Introduction

Retinal ischemia, which occurs as the consequence of a primary ocular disease or of a systemic disease, is a common cause of visual impairment and even irreversible blindness in the world. To counteract retinal ischemia, the tissue increases the production and release of angiogenic stimulators such as tumor necrosis factor-alpha (TNF-α) and vascular endothelial growth factor (VEGF) in order to promote angiogenesis that is the key event in pathogenesis [1, 2].

VEGF is a potent angiogenic and vasopermeability factor that has been initially identified in the bovine retina and later in the human retina [3, 4]. Intravitreal injections of recombinant human VEGF into healthy monkey eyes produce iris neovascularization and glaucoma [5]. Conversely, injection of VEGF antibodies to human vitreous causes regression of neovascularization [6].

TNF-α is a macrophage/monocyte-derived pluripotent polypeptide mediator that stimulates angiogenesis in the cornea and chorioallantoic membrane in vivo by its direct effect and also enhances the synthesis of VEGF, which is an angiogenic stimulator [7–10]. Additionally, TNF-α is a chemoattractant for leukocytes and causes breakdown of the blood-retinal barrier [11, 12].

In the 1950s, thalidomide [N(α).-phthalimidoglutarimide] (THD) was marketed as a sedative in Europe, but eventually withdrawn when it was found to be a potent teratogen. In 1961, shortly after its withdrawal, the anti-inflammatory properties were observed in erythema nodosum leprosum patients [13]. The anti-angiogenic effect of THD in a rabbit corneal neovascularization model [14] and inhibitor effect on TNF-α synthesis [15] were discovered in later years. To our knowledge the effect of THD in a model of retinal ischemia has not been studied. In our study we investigated the influence of THD treatment on TNF-α, VEGF production, and the histopathological changes in I/R injured guinea pig retinas.

Materials and methods

Experimental protocol

Twenty-one pigmented male guinea pigs weighing between 470 and 640 g were treated in accordance with the Association for Research in Vision and Ophthalmology in Experimental Research Center of our university upon the consent of the local ethics committee. Three groups (control, ischemia, and thalidomide/ischemia) including seven animals each were formed.

Induction of ischemia and reperfusion

Intramuscular ketamine HCl (50 mg/kg) and xylazine HCl (5 mg/kg) were used for the anesthesia of the animals. Topical anesthetic consisting of 0.005% proparacaine HCl was administered to both eyes of each animal. Pressure-induced retinal ischemia was induced by cannulating both anterior chambers with a 27-gauge needle connected to a bottle of normal saline and lifting the bottle rapidly to a height of 205 cm in order to raise the intraocular pressure to 150 mmHg. This lasted for 90 min, and reperfusion was established by lowering the saline bottle to eye level. Then the eyes were decanulated. In the control group, ischemia/reperfusion (I/R) was not induced, and no drug was administered. In the ischemia group, saline was administered orally via nasogastric tube with a dose of 2 ml/day along the experimental period. In the thalidomide/ischemia group, 300 mg/kg/day of THD was given in 2 ml saline via the same route 24 h before the ischemic insult and during the 7-day reperfusion period. At the 7th day of the experiment, the animals were reanesthetized, and both eyes of all the animals were rapidly enucleated; the animals were sacrificed by intracardiac thiopental sodium (50 mg/kg).

Biochemical assay

One eye of each animal was randomly selected for biochemical assays and the other for histopathological evaluation. The enucleated eyes for biochemical assays were dissected coronally through the pars plana. After removing the vitreous, the retinal tissue was gently peeled off from the choroidea and cut from the optic disc with fine forceps and scissors under an operating microscope and washed with phosphate buffered saline (PBS) (pH 7.4 molar 0.2). The retina was frozen, smashed, and homogenized in PBS at a dilution of 1/20. The supernatants were collected by centrifugation at 150 g for 10 min and stored at −80°C until the assay. Respectively, retinal VEGF and TNF-α levels were analyzed with RayBio Rat VEGF ELISA kit (cat. no.: ELR-VEGF-001c, USA), Biosource Rat TNF-α immunoassay kit (cat. no.: KRC3011, USA). The results were expressed as mean ± standard deviation.

Preparation of tissue samples

The eyes selected for histopathological examination were fixed in 10% formalin immediately after enucleation, and transverse sections passing through the optic nerve were obtained. The samples were embedded in paraffin wax, 5-μm-thick paraffin sections were prepared, and the specimens were stained with hematoxylin and eosin.

Histological changes

An Olympus BX50 light microscope was used for the histopathological evaluation of the tissue sections of three groups in a masked fashion. The quantification of the retinal ischemic damage was made by measuring the thickness of the inner plexiform layer (IPL) of the retina [16, 17]. The measurements were made with an ocular micrometer ×400 magnification within 0.5 mm from the optic nerve. Three measurements from adjacent locations in each nasal and temporal hemisphere were obtained, and a mean retinal thickness for each eye was obtained by averaging the six measurements. Polymorphonuclear leukocyte (PMNL) infiltration between the internal limiting membrane (ILM) and IPL was evaluated. The number of PMNLs was counted in ten high power fields (HPF)/slide. The results were expressed as mean ± standard deviation per HPF.

Statistical analysis

Statistical analyses were carried out by employing the Statistical Package for Social Sciences soft-ware 11.0 for Windows (SPSS, Inc., Chicago, IL). The Mann-Whitney U test was used, and a p value less than 0.05 was considered statistically significant.

Results

The mean retinal VEGF levels of the control, ischemia, and thalidomide/ischemia groups were 10.22 ± 2.58 pg/ml, 35.80 ± 5.97 pg/ml, and 19.01 ± 3.01 pg/ml, respectively (Fig. 1). The retinal VEGF levels were significantly elevated in the ischemia (p = 0.001) and thalidomide/ischemia (p = 0.001) groups compared with the control (Fig. 1). However, the retinal VEGF level was significantly lower in the thalidomide/ischemia group versus the ischemia group (p = 0.001) (Fig. 1).

Fig. 1
figure 1

Retinal VEGF levels from control, ischemia, and thalidomide/ischemia groups. Significant differences are demonstrated with *p < 0.05. The black lines in the box plot diagram show the median values of the groups

Full size image

The mean retinal TNF-α levels of the control, ischemia, and thalidomide/ischemia groups were 270.41 ± 69.77 pg/ml, 629.93 ± 146.41 pg/ml, and 340.93 ± 158.26 pg/ml, respectively (Fig. 2). The retinal TNF-α levels were significantly elevated in the ischemia group compared with the control (p = 0.001) (Fig. 2). There was no significant difference between the thalidomide/ischemia and the control group (p = 0.620) (Fig. 2). Also, the retinal TNF-α levels were significantly lower in the ischemia/thalidomide group versus the ischemia group (p = 0.017) (Fig. 2).

Fig. 2
figure 2

Retinal TNF-alpha levels from control, ischemia, and thalidomide/ischemia groups. Significant differences are demonstrated with *p < 0.05. The black lines in the box plot diagram show the median values of the groups

Full size image

Figure 3 shows the histological sections of retina in non-ischemic control eye (Fig. 3a), in retina subjected to I/R injury without treatment (Fig. 3b), and in I/R injured retina treated with THD (Fig. 3c), respectively. The mean thickness of the IPL in the control, ischemia, and thalidomide/ischemia groups were 18.92 ± 3.18, 29.28 ± 7.31, and 21.07 ± 3.49 μm, respectively (Fig. 4). In the statistical analysis, there was no difference between the ischemia and the thalidomide/ischemia group (p = 0.07) or between the control and the thalidomide/ischemia group (p = 0.318). However there was significant difference between the control and the ischemia group (p = 0.001). The mean PMNL infiltrations in the control, ischemia, and thalidomide/ischemia groups were 0.28 ± 0.48, 3.28 ± 2.28, and 0.85 ± 0.89, respectively. There was no statistical difference between the control and the thalidomide/ischemia groups (p = 0.259). However, there was a statistically significant difference between the control and the ischemia groups (p = 0.011) and between the ischemia and the thalidomide/ischemia groups (p = 0.038).

Fig. 3
figure 3

Hematoxylin and eosin staining of the retina. a. Control group retina. GCL: ganglion cell layer; IPL: inner plexiform layer; INL: inner nuclear layer; OPL: outer plexiform layer; ONL: outer nuclear layer; PL: photoreceptor layer. b. PMNLs infiltration in the internal limiting membrane (vertical arrows) and inner plexiform (horizontal arrow) layers of retina in the ischemia group. c. Retina from guinea pigs treated with oral thalidomide

Full size image
Fig. 4
figure 4

Inner plexiform layer thickness of control, ischemia, and thalidomide/ischemia groups. Significant differences are demonstrated with *p < 0.05. The black lines in the box plot diagram show the median values of the groups

Full size image

Discussion

Angiogenesis is tightly regulated by the action of both stimulatory and inhibitory factors that are in balance. Hypoxia shifts the balance in favor of angiogenesis. Many diseases such as diabetic retinopathy, occlusion of the central retinal artery, sickle cell anemia, and retinopathy of prematurity are driven by this unregulated angiogenesis.

VEGF is a potent angiogenic factor that regulates both physiological and pathological neovascularization [9, 10] secreted from pericytes, retinal endothelial, retinal pigment epithelial, and ganglion cells in response to hypoxia [2, 18, 19]. In our study, in the I/R injured groups we have also shown that the ischemia has increased the production of VEGF.

The antiangiogenic effects of THD initially became evident in the 1950s, but the mechanisms of its action were documented by subsequent studies. Blocking the action of basic fibroblast growth factor (bFGF) and reducing the release of matrix metalloproteinase-2 (MMP-2) are probable mechanisms of its action [14, 20]. In recent studies additional mechanisms were shown such as depletion of VEGF membrane receptors neuropilin-1, Flk-1 [21], and direct inhibition of VEGF secretion [22]. In our study we also demonstrated that THD decreased the VEGF synthesis.

Ischemia/reperfusion injury is a multistage process. Immediate hypoxia as well as return of oxygenated blood damages the tissue. Free radical generation and inflammatory-immunologic reactions are involved in the pathogenesis of I/R injury.

TNF-α increases nitric oxide (NO) production via induction of nitric oxide synthase (NOS)-2 in astrocytes [23]. NO interacts with \( \raise0.145em\hbox{${\scriptscriptstyle \bullet}$}O^{ - }_{2} \) and form peroxynitrite free radicals (·ONOO¯) that cause lipid peroxidation of cellular membranes in the mechanism of ischemic-reperfusion injury [24–26]. THD suppresses NO production via its direct effect and its indirect effect on TNF-α by increasing degradation of its mRNA [27, 28].

TNF-α leads to activation of PMNLs and upregulates cell adhesion molecules both on PMNLs and endothelium, followed by transmigration into the interstitium, release of toxic enzymes, and tissue damage, which is the inflammatory-immunologic reaction part of the I/R injury [29]. THD inhibits transmigration of PMNLs across human umbical vein endothelial cells [30]. The alternative way of its anti-inflammatory action is its inhibitor action on NF-κB (a multi-subunit transcription factor that takes part in the transcriptional regulation of many cytokines and regulation of inflammation mediator genes) activation induced by TNF [31, 32]. Our study demonstrates the first evidence that systemic THD reduces I/R injury in a guinea pig retina via reduction of TNF-α synthesis.

Thickness of the IPL has been selected for the quantification of ischemic retinal damage in the present study. IPL thickness has been associated with I/R injury in a number of studies [16, 17]. The IPL thicknesses have been observed to increase significantly in the ischemic group compared with the control. This result has been interpreted as the consequence of persisting retinal edema till the 7th day. Although statistically not significant, in the thalidomide/ischemia group compared with the ischemia group, the thickness of the IPL has been observed to be decreased. This insignificant decrement in the thickness of IPL in the thalidomide/ischemia groups may be interpreted as the consequence of the antiedematous effect of THD, which has been shown previously during early wallerian degeneration in neurons [33].

In conclusion, oral THD treatment was found to reduce PMNL infiltration, retinal edema, and VEGF and TNF-α synthesis of the retina in guinea pigs after ischemic insult. These effects of THD may have important therapeutic implications in diseases that are driven by ischemia compared to unregulated angiogenesis.