- Email: [email protected]
Low power laser treatment of the retina ameliorates neovascularisation in a transgenic mouse model of retinal neovascularisation
Experimental Eye Research 89 (2009) 791–800
Contents lists available at ScienceDirect
Experimental Eye Research journal homepage: www.elsevier.com/locate/yexer
Low power laser treatment of the retina ameliorates neovascularisation in a transgenic mouse model of retinal neovascularisation Paula K. Yu, Stephen J. Cringle, Ian L. McAllister, Dao-Yi Yu* Centre for Ophthalmology and Visual Science and the ARC Centre of Excellence in Vision Science, The University of Western Australia, 2 Verdun Street, Nedlands, Perth, Western Australia 6009, Australia
a r t i c l e i n f o
a b s t r a c t
Article history: Received 10 July 2008 Accepted in revised form 9 July 2009 Available online 16 July 2009
This study was designed to determine if low power laser therapy can achieve amelioration of vasoproliferation yet preserve useful vision in the treated area in a transgenic mouse model of retinal neovascularisation. The mice were anaesthetised and the pupils dilated for ERG and fundus fluorescein angiography on postnatal day 32. The left eyes were treated with approximately 85 laser spots (532 nm, 50 ms, 300 mm diameter) at a power level of 20 mW at the cornea. The eyes were examined using ERG and fluorescein angiography, one, four and six weeks later. Flat mounts of FITC-dextran infused retinas, retinal histology and PEDF immunohistochemistry was studied one or six weeks after laser treatment. In untreated eyes the expected course of retinal neovascularisation in this model was observed. However, retinal neovascularisation in the laser treated eye was significantly reduced. The laser parameters chosen produced only mild lesions which took 10–20 s to become visible. ERG responses were comparable between the treated and untreated eyes, and histology showed only partial loss of photoreceptors in the treated eyes. PEDF intensity corresponded inversely with the extent of neovascularisation. Low power panretinal photocoagulation can inhibit retinal neovascularisation and yet preserve partial visual function in this transgenic mouse model of retinal neovascularisation. Ó 2009 Elsevier Ltd. All rights reserved.
Keywords: photocoagulation VEGF oxygen PEDF FITC-dextran
1. Introduction Conventional laser photocoagulation is the established treatment for the sight-threatening consequences of ischemic retinal diseases, including diabetic retinopathy. The predominant mechanism by which this treatment is thought to work is through an improvement in the oxygenation status of the ischemic/hypoxic retina sufficient to ameliorate the hypoxia-induced upregulation of vascular endothelial growth factor (VEGF) that drives the vasoproliferative response to retinal ischemia (Wolbarsht and Landers, 1980; Aiello et al., 1994; Takagi et al., 1996; Ozaki et al., 1999; Stefansson, 2006). Although conventional laser photocoagulation is effective, it is also destructive, resulting in total visual loss in the treated area. Peripheral retina is essentially sacrificed to preserve central retina. If the condition progresses, the treated area may need to be progressively enlarged resulting in extensive loss of visual field and night vision.
* Corresponding author. Tel.: þ61 8 9381 0716; fax: þ61 8 9381 0700. E-mail addresses: [email protected] (P.K. Yu), [email protected]. edu.au (S.J. Cringle), [email protected] (I.L. McAllister), dyyu@cyllene. uwa.edu.au (D.-Y. Yu). 0014-4835/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.exer.2009.07.004
Although in recent decades there have been many studies attempting to reduce the level of retinal damage in retinal photocoagulation (Seiberth and Alexandridis, 1991; Bandello et al., 2001; Fong et al., 2007), it has been difficult to define a suitable damage level to achieve a given therapeutic benefit with minimum damage to the retina. Clinical outcomes vary dramatically between reports. Some report that there was no significant difference in outcome between lower and higher power, suggesting that lower laser power should be used (Seiberth and Alexandridis, 1991; Pahor, 1998; Bandello et al., 2001). However, other reports suggest that conventional laser photocoagulation should continue to be the standard therapy (Fong et al., 2007). One of the difficulties in such studies is that the exact mechanism by which laser therapy ameliorates the vasoproliferative response is not known. A strong candidate is improvement of the oxygen status in the inner retina in the treated area due to reduced oxygen consumption of the outer retina in the laser treated area. Only in animal studies can the relationship between laser therapy, intraretinal oxygen distribution and consumption, and the degree of retinal damage be determined. Recent data from studies in rabbits has confirmed an improvement in inner retinal oxygen levels following laser therapy and provided a dose–response relationship (Yu et al., 2005b). The rabbit was
792
P.K. Yu et al. / Experimental Eye Research 89 (2009) 791–800
chosen to take advantage of the avascular nature of the rabbit retina and the detailed knowledge of the normal oxygen requirements of the inner and outer retina (Yu and Cringle, 2004). Whilst the rabbit provides a useful model for studying the basic mechanisms of laser therapy, the absence of retinal vasculature in the vast majority of the retina makes the rabbit unsuitable for studies of retinal neovascularisation. Another candidate in the effectiveness of laser panretinal photocoagulation (PRP) in reducing neovascularisation may be the antiangiogenic pigmented epithelium derived factor (PEDF). PEDF is produced at high levels by retinal pigment epithelium (RPE) after photocoagulation in cell culture and has been shown to modulate neovascularisation in a rat model of laser induced choroidal neovascularisation (Ogata et al., 2002; Hattenbach et al., 2005). Injection of adenoviral vectored PEDF gene transfer in mouse models of proliferative retinopathy has been shown to cause regression of ocular neovascularisation by promoting apoptosis of cells within the neovascular lesions (Mori et al., 2002). Antiangiogenic activity of PEDF in the human retina has also been implied by low levels of PEDF in the vitreous of eyes with proliferative diabetic retinopathy (Ogata et al., 2001). The purpose of the present study was to directly test whether a significantly lower level of retinal damage than currently used in retinal photocoagulation therapy can produce a therapeutic benefit to ameliorate retinal neovascularisation, yet still preserve some vision in the treated area. The animal model chosen was the V-6 line of C57BL/6J (Rho/VEGF) transgenic mouse (Okamoto et al., 1997; Campochiaro and Hackett, 2003). Although our microelectrode based techniques for measuring intraretinal oxygen levels in rats (Yu et al., 1994) were recently improved to allow work in mice (Yu and Cringle, 2006), interpretation of laser induced oxygen changes is complicated by the spatial variation in oxygen distribution and the masking effect of likely regulation of the retinal circulation. We therefore chose to directly address the question whether or not low power laser therapy could prevent retinal neovascularisation. There is considerable supportive evidence that such an approach is feasible. There is solid epidemiological evidence that indicates that partial loss of photoreceptors (such as in retinitis pigmentosa (RP)) is protective against proliferation of retina vessels in ischemic retinal diseases since diabetic patients who also have RP do not develop diabetic retinopathy (Arden, 2001). We have recently measured the changes in the intraretinal oxygen environment in animal models of RP, and have demonstrated that the reduction of oxygen consumption by the depleted photoreceptor layer does not create a widespread increase in inner retinal oxygen tension, but it does create an increase in oxygen flux from the choroid to the middle retinal layers (Yu et al., 2000, 2004). Thus, it is possible that in RP patients there is only a moderate increase in oxygenation of the middle retinal layers, yet this is sufficient to protect against vascular proliferation in RP sufferers who are also diabetic. In clinical diabetic retinopathy the appearance, or worsening, of certain intraretinal lesions is a crucial risk factor for the development of ocular neovascularisation on the surface of the retina (The Diabetic Retinopathy Study Research Group, 1987), providing further evidence that it may be hypoxia of the middle retinal layers that stimulates the vasoproliferative response. We have recently demonstrated that partial destruction of the photoreceptor array by laser therapy can yield significant oxygenation benefits to the middle retinal layers (Yu et al., 2005b). Based on these observations, the experimental focus of this study was to investigate a new concept for photocoagulation therapy in which only limited improvement to the oxygenation status of the middle retinal layers by low level of laser photocoagulation is required. In this study we provide a structural and functional
assessment of the effects of low power photocoagulation therapy applied in a well established model of retinal neovascularisation. 2. Materials and methods 2.1. Animals We used the V-6 line of C57BL/6J (Rho/VEGF) transgenic mice with the rhodopsin promoter/VEGF gene incorporated (Okamoto et al., 1997; Campochiaro and Hackett, 2003). Prof Peter Campochiaro from Johns Hopkins University kindly provided us with the initial breeding stock, and the colony was re-derived by the Animal Resource Centre, Western Australia. 2.2. Procedures All procedures conformed to ARVO resolution for the Use of Animals in Ophthalmic and Vision Research. For each examination the mice were anaesthetised with ketamine (30–45 mg/g) and xylazine (3–4.5 mg/g). Mydriacyl (1%) was applied to both eyes to dilate the pupils. The initial examination was conducted at postnatal day 32 (P32). ERG assessment was conducted prior to colour fundus photograph and fluorescein angiography of both eyes. Laser application to the left eye was performed at P34. ERG and fluorescein angiography measurement were then repeated at P41, P62 and P76 (one week, four weeks and 6 weeks after laser treatment). A subgroup of animals at P41 and P76 were infused with FITCdextran and the retinas prepared for flat mounting and confocal microscopy. Another subgroup was used for histology. 2.3. Laser parameters We used a 532 nm clinical ophthalmic laser for photocoagulation therapy (OcuLight GL, Iridex Medical Instruments Inc, Mountainview, CA, USA). The laser energy was delivered via an operating microscope adaptor which was attached to an operating microscope (Zeiss OPSM6-SF Germany) and through a plano-concave contact lens to allow a clear view of the ocular fundus. Due to possible losses in the fiberoptic delivery system the actual power delivered to the eye was measured at the cornea. A range of laser parameters were tested in pilot experiments before deciding on a power of 20 mW at the cornea with 50 ms duration and a 300 mm spot size for use in this study. With these parameters there was no immediate burn visible but a light grey lesion became visible 10–20 s after laser application. For the panretinal photocoagulation therapy, as much of the retina was treated as possible, requiring approximately 85 spots. 2.4. Fundus fluorescein angiography
A fundus camera (Canon DF-60Dsi) set for a 40 field of view was adapted for mice fundus photography with the addition of a 20D lens in front of the existing lens system. Colour fundus photographs were taken of both eyes and then 0.01–0.02 ml of 10% sodium fluorescein was injected intraperitoneally for fluorescein angiography. It took approximately 20 s for the retinal vessels to become visible on angiography and images were then captured for at least 3 min and up to 10 min in some cases. Images obtained after 2.5 min are considered late phase as this is the time when the leakage spots (if present) were clearly visible. Although there was often a slight increase in size of the spots at later time points there was little change in the pattern of leaky spots. At later time points the clarity of the images became degraded due to sodium fluorescein diffusing into the vitreous, thus only images obtained at 2.5 min or shortly after were used for later analysis.
P.K. Yu et al. / Experimental Eye Research 89 (2009) 791–800
793
2.5. Electroretinogram (ERG)
2.9. Photoreceptor nuclei counting
All ERG measurements were performed with a commercial electroretinography suite (Espion, ColorBurst; Diagnosys LLC, Littleton, MA, USA). Procedures were based on a published method of ERG screening in mice using the same electrophysiology system (Dalke et al., 2004). Chlorided silver wires (125 mm in diameter) were used as electrodes. A ground reference electrode was inserted subcutaneously in the neck fold. Corneal electrodes were made from a coil of silver wire formed around a 25G needle and freshly chlorided before use. A drop of buffered saline solution was used to keep good electrical contact with the cornea. Corresponding reference electrodes were placed subcutaneously in the preauricular area. The mice were dark adapted for 20 min before ERG measurements commenced. A series of measurements were taken over a range of stimulus intensities against a scotopic background. The amplitude of the a- and b-waves was measured.
Digital images were taken from the Toluidine Blue stained epoxy sections of the treated and untreated eyes. Transverse semithin sections were taken from each eye at 500 mm intervals starting from optic disk region. Particular care was taken to ensure the retina was sectioned perpendicular to the curvature of the eye. Two to three sections were obtained from each eye spanning an overall distance of 1.5 mm from the optic disk. Images were taken using a 20 Plan Apo objective lens on a Nikon E800 microscope and analysed using Image Pro Plus (Media Cybernetics Inc, MD, USA). The images were first converted to gray scale and contrast adjusted to highlight the photoreceptor nuclei. The number of nuclei in the outer nuclear layer was counted. Only areas with single file of ganglion cell layer and without clustering of neovascularisation in the outer retina were included. Nuclei with dimensions of dark pixels at a width of 1–4 mm and a length up to 6 mm were counted. This excluded small fragments of partially sectioned photoreceptor nuclei from the count. The numbers of nuclei were expressed as per 500 mm measured from each eye and the percentage reduction in nuclei number compared to the untreated eye of the same animal.
2.6. Histology With the mice deeply anaesthetised, both eyes were enucleated. The corneas were punctured with a 25G needle and the eyes immersion fixed in 4% paraformaldehyde in 0.1 M phosphate buffer solution. The eyes were dissected for epoxy processing and frozen for cryosectioning. Semi-thin epoxy sections (1 mm) were cut and stained with Toluidine Blue and cryosections for immunohistochemistry staining.
2.10. FITC-dextran and confocal microscopy
7 mm thick cryosections were obtained from at least six pairs of treated and untreated eyes at each time point. After bringing the sections to room temperature, the sections were washed in three changes of phosphate buffered saline (PBS, 0.01 M, pH 7.20) of 5 min each and permeabilised in 3% H2O2 in methanol for 10 min. The sections were then washed in three changes of PBS Tween 20 (0.05%) and blocked for non-specific staining using 10% goat serum (in PBS) for an hour. The sections were then incubated with a monoclonal mouse anti-pigment-epithelium derived factor (MAB1059, Chemicon International; 1 in 100 dilution in PBS) for 2 h. After several washes with PBS Tween 20, the sections were incubated with a mixture of Alexa Fluor 488 tagged goat anti-mouse IgG antibodies (A11001, Invitrogen; 1:200 dilution) and bisBenzimide H 33342 trihydrochloride (B2261, Sigma–Aldrich; 1.2 mg/ml) in PBS for 2 h. The sections were washed thoroughly in a few changes of PBS over 30 min before being mounted in Hydramount for viewing and imaging.
At least 6 animals from P41 and P76 time points were successfully perfused with FITC-dextran according to the method of D’Amato et al. (1993) and studied using confocal microscopy. In brief, the mouse was anaesthetised and the thoracic cavity opened to expose the pumping heart. One millilitre of 50 mg FITC-dextran (2 million MW from Sigma) dissolved in phosphate buffered saline was slowly injected into the left ventricle and allowed to circulate through the mouse body. Successful labelling of the microvasculature in the eye could be predicted from the green appearance of the front paws and a green tint of the underside of the mouse tongue. The eyes were then enucleated, the cornea punctured with a 25G needle tip, and immersion fixed overnight in ice cold 4% paraformaldehyde in 0.1 M phosphate buffered solution. For confocal microscopy, the anterior chamber was removed and four radial slits made to the posterior globe to enable flat mounting. A drop of hydramount and a coverslip was used to flatten the specimen. Z-stacks were taken at 4 mm intervals to cover all visible microvasculature through the thickness of the retina from the four quadrants using a 10 objective lens on a Nikon E800 microscope and the 488 nm line on the Nikon C1 system.
2.8. Intensity measurement
2.11. Image analysis
Confocal images were obtained from the paired immunolabelled sections (treated and untreated from the same animal) using exactly the same settings of laser intensity and detector gain using the 40 Plan Apo oil lens at a single z-axis plane. (The ability of the confocal to obtain images at the same optical thickness can eliminate artefacts and inaccuracy in intensity measurement resulting from difference in section thickness.) Using Image Pro Plus (v5.10) intensity measurement function, an average of 68.6 5.21 spots at a set sample size of 3.48 mm2 were taken from each of the gray scale 12 bits images at the junction of apical retinal pigmented epithelium and the outer edge of photoreceptor outer segment. The mean intensity measurement from each sample point was obtained as a data point. The results were statistically analysed using the Mann–Whitney Rank Sum test. As background correction was not performed, the difference in intensity measurement is relative and not absolute values.
Images obtained from fundus fluorescein angiography and confocal microscopy were analysed using Image Pro Plus (v5.1). The number of leaky spots was counted for each eye at each time point using the selected late phase images. For images obtained from confocal microscopy, a projection of images from the same field of view was obtained. This was cross-referenced with a fluorescein angiogram taken before sacrifice to match precisely the region of interest. The area occupied by tortuous, non-uniform, unusual meshwork of vessels on the confocal image projection stack which cross-referenced positively with regions showing leakage on fluorescein angiogram was measured. The accumulated area occupied by these leaky vessels from each retina was compared with the area occupied by all vasculature in the region and expressed as a percentage. Typically, the posterior globe region within 1 mm radius of the optic disc imaged by fluorescein angiography was included in the measurement.
2.7. Immunohistochemistry
794
P.K. Yu et al. / Experimental Eye Research 89 (2009) 791–800
Fig. 1. Fluorescein angiogram. Fluorescein angiogram from the untreated (a, c, e, g) and laser treated (b, d, f, h) eye of a transgenic mouse taken on (a and b) day 32, before laser treatment; day 41 (c and d, 1 week after laser treatment), day 62 (e and f, 1 month after laser treatment) and day 76 (g and h, 6 weeks after laser treatment). Bright spots indicate locations of sodium fluorescein leakage.
P.K. Yu et al. / Experimental Eye Research 89 (2009) 791–800 Table 1 The average number of leaky spots SE as observed on fluorescein angiogram expressed as percentage of leaky spots before laser treatment. Time points
Untreated eye
P32 P41 P62 P76
100 125.9 18.30 75.6 8.27 58.5 10.56
795
3. Results 3.1. Evidence of reduction in retinal neovascularisation and leakage
Treated eye (21) (21) (6) (6)
100 22.9 4.83 33.01 15.02 20.8 14.26
(21) (21) (6) (6)
2.12. Statistical analysis All statistical testing was performed using the statistics program SigmaStat (SSPS Scientific; Chicago, IL). Student’s t test with an acceptance level of p < 0.05 was employed to determine any significant differences in retinal neovascularisation and leakage between the treated and untreated eyes and for the photoreceptor nuclear count measurements. All mean data is expressed as mean standard error.
3.1.1. Fundus fluorescein angiography Retinal neovascularisation and leakage was assessed in vivo using fundus fluorescein angiography. Fig. 1 shows fluorescein angiography images taken 2 days before laser treatment (P32), and one week (P41), 4 weeks (P62) and 6 weeks (P76) after laser treatment. The left eye was treated with laser and the right eye was left untreated. At P32, there were some leaky retinal vessels in both eyes as evidenced by fluorescein dye leakage in the late phase angiography. At P41 in this animal the fluorescein leakage appeared more severe in the untreated eye but the treated eye showed a reduction in fluorescein leakage. At P62 and P76 there were signs of reduced leakage in the untreated eye compared to that at P41. In the treated eye there was remarkably little evidence of fluorescein leakage when compared to the pre-treatment measurement at P32. Averaging such measurements across all animals at P32, the average number of
Fig. 2. A quadrant of projected images from flat mount of FITC-Dextran infused retina from mice sacrificed on day 41 (a and b, 1 week after laser treatment) and day 76 (c and d, 6 weeks after laser treatment). The corresponding untreated control retinas are shown in a and c. This P41 untreated specimen (a) clearly shows numerous round vascular structures (white arrow heads) distributed through the flat mount. By comparison, the contralateral eye that had been laser treated (b) had very few structures of that type. At 6 weeks postlaser treatment, unusual vascular structures may be seen in the untreated retina (c) and in the peripheral retina of the laser treated eye (d). However, these are denser in the untreated control retina (c) than the laser treated retina (d), especially comparing only the central 1 mm radius. These abnormal vascular structures appear to have a bigger diameter at P76 than at P41.
796
P.K. Yu et al. / Experimental Eye Research 89 (2009) 791–800
fluorescein leaky spots in the right eyes (18.1 2.8, n ¼ 21) was not significantly different (p ¼ 0.32) with that in the left eyes (23.3 4.5, n ¼ 21). Table 1 shows the averaged number of leaky spots as observed on fluorescein angiography at P41, P62 and P76 normalised to that at P32. In the untreated eyes there was no statistically significant difference at P41 (n ¼ 21, p ¼ 0.61), however, the number of leaky spots in the untreated eyes was significantly reduced from P32 levels to 75.6 8.3% at P62 and 58.5 10.6% at P76 (all n ¼ 6, p < 0.05). In the treated eyes the number of leaky spots was significantly less than pre-treatment levels at all time points (22.9 4.8% at P41, 33.0 15.0% at P62 and 27.6 14.3% at P76 (all p < 0.05)). It was also clearly evident that there was significantly less fluorescein leakage spots in the laser treated eye compared to the untreated eye at the same time point after laser treatment (all p < 0.05). 3.1.2. Flat mounted preparation of FITC-dextran infused retinas To assess the effects of low power laser treatment on the retinal neovascularisation histologically, flat mounted preparations of FITC-dextran infused retina were studied using confocal microscopy. With higher resolution and less aberration from overlying planes, retinal neovascularisation can clearly be identified. Fig. 2 shows a quadrant of projected images of retinas from each eye of FITC-dextran infused mice at P41 and P76. There were some clear differences between the treated and the untreated retina. In the untreated retina at P41, there were numerous new vessels distributed through the retina, while by comparison, the treated eye had relatively few structures of this type. Similar findings were also evident at P76 in which laser treated retina appeared to have less new vessels when compared with untreated retina. Quantitative data of neovascularisation from these flat mount studies were obtained in 6 animals at P41 and P76. Percentage area of the retina occupied by neovascularisation in the treated eyes was 4.65 1.65% at P41 and 5.80 1.82% at P76 which was significantly less than that in the untreated eyes 10.04 2.04% at P41 and 12.01 4.52% at P76 (p < 0.05). These data therefore support the results from the in vivo fluorescein angiography. 3.2. Retinal function Fig. 3 shows representative raw data of ERG measurements from untreated and laser treated eyes using different levels of flash intensity at different time points (P32, 41, 62 and 76). ERG responses were present and comparable through different time points between the treated and untreated eye. Fig. 4 shows a plot of the averaged a- and b-wave amplitudes. Prior to laser treatment, the ERGs from the left and right eyes were comparable (Fig. 4a). One week after laser treatment, however, there was a slight but significant (p < 0.05) drop in b-wave amplitude in the treated eye at stimulus levels of 2.25 and 5 cd s/m2 (Fig. 4b). One month after laser treatment, this drop in b-wave amplitude became insignificant (Fig. 4c). By six weeks, there was no difference seen in a-wave or b-wave amplitudes from the control and the treated eyes (Fig. 4d). In the longer term, the laser treatment did not appear to significantly impair visual function as indicated by ERG measurements. This is in contrast to the significant suppression of ERG amplitude with conventional PRP therapy (Capoferri et al., 1990). 3.3. Partial loss of the photoreceptors after low power laser treatments Representative images of retinal transverse sections in treated and untreated eyes at P41 and P76 are shown in Fig. 5. The layered structure within retina was well preserved in most regions, and there were some vessels in the outer nuclear layers indicating the presence of the retinal neovascularisation in both treated and
Fig. 3. Representative scotopic ERG recordings from untreated (a) and laser treated (b) eyes at different time points (P32, 41, 62 and 74 corresponding to columns from left to right). Numbers on the right indicate flash intensity level (cd s/m2). Scale bars on the bottom left hand corner equals to 0.2 mV (vertical) and 0.5 s (horizontal). ERG responses were comparable through different time points between the treated and untreated eye.
untreated groups. It was clearly evident that the number of new vessels varied with different regions, which is typical of this transgenic model. There was no obvious change in the inner retina, but there was thinning of the outer retina in the treated retina. Retinal pigmented epithelium showed evidence of laser effect including hypo-pigmentation, hyper-pigmentation, vacuolation and the presence of macrophages (Fig. 5b). With counting the nuclear number in the outer nuclear layer, percentage reduction in the number in the treated eyes was 18.9 4.88% at P41 and 11.5 4.09% at P76 when compared with the untreated eyes (all p < 0.05). 3.4. PEDF immunohistochemistry An increase in PEDF staining intensity was observed in the outer retina. The difference appears most marked at the border of RPE and photoreceptor outer segments. Intensity measurements (Table 2) taken at this region (Fig. 6) showed significant (p < 0.001) increase in intensity indicating increased PEDF protein presence in the treated retina. Significant increase in intensity was observed for both time points in the treated retina. Comparison in PEDF staining intensity between the untreated retinas at the two time points revealed a significant increase in intensity at 6 weeks post-treatment (Fig. 6) and may explain the reduction in leaky spots observed between 1 and 6-week post-treatment time points (Table 1). Comparison in PEDF staining intensity between the treated retinas at the two time points also revealed a significant increase in intensity at 6 weeks post-treatment (Table 2). 4. Discussion Panretinal photocoagulation is a common and well established therapy for a range of ischemic retinal diseases and diabetic retinopathy (Aiello et al., 1968; Branch retinal vein occlusion study
P.K. Yu et al. / Experimental Eye Research 89 (2009) 791–800
797
Fig. 4. Plots of averaged a- and b-wave amplitudes at a range of stimulus intensities at day 32 (a), 41 (b), 62 (c) and 76 (d) respectively. Standard errors are indicated and significant difference between treated and untreated eyes indicated with an asterisk.
group, 1984; Algvere et al., 1990; Early treatment diabetic retinopathy study research group, 1991; Arnarsson and Stefansson, 2000). The efficacy of such therapy is particularly well demonstrated in the treatment of proliferative diabetic retinopathy (Aiello et al., 1968; Early treatment diabetic retinopathy study research group, 1991). There may however be further room for improvement in terms of clinical outcomes. The degree of laser damage required to produce the therapeutic benefit is not known. One candidate for the ameliorative effects of laser therapy on vasoproliferation is the relief of retinal hypoxia due to destruction of the photoreceptors in the treated area (Stefansson et al., 1981, 1986; Landers et al., 1982; Pournaras et al., 1990; Zuckerman et al., 1993). This is then thought to lessen the production of the vasoproliferative factor VEGF due to the lessening of hypoxia-induced VEGF production (Wolbarsht and Landers, 1980; Boulton et al., 1998; Stefansson, 2001). It is not currently known how much reduction in retinal hypoxia is required or in which retinal layers such an effect is most critical. We have outlined a hypothesis that only a modest improvement in intraretinal oxygen levels may be required and that it is the middle retinal layers that may be most important. The supportive evidence comes from several sources. It is now well established that the retinas of patients with partial photoreceptor loss, such as in RP, are not susceptible to the proliferative effects of diabetic retinopathy (Arden, 2001). Although this may be of little comfort to the RP patient, it may point to an important factor in the pathogenesis of diabetic retinopathy. It had been suggested that partial loss of photoreceptors could result in increased oxygen levels in the
remaining retina due to a reduction in oxygen consumption of the depleted photoreceptor array (Stone et al., 1999). Recent work in animal models of RP failed to find a generalized increase in inner retinal oxygen levels following photoreceptor degeneration, but did demonstrate an increased oxygen flux into the middle retinal layers (Yu et al., 2000, 2004). This change in balance between the potential supply of oxygen from the choroidal and retinal vasculature to the middle retinal layers could provide an amelioration of ischemic insults to the retinal circulation. We were also able to show that laser therapy could be used to create similar oxygenation changes in otherwise normal animals (Yu et al., 2005b). To overcome the confounding effects of autoregulation of the retinal circulation we performed such measurements in the avascular region of the rabbit retina. We were able to quantify the reduction in outer retinal oxygen consumption for different powers, wavelengths, and modes of laser delivery. It was established that significant improvements of intraretinal oxygen level could be produced using laser powers that created very mild lesions as judged by their ophthalmoscopic appearance. The present study sought to investigate the effectiveness of low power laser therapy in a well established mouse model of retinal neovascularisation (Okamoto et al., 1997; Campochiaro and Hackett, 2003). The model is characterised by the early appearance of intraretinal neovascularisation (Ida et al., 2003). It is readily accepted that this is not a direct model of diabetic retinopathy, but it is comparable in that the vasoproliferation is driven by over-expression of VEGF. We therefore examined the ability of low power laser therapy to
Fig. 5. Epoxy semi-thin cross-sections from untreated (a and c) and treated (b and d) eyes at P41 (a and b) and P76 (c and d) stained with Toluidine Blue. The layered structure of the retina was retained in the treated retinae (b and d). However, some sparseness can be observed in the outer nuclear layer of the treated retinas suggesting a reduction in photoreceptor nuclei number. Changes such as hypo-pigmentation and vacuolation in the retinal pigmented epithelium (RPE) can be observed in the laser treated retina. A few pigment loaded macrophages are visible in the laser treated retina (arrow heads). Blood vessels may be seen within the outer nuclear layer (black arrows) in the treated and untreated retinas. Scale bar 100 mm.
ameliorate the vasoproliferative response in this model. This was confirmed by fluorescein fundus photography, by whole mount analysis of new vessel formation, and by increased presence of PEDF. In addition we sought to confirm that low power laser therapy had the potential for only partial destruction of the photoreceptor array, thus preserving some useful retinal function in the treated areas. The functional assessment was based on ERG amplitudes, which after an initial recovery period were not
Table 2 Intensity measurement (mean SE) from confocal images of PEDF immunostaining. (Intensity range from 0 to 4096). Statistically significant difference was found between all groups compared (p 0.001). Number in bracket indicates the number of measurements made for that group. Intensity
P41
P76
Untreated Treated
435.3 14.47 (727) 755.9 24.13 (685)
782.7 39.32 (429) 851.6 31.04 (451)
P.K. Yu et al. / Experimental Eye Research 89 (2009) 791–800
799
Fig. 6. PEDF immunohistochemistry. Confocal images taken from an untreated (left panel) and a treated (right panel) retina at P41 and P76 time points. PEDF has been labelled with an Alexa Fluor 488 secondary antibody and pseudo coloured green. The nuclei have been counterstained using bisBenzimide H and pseudo coloured blue. The treated retina showed stronger PEDF staining particularly at the border of the outer segments and retinal pigmented epithelium (RPE). The RPE layer has been enlarged for comparison purpose between the treatment group and different time points (Scale bar measures 20 mm). Untreated retina at P41 has weak PEDF labelling compared to the treated retina at the same time point. The untreated retina at P76 is much more intensely labelled than at P41. The treated retina at P76 is also more intensely labelled compared to all the other data points. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article).
significantly reduced in the treated eyes. Histological assessment of the treated retina showed only partial loss of the photoreceptor array. Whilst extrapolation of studies in mice to imply a given outcome in clinical studies should be done with extreme caution, it is comforting to note that in terms of intraretinal oxygen distribution and consumption these are remarkable similarities between measurements recently made in the mouse (Yu and Cringle, 2006) and the monkey (Yu et al., 2005a). Whilst intraretinal oxygen microelectrode work in the mouse is technically feasible, the use of such techniques in a model with such localized vascular dysfunction and localized changes in retinal structure is problematic. The more global measures of vascular leakage, vascular structure, and electrophysiological function provide more accurate markers of neovascularisation and retinal function. The present study cannot determine the relative importance of different mechanisms of the anti-proliferative effects of laser therapy in this model. Whilst an increase in intraretinal oxygen level is at the heart of our original hypothesis, the upregulation of PEDF may well be an important factor, as could the simple reduction of VEGF production due to partial loss of photoreceptors (the primary site of excess VEGF production in this model). Whatever the specific mechanisms involved it is comforting to note that the present study suggests that it may be feasible that low power laser therapy could have applications in the clinical management of a range of ischemic retinal diseases without the patient having to suffer complete loss of vision in the treated area. Acknowledgements The authors acknowledge the expert technical support of Mr Dean Darcey, Dr Joe Miller and Ms Zoya Gridneva. We also wish to acknowledge the provision of the mouse model by Professor Peter Campochiaro, Johns Hopkins University, USA. Grant support was provided by the National Health and Medical Research Council of Australia and the Australian Research Council Centre of Excellence in Vision Science. References Aiello, L.M., Beetham, W.P., Balodimos, M.C., Chazan, B.I., Bradley, R.F.,1968. Ruby laser photocoagulation in treatment of diabetic proliferating retinopathy. Preliminary Report. In: Goldberg, M.F., Fine, S.L. (Eds.), Symposium: Treatment of Diabetic Retinopathy. U.S. Government Printing Office, pp. 437–463 (Chapter 35).
Aiello, L.P., Avery, R.L., Arrigg, P.G., Keyt, B.A., Jampel, H.D., Shah, S.T., Pasquale, L.R., Thieme, H., Iwamoto, M.A., Park, J.E., Nguyen, H.V., Aiello, L.M., Ferrara, N., King, G.L., 1994. Vascular endothelial growth factor in ocular fluid of patients with diabetic retinopathy and other retinal disorders. N. Engl. J. Med. 331, 1480–1487. Algvere, P.V., Hallnas, K., Dafgard, E., Hoog, A., 1990. Panretinal photocoagulation aggravates experimental proliferative vitreoretinopathy. Graefes Arch. Clin. Exp. Ophthalmol. 228, 461–466. Arden, G.B., 2001. The absence of diabetic retinopathy in patients with retinitis pigmentosa: implications for pathophysiology and possible treatment. Br. J. Ophthalmol. 85, 366–370. Arnarsson, A., Stefansson, E., 2000. Laser treatment and the mechanism of edema reduction in branch retinal vein occlusion. Invest. Ophthalmol. Vis. Sci. 41, 877–879. Bandello, F., Brancato, R., Menchini, U., Virgili, G., Lanzetta, P., Ferrari, E., Incorvaia, C., 2001. Light panretinal photocoagulation (LPRP) versus classic panretinal photocoagulation (CPRP) in proliferative diabetic retinopathy. Semin. Ophthalmol. 16, 12–18. Boulton, M., Foreman, D., Williams, G., McLeod, D., 1998. VEGF localisation in diabetic retinopathy. Br. J. Ophthalmol. 82, 561–568. Branch retinal vein occlusion study group, 1984. Argon laser photocoagulation for macular oedema in branch vein occlusion. Am. J. Ophthalmol. 98, 271–282. Campochiaro, P.A., Hackett, S.F., 2003. Ocular neovascularization: a valuable model system. Oncogene 22, 6537–6548. Capoferri, C., Baqini, M., Chizzoli, A., Pece, A., Brancato, R., 1990. Electroretinographic findings in panretinal photocoagulation for diabetic retinopathy. A randomized study with blue-green argon and red krypton lasers. Graefes Arch. Clin. Exp. Ophthalmol. 228, 232–236. D’Amato, R., Wesolowski, E., Hodgson Smith, L.E., 1993. Microscopic visualization of the retina by angiography with high-molecular-weight fluorescein-labeled dextrans in the mouse. Microvasc. Res. 46, 135–142. Dalke, C., Loster, J., Fuchs, H., Gailus-Durner, V., Soewarto, D., Favor, J., NeuhauserKlaus, A., Pretsch, W., Gekeler, F., Shinoda, K., Zrenner, E., Meitinger, T., Hrabe de Angelis, M., Graw, J., 2004. Electroretinography as a screening method for mutations causing retinal dysfunction in mice. Invest. Ophthalmol. Vis. Sci. 45, 601–609. Early treatment diabetic retinopathy study research group, 1991. Early photocoagulation for diabetic retinopathy. ETDRS report number 9. Ophthalmology 98, 766–785. Fong, D.S., Strauber, S.F., Aiello, L.P., Beck, R.W., Callanan, D.G., Danis, R.P., Davis, M.D., Feman, S.S., Ferris, F., Friedman, S.M., Garcia, C.A., Glassman, A.R., Han, D.P., Le, D., Kollman, C., Lauer, A.K., Recchia, F.M., Solomon, S.D., 2007. Comparison of the modified Early Treatment Diabetic Retinopathy Study and mild macular grid laser photocoagulation strategies for diabetic macular edema. Arch. Ophthalmol. 125, 469–480. Hattenbach, L.O., Beck, K.F., Pfeilschifter, J., Koch, F., Ohrloff, C., Schacke, W., 2005. Pigment-epithelium-derived factor is upregulated in photocoagulated human retinal pigment epithelial cells. Ophthalmic Res. 37, 341–346. Ida, H., Tobe, T., Nambu, H., Matsumura, M., Uyama, M., Campochiaro, P.A., 2003. RPE cells modulate subretinal neovascularization, but do not cause regression in mice with sustained expression of VEGF. Invest. Ophthalmol. Vis. Sci. 44, 5430–5437. Landers, M.B., Stefansson, E., Wolbarsht, M.L., 1982. Panretinal photocoagulation and retinal oxygenation. Retina 2, 167–175. Mori, K., Gehlbach, P., Ando, A., McVey, D., Wei, L., Campochiaro, P.A., 2002. Regression of ocular neovascularization in response to increased expression of
800
P.K. Yu et al. / Experimental Eye Research 89 (2009) 791–800
pigment epithelium-derived factor. Invest. Ophthalmol. Vis. Sci. 2002, 2428–2434. Ogata, N., Wada, M., Otsuji, T., Jo, N., Tombran-Tink, J., Matsumura, M., 2002. Expression of pigment epithelium-derived factor in normal adult rat eye and experimental choroidal neovascularization. Invest. Ophthalmol. Vis. Sci. 43, 1168–1175. Ogata, N., Wada, M., Otsuji, T., Jo, N., Tombran-Tink, J., Matsumura, M., 2001. Pigment epithelium derived factor in the vitreous is low in diabetic retinopathy and high in rhegmatogenous retinal detachment. Am. J. Ophthalmol. 132, 378–382. Okamoto, N., Tobe, T., Hackett, S.F., Ozaki, H., Vinores, M.A., LaRochelle, W., Zack, D.J., Campochiaro, P.A., 1997. Transgenic mice with increased expression of vascular endothelial growth factor in the retina. Am. J. Pathol. 151, 281–291. Ozaki, H., Yu, A.Y., Della, N., Ozaki, K., Luna, J.D., Yamada, H., Hackett, S.F., Okamoto, N., Zack, D.J., Semenza, G.L., Campochiaro, P.A., 1999. Hypoxia inducible factor-1alpha is increased in ischemic retina: temporal and spatial correlation with VEGF expression. Invest. Ophthalmol. Vis. Sci. 40, 182–189. Pahor, D., 1998. Visual field loss after argon laser panretinal photocoagulation in diabetic retinopathy: full-versus mild-scatter coagulation. Int. Ophthalmol. 22, 313–319. Pournaras, C.J., Tsacopoulos, M., Strommer, K., Gilodi, N., Leuenberger, P.M., 1990. Scatter photocoagulation restores tissue hypoxia in experimental vasoproliferative microangiopathy in miniature pigs. Ophthalmology 97, 1329–1333. Seiberth, V., Alexandridis, E., 1991. Function of the diabetic retina after panretinal argon laser photocoagulation. Influence of the intensity of the coagulation spots. Ophthalmologica 202, 10–17. Stefansson, E., 2001. The therapeutic effects of retinal laser treatment and vitrectomy. A theory based on oxygen and vascular physiology. Acta Ophthalmol. Scand. 79, 435–440. Stefansson, E., 2006. Ocular oxygenation and the treatment of diabetic retinopathy. Surv. of Ophthalmol 51, 364–380. Stefansson, E., Hatchell, D.L., Fisher, B.L., Sutherland, F.S., Machemer, R., 1986. Panretinal photocoagulation and retinal oxygenation in normal and diabetic cats. Am. J. Ophthalmol. 101, 657–664. Stefansson, E., Landers, M.B., Wolbarsht, M.L., 1981. Increased retinal oxygen supply following pan-retinal photocoagulation and vitrectomy and lensectomy. Trans. Am. Ophthalmol. Soc. LXXIX, 307–334.
Stone, J., Maslim, J., Valter-Kocsi, K., Mervin, K., Bowers, F., Chu, Y., Barnett, N., Provis, J., Lewis, G., Fisher, S.K., Bisti, S., Gargini, C., Cervetto, L., Merin, S., Pe’er, J., 1999. Mechanisms of photoreceptor death and survival in mammalian retina. Prog. Retin. Eye Res. 18, 689–735. Takagi, H., King, G.L., Ferrara, N., Aiello, L.P., 1996. Hypoxia regulates vascular endothelial growth factor receptor KDR/Flk gene expression through adenosine A2 receptors in retinal capillary endothelial cells. Invest. Ophthalmol. Vis. Sci. 37, 1311–1321. The Diabetic Retinopathy Study Research Group, 1987. Indications for photocoagulation treatment of diabetic retinopathy: diabetic Retinopathy Study Report no. 14. Int. Ophthalmol. Clin. 27, 239–253. Wolbarsht, M.L., Landers III, M.B., 1980. The rationale of photocoagulation therapy for proliferative diabetic retinopathy: a review and a model. Ophthalmic Surg. 11, 235–245. Yu, D.-Y., Cringle, S.J., 2004. Low oxygen consumption in the inner retina of the visual streak of the rabbit. Am. J. Physiol. (Heart C) 286, H419–H423. Yu, D.-Y., Cringle, S.J., 2006. Oxygen distribution in the mouse retina. Invest. Ophthalmol. Vis. Sci. 47, 1109–1112. Yu, D.-Y., Cringle, S.J., Alder, V.A., Su, E.N., 1994. Intraretinal oxygen distribution in rats as a function of systemic blood pressure. Am. J. Physiol. 36, H2498–H2507. Yu, D.-Y., Cringle, S.J., Su, E.-N., 2005a. Intraretinal oxygen distribution in the monkey retina and the response to systemic hyperoxia. Invest. Ophthalmol. Vis. Sci., 4728–4733. Yu, D.-Y., Cringle, S.J., Su, E.-N., Yu, P.K., Humayun, M.S., Dorin, G., 2005b. Laser induced changes in intraretinal oxygen distribution in pigmented rabbits. Invest. Ophthalmol. Vis. Sci. 46, 988–999. Yu, D.-Y., Cringle, S.J., Su, E.N., Yu, P.K., 2000. Intraretinal oxygen levels before and after photoreceptor loss in the RCS rat. Invest. Ophthalmol. Vis. Sci. 41, 3999–4006. Yu, D.-Y., Cringle, S.J., Valter, K., Walsh, N., Lee, D., Stone, J., 2004. Photoreceptor death, trophic factor expression, retinal oxygen status and photoreceptor function in the P23H rat: stress and protection in a slow degeneration. Invest. Ophthalmol. Vis. Sci. 45, 2013–2049. Zuckerman, R., Cheasty, J.E., Wang, Y., 1993. Optical mapping of inner retinal tissue PO2. Curr. Eye Res. 12, 809–825.
Contents lists available at ScienceDirect
Experimental Eye Research journal homepage: www.elsevier.com/locate/yexer
Low power laser treatment of the retina ameliorates neovascularisation in a transgenic mouse model of retinal neovascularisation Paula K. Yu, Stephen J. Cringle, Ian L. McAllister, Dao-Yi Yu* Centre for Ophthalmology and Visual Science and the ARC Centre of Excellence in Vision Science, The University of Western Australia, 2 Verdun Street, Nedlands, Perth, Western Australia 6009, Australia
a r t i c l e i n f o
a b s t r a c t
Article history: Received 10 July 2008 Accepted in revised form 9 July 2009 Available online 16 July 2009
This study was designed to determine if low power laser therapy can achieve amelioration of vasoproliferation yet preserve useful vision in the treated area in a transgenic mouse model of retinal neovascularisation. The mice were anaesthetised and the pupils dilated for ERG and fundus fluorescein angiography on postnatal day 32. The left eyes were treated with approximately 85 laser spots (532 nm, 50 ms, 300 mm diameter) at a power level of 20 mW at the cornea. The eyes were examined using ERG and fluorescein angiography, one, four and six weeks later. Flat mounts of FITC-dextran infused retinas, retinal histology and PEDF immunohistochemistry was studied one or six weeks after laser treatment. In untreated eyes the expected course of retinal neovascularisation in this model was observed. However, retinal neovascularisation in the laser treated eye was significantly reduced. The laser parameters chosen produced only mild lesions which took 10–20 s to become visible. ERG responses were comparable between the treated and untreated eyes, and histology showed only partial loss of photoreceptors in the treated eyes. PEDF intensity corresponded inversely with the extent of neovascularisation. Low power panretinal photocoagulation can inhibit retinal neovascularisation and yet preserve partial visual function in this transgenic mouse model of retinal neovascularisation. Ó 2009 Elsevier Ltd. All rights reserved.
Keywords: photocoagulation VEGF oxygen PEDF FITC-dextran
1. Introduction Conventional laser photocoagulation is the established treatment for the sight-threatening consequences of ischemic retinal diseases, including diabetic retinopathy. The predominant mechanism by which this treatment is thought to work is through an improvement in the oxygenation status of the ischemic/hypoxic retina sufficient to ameliorate the hypoxia-induced upregulation of vascular endothelial growth factor (VEGF) that drives the vasoproliferative response to retinal ischemia (Wolbarsht and Landers, 1980; Aiello et al., 1994; Takagi et al., 1996; Ozaki et al., 1999; Stefansson, 2006). Although conventional laser photocoagulation is effective, it is also destructive, resulting in total visual loss in the treated area. Peripheral retina is essentially sacrificed to preserve central retina. If the condition progresses, the treated area may need to be progressively enlarged resulting in extensive loss of visual field and night vision.
* Corresponding author. Tel.: þ61 8 9381 0716; fax: þ61 8 9381 0700. E-mail addresses: [email protected] (P.K. Yu), [email protected]. edu.au (S.J. Cringle), [email protected] (I.L. McAllister), dyyu@cyllene. uwa.edu.au (D.-Y. Yu). 0014-4835/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.exer.2009.07.004
Although in recent decades there have been many studies attempting to reduce the level of retinal damage in retinal photocoagulation (Seiberth and Alexandridis, 1991; Bandello et al., 2001; Fong et al., 2007), it has been difficult to define a suitable damage level to achieve a given therapeutic benefit with minimum damage to the retina. Clinical outcomes vary dramatically between reports. Some report that there was no significant difference in outcome between lower and higher power, suggesting that lower laser power should be used (Seiberth and Alexandridis, 1991; Pahor, 1998; Bandello et al., 2001). However, other reports suggest that conventional laser photocoagulation should continue to be the standard therapy (Fong et al., 2007). One of the difficulties in such studies is that the exact mechanism by which laser therapy ameliorates the vasoproliferative response is not known. A strong candidate is improvement of the oxygen status in the inner retina in the treated area due to reduced oxygen consumption of the outer retina in the laser treated area. Only in animal studies can the relationship between laser therapy, intraretinal oxygen distribution and consumption, and the degree of retinal damage be determined. Recent data from studies in rabbits has confirmed an improvement in inner retinal oxygen levels following laser therapy and provided a dose–response relationship (Yu et al., 2005b). The rabbit was
792
P.K. Yu et al. / Experimental Eye Research 89 (2009) 791–800
chosen to take advantage of the avascular nature of the rabbit retina and the detailed knowledge of the normal oxygen requirements of the inner and outer retina (Yu and Cringle, 2004). Whilst the rabbit provides a useful model for studying the basic mechanisms of laser therapy, the absence of retinal vasculature in the vast majority of the retina makes the rabbit unsuitable for studies of retinal neovascularisation. Another candidate in the effectiveness of laser panretinal photocoagulation (PRP) in reducing neovascularisation may be the antiangiogenic pigmented epithelium derived factor (PEDF). PEDF is produced at high levels by retinal pigment epithelium (RPE) after photocoagulation in cell culture and has been shown to modulate neovascularisation in a rat model of laser induced choroidal neovascularisation (Ogata et al., 2002; Hattenbach et al., 2005). Injection of adenoviral vectored PEDF gene transfer in mouse models of proliferative retinopathy has been shown to cause regression of ocular neovascularisation by promoting apoptosis of cells within the neovascular lesions (Mori et al., 2002). Antiangiogenic activity of PEDF in the human retina has also been implied by low levels of PEDF in the vitreous of eyes with proliferative diabetic retinopathy (Ogata et al., 2001). The purpose of the present study was to directly test whether a significantly lower level of retinal damage than currently used in retinal photocoagulation therapy can produce a therapeutic benefit to ameliorate retinal neovascularisation, yet still preserve some vision in the treated area. The animal model chosen was the V-6 line of C57BL/6J (Rho/VEGF) transgenic mouse (Okamoto et al., 1997; Campochiaro and Hackett, 2003). Although our microelectrode based techniques for measuring intraretinal oxygen levels in rats (Yu et al., 1994) were recently improved to allow work in mice (Yu and Cringle, 2006), interpretation of laser induced oxygen changes is complicated by the spatial variation in oxygen distribution and the masking effect of likely regulation of the retinal circulation. We therefore chose to directly address the question whether or not low power laser therapy could prevent retinal neovascularisation. There is considerable supportive evidence that such an approach is feasible. There is solid epidemiological evidence that indicates that partial loss of photoreceptors (such as in retinitis pigmentosa (RP)) is protective against proliferation of retina vessels in ischemic retinal diseases since diabetic patients who also have RP do not develop diabetic retinopathy (Arden, 2001). We have recently measured the changes in the intraretinal oxygen environment in animal models of RP, and have demonstrated that the reduction of oxygen consumption by the depleted photoreceptor layer does not create a widespread increase in inner retinal oxygen tension, but it does create an increase in oxygen flux from the choroid to the middle retinal layers (Yu et al., 2000, 2004). Thus, it is possible that in RP patients there is only a moderate increase in oxygenation of the middle retinal layers, yet this is sufficient to protect against vascular proliferation in RP sufferers who are also diabetic. In clinical diabetic retinopathy the appearance, or worsening, of certain intraretinal lesions is a crucial risk factor for the development of ocular neovascularisation on the surface of the retina (The Diabetic Retinopathy Study Research Group, 1987), providing further evidence that it may be hypoxia of the middle retinal layers that stimulates the vasoproliferative response. We have recently demonstrated that partial destruction of the photoreceptor array by laser therapy can yield significant oxygenation benefits to the middle retinal layers (Yu et al., 2005b). Based on these observations, the experimental focus of this study was to investigate a new concept for photocoagulation therapy in which only limited improvement to the oxygenation status of the middle retinal layers by low level of laser photocoagulation is required. In this study we provide a structural and functional
assessment of the effects of low power photocoagulation therapy applied in a well established model of retinal neovascularisation. 2. Materials and methods 2.1. Animals We used the V-6 line of C57BL/6J (Rho/VEGF) transgenic mice with the rhodopsin promoter/VEGF gene incorporated (Okamoto et al., 1997; Campochiaro and Hackett, 2003). Prof Peter Campochiaro from Johns Hopkins University kindly provided us with the initial breeding stock, and the colony was re-derived by the Animal Resource Centre, Western Australia. 2.2. Procedures All procedures conformed to ARVO resolution for the Use of Animals in Ophthalmic and Vision Research. For each examination the mice were anaesthetised with ketamine (30–45 mg/g) and xylazine (3–4.5 mg/g). Mydriacyl (1%) was applied to both eyes to dilate the pupils. The initial examination was conducted at postnatal day 32 (P32). ERG assessment was conducted prior to colour fundus photograph and fluorescein angiography of both eyes. Laser application to the left eye was performed at P34. ERG and fluorescein angiography measurement were then repeated at P41, P62 and P76 (one week, four weeks and 6 weeks after laser treatment). A subgroup of animals at P41 and P76 were infused with FITCdextran and the retinas prepared for flat mounting and confocal microscopy. Another subgroup was used for histology. 2.3. Laser parameters We used a 532 nm clinical ophthalmic laser for photocoagulation therapy (OcuLight GL, Iridex Medical Instruments Inc, Mountainview, CA, USA). The laser energy was delivered via an operating microscope adaptor which was attached to an operating microscope (Zeiss OPSM6-SF Germany) and through a plano-concave contact lens to allow a clear view of the ocular fundus. Due to possible losses in the fiberoptic delivery system the actual power delivered to the eye was measured at the cornea. A range of laser parameters were tested in pilot experiments before deciding on a power of 20 mW at the cornea with 50 ms duration and a 300 mm spot size for use in this study. With these parameters there was no immediate burn visible but a light grey lesion became visible 10–20 s after laser application. For the panretinal photocoagulation therapy, as much of the retina was treated as possible, requiring approximately 85 spots. 2.4. Fundus fluorescein angiography
A fundus camera (Canon DF-60Dsi) set for a 40 field of view was adapted for mice fundus photography with the addition of a 20D lens in front of the existing lens system. Colour fundus photographs were taken of both eyes and then 0.01–0.02 ml of 10% sodium fluorescein was injected intraperitoneally for fluorescein angiography. It took approximately 20 s for the retinal vessels to become visible on angiography and images were then captured for at least 3 min and up to 10 min in some cases. Images obtained after 2.5 min are considered late phase as this is the time when the leakage spots (if present) were clearly visible. Although there was often a slight increase in size of the spots at later time points there was little change in the pattern of leaky spots. At later time points the clarity of the images became degraded due to sodium fluorescein diffusing into the vitreous, thus only images obtained at 2.5 min or shortly after were used for later analysis.
P.K. Yu et al. / Experimental Eye Research 89 (2009) 791–800
793
2.5. Electroretinogram (ERG)
2.9. Photoreceptor nuclei counting
All ERG measurements were performed with a commercial electroretinography suite (Espion, ColorBurst; Diagnosys LLC, Littleton, MA, USA). Procedures were based on a published method of ERG screening in mice using the same electrophysiology system (Dalke et al., 2004). Chlorided silver wires (125 mm in diameter) were used as electrodes. A ground reference electrode was inserted subcutaneously in the neck fold. Corneal electrodes were made from a coil of silver wire formed around a 25G needle and freshly chlorided before use. A drop of buffered saline solution was used to keep good electrical contact with the cornea. Corresponding reference electrodes were placed subcutaneously in the preauricular area. The mice were dark adapted for 20 min before ERG measurements commenced. A series of measurements were taken over a range of stimulus intensities against a scotopic background. The amplitude of the a- and b-waves was measured.
Digital images were taken from the Toluidine Blue stained epoxy sections of the treated and untreated eyes. Transverse semithin sections were taken from each eye at 500 mm intervals starting from optic disk region. Particular care was taken to ensure the retina was sectioned perpendicular to the curvature of the eye. Two to three sections were obtained from each eye spanning an overall distance of 1.5 mm from the optic disk. Images were taken using a 20 Plan Apo objective lens on a Nikon E800 microscope and analysed using Image Pro Plus (Media Cybernetics Inc, MD, USA). The images were first converted to gray scale and contrast adjusted to highlight the photoreceptor nuclei. The number of nuclei in the outer nuclear layer was counted. Only areas with single file of ganglion cell layer and without clustering of neovascularisation in the outer retina were included. Nuclei with dimensions of dark pixels at a width of 1–4 mm and a length up to 6 mm were counted. This excluded small fragments of partially sectioned photoreceptor nuclei from the count. The numbers of nuclei were expressed as per 500 mm measured from each eye and the percentage reduction in nuclei number compared to the untreated eye of the same animal.
2.6. Histology With the mice deeply anaesthetised, both eyes were enucleated. The corneas were punctured with a 25G needle and the eyes immersion fixed in 4% paraformaldehyde in 0.1 M phosphate buffer solution. The eyes were dissected for epoxy processing and frozen for cryosectioning. Semi-thin epoxy sections (1 mm) were cut and stained with Toluidine Blue and cryosections for immunohistochemistry staining.
2.10. FITC-dextran and confocal microscopy
7 mm thick cryosections were obtained from at least six pairs of treated and untreated eyes at each time point. After bringing the sections to room temperature, the sections were washed in three changes of phosphate buffered saline (PBS, 0.01 M, pH 7.20) of 5 min each and permeabilised in 3% H2O2 in methanol for 10 min. The sections were then washed in three changes of PBS Tween 20 (0.05%) and blocked for non-specific staining using 10% goat serum (in PBS) for an hour. The sections were then incubated with a monoclonal mouse anti-pigment-epithelium derived factor (MAB1059, Chemicon International; 1 in 100 dilution in PBS) for 2 h. After several washes with PBS Tween 20, the sections were incubated with a mixture of Alexa Fluor 488 tagged goat anti-mouse IgG antibodies (A11001, Invitrogen; 1:200 dilution) and bisBenzimide H 33342 trihydrochloride (B2261, Sigma–Aldrich; 1.2 mg/ml) in PBS for 2 h. The sections were washed thoroughly in a few changes of PBS over 30 min before being mounted in Hydramount for viewing and imaging.
At least 6 animals from P41 and P76 time points were successfully perfused with FITC-dextran according to the method of D’Amato et al. (1993) and studied using confocal microscopy. In brief, the mouse was anaesthetised and the thoracic cavity opened to expose the pumping heart. One millilitre of 50 mg FITC-dextran (2 million MW from Sigma) dissolved in phosphate buffered saline was slowly injected into the left ventricle and allowed to circulate through the mouse body. Successful labelling of the microvasculature in the eye could be predicted from the green appearance of the front paws and a green tint of the underside of the mouse tongue. The eyes were then enucleated, the cornea punctured with a 25G needle tip, and immersion fixed overnight in ice cold 4% paraformaldehyde in 0.1 M phosphate buffered solution. For confocal microscopy, the anterior chamber was removed and four radial slits made to the posterior globe to enable flat mounting. A drop of hydramount and a coverslip was used to flatten the specimen. Z-stacks were taken at 4 mm intervals to cover all visible microvasculature through the thickness of the retina from the four quadrants using a 10 objective lens on a Nikon E800 microscope and the 488 nm line on the Nikon C1 system.
2.8. Intensity measurement
2.11. Image analysis
Confocal images were obtained from the paired immunolabelled sections (treated and untreated from the same animal) using exactly the same settings of laser intensity and detector gain using the 40 Plan Apo oil lens at a single z-axis plane. (The ability of the confocal to obtain images at the same optical thickness can eliminate artefacts and inaccuracy in intensity measurement resulting from difference in section thickness.) Using Image Pro Plus (v5.10) intensity measurement function, an average of 68.6 5.21 spots at a set sample size of 3.48 mm2 were taken from each of the gray scale 12 bits images at the junction of apical retinal pigmented epithelium and the outer edge of photoreceptor outer segment. The mean intensity measurement from each sample point was obtained as a data point. The results were statistically analysed using the Mann–Whitney Rank Sum test. As background correction was not performed, the difference in intensity measurement is relative and not absolute values.
Images obtained from fundus fluorescein angiography and confocal microscopy were analysed using Image Pro Plus (v5.1). The number of leaky spots was counted for each eye at each time point using the selected late phase images. For images obtained from confocal microscopy, a projection of images from the same field of view was obtained. This was cross-referenced with a fluorescein angiogram taken before sacrifice to match precisely the region of interest. The area occupied by tortuous, non-uniform, unusual meshwork of vessels on the confocal image projection stack which cross-referenced positively with regions showing leakage on fluorescein angiogram was measured. The accumulated area occupied by these leaky vessels from each retina was compared with the area occupied by all vasculature in the region and expressed as a percentage. Typically, the posterior globe region within 1 mm radius of the optic disc imaged by fluorescein angiography was included in the measurement.
2.7. Immunohistochemistry
794
P.K. Yu et al. / Experimental Eye Research 89 (2009) 791–800
Fig. 1. Fluorescein angiogram. Fluorescein angiogram from the untreated (a, c, e, g) and laser treated (b, d, f, h) eye of a transgenic mouse taken on (a and b) day 32, before laser treatment; day 41 (c and d, 1 week after laser treatment), day 62 (e and f, 1 month after laser treatment) and day 76 (g and h, 6 weeks after laser treatment). Bright spots indicate locations of sodium fluorescein leakage.
P.K. Yu et al. / Experimental Eye Research 89 (2009) 791–800 Table 1 The average number of leaky spots SE as observed on fluorescein angiogram expressed as percentage of leaky spots before laser treatment. Time points
Untreated eye
P32 P41 P62 P76
100 125.9 18.30 75.6 8.27 58.5 10.56
795
3. Results 3.1. Evidence of reduction in retinal neovascularisation and leakage
Treated eye (21) (21) (6) (6)
100 22.9 4.83 33.01 15.02 20.8 14.26
(21) (21) (6) (6)
2.12. Statistical analysis All statistical testing was performed using the statistics program SigmaStat (SSPS Scientific; Chicago, IL). Student’s t test with an acceptance level of p < 0.05 was employed to determine any significant differences in retinal neovascularisation and leakage between the treated and untreated eyes and for the photoreceptor nuclear count measurements. All mean data is expressed as mean standard error.
3.1.1. Fundus fluorescein angiography Retinal neovascularisation and leakage was assessed in vivo using fundus fluorescein angiography. Fig. 1 shows fluorescein angiography images taken 2 days before laser treatment (P32), and one week (P41), 4 weeks (P62) and 6 weeks (P76) after laser treatment. The left eye was treated with laser and the right eye was left untreated. At P32, there were some leaky retinal vessels in both eyes as evidenced by fluorescein dye leakage in the late phase angiography. At P41 in this animal the fluorescein leakage appeared more severe in the untreated eye but the treated eye showed a reduction in fluorescein leakage. At P62 and P76 there were signs of reduced leakage in the untreated eye compared to that at P41. In the treated eye there was remarkably little evidence of fluorescein leakage when compared to the pre-treatment measurement at P32. Averaging such measurements across all animals at P32, the average number of
Fig. 2. A quadrant of projected images from flat mount of FITC-Dextran infused retina from mice sacrificed on day 41 (a and b, 1 week after laser treatment) and day 76 (c and d, 6 weeks after laser treatment). The corresponding untreated control retinas are shown in a and c. This P41 untreated specimen (a) clearly shows numerous round vascular structures (white arrow heads) distributed through the flat mount. By comparison, the contralateral eye that had been laser treated (b) had very few structures of that type. At 6 weeks postlaser treatment, unusual vascular structures may be seen in the untreated retina (c) and in the peripheral retina of the laser treated eye (d). However, these are denser in the untreated control retina (c) than the laser treated retina (d), especially comparing only the central 1 mm radius. These abnormal vascular structures appear to have a bigger diameter at P76 than at P41.
796
P.K. Yu et al. / Experimental Eye Research 89 (2009) 791–800
fluorescein leaky spots in the right eyes (18.1 2.8, n ¼ 21) was not significantly different (p ¼ 0.32) with that in the left eyes (23.3 4.5, n ¼ 21). Table 1 shows the averaged number of leaky spots as observed on fluorescein angiography at P41, P62 and P76 normalised to that at P32. In the untreated eyes there was no statistically significant difference at P41 (n ¼ 21, p ¼ 0.61), however, the number of leaky spots in the untreated eyes was significantly reduced from P32 levels to 75.6 8.3% at P62 and 58.5 10.6% at P76 (all n ¼ 6, p < 0.05). In the treated eyes the number of leaky spots was significantly less than pre-treatment levels at all time points (22.9 4.8% at P41, 33.0 15.0% at P62 and 27.6 14.3% at P76 (all p < 0.05)). It was also clearly evident that there was significantly less fluorescein leakage spots in the laser treated eye compared to the untreated eye at the same time point after laser treatment (all p < 0.05). 3.1.2. Flat mounted preparation of FITC-dextran infused retinas To assess the effects of low power laser treatment on the retinal neovascularisation histologically, flat mounted preparations of FITC-dextran infused retina were studied using confocal microscopy. With higher resolution and less aberration from overlying planes, retinal neovascularisation can clearly be identified. Fig. 2 shows a quadrant of projected images of retinas from each eye of FITC-dextran infused mice at P41 and P76. There were some clear differences between the treated and the untreated retina. In the untreated retina at P41, there were numerous new vessels distributed through the retina, while by comparison, the treated eye had relatively few structures of this type. Similar findings were also evident at P76 in which laser treated retina appeared to have less new vessels when compared with untreated retina. Quantitative data of neovascularisation from these flat mount studies were obtained in 6 animals at P41 and P76. Percentage area of the retina occupied by neovascularisation in the treated eyes was 4.65 1.65% at P41 and 5.80 1.82% at P76 which was significantly less than that in the untreated eyes 10.04 2.04% at P41 and 12.01 4.52% at P76 (p < 0.05). These data therefore support the results from the in vivo fluorescein angiography. 3.2. Retinal function Fig. 3 shows representative raw data of ERG measurements from untreated and laser treated eyes using different levels of flash intensity at different time points (P32, 41, 62 and 76). ERG responses were present and comparable through different time points between the treated and untreated eye. Fig. 4 shows a plot of the averaged a- and b-wave amplitudes. Prior to laser treatment, the ERGs from the left and right eyes were comparable (Fig. 4a). One week after laser treatment, however, there was a slight but significant (p < 0.05) drop in b-wave amplitude in the treated eye at stimulus levels of 2.25 and 5 cd s/m2 (Fig. 4b). One month after laser treatment, this drop in b-wave amplitude became insignificant (Fig. 4c). By six weeks, there was no difference seen in a-wave or b-wave amplitudes from the control and the treated eyes (Fig. 4d). In the longer term, the laser treatment did not appear to significantly impair visual function as indicated by ERG measurements. This is in contrast to the significant suppression of ERG amplitude with conventional PRP therapy (Capoferri et al., 1990). 3.3. Partial loss of the photoreceptors after low power laser treatments Representative images of retinal transverse sections in treated and untreated eyes at P41 and P76 are shown in Fig. 5. The layered structure within retina was well preserved in most regions, and there were some vessels in the outer nuclear layers indicating the presence of the retinal neovascularisation in both treated and
Fig. 3. Representative scotopic ERG recordings from untreated (a) and laser treated (b) eyes at different time points (P32, 41, 62 and 74 corresponding to columns from left to right). Numbers on the right indicate flash intensity level (cd s/m2). Scale bars on the bottom left hand corner equals to 0.2 mV (vertical) and 0.5 s (horizontal). ERG responses were comparable through different time points between the treated and untreated eye.
untreated groups. It was clearly evident that the number of new vessels varied with different regions, which is typical of this transgenic model. There was no obvious change in the inner retina, but there was thinning of the outer retina in the treated retina. Retinal pigmented epithelium showed evidence of laser effect including hypo-pigmentation, hyper-pigmentation, vacuolation and the presence of macrophages (Fig. 5b). With counting the nuclear number in the outer nuclear layer, percentage reduction in the number in the treated eyes was 18.9 4.88% at P41 and 11.5 4.09% at P76 when compared with the untreated eyes (all p < 0.05). 3.4. PEDF immunohistochemistry An increase in PEDF staining intensity was observed in the outer retina. The difference appears most marked at the border of RPE and photoreceptor outer segments. Intensity measurements (Table 2) taken at this region (Fig. 6) showed significant (p < 0.001) increase in intensity indicating increased PEDF protein presence in the treated retina. Significant increase in intensity was observed for both time points in the treated retina. Comparison in PEDF staining intensity between the untreated retinas at the two time points revealed a significant increase in intensity at 6 weeks post-treatment (Fig. 6) and may explain the reduction in leaky spots observed between 1 and 6-week post-treatment time points (Table 1). Comparison in PEDF staining intensity between the treated retinas at the two time points also revealed a significant increase in intensity at 6 weeks post-treatment (Table 2). 4. Discussion Panretinal photocoagulation is a common and well established therapy for a range of ischemic retinal diseases and diabetic retinopathy (Aiello et al., 1968; Branch retinal vein occlusion study
P.K. Yu et al. / Experimental Eye Research 89 (2009) 791–800
797
Fig. 4. Plots of averaged a- and b-wave amplitudes at a range of stimulus intensities at day 32 (a), 41 (b), 62 (c) and 76 (d) respectively. Standard errors are indicated and significant difference between treated and untreated eyes indicated with an asterisk.
group, 1984; Algvere et al., 1990; Early treatment diabetic retinopathy study research group, 1991; Arnarsson and Stefansson, 2000). The efficacy of such therapy is particularly well demonstrated in the treatment of proliferative diabetic retinopathy (Aiello et al., 1968; Early treatment diabetic retinopathy study research group, 1991). There may however be further room for improvement in terms of clinical outcomes. The degree of laser damage required to produce the therapeutic benefit is not known. One candidate for the ameliorative effects of laser therapy on vasoproliferation is the relief of retinal hypoxia due to destruction of the photoreceptors in the treated area (Stefansson et al., 1981, 1986; Landers et al., 1982; Pournaras et al., 1990; Zuckerman et al., 1993). This is then thought to lessen the production of the vasoproliferative factor VEGF due to the lessening of hypoxia-induced VEGF production (Wolbarsht and Landers, 1980; Boulton et al., 1998; Stefansson, 2001). It is not currently known how much reduction in retinal hypoxia is required or in which retinal layers such an effect is most critical. We have outlined a hypothesis that only a modest improvement in intraretinal oxygen levels may be required and that it is the middle retinal layers that may be most important. The supportive evidence comes from several sources. It is now well established that the retinas of patients with partial photoreceptor loss, such as in RP, are not susceptible to the proliferative effects of diabetic retinopathy (Arden, 2001). Although this may be of little comfort to the RP patient, it may point to an important factor in the pathogenesis of diabetic retinopathy. It had been suggested that partial loss of photoreceptors could result in increased oxygen levels in the
remaining retina due to a reduction in oxygen consumption of the depleted photoreceptor array (Stone et al., 1999). Recent work in animal models of RP failed to find a generalized increase in inner retinal oxygen levels following photoreceptor degeneration, but did demonstrate an increased oxygen flux into the middle retinal layers (Yu et al., 2000, 2004). This change in balance between the potential supply of oxygen from the choroidal and retinal vasculature to the middle retinal layers could provide an amelioration of ischemic insults to the retinal circulation. We were also able to show that laser therapy could be used to create similar oxygenation changes in otherwise normal animals (Yu et al., 2005b). To overcome the confounding effects of autoregulation of the retinal circulation we performed such measurements in the avascular region of the rabbit retina. We were able to quantify the reduction in outer retinal oxygen consumption for different powers, wavelengths, and modes of laser delivery. It was established that significant improvements of intraretinal oxygen level could be produced using laser powers that created very mild lesions as judged by their ophthalmoscopic appearance. The present study sought to investigate the effectiveness of low power laser therapy in a well established mouse model of retinal neovascularisation (Okamoto et al., 1997; Campochiaro and Hackett, 2003). The model is characterised by the early appearance of intraretinal neovascularisation (Ida et al., 2003). It is readily accepted that this is not a direct model of diabetic retinopathy, but it is comparable in that the vasoproliferation is driven by over-expression of VEGF. We therefore examined the ability of low power laser therapy to
Fig. 5. Epoxy semi-thin cross-sections from untreated (a and c) and treated (b and d) eyes at P41 (a and b) and P76 (c and d) stained with Toluidine Blue. The layered structure of the retina was retained in the treated retinae (b and d). However, some sparseness can be observed in the outer nuclear layer of the treated retinas suggesting a reduction in photoreceptor nuclei number. Changes such as hypo-pigmentation and vacuolation in the retinal pigmented epithelium (RPE) can be observed in the laser treated retina. A few pigment loaded macrophages are visible in the laser treated retina (arrow heads). Blood vessels may be seen within the outer nuclear layer (black arrows) in the treated and untreated retinas. Scale bar 100 mm.
ameliorate the vasoproliferative response in this model. This was confirmed by fluorescein fundus photography, by whole mount analysis of new vessel formation, and by increased presence of PEDF. In addition we sought to confirm that low power laser therapy had the potential for only partial destruction of the photoreceptor array, thus preserving some useful retinal function in the treated areas. The functional assessment was based on ERG amplitudes, which after an initial recovery period were not
Table 2 Intensity measurement (mean SE) from confocal images of PEDF immunostaining. (Intensity range from 0 to 4096). Statistically significant difference was found between all groups compared (p 0.001). Number in bracket indicates the number of measurements made for that group. Intensity
P41
P76
Untreated Treated
435.3 14.47 (727) 755.9 24.13 (685)
782.7 39.32 (429) 851.6 31.04 (451)
P.K. Yu et al. / Experimental Eye Research 89 (2009) 791–800
799
Fig. 6. PEDF immunohistochemistry. Confocal images taken from an untreated (left panel) and a treated (right panel) retina at P41 and P76 time points. PEDF has been labelled with an Alexa Fluor 488 secondary antibody and pseudo coloured green. The nuclei have been counterstained using bisBenzimide H and pseudo coloured blue. The treated retina showed stronger PEDF staining particularly at the border of the outer segments and retinal pigmented epithelium (RPE). The RPE layer has been enlarged for comparison purpose between the treatment group and different time points (Scale bar measures 20 mm). Untreated retina at P41 has weak PEDF labelling compared to the treated retina at the same time point. The untreated retina at P76 is much more intensely labelled than at P41. The treated retina at P76 is also more intensely labelled compared to all the other data points. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article).
significantly reduced in the treated eyes. Histological assessment of the treated retina showed only partial loss of the photoreceptor array. Whilst extrapolation of studies in mice to imply a given outcome in clinical studies should be done with extreme caution, it is comforting to note that in terms of intraretinal oxygen distribution and consumption these are remarkable similarities between measurements recently made in the mouse (Yu and Cringle, 2006) and the monkey (Yu et al., 2005a). Whilst intraretinal oxygen microelectrode work in the mouse is technically feasible, the use of such techniques in a model with such localized vascular dysfunction and localized changes in retinal structure is problematic. The more global measures of vascular leakage, vascular structure, and electrophysiological function provide more accurate markers of neovascularisation and retinal function. The present study cannot determine the relative importance of different mechanisms of the anti-proliferative effects of laser therapy in this model. Whilst an increase in intraretinal oxygen level is at the heart of our original hypothesis, the upregulation of PEDF may well be an important factor, as could the simple reduction of VEGF production due to partial loss of photoreceptors (the primary site of excess VEGF production in this model). Whatever the specific mechanisms involved it is comforting to note that the present study suggests that it may be feasible that low power laser therapy could have applications in the clinical management of a range of ischemic retinal diseases without the patient having to suffer complete loss of vision in the treated area. Acknowledgements The authors acknowledge the expert technical support of Mr Dean Darcey, Dr Joe Miller and Ms Zoya Gridneva. We also wish to acknowledge the provision of the mouse model by Professor Peter Campochiaro, Johns Hopkins University, USA. Grant support was provided by the National Health and Medical Research Council of Australia and the Australian Research Council Centre of Excellence in Vision Science. References Aiello, L.M., Beetham, W.P., Balodimos, M.C., Chazan, B.I., Bradley, R.F.,1968. Ruby laser photocoagulation in treatment of diabetic proliferating retinopathy. Preliminary Report. In: Goldberg, M.F., Fine, S.L. (Eds.), Symposium: Treatment of Diabetic Retinopathy. U.S. Government Printing Office, pp. 437–463 (Chapter 35).
Aiello, L.P., Avery, R.L., Arrigg, P.G., Keyt, B.A., Jampel, H.D., Shah, S.T., Pasquale, L.R., Thieme, H., Iwamoto, M.A., Park, J.E., Nguyen, H.V., Aiello, L.M., Ferrara, N., King, G.L., 1994. Vascular endothelial growth factor in ocular fluid of patients with diabetic retinopathy and other retinal disorders. N. Engl. J. Med. 331, 1480–1487. Algvere, P.V., Hallnas, K., Dafgard, E., Hoog, A., 1990. Panretinal photocoagulation aggravates experimental proliferative vitreoretinopathy. Graefes Arch. Clin. Exp. Ophthalmol. 228, 461–466. Arden, G.B., 2001. The absence of diabetic retinopathy in patients with retinitis pigmentosa: implications for pathophysiology and possible treatment. Br. J. Ophthalmol. 85, 366–370. Arnarsson, A., Stefansson, E., 2000. Laser treatment and the mechanism of edema reduction in branch retinal vein occlusion. Invest. Ophthalmol. Vis. Sci. 41, 877–879. Bandello, F., Brancato, R., Menchini, U., Virgili, G., Lanzetta, P., Ferrari, E., Incorvaia, C., 2001. Light panretinal photocoagulation (LPRP) versus classic panretinal photocoagulation (CPRP) in proliferative diabetic retinopathy. Semin. Ophthalmol. 16, 12–18. Boulton, M., Foreman, D., Williams, G., McLeod, D., 1998. VEGF localisation in diabetic retinopathy. Br. J. Ophthalmol. 82, 561–568. Branch retinal vein occlusion study group, 1984. Argon laser photocoagulation for macular oedema in branch vein occlusion. Am. J. Ophthalmol. 98, 271–282. Campochiaro, P.A., Hackett, S.F., 2003. Ocular neovascularization: a valuable model system. Oncogene 22, 6537–6548. Capoferri, C., Baqini, M., Chizzoli, A., Pece, A., Brancato, R., 1990. Electroretinographic findings in panretinal photocoagulation for diabetic retinopathy. A randomized study with blue-green argon and red krypton lasers. Graefes Arch. Clin. Exp. Ophthalmol. 228, 232–236. D’Amato, R., Wesolowski, E., Hodgson Smith, L.E., 1993. Microscopic visualization of the retina by angiography with high-molecular-weight fluorescein-labeled dextrans in the mouse. Microvasc. Res. 46, 135–142. Dalke, C., Loster, J., Fuchs, H., Gailus-Durner, V., Soewarto, D., Favor, J., NeuhauserKlaus, A., Pretsch, W., Gekeler, F., Shinoda, K., Zrenner, E., Meitinger, T., Hrabe de Angelis, M., Graw, J., 2004. Electroretinography as a screening method for mutations causing retinal dysfunction in mice. Invest. Ophthalmol. Vis. Sci. 45, 601–609. Early treatment diabetic retinopathy study research group, 1991. Early photocoagulation for diabetic retinopathy. ETDRS report number 9. Ophthalmology 98, 766–785. Fong, D.S., Strauber, S.F., Aiello, L.P., Beck, R.W., Callanan, D.G., Danis, R.P., Davis, M.D., Feman, S.S., Ferris, F., Friedman, S.M., Garcia, C.A., Glassman, A.R., Han, D.P., Le, D., Kollman, C., Lauer, A.K., Recchia, F.M., Solomon, S.D., 2007. Comparison of the modified Early Treatment Diabetic Retinopathy Study and mild macular grid laser photocoagulation strategies for diabetic macular edema. Arch. Ophthalmol. 125, 469–480. Hattenbach, L.O., Beck, K.F., Pfeilschifter, J., Koch, F., Ohrloff, C., Schacke, W., 2005. Pigment-epithelium-derived factor is upregulated in photocoagulated human retinal pigment epithelial cells. Ophthalmic Res. 37, 341–346. Ida, H., Tobe, T., Nambu, H., Matsumura, M., Uyama, M., Campochiaro, P.A., 2003. RPE cells modulate subretinal neovascularization, but do not cause regression in mice with sustained expression of VEGF. Invest. Ophthalmol. Vis. Sci. 44, 5430–5437. Landers, M.B., Stefansson, E., Wolbarsht, M.L., 1982. Panretinal photocoagulation and retinal oxygenation. Retina 2, 167–175. Mori, K., Gehlbach, P., Ando, A., McVey, D., Wei, L., Campochiaro, P.A., 2002. Regression of ocular neovascularization in response to increased expression of
800
P.K. Yu et al. / Experimental Eye Research 89 (2009) 791–800
pigment epithelium-derived factor. Invest. Ophthalmol. Vis. Sci. 2002, 2428–2434. Ogata, N., Wada, M., Otsuji, T., Jo, N., Tombran-Tink, J., Matsumura, M., 2002. Expression of pigment epithelium-derived factor in normal adult rat eye and experimental choroidal neovascularization. Invest. Ophthalmol. Vis. Sci. 43, 1168–1175. Ogata, N., Wada, M., Otsuji, T., Jo, N., Tombran-Tink, J., Matsumura, M., 2001. Pigment epithelium derived factor in the vitreous is low in diabetic retinopathy and high in rhegmatogenous retinal detachment. Am. J. Ophthalmol. 132, 378–382. Okamoto, N., Tobe, T., Hackett, S.F., Ozaki, H., Vinores, M.A., LaRochelle, W., Zack, D.J., Campochiaro, P.A., 1997. Transgenic mice with increased expression of vascular endothelial growth factor in the retina. Am. J. Pathol. 151, 281–291. Ozaki, H., Yu, A.Y., Della, N., Ozaki, K., Luna, J.D., Yamada, H., Hackett, S.F., Okamoto, N., Zack, D.J., Semenza, G.L., Campochiaro, P.A., 1999. Hypoxia inducible factor-1alpha is increased in ischemic retina: temporal and spatial correlation with VEGF expression. Invest. Ophthalmol. Vis. Sci. 40, 182–189. Pahor, D., 1998. Visual field loss after argon laser panretinal photocoagulation in diabetic retinopathy: full-versus mild-scatter coagulation. Int. Ophthalmol. 22, 313–319. Pournaras, C.J., Tsacopoulos, M., Strommer, K., Gilodi, N., Leuenberger, P.M., 1990. Scatter photocoagulation restores tissue hypoxia in experimental vasoproliferative microangiopathy in miniature pigs. Ophthalmology 97, 1329–1333. Seiberth, V., Alexandridis, E., 1991. Function of the diabetic retina after panretinal argon laser photocoagulation. Influence of the intensity of the coagulation spots. Ophthalmologica 202, 10–17. Stefansson, E., 2001. The therapeutic effects of retinal laser treatment and vitrectomy. A theory based on oxygen and vascular physiology. Acta Ophthalmol. Scand. 79, 435–440. Stefansson, E., 2006. Ocular oxygenation and the treatment of diabetic retinopathy. Surv. of Ophthalmol 51, 364–380. Stefansson, E., Hatchell, D.L., Fisher, B.L., Sutherland, F.S., Machemer, R., 1986. Panretinal photocoagulation and retinal oxygenation in normal and diabetic cats. Am. J. Ophthalmol. 101, 657–664. Stefansson, E., Landers, M.B., Wolbarsht, M.L., 1981. Increased retinal oxygen supply following pan-retinal photocoagulation and vitrectomy and lensectomy. Trans. Am. Ophthalmol. Soc. LXXIX, 307–334.
Stone, J., Maslim, J., Valter-Kocsi, K., Mervin, K., Bowers, F., Chu, Y., Barnett, N., Provis, J., Lewis, G., Fisher, S.K., Bisti, S., Gargini, C., Cervetto, L., Merin, S., Pe’er, J., 1999. Mechanisms of photoreceptor death and survival in mammalian retina. Prog. Retin. Eye Res. 18, 689–735. Takagi, H., King, G.L., Ferrara, N., Aiello, L.P., 1996. Hypoxia regulates vascular endothelial growth factor receptor KDR/Flk gene expression through adenosine A2 receptors in retinal capillary endothelial cells. Invest. Ophthalmol. Vis. Sci. 37, 1311–1321. The Diabetic Retinopathy Study Research Group, 1987. Indications for photocoagulation treatment of diabetic retinopathy: diabetic Retinopathy Study Report no. 14. Int. Ophthalmol. Clin. 27, 239–253. Wolbarsht, M.L., Landers III, M.B., 1980. The rationale of photocoagulation therapy for proliferative diabetic retinopathy: a review and a model. Ophthalmic Surg. 11, 235–245. Yu, D.-Y., Cringle, S.J., 2004. Low oxygen consumption in the inner retina of the visual streak of the rabbit. Am. J. Physiol. (Heart C) 286, H419–H423. Yu, D.-Y., Cringle, S.J., 2006. Oxygen distribution in the mouse retina. Invest. Ophthalmol. Vis. Sci. 47, 1109–1112. Yu, D.-Y., Cringle, S.J., Alder, V.A., Su, E.N., 1994. Intraretinal oxygen distribution in rats as a function of systemic blood pressure. Am. J. Physiol. 36, H2498–H2507. Yu, D.-Y., Cringle, S.J., Su, E.-N., 2005a. Intraretinal oxygen distribution in the monkey retina and the response to systemic hyperoxia. Invest. Ophthalmol. Vis. Sci., 4728–4733. Yu, D.-Y., Cringle, S.J., Su, E.-N., Yu, P.K., Humayun, M.S., Dorin, G., 2005b. Laser induced changes in intraretinal oxygen distribution in pigmented rabbits. Invest. Ophthalmol. Vis. Sci. 46, 988–999. Yu, D.-Y., Cringle, S.J., Su, E.N., Yu, P.K., 2000. Intraretinal oxygen levels before and after photoreceptor loss in the RCS rat. Invest. Ophthalmol. Vis. Sci. 41, 3999–4006. Yu, D.-Y., Cringle, S.J., Valter, K., Walsh, N., Lee, D., Stone, J., 2004. Photoreceptor death, trophic factor expression, retinal oxygen status and photoreceptor function in the P23H rat: stress and protection in a slow degeneration. Invest. Ophthalmol. Vis. Sci. 45, 2013–2049. Zuckerman, R., Cheasty, J.E., Wang, Y., 1993. Optical mapping of inner retinal tissue PO2. Curr. Eye Res. 12, 809–825.