EA042590B1 - GENE THERAPY FOR IMPROVED VISION - Google Patents

Description

In many mammalian species, including mice and humans, the number of rod photoreceptors that provide vision in dim light conditions greatly exceeds the number of cone photoreceptors (Curcio et al., 2000). However, in the industrialized world, where lighting allows the cones to function throughout the day, the vision provided by the rods is less important. Many patients with non-functioning rods from birth are discovered only incidentally and, in fact, cannot recognize their abnormal vision (Dryja, 2000). In contrast, with cone dysfunction, patients are always symptomatic and often suffer visual impairment depending on the degree of their cone dysfunction. However, in some conditions only (or mostly) the cones are inactive or dysfunctional, while the rods remain relatively intact. For example, achromatopsia is a severe hereditary retinal dystrophy with completely non-functioning cones from birth, but presumably normally functioning rods (Hess et al., 1986; Nishiguchi et al., 2005). An association has been found between mutations in several genes, including CNGA3 (Kohl et al. 1998); and PDE6C (Chang et al., 2009; Thiadens et al., 2009) and disease. Each of the disease-causing genes encodes an important component of the cone phototransduction cascade, which converts light into an electrical signal, causing hyperpolarization of the photoreceptor cell. In age-related macular degeneration (AMD), visual impairment is caused primarily by degeneration in the macula fovea, which contains a large number of cones. Thus, patients lose central vision and visual acuity, but often have a relatively well-preserved peripheral macula and therefore have some useful residual vision that is limited to a small number of cones outside the fovea. Rods are very sensitive to light, which allows them to perceive a small amount of light in twilight conditions. Cones, on the other hand, are less sensitive, but are able to process large amounts of light and continuously transmit visual signals in daylight.

This functional difference is partly due to the efficiency of the photosignal deactivation apparatus, the GTPase complex, consisting of RGS9, R9AP (also known as RGS9BP) and Ge5. RGS9 is the catalytic component that hydrolyzes GTP coupled to the G protein, while R9AP and Ge5 are the main constituent subunits (Burns et al., 2009; Burns et al., 2010). Importantly, R9AP binds the complex to the disc membrane in the outer segment of the photoreceptor, where phototransduction signaling also occurs (Baseler et al., 2002). Expression of R9AP determines the amount of the GTPase complex, so that RGS9 produced in excess of R9AP is likely to be rapidly cleaved (Martyemyanov et al., 2009). Overexpression of R9AP in mouse rods is sufficient to increase GTPase activity and, in fact, to accelerate their deactivation kinetics, as evidenced by individual cell signaling (Krispel et al., 2006). RGS9 expression in cones has been estimated to be 10 times higher than in rods (Cowan et al., 1998; Zhang et al., 2003). This underlies the ability of cones to quickly recover from exposure to light and thus remain functional under continuous light stimulus. It also allows the cones to respond to faster stimulation. Indeed, from a clinical point of view, patients with delayed phototransduction cascade deactivation caused by genetic defects in mRGS9 or R9AP, i.e., bradiopsic, have severe impairment of cone-mediated vision, including day blindness and reduced ability to see moving objects Nishiguchi et al., 2004; Michaelides et al., 2010). At the same time, vision provided by rods is less affected by the same mutation.

Some macular degeneration diseases, such as age-related macular degeneration (AMD) and hereditary macular degeneration diseases, also show cone dysfunction but normal or less impaired rod function. Macular degeneration is the leading cause of blindness in the developed world, and as its incidence quadruples every decade of life, the incidence of MPAD is expected to rise in the coming years as life expectancy increases. Drugs for the treatment of ADHD already account for more than 1% of the total drug budget of the UK National Health Service. While patients with progressive VAD can be trained to fix outside the fovea, the slow recovery rate and low bleaching threshold (bleaching is the isomerization of 11-cis-retinal to trans-retinal leading to a conformational change in rhodopsin and subsequent release of trans-retinal, while the photopigment loses color) rods limit the quality of the resulting vision.

- 1 042590

The essence of the invention

Using cone-deficient mice, we demonstrated that AAV-mediated overexpression of the RGS9 anchor protein (R9AP), a critical component of the GTPase complex that mediates deactivation of the phototransduction cascade, results in photodesensitization and a photopic shift in the rod-generated electroretinogram. Treatment allows rods to respond to brighter light (up to ~2.0 log) than untreated cells, at the expense of scotopic function (twilight vision). Chip-multielectrode measurements using treated retinas showed that retinal ganglion cells also displayed photopic rod shift, showing stepped responses at photopic (daylight) light levels. The contrast sensitivity function, measured by quantifying head tracking movements in response to rotating sinusoidal gratings, showed a 25-fold improvement in sensitivity under indoor lighting conditions and a faster response to repetitive stimuli. In addition, biochemical measurement of the amounts of decolorized rhodopsin in these mice indicated that the rhodopsin conversion cycle did not limit rod function.

We also expressed the ArchT light-driven proton pump (Han et al., 2011) in rod photoreceptors. AAV8 particles carrying ArchT-EGFP under the control of the rhodopsin (Rho) promoter were administered subretinally to adult mice. Rho-ArchT-EGFP expression was limited to the rod photoreceptor membrane. ArchT expression allowed extremely fast light responses, while the rods' own response was maintained and was comparable to that observed in non-transfected rod photoreceptors. Overall, ArchT expression did not alter the ability of rod photoreceptors to respond to scotopic stimuli, but it conferred additional ability to respond through rapid, non-bleaching responses to higher light levels. We also found that transduced rods were able to reliably withstand this fast cone-like signaling, as ArchT-expressing rods triggered sustained retinal ganglion cell (RGC) firing at high light intensities and at frequencies approaching frequencies of cone photoreceptors. Expression of Rho-ArchT-EGFP in non-cone vision CNGA3-/- and PDE6C-/- mice also increased the sensitivity of these mice to bright light exposure and provided these mice with fast vision.

The maximum exposure frequency that ArchT-expressing mice could detect was the same as for cone photoreceptors.

Together, these results show that after transduction of healthy rod photoreceptors with genes encoding or light-sensitive proteins characteristic of cones or with genes encoding molecules that increase the rate of endogenous rod signaling, rods behave more like cones and therefore can compensate cone dysfunction. This has implications for the treatment of a number of visual impairments in which cone function is reduced but at least some healthy rods are retained. This is in stark contrast to previous approaches (Busskamp et al., 2010; US Patent Publication No. 2012258530) where the goal was to restore lost cone function. Modifying rod function in accordance with the present invention is more beneficial because cases where cones are non-functional but can be restored (e.g., in the early stages of retinal degeneration when photoreceptor function is lost but the photoreceptor-bipolar synapse may be intact) are rare, then as cases where cone dysfunction is more severe or worsened and cannot be restored, or where cones are completely lost but at least some healthy rod photoreceptors are retained, are common (see above). In addition, the present invention allows the creation of a pseudo-fovea, a small area of cone-like rods, thereby improving vision in conditions where foveal cones have been lost or are dysfunctional.

Thus, the invention provides a vector containing a nucleic acid encoding a gene product that is photosensitive and/or enhances endogenous photosensitive signaling in a photoreceptor cell, for use in a method for improving vision in a patient with dysfunction and/or degradation of cone photoreceptors, by introducing said nucleic acid into healthy rod photoreceptors in the patient's retina and expressing said gene product therein such that the range of light intensities to which the rod photoreceptor responds is increased and/or the rate at which the rod photoreceptor responds is increased into the world.

The invention also provides a vector as defined above, a host cell containing said vector, and methods of treatment carried out using such a vector.

- 2 042590

Brief description of graphic materials

Fig. 1: ArchT expression in rod photoreceptors results in fast, light-induced currents.

A) Top panels: AAV8-mediated transduction of ArchT-EGFP (green, see left panels) under the control of the rhodopsin promoter. No match with cones was observed (purple: cone arrestin, see middle panels, white: DAPI, see right panels). Lower panels: Expression specificity is also observed at the level of synapses.

B) ArchT-EGFP is localized on the membrane of rod photoreceptors, including inner and outer segments.

C) Fluorescence quantification at rod ends expressing ArchT-EGFP (green, left peak) and cone arrestin-positive cone ends (purple, right peak) shows two distinct bands corresponding to the sublayer where rod and cone synapses are located, respectively (n =22).

D) Recordings from single cells, from cell bodies of ArchT-expressing rod photoreceptors. The currents mediated by intrinsic phototransduction of the rods (top curve) in response to 10 ms 530 nm light pulses (green, see vertical bars) were preserved. The currents generated by the ArchT were faster (lower trace). Scale bars: a) top panels: 50 µm; a) bottom panels and b): 10 µm.

Fig. 2: ArchT expression elicits high frequency responses in rods and rapid signal transduction to retinal ganglion cells.

A) Rod-specific, light-induced currents in uninjected retinas C57BL6.

B) Currents mediated by ArchT are able to follow much higher stimulation frequencies.

C) Responses are time-limited by the stimulus (green vertical bars).

D) ArchT-expressing rods respond to stimulation rates up to 80 Hz without interruption, while rod-specific responses subside at -20 Hz.

E) Summary data showing that ArchT expression does not alter the intrinsic response of rod photoreceptors, while ArchT-mediated responses start at brighter light levels.

E) Chip-multielectrode recordings of ArchT-expressing PDE6C-/- retinas. Rod-specific responses failed to reliably generate retinal ganglion cell impulses above 20 Hz. In contrast, ArchT-mediated rod activation resulted in rapid generation of retinal ganglion cells to levels comparable to cone photoreceptors.

Fig. 3: ArchT-mediated rod activation triggers a behavioral response to high intensity light stimuli.

A) Upper panels: a schematic drawing for the development of conditioned fear reflexes. Briefly, a visual stimulus was combined with an electrical shock. 24 h later, the fading behavior was tested under new conditions. Lower panels: uninjected CNGA3-/- and PDE6C-/- mice failed to learn the task (left-hand band sets in each plot). However, ArchT expression successfully triggered freezing behavior in mice (right-hand band sets in each graph).

B) Optomotor testing. ArchT-expressing mice were able to track stimuli at frequencies comparable to those at which stimuli are reliably tracked by cones.

Fig. 4: AAV-mediated R9AP overexpression in rods and accelerated awave deactivation in Cnga3-/- mice.

A. Increased expression of RGS9 in Cnga3-/- eye treated with rRAAV2/8.Rho.mR9ap. Overexpression of R9AP leads to increased immunoreactivity towards RGS9 (red) across the entire photoreceptor layer in the treated eye (left) compared to the untreated eye (right). Western blotting demonstrates overexpression of RGS9 in both the retina and retinal pigment epithelium (RPE) with overexpression of R9AP (bottom) in the eye. A small amount of the RGS9 protein was also found in the RPE of the treated eye. This may reflect the controlled release of excess protein contained in the phagocytosed disc membrane. The scale bar indicates 25 µm.

B. Increased rate of a-wave amplitude recovery in the Cnga3-/- eye treated with rAAV2/8.Rho.mR9ap and rAAV2/8.CMV.mR9ap. Representative ERG curves for the test flash (black curves, see curves with a peak in the middle of the time period) and for the 2nd flash (red curves), presented in intervals between stimuli (CMI) for 2 s, from the treated (upper) and untreated (lower) eyes from the same animal. Note that the second flash produces a small a-wave (arrow) that is clearly visible in the treated eye, while no a-wave is visible (arrow) in the untreated eye (the second of the pair). Graph of α-wave recovery for various UTIs in treated and untreated eyes. Eyes injected with rAAV2/8.CMV.mR9ap (n = 5) or rAAV2/8.Rho.mR9ap (n = 7) have faster recovery kinetics than untreated eyes (n = 5), which is most noticeable with more short

- 3 042590

IC. Data are presented as mean ± standard error of the mean. IE: overexpression.

Fig. 5: Acquisition of photopic function by rods via R9AP overexpression in Cnga3-/- mice.

A. Threshold elevation and 6 Hz ERG photopic shift by R9AP overexpression in rods of Cnga3 -/- mice. Representative 6 Hz ERG curves from a Cnga3-/- mouse treated with rAAV2/8.CMV.mR9ap in one eye and untreated in the other eye (upper panel). The ERG curves are leveled from the responses to the weakest flash (-6.0 log cd s/m ² ) to the responses to the brightest flash (2.0 log cd s/m ² , bottom) from top to bottom with a step of 0, 5 log cd s/m ² . Note that the lower threshold for flare intensity at which responses occur has increased, due to the higher threshold for flare responses for brighter flares. This results in a photopic shift in retinal function in the rAAV2/8.CMV.mR9ap treated eye. Summary of 6 Hz EGR results demonstrating a photopic shift in retinal function after treatment with rAAV2/8.CMV.mR9ap or rAAV2/8.Rho.mR9ap (bottom panel). ERG responses obtained from rod-deficient Gnat1-/- mice represent cone-mediated function. Meanwhile, responses from C57BL6 mice have been obtained for both rod and cone photoreceptors. Data are presented as % amplitude relative to the maximum response and are presented as mean ± standard error of the mean. ERGs were recorded for Cnga3 -/- mice treated with rAAV2/8.CMV.mR9ap (Cnga3 -/- CMV.R9ap; N=8), Cnga3 -/- mice treated with rAAV2/8.Rho. mR9ap (Cnga3-/- Rho.R9ap; N=6), untreated Cnga3-/- mice (untreated Cnga3-/-, N=8), untreated Gnat1-/- mice (Gnat1-/- untreated, N=6) and untreated C57BL6 mice (C57BL6 untreated, N=6).

B. Increased retinal responses to long flashes in a Cnga3-/- eye treated with rAAV2/8.CMV.mR9ap. The unfilled box indicates the duration of the flash. Note that in the eye treated with rAAV2/8.CMV.mR9ap, responses are found with increasing light stimulus duration. In contrast, the untreated contralateral eye exhibits little or no response when recording at the same time under identical conditions.

B. Enhancement of retinal function under photopic conditions in rAAV2/8.CMV.mR9ap treated Cnga3-/- eyes. Note that the treated eye shows responses under photopic recording conditions (white background light 20 cd/m ² ), while the untreated contralateral eye, recorded simultaneously, remains unresponsive.

Fig. 6: Efficient transmission of altered photoreceptor signal to bipolar cells in eyes overexpressing R9AP.

A. Representative ERG curves. ERG was recorded after injection of rAAV2/8.CMV.mR9ap (red curve, lower curve) in Cnga3 −/− mice using a saturating flash (1.9 log cd s/m ² ). The contralateral eye served as the untreated control (black curve, upper curve).

B. Delayed activation of bipolar cells in response to a flash. The time to peak of the a-wave and b-wave was measured by ERG responses to a saturating flash (1.9 log cd s/m ² ) in treated and untreated eyes.

B. Response intensity curve for a-wave and b-wave amplitudes recorded in Cnga3-/- mice (N=5) with one eye treated with rAAV2/8.CMV.mR9ap (red curves) and the other the eye that remained untreated (black curves). All data are presented as mean ± standard error of the mean. IE: overexpression.

Fig. 7: Acquisition of stable visual perception after R9AP overexpression in Cnga3-/- mice.

A. Improved contrast sensitivity function as measured by optokinetic responses. In Cnga3-/- mice whose left eye was treated with rAAV2/8.Rho.mR9ap, contrast sensitivity function (CSF) was differentially measured by head tracking clockwise (representing the treated left eye) and counterclockwise (representing untreated right eye) versus sinusoidal gratings. The FCF for treated eyes (red curve) was better than that for untreated eyes (blue curve), which was similar to that of intact Cnga3 −/− mice (black curve, mean of both eyes). Note that the FCF for the treated eye is equivalent if not slightly better than that of intact wild-type controls (green, average of both eyes). N=5 for all groups. All data are presented as mean ± standard error of the mean. IE: overexpression.

B. Steady levels of rhodopsin after prolonged exposure to the optomotor test. Representative optical absorption recording of an ocular sample using a scanning spectrophotometer (left panel within the green dotted box). Subtraction from absorbance of eye samples measured after complete photobleaching (blue curve, upper curve between 300 and 400 nm),

- 4 042590 absorbance of ocular samples measured to complete photobleaching (red curve) showed that λmin peaks at about 380 nm, which corresponds to the released photoproducts, in combination with λmax peaks at about 500 nm, which is the amount of bleached rhodopsin in the sample (right panel inside green dotted box). Rhodopsin decolorization rate was assessed by measuring the difference spectrum (λmax) in fully dilated, treated or untreated eyes with rAAV2/8.Rho.mR9ap in Cnga3-/ mice after 5 min exposure to 7.0 mW white light (lower left, mean ± standard error of the mean). The amount of rhodopsin was also measured after exposing Cnga3 −/− mice to an optomotor test for up to 120 minutes (bottom right; N=3 for each time point) following unilateral injection of rAAV2/8.Rho.mR9ap. The gray area shows the amounts of rhodopsin (mean ± SD) recorded in intact Cnga3-/- mice (N=8) after overnight dark adaptation. The dotted line indicates the mean value. Note that the amount of rhodopsin remains stable for at least 2 h in both rAAV2/8.Rho.mR9ap-treated and untreated eyes of Cnga3-/- mice. Data with error bars were presented as mean ± standard error of the mean.

Fig. 8: R9AP overexpression increases the rate of rod photoresponse recovery in Pde6c-/- mice.

The time constant (σ) for 50% a-wave amplitude recovery was reduced by ~50% in Pde6c-/- eyes injected with rAAV2/8.CMV.mR9ap (σ = ~5.75 s) compared to untreated contralateral eyes (σ = ~11.46 s), which is consistent with accelerated deactivation of phototransduction after treatment. N=6. Data with error bars were presented as mean ± standard error of the mean.

Fig. 9: R9AP overexpression results in a photopic shift in the response intensity curve in Pde6c-/- mice.

Eyes injected with rAAV2/8.CMV.mR9ap showed a photopic shift of 6 Hz ERG responses in response to increasing flash intensities compared to untreated contralateral eyes in Pde6c −/− mice (N=6). Data are presented as % amplitude relative to the maximum response and are presented as mean ± standard error of the mean.

Fig. 10: Consistent effect of R9AP overexpression without clear evidence of retinal degeneration 5 months after rAAV2/8.CMV.mR9ap injection in Cnga3-/- mice.

The response profile, normalized to the maximum amplitude, confirmed the presence of a photopic shift in the response intensity curve in response to the 6 Hz flare (top). The same data without normalization did not provide evidence of amplitude reduction in the treated eye (bottom). N=5. Data are presented as mean ± standard error of the mean.

Fig. 11: Treatment of wild-type mice with rAAV2/8.Rho.mR9ap showed no obvious effect on the 6 Hz ERG response intensity curve.

The rAAV2/8.Rho.mR9ap treated eyes did not show a 6 Hz ERG shift in the response intensity curve compared to that of the untreated contralateral eyes in C57BL6 mice (N=5). Data are presented as mean ± standard error of the mean.

Fig. 12: No enhancement of visual perception after R9AP overexpression in C57BL6 mice.

In C57BL6 mice whose left eye alone was treated with rAAV2/8.Rho.mR9ap, contrast sensitivity function (CSF) was differentially measured by clockwise (representing treated left eye) and counterclockwise (representing untreated) head tracking movements. right eye) against sinusoidal gratings. HFC for both treated eyes (pink curve) and untreated eyes (blue curve) showed similar results, which were similar to those for intact C57BL6 mice (green curve, mean of both eyes). N=5 for all groups. All data are presented as mean ± standard error of the mean. IE: overexpression.

Detailed Description of the Invention

The vector of the invention contains a nucleic acid, the expression of which produces a gene product, typically a protein, that will carry out the treatment as described herein, operably linked to a promoter to form an expression cassette.

Nucleic acids and gene products.

The vector of the invention contains a nucleic acid encoding a gene product that is photosensitive and/or that alters endogenous photosensitive signaling in a photoreceptor cell and causes a rod transduced with the nucleic acid of the invention to behave more like a cone by expanding the range of light intensities to which the rod photoreceptor responds, and/or an increase in the rate at which the rod photoreceptor responds to light. Thus, the protein itself can be directly photosensitive, for example, it can change the conductivity of the membrane in rods in such a way that this leads to

- 5 042590 hyperpolarization (outgoing current flow) during light stimulation. Such proteins may be, for example, light-sensitive or light-regulated G-linked membrane proteins, ion channels, ion pumps or ion transporters. Preferred photosensitive proteins include ArchT, Jaws (cruxhalorhodopsin) (Chuong et al., 2014) and iClC2. Alternatively, the protein itself may not be directly sensitive to light, but may indirectly alter endogenous light-sensitive signaling in the rod photoreceptor. Examples of such proteins are elements of the RGS9 complex, in particular R9AP (also known as RGS9BP) and GB5. Alternatively, the nucleic acid may encode any other gene product that increases the rate of endogenous rod signaling. In all these cases, the sequence may encode a wild-type protein, or a mutant or modified or truncated version that retains the activity of the wild-type protein. The nucleic acid can also be codon-optimized for expression in the target cell type.

Upon expression of the gene product, the rods will show a stronger and/or more rapid change in response to light exposure than non-transduced rods at higher than normal light intensity. Examples include improved rate of change and/or faster activation/inactivation kinetics. Therefore, rods transduced in accordance with the invention will therefore respond more strongly and/or rapidly to illumination in the mesopic and/or photopic range than non-transduced rods. Preferably, the response of rods to scotopic light conditions is unaffected or not significantly affected, i. e. rods gain the ability to respond strongly and/or quickly to brighter light without losing the ability to respond to dim light.

Promoters and other regulatory elements.

In an expression construct, the nucleic acid encoding the gene product is usually operably linked to a promoter. The promoter may be constitutive, but is desirably a photoreceptor-specific or photoreceptor-preferred promoter, more preferably a rod-specific or rod-preferred promoter such as the rhodopsin (Rho) promoter, the retinal neuron-specific leucine clasp protein (NRL) promoter, or phosphodiesterase 6B (PDE6B) promoter. The promoter region included in the expression cassette may be of any length as long as it is effective in directing expression of the gene product, desirably photoreceptor-specific or photoreceptor-preferred expression, or rod-specific or preferred expression.

By photoreceptor-specific promoter is meant a promoter that drives expression only or to a greater extent only in photoreceptors, such as one that enhances expression at least a hundredfold in photoreceptors compared to any other cell type. By rod-specific promoter is meant a promoter that drives expression only or to a greater extent only in photoreceptors, such as one that enhances expression at least a hundredfold in photoreceptors over any other cell type, including cones. By photoreceptor-preferred promoter is meant a promoter that drives expression preferentially in photoreceptors, but can also drive expression in other tissues to some extent, such as one that enhances expression at least two-fold, at least five-fold, at least ten-fold, at least twenty-fold or at least fifty-fold in photoreceptors compared to any other cell type. By rod-preferred promoter is meant a promoter that drives expression preferentially in photoreceptors, but can also drive expression in other tissues to some extent, such as one that enhances expression at least two-fold, at least five-fold, at least ten-fold, at least twenty-fold or at least fifty-fold in photoreceptors compared to any other cell type, including cones.

One or more other regulatory elements, such as enhancers, as well as a promoter, may also be present.

Vectors.

The vector of the invention may be of any type, for example it may be a plasmid vector or minicircular DNA.

Typically, the vectors of the invention, however, are viral vectors. The viral vector may, for example, be based on the herpes simplex virus, adenovirus or lentivirus. The viral vector may be an adeno-associated virus (AAV) vector or a derivative thereof. The viral vector derivative may be a chimeric, shuffled or capsid modified derivative.

The viral vector may contain an AAV genome from a naturally derived AAV serotype, isolate, or clade. The serotype may be, for example, AAV2, AAV5 or AAV8.

The effectiveness of gene therapy in general, as a rule, depends on the adequate and efficient delivery of nested DNA. This process is usually mediated by viral vectors. In gene therapy, usually

- 6 042590 adeno-associated viruses (AAV), which are members of the parvovirus family, are used. The wild-type AAV, containing viral genes, inserts its genomic material into chromosome 19 of the host cell. The single-stranded AAV DNA genome contains two inverted terminal repeats (ITRs) and two open reading frames containing structural (cap) and packaging (rep) genes.

For therapeutic purposes, the only sequences required in cis other than the therapeutic gene are the ITRs. Therefore, the AAV virus is modified: viral genes are removed from the genome, producing recombinant AAV (rAAV). This one contains only the therapeutic gene, two ITRs. Deletion of viral genes renders rAAV unable to actively insert its genome into the DNA of the host cell. Instead, rAAV genomes fuse through the ITR to form circular, episomal structures, or insert into pre-existing chromosomal breaks. For virus production, the structural and packaging genes, now removed from rAAV, are delivered in trans, in the form of a helper plasmid. AAV is a particularly attractive vector because it is generally non-pathogenic in general; most people have been infected with this virus during their lifetime without any side effects. Immune immunity of ocular tissue, a result of anatomical barriers and immunomodulatory factors, renders the eye largely immune to adverse immunological responses that AAVs can elicit in other tissues (Taylor 2009).

AAV vectors are limited to a relatively small packaging capacity of approximately 4.8 kbp. and slow onset of expression after transduction. Despite these minor shortcomings, AAV has become the most commonly used viral vector for retinal gene therapy.

Most vector constructs are based on AAV serotype 2 (AAV2). AAV2 binds to target cells via the proteoglycan heparin sulfate receptor. The AAV2 genome, like all AAV serotypes, can be enclosed in a number of different capsid proteins. AAV2 may be packaged in its natural AAV2 capsid (AAV2/2) or it may be pseudotyped with other capsids (e.g., the AAV2 gene in the AAV1 capsid, AAV2/1, the AAV2 gene in the AAV5 capsid, the AAV2/5 and AAV2 gene in the AAV8 capsid; AAV2/8).

rAAV transduces cells by serotype-specific receptor-mediated endocytosis. The main factor affecting the kinetics of rAAV transgene expression is the shedding rate of the capsid of the viral particle in the endosome. This, in turn, depends on the type of capsid enclosing the genetic material (ibid.). After capsid shedding, the linear single-stranded rAAV genome is stabilized by forming a double-stranded molecule through de novo synthesis of the complementary strand. The use of self-complementary DNA can bypass this step by producing double-stranded transgenic DNA. Natkunarajah et al. (2008) found that self-complementary AAV2/8 gene expression has a faster onset and higher amplitude compared to single-stranded AAV2/8. Thus, by bypassing the time delay associated with second strand synthesis, gene expression levels are increased compared to transgene expression from standard single strand constructs. Subsequent studies investigating the effect of self-complementary DNA in other AAV pseudotypes (eg AAV2/5) produced similar results. One caveat to this method is that since AAV has a packaging capacity of about 4.8 kb, the self-complementary recombinant genome must be suitably sized (i.e. 2.3 kb or less).

In addition to changing the packaging capacity, pseudotyping of the AAV2 genome with other AAV capsids can change the cell specificity and kinetics of transgene expression. AAV2/8 is reported to transduce photoreceptors more efficiently than either AAV2/2 or AAV2/5 (Natkunarajah et al., 2008).

Therefore, the vector of the invention may contain an adeno-associated virus (AAV) genome or a derivative thereof.

The AAV genome is a polynucleotide sequence that encodes the functional elements necessary for the production of an AAV viral particle. These functional elements include those that operate in the AAV replication and packaging cycle in the host cell, including the encapsulation of the AAV genome into an AAV viral particle. Naturally occurring AAV viruses are replication-deficient and rely on the provision of in trans helpers to complete the replication and packaging cycle. Accordingly, and with the additional deletion of the AAV rep and cap genes, the AAV vector genome of the invention is replication-deficient.

The AAV genome can be in single stranded form, either sense or antisense, or alternatively double stranded. The use of the double-stranded form makes it possible to bypass the DNA replication step in the target cell and thus accelerate the expression of the transgene. The AAV genome may be from any naturally occurring AAV serotype or isolate or clade. As the person skilled in the art knows, naturally occurring AAV viruses can be classified according to different biological systems.

Typically, AAV viruses are referred to in terms of their serotype. The serotype corresponds to a variant of the AAV subspecies, which, due to its expression profile of capsid surface antigens, has a characteristic

- 7 042590 ractory reactivity, which can be used to distinguish it from other subspecies. In general, a virus having a specific AAV serotype does not effectively cross-react with neutralizing antibodies specific for any other AAV serotype. AAV serotypes include AAV1, AAV2, AAV3, AAV4, AAV5, AAV6, AAV7, AAV8, AAV9, AAV10 and AAV11 as well as recombinant serotypes such as Rec2 and Rec3 recently identified in primate brains. In the vectors of the invention, the genome can be derived from any AAV serotype. The capsid can also be derived from any AAV serotype. The genome and capsid may be derived from the same serotype, or from different serotypes.

Preferably, the genome in the vectors of the invention is derived from AAV serotype 2 (AAV2), AAV serotype 4 (AAV4), AAV serotype 5 (AAV5), or AAV serotype 8 (AAV8). Most preferably, the genome is derived from AAV2, however, other serotypes of particular interest for use in the invention include AAV4, AAV5 and AAV8, which efficiently transduce ocular tissue such as the retinal pigment epithelium. Preferably, the capsid is derived from AAV5 or AAV8, in particular from AAV8.

Reviews of AAV serotypes can be found in Choi et al. (Curr Gene Ther. 2005; 5(3); 299-310) and Wu et al. (Molecular Therapy. 2006; 14(3), 316-327). Sequences of AAV genomes or AAV genome elements, including ITR, rep or cap gene sequences for use in the invention, can be obtained from the following sequence identification numbers for complete AAV genomes: adeno-associated virus 1 NC_002077, AF063497; adeno-associated virus 2 NC_001401; adeno-associated virus 3 NC_001729; adeno-associated virus 3B NC_001863; adeno-associated virus 4 NC_001829; adeno-associated virus 5 Y18065, AF085716; adeno-associated virus 6 NC_001862; avian AAV ATCC VR-865 AY186198, AY629583, NC_004828; avian strain AAV DA-1 NC_006263, AY629583; bullish AAV NC_005889, AY388617.

AAV viruses may also be referred to in terms of clades or clones. It refers to the phylogenetic relationship of naturally derived AAV viruses, and generally to the phylogenetic group of AAV viruses that can be traced back to a common ancestor and that includes all of its descendants. In addition, AAV viruses may be referred to in terms of a specific isolate, ie a genetic isolate of a particular AAV virus found in nature. The term genetic isolate describes a population of AAV viruses that has undergone limited genetic mixing with other naturally occurring AAV viruses, thereby defining a recognizably distinct population at the genetic level.

Examples of AAV clades and isolates that may be used in the invention include:

clade A: AAV1 NC_002077, AF063497, AAV6 NC_001862, Hu. 48 AY530611, Hu 43 AY530606, Hu 44 AY530607, Hu 46 AY530609, clade B: Hu. 19 AY530584, Hu. 20 AY530586, Hu 23 AY530589, Hu22 AY530588, Hu24 AY530590, Hu21 AY530587, Hu27 AY530592, Hu28 AY530593, Hu 29 AY530594, Hu63 AY530624, Hu64 AY530625, Hu13 AY530578, Hu56 AY530618, Hu57 AY530619, Hu49 AY530612, Hu58 AY530620, Hu34 AY530598, Hu35 AY530599, AAV2 NC_001401, Hu45 AY530608, Hu47 AY530610, Hu51 AY530613, Hu52 AY530614, Hu Т41 AY695378, Hu S17 AY695376, Hu Т88 AY695375, Hu Т71 AY695374, Hu Т70 AY695373, Hu Т40 AY695372, Hu Т32 AY695371, Hu Т17 AY695370, Hu LG15 AY695377, клада В: Hu9 AY530629, Hu10 AY530576, Hu11 AY530577, Hu53 AY530615, Hu55 AY530617, Hu54 AY530616, Hu7 AY530628, Hu18 AY530583, Hu15 AY530580, Hu16 AY530581, Hu25 AY530591, Hu60 AY530622, Ch5 AY243021, Hu3 AY530595, Hu1 AY530575, Hu4 AY530602 Hu2, AY530585, Hu61 AY530623, клада Г: Rh62 AY530573, Rh48 AY530561, Rh54 AY530567, Rh55 AY530568, Cy2 AY243020, AAV7 AF513851, Rh35 AY243000, Rh37 AY242998, Rh36 AY242999, Cy6 AY243016, Cy4 AY243018, Cy3 AY243019, Cy5 AY243017, Rh13 AY243013, клада Д: Rh38 AY530558, Hu66 AY530626, Hu42 AY530605, Hu67 AY530627, Hu40 AY530603, Hu41 AY530604, Hu37 AY530600, Rh40 AY530559, Rh2 AY243007, Bb1 AY243023, Bb2 AY243022, Rh10 AY243015, Hu17 AY530582, Hu6 AY530621 , Rh25 AY530557, Pi2 AY530554, Pi1 AY530553, Pi3 AY530555, Rh57 AY530569, Rh50 AY530563, Rh49 AY530562, Hu39 AY530601, Rh58 AY530570, Rh61 AY530572, Rh52 AY530565, Rh53 AY530566, Rh51 AY530564, Rh64 AY530574, Rh43 AY530560, AAV8 AF513852, Rh8 AY242997, Rh1 AY530556, клада E: Hu14 (AAV9) AY530579, Hu31 AY530596, Hu32 AY530597, Clonal Isolate AAV5 Y18065, AF085716, AAV 3 NC_001729, AAV 3B NC_001863, AAV4 NC_001829, Rh34 AY243001, Rh33 AY243002, Rh32 AY243003.

The person skilled in the art can select the appropriate AAV serotype, clade, clone or isolate for use in the present invention based on the prior art. However, it should be understood that the invention also encompasses the use of the AAV genome of other serotypes that have not yet been identified or characterized. The AAV serotype determines the tissue specificity of the infection (or tropism) of the AAV virus. Accordingly, the preferred AAV serotypes for use as AAV viruses administered to patients in accordance with the invention are those

- 8 042590 which have a natural tropism for rod photoreceptors or a high infection efficiency of rod photoreceptors.

Typically, the AAV genome of a naturally derived AAV serotype or isolate or clade contains at least one inverted terminal repeat (ITR) sequence. The vectors of the invention typically contain two ITRs, preferably one at each end of the genome. The ITR sequence acts in cis to provide a functional origin of replication and allows integration and excision of the vector from the cell's genome. Preferred ITR sequences are AAV2 sequences and variants thereof. The AAV genome typically contains packaging genes, such as the rep and/or cap genes, which encode the packaging functional elements for the AAV virus particle. The rep gene encodes one or more proteins: Rep78, Rep68, Rep52 and Rep40, or variants thereof. The cap gene encodes one or more capsid proteins, such as: VP1, VP2 and VP3, or variants thereof. These proteins form the capsid of the AAV viral particle. Variants of the capsid are discussed below.

Preferably the AAV genome will be derivatized for administration to patients. Such derivatization is standard in the art, and the present invention encompasses the use of any known derivative of the AAV genome, and derivatives that can be obtained using methods known in the art. Derivatization of the AAV genome and AAV capsid is discussed, for example, in Choi et al. and Wu et al., mentioned above.

Derivatives of the AAV genome include any truncated or modified form of the AAV genome that allows in vivo expression of the Rep-1 transgene from the vector of the invention. As a rule, it is possible to significantly truncate the AAV genome in order to create a minimal viral sequence and still retain the above function. For safety reasons, this is preferred in order to reduce the risk of recombination of the vector with wild-type virus, as well as to prevent the induction of a cellular immune response, which can result from the presence of viral gene proteins in the target cell.

Typically, the derivative will contain at least one inverted terminal repeat (ITR) sequence, preferably more than one ITR, such as two or more ITRs. One or more ITRs may be derived from AAV genomes having different serotypes, or may be a chimeric or mutated ITR. A preferred ITR mutant is one that has a trs (treminal resolution site) deletion. This deletion allows the genome to continue to replicate to form a single-stranded genome that contains both the coding and complementary sequence, i.e. self-complementary AAV genome. This allows the DNA replication step to be skipped in the target cell and thus allows accelerated expression of the transgene.

One or more ITRs will preferably flank an expression cassette containing a promoter and a transgene of the invention. The insertion of one or more ITRs is preferred to facilitate packaging of the vector of the invention into viral particles. In preferred embodiments, the ITR elements will be the only sequences conserved in the derivative from the native AAV genome. Thus, the derivative preferably does not contain the rep and/or cap genes of the native genome, and any other sequences of the native genome. This is preferred for the reasons described above, as well as to reduce the possibility of integration of the vector into the genome of the host cell. In addition, reducing the size of the AAV genome allows for more flexibility in inserting other sequence elements (such as regulatory elements) into the vector in addition to the transgene. Referring to the AAV2 genome, the following parts can be deleted in a derivative of the invention: one inverted terminal repeat (ITR) sequence, replication (rep) and capsid (cap) genes. However, in some embodiments of the invention, including in vitro embodiments of the invention, the derivatives may additionally contain one or more rep and/or cap genes, or other viral sequences of the AAV genome.

The derivative may be a chimeric, shuffled, or capsid-modified derivative of one or more naturally occurring AAV viruses. The invention encompasses the placement of capsid protein sequences from different AAV serotypes, clades, clones or isolates within the same vector. The invention encompasses the packaging of the genome of one serotype into a capsid of another serotype, i.e. pseudotyping.

Chimeric, shuffled, or capsid-modified derivatives are generally chosen to provide one or more desired functionality to the viral vector. Thus, these derivatives may exhibit increased gene delivery efficiency, reduced immunogenicity (humoral or cellular), altered tropism range, and/or improved targeting to a specific cell type compared to an AAV viral vector containing a native AAV genome, such as the AAV2 genome. Increased efficiency of gene delivery can be achieved by improved binding to a receptor or co-receptor on the cell surface, improved entry into the cell, improved movement within the cell and entry into the nucleus, improved shedding of the capsid of the viral particle, and improved single-stranded genome conversion.

- 9 042590 into double stranded form. Increased efficacy may also refer to an altered range of tropism or targeting a specific cell population so that the vector dose is not diluted by administration to tissues where it is not required.

Chimeric capsid proteins include those produced by recombination between two or more capsid-coding sequences of naturally occurring AAV serotypes. This can be done, for example, using a genetic marker rescue method in which non-infectious capsid sequences of one serotype are co-transfected with capsid sequences of another serotype, and directional selection is used to select capsid sequences having the desired properties. The capsid sequences of different serotypes can be altered by homologous recombination within the cell to produce new chimeric capsid proteins. Chimeric capsid proteins also include those created by constructing sequences of capsid proteins to transfer specific domains, surface loops, or single amino acid residues of a capsid protein between two or more capsid proteins, for example, between two or more capsid proteins of different serotypes.

Shuffled or chimeric capsid proteins can also be generated by DNA shuffling, or by error-producing PCR. AAV capsid fusion genes can be created by randomly fragmenting sequences of related AAV genes, such as those encoding capsid proteins, from many different serotypes, and then assembling the fragments in a self-priming polymerase reaction, which can also lead to crossover in areas of sequence homology. A library of hybrid AAV genes created in this way, i.e. by shuffling the capsid genes of several serotypes, can be screened to identify viral clones having the desired functionality. Similarly, error-producing PCR can be used to randomly mutate the AAV capsid genes to create a heterogeneous library of variants that can then be selected for a desired property.

Capsid gene sequences can also be genetically modified to introduce specific deletions, substitutions, or insertions relative to the native wild type sequence. In particular, capsid genes can be modified by inserting an unrelated protein or peptide sequence into the open reading frame of the capsid-coding sequence or at the N- and/or C-terminus of the capsid-coding sequence.

The unrelated protein or peptide may advantageously be one that acts as a ligand for a particular cell type, thereby providing improved binding to the target cell, or improved specificity in targeting the vector to a particular cell population.

The unrelated protein may also be one that assists in the purification of the viral particle as part of the manufacturing process, i.e. epitope or affinity tag. The insertion site, as a rule, is chosen so as not to interfere with other functions of the viral particle, for example, the function of penetration into cells, the function of moving viral particles in the cell. One skilled in the art can determine suitable sites for insertion based on knowledge of the prior art. Specific sites are described in Choi et al., mentioned above.

The invention further includes providing the AAV genome sequences in a different order and configuration than the native AAV genome. The invention also encompasses the replacement of one or more AAV sequences or genes with sequences from another virus or chimeric genes consisting of sequences from more than one virus. Such chimeric genes may consist of sequences from two or more related viral proteins from different viral species.

The vector of the invention takes the form of a viral vector containing promoters and expression constructs of the invention.

The invention also provides an AAV virus particle containing the vector of the invention. The AAV particles of the invention comprise transcapsidated forms wherein the AAV genome or derivative having the ITR of one serotype is packaged into a capsid of another serotype. The AAV particles of the invention also include mosaic forms, wherein a mixture of unmodified capsid proteins from two or more different serotypes constitutes the viral envelope. The AAV particle also includes chemically modified forms bearing ligands adsorbed on the surface of the capsid. For example, such ligands may include antibodies to target a particular receptor on the cell surface.

The invention further provides a host cell containing the AAV vector or virus particle of the invention.

The vectors of the invention can be prepared by standard methods known in the art for providing vectors for gene therapy. Thus, conventional and publicly available methods of transfection, packaging and purification can be used to prepare a suitable vector form.

As discussed above, the vector of the invention may contain the entire genome of a naturally occurring AAV virus in addition to the promoter of the invention, or a variant thereof. However, usually

- 10 042590 uses a derivatized gene, for example a derivative that has at least one inverted terminal repeat (ITR) sequence, but which may lack any AAV genes such as rep or cap.

In such embodiments, in order to allow the assembly of the derivatized genome into an AAV viral particle, additional genetic constructs will be provided to provide AAV and/or helper virus functions in a host cell in combination with the derivatized genome. These additional constructs typically contain genes encoding structural AAV capsid proteins, ie. cap, VP1, VP2, VP3, and genes encoding other functional elements required for the AAV life cycle, such as rep. The choice of structural capsid proteins provided in the supplementary construct will be determined by the serotype of the packaged viral vector.

A particularly preferred packaged viral vector for use in the invention comprises a derivatized AAV2 gene in combination with AAV5 or AAV8 capsid proteins. As mentioned above, AAV viruses are incapable of replication and, therefore, the accessory functions of the virus, preferably the accessory functions of the adenovirus, are usually also provided in one or more additional constructs in order to allow AAV replication.

All of the above additional constructs may be provided as plasmids or other episomal elements in the host cell, or alternatively, one or more constructs may be integrated into the host cell's genome.

Pharmaceutical compositions, dosages and treatments.

The vector of the invention may be formulated into pharmaceutical compositions. In addition to the vector, these compositions may contain a pharmaceutically acceptable excipient, carrier, buffer, stabilizer, or other substances well known to those skilled in the art. Such substances must be non-toxic and must not interfere with the effectiveness of the active ingredient. The exact nature of the carrier or other substance can be determined by the skilled person according to the route of administration, ie herein: direct retinal, subretinal or intravitreal injection.

The pharmaceutical composition is usually provided in liquid form. Liquid pharmaceutical compositions generally include: a liquid carrier such as water, petroleum, animal or vegetable oils, mineral oil or synthetic oil. May be included: physiological saline, magnesium chloride, dextrose or other saccharide solution, or glycols such as ethylene glycol, propylene glycol or polyethylene glycol. In some cases, a surfactant such as 0.001% pluronic acid (PF68) may be used.

For injection at the site of injury, the active ingredient will be presented in the form of an aqueous solution that is free of pyrogens and has the appropriate pH, isotonicity and stability. Those skilled in the art are well able to prepare suitable solutions using, for example, isotonic vehicles such as sodium chloride injection, Ringer's solution for injection, Ringer's lactose solution for injection, Hartmann's solution. If necessary, preservatives, stabilizers, buffers, antioxidants and/or other additives may be included.

For sustained release, the vector can be incorporated into a pharmaceutical composition that is formulated for slow release, such as microcapsules formed from biocompatible polymers or liposomal carrier systems, according to methods known in the art. The vectors and/or pharmaceutical compositions of the invention may be packaged in a kit.

In general, direct retinal, subretinal or intravitreal delivery of the vectors of the invention is generally preferred, usually by injection. Thus, delivery to the retina, subretinular space or intravitreal space is preferred. Vectors can also be introduced into rod photoreceptors in vitro followed by cell transplantation into the retina.

The vectors and/or pharmaceutical compositions of the invention may also be used in combination with any other therapy for the treatment or prevention of visual impairment. For example, they can be used in combination with known therapies that use VEGF antagonists, such as anti-VEGF antibodies such as bevacizumab or ranibizumab, or soluble receptor antagonists such as aflibicept, to treat ADHD or other ocular conditions such as described in this document.

Doses and dosage regimens can be determined within the ordinary skill of the physician responsible for administering the composition. The dose of the vector according to the invention can be determined according to various parameters, especially according to information about the age, weight and condition of the patient to be treated; routes of administration; and the required drug regimen. Again, the physician will be able to determine the appropriate route of administration and dosage for any particular patient.

A typical single dose is 10 ¹⁰ to 10 ¹² genomic particles, depending on the amount

- 11 042590 retinal tissue that requires transduction. A genomic particle is defined herein as an AAV capsid that contains a single-stranded DNA molecule that can be quantified using a sequence-specific method (such as real-time PCR). This dose may be given as a single dose, but may be duplicated for the other eye, or for cases in which the vector may not reach the correct area of the retina for any reason (eg, due to a surgical complication). Treatment is preferably a single long term treatment for each eye, but repeated injections may be considered, eg in subsequent years and/or with different AAV serotypes.

Treatment.

The vectors of the invention can be used to treat any ocular pathology in which dysfunction, degeneration, or absence of cones occurs, but at least some healthy rods remain. Cone function may be completely or partially lost, for example, at least 10%, at least 25%, at least 50%, at least 75%, at least 80%, at least 90%, or more. Healthy rods are rods that are able to function fully or partially, for example, at least 10%, at least 25%, at least 50%, at least 75%, or at least 90% from the normal function of the rods in terms of light perception at scotopic levels.

Pathologies that can be treated using the vectors of the invention thus include macular degeneration, achromatopsia and Leber's amaurosis. Macular degeneration can be age-related macular degeneration (AMD), such as wet or neovascular MPD, or geographic atrophy, hereditary macular degeneration, or hereditary cone degeneration. In some embodiments, the invention will result in the creation of a pseudo-fovea, a small area of cone-like rods that improves vision in conditions where foveal cones have been lost or are dysfunctional.

In general, as a rule, patients who will be treated with the vectors of the invention will be human patients. They can be male or female and of any age.

The following examples illustrate the invention.

Examples

Example 1 - methods for ArchT experiments.

Animals.

Wild type mice (C57BL/6J) were purchased from Harlan Laboratories (Blackthorn, UK). Mice CNGA3-/- and PDE6C-/- were bred in the laboratory. All mice were maintained under intermittent lighting conditions (12 h light-dark); the illumination in the cage was 7 fc during the light cycle. All experiments were approved by local committees for the care and use of animals (UKL, London, UK), and were in accordance with the guidelines for the care and use of animals adopted by the Society of Neurology and the Association for the Study of Vision and Ophthalmology (Rockville, Maryland).

Plasmid constructs, virus production and injection procedures.

The transgenic construct (ArchT-EGFP) was kindly provided by Prof. Ed Boyden (MIT, USA) and contains the cDNA sequence of the ArchT gene fused to the EGFP fluorescent protein. Plasmids were packaged in AAV8 to obtain recombinant viral vectors AAVAAV8.hRho.ArchT-EGFP. The recombinant AAV8 vector was produced using the triple transient transfection method as previously described. The plasmid construct, AAV serotype specific packaging plasmid and helper plasmid were mixed with polyethyleneimine (Polysciences Inc.) to form transfection complexes, which were then added to 293T cells and left for 72 hours. Cells were harvested, concentrated and lysed to release the vector. AAV8 was purified using Sepharose AVB columns (GE Healthcare). Both were washed in 1x PBS (phosphate buffered saline) and concentrated to a volume of 100-150 µl. Viral particle titers were determined by comparative dot-blot hybridization of DNA prepared from purified viral stocks and established plasmid controls. Purified vector concentrates that were used in all experiments had 5x10 ¹² viral particles/ml. Subretinal injections were performed as previously described by our group and consisted of double injections of 2 μl each.

Immunohistochemistry.

Animals were euthanized, eyeballs removed, and cornea, lens, and iris removed. To obtain sections of the retina, the eye cups were fixed in 4% paraformaldehyde (PPD) for 1 h at room temperature, before introducing the optimal cutting temperature (OST) into the medium. 30 μm cryosections were cut in sagittal orientation, washed with PBS and blocked in 10% normal goat serum (NSC), 3% bovine serum albumin (BSA) and 0.1% Triton-X100. Appropriate samples were incubated with primary antibodies in blocking solution at 4°C overnight using rabbit anti-cone arrestin (diluted 1:500). After washing with PBS, the appropriate combination of secondary antibodies (all diluted 1:500, life technologies) was applied, including goat-anti-rabbit Alexa Fluor 546 (catalog

- 12 042590 A11035), goat-anti-mouse Alexa Fluor 633 (Cat. No. A21052) and Streptavidin, Alexa Fluor 633 conjugate (Cat. No. S21375) to label samples before they were restained (after decolorized) with DAPI and fixed using DAKO fluorescence fixing medium (DAKO, S3023, Denmark). Images were obtained using confocal microscopy (Leica DM5500Q).

Registration of the signal of a single photoreceptor using a suction pipette electrode.

The animals were dark adapted for 12 h prior to the start of the experiments. Mice were overdosed with the ketamine-dormitor anesthetic mixture via the intraperitoneal cavity to induce terminal surgical planar anesthesia. The mice were then sacrificed by dislocation of the cervical vertebrae and their eyeballs were removed. His eyes were cut open in the dim distant red light. The isolated retinas were placed in a 1% low melting point agarose solution and then cut using a vibrotome (leica) into 230 µm thick facial sections. Sections were mounted in a recording chamber and perfused with Ames medium saturated with carboxylic acid (95% O ₂ 5% CO2) containing 100 μM 9-cis retinal (Sigma) and 0.2% BSA (Sigma). The temperature of the perfusion solution was maintained at 37°C using a built-in heating element under feedback control (Scientifica). Low resistance patch pipettes (1-2 mΩ) were made from filamentous borosilicate glass capillaries (Harvard Apparatus Ltd) using a Narishige PC-10 vertical puller. The pipettes were filled with external solution, placed on the headstage, and slight pressure was applied to the tip (~30 mbar). Using infrared light and an imaging microscope, a pipette was placed on the surface of the retina and then plunged about 50 µm into the section until the photoreceptor segments were intact and neatly positioned. A slight negative pressure was applied to the pipette tip as it moved slowly through the retinal tissue, using a 100 ms, 10 mV pulse to control the resistance across the entire pipette tip. When the resistance increased to ~20-30 mΩ, photosensitive responses were elicited. Light stimuli from an LED light source (peak wavelength 530 nm) connected to a liquid light guide were applied through a microscope objective (Olympus). Neutral density filters were used to precisely control the intensity of the light stimulus. Light stimuli consisted of square wave pulses programmed using P-Clamp software (Molecular Devices) and delivered via a DAC board (Axon Instruments) in connection with an LED driver (Thorlabs).

Electrophysiological data recording was performed using a Multiclamp 700B amplifier (Molecular Devices). The data were digitized at a frequency of 20 kHz.

CHME registration of signals.

The animals were dark adapted for 12 h prior to the start of the experiments. Mice were overdosed with the ketamine-dormitor anesthetic mixture via the intraperitoneal cavity to induce terminal surgical planar anesthesia. The mice were then sacrificed by dislocation of the cervical vertebrae and their eyeballs were removed. Eyes were opened in carbogen-saturated (95% oxygen, 5% carbon dioxide) Ames' medium (Sigma), under dim red light. The cornea and lens were removed, and efforts were made to remove as much of the vitreous as possible from the surface of the retina. RPE was separated from the retina, and a flat segment (petal) 1-3 mm in size was cut off from the cup of the retina. This retinal segment was placed with the ganglion cell side on the surface of the multielectrode chip, and a circle with a handle made of inert platinum wire (Sigma) and nylon was used to hold the segment in position. In all signal recordings, tissue was perfused with carbogen-saturated Ames' medium (Sigma) maintained at 36.5°C. For recordings that included scotopic or mesopic conditions, the perfusion medium was prepared to contain 100 μM 9-cis-retinal (Sigma) in 0.2% BSA (Sigma). To record the extracellular potentials of ganglion cells, a perforated recording chip with 60 electrodes was used, consisting of tungsten electrodes located at a distance of 100 μm from each other (Multi Channel Systems). The voltage changes were amplified and digitized at 50 kHz by the MC Card system using the MC Rack software (Multichannel Systems).

Analysis of electrophysiological data.

Electrophysiological data were analyzed using custom macros in IgorPro 6. Synaptic currents and potentials were detected using an amplitude threshold algorithm, with the threshold for event detection set to twice the standard deviation of the baseline noise (typically around 10 pA). The detected currents and potentials were manually confirmed by carefully checking all electrophysiological data.

Development of conditioned reflexes by fear (Fear conditioning).

The mice were trained and tested using a commercially available fear conditioning system (Med Associates). To ensure the conditions of the blind experiment, the experimenter performing the preparation and testing was always ignorant of the mouse strain and

- 13 042590 for treatment. Briefly, the setup consisted of a conditioning chamber (20 x 30 cm) with a stainless steel grating floor placed inside a small soundproof room. The behavior of the mouse was constantly monitored during training and testing using a built-in infrared digital video camera (reception rate of 30 fps) and infrared illumination. Video Freeze software (Med Associates) was used to control the delivery of light stimuli and electrical shocks. The light stimulus consisted of one LED (530 nm, Thorlabs) 5 Hz 50 ms flash at full brightness generated via the Arduino interface (Arduino software); the lamp was located on the side panel of the chamber for the development of conditioned reflexes. To ensure that the conditions under which training and testing took place differed, a floor panel and curved wall panels were inserted into the chamber for the testing session. Background white light was used to reduce the likelihood of rod activation, and pupils were dilated with tropicamide drops to increase the amount of light reaching the mouse's retina.

Mice were placed inside the chamber and a single conditioning session was performed, consisting of 6 paired 5-s light stimuli combined with a 2-s shock of 0.65 mA. The interval between stimuli was pseudo-randomized (mean interval 90 s). After training, the mice were returned to their home cage. 24 h after training, the mice were tested for visual cues recall. The mice were placed in the test chamber and observed for 360 seconds. A light stimulus for the development of a conditioned reflex was applied continuously during the last 120 seconds of the test session. All data were automatically acquired and evaluated using VideoFreeze software (Med Associates). Briefly, the software is calibrated before placing the animal in the chamber. The software then measures the pixel changes that occur between each video frame. The motion threshold was set as low as possible (20 motion index units) and the continuous fade count was set equal to the frame rate to provide the most sensitive motion reading. To assess recall using light cues, the percentage fading time was averaged over 2 min immediately before and after the start of the light stimulus.

Statistical significance was assessed using one-way analysis of variance. Results are presented as mean ± standard error of the mean (SEM).

Optomotor research.

Visual acuity was measured by observing the optomotor responses of mice to rotating sinusoidal gratings (OptoMotry, Cerebral Mechanics). The protocol used results in independent measurements of the acuity of the right and left eyes, based on the unequal sensitivity of the two eyes to the rotation of the pattern, since only movement in the direction from the temple to the nose causes a tracking response. As a result, the right and left eyes are most sensitive to counterclockwise (CW) and clockwise (CW) rotation, respectively. Stimuli of different temporal frequencies were used to determine the threshold at which a response was present. A double-blind, two-way forced selection procedure was applied in which the observer was blind in the direction of rotation of the pattern, whether it was a treated or untreated CNGA3-/- or PDE6C-/- ArchT mouse, or an age-matched wild-type control animal (C57BL6). Visual acuity was measured for both eyes of the test animal and averaged, or analyzed separately for each eye after 4 trials were conducted over 4 separate days. The measurement was carried out on injected mice 3-10 weeks after treatment of these mice, together with control isogenic age-matched mice.

Example 2 Expression of ArchT in rod photoreceptors allows response with rapid responses without decolorization.

The vision provided by the rods is optimized for low light levels, including single photon detection. However, rods cannot match cones in terms of the speed of starting and recovering responses to light (Fu et al., 2007, Pugh et al., 1999). This functional difference, useful for providing reliable vision in different environments, becomes debilitating when cone-mediated vision is lost in pathologies such as age-related macular degeneration, where densely packed cones in the fovea undergo degradation (de Jong 2006). It has been investigated whether rods can respond and recover more quickly in response to stimuli; this may help alleviate functional impairment caused by cone loss.

The fast light-driven proton pump (ArchT) (Han et al., 2011) was expressed in rod photoreceptors. AAV8 particles carrying ArchT-EGFP under the control of the rhodopsin (Rho) promoter were administered subretinally to adult mice. Rho-ArchT-EGFP expression was membrane-limited to rod photoreceptors (Fig. 1A-B). The synaptic ends of rods expressing Rho-ArchTEGFP and cones can be easily distinguished after immunohistochemistry (Fig. 1B). Quantitative PCR on the sorted cone population confirmed that Rho-ArchT-EGFP expression was specific to the rod population and no clear signs of toxicity were observed by 6 months after AAV8 injection. The expression of ArchT allowed for extremely rapid responses to

- 14 042590 light, while the intrinsic response of the rods was maintained and was comparable to that observed in non-transduced rod photoreceptors (Fig. 1D). Light-induced currents recorded from rods expressing ArchT exhibited significantly faster kinetics than rod intrinsic currents in all mouse models tested (Fig. 2A-B). This kinetics allowed the light-induced currents to be amplified up to 80 Hz, well beyond the limits of both rods, which die at -20 Hz (FIGS. 2A-B), and cones (Fu et al., 2007).

Surprisingly, ArchT expression did not alter the characteristics of rod photoreceptors, but in addition it conferred additional ability to respond with rapid responses without decolorization (FIG. 2E).

Example 3 ArchT-expressing rods triggered sustained GCS pulses at high light intensities and at frequencies approaching those of cone photoreceptors.

It was then investigated whether the chopstick-driven neural pathway could keep up with the visual perception provided by the faster-than-normal chopsticks. The rod and cone pathways share some similarities and some significant differences, and it is therefore not clear whether the rod neural pathways can reliably sustain fast cone-like transmission (Wassle et al., 2004). Rods have been shown to make direct contact with OFF-bipolar cone cells (depolarization in these cells is mediated by outgoing cationic current) (Soucy et al., 1998 and Hack et al., 1999) and paired flash stimulation suggests that this alternative pathway may be as fast as the cone-OFF-bipolar cell pathway (Li et al., 2010). However, it is not clear how robust this response might be, and whether ON-bipolar rod cells can also support rapid transmission. Furthermore, the synaptic ends of the rods have a different size and ultrastructural organization compared to the cones, and the bipolar rod cells do not directly contact the retinal ganglion cells (RGCs), but only via a pathway involving AII amacrine cells (Wassle et al., 2004). To investigate the maximum rate that the rod signaling pathway can achieve, multi-electrode recording of the signals received from glucocorticosteroids in mouse models with non-functioning cones was performed to determine the outgoing glucocorticosteroid signal mediated by rods. Rod-induced responses in RGCs in non-transduced retinas faded at high light levels and could not follow stimulation frequencies higher than -20 Hz (Fig. 2E).

In contrast, ArchT-expressing rods elicited GCS pulses at high light intensities and at frequencies approaching those of cone photoreceptors (Fig. 2E).

Example 4 - Expression of Rho-ArchT-EGFP enhanced the sensitivity of mice that had lost cone-mediated vision to bright light stimuli.

For this faster-than-normal visual perception provided by rods to be useful, it has been observed that mice must be able to use ArchT-mediated currents to respond reliably to bright and fast stimuli. CNGA3-/- and PDE6C-/- mice lacking cone-mediated vision (Biel et al., 1999 and Change et al., 2009) failed to learn through fear conditioning when bright light stimuli were paired with and stopped along with a light shock to the legs (Fig. 3A). However, the expression of Rho-ArchT-EGFP expanded the sensitivity of these mice to bright light stimuli, allowing learning based on the association between visual stimulus and electric shock (Fig. 3A). Finally, the possibility was tested whether ArchT expression could confer rapid visual perception in CNGA3-/- and PDE6C-/- mice. Evaluation of the speed of visual perception using optomotor testing (Umino et al., 2008) showed that mice expressing ArchT were able to track stimuli faster than mice that did not receive subretinal viral injections and than mice that received a vector containing only GFP (Fig. 3B). The maximum stimulus frequency that ArchT expressing mice could track was similar to that of cone photoreceptors (Fig. 3B).

Together, these results suggest that rods can fire faster than their own phototransduction cascade and that the neural pathways triggered by rods can support faster signaling. Importantly, synaptic release from rods does not require large voltage fluctuations, but small currents may instead cause sufficient voltage fluctuations to significantly alter their synaptic transmission (Cangiano et al., 2012). This extends the applicability of the invention to light levels that are several times lower than the average light levels required for optogenetic control of activity in most other neurons (Han et al., 2011).

Example 5 - Methods for RPAP Experiments.

Animals.

C57BL6 mice (Harlan, UK), Cnga3-/- (J.R. Heckenlively, University of Michigan), Pde6c-/- (J.R. Heckenlively, University of Michigan, Michigan) (Chang et al., 2009) and Gnat1-/- (J. Lem , Tufts University School of Medicine, Massachusetts) (Calvert et al., 2000) were kept in a vivarium in

- 15 042590

University College London. Adult male and female animals were 6-12 weeks old at the time of virus injection and were used for experiments at least 2 weeks after injection to ensure sufficient expression of R9AP. All mice used were between 2 and 6 months of age and were similar in age between groups in this experiment. All experiments were performed in accordance with the policy on the use of animals and humans in neurobiological research and in accordance with the AISO (Association for the Study of Vision and Ophthalmology) specification on the use of animals in ophthalmic and vision research. Animals were maintained on a standard 12/12 hour day-night cycle.

Plasmid constructs and production of recombinant AAV8.

The mouse R9ap cDNA was PNR amplified from mouse retinal cDNA using primers that were designed to cover the entire coding region. The R9ap cDNA was cloned into a vector between the promoter (CMV promoter (cytomygalovirus) or bovine rhodopsin promoter) and the SV40 polyadenylation site. These plasmids were used to create two pseudotyped AAV2/8 viral vectors, rAAV2/8.CMV.mR9ap and rAAV2/8.Rho.mR9ap, as described below.

The recombinant AAV2/8 vector was produced by a triple transient transfection method as described previously (Gao et al., 2002). The plasmid construct, AAV serotype-specific packaging plasmid and helper plasmid were mixed with polyethyleneimine to form transfection complexes, which were then added to 293T cells and left for 72 hours. Cells were harvested, concentrated and lysed to isolate the vector. AAV2/8 was purified by affinity chromatography and concentrated using ultrafiltration columns (Sartorius Stedim Biotech, Göttingen, Germany), washed in PBS and concentrated to a volume of 100-150 μl. Titers of viral particles were determined using dot-blot hybridization or real-time PCR. The concentration of purified vectors used was 1-2x10 ¹² viral particles/ml.

Electroretinogram (ERG).

ERGs were recorded for both eyes after the mice were dark adapted overnight using a commercially available system (Espion E2, Diagnosys LLC, Lowell, MA). Before recording, animals were anesthetized with an intraperitoneal injection of 0.007 ml/g of a mixture of medetomidine hydrochloride (1 mg/ml), ketamine (100 mg/ml) and water in a ratio of 5:3:42. The pupils were fully dilated using 2.5% phenylephrine and 1.0% tropicamide. Midline subdermal ground electrodes and mouth reference electrodes were placed first, followed by silver plus electrodes that were allowed to lightly touch the center of the cornea under dim red illumination. A drop of Viscotears Liquid Gel 0.2% (Dr. Robert Winzer Pharma/OPD Laboratories, Watford, UK) was placed on the plus electrodes to keep the cornea hydrated during recording and the mice were allowed to undergo further dark adaptation for 5 min. . The band pass filter cutoff frequencies were 0.312 Hz and 1000 Hz. The photoresponse recovery rate was measured using the concept of a paired flash, in which pairs of flashes with the same saturation intensity (1.8 log cd s/m ² ) were separated by different intervals between stimuli (IMS; 0.5, 1, 2, 4, 8, 16, 32, 64 s). In this concept, the 1st flash completely suppresses the electrical responses of the rod mechanisms, which makes it possible to observe the rate of functional recovery of the rod function by applying the 2nd flash with a different IC. Sufficient time (150 s) was allowed between pairs of flashes to allow full recovery from the first flash. The observed a-wave amplitude recovery should then reflect the rate of rod deactivation in animals lacking cone function, since the flash should only decolorize a fraction (0.02%) of rhodopsin (Lyubarsky et al., 2004 and Weymouth, AE & Vingrys 2008). A series of 6 Hz flashes was performed with scotopic intensity, as previously reported, with several modifications (Seeliger et al., 2001). 17 flash intensity steps were used from -6 to 2 log cd s/m ² separated by 0.5 log units from each other. For each stage, after 10 seconds of adaptation, 600 ms sweeps were averaged 20 times using the same burst condition. A series of dark-adapted responses was also obtained using longer flashes of 20, 100 and 200 ms duration at 83.3 cd/m ² . Standard single-flash scotopic recordings were obtained on dark-adapted animals at the following increasing light intensities: -6, -5, 4, -3, -2, -1, 0, 1.0, 1.5, and 1.9 log cd s / m ² . Recordings in response to photopic flashes were made at 5 minute light adaptation intervals with a background light intensity of 20 cd/m ² , which was also used as background lighting during the recording. The following photopic illumination intensities were used: -2, -1, 0, 1, 1.5 and 1.9 log cd s/m ² .

Histology.

Both eyes were rapidly removed and frozen in liquid nitrogen from a Cnga3-/- mouse six weeks after a unilateral subretinal injection of rAAV2/8.Rho.mR9ap. After cryo-treatment of the eye in OCT medium (RA Lamb, Eastbourne, UK), the eyes were cut into 15 µm transverse sections and air-dried for 15-30 min. For immunohistochemistry, sections were pre-blocked in PBS containing normal donkey serum (2%) and bovine serum albumin (2%) 1 h prior to incubation with anti-RGS9 antibody (1:500, Santa Cruz Biotechnology, Santa Cruz, CA) V

- 16 042590 for 2 hours at room temperature. After rinsing in PBS 2 x 15 min, sections were incubated with the appropriate secondary labeled Alexa 546 antibody (Invitrogen, Carlsbad, CA) for 2 h at room temperature (RT), washed and counterstained with Hoechst 33342 (Sigma-Aldrich, Gillingham, UK). ). Retinal sections were viewed with a confocal microscope (Leica TCS SP2, Leica Microsystems, Wetzlar, Germany).

Western blotting.

Eyes were harvested from Cnga3 -/- mice 4 weeks after unilateral subretinal injection of rAAV2/8.Rho.mR9ap. After separating the neuronal part of the retina from the RPE/choroid/sclera complex, the tissues were homogenized in RIPA buffer and left on ice for 20 min. Samples were centrifuged at 16,000 g for 30 min at 4°C and stored at -20°C until use. Western blot was performed using known protocols.

Optomotor responses and contrast sensitivity function.

Contrast sensitivity and visual acuity of treated and untreated eyes were measured by observing mouse optomotor responses to rotating sinusoidal gratings (OptoMotry™, Cerebral Mechanics, Lethbridge, Alberta, Canada). The protocol used yields independent measurements of right and left eye acuity based on the unequal sensitivity of the two eyes to pattern rotation: the right and left eyes are driven primarily by counterclockwise and clockwise rotations, respectively (Douglas et al., 2005). The mouse was placed on a small island, isolated from the floor, in an enclosed space surrounded by four rotating sine-array monitors with an average illumination of 62 cd/m ² . A double-blind, two-way forced selection procedure was applied in which the observer was blind in the direction of rotation of the pattern, whether it was a treated or untreated Cnga3 -/- mouse, or an age-matched wild-type control animal (C57BL6). Contrast sensitivity, measured at 0.128, 0.256, 0.383, 0.511 cycles/degree, shown at 6 Hz, was defined as 100 divided by the lowest percent contrast, giving a threshold response. Both eyes of each mouse were tested four times on independent days. The data were projected onto a Campbell-Robson contrast sensitivity plot using sinusoidal gratings representing relative spatial frequencies.

Measurement of rhodopsin.

After the mice were fully dark-adapted overnight, the mice were anesthetized and the pupils were fully dilated to assess the rate of visual pigment bleaching. The mice were then placed in a light box with a light source (7.0 mW) directly illuminating the eye for 5 minutes before the eyes were harvested. In another experiment, mice were exposed to identical conditions, the same as for measuring contrast sensitivity, for different periods of time (0, 30, 60, 120 min). The mice's eyes were removed at each time point and placed in 250 μl of phosphate buffered saline, and frozen in liquid nitrogen in a light-opaque tube, and stored at -20°C until use. Some eyes were harvested in the dark under red light after dark adaptation of the mice during the night. Spectrophotometric measurement of rhodopsin was performed as previously reported with minor modifications (Douglas et al., 1995). Briefly, the samples were thawed at room temperature and homogenized. This and all subsequent operations were performed under dim red lighting, which minimally discolored visual pigments. 50 µl of n-dodecyl-v-O-maltoside (200 mM, Sigma-Aldrich, St. Louis, MO) was added to each sample and the resulting mixture was spun for 2 hours at room temperature followed by a 10-minute centrifugation (23,000 g ) at 4°C. The supernatant was removed and placed in a quartz cuvette in a Shimadzu UV-2101PC spectrophotometer (Shimadzu, Kyoto, Japan). After an initial scan of the unbleached extract in the range from 300 to 700 nm, the sample was illuminated with monochromatic light (502 nm) for 3 min, which, as has already been shown, is sufficient for complete decolorization of rhodopsin (Longbottom et al., 2009), and a second scanning. All absorption spectra were zeroed at 700 nm. A difference spectrum was plotted using curves obtained before and after bleaching, and maximum optical densities were determined at about 500 nm, representing the amount of visual pigment released.

Example 6 - Overexpression of R9AP in rods and an increased rate of photoreceptor deactivation.

RGS9, Ge5, and R9AP are mandatory members of the GTPase regulatory complex. To study the effect of AAV-mediated R9AP overexpression on the GTPase complex in rods, the amount and distribution of RGS9 was examined after subretinal injection of rAAV2/8.Rho.mR9ap in Cnga3-/- mice. These mice have normally functioning rods but non-functioning cones and serve as a model for achromatopsia. After four weeks, the treated retina showed increased immunoreactivity against RGS9 across the entire photoreceptor layer in treated retinas compared to untreated retinas (FIG. 4A). Western blot analysis further confirmed increased expression of the RGS9 protein in the treated

- 17 042590 retina (Fig. 4B). These results showed that overexpression of R9AP using AAV2/8 effectively increases the level of the RGS9 catalytic component and the GTPase complex.

The functional effect of AAV2/8-mediated R9AP overexpression on rod phototransduction was then studied using paired flash ERG. (Lyubarsky and Pugh 1996). In this concept, a pair of flashes of the same intensity was given at a variable interval between stimuli and the recovery of the second response relative to the first was measured. In the rod photoreceptor pathway, the rate of recovery of the a-wave (derived from photoreceptors) depends on the rate of deactivation. It was found that the time constant (σ) for 50% a-wave amplitude recovery was reduced by about 60% in Cnga3-/- eyes injected with rAAV2/8.CMV.mR9ap (σ = ~2.99 s) compared to untreated eyes (σ = ~7.38 s; Fig. 4B). Similarly, an increased a-wave recovery rate was observed using the rhodopsin promoter (rAAV2/8.Rho.mR9ap; σ = ~2.74 s; Fig. 4B) in the same mouse strain (Cnga3-/-), or the same virus (rAAV2/8.CMV.mR9ap) in another strain of cone-defective mice (Pde6c-/-; Fig. 8). These observations indicated that subretinal injection of rRAAV2/8.CMV.mR9ap or rAAV2/8.Rho.mR9ap can significantly increase the rate of deactivation of rod phototransduction by increasing the amount of RGS9 and the GTPase complex.

Example 7 - Photopic shift of rod function via R9AP overexpression.

To investigate whether the increased rate of deactivation achieved by overexpression of R9AP and the GTPase complex in rods could alter the operating range of photoreceptor function, ERGs were recorded with a 6 Hz dark-adapted flash using gradually increasing flash intensities. Eyes treated with rAAV2/8.CMV.mR9ap or rAAV2/8.Rho.mR9ap showed increased responses to brighter flashes compared to untreated eyes. This resulted in an increase in the upper response threshold to ~2 log units (Fig. 5A), with little effect on the maximum photoresponse (151 ± 17 μV in treated versus 162 ± 29 μV in untreated eyes, mean ± standard error of the mean). As expected, this photopic shift in the operating range of the rods was accompanied by a reciprocal increase in the lower response threshold to ~1.5 log units. Meanwhile, the upper threshold for ERG responses in wild-type and Gnat1-/- eyes, both those with functional cones, was increased by ~4.0 log units compared to the upper threshold for untreated Cnga3-/- eyes. Similar results were obtained when rAAV2/8.Rho.mR9ap was administered to Pde6c-/- mice (FIG. 9). As a consequence, the treatment allowed the rods to respond to flashes of longer duration (FIG. 5B) and to flashes with cone-isolating background lighting (FIG. 5C). These include conditions under which untreated rods are practically unresponsive.

Taken together, these results indicate that overexpression of R9AP in rods leads to a decrease in their sensitivity and allows cells to acquire photopic function in exchange for scotopic function. This therapeutic induction of a photopic shift in rod function continued for at least 5 months without obvious signs of retinal degeneration (Fig. 10). Meanwhile, treatment of wild-type mice with the same viral vectors showed no measurable change in retinal function (FIG. 11).

Example 8 - The bipolar rod pathway accommodates altered-function rod signaling.

In this work, it was found that overexpression of R9AP in rods leads to a faster kinetics of photoreceptor deactivation and allows the neuron to respond to more photons. Meanwhile, accelerated deactivation should also lead to a shorter duration of neurotransmitter release at the synaptic end of the photoreceptor. Therefore, it was evaluated whether treatment affects downstream bipolar rod signaling. First, the rate and extent of signaling from photoreceptors to bipolar cells in response to a short single flash was studied by measuring the time to peak and amplitude of the a-wave (outgoing from photoreceptors) and b-wave (outgoing from bipolar cells) using ERG (Fig. 6A ). In general, for treated and untreated eyes, a slightly smaller, but almost identical response intensity curve was observed for both waves - a-wave and b-wave. The small difference observed may reflect either a true consequence of accelerated photoreceptor deactivation, or simply neuronal damage caused by subretinal injection. The time to peak of the a-wave denotes the point at which the b-wave triggered by the bipolar cell becomes detectable. We also observed small delays in time to the peak of the a-wave and b-wave, indicating a slight delay in the transmission of neuronal signals from photoreceptors to bipolar cells. However, a relatively large variation in ERG responses between individuals, indicating a slight delay or decrease in rod response, does not necessarily lead to visual dysfunction (Birch and Anderson 1992).

Thus, these results show that, in principle, bipolar cells adapt almost completely to changes in photoreceptor function and, importantly, show an appropriate dose-response relationship.

- 18 042590

Example 9 - R9ap overexpression leads to improved contrast sensitivity function.

It was then asked whether the photopic shift in rod function mediated by R9AP overexpression could consequently be translated into improved visual performance under light exposure by measuring the optokinetic response to rotating sinusoidal gratings under the brightest possible recording conditions using a standard computer monitor. (62 cd/ ^m2 ) (Carvalho et al., 2011). The unique advantage of this behavioral test is that the visual function of each eye can be studied separately; the function of the right eye can be examined by responding to counterclockwise rotating gratings (CCW) and that of the left eye by clockwise (CW) stimuli (Douglas et al., 2005). Spatial contrast sensitivity function (SCF) was studied using a fixed temporal frequency of 6.0 Hz and Cnga3 −/− mice were found to have reduced SCF compared to wild-type mice (FIG. 7). FCF, a function of contrast sensitivity and visual acuity, reflected the estimated range of visual perception (animals could apparently perceive gratings under the curve, but not on it; Fig. 7A). Very curiously, an 8.0-fold (P=0.005) and a 5.4-fold (P=0.011) increase in sensitivity was observed using both 0.128 and 0.256 cycles/degree (c/g) gratings, respectively, when contrast sensitivity of the treated and untreated eye (Fig. 7A, left). No clear change in contrast sensitivity was observed for the 0.383 (P=0.056) and 0.511 (P=0.111) gratings (c/g). Interestingly, the average sensitivity of the treated eye in Cnga3 -/- mice exceeded the average sensitivity of wild-type controls with normally functioning cones. However, when wild-type mice were treated with the same viral construct, FCC did not differ between treated and untreated eyes (FIG. 12).

Having established that R9AP overexpression leads to increased visual performance when viewing the monitor's brightest settings, it was proposed to determine whether this increase in visual perception is sustainable. This is of great concern given that regeneration of visual pigment in rods is known to be significantly slower than regeneration in cones (Wang and Kefalov 2011). First, it was assessed whether the treatment had an effect on changing the rate of decolorization of the visual pigment. It was found that exposure to bright light in treated and untreated eyes for 5 minutes did not produce any difference in the amounts of residual visual pigment bleaching (FIG. 7B lower left panel). Second, the amount of decolorized rhodopsin in the eye was studied after exposing Cnga3 −/− mice for varying amounts of time to the same experimental parameter used to measure the FCF. The results showed that the amount of visual pigment remained stable without showing a decrease over 2 hours of exposure to visual stimuli, similarly for treated and untreated eyes (lower right panel in Fig. 7B). These results showed that the gain in visual perception in treated Cnga3 -/- mice is maintained by a sufficient amount of rhodopsin molecules and is stable.

Claims (14)

CLAIM 1. A method for improving vision in a patient with dysfunction and/or degradation of cone photoreceptors, comprising: providing an adeno-associated virus (AAV) vector containing a nucleic acid encoding ArchT protein and/or R9AP protein, and introducing said vector into healthy rod photoreceptors in the patient's retina to express said nucleic acid encoding ArchT protein and/or R9AP protein, in them in such a way that the range of illumination intensities to which the photoreceptor-rod responds increases, and / or the speed with which the photoreceptor-rod reacts to light increases, where the capsid of the AAV vector is derived from AAV8, and the nucleic acid is expressed under the control of the promoter rod-specific. 2. The method of claim 1, wherein the genome of the AAV vector is derived from AAV2. 3. The method according to any one of the preceding claims, wherein the patient suffers from macular degeneration, achromatopsia or Leber's amaurosis. 4. The method of claim 3, wherein the macular degeneration may be age-related macular degeneration (AMD), such as hereditary macular degeneration, or hereditary cone degeneration. 5. The method according to claim 4, where the ARTD is a wet (neovascular) URLD or geographic retinal atrophy. 6. A method according to any one of the preceding claims, wherein the rod photoreceptor signaling is extended into the mesopic and/or photopic range of illumination. 7. The method according to any one of the preceding claims, wherein the rod exhibits improved rate of change and/or accelerated activation/inactivation kinetics. 8. The method according to any one of the preceding claims, wherein the vector is introduced into rod photoreceptors in vitro followed by transplantation into the retina. 9. A method according to any one of the preceding claims, wherein the patient's mesopic and/or photopic vision is improved. 10. The method of any one of the preceding claims, wherein said promoter is a rhodopsin (Rho) promoter, a retinal neuron-specific leucine zipper (NRL) protein promoter, or a phosphodiesterase 6B (PDE6B) promoter. 11. Use of an adeno-associated virus (AAV) vector in the manufacture of a medicament for the treatment of dysfunction and/or degradation of cone photoreceptors, wherein the AAV vector contains an expression cassette containing a nucleic acid encoding a gene product operably linked to a rod-specific promoter, wherein said gene product is ArchT or R9AP and the AAV vector capsid is derived from AAV8. 12. Use according to claim 11, wherein said promoter is a rhodopsin (Rho) promoter, a leucine zipper protein specific for retinal neurons (NRL), or a phosphodiesterase 6B (PDE6B) promoter. 13. Use according to claim 11, wherein the AAV vector contains a nucleic acid encoding an ArchT protein or an R9AP protein, wherein the AAV vector capsid is derived from AAV8 and said nucleic acid is under the control of a rod-specific promoter. 14. Use according to any one of claims 11-13, wherein the genome of the AAV vector is derived from AAV2.