP is currently the only clinically approved therapeutic option for patients and is achieved through drugs and/or surgery that target the aqueous humor outflow pathways in the anterior segment of the eye (Lusthaus and Goldberg, 2019; Weinreb et al., 2014). These treatments can significantly slow, but not stop, disease progression. Many glaucoma patients continue to progress and lose vision, in spite of achieving optimal IOP levels under IOP-lowering therapies and there is no evidence that glaucoma patients spontaneously recover vision over the long term (Chang and Goldberg, 2012; Hood, 2019; Kass et al., 2002; Pascale et al., 2012). Moreover, other patients develop “normal tension” glaucoma in which the IOP never rises and stays within normal levels (Killer and Pircher, 2018; Leung and Tham, 2022).

Since the only approved treatments regulate IOP, these approaches can only indirectly affect RGC survival, highlighting the need for new therapies that directly target RGCs and protect or cure them from the stress-induced injury that causes glaucoma.

Aging is the biggest risk factor for glaucoma in the Caucasian population, the risk of glaucoma increases from 9% in 18–39 years of age to 31% in 65–84 years of age (Allison et al., 2020). How age-related changes in the retina increase the susceptibility to developing glaucoma, however, is not clear. As glaucoma is an age-associated disease, one novel therapeutic approach we have taken has been to reverse cellular aging by restoring gene expression patterns to a more resilient, youthful cell that can withstand glaucoma-inducing stress and regain lost functions. This was demonstrated in our group's previous work, the first in vivo application of epigenetic reprogramming in the retina, using three of the four Yamanaka transcription factors Oct4, Sox2, and Klf4 (OSK) (Lu et al., 2020).

Yamanaka's seminal work discovered that overexpression of Oct4, Sox2, Klf4, and c-Myc (OSKM) induced dedifferentiation of mature mouse fibroblasts that were reprogrammed into induced pluripotent stem cells (iPSCs) (Takahashi and Yamanaka, 2006), which coincided with resetting the epigenetic clock back to zero (Horvath, 2013). Due to health-related issues associated with cell identity loss and tumor formation resulting from continuous OSKM expression in vivo (Abad et al., 2013; Ohnishi et al., 2014), a standard protocol is to induce OSKM cyclically for 2 days in a week to avoid animal death and achieve functional improvements (Browder et al., 2022; Ocampo et al., 2016), unless the level of induction was kept ultra low (Alle et al., 2022). Its safety for therapeutic translation has been questioned due to the oncogenic feature of c-Myc (Senis et al., 2018). Our previous work, however, has clearly demonstrated that removing oncogene c-Myc from the cocktail, using only OSK, prevents cell identity loss in vitro or in vivo (Lu et al., 2020).

We utilized an AAV2-tTA; TRE-OSK-inducible system, and reported the reversal of age in RGCs in older mice based on transcriptome and DNA methylome changes, which led to a dramatic improvement in RGC function and visual acuity. In addition, OSK reprogramming restored axon regeneration in an optic nerve crush model system. This study represented the first instance of successful in vivo partial epigenetic reprogramming without cell proliferation, toxicity, or an increased risk of tumor formation (Lu et al., 2020).

Induction of IOP in young mice induced glaucoma with corresponding vision loss. Remarkably, if damaged RGCs in mice with glaucoma subsequently received OSK reprogramming, visual function was improved, indicating reprogramming was rejuvenating dysfunctional but not dead RGCs (Lu et al., 2020). Further study of the epigenetic changes in these mice revealed that injury of RGCs through optic nerve crush in young mice induced changes in CpG methylation sites that were similar to those seen during aging. These data indicate that injury accelerates aging, and this is a cause of their dysfunction. This idea was later tested directly and confirmed in mice (Xu et al., 2022) and in humans (Poganik et al., 2023).

However, past research only examined the restoration of visual function after a 4-week treatment of AAV-OSK in mice with glaucoma and there was only a partial recovery (Lu et al., 2020). We did not examine the maximal vision improvement potential and the long-term implications of RGC reprogramming on visual function and did not test if the benefits can also be achieved through inducible OSK AAV that would be more clinically relevant. In this article, we present the first comprehensive long-term functional tracking study of in vivo OSK epigenetic reprogramming. We determine the duration of the reprogramming effect, demonstrate the validity of a doxycycline (DOX)-inducible dual AAV system, and establish its effectiveness and safety in mice. Our findings establish that epigenetic reprogramming of RGCs is a viable and sustainable approach for recovering lost vision in glaucoma and potentially other age-related disorders.

Materials and Methods

Animals

C57BL6/J wild-type female mice were purchased from the Jackson Laboratory (000664). All animal work was approved by the Institutional Animal Care and Use Committees (IACUCs) at Schepens Eye Research Institute of Mass Eye and Ear according to appropriate animal welfare regulations. Animals were housed under 12-h light–12-h dark cycles (6:00/18:00), at an ambient temperature of 70°F–72°F (21°C–22°C) and 40%–50% humidity.

Animal anesthesia

Mice were anesthetized by intraperitoneal injection of a mixture of ketamine (100 mg/kg, Ketaset; Fort Dodge Animal Health) and xylazine (20 mg/kg, TranquiVed; Vedco, Inc.) supplemented by topical application of proparacaine (0.5%; Bausch & Lomb) followed by a drop of 1% tropicamide (Bausch & Lomb) to dilate the pupil. Genteal (Alcon) was applied to the untreated eye to prevent corneal dryness.

Microbead-induced model of elevated IOP

The elevation of IOP was induced unilaterally by injection of magnetic microbeads (4.5-μm diameter, Dynabeads M-450 Epoxy; Invitrogen) into the anterior chamber of the right eye of each animal under a surgical microscope, as previously reported (Ito et al., 2016; Tan et al., 2022). Briefly, the magnetic beads were prepared at a concentration of 1.6 × 10⁶ beads/μL in phosphate-buffered saline (PBS) and 2.5 μL of the bead suspension was injected into the anterior chamber of the right eyes using a microsyringe pump and a microneedle with a faceted bevel. A hand-held magnet was used to attract the magnetic microbeads to the iridocorneal angle to prevent the drainage of aqueous humor from the anterior chamber. Control mice were injected with an equal volume of sterile saline. Any mice that developed signs of inflammation (clouding of the cornea, edematous cornea, etc.) were excluded from the study.

IOP measurements

IOPs were measured with a rebound TonoLab tonometer (Colonial Medical Supply) as previously described (Lu et al., 2020). Mice were anesthetized with isoflurane in 100% oxygen with a precision vaporizer. IOP measurement was initiated within 2–3 minutes after the loss of a toe or tail pinch reflex. Anesthetized mice were placed on a platform, and the tip of the pressure sensor was placed ∼1/8 of an inch from the central cornea. Average IOP was displayed automatically after six measurements after elimination of the highest and lowest values. The machine-generated mean was considered as 1 reading, and 8–10 readings were obtained for each eye. Following injections, IOP was taken every 3–4 days. All IOPs were taken in the morning between 08:00 and 12:00 owing to the variation of IOP throughout the day.

Optomotor acuity assessment

The visual acuity of mice was assessed using an automated optomotor reflex-based spatial frequency threshold test with the Optodrum (Striatech). The mice were positioned on a pedestal at the center of an area surrounded by four computer monitors arranged in a quadrangle. These monitors displayed a moving vertical black and white sinusoidal grating pattern, which was adjusted based on the mouse's movement to maintain a consistent distance from the pattern. The software used captured the mice's outline, while nose and tail pointers were utilized to automatically evaluate their tracking behavior. Only the bead- or saline-injected eyes were assessed. Tracking behavior was only recorded when the mice were stationary. The contrast level was kept constant at 99.27% and the rotation speed at 12°s⁻¹. The cycle per degree was adjusted using a preprogrammed staircase method. Confirming the final outcome required two positive trials and three negative trials at the next higher (more difficult) cycle per degree.

Pattern electroretinogram

Pattern electroretinogram (pERG) was assessed using the Celeris Pattern Stimulator for rodents (Diagnosys) as previously described (Lu et al., 2020). Briefly, the mice were kept under dim red light throughout the procedure. To maintain their body temperature at 37°C, mice were kept on a built-in warming plate. Both eyes are treated with GenTeal (Alcon) and the light-guided pattern stimulator was placed directly on the ocular surface of the bead- or saline-injected eye. A flash stimulator is placed on the contralateral eye. A black and white reversing checkerboard pattern with a check size of 1° was displayed on the pattern stimulator. The visual stimuli were presented at 98% contrast and constant mean luminance of 50 cd/m², with a spatial frequency of 0.05 cycles per degree and a temporal frequency of 1 Hz. A total of 300 complete contrast reversals of pERG were repeated twice in each eye and the 600 cycles were segmented, averaged, and recorded. The averaged pERGs were analyzed to evaluate the N1–P1 response amplitude.

Optical coherence tomography

Full retinal optical coherence tomography (OCT) scans were taken with Bioptigen Envisu R-Class OCT (Leica Microsystems). The optic nerve was centered, and radial scans were performed on all eyes. Representative OCT images of the retina were taken to measure total retinal thickness at distances of 100–600 μm on both sides of the optic nerve head using ImageJ. The total retinal thickness measurements were taken from the RGC layer to the choroid. Uninjected left eyes of the saline group were used as additional controls.

Hematoxylin and eosin staining and imaging

Mouse eyes were fixed in 10% formalin overnight, embedded in methacrylate and sectioned. Retinal sections were stained with Hematoxylin and Eosin (H&E) by the morphology core at Schepens Eye Research Institute. Stained tissue sections were imaged using a Nikon E800 microscope paired with an Olympus DP70 camera.

AAV2 production and injection

Vectors of AAV-TRE-OSK and AAV-CMV-rtTA were made as described previously (Lu et al., 2020) and then packaged into AAVs of serotype 2/2 (titers: >5 × 10¹² genome copies/mL).

Intravitreal injections

A conjunctival incision was made just anterior to the eye equator using a scalpel. A sclerotomy was performed by advancing a 27-gauge needle through the conjunctival incision. A 33-gauge needle connected to a Hamilton syringe was inserted through the sclerotomy to deliver the AAVs (1 μL) intravitreally. For glaucoma model, intravitreal injection of AAVs was performed 4 weeks after anterior chamber injection of magnetic microbeads or saline.

In vivo DOX induction of OSK

Induction of the Tet-On AAV2 systems in the retina was performed by administration of DOX (1 mg/mL) (USP grade; MP Biomedicals 0219895505) in the drinking water.

Immunofluorescence

Immediately following euthanasia, eyes were enucleated and fixed in 4% paraformaldehyde (Thermo Fisher Scientific) for 2 hours at room temperature. After fixation, eyes were washed in PBS and dissected. Retinas were placed in 200 μL blocking solution (0.5% Triton X-100, 1% bovine serum albumin, 10% donkey serum) in a 96-well plate for 1 hour on the shaker. Retinas were then incubated with 100 μL primary antibodies in blocking buffer (Mouse anti-Brn-3a antibody [MAB1585; Sigma-Aldrich] at 1:200; Goat anti-mouse Klf4 antibody [AF3158; R&D Systems] at 1:100) for 48 hours at 4°C on the shaker. Retinas were washed three times in PBS with 0.3% Triton X-100 (PBST) for 5 minutes each on the shaker. Retinas were then incubated with 100 μL secondary antibodies in blocking buffer [Donkey Anti-Mouse IgG (H+L) Alexa Fluor^® 594 AffiniPure conjugated antibody, 715-585-150; Jackson Immuno Research Laboratories] at 1:800 and Donkey anti-goat IgG (H+L) Alexa Fluor™ 647 conjugated antibody (A-21447; Thermo Fisher Scientific) at 1:500 for 48 hours at 4°C on the shaker. Retinas were then washed three times in PBST for 5 minutes each on the shaker.

After washing, nuclear staining was performed using 4′,6-diamidin-2-phenylindol (1:50) for 20 minutes. Retinas were washed three times in PBST for 5 minutes on the shaker and mounted in Permount mounting media and sealed. The slides were stored in the fridge protected for light. The whole-mount stained retinas were divided into four quadrants and three parts, the center, midperiphery, and periphery (Fig. 2 and Supplementary Fig. S3). Images, three per quadrant and one in each part were taken using the 40 × oil immersion objective of the Leica TCS SP8 confocal microscope system (Leica Microsystems, Wetzlar, Germany). For illustration, maximum-intensity projections of image stacks were created and Klf4 images were despeckled, small particles removed. Channels of Klf4 and Brn3a were merged using Fiji software.

FIG. 2.

Inducibility and stringency of Tet-On AAV system in RGCs. (a) Schematic illustration of the experimental timeline; (b–d) expression of Oct4, Sox2, and Klf4 from the whole retina measured by qPCR (relative to AAV2-TRE-OSK injected eye without DOX treatment) 2 weeks after IVT injection of AAV2-TRE-OSK or AAV2-TRE-OSK; CMV-rtTA and treated with or without DOX; (e, i) representative confocal images from retinal flat-mounts stained with anti-Brn3a (red), an RGC-specific marker, and Klf4 (green). Images from retinal center, midperiphery, and periphery, 2 weeks after IVT injection of AAV2-TRE-OSK; CMV-rtTA and treated +DOX (e) and after additional 3–4 weeks DOX withdrawal (i). Scale bars are 50 μm. (f–h) Expression of Oct4, Sox2, and Klf4 from the whole retina (of which 0.1% cells are RGCs) measured by qPCR (relative to AAV2-TRE-OSK-injected eye without DOX) 2 weeks after IVT injection of TRE-OSK or TRE-OSK-rtTA and treated with or without DOX, followed by 3–4 weeks DOX withdrawal. IVT, intravitreal; qPCR, quantitative PCR; RGC, retinal ganglion cell. *p < 0.05; **p < 0.01.

RNA isolation and complementary DNA synthesis

For RNA isolation from the total retina, eyes were dissected in PBS and placed into RNAlater stabilization solution (Catalog No. AM7022; Thermo Fisher) on ice. RNA isolation was performed using the RNeasy Micro Kit (Catalog No. 74004; Qiagen), according to the manufacturer's protocol. RNA concentration and purity were measured in a spectrophotometer (NanoDrop). Complementary DNA (cDNA) synthesis from isolated RNA was performed using the Invitrogen Superscript IV VILO master Mix Kit (Catalog No. 11756050; Thermo Fisher) according to the manufacturer's instructions. The cDNA was stored at −20°C until quantitative PCR (qPCR) analysis.

Quantitative real-time PCR

Real-time PCR was performed using the PowerUp™ SYBR™ Green Master Mix (Catalog No. A25742; Thermo Fisher) containing SYBR Green DNA dye for quantitative measurement. PCR reactions were set up as duplicates and contained 0.5 μM of each primer, a constant amount of cDNA (10 ng) and PowerUp SYBR Green Master Mix in a total volume of 10 μL. Real-time PCR was performed in five individual experiments using different cDNA samples. Each experiment included nontemplate control, genomic DNA control. The gene expression is depicted relative to AAV2-TRE-OSK without rtTA and DOX treatment. All of the primers used are listed in Table 1. PCR cycles consisted of a denaturation step at 95°C for 2 minutes, followed by 40 cycles of 95°C for 5 seconds, and 60°C for 30 seconds. Each sample was subjected to melting curve analysis to verify the specificity of the amplification. Hypoxanthine–guanine phosphoribosyltransferase (Hprt1), peptidylprolyl isomerase A (Ppia), beta-2 microglobulin (B2M), and heat shock protein 90-beta 1 (Hsp90ab1) were used as reference genes.

Table 1.

Sequences of the Quantitative PCR Primers

Gene symbol	Gene name	Forward sequence (5′-3′)	Reverse sequence (5′-3′)
Klf4	Kruppel-like factor 4	GCGAGTCTGACATGGCTGT	GTTCCTCACGCCAACGGTTA
Oct4	Octamer-binding transcription factor 4	TACTGTGGACCTCAGGTTGGA	CTTTCATGTCCTGGGACTCCTC
Sox2	SRY-box 2	GAAGGATAAGTACACGCTTCCC	GTTGCTCCAGCCGTTCAT
Hprt1	Hypoxanthine phosphoribosyltransferase 1	TCAGTCAACGGGGGACATAAA	GGGGCTGTACTGCTTAACCAG
PPIA	Peptidylprolyl isomerase A	GTCAACCCCACCGTGTTCTTC	ACTTGCCACCAGTGCCATTATG
B2M	Beta-2 microglobulin	TTCTGGTGCTTGTCTCACTGA	CAGTATGTTCGGCTTCCCATTC
Hsp90ab1	Heat shock protein 90-beta 1	TACTCGGCTTTCCCGTCAAG	CCTGAGGGTTGGGGATGATG

The mean of all four reference genes (HPRT, PPIA, HSP90ab, B2M) were used for normalization of all samples. Expression levels were normalized to untreated left eye controls using the ΔΔCT method. Fold changes were calculated with respect to saline-injected control eyes.

Statistical analyses

The Shapiro–Wilk normality test was used to test for the Gaussian distribution of the data. To assess statistical significance in multiple comparisons, in case of parametric datasets an ordinary one-way analysis of variance (ANOVA) was performed, followed by Sidak post hoc analysis when appropriate. In case of nonparametric datasets, a Kruskal–Wallis test followed by a Dunn's post hoc test when appropriate was performed. A p ≤ 0.05 was considered statistically significant, resulting in a 95% confidence level. Longitudinal statistical analysis using R was performed following the visual function of the individual mice over time. Significance levels of visual differences were assessed in comparison to the glaucomatous time point. All data were presented as individual means or grouped values with standard deviation or standard error of mean. Significance levels: ^nsp > 0.05, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, and ****p ≤ 0.0001, respectively. Statistical analysis was performed with GraphPad Prism version 9 for Windows (GraphPad Software, La Jolla) and R (Version 1.4.1106). Illustrations in the figures are prepared using Powerpoint and BioRender.

Results

Eight weeks of inducible OSK expression fully restores vision loss from glaucoma

Glaucoma was induced in C57BL/6J mice by an intracameral injection of magnetic microbeads, resulting in a significant increase of IOP for up to 4 weeks (Fig. 1a, b). The optomotor response (OMR), a reflexive head movement triggered in unrestrained mice exposed to globally rotating striped patterns, was used to assess visual acuity (Fig. 1c). OMR measurements were taken 1 week before (baseline, day −7) and 4 weeks after microbead injection, where as expected, mice displayed a significant decrease in visual acuity due to elevated pressure-induced damage (Fig. 1d). Therefore, for each individual mouse, we know their original “normal” baseline level of visual acuity and the magnitude of vision lost by the induction of glaucoma, before AAV or Dox treatment.

FIG. 1.

Full recovery of visual function through inducible OSK AAVs in microbead-induced glaucomatous animals. (a) Schematic illustration of the experimental timeline. Visual acuity measurements, by OMR, were made at baseline (day −7), after induction of glaucoma (day 28), and after AAV-OSK treatment (week 4 and 8). (b) IOP measurements for the first 4 weeks after microbeads or saline. (c) Schematic illustration of the OMR setup. Reflexive head movements of each mouse were tracked in response to the rotation of a moving striped pattern that increased in spatial frequency (cyc/deg). (d) Optomotor acuity was measured at D-7 (baseline) and 4 weeks postbeads' injection, and 4 and 8 weeks post-AAV treatment in four groups of mice: Saline (nonglaucomatous control); Beads (Tet-ON OSK, no DOX); Beads (Tet-On OSK, plus DOX); Beads (Tet-Off OSK, continuous OSK expression) (n = 15, 7, 8, 10 mice per group). Two experiments each with saline control are combined, see individual batch results in Supplementary Figure S1. Data are presented as mean ± SEM *p < 0.05, ****p < 0.0001. DOX, doxycycline; IOP, intraocular pressure; OMR, optomotor response; OSK, Oct4, Sox2, and Klf4; SEM, standard error of mean.

The glaucomatous damage due to IOP elevation as a result of microbead injections has been shown to be irreversible in the absence of treatments by previous studies (Frankfort et al., 2013; Pan et al., 2023; and Supplementary Table S1). After glaucomatous damage was confirmed (postbead OMR assessment), DOX-inducible OSK vectors containing a 1:3 mixture of AAV2-TRE-OSK, together with either AAV2-tTA (Tet-Off) or AAV2-rtTA (Tet-On) were injected intravitreally. DOX was added to the drinking water to induce AAV2-rtTA and TRE-OSK expression. We have previously shown that AAV2 transduction of RGCs is not toxic and does not affect the OMR response as compared with a PBS control (Lu et al., 2020).

In the AAV2-tTA; TRE-OSK mice (DOX independent, continuous OSK), loss of visual acuity from glaucoma was restored to healthy levels after 8 weeks of treatment (Fig. 1d). In the AAV2-rtTA; TRE-OSK group (DOX-dependent OSK expression), the visual acuity of DOX untreated mice (uninduced OSK) continued to decline due to glaucoma, while the visual acuity in mice treated with DOX (induced OSK) was fully restored to healthy levels by 8 weeks (Fig. 1d and Supplementary Fig. S1 and S2), suggesting the beneficial effect of OSK can be achieved using a DOX-inducible dual AAV system.

OSK expression in RGCs returns to baseline after removal of DOX

To investigate the duration of the reprogramming effect, we utilized the Tet-On system, which allows for the cessation of expression from the AAV vector after removing DOX. However, a comprehensive analysis of the time required for the Tet-On AAV system to shut down in neurons has not been conducted, nor has the extent of expression regression to baseline or the presence of residual expression resulting from leaky binding of rtTA been determined.

To answer these questions, we designed a thorough study where four conditions were included, AAV2-TRE-OSK alone; AAV2-TRE-OSK together with AAV2-rtTA either with DOX treatment or without DOX treatment. These groups were analyzed using qPCR and immunofluorescence staining of retinal flat-mounts 2 weeks after intravitreal injection of OSK vectors +/− DOX and 3–4 weeks post-DOX withdrawal (Fig. 2a).

To induce the OSK expression, DOX was delivered through drinking water. Based on the qPCR analysis, 2 weeks of induction with DOX significantly increased OSK expression, whereas the transcription of these genes returned to baseline 4 weeks post-DOX removal (Fig. 2b–e). Similarly, retinal whole mounts from mice treated with AAV2-TRE-OSK together with AAV2-rtTA stained with the RGC-specific marker Brn3a and Klf4 after 2 weeks of DOX treatment showed that Klf4 protein was degraded 4 weeks after DOX withdrawal at most retina regions (Fig. 2f, g), except a small region close to the injection site where RGCs likely received more AAV copies during injection (Supplementary Fig. S3a–c). The Klf4 protein expression at this small region takes another 2 weeks or more to be degraded (Supplementary Fig. 3d).

Sustained improvement in visual acuity following continuous or cyclic expression of OSK

We and others have shown there is no spontaneous recovery of visual function within 3 months after induction of glaucoma (Supplementary Table S1), but to further confirm this over a longer follow-up period, we first used pERG that is specific for RGCs and measures their electrical activity in response to a changing visual pattern projected on the retina. This assay is highly sensitive to small changes in RGCs activity and is therefore the best method to detect any evidence of “spontaneous” recovery of RGC function after induction of glaucoma. Mice treated with a control AAV2 (empty vector) show no recovery of RGC function as determined by pERG out to 24 weeks post-AAV injection (Supplementary Fig. S2). Since an improvement in electrical activity is necessary (but not sufficient) for a recovery of vision from glaucoma, this basically rules out the possibility of the spontaneous recovery of visual acuity as measured by OMR activity in mice with glaucoma.

To determine the long-term effects of OSK treatment on visual acuity, we performed OMR monthly for 1 year in three groups of mice (i) the saline-injected control group, (ii) the glaucoma group treated with Tet-Off OSK (continuously on OSK), and (iii) the glaucoma group treated with Tet-On OSK (DOX-inducible OSK) (Fig. 3a, d). The level of retinal damage induced by the microbead model of glaucoma is dependent upon the level of elevated IOP, which varies from mouse to mouse. To ensure the different groups of mice received comparable amounts of pressure-induced retinal damage, the mice were distributed between groups based upon their cumulative IOP before receiving the AAV injection of Tet-Off or Tet-On OSK (Fig. 3b, c).

FIG. 3.

Long-term, safe restoration of visual function in mice. (a) Schematic illustration of the experimental timeline for the 1-year vision follow-up. (b) IOP measurements taken by rebound tonometry in mice injected with microbeads or saline. Data are presented as mean IOP ± SD, saline group: n = 10, microbeads for cyclic OSK: n = 8; microbeads for continuous OSK: n = 10. (c) The cumulative IOP was significantly elevated in mice that received microbeads as compared with control mice receiving saline. No difference of cumulative IOP elevation between the bead-treated groups (cyclic vs. continuous OSK). (d) A schematic illustration of the continuous (tTA; TRE-OSK—green) and the dox-inducible cyclic (rtTA; TRE-OSK—orange) AAV vectors. (e) Optomotor acuity follow-up with the nonglaucomatous (saline) and glaucomatous (beads) eyes, either treated with cyclic OSK or continues OSK AAVs. (f, g) Longitudinal statistical analysis of visual acuity in the continuous (f) or cyclic (g) OSK-treated groups comparing the visual acuity at the damage baseline (glaucoma) up to 12 months of treatment. SD, standard deviation. Data are presented as mean ± SEM. *p < 0.05; **p < 0.01; ***p < 0.001.

To determine the long-term effects of continuous OSK expression, we maintained OSK expression in the Tet-Off group for 12 months (Fig. 3e, f). One of the advantages of in vivo monitoring of visual acuity through OMR is that we can sequentially assess vision in the same mouse over the course of the experiment. In other words, we can determine visual acuity at baseline before the experiments start, after glaucomatous retinal damage has occurred (4 weeks postbead injection and before starting OSK treatment), and then monthly during treatment. In the Tet-Off OSK group (continuously on OSK), this approach revealed the peak improvement in visual acuity after 3 months of OSK treatment as compared with immediately following 4 weeks of elevated IOP (Fig. 3f). Remarkably, the improved visual acuity was even significantly better than the baseline level of vision before the experiment was started. Continued monthly assessment of visual acuity demonstrated the significant improvement in vision was sustained for 11 months (Fig. 3f and Supplementary Fig. S4a, b).

At the final acuity measurement at 12 months, the age of the mice was over 15 months old, a time when mice display an age-related loss of retinal cell function that coincides with an overall decline in vision. Therefore, it is highly possible that at this time, visual acuity is declining, not due to the loss of the OSK reprogramming effect on RGCs, but due to the age-related decline of other retinal layers (e.g., cornea, photoreceptors, retinal pigment epitheliums) that did not receive OSK treatment (Ferdous et al., 2021).

Similar to the Tet-Off OSK group (continuously on OSK), the Tet-On OSK group (DOX-inducible OSK) also displayed a significant improvement in visual acuity when compared with immediately following 4 weeks of elevated IOP and this improvement was observed as early as 1 month after initiating treatment (Fig. 3g). At 2 months postinitiation of treatment, when improved visual acuity reached a healthy level, we removed DOX from the drinking water to determine the duration of improved vision once OSK expression was stopped. After switching off OSK, visual acuity gradually diminished. It is possible that this reduction can be prevented if we turned off OSK expression after 3 months instead of 2 months, between which there was a further improvement. After the 8-month follow-up data of the mice, it became evident that continuous OSK treatment yielded greater improvements in visual function as compared with the initial 2-month DOX-induced OSK treatment. We then reintroduced DOX to this group at 9 months post-AAV treatment and their visual acuity increased to the level of healthy control eyes and glaucomatous eyes receiving continuous OSK expression (Fig. 3e).

Although not reaching statistical significance, six out of eight mice had vision improved during the 2-month OSK reinduction (Supplementary Fig. S4c, d), against the normal aging trend.

The long-term functional tracking of glaucomatous eyes treated with OSK treatment provides encouraging results. It demonstrates that OSK-mediated epigenetic reprogramming rejuvenates dysfunctional RGCs and improves visual acuity, which is maintained at a healthy level for a prolonged period of time. Furthermore, the improved vision is directly associated with the expression of OSK.

Long-term OSK treatment of glaucoma does not cause tumors or changes in retinal structure

In our previous study (Lu et al., 2020), we provided evidence demonstrating that continuous expression of OSK in RGCs of healthy, naive mice for up to 15 months did not result in tumor formation or any structural changes in the retina. However, it remained unknown whether continuous OSK expression would remain safe in a disease setting. Therefore, we examined the long-term effects on health and vision of continuous and cyclic OSK expression in the retina of glaucomatous mice.

Twenty-one months after intravitreal injections of AAV for continuous (tTA; TRE-OSK) and cyclic (rtTA; TRE-OSK) OSK expression (Fig. 4a), the mice exhibited no difference in body weights between the groups, indicating that long-term OSK expression in the eye does not cause severe disease and obvious tumors (Fig. 4b). Importantly, expression from the AAV is still strong after 21 months postinjection (Supplementary Fig. S5).

FIG. 4.

Retina health and animal body weight at 21 months post-AAV treatment. (a) Schematic illustration of the experimental timeline and endpoint measurements. (b) Body weight (g) measured at 21 months post-AAV treatment revealed no significant differences between the saline and beads-injected AAV-injected groups. (c) Representative retinal cross-section B-scan images (retinal layers: GCL, ganglion cell layer; INL, inner nuclear layer; ONL, outer nuclear layer and choroid). (d) Quantification of total retinal thickness shows no changes in retinal structure and no tumor development in any group (n = 4 uninjected eye, n = 4 saline injected, n = 6 continuous OSK, n = 6 cyclic OSK). (e) Representative H&E-stained retinal cross-sections (retinal layers: GCL; INL; ONL; PR, photoreceptor; RPE, retinal pigment epithelium and choroid). Scale bar = 20 μm. H&E, Hematoxylin and Eosin.

To assess the ocular health implications of long-term OSK expression, we performed OCT. OCT is a noninvasive and highly sensitive diagnostic technique that provides in vivo cross-sectional images of the entire retina. Analysis of the complete retina revealed no discernible changes in retinal structure (Fig. 4c). Furthermore, quantification of total retinal thickness showed no significant differences between the control groups (uninjected, naive eyes and saline-injected eyes) and the groups injected with beads+AAV (Fig. 4d). Histological examination of H&E-stained retinal sections showed no detectable change in retinal architecture (Fig. 4e).

In summary, sustained OSK expression over the long term does not induce alterations in retinal thickness or present any discernible evidence of retinal tumor development or malignant transformation, as indicated by OCT measurements. Furthermore, no observable changes in retinal architecture were identified through histological examination.

Conclusion and Future Directions

In our previous work, we developed a novel form of gene therapy using three reprogramming factors, OSK that reversed the age of RGCs by restoring youthful transcriptome and methylome patterns. Importantly, this rejuvenating effect allowed the RGCs to recover functions that were lost with aging, such as axon regeneration and visual acuity in mice, or that were lost in a NAION disease model in non-human primates (Ksander et al., 2023).

In the current study, we performed year-long functional tracking of mice with experimentally induced glaucoma that were treated with AAV2-OSK to induce epigenetic reprogramming in RGCs. Our results indicated that vision loss due to glaucoma can be safely reversed. In the AAV2-tTA; TRE-OSK mice (OSK continuously on), vision loss was restored to healthy levels after 2 months of treatment and remained close to healthy levels for 11 months. In the AAV2-rtTA; TRE-OSK group (DOX inducible OSK), for negative control mice that received no DOX/no OSK expression, their vision continued to decline; however, mice that received DOX that triggered OSK expression restored vision to healthy levels by 8 weeks, suggesting that the beneficial effects of OSK treatment can be achieved using a DOX-inducible dual AAV system. When DOX was withdrawn and OSK turned off in these mice, although significance was compromised compared with continuous ON group, the absolute vision performance stays higher than its glaucoma baseline until 11 months post-AAV injection.

At the 12-month follow-up, the mice had reached 15–16 months of age, which may represent the upper limit at which reprogramming solely RGCs may be sufficient for restoring visual function. As part of the aging process, all retinal cells undergo age-related functional decline. In our previous study, we demonstrated that OSK reprogramming effectively improved age-related vision loss in 12-month-old mice, but not when they reached 18 months of age (Lu et al., 2020).

Our hypothesis is that OSK epigenetic interventions hold promise for rejuvenating RGC function compromised by both aging and glaucoma. Nevertheless, considering that we reprogrammed RGCs alone and the broader age-related functional decline affecting various other retinal cell types, it is probable that there is an age threshold beyond which exclusive RGC reprogramming may fall short of restoring visual function. This same challenge may potentially occur in glaucoma patients, where advancing age could eventually render RGC reprogramming insufficient for countering the age-related dysfunction in other retinal cell types. To address this limitation, a promising avenue could involve targeting and revitalizing the function of additional retinal cell populations to achieve more comprehensive visual restoration.

Importantly, there were no adverse effects observed in mice expressing OSK over the entire course of the year-long experiment. This further indicates that in vivo epigenetic reprogramming holds great therapeutic potential in humans, not only for glaucoma but also for other age-related diseases of the retina, such as nonarteritic anterior ischemic optic neuropathy, age-related macular degeneration, and diabetic retinopathy, as well as nonocular age-related diseases in other tissues.

While we cannot rule out that OSK reprogramming caused some nonrejuvenating off-target transcriptome changes, the visual acuity and histological data suggest the RGC identity and function is largely preserved even with continuous OSK expression. We have previously shown that OSK-expressing RGCs do not proliferate and remain a postmitotic feature (Lu et al., 2020). Do postmitotic neuronal cells possess a natural barrier that blocks reprogramming from progressing further? Reik group reported that enhancers of cellular identity genes remain hypomethylated during partial reprogramming, which allows the cells to swiftly regain their identity once reprogramming factors are withdrawn (Gill et al., 2022). It is possible that without cellular division, the methylation status of promoter and enhancer regions of cellular identity genes is very persistent and therefore highly resistant to the effects of reprogramming genes. This would suggest that postmitotic RGCs and other neuronal cells of the central nervous system offer a perfect system to examine the maximum rejuvenation effect of reprogramming factors.

OSK-mediated epigenetic reprogramming of young mice with glaucoma enhanced RGC cellular functions and restored visual function back to normal levels after 2 months of treatment. It is surprising that a retinal injury in a young mouse responds so well to an age-reversing epigenetic reprogramming therapy. Our previous work showed that when retinal neurons from young animals experience an acute physical injury, their global DNA methylation patterns shift to resemble the pattern seen in aged animals, with a corresponding increase in DNA methylation age and a loss of sensory function (Lu et al., 2020). The fact that the DNA demethylation enzymes Tet1 and Tet2 are essential for functional rejuvenation points to the critical role of restoring a youthful DNA methylome (Lu et al., 2020).

Even less clear is how OSK-induced epigenetic rejuvenation is capable of identifying age-associated methylation sites, while avoiding methylation sites not associated with aging. We hypothesize there are DNA and/or chromatin marks that facilitate a gene expression reset, consistent with the Information Theory of Aging (Lu et al., 2023). Deciphering the rejuvenation mechanism of OSK-mediated epigenetic reprogramming requires a careful profiling of epigenetic modifiers and epigenetic events that occur during reprogramming and validating each site to determine whether it is necessary through molecular or functional readouts.

OSK are transcription factors that primarily function within the nucleus and can be delivered to specific tissues using viral vectors, yet they are not optimal for achieving comprehensive rejuvenation throughout the entire body due to viral tropism, resulting in uneven distribution among internal organs, as shown by our recent demonstration of age-reversal in muscle and kidney through whole-body AAV-OSK IV injection (Yang et al., 2023). Secretory factors and chemicals that promote rejuvenation would have a natural advantage by being able to access multiple organs through the bloodstream. It would be a significant breakthrough if epigenetic reprogramming could be achieved using secretory factors present in young blood, or through chemicals, such as those used to make induced pluripotent stem cells (iPSC). Our results indicate that the eye, particularly in the context of a glaucoma model, represents an ideal system and condition to evaluate the efficiency and safety of novel interventions to rejuvenate tissues.

Authors' Contributions

B.R.K. conceived the study with D.A.S., M.M.K., Y.R.L., J.M., J.M.C., M.W., K.B., and S.R.-L., M.M.K., E.H., C.H., M. Shrestha. Y.R.L. performed IOP, OMR, and OCT measurements. M.G-K. supervised the microbead-induced glaucoma studies and M. Shrestha and Y.G. performed the microbead injections and intravitreal injections of AAV. N.R. performed the Tet-ON system inducibility study with S.S., M.M.K., Y.R.L, and J.C. performing the OCT analysis. N.R. conducted the qPCR and confocal imaging. D.K. imaged retinas H&E sections. A.K. performed the pERG. M.M.K., M. Shah, and Y.L. created illustration. Y.R.L., M.M.K., N.S., D.A.S., and B.R.K., co-wrote the article with input from all authors.

Author Disclosure Statement

D.A.S. is a consultant, board member, or shareholder in the following companies, or is an inventor of IP owned or licensed by such companies: Life Biosciences (developing reprogramming medicines), InsideTracker, Zymo, EdenRoc Sciences/Cantata/Metrobiotech, Galilei, Immetas, Animal Biosciences, and Tally Health. For additional details see https://sinclair.hms.harvard.edu/david-sinclairs-affiliations. Y.R.L., and D.A.S. are inventors of patent applications licensed to Life Biosciences, in which Y.R.L. and D.A.S. have equity and D.A.S. sits on the board. B.R.K. receives funding from Life Biosciences for this work. J.M.C., M.W., K.B., and S.R.-L. are past or current employees of Life Biosciences and have equity.

Funding Information

Margarete M. Karg was supported by Walter Benjamin Postdoctoral Fellowship (German Research Council). Yuancheng Ryan Lu was supported by AFAR Glenn Postdoc Fellowship (from Michael Shen, MIT’13) and Life Science Research Foundation Postdoc Fellowship (from Ming Lei—Chang Luo Life Science Fund). Bruce R. Ksander was partially supported by Life BioScience Sponsored Research Agreement. Meredith Gregory-Ksander was supported by NEI award R01EY032762. NEI Core Grant P30EYE003790 supports the SERI Morphology Core and SERI animal facility. Animal work was supported by Life Biosciences.

Supplementary Material

Supplementary Figure S1

Supplementary Figure S2

Supplementary Figure S3

Supplementary Figure S4

Supplementary Figure S5

Supplementary Table S1