Superior Cervical Ganglionectomy Induces Changes in Growth Factor Expression in the Rat Retina

Abstract

purpose. To determine whether sympathetic nerves regulate expression of known angiogenic growth factors.

methods. Surgical sympathectomy (SNX) was used to remove sympathetic innervation to the eye. Real-time PCR was used to measure steady state mRNA expression of VEGF, VEGFR-2, angiopoietin-1, and Tie2. Western blot analysis was performed to assess protein expression. Blood–retinal barrier (BRB) permeability surface area product (PS) was measured using enhanced MRI on a separate group of control and SNX rats.

results. mRNA of both VEGF and VEGFR-2 decreased significantly at 6 weeks after SNX. VEGF protein expression also decreased significantly. VEGFR-2 protein was unchanged. Both angiopoietin-1 and Tie2 mRNA expression increased significantly after SNX. Immunoblot analysis showed that angiopoietin-1 protein expression coincided with its mRNA expression. Tie2 protein expression was unaffected. Sympathetic denervation did not significantly increase BRB PS.

conclusions. Surgical SNX results in a reduction in VEGF gene and protein expression, while increasing protein expression of Ang-1. These findings suggest that sympathetic nerves regulate the expression of angiogenic growth factors in rat retina.

Diabetic retinopathy is the number one cause of blindness in working-age adults (ages 25–60 years). ¹ Characteristics of retinopathy include capillary basement membrane thickening, pericyte dropout, microaneurysms, and neovascularization. ² However, the pathogenic mechanism between these histopathologic changes and diabetes has yet to be elucidated. Another characteristic of diabetes is substantial changes to sympathetic nerves. ^{3 4} We have shown that superior cervical ganglionectomy in nondiabetic rats causes increased density of capillaries in the outer plexiform layer of the rat retina, ⁵ as well as increased basement membrane thickening and pericyte loss. ⁶ β-Adrenergic receptors were found to be responsible for the changes in capillary density, evidenced by the fact that administration of the antagonist propranolol induced changes similar to those caused by surgical denervation. ⁷  

A potential mechanism for the vascular remodeling noted after sympathectomy (SNX) would be alterations in the regulation of angiogenic growth factors. 

VEGF is an angiogenic growth factor and a specific mitogen for vascular endothelial cells. ⁸ VEGF mRNA and protein expression are upregulated in diseases in which neovascularization is prominent, such as solid tumors and diabetic retinopathy. ^{9 10} Transgenic knockout models with deletion of one copy of the VEGF gene lead to embryonic lethality, suggesting that VEGF is necessary for vasculogenesis, as well as angiogenesis. ^{11 12} VEGF binds two receptor tyrosine kinases, flt-1 (VEGFR-1) and flk-1 (VEGFR-2), with high affinity. ¹³ VEGF has also been shown to have negative effects on the blood–retinal barrier (BRB), such as decreasing expression of the proteins essential for tight junction maintenance. ¹⁴ However, to our knowledge, the role of VEGF after loss of sympathetic innervation has never been investigated and it could be central in the vascular remodeling observed after SNX. 

Another receptor tyrosine kinase that is largely endothelial cell specific is the receptor Tie2. ¹⁵ Two ligands, angiopoietin (Ang)-1 and -2, are capable of binding Tie2. ¹⁶ Ang-1 has been shown to induce the autophosphorylation of Tie2 in cultured endothelial cells, ¹⁷ whereas Ang-2 acts as the natural antagonist by preventing binding and activation of Tie2 via Ang-1. ¹⁸ Ang-1-mediated activation of Tie2 regulates endothelial cell survival and blood vessel maturation and acts as an anti-inflammatory and protects against cardiac allograft arteriosclerosis. ¹⁹ Whereas VEGF is thought to play a primary role in initiating the angiogenic response, the phenotype of Ang-1 null embryos suggests a role for Ang-1 as a vascular stabilizer, particularly involved in the interactions between endothelial cells and supporter cells such as pericytes, ¹⁵ while also attenuating vascular leakage. ²⁰  

To characterize the expression levels of these two growth factor systems, female Sprague-Dawley rats were sympathectomized, followed by extraction of retinas for experimental use. Retinal extracts were then subjected to real-time polymerase chain reaction (RT-PCR) and Western blot analysis to elucidate the gene and protein activities of these growth factor systems. The effect of SNX on BRB integrity was evaluated by MRI. 

Methods

Surgical SNX

Female Sprague-Dawley rats were anesthetized intraperitoneally at postnatal day 60 with a mixture of ketamine (60 mg/kg), atropine (0.54 mg/kg), and xylazine (8 mg/kg). The right superior cervical ganglion was removed aseptically by using previously described methods. ²¹ Right eye ptosis was used to confirm denervation, and only rats displaying good ptosis were used for experiments. Retinal samples were taken 1, 3, and 6 weeks after SNX. The contralateral (left) eye served as an intra-animal control. All surgical procedures were approved by Institutional Animal Care and Use Committee at Southern Illinois University-Carbondale and conformed to the ARVO statement for the Use of Animals in Ophthalmic and Vision Research and NIH guidelines. 

RNA Isolation and Reverse Transcription

RNA was isolated from retinal samples of six to eight rats at each time point (TriReagent; Molecular Research Center, Inc., Cincinnati, OH). RNA was isolated using chloroform and isopropanol. RNA purity was detected by agarose gel electrophoresis, and RNA concentration was measured spectrophotometrically. Reverse transcription of 1 μg RNA was performed for cDNA synthesis (Improm II Kit; Promega, Madison, WI). The reaction mixture consisted of diethyl pyrocarbonate (DEPC) water, 5× reaction buffer, 25 mM MgCl₂, 10 mM dNTP, and 20 units RNAsin. Strands were extended for 60 minutes at 42°C, and the reverse transcriptase enzyme was heat-inactivated at 70°C for 15 minutes. RNase A inhibitor (0.2 μL; 10 mg/mL) was added, followed by incubation for 30 minutes at 37°C. Samples were stored at −20°C for real-time PCR. 

Real-Time Polymerase Chain Reaction Analysis of Gene Expression

Real-time PCR primers to detect rat VEGF, rat flk-1/KDR, Ang-1, and Tie2 were designed on computer GCG Software Prime (Accelrys, Campbell, CA). Primers were chosen to generate an amplicon smaller than 150 bp. The sequences of the PCR primer pairs (5′–3′) that were used for each gene are listed in Table 1 . Real-time PCR reactions and calculations were performed as described previously. ⁶ A 1.2% agarose gel was used to confirm correct amplicon size. 

Western Blot Analysis

Western blot analysis was performed to determine whether retinal protein expression of rat VEGF, rat VEGFR-2, Ang-1, or Tie2 is altered after SNX. Female Sprague-Dawley rats were euthanatized with 150 mg/kg pentobarbital, and the retina was dissected from the eye and lysed in buffer (1 mM Tris-HCl [pH 7.4], 10 mL 10% Igepal-40, 2.5 mL 10% Na-deoxycholate, and 1 mL 100 mM EDTA). A protein assay was performed using the kit solution to determine protein concentration (Bio-Rad, Hercules, CA), and 30 μg protein was loaded into each well of a 4% to 12% precast Tris-glycine gel (Invitrogen, Carlsbad, CA). Proteins were then transferred to a nitrocellulose membrane at 35 V for 2 hours. After transfer, nonspecific activity was blocked by incubation of the membrane in 2% dry milk in buffer (1 mM Tris [pH 7.5], 150 mM NaCl, and 0.05% Tween 20) for 1 hour at room temperature. Primary antibodies to VEGF (1:500; Chemicon, Temecula, CA), VEGFR-2 (1:250; Chemicon), Ang-1 (1:250; R&D Systems, Minneapolis, MN), or Tie2 (1:250; Chemicon) were applied overnight at 4°C with shaking. Membranes were then probed with secondary antibodies conjugated to horseradish peroxidase (anti-mouse and anti-rabbit; 1:5000; Promega, and anti-goat; 1:5000; Calbiochem, Darmstadt, Germany). Bands were enhanced by chemiluminescence (ECL Reagent; GE Healthcare, Buckinghamshire, UK) and photographed (Image Station 2000r; Eastman Kodak, Rochester, NY). Membranes were reprobed for actin (1:500; Santa Cruz Biotechnology, Santa Cruz, CA) to confirm equal loading between samples. Protein bands were then analyzed and normalized to actin using the system software for densitometry. 

MRI Examination

In urethane-anesthetized (0.083 mL of a 36% solution of urethane/20 g animal weight, intraperitoneally, freshly made daily; Aldrich, Milwaukee, WI) control and SNX rats, dynamic contrast-enhanced MRI data were generated and analyzed similarly to that previously described. ^{22 23} MRI data were acquired on a 4.7-T system, with a two-turn surface coil (1.5 cm diameter) placed over the eye and a spin-echo imaging sequence (repetition time [TR], 1 second; echo time [TE], 22.7 ms; number of acquisitions [NA], 1; matrix size, 128 × 256; slice thickness, 1 mm; field of view, 32 × 32 mm²; sweep width, 25,000 Hz; 2 minutes per image). Twelve sequential 2-minute images were acquired as follows: Three control images before injection of contrast agent and nine images during and after a 6-second Gd-DTPA bolus injection. The dose of contrast agent (Gd-DTPA, Magnevist; Berlex Laboratories, Wayne, NJ) was 0.1 mM Gd-DTPA per liter per kilogram. In each animal Gd-DTPA was injected at the same phase encode step collected near the beginning of the fourth image. After the MRI examination, the anesthetized animals were humanely killed. 

For each pixel, the fractional signal enhancement, E, was calculated: E = (S(t) − S0)/S0, where S(t) is the pixel signal intensity at time t after addition of contrast, and S0 is the precontrast signal intensity (measured from the average of the three control images) at the same pixel spatial location. Because no increase in vitreous signal intensity was found after injection of Gd-DTPA (see the Results section, Fig. 6 ), we did not calculate a BRB permeability surface area product in this study. For calculation of signal enhancement, a region-of-interest (ROI) was chosen that contained the entire vitreous space. The area of this ROI and the mean E within the ROI were measured at each postcontrast time point. Note that in some rats, subtle movement generated ghosting artifacts in the phase encode direction. Such artifacts are discernible as random noise by visual inspection and can reduce precision. To address this, some points in each rat were eliminated. In this study, of the nine scans per rat, we found in the SNX group large ghosting artifacts in two scans of rat 1, five scans of rat 2, one scan of rat 3, and one scan of rat 4. These points were not included in Figure 6 . 

Statistical Analysis

All data are expressed as the mean ± SEM. Statistics were analyzed on computer (Prism software; Graph Pad, San Diego, CA). Paired t-tests were used to compare the data from the sympathectomized retina to that of the contralateral retina. P < 0.05 was considered significant. A Mann-Whitney nonparametric test was used for the results of the MRI experiments, with P < 0.05 considered significant. 

Results

Effect of SNX on Time-Dependent Changes in VEGF/VEGFR-2 Gene Expression

SNX significantly decreased VEGF steady state mRNA expression by 33% compared with the contralateral retina 6 weeks after surgery (Fig. 1C ; P < 0.05). mRNA levels for VEGF were increased at 1- and 3-weeks after SNX (Figs. 1A 1B) ; however, these increases were not statistically significant. Similar to that of its ligand, VEGFR-2 mRNA was significantly decreased (38%) at the 6-week time point (Fig. 2C) . Steady state mRNA expression of VEGFR-2 was unaffected by SNX at 1 and 3 weeks after surgery (Figs. 2A 2B) . 

Effect of Superior Cervical Ganglionectomy on VEGF Protein Expression

Because SNX caused significant reduction in steady state mRNA levels of VEGF and VEGFR-2 at the 6-week time point, protein expression of this ligand-receptor system was also evaluated. SNX also significantly (58%) downregulated VEGF protein expression, correlating with the decrease in mRNA levels (Fig. 3 ; P < 0.05). Whereas SNX affected both the receptor and ligand mRNA expression, it did not produce substantial changes in VEGFR-2 protein at 6 weeks (data not shown). 

Effect of Superior Cervical Ganglionectomy on Time-Dependent Ang-1 and Tie2 Steady State mRNA Levels

SNX induced significant changes in gene expression of both Ang-1 and its receptor Tie2 6 weeks after superior cervical ganglionectomy. Ang-1 mRNA was increased by 50% (Fig. 4A ; P < 0.05), whereas Tie2 mRNA was upregulated by 64% (Fig. 4B ; P < 0.05). 

Effect of SNX on Ang-1 Protein Expression

Superior cervical ganglionectomy significantly increased Ang-1 protein expression by 58% 6 weeks after surgery, in concert with the induction on steady state mRNA expression (Fig. 5 ; P < 0.05). Just as was the case with VEGFR-2, Tie2 protein expression was not altered by SNX (data not shown). 

Effect of SNX on the BRB

As summarized in Figure 6 , BRB fractional signal enhancement (E) for control and SNX rats at the 6-week time point were not significantly different (P > 0.05). 

Discussion

In this study, we demonstrated that loss of sympathetic innervation to the rat retina induces changes in gene expression of both the VEGF/VEGFR-2 and the Ang-1/Tie2 angiogenic ligand-receptor axis, without compromising the BRB. These changes in growth factor expression suggest a potential mechanism for the observed increase in capillary density in the outer nuclear layer of the rat retina at 6 weeks after SNX. 

VEGF is reported to be responsible for increased growth of blood vessels, as an initiator of vascular growth, stimulating proliferation and migration of vascular endothelial cells. ²⁴ However, SNX significantly decreased steady state mRNA and protein expression at the 6-week postsurgery time point. Because the increased capillary density was reported at 6 weeks after surgery, we originally considered the possibility that the increased VEGF expression needed for vascular remodeling had occurred before the 6-week time point. This hypothesis was wrong; VEGF mRNA levels were unchanged at both 1 and 3 weeks after superior cervical ganglionectomy. It has been reported that sympathetic nervous system stimulation, via norepinephrine, induces VEGF gene expression in adipocytes. ²⁵ This correlates well with the results observed in the SNX model, given that superior cervical ganglionectomy eliminates sympathetic communication, thus inhibiting any sympathetic VEGF gene regulation, resulting in decreased gene expression. In addition, MRI examination revealed no significant changes in the BRB, which supports a downregulation of VEGF, because VEGF has been shown to compromise BRB function. ¹⁴ Taken together, these results suggest the possibility that another factor is responsible for increased capillary density of the outer nuclear layer induced by SNX. 

Ang-1 may be the factor that causes vascular remodeling of the outer retina. Although it has been shown that Ang-1 and Tie2 are capable of regulating cell proliferation and chemotactic migration of fibroblasts, ²⁶ Ang-1 is primarily thought to have a maturation effect on the vasculature. Ang-1 is especially hypothesized to play a role in the accumulation and function of pericytes. ¹⁵ Because Ang-1 mRNA and protein expression are increased at 6 weeks after SNX, they may be providing stabilizing effects on the remodeled vasculature. We have previously reported that SNX results in pericyte dropout, as well as decreased steady state mRNA levels of PDGF-BB, ⁶ which is also hypothesized to regulate pericyte growth, function, and survival. ²⁷ Therefore, it is possible that Ang-1 and Tie2 expression are upregulated to compensate for the decreased activity of PDGF-BB and loss of pericytes. The cellular mechanisms by which sympathetic nerves may regulate Ang-1 is unknown. Ang-1 transcription is induced by the transcription factor ESE-1 of the ETS family of transcription factors. ²⁸ There is also evidence that AML-1 of the runt transcription factor family regulates constitutive expression of Ang-1 in hematopoietic cells. ²⁹ Future studies in vitro are needed to assess the role of sympathetic nerve regulation of these transcription factors. 

Other factors may be involved in the vascular remodeling noted in the outer retina after SNX. Because it is clear that diabetes involves several complications that involve altered vascular changes, other compensatory mechanisms may be in place to elicit the observed changes. It is not likely that it is due to changes in parasympathetic or sensory innervation, as we have previously shown that neither para-SNX nor sensory denervation elicits vascular remodeling in the eye.³⁰ There has also been some suggestion that nerve growth factor (NGF), whose levels would be increased after SNX, may elicit angiogenic effects. Using real-time PCR and Western blot analysis, we have not seen gene or protein expression changes in NGF or either of its receptors, TrkA and p75 (Steinle et al., unpublished data, 2004). 

The purpose of this study was to elucidate a mechanism for the increased capillary density observed after superior cervical ganglionectomy. Although differentially expressed, the data suggest that both VEGF and Ang-1 are regulated by the sympathetic nervous system. Given that sympathetic nerves are compromised in diabetes, ³ further understanding of the role that sympathetic innervation has on angiogenic growth factor expression could be useful. Because SNX and diabetic retinopathy show similar histologic similarities, changes in sympathetic regulation of angiogenic growth factors may play a role in retinopathy. Therefore, the SNX model is a novel approach to the investigation of complications involved in retinopathy and could provide new avenues for therapeutic development. 

 Supported by American Heart Association Grant 0430344Z (JJS), National Eye Institute Grant R01 EY013831 (BAB), and Juvenile Diabetes Research Foundation Grant 1-2003-696 (BAB).

 Submitted for publication May 25, 2005; revised July 21, August 5, and August 24, 2005; accepted October 26, 2005.

 Disclosure: L. A. Wiley, None; B. A. Berkowitz, None; J. J. Steinle, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Primers Used for the Real-Time PCR Experiments

Primer	Sequence (5′→3′)	Sense/Antisense
Rat VEGF	ACGAAGCGCAAGAAATCCC	Sense
Rat VEGF	TTAACTCAAGCTGCCTCGCC	Antisense
Rat VEGFR-2	TAGCACGACAGAGACTGTGAGG	Sense
Rat VEGFR-2	TGAGGTGAGAGAGATGGGTAGG	Antisense
Angiopoietin-1	CCATGCTTGAGATAGGAACCAG	Sense
Angiopoietin-1	TTCAAGTCGGGATGTTTGATTT	Antisense
Tie2	CGGCTTAGTTCTCTGTGGAGTC	Sense
Tie2	GGCATCAGACACAAGAGGTAGG	Antisense
GAPDH	TCCACCACCCTGTGCTGTA	Sense
GAPDH	ACCACAGTCCATGCCATCAC	Antisense

 GAPDH was used as a housekeeping gene, and all PCR data were normalized to its expression.

Figure 6.

 

Summary of BRB fractional signal enhancement (E) from control and SNX rats. No significant difference (P > 0.05) between groups was found (n = 4 for each group). Solid squares: control; solid diamonds: SNX.

Figure 1.

 

Steady state mRNA expression of VEGF at (A) 1, (B) 3, and (C) 6 weeks after SNX. Increase in VEGF mRNA seen at 1- and 3-week time points (n = 5 and n = 6, respectively). Significant decrease in VEGF mRNA (*P < 0.05) 6 weeks after superior cervical ganglionectomy (n = 6).

Figure 2.

 

Steady state mRNA levels of VEGFR-2 at (A) 1, (B) 3, and (C) 6 weeks after SNX. mRNA expression was significantly decreased (*P < 0.05) at 6 weeks after SNX (n = 7), whereas mRNA levels were unchanged at the 1- and 3-week time points (n = 6 and n = 5, respectively).

Figure 3.

 

VEGF protein expression in retinal lysates 6 weeks after surgery. (A) Representative Western blot of VEGF and actin. +, surgery; −, nonsurgical control. (B) Results of densitometry after VEGF normalization to actin (n = 5). VEGF protein, similar to the mRNA, was significantly decreased (*P < 0.05) 6 weeks after surgery.

Figure 4.

 

Results of real-time PCR for Ang-1 (A) and Tie2 (B) 6 weeks after SNX. Both Ang-1 (*P < 0.05) and Tie2 (*P < 0.05) were significantly increased compared with the contralateral eye at the 6-week time point (n = 7 and n = 6, respectively).

Figure 5.

 

Angiopoietin-1 protein expression in retinal lysates. (A) Representative blot of Ang-1 and actin at 6 weeks after SNX. +, surgery; −, nonsurgical control. (B) The Angiopoietin-1 protein level was significantly increased (*P < 0.05) compared with that of the contralateral eye, after normalization to actin (n = 5).

The authors thank Robin Roberts and Hongmei Luan for help with the MRI experiments.