Activating the Wnt/β-Catenin Pathway Did Not Protect Immature Retina from Hypoxic-Ischemic Injury

Abstract

Purpose: Visual loss associated with hypoxic-ischemic (HI) brain damage is the most common cause of visual impairment in children of developed countries. A neuroprotective role for Wnt/β-catenin signaling has been demonstrated in several neurodegenerative disorders. The association of Wnt signaling with HI injury in immature retina has not been established.

Methods: On postnatal day 7 (P7), HI was induced by unilateral common carotid artery ligation followed by 8% oxygen hypoxia for 2 hours. The pups received intravitreous injection (i.v.i) of PBS, Dickkopf-1 (DKK-1, the negative modulator of Wnt/β-catenin pathway) antisense (AS) or sense (S) oligonucleotides at various concentrations for pretreatment (24 and 1 hour before HI) or post treatment (1 and 4 hours after HI). For chronic treatments, animals received repeated intraperitoneal (i.p.) injection of DKK-1-AS, DKK-1-S, lithium chloride (LiCl), or vehicles after HI. The retinal injury was assessed by electroretinography (ERG, P21, and P30) and immunohistochemical staining (P8 or P30).

Results: Pretreatment with DKK-1-AS (i.v.i.) attenuated DKK-1 and enhanced β-catenin expression, but did not protect immature retina against HI injury at both pathological and functional levels. Post treatment with DKK-1-AS (i.v.i. or i.p.) also did not rescue HI retinopathy. Chronic systemic LiCl treatment did not decrease Müller cell activation or neuronal damage in HI retinal injury.

Conclusions: Our data demonstrated that DKK-1 inhibition or chronic lithium treatment did not protect the immature retina from HI injury. It is speculated that the enhanced canonical Wnt/β-catenin signaling is not sufficient to protect the immature retina from HI injury.

With advances in perinatal care, survival rates for infants with hypoxic-ischemia (HI) injury are increasing. The prevalence of visual impairment in children with HI injury ranges from 66% to 94%.^1,2 Visual loss associated with brain damage is currently the most common cause of visual impairment in children of developed countries, placing a major burden on ophthalmologic and educational services in these countries.³ 

Most studies have emphasized ‘cerebral visual impairment' in neurologically damaged children.⁴ These findings were presumably attributed to transsynaptic degeneration of optic axons as a result of bilateral occipital damage. However, our studies in rat pups have demonstrated that HI caused damage in immature retina per se at both pathological and functional level. The injuries were rapid and extensive in retinal ganglion cells (RGC), inner plexiform layer (IPL), and inner nuclear layer (INL). The pathological evidence of retinal injury corresponded to the marked suppression of the amplitude of b-waves in electroretinography (ERG) even on P60. Our study provides the evidence that retinal injury contributes significantly to the visual impairments with HI brain injury in neonates. Furthermore, the immature retinas appear to be more susceptible to HI injury, in both temporal and spatial aspects, than the mature rats.⁵ 

Wnt signaling pathways regulate cellular proliferation, survival, and differentiation in the eye.⁶ The molecular signaling cascades of the canonical Wnt pathway have been extensively characterized. In the absence of Wnt ligand, the central mediator β-catenin is phosphorylated by the glycogen synthase kinase-3β (GSK-3β) and is rapidly degraded. Binding of Wnt ligand to the coreceptors Frizzled and low-density lipoprotein receptor-related protein (LRP) inactivates GSK-3β phosphorylation, allowing β-catenin to translocate into the nucleus and initiates transcription of Wnt target genes. The Wnt pathway is negatively modulated by the secreted protein Dickkopf-1 (DKK-1), which binds to LRPs, thus preventing their interaction with Wnts.⁷ Lithium, a well-known mood-stabilizer and neuroprotective agent, mainly acts through inhibition of GSK-3β.^8,9 

A neuroprotective role for Wnt signaling has been demonstrated in animal and cellular models of neurodegenerative disorders, retinal degeneration, and cerebral ischemia.^8–10 Wnt/β-catenin signaling promotes proliferation of Müller cell–derived retinal progenitors and neural regeneration after damage in adult mice.¹¹ Activation of Wnt signaling by Norrin significantly decreased apoptotic death of retinal neurons following excitotoxicity in adult rodents.¹² Although Wnt signaling is known to mediate multiple biological and pathological processes, its association with HI injury in immature retina has not been established. In the present study, we investigated if activation of Wnt/β-catenin signaling by blocking DKK-1 or treatment with lithium can protect the HI retinopathy in rat pups. 

Materials and Methods

Animals

This study was approved by the Animal Care Committee at Kaohsiung Chang Gung Memorial Hospital and conducted in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Ten to 12 Sprague-Dawley pups per dam were used and housed with a 12/12-hour light/dark schedule in a temperature- and humidity-controlled colony room. The pups were housed with their dams until weaning on postnatal (P) day 21, and then housed in groups of 4 to 5 per cage. 

Hypoxic-Ischemia Injury Model

Postnatal day 7 rat pups of both sex were anesthetized with fluothane, and the right common carotid artery was surgically exposed and permanently ligated with 5-0 surgical silk. After surgery, the pups were returned to the dam for a 1-hour recuperation period before HI. The animals were then placed in airtight 500 mL containers partially submerged in a 37°C water bath through which humidified 3 l/min 8% oxygen (balance, nitrogen) was maintained for 2 hours.⁵ After completion of HI, the rat pups were returned to their cages. The sham control rats had anesthesia and surgical exposure without vessel occlusion and hypoxia. 

Intravitreous Delivery of Anti–DKK-1 Oligonucleotides

The pups were intravitreous infused with PBS, DKK-1 antisense oligonucleotides (DKK-1-AS: 5-TAC-AGA-TCT-TGG-ACC-AGA-3) or sense oligonucleotide (DKK-1-S: 5-TCT-GGT-CCA-AGA-TCT-GAT-3, Bio Basic, Inc., Amherst, NY, USA), end-capped phosphorothioate carrying a fluorescence-emitting port, 1 ul in both eyes¹³ The DKK-1 oligonucleotides were infused at various concentrations for pretreatment (24 and 1 hour before HI) or post treatment (1 and 4 hours after HI). The retinas were collected and the damage was determined on P8 or P30. 

Intraperitoneal Administration of Drug

Separate groups of animals were treated by repeated intraperitoneal (i.p.) injection of DKK-1-AS, DKK-1-S oligonucleotides (50 μg/Kg), or normal saline (NS) immediately after HI, and at 6 and 24 hours post-HI. For chronic lithium treatment, the animals received daily i.p. injection of lithium chloride (LiCl, 30 mg/kg/day) or normal saline (NS) from 1 day prior to HI to P30. Animals were killed at P8 or P30, and the eyes were processed for histology and immunohistochemistry assessment. 

Assessment of Retinal Injury

Following deep anesthesia with i.p. overdose chloral hydrate, rats were perfused transcardially with NS followed by 4% paraformaldehyde. The eyes were fixed in the same fixative for 3 to 7 days. Eye blocks were embedded in paraffin, cut at 4 (for P8 rats) or 8 μm (for P30 rats), and stained with hematoxylin and eosin (HE). Images of HE-stained sections at ×200 magnification per visual field (0.145 mm²) were scanned using a microscope (Nikon, Tokyo, Japan) and analyzed by a computerized software (ImagePro Plus 6.0; Media Cybernetics, Bethesda, MD, USA).⁵ Retinal ganglion cells were counted along a constant length of 100 μm in 2 nonoverlapping views within central areas (100 μm from the optic disc) from each section by two independent observers.^5,14 The inner retinal thickness also was measured as the combined thickness of IPL and INL. 

Functional Evaluation by Electroretinography (ERG)

Full-field scotopic flash ERGs (RETIport ERG; Roland Consult, Brandenburg, Germany) were recorded from both eyes separately on P21 and P30. The pupils were dilated by 1% Mydriacyl (Alcon, Puurs, Belgium) and 1% Cyclogyl (Alcon), and the eyes were dark-adapted for 1 hour. After the rats were anesthetized by intramuscular injection of diluted 10% atropine and 6% Choloral hydrate i.p., the recording Ag:AgCl electrode was placed on the cornea with 0.5% methyl cellulose as a conductive medium. A reference electrode was inserted to the subcutaneous area of the forehead, and a ground electrode was inserted to subcutaneous tissue on back. The luminance of the stimulus was 3 cds/m², with a duration of 10 ms Scotopic 0-dB ERGs were recorded with a standard white flash and a dark background. Twenty responses elicited by identical flashes applied at 10-second intervals were averaged in the dark-adapted state.¹⁵ The amplitudes and the implicit times of the a- and b-waves were measured and averaged separately. 

Immunohistochemical Staining

Paraffin sections were dewaxed, hydrated through graded alcohols, and placed in PBS. After being blocked by 5% goat serum, the sections were incubated with various primary antibodies followed by a 60-minute incubation at room temperature with secondary Ab (Santa Cruz Biotechnology, Santa Cruz, CA, USA). The primary antibodies used were synaptophysin (1:1,000; Santa Cruz Biotechnology), β-Catenin(1:50; Cell signaling, Danvers, MA, USA), glial fibrillary acidic protein (GFAP; 1:200; Chemicon, Temecula, CA, USA), DKK-1 (1:100; Santa Cruz Biotechnology), and ED1 (1: 500, Serotec, Raleigh, NC, USA). The sections were developed in 3′3-diaminobenzidine (Sigma-Aldrich Corp., St. Louis, MO, USA) with the Vectastatin ABC system (Vector Laboratories, Burlingame, CA, USA). In every experiment, a control with the primary antibody omitted was used as a negative control. 

TUNEL Staining

The TUNEL stain was performed using an ApopTag fluorescein direct in situ apoptosis detection kit (Chemicon, Temecula, CA, USA) counterstained with Hoechst 33258 (Invitrogen, Carlsbad, CA, USA). Fluorescent images are acquired using a fluorescence microscope system (Nikon). 

Western Blotting Analysis

For whole retinal analysis, tissue was homogenized in cold lysis buffers, 40-μg samples were resolved in 10% SDS-PAGE, and blotted electrophoretically to nitrocellulose membranes.¹⁶ Membranes were blocked with 5% nonfat dry milk and subsequently incubated with primary antibodies, and immunoreactivity was detected by horseradish-conjugated secondary antibody, and visualized with enhanced chemiluminescence. The primary antibodies used were: anti–DKK-1 antibody (1:1000; Cell Signaling) and β-actin antibody (1:2000; Chemicon). For densitometry, data were normalized for the β-actin expression and the ratio of the study eyes to the sham controls was calculated. 

Statistics

Statistical analyses were performed using GraphPad Prism 4 software (GraphPad Software, San Diego, CA, USA). Statistical comparisons were performed using one-way ANOVA or two-tailed Student's t-tests. Values were considered significant at P less than 0.05, and data were presented as means ± SEM. 

Results

Intravitreous Injection of DKK-1-AS Oligonucleotides Attenuated DKK-1 and Enhanced β-Catenin Expression After HI injury in Immature Retina

We observed very few DKK-1 immunodensity in the sham controls. There were progressive increases of DKK-1 immunoreactivity in the cytoplasm of RGC and INL of the ipsilateral retina from 3 hours post-HI to P8, with the peak appeared on 6 hours post-HI (Fig. 1). 

Figure 1

 

Hypoxia-ischemia (HI) caused transient DKK-1 expression in the inner retina. There was a progressive increase of DKK-1 immunoreactivity (arrows) in the ipsilateral inner retina as compared with the sham controls (a) from 3 (b), 6 (c), 12 (d), 18 (e) to 24 hours (f) post-HI. Then a gradual decrease of DKK-1 immunodensity was observed on P9 (g) and P11 (h). The inset showed that the DKK-1 was mainly distributed in the cytoplasm of the RGC and INL. Scale bars: 100 μm; 50 μm (inset). n = 3 for each time point.

The rat pups were pretreated with intravitreous injection of 7.5 μg/mL DKK-1-AS, DKK-1-S, or PBS 24 and 1 hour before HI. There was obvious fluorescent staining for oligonucleotides scattering in the RGC and INL after intravitreous injection of DKK-1-AS or DKK-1-S (Figs. 2a, 2b). The DKK-1 immunostaining was markedly decreased in the DKK-1-AS–treated group as compared with the DKK-1-S–, PBS-treated, and HI-only groups 6 hours after HI (Figs. 2c–f). Western blotting also showed that the protein levels of DKK-1 were increased in the HI-only group 24 hours after HI injury. The DKK-1-AS treatment significantly decreased the DKK-1 protein expression as compared with the DKK-1-S– and HI-only groups (Fig. 2k). Corresponding to the attenuated DKK-1 expression after treatment, β-catenin expression was prominently increased in the DKK-1-AS–treated group as compared with the DKK-1-S–, PBS-treated and HI-only groups 24 hours after HI (Figs. 2g–j). 

Figure 2

 

Intravitreous injection (i.v.i.) of DKK-1-AS oligonucleotides attenuated DKK-1 and enhanced β-catenin expression after HI injury in immature retina. The rat pups were pretreated with i.v.i. of 7.5 μg/mL DKK-1-AS, DKK-1-S, or PBS 24 hours and 1 hour before HI. Obvious DKK-1-S (a) and DKK-1-AS (b) oligonucleotide fluorescence (arrowheads) was detected in the RGC and INL 6 hours after the last i.v.i. The DKK-1 immunostaining (arrows) was markedly decreased in the DKK-1-AS–treated (d) group as compared with the DKK-1-S- (c), PBS-treated (e), and HI only groups (f) at 6 hours post-HI. Western blotting also showed that the protein levels of DKK-1 were increased in the HI-only group 24 hrs after HI injury. The DKK-1-AS treatment significantly decreased the DKK-1 protein expression as compared with the DKK-1-S- and HI-only groups (k). β-catenin expression was also prominently increased in the inner retinas of DKK-1-AS–treated (h) group as compared with the DKK-1-S– (g), PBS-treated (i) and HI-only groups (j) 24 hours after HI. Scale bars: 50 μm in (a, b); 100 μm in (c–g). *P < 0.05.

Pretreatment With Intravitreous Injection of DKK-1-AS Oligonucleotides Did Not Protect Immature Retina From HI Injury at Both Pathological and Functional Levels

To assess the long-term effects of DKK-1 blockade on functional levels, ERG examination was performed on P21 and P30. The rat pups were pretreated with intravitreous injection of oligonucleotides or PBS 24 and 1 hour in both eyes before HI as described above. There were no significant differences in the a- and b-wave amplitude of contralateral retinas between the HI groups and the sham control (Fig. 3a). However, the b-wave, not a-wave, amplitude of ipsilateral retinas was markedly depressed in the HI-group as compared with the sham control on both P21 (P < 0.01) and P30 (P < 0.05, Fig. 3b). Therefore, further pathological and functional assessments were performed in ipsilateral eyes only. There were no significant differences in b-wave amplitude between the DKK-1-AS–, DKK-1-S–, PBS-treated and HI-only groups. Corresponding to the functional outcome, the pathology demonstrated markedly decreased inner retinal thickness in all the four HI groups on P30 (Figs. 3c–g). The number of TUNEL(+) cell in the inner retina did not differ between all the treated and HI-only groups 6 hours after HI (Figs. 3h–k). At 24 hours post-HI, there was prominent microglia activation in both the DKK-1 oligonucleotide- and PBS-treated groups (Figs. 3l–o). 

Figure 3

 

Pretreatment with i.v.i of DKK-1-AS oligonucleotides at various doses did not protect immature retina from HI injury at both pathological and functional levels. Groups data of ERG showed that there was no alteration of a- and b- wave amplitudes of contralateral retinas after HI injury (a). But the b-wave, not a-wave amplitude was significantly decreased in the ipsilateral eyes of all the HI groups as compared with the sham controls on both P21 (P < 0.01) and P30 (P < 0.05). The rat pups received i.v.i of 7.5 μg/mL DKK-1-AS oligonucleotides (7.5 AS group) had comparable b-wave amplitude with those received 7.5 μg/mL DKK-1-S oligonucleotides (7.5S group), PBS (PBS group), or nothing (HI-only group [a, b]). Representative retinal histologic sections from P30 showed prominently decreased RGC and inner retinal thickness in DKK-1-AS- (d), DKK-1-S- (c), PBS-treated (e), and HI only (f) groups, as compared with sham controls (g). At 6 hours post-HI, there were prominent TUNEL(+) cells in the inner retina of the DKK-1-AS- (i), DKK-1-S- (h), PBS-treated (j), and HI only groups (k). There were also markedly ED1 immunostaining (arrows) in the RGC INL of all the DKK-1-AS– (m), DKK-1-S– (l), PBS-treated (n), and HI only groups (o) at post-HI 24 hours. To investigate the dose-dependent effects, the rat pups received i.v.i of various doses of DKK-1-AS and DKK-1-S (7.5, 10, 100 μg/mL) 24 and 1 hour before HI. The quantitative data at 24 hours post-HI showed that there were no significant differences in the RGC counts (p) and inner retinal thickness (q) between the DKK-1-AS–, DKK-1-S–, and PBS-treated, and HI only groups. *P < 0.05, **P < 0.01, ***P < 0.001 as compared with sham controls. Scale bars: 100 μm (n = 7–16 for each group).

To investigate the dose-dependent effects, the rat pups received various doses of DKK-1-AS and DKK-1-S (7.5, 10, 100 μg/mL) intravitreously 24 and 1 hour before HI. There were no significant differences in RGC counts and inner retina thickness between the DKK-1-AS–, DKK-1-S–, PBS-treated and HI-only groups at 24 hours post-HI (Figs. 3p–q). 

Posttreatment With Intravitreous Injection of Very High Dose DKK-1 as Oligonucleotides Provide Marginal Protection Against HI Injury in Immature Retina

The rat pups received intravitreous injection of DKK-1-AS, DKK-1-S (1000 μg/mL), or PBS 1 and 4 hours after HI. On P8, the RGC counts were significantly decreased in the DKK-1-S–, PBS–treated and HI only groups, as compared with the sham controls. There was no significant difference between the DKK-1-AS posttreated group and the sham controls, indicating a nonsignificant trend of protection for the very high dose DKK-1-AS treatment (Fig. 4). 

Figure 4

 

Posttreatment with i.v.i of very high dose DKK-1-AS oligonucleotides provide marginal protection against HI injury in immature retina. The rat pups received i.v.i of DKK-1-AS, DKK-1-S (1000 μg/mL) or PBS 1 and 4 hours after HI. Representative retinal histologic sections on P8 showed prominent reduction of inner retinal thickness in DKK-1-S– (a), DKK-1- AS– (b), PBS-treated (c), and HI only (d) groups, as compared with sham controls (e). The RGC counts were significantly decreased in the DKK-1-S–, PBS-treated, and HI only groups, as compared with the sham controls. There was no significant difference between the DKK-1-AS–treated group and the sham controls, indicating a nonsignificant trend of protection for the very high dose DKK-1-AS treatment (f). *P < 0.05, **P < 0.01, ***P < 0.001 as compared with sham controls. Scale bar: 100 μm. (n = 7–16 for each group).

Systemic Treatment of DKK-1-AS Did Not Protect Immature Retinal From HI Injury

The rat pups received i.p. injection of DKK-1-AS, DKK-1-S (50 μg/Kg), or NS immediately after HI, and at 6 and 24 hours post-HI. There was significantly lower amplitude of b-waves in the DKK-1-AS–, DKK-1-S–, and NS-treated groups than the sham controls on P21 (P < 0.05) and P30 (P < 0.01). However, no difference was found between the three HI groups (Fig. 5a). The long-term retinal dysfunction after HI injury corresponded to the pathological changes. As compared with sham controls, the ipsilateral RGC counts were significant decreased in all the three HI groups. No significant difference was found between the DKK-1-AS–, DKK-1-S–, and NS-treated groups (Figs. 5b–f). 

Figure 5

 

Systemic treatment of DKK-1-AS oligonucleotides did not protect immature retinal from HI injury. The rat pups received i.p. injection of DKK-1-AS, DKK-1-S (50 μg/Kg), or NS immediately after HI, and at 6 and 24 hours post-HI. There was significantly lower amplitude of b-waves in the DKK-1-AS–, DKK-1-S–, and NS-treated groups than the sham controls on P21 (P < 0.05) and P30 (P < 0.01). However, no difference was found between the three treated-HI groups (a). The histology studies showed the prominently decreased inner retinal thickness in the DKK-1-S– (b), DKK-1-AS– (c), and NS-treated (d) groups as compared with sham controls (e). All the three treated HI groups had significantly decreased RGC counts than the sham controls. No difference was found between the DKK-1-AS–, DKK-1-S–, and NS-treated groups (f). *P < 0.05, **P < 0.01, ***P < 0.001 as compared with sham controls. Scale bar: 100 μm. (n = 5–12 for each group).

Chronic Systemic Lithium Treatment Did Not Attenuate Müller Cell Activation and Neuronal Damage in HI Retinal Injury

The rats received i.p. LiCl or NS daily from P6 to P30. The body weight gains were not different between LiCl and NS groups (data not shown). The ERG showed that both LiCl- and NS-treated groups had comparable a-wave amplitude as the sham controls. However, there was significantly decreased amplitude of b-waves in HI groups than the sham controls on P21 and P30 (both P < 0.001). No significant differences were observed between the LiCl- and NS-treated HI groups (Figs. 6a, 6b). 

Figure 6

 

Chronic systemic lithium treatment did not attenuate Müller cell activation and neuronal damage in HI retinal injury. Group data of ERG showed that no difference of a- wave (a) but significantly lower amplitude of b-wave (b) in the NS- and LiCl-treated HI groups, as compared with the sham controls on P21 and P30. Representative retinal sections showed prominent inner retina thinning in NS- (c), and LiCl -treated HI (d) groups as compared with sham control (e) on P30. Group data showed that the both NS- or LiCl-treated HI rats had significantly decreased RGC counts (f), IPL (g), and INL (h) thickness than the sham controls. There was dense synaptophysin immunostaining in the IPL and outer plexiform layer (OPL) of the sham retinas (k), but only scanty immunodensity in the IPL was found in NS- (i) and lithium-treated (j) retinas on P30. Glial fibrillary acidic protein immunoreactivity was restricted to the astrocytes of RGC layer in sham retinas (n), but was expressed throughout the whole retina markedly in the NS- (l) and LiCl-treated (m) retinas on P30. ***P < 0.001 as compared with sham group. Scale bars: 100 μm. (n = 12–21 for each group).

The pathological studies demonstrated significantly decreased RGC counts, IPL, and INL thickness in both the LiCl- and NS-treated HI groups, as compared with sham controls on P30 (Figs. 6c–h). The synaptophysin immunostaining showed obvious IPL thinning in both HI groups (Figs. 6i–k). Müller cells activation, as demonstrated by GFAP immunostaining, also was prominently increased throughout the whole retinal in the LiCl- and NS-treated HI groups as compared with the sham controls (Figs. 6l–n). 

Discussion

Our data demonstrated that the inhibition of DKK-1 or chronic lithium treatment did not protect the immature retina from HI injury at both pathological and functional levels. A growing body of evidence shows that inhibition of the Wnt pathway contributes to the pathophysiology of neuronal damage in various models of acute and chronic neurodegenerative disorders.^7,17,18 Induction of DKK-1 is required for the development of ischemic and excitotoxic neuronal death in the brain.^7,17 However, Wnt-mediated protection in retina is injury-specific (i.e., against elevated pressure-induced apoptosis but not for hypoxia or oxidative injury).⁶ Furthermore, DKK-1 treatment ameliorates retinal inflammation, vascular leakage, and ischemic areas in the models of ischemic retinal diseases, including diabetic retinopathy and retinopathy of prematurity.^19,20 DKK-1 also blocked the generation of reactive oxygen species induced by high glucose or TNF-alpha treatment in retinal capillary endothelial cells.¹⁹ These observation suggested that the Wnt pathways mediate multiple and complex biological and pathological processes in retinopathy.^21,22 

The neuroprotection by lithium treatment mainly rescues the canonical Wnt pathway.⁹ Intraperitonal or intravitreal treatment with LiCl markedly increased the retinal levels of β-catenin, but did not improve capillary repair in mice with oxygen-induced retinopathy, a model for retinopathy of prematurity.²³ In the present study, we also found that chronic systemic lithium treatment did not attenuate Müller cell activation and neuronal damage in HI retinal injury. It should be noted that lithium can also act on neuronal signal transmission and neurotransmitter release through disturbing the homeostasis of sodium, magnesium, and calcium.²⁴ In addition, lithium influences a broad array of transcription factors and intracellular signaling pathways, in particular that of phosphoinositide 3-kinase, protein phosphatase 2A, and monophosphatase.^8,24 Some of these pathways might have interfered with the neuroprotection of lithium in immature retina. 

Our data strongly indicated the involvements of additional mechanisms in the HI injury of immature retina (i.e., the β-catenin-independent Wnt pathway). Inside the cell, the Wnt signal can activate three pathways: one canonical (Wnt/β-catenin) and two noncanonical (Wnt/PCP and Wnt/Ca²⁺).²⁰ Frizzled-2 and Wnt5a, which belong to the noncanonical Wnt pathway, have been demonstrated to be responsible for Ca²⁺ overload in hypoxia/reoxygenation injury.²⁵ Wnt5a was highly expressed in ischemic muscles and endothelial cells, and exogenous Wnt5a attenuated revascularization in hind limb ischemia.²⁶ Interestingly, activation of all three Wnt-signaling pathways by norrin and inhibition of Wnt-canonical signaling by DKK1 result in similar retinal rescue in retinopathy of prematurity, indicating that both canonical and noncanonical Wnt-signaling pathways play significant roles in immature retina.²⁰ 

Another complexity is that the regulation of one Wnt intracellular pathway by others. It is generally accepted that the noncanonical pathways antagonize the Wnt/β-catenin pathway²⁷ and inhibiting canonical signaling can activate Wnt/PCP signaling.²⁸ These seemingly contradictory results can be explained by the complexity of Wnt signaling in retina development and retinopathy.^21,22 Both canonical and noncanonical Wnt signal transductions have been implicated in new blood vessel formation in retina, depending on the cellular context.²⁰ The DKK-1 knockout mice may delineate the specific role of Wnt/β-catenin pathway in HI retinopathy. Unfortunately, there are currently no viable DKK-1 knockout mice. 

Another possibility for the lack of therapeutic effects in our study may be related with the sequence-independent effects of antisense oligonucleotides. For improvement of stability and cell penetration, oligonucleotides structure was corrected by many chemical modifications. It may result in a variety of nonantisense activities of oligonucleotides.²⁹ However, there was markedly attenuated DKK-1 expression and enhanced β-catenin expression in the DKK-1-AS–treated group, as compared with the DKK-1-S–, and PBS-treated groups after HI injury. Furthermore, the systemic treatment by lithium acting via inhibiting GSK-3β, the downstream of the DKK-1 in the Wnt pathway, also did not rescue the HI retinopathy in the present study. These data suggested that activating Wnt/β-catenin signaling may not be sufficient for the protection of immature retina in HI injury. 

In summary, blocking DKK-1 or chronic lithium treatment failed to protect the HI retinopathy in immature rats. It is tempting to speculate that the increase in activity of canonical Wnt/β-catenin signaling is essential, but not sufficient to protect the immature retina from HI injury. Still, other and as of yet unidentified pathways might be directly or indirectly activated in Wnt-signaling during HI retinopathy. Furthermore, reduction of DKK-1 can cause severe eye developmental defect,³⁰ suggesting Wnt-secreted glycoproteins are morphogens in early life. Whether the similar pathways in HI retinopathy might be also involved during the development of retina appears to be a very interesting target for further studies. 

Acknowledgments

Supported by grants from Kaohsiung Chang Gung Memorial Hospital (Kaohsiung, Taiwan), Chang Gung University College of Medicine (CMRPG8A0721; Kaohsiung, Taiwan), and National Science Council (NSC 102-2314-B-182A-110-MY3; Taipei, Taiwan). 

Disclosure: H.-M. Huang, None; C.-C. Huang, None; F.-S. Wang, None; P.-L. Hung, None; Y.-C. Chang, None