Skip to main content
NCBI home page
Molecular Therapy logoLink to Molecular Therapy
. 2015 Jun 2;23(8):1308–1319. doi: 10.1038/mt.2015.68

CNTF Gene Therapy Confers Lifelong Neuroprotection in a Mouse Model of Human Retinitis Pigmentosa

Daniel M Lipinski 1,2,3, Alun R Barnard 1,2, Mandeep S Singh 1,2, Chris Martin 4, Edward J Lee 5, Wayne I L Davies 1,6, Robert E MacLaren 1,2,5,*
PMCID: PMC4539573  EMSID: EMS64286  PMID: 25896245

Abstract

The long-term outcome of neuroprotection as a therapeutic strategy for preventing cell death in neurodegenerative disorders remains unknown, primarily due to slow disease progression and the inherent difficulty of assessing neuronal survival in vivo. Employing a murine model of retinal disease, we demonstrate that ciliary neurotrophic factor (CNTF) confers life-long protection against photoreceptor degeneration. Repetitive retinal imaging allowed the survival of intrinsically fluorescent cone photoreceptors to be quantified in vivo. Imaging of the visual cortex and assessment of visually-evoked behavioral responses demonstrated that surviving cones retain function and signal correctly to the brain. The mechanisms underlying CNTF-mediated neuroprotection were explored through transcriptome analysis, revealing widespread upregulation of proteolysis inhibitors, which may prevent cellular/extracellular matrix degradation and complement activation in neurodegenerative diseases. These findings provide insights into potential novel therapeutic avenues for diseases such as retinitis pigmentosa and amyotrophic lateral sclerosis, for which CNTF has been evaluated unsuccessfully in clinical trials.

Introduction

Identifying means of treating neurodegeneration remains one of the unmet challenges of human health. The unique optical properties of the eye make it an ideal organ in which to assess the therapeutic potential of neuroprotective compounds, where the survival of retinal neurons can be visualized directly through the pupil.

Several compounds have been demonstrated to have some efficacy in the preservation of retinal neurons, including brain-derived neurotrophic factor,1 glial cell line-derived neurotrophic factor,2,3 and ciliary neurotrophic factor (CNTF). CNTF in particular has been shown to be highly protective against retinal ganglion cell death in models of oxidative stress and experimental glaucoma,4,5,6,7,8 and against photoreceptor loss in several animal models of retinal disease, including retinitis pigmentosa (RP).9,10,11 A number of studies have previously demonstrated that sustained expression of CNTF following recombinant adeno-associated virus (rAAV)-mediated gene delivery results in robust preservation of photoreceptor cell bodies for a period of months in rodent models of RP.12,13,14 Although these studies observed that CNTF treatment led to anatomical preservation of photoreceptors, the majority noted significant electrophysiological suppression, indicating that CNTF may not be an appropriate neuroprotective agent for clinical use. While initially promising, several factors complicate the interpretation of previously reported findings, including the commencement of treatment prior to photoreceptor degeneration, the inability to effectively control CNTF dosing, and a reliance on electrophysiological measures of retinal function, rather than assessments of visually guided behavior, to determine the extent of functional vision remaining.

We have shown previously that degenerative retina harvested from a reporter mouse with intrinsically fluorescent cone photoreceptors can be used to assess short-term survival of retinal neurons ex vivo in response to the presence of neuroprotective proteins.15 However, due to the slowly progressing nature of neurodegenerative disorders such as RP, and the short-term nature of previous studies, it remains unclear whether such neuroprotection might be maintained throughout the lifetime of an animal, and critically, whether preserved neurons retain function. To address these questions, we employed a novel method of in vivo analysis to quantify cone photoreceptor survival in response to rAAV-mediated CNTF expression over a period of seven months.16,17,18 Critically, we utilized variable dosing in order to preserve cone photoreceptors at a stage when rod degeneration was already well advanced, closely mimicking the ocular phenotype of many RP patients when presenting to the clinic. By combining standard assessments of electrophysiological function with visually guided behavioral testing, we aimed to determine the extent of functional vision remaining following long-term CNTF treatment while correlating our observations to evidence of higher signal processing in the visual cortex. Finally, the mechanisms underpinning CNTF-mediated neuroprotection were explored through transcriptome (RNA-seq) analysis, revealing a novel pathway by which neurons are prevented from degeneration in spite of significant cellular stress.

Results

To model accurately cone survival and function in response to long-term CNTF therapy, C57B/6.129 Rhotm1Phm mice homozygous for knockout of rhodopsin (Rho) were crossed with C57BL/6JTgOPN1LW-EGFP85933Hue transgenic mice that express enhanced green fluorescent protein (EGFP) in cone photoreceptors.16,17 In this model of end-stage RP (termed herein, Rho-/-, Tg(OPN1LW-EGFP)+/0), apoptosis of rod photoreceptors due to the absence of rhodopsin protein occurs by postnatal week (PW) 8 and is followed by the secondary loss of intrinsically fluorescent cones, leading to total photoreceptor degeneration.19 The utilization of an animal model whereby cone photoreceptors degenerate secondary to advanced rod loss reflects the disease phenotype most commonly observed in RP patients (rod-cone dystrophy). Furthermore, the model closely recapitulates the clinical scenario in which neuroprotection would most likely be applied, where the majority of rod photoreceptors have degenerated and the priority is to preserve cone mediated vision irrespective of the etiology underlying photoreceptor loss. This approach is particularly relevant to RP, where the cause of photoreceptor loss is incompletely understood in 50% of autosomal dominant and 30% of autosomal recessive cases.18,20

A recombinant adeno-associated virus serotype 2 (rAAV2/2)-based vector was manufactured (1 × 1013 genome particles (gp) per ml) to express human CNTF (hCNTF) protein modified for extracellular release through addition of an upstream human nerve growth factor secretion signal. Protein secretion from the resultant rAAV2/2.hCNTF vector was validated in vitro by the transduction of HEK293 cells (MOI = 1,000), resulting in high levels of hCNTF protein being detectable in the media by enzyme-linked immunosorbent assay (ELISA; Supplementary Figure S1a). In vivo administration rAAV2/2.hCNTF into the vitreous cavity (intravitreal injection) of PW4 Rho-/-, Tg(OPN1LW-EGFP)+/0 mice (n = 3 per group) similarly resulted in high levels of secreted hCNTF protein, most likely from ganglion and Müller glia cells which are known to be transduced efficiently by AAV2-based vectors. Importantly, we observed that altering the number of genome particles administered per eye could broadly control the effective dose of hCNTF protein (Supplementary Figure S1b).

CNTF prevents development of RP in a dose-dependent manner

Rho-/-, Tg(OPN1LW-EGFP)+/0 mice received an intravitreal injection of rAAV2/2.hCNTF at low (2 × 108 gp; n = 8 mice), medium (2 × 109 gp; n = 6 mice), or high (2 × 1010 gp; n = 5 mice) dose prior to significant cone photoreceptor loss (PW4). The contralateral eye in all mice received an equivalent volume (2 μl) sham injection of phosphate-buffered saline (PBS) to control for the effects of the surgical intervention.

Four weeks postinjection (PW8), the retinae of injected mice were imaged with a confocal scanning laser ophthalmoscope (cSLO), which allows noninvasive assessment of retinal pathology by direct visualization of the fundus through the pupil. Specifically, the cSLO autofluorescence (AF) imaging mode utilizes a laser with an excitation wavelength proximate to the absorption peak of EGFP, allowing the visualization of EGFP-expressing cells within the retina.21,22 AF imaging of the Rho-/-, Tg(OPN1LW-EGFP)+/0 mouse fundus revealed a dot-like pattern (Figure 1a) corresponding to the expected distribution of EGFP-expressing cone photoreceptors across the retina. Postmortem lectin staining (Figure 1c) of retinas isolated from Rho-/-, Tg(OPN1LW-EGFP)+/0 mice immediately following AF imaging confirmed that individual “dot-like” signals observed in vivo corresponded to single EGFP-expressing cone photoreceptors (Figure 1a–c). Using vascular landmarks to define a standardized region of interest within each eye that would remain constant throughout the life of an experimental animal, it was possible to quantify accurately cone photoreceptor numbers and to assess survival longitudinally in response to hCNTF treatment (Figure 1d–g; Supplementary Figure S1c–j).

Figure 1.

Figure 1

Open in a new tab

Long-term quantification of intrinsically fluorescent cone photoreceptors through repetitive in vivo imaging following rAAV2/2.hCNTF treatment. (a) Autofluorescence (AF) image of Rho-/- Tg(OPN1LW-EGFP)+/0 mouse fundus showing cone photoreceptors represented as punctate white dots. (a') Increase magnification AF image demonstrating a distinctive group of cone photoreceptors (aligned β to γ) and individual cones (white arrows) at the nasal vascular bifurcation (α, dotted line). (b) Near infrared reflectance (NIR) fundus image detailing the retina vasculature and (b') magnified NIR image detailing the nasal vascular bifurcation (α, dotted line). (c) flat mounted post-mortem retina with retinal vasculature counterstained with lectin. (c') magnified image of the retinal flat mount demonstrating the distribution of cone photoreceptors (green dots) at the nasal bifurcation (α, dotted line). Note that the pattern of cones in panel (c') correlates precisely with that observed by in vivo AF imaging (a'); including a distinctive group of cone photoreceptors (aligned β to γ) and individual cones (white arrows). (d–g) Repetitive AF fundus imaging of a Rho-/-, Tg(OPN1LW-EGFP)+/0 mouse treated with high dose rAAV2/2.hCNTF from postnatal week (PW) 8 to PW30. (d'–g') Magnified AF images showing how retinal vascular l lifetime of each animal. Cone photoreceptors (white dots) can be quantified (red dot overlay) within the regions of interest allowing highly reproducible longitudinal assessment of cone survival. Survival is expressed as a function (%) of cone photoreceptor numbers at baseline (PW8). All scale bars ~100 µm. (h) Survival of cone photoreceptors expressed as a function (%) of cone photoreceptor numbers at baseline (PW8) for each treatment group: high dose (dark blue), medium dose (bright blue), low dose (pale blue) or PBS sham (grey). hCNTF treatment has a significant effect on cone photorecptor survival: *P < 0.05; **P < 0.01; ***P < 0.001, two-way repeated measures analysis of variance with dose and time as factors. High-dose group n = 5, medium-dose group n = 6, low-dose group n = 8, at PW30.

AF imaging demonstrated a comparable distribution of EGFP-expressing cone photoreceptors in all treatment and control groups at PW8 (Figure 1h; Supplementary Figure 2a–d). Cone photoreceptor numbers declined rapidly in the PBS sham treated-group over the experimental time course, with cone numbers reaching zero at PW24 (Figure 1h, Supplementary Figure S2d). Consistent with the absence of detectable hCNTF protein (Supplementary Figure S1b), cone photoreceptor numbers in eyes that received a low-dose (2 × 108gp) intravitreal injection of rAAV2/2.hCNTF were not significantly different to paired PBS sham-injected eyes at any time point (Figure 1h, Supplementary Figure S2c; P ≥ 0.05, repeated measures two-way analysis of variance (ANOVA) with Bonferroni post-test, n = 8). Near-infrared reflectance (NIR) imaging revealed focal geographic changes in the retinal pigment epithelium consistent with loss of the overlying photoreceptors, a common clinical observation in late-stage RP (Figure 2a–d, Supplementary Figure S2c,d). This observation is supported by histology conducted at PW30 that demonstrated the absence of both rod and cone photoreceptors in PBS sham treated eyes (Figure 2i).

Figure 2.

Figure 2

Open in a new tab

Near infrared reflectance (NIR) and autofluorescence (AF) fundus imaging of high dose rAAV2/2.hCNTF-treated and PBS sham-treated mice at PW8 and PW30. Sham-treated controls demonstrate areas of increased reflectance indicative of retinal pigment epithelium (RPE) atrophy (a,c) and loss of EGFP-expressing cone photoreceptors (b,d) by PW30. High-dose rAAV2/2.hCNTF-treated eyes show no signs of RPE atrophy (e,g) and minimal loss of cone photoreceptors (f,h) over the same period. Representative histology of phosphate-buffered saline (PBS) sham treated (i) retina at PW30 demonstrates and absence of both rod and cone photoreceptors. Histology of high dose rAAV2/2.hCNTF treated (j–o) retina at PW30 demonstrating preservation of cone (GFP-expressing) and rod (non-GFP-expressing) photoreceptors in the outer nuclear layer (ONL). (j,k) Cone photoreceptors (l–o) express Mws-opsin (arrow head) in the correct cellular compartment (outer segment (OS)) with no evidence of opsin mis-localization in the cell body (*). Cone photoreceptor morphology indicates compression between the inner nuclear layer (INL) and RPE due to the loss of multiple nuclear rows following rod degeneration.

Cone numbers in eyes treated with a medium (2 × 109 gp) or high (2 × 1010 gp) dose of rAAV2/2.hCNTF vector were significantly greater in comparison to contralateral control eyes from PW12 onwards (Figure 1h; Supplementary Figure S2a,b). Neuroprotection of cone photoreceptors was dose dependent, with greatest survival being observed in high dose treated eyes (62.38 ± 1.95%) compared to medium dose-treated eyes (44.90 ± 1.69%; PW30 versus PW8). The rate of cone photoreceptor loss was near-zero from PW15 onwards in both treatment groups (Figure 1h; 0.42–0.69%/week); no RP phenotype was observed in medium or high dose-treated eyes (Figure 2e–h, Supplementary Figure S2a,b). Histology at PW30 revealed preservation of both cone (GFP-expressing) and rod (non GFP-expressing) photoreceptors in eyes receiving high dose rAAV2/2.hCNTF vector (Figure 2j). The absence of EGFP colocalization with medium wavesensitive (Mws) cone opsin expression strongly indicates the presence of discrete cone outer segments (Figure 2k–o), where the connecting cilia effectively prohibits the trafficking of EGFP protein from the photoreceptor cell body the outer segments when present.

CNTF preserves visual function despite mild suppression of electrophysiological responses

rAAV2/2.hCNTF-treated mice were examined at PW8 by electroretinography (ERG), a technique that utilizes a corneal electrode to indirectly measure the massed electrical response of the retina following stimulation with light. In line the with evidence of previous studies that demonstrated reduced cone sensitivity in the presence of CNTF protein, cone photoreceptor-mediated responses was observed to be suppressed in a dose-dependent manner23 (Supplementary Figure S3a–f). Peak ERG recordings fell below the mean amplitude of background noise (~20 μV) in all groups by PW12 and were consequently considered to be unrecordable. While the reduction in peak ERG amplitude coincided with a period of significant cone photoreceptor loss in all groups, medium (2 × 109 gp) and high (2 × 1010 gp) dose treated eyes demonstrated significant preservation of cones beyond PW12, potentially indicating the absence of cone-mediated vision despite preservation of cone photoreceptor cell bodies (Figure 2j–o). We therefore explored visual function in more detail with further electrophysiological and behavioral testing at PW30.

The ability of remaining cones to relay signals through the visual pathway to the brain was assessed directly by performing laser speckle imaging (LSI) of the visual cortex during stimulation of the retina by light. LSI is a noncontact imaging technique with high temporal and spatial resolution that can detect changes in random speckle patterns that are linearly associated with increased movement of red blood cells in the microvasculature (Figure 3c). Herein, LSI was used to measure changes in the cortical blood flow (CBF) in the visual cortex following monocular stimulation with brief (10 ms) flashes of 14.5cd monochromatic light (510 ± 3 nm), where the wavelength is proximate to the maximal peak of absorption of the murine M-cone. The ability of the chosen stimulation parameters to elicit a cone-mediated response was validated on PW6 untreated Rho-/-, Tg(OPN1LW-EGFP)+/0 mice (n = 3). Changes in CBF were detected primarily in the visual cortex contralateral to the eye stimulated, indicating appropriate down-stream processing of the visual stimulus (Figure 3a–c). Cortical LSI at PW30 during stimulation of eyes treated with a medium dose (2 × 109 gp) of rAAV2/2.hCNTF vector revealed increased, but not significant, CBF changes in the contralateral visual cortex (Figure 3d–f; P = 0.20, one-tailed Wilcoxon matched-pairs, n = 6). By contrast, LSI during stimulation of eyes that received a high dose (2 × 1010 gp) of rAAV2/2.hCNTF demonstrated an increased CBF in the contralateral visual cortex that was significantly greater than the response elicited during stimulation of the sham-treated control eyes (Figure 3c,g–i; P = 0.012, one-tailed Wilcoxon matched-pairs, n = 4). When compared to baseline LSI CBF measurements obtained from PW6 animals (+0.819%), the magnitude of CBF change elicited following stimulation of high (+0.509%) dose CNTF-treated eyes broadly correlates to the percentage cone photoreceptor survival observed at PW30 (62.38 ± 1.95% of PW8). Cortical LSI strongly indicated that surviving cones remained light sensitive and were able to relay signals to the brain. The extent to which changes in cortical blood flow corresponded to the processing of useful vision was assessed by measurement of optomotor response (OMR). OMR measures the number of involuntary head-tracking movements an experimental animal makes in response to a full-field visual stimulus, classically consisting of a brightly lit (~1,000 lux) rotating drum with vertically orientated black and white stripes of a defined width and contrast (100% contrast; 0.2 cycles/degree) into which the animal is placed (Figure 4a). OMR is characterized by a slow-phase head-tracking motion in the direction of the drum's rotation, followed by a rapid-phase reorientation of the head to a central position. The slow phase head-tracking movement is driven by each eye independently and relies on the direction of the drums rotation (Figure 4a).24,25,26 When experimental animals were examined at PW30, the number of head tracking responses per minute recorded from sham treated eyes was not significantly above zero in any group (0.33 ± 0.05/minute). The number of head tracking responses elicited by stimulation of medium dose (2.11 ± 0.31/minute) and high dose treated eyes (6.92 ± 0.92/minute) were significantly greater than paired sham eyes (Figure 4b,c; P ≤ 0.001, Kruskal-Wallis test with Dunn's multiple comparisons, n = 4), while linear regression analysis of OMR responses versus residual cone number at PW30 revealed a significant positive correlation (r2 = 0.80, F = 33, P = 0.0004) (Supplementary Figure S4a).

Figure 3.

Figure 3

Open in a new tab

Assessment of cone photoreceptor function in PW30 rAAV2/2.hCNTF treated mice through laser speckle imaging of the visual cortex. (a,b) change in cortical blood flow (CBF) in untreated Rho-/-TgOPN1LW-EGFP+/- mice at PW6 demonstrating that laser speckle imaging (LSI) detect changes in micro-vascular blood flow primarily in the visual cortex following stimulation with a 510 nm flicker stimulus. (c) Schematic representation of the visual pathway in mice demonstrating that visual input is primarily processed in the visual cortex located in the hemisphere contralateral to the eye stimulated. White arrow = superior sagital sinus; arrow head = lambda; OS = oculus sinister (left); OD = oculus dexter (right). Changes in CBF were examined in (d,e) medium and (g–i) high dose treated Rho-/-, Tg(OPN1LW-EGFP)+/0 mice at PW30 following independent stimulation of PBS-treated (d,g) or rAAV2/2.hCNTF-treated (e,h) treated eyes, revealing significant differences in CBF from high-dose-treated eyes (i) compared to sham controls. ns, not significant. *P < 0.05, Wilcoxon matched-paired signed rank test. LSI: high dose n = 3; medium dose n = 6.

Figure 4.

Figure 4

Open in a new tab

Cone-photoreceptor mediated vision is preserved only in rAAV2/2.hCNTF treated eyes at PW30. (a) schematic representation of the optomotor response (OMR) in relation to the direction of the drums rotation. (b) Number of head-tracking responses (three tests per animal) from individual mice at PW30. Error = SEM; X = zero responses; L = low dose, M = medium dose, H = high dose. (c) Mean group responses to OMR testing at PW30 demonstrating significantly greater number of head-tracks from medium (bright blue) and high dose (dark blue) compared to paired PBS sham-treated controls. ns, not significant. ***P < 0.001, one-way analysis of variance. Low dose n = 5, medium dose n = 6, high dose n = 4.

Collectively, these data strongly indicate that despite the apparent suppression of electrophysiological function as assessed by ERG, cone photoreceptors preserved by hCNTF-mediated neuroprotection remain sensitive to light, and are able to usefully relay signals via the optic nerve to central visual pathways.

CNTF upregulates complement factor proteins and inhibitors of proteolysis

Transcriptome analysis of 23,365 mouse genes was undertaken to determine the mechanisms underlying long-term neuroprotection of cone photoreceptors in response to CNTF. Transcriptomes of treated eyes were compared at each dose to paired sham treated controls (n = 4 eyes per group) by next-generation sequencing of total mRNA extracted from whole eyes at PW30. The false discovery rate was fixed at 2% following the Benjamini-Hochberg procedure and all samples were normalized with respect to the effective library size. When comparing the transcriptomes of high dose treated (2 × 1010 gp) eyes with paired PBS sham-treated controls, 1,533 genes were found to show significant differential expression (P < 0.05; Supplementary Figure S4b). Of those, 460 genes demonstrated a >2-fold change in their expression levels and were subsequently grouped based by cellular function using gene ontology (Figure 5a,b). Serine-, cysteine-, and metallo-type peptidase inhibitors (17 genes; Supplementary Table S1.1) comprised the most transcriptionally upregulated family, with members demonstrating up to 89-fold greater expression in high dose hCNTF-treated eyes compared to paired sham-treated controls. Members of this group encode intracellular and extracellular peptidase inhibitors, dysfunction of which have previously been associated with the development of retinal disease; such as tissue inhibitor of metalloproteinase 3 (Timp3, +3.49-fold), mutations in which are linked to autosomal dominant Sorsby's fundus dystrophy.27 Transcriptional regulation of peptidase inhibitors is not well understood, but activation appears to be signal transducer and activator of transcription 3 (Stat3) mediated (+2.46-fold expression) and require members of the CCAAT/enhancer binding protein (Cebp) family of transcription factors, three of which (Cebpα, Cebpβ, and Cebpδ) were also significantly upregulated (Supplementary Table S1.2).28,29

Figure 5.

Figure 5

Open in a new tab

RNAseq analysis comparing the transcriptional profile of rAAV2/2.hCNTF eyes with contralateral PBS sham treated controls revealed upregulation of proteolysis inhibitors and complement. (a) Hierarchical clustering of transcriptome output from low-, medium-, and high-dose rAAV2/2.hCNTF-treated mice showing altered gene expression relative to paired PBS sham-treated controls (n = 4 mice per group) at PW30, where the solid blue line in each column represents fold-change compared to baseline (dotted blue line). Two clusters relating to genes with the highest transcriptional changes are enlarged: the upper panel shows downregulation of genes involved in ion transport and visual cycle; the lower panel shows widespread upregulation of genes encoding serine-, cysteine-, and metallopeptidase inhibitors. (b) Clustering of genes based function using genome search meta analysis ontology visualized using the enrichment map plugin for cytoscape.

Several members of the classical and alternative complement cascades (16 genes, Supplementary Table S1.3) were significantly up regulated (up to 10-fold) including various complement component (e.g., C3, C4a) and complement factor genes (e.g., Cfb). Numerous modulators of cytokine signaling were upregulated (19 genes, up to eightfold; Supplementary Table S1.4), including suppressors of cytokine signaling 3 (SOCS3) and 5 (SOCS5), which are known to be expressed in photoreceptors and reduce stress-induced inflammatory responses.30,31

Downregulation (1.7-fold) of lecithin retinol acyltransferase, a critical component of the vitamin-A cycle, and several genes encoding ion channels and G protein-coupled receptors expressed in inner retinal neurons was observed in medium and high dose rAAV2/2.hCNTF treated eyes (Supplementary Table S1.5) indicating widespread alteration of phototransduction in response to CNTF treatment, in line with previous studies.23

Discussion

Through exploitation of the eye's unique optical properties, we were able to assess progressive neuronal degeneration in real-time and accurately quantify the survival of intrinsically fluorescent cone photoreceptors through repetitive in vivo imaging. This work demonstrates for the first time the ability of a neurotropic compound to provide life-long protection in a model of degenerative disease.

Our data demonstrate that dose was observed to be a critical factor in the therapeutic effectiveness of CNTF-mediated neuroprotection, and that cone photoreceptors can be preserved by rAAV-mediated hCNTF secretion beyond the point at which total outer-retinal degeneration would be expected without intervention. Stabilization of cone photoreceptor numbers was observed in medium (44.90 ± 1.69%) and high (62.38 ± 1.95%) dose treatment groups suggesting that protection may be maintained long-term. These findings are significant in light of recent clinical trials utilizing encapsulated cell technology to provide sustained CNTF delivery, where photoreceptor preservation has been demonstrated in the short to medium term (up to 5 years), but the likely outcomes of long-term neuroprotection remain unknown due to the slow progressive nature of RP in humans.32,33

In the present study, cell survival was observed to be dose-dependent, yet so to was physiological suppression of function as assessed by electroretinography (ERG); a dose was not observed at which neuroprotection was achieved without functional suppression, as had been previously proposed.34 The number of photoreceptors preserved, and the sensitivity of each photoreceptor to light, appears to be directly proportional to the bioavailability of hCNTF protein. This strongly suggests that the mechanisms underlying functional suppression and cell survival may be inextricably linked, raising complex questions with regard to the clinical application of hCNTF in the treatment of retinal disorders. In particular, whether suppression of visual function during the early stages of a progressive degenerative disease, when patients might otherwise have good visual acuity, is outweighed by the potential to extend vision beyond the point at which blindness would have occurred without therapy?

The time point of intervention is another factor that is likely critical to successfully preserving cone photoreceptors and maintaining function. In the present study, we aimed to model closely the clinical scenario whereby a significant proportion of rod photoreceptors are absent at the point of maximum CNTF expression. This was achieved by timing rAAV.hCNTF administration so that the peak of transgene expression would not occur until PW8-10, by which point rod degeneration is well advanced in the rhodopsin knockout mouse model.17 Our findings that a significant number of cone photoreceptor can be preserved, even when intervention occurs once retinal degeneration is relatively advanced, supports recent trials conducted in patients with end-stage RP that observed robust cone photoreceptor preservation.32,33 While our findings are promising, it is worth noting that the histology presented herein clearly demonstrates the preservation of rod photoreceptors (non-GFP-expressing cells) in addition to cones following CNTF treatment. It is unclear from the present study whether preservation of rods is a requirement for cone survival, or how this observation might correlate to the human eye, where cones are concentrated in the fovea to the exclusion of rod photoreceptors.

Interestingly, despite the early point of intervention, preclinical studies uniformly failed to observe functional preservation, suggesting that maintenance of vision may not be possible even when treatment is given prior to the onset of retinal degeneration. In line with those studies, we initially utilized ERG as a primary assessment of residual electrophysiological function, and observed that measurements became unrecordable in all groups by PW12. As ERG is an indirect measurement of retinal function, dependent upon the propagation of an electrical signal of sufficient amplitude and uniformity to be detected by electrodes placed on the corneal surface, the absence of recordable cone ERGs beyond PW12 was thought to not necessarily reflect an absence of functional vision. Indeed, it is quite possible for small areas of the retina to be light sensitive and yet not propagate a detectable signal, a principle best highlighted by studies focusing on photoreceptor transplantation and optogenetic approaches to restore light sensitivity to degenerate retinae, where functional vision can be partly restored and yet ERG responses almost certainly remain absent.35,36,37,38

A contributing factor to the absence of ERG responses observed in the present study beyond PW12 may be suppression of outer segment formation, as has been noted with rod photoreceptors following CNTF treatment.39 Studies into the effects of CNTF on cone outer segment formation are limited; however, we observed the persistence of defined opsin-containing cone outer segments at PW30, strongly indicating that cones may remain light sensitive despite an absent ERG.40,41,42 To examine the hypothesis that cone photoreceptors retained the ability to relay interpretable signals to the visual cortex and pretectal area of the brain, we performed laser speckle imaging (LSI). LSI during unilateral ocular stimulation proved a robust method by which to assess the light sensitivity of remaining cones, crucially allowing the delineation of responses from treated and untreated eyes separately within each animal. Cortical imaging revealed that stimulation of high dose rAAV.hCNTF-treated eyes elicited a significant change in cortical blood flow in the contralateral visual cortex, strongly suggesting that residual cone photoreceptors were receptive to light and remained correctly synapsed. Importantly, the propagation of an optomotor response revealed that the signals received by the visual cortices encoded physiologically relevant information. Head tracking responses were not elicited following stimulation of sham-treated control eyes, supporting the evidence that intrinsically photoreceptive retinal ganglion cells, which are present in the inner retina even in end-stage degeneration, do not contribute to OMR.43 Together these findings strongly suggest that functional vision may have been maintained in previous AAV-mediated CNTF gene therapy studies, but that due to the reliance on ERG as a primary outcome measure for retinal sensitivity, it may have remained undetected. The findings are also of particular relevance to clinical trials expressing hCNTF protein using encapsulated cell technologies, as they demonstrate that despite short-term functional suppression, useful vision may still be preserved in end-stage disease and beyond the point at which at which blindness may have otherwise occurred.

The impact of CNTF dosing appears to be critical not only in the treatment of RP, where photoreceptors are functionally suppressed in line with increased dose, but is an important consideration when evaluating treatments for other neurodegenerative diseases. In particular, past unsuccessful clinical trials aimed at preventing motor neuron loss in ALS demonstrated an absence of therapeutic efficacy and significant incidence of adverse events in patients treated systemically with high doses of CNTF.44,45 These findings indicate that a more targeted approach may be required than offered by current neuroprotective strategies, whereby the beneficial mechanisms underlying cell survival are isolated from those resulting in adverse effects. In this study, we were able to reveal unique insights into the potential mechanisms of CNTF-mediated neuroprotection following sustained expression, which until now have remained incompletely understood, by conducting a transcriptome analysis of surviving neurons in end-stage retinal disease. A transcriptome analysis of rAAV2/2.hCNTF-treated versus sham-treated eyes revealed several important findings. First, significant overexpression of the Stat3 gene was observed in medium and high dose rAAV2/2.hCNTF-treated eyes, consistent with previous studies showing that CNTF signals via a receptor complex of Cntfr/Lifrβ/Gp130 activity of mediated by Stat3.42,46,47,48

Second, several families genes encoding inhibitors of proteolysis were robustly overexpressed (up to 89-fold) indicating a key role in hCNTF-mediated neuroprotection. Activation of members of the serine protease inhibitor (Serpin) family (11 genes) in particular are closely linked to Stat3 expression levels and have proved to be antiapoptotic when overexpressed in tumors.49 Inhibitors of proteolysis may play a crucial role in retinal disease, including Timp3 and Serpin1b, which act to prevent degradation of extracellular matrix components and Bruch's membrane, respectively, and are implicated in Sorsby's fundus dystrophy and age-related macular degeneration.27,50,51 With the physical degradation of neurons, such as photoreceptors, in neurodegenerative diseases being typically mediated by serine-cysteine proteases, lysosomal proteases, and complement mediate lysis, we hypothesis that the overexpression of endogenous inhibitors of proteolysis observed herein may provide a critical defense mechanism against neuronal apoptosis in response to cellular stress. We propose that direct overexpression of secreted proteolysis inhibitors may provide a novel therapeutic avenue for the prevention of neuronal cell death in degenerative diseases. Indeed, utilizing a targeted approach aimed at preventing photoreceptor degeneration in RP through direct inhibition of cell body and extracellular matrix degradation may enable photoreceptors to be preserved without physiological suppression of function. In the case of ALS, preventing degeneration of cortical motor neurons directly through inhibition of proteolysis may allow treatment efficacy to be achieved without the significant adverse effects observed in clinical trials when CNTF was delivered systemically at high dose.44,45

Last, a sustained innate immune response characterized by significant upregulation (up to 10-fold) of several members of the classical and alternative complement cascades (16 genes) was observed in medium and high dose-treated eyes, including two complement factors associated with age-related macular degeneration (encoded by the Cfi gene) and retinal light damage (encoded by the Cfd gene).52,53 Interestingly, no complement or cytokine upregulation was observed in low-dose-treated eyes at PW30, indicating that the innate immune response ceases once photoreceptor degeneration is complete. Although the role of several upregulated protease inhibitors remains unknown, the expression of genes that encode proteolysis inhibitors may also be particularly relevant to modifying innate immune responses occurring in the central nervous system, where various components (e.g., C3, C4 and Cfb proteins) are activated only following proteolytic cleavage. Serpin3k, for example, is known to have strong anti-inflammatory and antiangiogenic properties, inhibiting tumor necrosis factor-α (Tnf-α), intracellular adhesion molecule 1 (Icam-1), and vascular endothelial growth factor (Vegf) through blockade of the Wnt pathway.54 In addition to preventing retinal necrotic cell death, Serpin3k also significantly inhibits rod outer segment generation, and thus its upregulation by hCNTF may also contribute to reduced retinal sensitivity.54,55,56

In summary, this study has demonstrated that the unique optical properties of the eye allow for neuronal survival in response to a therapeutic intervention to be assessed longitudinally and with great accuracy through repetitive in vivo imaging. Herein, we have demonstrated that sustained hCNTF expression leads to life-long preservation of cone photoreceptors, and that despite limited electrophysiological suppression, useful vision was maintained until the end-stages of degeneration. Importantly, this study demonstrated that robust cell survival is directly linked to the dose of CNTF available. Transcriptome analyses strongly indicate that Stat3-mediated overexpression of intracellular and extracellular protease inhibitors underlie cone preservation in RP following hCNTF treatment, possibly through direct inhibition of cellular and extracellular matrix degradation. We propose that overexpression of specific endogenous protease inhibitors may provide a novel therapeutic avenue for the protection of neurons against cell death in RP/age-related macular degeneration and other neurodegenerative disorders, such as ALS.

Materials and Methods

Mice and anesthesia. All animal experiments were performed in compliance with the ARVO statement for the Use of Animals in Ophthalmic and Vision Research and were conducted under a valid UK Home Office licence. Rhodopsin knockout mice (C57B/6.129 Rhotm1Phm, referred to herein as Rho-/-) have been described previously17and were obtained under a material transfer agreement as a kind gift from Jane Farrar, Trinity College Dublin, Ireland. OPN1LW-EGFP mice (C57BL/6JTgOPN1LW-EGFP85933Hue, referred to herein as B6.Tg(OPN1LW-EGFP) have been described previously16 and were obtained under an material transfer agreement from the Mutant Mouse Regional Resource Centres, National Institute of Health, USA, with the kind help of Dr Rachel Pearson, University College London Institute of Ophthalmology, London, UK. Mice that express EGFP in cone photoreceptors and have a primary rod specific degeneration (Rho-/-, Tg(OPN1LW-EGFP)+/0) were created through crossing of B6.Tg(OPN1LW-EGFP)+/+ mice (homozygous for the OPN1LW-EGFP transgene insertion) with Rho-/- (homozygous for a targeted rhodopsin knockout), followed by backcrossing of F1 progeny (Rho+/-, Tg(OPN1LW-EGFP)+/0) to the parental Rho-/- line. Mice were phenotyped at weaning by AF imaging, and genotyped by PCR. All mice were housed under standard 12 hour: 12 hour light/dark cycle with food and water available ad libitum. General anesthesia was induced by a single intraperitoneal injection of Dormitor (medetomidine hydrochloride, 1 mg/kg body weight) and ketamine (60 mg/kg body weight) and the pupils fully dilated with 1%tropicamide and 2.5% phenylephrine hydrochloride eye drops (both Bausch & Lomb, Kingston-Upon-Thames, UK). Where appropriate, anesthesia was reversed following procedures by intraperitoneal injection of antisedan (atipamezole, 5 mg/kg body weight).

Vector construction. An adeno-associate virus serotype 2 vector backbone plasmid (pUF11), was kindly provided by William W. Hauswirth, University of Florida, USA. cDNA clones of human neuronal growth factor (NGF; NM_002506) and human ciliary neurotrophic factor (CNTF, NM_000614) were purchased from Origene (Rockville, MD). PCR was used to amplify the NGF secretion signal (sense primer, GTACGCGGCCGCATGTCCATGTTGTTC; antisense primer, GCTCTGTGAAAGCTGAGTGTGGTTCC) and the CNTF coding region minus the translation start codon and surrounding sequence with identity to the kozak consensus sequence (sense primer, GGAACCACACTCAGCTTTCACAGAGC, antisense primer, TCTAGTCGACCTACATTTTCTTG) from the respective constructs. The two fragments were subsequently combined using a “swift PCR for ligating in vitro construted exons” (SPLICE) reaction57 and ligated into the pUF11 backbone using synthetic NotI and AccI restriction sites (underlined). Fidelity of the final AAV2.CBA.hCNTF construct was confirmed by full-length sequencing in both orientations.

Virus production and titration. Recombinant AAV serotype 2 (rAAV2/2) virus packaging the CNTF construct (termed rAAV2/2.hCNTF throughout) was produced by transient cotransfection of HEK293 cells seeded in cell factories (HYPERflask; Corning, Tewksbury, MA) followed by purification using iodixanol gradient centrifugation, as previously reported.58,59 Purified rAAV2/2.hCNTF virus was concentrated by buffer exchange (Amicon Ultra-15, Millipore, Billerica, MA) that removed residual iodixanol before being resuspended in sterile phosphate-buffered saline (PBS) to a total volume of 100 μl and aliquoted into preblocked tubes (0.01% BSA). The virus was titrated by quantitative PCR (qPCR) using primers designed to amplify a 120 bp fragment within the poly-A region, using vector plasmid and virus of known titer as standards. The titer of rAAV2/2.hCNTF virus was calculated to be 2 × 1013 genome particles (gp) per ml.

Intraocular injection. The pupils of 4-week-old Rho-/-, Tg(OPN1LW-EGFP)+/0 mice were dilated as above and a gel lubricant (Viscotears, Novartis, Frimley, UK) was applied prior to the positioning of a 6 mm circular cover slip over the cornea to allow good visualization of the retina when viewed under a surgical microscope (M620 F20, Leica, Wetzlar, Germany). Injections were performed by advancing a Hamilton syringe with a 10 mm 34-gauge needle (65N, Hamilton AG) trans-sclerally through the neural retina into the vitreous. Two microliters of vector suspension or buffer (PBS) were delivered into the vitreous cavity close to the posterior pole; reflux was minimized by allowing the intraocular pressure to normalize prior removing the needle. Orientation and stabilization of the eye was maintained throughout by holding the superior rectus muscle with notched forceps.

Enzyme-linked immunosorbent assays (ELISA). HEK293 cells were seeded in six-well plates (1 × 106 cells per well) and transduced with 2 μl rAAV2/2.hCNTF (1 × 1013 gp/ml) virus. Three days post-transduction, media was harvested and assayed using a commercially available human CNTF-specific ELISA (R&D systems, Abingdon, UK) according to the manufacturer's instructions. In vivo assessment of hCNTF expression was carried out by intravitreal injection of PW4 Rho-/-, Tg(OPN1LW-EGFP)+/0 mice with 2 × 108 gp (low dose), 2 × 109 gp (medium dose), or 2 × 1010 gp (high dose) rAAV2/2.hCNTF vector. Four weeks were allowed for genome integration and vector expression at which time (PW8) eyes were harvested and flash frozen with liquid nitrogen. Globes were thawed in 100 μl PBS and homogenized prior to hCNTF protein levels being assessed using a commercially available human CNTF-specific ELISA (R&D systems, Abingdon, UK) as stated above.

Autofluorescence (AF) imaging and cone quantification. The ocular fundi of each mouse were imaged using a confocal scanning laser ophthalmoscope (cSLO; Spectralis HRA, Heidelberg Engineering, Heidelberg, Germany) at weeks 8, 10, 12, 15, 18, 21, 24, and 30, as previously described.22 Briefly, following dilation, a contact lens was placed on the cornea using a viscous coupling gel (0.3% w/v hypromellose, Matindale Pharmaceuticals, Romford, UK) to prevent cataract formation and to improve and standardize image quality. The mouse was positioned on an imaging platform and the NIR mode (820 nm laser) used to achieve camera alignment at the confocal plane of the neural retina. Intrinsically, EGFP expressing cone photoreceptors were imaged using the autofluorescence (AF) mode (480 nm excitation) using a 55° lens and high-resolution images (1,536 × 1,536 pixels) were recorded at a standardized detector sensitivity with automated real-time averaging. Cone numbers were quantified manually from each image using ImageJ software,60 using vascular landmarks in the inferior retina to standardize the region of interest for each eye across the time course.

Electroretinography (ERG). ERG recording was carried out using an Espion E2 system (Diagnosys LLC, Cambridge, UK) according to a standard protocol at PW8, PW10, and PW12, as described previously.61 Following a period of light adaptation (>10 minutes, polychromatic white light, 30 cd/m2) brief (4 ms) white flashes were delivered at 3, 10, and 25 cd.s/m2 (20 flashes per intensity) and the b-wave amplitudes quantified with the Espion software (Diagnosys LLC, Cambridge, UK).

OMR. A custom built optomotor system was produced consisting of a rotating cylinder which allowed speed and grating size to be precisely regulated. Mice were placed on a raised platform in the centre of the drum which was illuminated from above by a bright white stimuls (~1,000 lux) and acclimatized to the drum for 1 minute during which the drum remained stationary and the mice were free to explore the platform. Each experimental run consisted of a 1-minute clockwise rotation and a 1-minute anticlockwise rotation divided into alternating 30-second periods. The experimental run was repeated independently three times at each time point for each animal, and the number of responses was averaged. The rotation of the square-wave grating elicited head-tracking responses, with a single response consisting of a slow head-tracking motion in the direction of the drum's rotation followed by a rapid repositioning of the head to a central position. The behavior of each mouse in response to the rotating drum was recorded using a digital camera mounted directly above the central platform; the number of head-tracking responses was quantified manually with the scorer blinded with regards to treatment.

Laser speckle cortical imaging. Cortices were imaged of 30-week-old Rho-/-, Tg(OPN1LW-EGFP)+/0 mice that had received rAAV2/2.hCNTF treatment in one eye and PBS sham treatment in the contralateral eye (n = 4, high dose; n = 6, medium dose; n = 8 low dose). The positive control group consisted of mice (n = 3) of identical background at an age (6 weeks old) when cone degeneration had not yet commenced. Mice were dark adapted for 8 hours prior to imaging and all preparatory steps conducted in a dark room under dim red illumination. Mice were anesthetized and the pupils fully dilated as described above. The head was secured in a stereotactic frame, and the scalp was resected to reveal the cranium over the visual cortices. Transcranial imaging was performed using the Speckle Contrast Imager (moorFLPI-2, Moor Instruments, Delaware) and data were acquired at 25Hz. Core body temperature was monitored by rectal thermometry and maintained throughout by means of a feedback heat pad wrapped around the body of the experimental animal. Changes in CBF were measured following the stimulation of each eye independently, where the contralateral eye was covered throughout to maintain dark adaptation. The eye to be examined first was randomized for each mouse and the investigators blinded as to which eye had received rAAV.hCNTF treatment. Each eye was stimulated by brief (10 ms) flashes of 14.5 cd 510 ± 3 nm monochromatic light at 5 Hz for 1 second (Grass PS33 photic stimulator with LED stimulus). Stimulation was repeated 10 times per eye with an interstimulus interval of 30-second (5 minute total time). The light stimulus was immediately transferred to the contralateral eye for examination, during which time the mouse and imaging equipment remained static. Data were processed using MATLAB R2012b (version 8.0.0.783). After down-sampling the data to 5 Hz, regions of interest were selected for the bilateral visual cortices from which the time-series changes where extracted and averaged over the pixels within each region of interest. Individual time-series were then smoothed using a Chebyshev type 1 filter. The percentage change in CBF relative to a 10-second prestimulus baseline period was calculated for left and right hemispheres. The responses of both hemispheres were summed (contralateral + ipsilateral) for each eye examination, and the magnitude change in CBF compared between rAAV.hCNTF-treated and PBS sham-treated eyes.

RNA sequencing and analysis. Following euthanasia, murine eyes were enucleated and snap frozen in RNAlater (Ambion, Paisley, UK) on liquid nitrogen and stored at −80 °C until processing. Whole eyes were homogenized and total RNA was extracted (RNA tissue mini kit, Qiagen, Manchester, UK). Contaminating DNA was removed by DNase I treatment following an ‘on column' digestion protocol (RNase-Free DNase, Qiagen). Poly-adenylated mRNA was purified from 2 μg total RNA per sample by magnetic isolation (NEBNextPoly(A) mRNA magnetic isolation module, New England Biolabs, Hitchin, UK). A library preparation protocol was performed using NEBNext mRNA Library Prep Master Mix Set for Illumina Kit (New England Biolabs) with the following modifications: Following fragmentation, RNA underwent additional clean up (Ampure XP, 2.8× volume ratio, Ambion, Paisley, UK) and washing (2 × 80% ehtanol) steps prior to elution in EB buffer. Reverse transcription (RT) was carried out with Superscript II (Invitrogen, Paisley, UK) with post-RT clean up (Ampure XP, 1.2× volume ration) and wash (2 × 80% EtOH) steps. cDNA was end-repaired, A-tailed, and adapter ligated (custom adapter) prior to PCR amplification (12 cycles). Prepared libraries were multiplexed (4×) and assessed for quality prior to paired-end sequencing on the IlluminaHiSeq 2000 platform with read length of 50 bp. Output FASTQ files were assessed for quality (SAMStat and FastQC:Read QC), groomed (FASTQ groomer62) prior to adapter removal (Cutadapt) and aligned to the reference genome (Mus musculus, UCSC, mm9.fa) in a paired-end manner (TopHat for Illumina63). Output of accepted hits was converted to SAM format (SAM Tools) and hits assigned to genomic features by comparison to the reference genome with HiSeq-Count. Differential expression was called using the DESeq R package (Bioconductor) with normalization for effective library size and false discovery rate set at 2%. Differentially expressed genes were considered as those with minimum twofold (1 2log-fold) change in expression level. Heatmap outputs were created with RColorBrewer and gplots packages for R.

Statistics. Anderson–Darling tests for normality on full data sets for each experiment revealed that the data were sampled from a normally distributed population in the following cases: In vivo cone quantification, A2 = 0.509, P = 0.1983; low-dose ERG, A2 = 0.3402, P = 0.4824; medium-dose ERG, A2 = 0.4483, P = 0.2785; high-dose ERG, A2 = 0.7228, P = 0.0594. Plotting of residuals versus fitted means for each experimental group demonstrated that the variance was equal within each group (plots not shown). Repetitive measures two-way ANOVA was used to analysis in vivo cone quantification with dose (independent variable) and time (repeated measure) as factors, and cone number as the dependent variable. Repetitive measures two-way ANOVA were used to analyze ERG responses with flash intensity (independent variable) and time (repetitive measure) as factors, with photopic b-wave amplitude the dependent variable in each case, where each dose was analyzed independently. Normality was not confirmed for optomotor response data: A2 = 4.5536, P ≤ 0.0005. Nonparametric Kruskal-Wallis test with Dunn's multiple comparisons test was performed to compare mean head tracking responses elicited from treated and untreated eyes at each dose. Due to low sample size, variance could not be plotted for groups when assessing CNTF levels following in vitro or in vivo vector transduction. Nonparametric Mann-Whitney U-tests were applied in both instances to compare mean CNTF levels in transduced versus sham transduced cell supernatants (in vitro) or harvested eyes (in vivo). Bonferroni post-tests were applied to all ANOVAs. The significance (P) level for all tests was set at 0.05. Controls. ns, not significant. ***P < 0.001, one-way ANOVA. Low dose n = 5, medium dose n = 6, high dose n = 4.

SUPPLEMENTARY MATERIAL Figure S1. Quantification of cone photoreceptor survival using vascular landmarks to define a constant region of interest. Figure S2. Long-term comparison of rAAV2/2.hCNTF vector dosing on cone photoreceptor survival in Rho-/-, Tg(OPN1LW-EGFP)+/0 mice. Figure S3. Electrophysiological suppression of photoreceptor function is dependent on hCNTF dose. Figure S4. Correlation between cone photoreceptor survival and visually guided behavioral responses. Table S1. Selected genes grouped by gene ontology.

Acknowledgments

This research was funded by Fight for Sight, the Wellcome Trust, the Health Foundation, the Medical Research Council, the Royal College of Surgeons of Edinburgh, the Oxford Stem Cell Institute and the NIHR Ophthalmology and Oxford Biomedical Research Centre (all bodies UK based). W.I.L.D. was funded by an Australian Research Council (ARC) Future Fellowship (FT110100176) and an ARC Discovery Project grant (DP140102117). C.M. was supported by a Royal Society University Research Fellowship. D.M.L. was sponsored in part by a Fulbright-Fight for Sight Research Scholarship. We thank the High-Throughput Genomics Group at the Wellcome Trust Centre for Human Genetics (funded by Wellcome Trust grant reference 090532/Z/09/Z and MRC Hub grant G0900747 91070; University of Oxford) for the generation of the sequencing data. The authors would also like to thank Oleksandr Moskalenko (High Performance Computing Centre, University of Florida, USA) for assistance in the analysis of RNA-Seq data. D.M.L. would like to thank Bally (currently, University of Florida, USA) for helpful contributions throughout the preparation of this manuscript. The authors declare that they have no relevant conflicts of interest.

Supplementary Material

Supplementary Figure S1

Quantification of cone photoreceptor survival using vascular landmarks to define a constant region of interest.

Click here for additional data file. (11.4MB, tiff)
Supplementary Figure S2

Long-term comparison of rAAV2/2.hCNTF vector dosing on cone photoreceptor survival in Rho-/-, Tg(OPN1LW-EGFP)+/0 mice.

Click here for additional data file. (11.4MB, tiff)
Supplementary Figure S3

Electrophysiological suppression of photoreceptor function is dependent on hCNTF dose.

Click here for additional data file. (11.4MB, tiff)
Supplementary Figure S4

Correlation between cone photoreceptor survival and visually guided behavioral responses.

Click here for additional data file. (11.4MB, tiff)
Supplementary Table S1

Selected genes grouped by gene ontology.

Click here for additional data file. (60KB, xls)

References

  1. Okoye, G, Zimmer, J, Sung, J, Gehlbach, P, Deering, T, Nambu, H et al. (2003). Increased expression of brain-derived neurotrophic factor preserves retinal function and slows cell death from rhodopsin mutation or oxidative damage. J Neurosci 23: 4164–4172. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. McGee Sanftner, LH, Abel, H, Hauswirth, WW and Flannery, JG (2001). Glial cell line derived neurotrophic factor delays photoreceptor degeneration in a transgenic rat model of retinitis pigmentosa. Mol Ther 4: 622–629. [DOI] [PubMed] [Google Scholar]
  3. Buch, PK, MacLaren, RE, Durán, Y, Balaggan, KS, MacNeil, A, Schlichtenbrede, FC et al. (2006). In contrast to AAV-mediated Cntf expression, AAV-mediated Gdnf expression enhances gene replacement therapy in rodent models of retinal degeneration. Mol Ther 14: 700–709. [DOI] [PubMed] [Google Scholar]
  4. Ji, JZ, Elyaman, W, Yip, HK, Lee, VW, Yick, LW, Hugon, J et al. (2004). CNTF promotes survival of retinal ganglion cells after induction of ocular hypertension in rats: the possible involvement of STAT3 pathway. Eur J Neurosci 19: 265–272. [DOI] [PubMed] [Google Scholar]
  5. Leaver, SG, Cui, Q, Bernard, O and Harvey, AR (2006). Cooperative effects of bcl-2 and AAV-mediated expression of CNTF on retinal ganglion cell survival and axonal regeneration in adult transgenic mice. Eur J Neurosci 24: 3323–3332. [DOI] [PubMed] [Google Scholar]
  6. MacLaren, RE, Buch, PK, Smith, AJ, Balaggan, KS, MacNeil, A, Taylor, JS et al. (2006). CNTF gene transfer protects ganglion cells in rat retinae undergoing focal injury and branch vessel occlusion. Exp Eye Res 83: 1118–1127. [DOI] [PubMed] [Google Scholar]
  7. Maier, K, Rau, CR, Storch, MK, Sättler, MB, Demmer, I, Weissert, R et al. (2004). Ciliary neurotrophic factor protects retinal ganglion cells from secondary cell death during acute autoimmune optic neuritis in rats. Brain Pathol 14: 378–387. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Pease, ME, Zack, DJ, Berlinicke, C, Bloom, K, Cone, F, Wang, Y et al. (2009). Effect of CNTF on retinal ganglion cell survival in experimental glaucoma. Invest Ophthalmol Vis Sci 50: 2194–2200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Tao, W, Wen, R, Goddard, MB, Sherman, SD, O'Rourke, PJ, Stabila, PF et al. (2002). Encapsulated cell-based delivery of CNTF reduces photoreceptor degeneration in animal models of retinitis pigmentosa. Invest Ophthalmol Vis Sci 43: 3292–3298. [PubMed] [Google Scholar]
  10. LaVail, MM, Yasumura, D, Matthes, MT, Lau-Villacorta, C, Unoki, K, Sung, CH et al. (1998). Protection of mouse photoreceptors by survival factors in retinal degenerations. Invest Ophthalmol Vis Sci 39: 592–602. [PubMed] [Google Scholar]
  11. Chong, NH, Alexander, RA, Waters, L, Barnett, KC, Bird, AC and Luthert, PJ (1999). Repeated injections of a ciliary neurotrophic factor analogue leading to long-term photoreceptor survival in hereditary retinal degeneration. Invest Ophthalmol Vis Sci 40: 1298–1305. [PubMed] [Google Scholar]
  12. Bok, D, Yasumura, D, Matthes, MT, Ruiz, A, Duncan, JL, Chappelow, AV et al. (2002). Effects of adeno-associated virus-vectored ciliary neurotrophic factor on retinal structure and function in mice with a P216L rds/peripherin mutation. Exp Eye Res 74: 719–735. [DOI] [PubMed] [Google Scholar]
  13. Schlichtenbrede, FC, MacNeil, A, Bainbridge, JW, Tschernutter, M, Thrasher, AJ, Smith, AJ et al. (2003). Intraocular gene delivery of ciliary neurotrophic factor results in significant loss of retinal function in normal mice and in the Prph2Rd2/Rd2 model of retinal degeneration. Gene Ther 10: 523–527. [DOI] [PubMed] [Google Scholar]
  14. Liang, FQ, Dejneka, NS, Cohen, DR, Krasnoperova, NV, Lem, J, Maguire, AM et al. (2001). AAV-mediated delivery of ciliary neurotrophic factor prolongs photoreceptor survival in the rhodopsin knockout mouse. Mol Ther 3: 241–248. [DOI] [PubMed] [Google Scholar]
  15. Lipinski, DM, Singh, MS and MacLaren, RE (2011). Assessment of cone survival in response to CNTF, GDNF, and VEGF165b in a novel ex vivo model of end-stage retinitis pigmentosa. Invest Ophthalmol Vis Sci 52: 7340–7346. [DOI] [PubMed] [Google Scholar]
  16. Fei, Y and Hughes, TE (2001). Transgenic expression of the jellyfish green fluorescent protein in the cone photoreceptors of the mouse. Vis Neurosci 18: 615–623. [DOI] [PubMed] [Google Scholar]
  17. Humphries, MM, Rancourt, D, Farrar, GJ, Kenna, P, Hazel, M, Bush, RA et al. (1997). Retinopathy induced in mice by targeted disruption of the rhodopsin gene. Nat Genet 15: 216–219. [DOI] [PubMed] [Google Scholar]
  18. Hartong, DT, Berson, EL and Dryja, TP (2006). Retinitis pigmentosa. Lancet 368: 1795–1809. [DOI] [PubMed] [Google Scholar]
  19. Hobson, AH, Donovan, M, Humphries, MM, Tuohy, G, McNally, N, Carmody, R et al. (2000). Apoptotic photoreceptor death in the rhodopsin knockout mouse in the presence and absence of c-fos. Exp Eye Res 71: 247–254. [DOI] [PubMed] [Google Scholar]
  20. Lipinski, DM, Thake, M and MacLaren, RE (2013). Clinical applications of retinal gene therapy. Prog Retin Eye Res 32: 22–47. [DOI] [PubMed] [Google Scholar]
  21. Beck, SC, Schaeferhoff, K, Michalakis, S, Fischer, MD, Huber, G, Rieger, N et al. (2010). In vivo analysis of cone survival in mice. Invest Ophthalmol Vis Sci 51: 493–497. [DOI] [PubMed] [Google Scholar]
  22. Charbel, IP, Singh, MS, Lipinski, DM, Chong, NV, Delori, FC, Barnard, AR et al. (2012). Optimization of in vivo confocal autofluorescence imaging of the ocular fundus in mice and its application to models of human retinal degeneration. Invest Ophthalmol Vis Sci 53: 1066–1075. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Wen, R, Tao, W, Li, Y and Sieving, PA (2012). CNTF and retina. Prog Retin Eye Res 31: 136–151. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Hobbelen, JF and Collewijn, H (1971). Effect of cerebro-cortical and collicular ablations upon the optokinetic reactions in the rabbit. Doc Ophthalmol 30: 227–236. [DOI] [PubMed] [Google Scholar]
  25. Harvey, RJ, De'Sperati, C and Strata, P (1997). The early phase of horizontal optokinetic responses in the pigmented rat and the effects of lesions of the visual cortex. Vision Res 37: 1615–1625. [DOI] [PubMed] [Google Scholar]
  26. Douglas, RM, Alam, NM, Silver, BD, McGill, TJ, Tschetter, WW and Prusky, GT (2005). Independent visual threshold measurements in the two eyes of freely moving rats and mice using a virtual-reality optokinetic system. Vis Neurosci 22: 677–684. [DOI] [PubMed] [Google Scholar]
  27. Weber, BH, Vogt, G, Pruett, RC, Stöhr, H and Felbor, U (1994). Mutations in the tissue inhibitor of metalloproteinases-3 (TIMP3) in patients with Sorsby's fundus dystrophy. Nat Genet 8: 352–356. [DOI] [PubMed] [Google Scholar]
  28. Zhao, H, Friedman, RD and Fournier, RE (2007). The locus control region activates serpin gene expression through recruitment of liver-specific transcription factors and RNA polymerase II. Mol Cell Biol 27: 5286–5295. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Udofa, EA, Stringer, BW, Gade, P, Mahony, D, Buzza, MS, Kalvakolanu, DV et al. (2013). The transcription factor C/EBP-β mediates constitutive and LPS-inducible transcription of murine SerpinB2. PLoS One 8: e57855. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Ozawa, Y, Nakao, K, Kurihara, T, Shimazaki, T, Shimmura, S, Ishida, S et al. (2008). Roles of STAT3/SOCS3 pathway in regulating the visual function and ubiquitin-proteasome-dependent degradation of rhodopsin during retinal inflammation. J Biol Chem 283: 24561–24570. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Takase, H, Yu, CR, Liu, X, Fujimoto, C, Gery, I and Egwuagu, CE (2005). Induction of suppressors of cytokine signaling (SOCS) in the retina during experimental autoimmune uveitis (EAU): potential neuroprotective role of SOCS proteins. J Neuroimmunol 168: 118–127. [DOI] [PubMed] [Google Scholar]
  32. Birch, DG. Long-term follow-up of patients with retinitis pigmentosa (RP) receiving sustained-release CNTF through intraocular encapsulated cell technology implants. ARVO Annual Meeting, Seattle, WA, 2013.
  33. Birch, DG, Weleber, RG, Duncan, JL, Jaffe, GJ and Tao, W; Ciliary Neurotrophic Factor Retinitis Pigmentosa Study Groups (2013). Randomized trial of ciliary neurotrophic factor delivered by encapsulated cell intraocular implants for retinitis pigmentosa. Am J Ophthalmol 156: 283–292.e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. McGill, TJ, Prusky, GT, Douglas, RM, Yasumura, D, Matthes, MT, Nune, G et al. (2007). Intraocular CNTF reduces vision in normal rats in a dose-dependent manner. Invest Ophthalmol Vis Sci 48: 5756–5766. [DOI] [PubMed] [Google Scholar]
  35. Macé, E, Caplette, R, Marre, O, Sengupta, A, Chaffiol, A, Barbe, P et al. (2015). Targeting channelrhodopsin-2 to ON-bipolar cells with vitreally administered AAV Restores ON and OFF visual responses in blind mice. Mol Ther 23: 7–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Gaub, BM, Berry, MH, Holt, AE, Reiner, A, Kienzler, MA, Dolgova, N et al. (2014). Restoration of visual function by expression of a light-gated mammalian ion channel in retinal ganglion cells or ON-bipolar cells. Proc Natl Acad Sci USA 111: E5574–E5583. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Singh, MS, Charbel Issa, P, Butler, R, Martin, C, Lipinski, DM, Sekaran, S et al. (2013). Reversal of end-stage retinal degeneration and restoration of visual function by photoreceptor transplantation. Proc Natl Acad Sci USA 110: 1101–1106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Francis, PJ, Mansfield, B and Rose, S (2013). Proceedings of the First International Optogenetic Therapies for Vision Symposium. Transl Vis Sci Technol 2: 4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Wen, R, Song, Y, Kjellstrom, S, Tanikawa, A, Liu, Y, Li, Y et al. (2006). Regulation of rod phototransduction machinery by ciliary neurotrophic factor. J Neurosci 26: 13523–13530. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Li, Y, Tao, W, Luo, L, Huang, D, Kauper, K, Stabila, P et al. (2010). CNTF induces regeneration of cone outer segments in a rat model of retinal degeneration. PLoS One 5: e9495. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Beltran, WA, Wen, R, Acland, GM and Aguirre, GD (2007). Intravitreal injection of ciliary neurotrophic factor (CNTF) causes peripheral remodeling and does not prevent photoreceptor loss in canine RPGR mutant retina. Exp Eye Res 84: 753–771. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Rhee, KD, Nusinowitz, S, Chao, K, Yu, F, Bok, D and Yang, XJ (2013). CNTF-mediated protection of photoreceptors requires initial activation of the cytokine receptor gp130 in Müller glial cells. Proc Natl Acad Sci USA 110: E4520–E4529. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Cahill, H and Nathans, J (2008). The optokinetic reflex as a tool for quantitative analyses of nervous system function in mice: application to genetic and drug-induced variation. PLoS One 3: e2055. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. ALS CNTF Treatment Study Group (1996). A double-blind placebo-controlled clinical trial of subcutaneous recombinant human ciliary neurotrophic factor (rHCNTF) in amyotrophic lateral sclerosis. Neurology 46: 1244–1249. [DOI] [PubMed] [Google Scholar]
  45. Bongioanni, P, Reali, C and Sogos, V (2004). Ciliary neurotrophic factor (CNTF) for amyotrophic lateral sclerosis/motor neuron disease. Cochrane Database Syst Rev CD004302. [DOI] [PMC free article] [PubMed]
  46. Davis, S, Aldrich, TH, Ip, NY, Stahl, N, Scherer, S, Farruggella, T et al. (1993). Released form of CNTF receptor alpha component as a soluble mediator of CNTF responses. Science 259: 1736–1739. [DOI] [PubMed] [Google Scholar]
  47. Kassen, SC, Thummel, R, Campochiaro, LA, Harding, MJ, Bennett, NA and Hyde, DR (2009). CNTF induces photoreceptor neuroprotection and Müller glial cell proliferation through two different signaling pathways in the adult zebrafish retina. Exp Eye Res 88: 1051–1064. [DOI] [PubMed] [Google Scholar]
  48. Peterson, WM, Wang, Q, Tzekova, R and Wiegand, SJ (2000). Ciliary neurotrophic factor and stress stimuli activate the Jak-STAT pathway in retinal neurons and glia. J Neurosci 20: 4081–4090. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Ahmed, ST and Darnell, JE Jr (2009). Serpin B3/B4, activated by STAT3, promote survival of squamous carcinoma cells. Biochem Biophys Res Commun 378: 821–825. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Chong, NH, Keonin, J, Luthert, PJ, Frennesson, CI, Weingeist, DM, Wolf, RL et al. (2005). Decreased thickness and integrity of the macular elastic layer of Bruch's membrane correspond to the distribution of lesions associated with age-related macular degeneration. Am J Pathol 166: 241–251. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Sorsby, A and Mason, ME (1949). A fundus dystrophy with unusual features. Br J Ophthalmol 33: 67–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. van de Ven, JP, Nilsson, SC, Tan, PL, Buitendijk, GH, Ristau, T, Mohlin, FC et al. (2013). A functional variant in the CFI gene confers a high risk of age-related macular degeneration. Nat Genet 45: 813–817. [DOI] [PubMed] [Google Scholar]
  53. Rohrer, B, Guo, Y, Kunchithapautham, K and Gilkeson, GS (2007). Eliminating complement factor D reduces photoreceptor susceptibility to light-induced damage. Invest Ophthalmol Vis Sci 48: 5282–5289. [DOI] [PubMed] [Google Scholar]
  54. Zhang, B, Hu, Y and Ma, JX (2009). Anti-inflammatory and antioxidant effects of SERPINA3K in the retina. Invest Ophthalmol Vis Sci 50: 3943–3952. [DOI] [PubMed] [Google Scholar]
  55. Zhang, B, Abreu, JG, Zhou, K, Chen, Y, Hu, Y, Zhou, T et al. (2010). Blocking the Wnt pathway, a unifying mechanism for an angiogenic inhibitor in the serine proteinase inhibitor family. Proc Natl Acad Sci USA 107: 6900–6905. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Zhang, B and Ma, JX (2008). SERPINA3K prevents oxidative stress induced necrotic cell death by inhibiting calcium overload. PLoS One 3: e4077. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Davies, WL, Carvalho, LS and Hunt, DM (2007). SPLICE: a technique for generating in vitro spliced coding sequences from genomic DNA. Biotechniques 43: 785–789. [DOI] [PubMed] [Google Scholar]
  58. Zolotukhin, S (2005). Production of recombinant adeno-associated virus vectors. Hum Gene Ther 16: 551–557. [DOI] [PubMed] [Google Scholar]
  59. Jacobson, SG, Acland, GM, Aguirre, GD, Aleman, TS, Schwartz, SB, Cideciyan, AV et al. (2006). Safety of recombinant adeno-associated virus type 2-RPE65 vector delivered by ocular subretinal injection. Mol Ther 13: 1074–1084. [DOI] [PubMed] [Google Scholar]
  60. Abramoff, MD, Magelhaes, PJ and Ram, SJ (2004). Image processing with ImageJ. Biophotonics International 11: 36–42. [Google Scholar]
  61. Lipinski, DM, Yusuf, M, Barnard, AR, Damant, C, Charbel Issa, P, Singh, MS et al. (2011). Characterization of a dominant cone degeneration in a green fluorescent protein-reporter mouse with disruption of Loci associated with human dominant retinal dystrophy. Invest Ophthalmol Vis Sci 52: 6617–6623. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Blankenberg, D, Gordon, A, Von Kuster, G, Coraor, N, Taylor, J and Nekrutenko, A; Galaxy Team (2010). Manipulation of FASTQ data with Galaxy. Bioinformatics 26: 1783–1785. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Trapnell, C, Pachter, L and Salzberg, SL (2009). TopHat: discovering splice junctions with RNA-Seq. Bioinformatics 25: 1105–1111. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Figure S1

Quantification of cone photoreceptor survival using vascular landmarks to define a constant region of interest.

Click here for additional data file. (11.4MB, tiff)
Supplementary Figure S2

Long-term comparison of rAAV2/2.hCNTF vector dosing on cone photoreceptor survival in Rho-/-, Tg(OPN1LW-EGFP)+/0 mice.

Click here for additional data file. (11.4MB, tiff)
Supplementary Figure S3

Electrophysiological suppression of photoreceptor function is dependent on hCNTF dose.

Click here for additional data file. (11.4MB, tiff)
Supplementary Figure S4

Correlation between cone photoreceptor survival and visually guided behavioral responses.

Click here for additional data file. (11.4MB, tiff)
Supplementary Table S1

Selected genes grouped by gene ontology.

Click here for additional data file. (60KB, xls)

Articles from Molecular Therapy are provided here courtesy of The American Society of Gene & Cell Therapy

RESOURCES