Topical Administration of Somatostatin Prevents Retinal Neurodegeneration in Experimental Diabetes

A total of 24 male Sprague-Dawley rats were obtained from Charles River Laboratories (Senneville, Quebec, Canada). A minimum acclimatization period of 14 days preceded treatment. STZ (60 mg/kg) was administered to 8-week-old (n = 16) rats by intravenous injection via tail vein on day −2 prior to starting SST treatment. The remaining eight rats were injected with vehicle and served as a nondiabetic control group. Animals with blood glucose levels >250 mg/dL the day after STZ administration were considered diabetic. On day −1, all animals were weighed and randomly assigned to treatment groups (vehicle or SST) using a computer-based randomization procedure. Randomization was performed by stratification using body weight as the parameter.

SST or vehicle eye drops were administered directly onto the superior corneal surface of each eye using a syringe. One drop (20 µL) of SST (1%; 10 mg/mL) or vehicle (20 µL of 0.9% sodium chloride) was administered once daily for 14 days. On day 15, the animals’ eyes were instilled with a drop of SST or vehicle ~1 h prior to necropsy. Therefore, the elapsed time between the onset of diabetes mellitus until necropsy was 16 days.

A dose-response experiment assessing the hallmarks of retinal neurodegeneration (apoptosis and glial activation) was performed using SST eye drops at 10, 2, and 0.5 mg/mL (Table 1).

The animals used in this study were treated in accordance with the Association for Research in Vision and Ophthalmology Statement.

A pharmacokinetic analysis after a single topical dose of radiolabeled ¹²⁵I-somatostatin was performed. For this purpose, seven rats were anesthetized and placed in a supine position. After they were placed in the experimental area, 10 µL (2.5 µCi) ¹²⁵I-somatostatin was applied in each eye with a micropipette. The animals were killed at 15, 30, and 60 min and 4, 6, 12 and 24 h after the application of the labeled SST solution. After the animals were killed, the eyes were obtained and gently washed in a volume of 20 mL physiologic saline. A small incision was performed in the rear portion of the eye ball, and the eye content was extracted. The vitreous humor, retina, and choroid-RPE were collected. The sclera was placed in different tubes for independent measurements. All the samples were weighted and counted in the γ counter in order to obtain the percentage of injected dose for each gram of tissue. Samples of blood were also obtained.

In addition, absorption of carboxyfluorecein-labeled SST (CF-SST) eye drops through the ocular structures was also evaluated in 8-week-old Sprague-Dawley rats. For this purpose, three rats were treated with vehicle eye drops in both eyes. Three rats were treated with the fluorescent label (CF) eye drops. The eye drops of CF-SST (10 mg/mL; one drop) were administered to the right eye, and the vehicle was administered to the left eye of three rats. One animal of each group was killed at 10, 30, and 60 min after the application of the corresponding eye drop. The eyes were enucleated, fixed in 4% paraformaldehyde, submerged in optimum cutting temperature Tissue-teck, frozen in liquid nitrogen, and stored at −80°C until microtome sectioned. The whole eyes were cryosectioned to 50-μm sections, mounted with antifade-mounting medium with DAPI, and visualized in fluorescent and confocal microscopes (Olympus). All images were taken with the same settings.

Slit lamp biomicroscopy, indirect ophthalmoscopy, and intraocular pressure measurement were performed in all animals at baseline and at days 7 and 13 posttreatment.

Electroretinogram evaluations were performed on day –1 before treatment and on day 14 as follows: the animals were dark adapted for at least 1 h prior to electroretinography (ERG) recording and then anesthetized with isoflurane to maintain the anesthesia prior to and during the procedure. Tropicamide (1%) was applied to each eye prior to the test. A contact lens electrode was placed on the surface of each eye, and a needle electrode was placed cutaneously on the head between the two eyes. A cutaneous ground electrode was placed near the base of the tail. Carboxymethylcellulose (1%) drops were applied to the interior surface of the contact lens electrodes prior to their placement on the eyes. A topical anesthetic (proparacaine, 0.5%) was applied to the eyes. The amplitude and implicit time of the b-wave after scotopic flash stimuli were measured. b-wave amplitude and implicit time were measured as defined by the International Society for Clinical Electrophysiology of Vision (28).

Eyes were prepared for histopathological examination by being embedded in paraffin wax, sectioned, and stained with hematoxylin-eosin. For each eye, a total of three sagittal sections (one section through the optic nerve, one section ~500 µm medially, and another section ~500 µm laterally from the optic nerve) were evaluated.

Sections of 7-μm thicknesses were obtained from the paraffin blocks and fixed on highly adherent slides (Visionbiosystems; Newcastle upon Tyne, U.K.). They were desparaffined in xylol; rehydrated in 100% ethanol, 90% ethanol, and 75% ethanol; and incubated for 10 min in PBS. Unspecific unions were blocked by incubating the samples for 1 h in PBS 1% BSA and 0.05% Triton X-100. Thereafter, the primary rabbit anti-human glial fibrillar acidic protein (GFAP) antibody (Sigma, Madrid, Spain) diluted in the blocking buffer (1:100) was incubated for 36 h at 4°C. After three washings for 5 min with PBS, the sections were incubated with a secondary human IgG antibody labeled with Alexa Fluor 568 (Invitrogen, Eugene, OR) for 1 h at room temperature. The labeled sections were washed and mounted with fluorescent medium containing DAPI for identifying cell nuclei (Vector Laboratories, Burlingame, CA). Positive sections for GFAP were captured in a confocal microscope (FV1000; Olympus, Hamburg, Germany). Optical sections were obtained with a 568-nm laser for Alexa Fluor 568 and a 405-nm laser for the DAPI (image resolution 1,024 × 1,024 pixels). The degree of glial activation was evaluated by using a score system based on the extent of GFAP staining (29). This scoring system is as follows: negligible staining (score 0), Müller cell endfeet region/ganglion cell layer (GCL) only (score 1), Müller cell endfeet region/GCL plus a few proximal processes (score 2), Müller cell endfeet plus many processes but not extending to outer nuclear layer (ONL) (score 3), Müller cell endfeet plus processes throughout with some in the ONL (score 4), and Müller cell endfeet plus a lot of dark processes from GCL to the outer margin of ONL (score 5).

The In Situ Cell Death Detection kit (Roche Diagnostics, Mannheim, Germany) was used to evaluate the presence of apoptotic cells. Sections were deparaffined in xylol; rehydrated in 100% ethanol, 90% ethanol, and 75% ethanol; and incubated for 10 min in PBS. Autofluorescence was eliminated after incubation with 0.2% potassium permanganate for 20 min and 1% oxalic acid for 30 s before the TUNEL procedure was started. Positive and negative controls were processed at the same time. Three confocal images were recorded (×40) corresponding to a surface of 317.13 × 317.13 μm for each section. Staining with propidium iodide was done to examine nucleus morphology and discard false positives. The total number of nuclei and nuclei positive for TUNEL were counted with the ImageJ program.

Proapoptotic signaling pathways were investigated by analyzing several proapoptotic molecules (FasL, caspase-8, total Bid, truncated Bid, and active caspase-3), as well as antiapoptotic markers (BclxL), by Western blot as previously described (30).

Quantification of glutamate was performed by liquid chromatography coupled to mass spectrometry. Chromatographic separation was performed on an Agilent 1200 series (Waldbronn, Germany) using an Ascentis Express HILIC column, 50 × 2.1 mm with 2.7 µm particle size from Supelco (Belfonte, PA), maintained at 25°C throughout the analysis, a mobile phase acetonitrile (ACN), and water (50 mmol/L ammonium acetate) with a flow rate of 0.6 mL/min. The volume injected was 10 µL. The mobile phase involved a gradient starting at 87% of ACN, which was maintained for 3 min. Then, from min 3 to 10 the ACN content was decreased to 20% and increased again to 87% at min 12.5. Glutamate was eluted at 6.5 min. The mass detection system was an Agilent 6410 Triple Quad (Santa Clara, CA) using positive electrospray ionization with a gas temperature of 350°C, gas flow rate of 12 L/min, nebulizer pressure of 45 ψ, capillary voltage of 3,500 V, fragmentor of 135 V, and collision energy of 10 V.

Total RNA was extracted from retinas with the Rneasy Mini kit with DNAase digestion (Qiagen; IZASA, Madrid, Spain). One microgram of total RNA was used for reverse transcription with TaqMan and random primer reagents (Applied Biosystems, Madrid, Spain). Real-time PCR was performed with the GLAST specific assay for exon 3–4 boundary Rn00570130_m1 (Applied Biosystems), and results were normalized to the expression level of β-actin (Rn00667969_m1) as a housekeeping gene. Relative quantification values were obtained using an ABI Prism 7000 SDS software (Applied Biosystems).

Sections of 4 μm were deparaffined in xylene and hydrated in a graded ethanol series. Then, sections were antigen retrieved (sodium citrate 10 mmol/L, pH 6.0) and washed in PBS. Next, sections were incubated in blocking solution (2% BSA, Tween 0.05% PBS) for 1 h at room temperature followed by incubation with primary antibody anti-GLAST (1:200; Abcam, Cambridge, U.K.). After washing, sections were incubated with a fluorescent anti-rabbit Alexa Fluor 488 as a secondary antibody (Life Technologies, Madrid, Spain) in blocking solution for 1 h, washed, and mounted in Vectashield (Vector Laboratories). DAPI was used for nuclear staining. Quantification of fluorescence intensity of images was performed as previously explained for GFAP.

Normal distribution of the variables was evaluated using the Kolmogorov-Smirnov test. Data were expressed either as means ± SD or as median (range). Comparisons of continuous variables between groups were performed using the unpaired Student t test or Mann-Whitney U test. For comparisons within groups (ERG of day 14 vs. ERG of day −1), the paired Student t test was used. Levels of statistical significance were set at P < 0.05.

The results of pharmacokinetic study are displayed in Fig. 1. After topical ocular administration of ¹²⁵I-somatostatin, the maximum postdose concentration (5% of applied dose per gram of tissue) was detected at 1 h. However, at 15 min most of the ¹²⁵I-somatostatin had already reached the retina. This rapid absorption argues against the permeation of SST through the cornea. For further exploration of this issue, a pharmacokinetic study using eye drops containing CF-SST was performed. This study clearly showed that SST contained in the eye drops reached the retina not through the cornea but by the trans-scleral route, thus bypassing the anterior chamber (Supplementary Fig. 1).

No significant differences in the cornea, pupil size, or intraocular pressure were observed among the three groups at baseline or during follow-up.

ERG was performed to assess retinal function. b-wave amplitude was significantly reduced at day 14 in the diabetic group treated with vehicle (Fig. 2A). By contrast, in the control group and in the diabetic group treated with SST eye drops no reduction of b-wave amplitude was observed. b-wave implicit time increased at day 14 in diabetic rats treated with vehicle (Fig. 2B). However, in the control group and in the diabetic group treated with SST eye drops the implicit time did not increase at day 14. Therefore, SST treatment prevented ERG abnormalities caused by diabetes mellitus.

The retinas of nondiabetic control rats revealed GFAP immunolabeling in cells and their processes in the inner layer of the retina, especially in the inner limiting membrane, and GCL (GFAP score 0–1) (Fig. 3). According to their location and morphology, these GFAP-positive cells were interpreted as being retinal astrocytes. However, a slight GFAP immunofluorescence in Müller cells could be detected. In the retina of diabetic rats treated with placebo, GFAP expression was prominent along the inner limiting membrane, in Müller cell endfeet, and in Müller cell radial fibers extending through both the inner and outer retina (GFAP score 1–3). Diabetic rats treated with SST eye drops presented significantly lower GFAP score (0–1) than diabetic rats treated with placebo (Fig. 3).

Representative images of the presence of apoptotic cells in the retinas of the three experimental groups are shown in Fig. 4A. Two weeks after the induction of diabetes mellitus with STZ, the whole percentage of retinal apoptotic cells, as well as the percentage of apoptotic cells in retinal layers, was significantly higher in comparison with that observed in retinas from age-matched nondiabetic controls (Fig. 4B). In all groups, apoptosis was highest in the GCL. Diabetic rats treated with SST eye drops presented a significantly lower ratio of apoptosis in all retinal layers than diabetic rats treated with placebo and similar to nondiabetic rats (Fig. 4B).

The effectiveness of SST eye drops at concentrations of 2 mg/mL was similar to that of eye drops at 10 mg/mL concentrations in preventing glial activation and apoptosis. However, the effect of 0.5 mg/mL was significantly lower than that obtained using either 2 or 10 mg/mL (Fig. 5).

The results of proapoptotic and survival molecules measured in the retina of the three rat groups studied are shown in Fig. 6. The expression of FasL, a proapoptotic component of the death receptor apoptotic pathway, was significantly increased in neuroretinas from diabetic rats compared with control rats. The interaction of FasL with the death receptor Fas/CD95 assembles the death-inducing signaling complex (31). This includes the recruitment of caspase-8. Activation of caspase-8, monitored by the presence of its 18-kDa active fragment, was significantly increased in neuroretinas of diabetic rats compared with controls. These data were reinforced by the detection of elevated levels of truncated Bid fragment in diabetic samples. As a result, neuroretina from diabetic rats displayed a significant activation of the executer caspase-3, monitored by the presence of its 17-kDa active fragment. Treatment with SST eye drops prevented the upregulation of all of these proapoptotic molecules.

On the other hand, a reduction in BclxL, an antiapoptotic member of the Bcl2 family, was found in diabetic retinas in comparison with control retinas. When treated with SST eye drops, BclxL expression was greatly improved (Fig. 6).

Glutamate levels (arbitrary units, median [range]) in the diabetic retinas were higher than in nondiabetic retinas (117 [57–412] vs. 47 [27–300]; P = 0.012). In diabetic rats treated with SST eye drops, glutamate concentration (40 [21–127]) was significantly decreased in comparison with diabetic rats treated with vehicle (P = 0.021) and similar to control rats (P = 0.51).

Gene and protein expression of GLAST were downregulated in retinas of diabetic rats (Fig. 7). In diabetic rats treated with SST, this downregulation was prevented and GLAST protein levels were even increased in comparison with nondiabetic rats (Fig. 7).

In the current study, we provide the first evidence that topical administration of SST prevents retinal neurodegeneration in STZ-DM rats. SST eye drops prevented b-wave abnormalities in the ERG (reduction in amplitude and increase in implicit time), which are considered sensitive indicators of DR (32). In addition, topical administration of SST abrogated the characteristic hallmarks of neurodegeneration (glial activation and apoptosis) caused by short-term diabetes mellitus. In this regard, it should be noted that the degree of glial activation and apoptosis in diabetic rats treated with SST eye drops was very similar to that found in nondiabetic rats.

Glial activation is a prominent event in retinal neurodegeneration. Retinal astrocytes normally express GFAP, while Müller cells do not. However, during the course of diabetes mellitus there is an aberrant expression of GFAP by Müller cells (33). Because Müller cells produce factors capable of modulating blood flow, vascular permeability, and cell survival and because their processes surround all the blood vessels in the retina, it seems that these cells play a key role in the pathogenesis of retinal microangiopathy in the diabetic eye (33,34). In this paper, we have found that SST eye drops were able to significantly prevent glial activation or, in other words, most of the aberrant expression of GFAP in Müller cells.

Apart from glial activation, apoptosis is the other hallmark of retinal neurodegeneration. As previously reported (35), we found a significant rate of apoptosis in retinas from STZ-induced diabetic rats in comparison with nondiabetic control rats. In addition, we found a misbalance between apoptotic and survival signaling pathways. In this regard, we demonstrated that the activation of the death receptor pathway, monitored by the elevation in the expression of FasL and the activation of caspase-8, was significantly higher in neuroretina from STZ-induced diabetic rats than in nondiabetic controls. Importantly, our data clearly show increased Bid cleavage in the neuroretina of diabetic rats. This result suggests that in the diabetic neuroretina, the apoptotic signals emerging from the death receptor pathway are amplified to additional activation of the intrinsic apoptotic signaling pathway, resulting in a stronger activation of the executer caspase-3. On the other hand, a significant downregulation of BclxL (an antiapoptotic molecule belonging to the Bcl2 family) was found in diabetic retinas. Notably, topical administration of SST was able to prevent all these abnormalities in proapoptotic/survival signaling.

Apart from the prevention of apoptosis and glial activation, SST eye drops significantly reduced glutamate retinal levels. Rapid removal or inactivation of glutamate is necessary to maintain the normal function of the retina and for avoiding excitotoxicity. Elevated levels of glutamate in the retina have been found in experimental models of diabetes mellitus (36–38), as well as in the vitreous fluid of diabetic patients (12,39). The reasons why diabetes mellitus facilitates extracellular accumulation of glutamate include 1) an increase of glutamate production by glial cells due to the loss of the Müller cell–specific enzyme glutamine synthetase, which converts glutamate to glutamine (36,37); 2) a reduction in the retinal ability to oxidize glutamate to α-ketoglutarate (37); 3) the death of neurons with subsequent release of intracellular contents; and 4) the impairment of glutamate removal from extracellular space by the glial cells (12). Regarding this later point, in the current study we provide evidence that GLAST, the main glutamate transporter expressed by Müller cells (40), is significantly decreased in STZ-induced diabetic rats. More importantly, SST eye drops not only completely prevent GLAST downregulation induced by diabetes mellitus but even lead to a significant increase in GLAST expression (mRNA and protein). These findings can contribute to our understanding of the underlying mechanisms involved in the retinal reduction of AMPA-induced neurotoxicity reported for SST and SST analogs administered intravitreally (25).

The study of the mechanisms linking neurodegeneration with early vascular abnormalities is crucial for understanding how neuroprotection could also exert beneficial effects in diabetic microangiopathy. SST may inhibit angiogenesis directly through SST receptors present on endothelial cells (41) and also indirectly through the downregulation of vascular endothelial growth factor (VEGF) (42,43) or by the inhibition of postreceptor signaling events not only of VEGF but also of other peptide growth factors such as IGF-1, epidermal growth factor, and platelet-derived growth factor (44). The potential effects of SST in abrogating microvascular impairment are beyond the scope of the current study, and specific studies on this issue are urgently needed.

Given the essential role of neurodegeneration in the pathogenesis of DR, it is reasonable to hypothesize that therapeutic strategies based on the neuroprotective effects of SST would be effective in preventing or arresting DR development. The observation that circulating growth hormone–IGF-1 production is reduced by SST has been the basis for proposing systemic administration of SST analogs for treating DR. However, this concept has not been supported by clinical intervention trials. The main concerns of systemic administration of SST are as follows: 1) the paracrine effects of SST synthesized by the retina involve SSTR subtypes other than SSTR2, which mediates growth hormone inhibition and for which octreotide presents high affinity, and 2) SST analogs do not cross the blood-retinal barrier and would have access only where there is disruption of the blood-retinal barrier, thus limiting the amount of the drug reaching the retinal target tissues. For these reasons and given that a downregulation of retinal production of SST occurs in diabetic retina, a replacement using local administration of the natural peptide seems a reasonable approach. In this regard, the neuroprotective effects of topical administration of brimonidine and nerve growth factor have already been reported in experimental models (45–47). These findings open up the possibility of developing a topical therapy targeting the early stages of DR in which the use of the only currently established therapies such as laser photocoagulation or intravitreal injections of corticosteroids or anti-VEGF agents are inappropriately invasive. In addition, topical administration of drugs limits their action to the eye and minimizes the associated systemic effects, resulting in higher patient compliance (48). Therefore, topical therapies could revolutionize the care of diabetic patients (49). In this regard, a phase II-III, randomized controlled clinical trial (EUROCONDOR-278040) to assess the efficacy of SST and brimonidine administered topically to prevent or arrest DR has been approved by the European Commission in the setting of the FP7-HEALTH-2011.

Finally, the route by which SST contained in the eye drops reaches the retina deserves a brief comment. The pharmacokinetic study revealed a rapid absorption of ¹²⁵I-SST, which was unexpected for a compound that reached the retina diffusing through the cornea. The study performed using CF-SST showed that the absorption of SST eye drops was mainly by the trans-scleral route.

In summary, we provide first evidence that topical administration of SST has a potent effect in preventing the retinal neurodegenerative process that occurs in the early stages of DR. A preventive effect on the impairment of survival/apoptotic signaling induced by diabetes mellitus and a significant reduction in glutamate-induced excitotoxicity are among the mechanisms by which SST exerts its beneficial actions.

This study was supported by grants from the Ministerio de Ciencia e Innovación (SAF2012-35562 and SAF2009-08114), from the European Foundation for the Study of Diabetes, and from the 7th Framework Programme (EUROCONDOR, FP7-278040).

No potential conflicts of interest relevant to this article were reported.

C.H. designed the project, led the analysis, wrote the manuscript, reviewed and edited the manuscript, and approved the final version of the manuscript. M.G.-R. and L.C. led the analysis, reviewed the manuscript, and approved the final version of the manuscript. J.F.-C., J.F.-S., and B.P. contributed to discussion, reviewed the manuscript, and approved the final version of the manuscript. A.G.-R. led the analysis, reviewed the manuscript, and approved the final version of the manuscript. A.M.V. contributed to discussion, reviewed the manuscript, and approved the final version of the manuscript. R.S. designed and coordinated the project, wrote the manuscript, reviewed and edited the manuscript, and approved the final version of the manuscript. C.H. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

The authors thank Dr. O. Yanes (CIBERDEM) for glutamate quantification and Dr. A.R. Carvalho (Ophthalmologic Research Unit, Vall d’ Hebron Research Institute) for her technical assistance in obtaining the images of pharmacokinetic analysis.