Abstract
purpose. To determine whether loss of sympathetic innervation alters basement membrane thickness and pericyte loss.
methods. Sympathetic innervation to the eye was destroyed by surgical removal of the right superior cervical ganglion in rats. Basement membrane changes were assessed by real-time PCR and electron microscopy. The number of pericytes was measured by immunofluorescent staining for NG2 proteoglycan. Steady-state mRNA levels were also evaluated for platelet-derived growth factor-BB (PDGF-BB).
results. Loss of sympathetic innervation caused a significant increase in steady state mRNA levels of fibronectin and a 15% increase in laminin-β1 mRNA 3 weeks after surgical sympathectomy. Protein expression also increased at this point. In addition, capillary basement membrane thickness increased significantly. NG2 proteoglycan staining decreased significantly in pericytes in the sympathectomized rat retina. Steady state mRNA for PDGF-BB decreased significantly 6 weeks after surgery.
conclusions. Sympathetic nerves may be compromised in diabetes, and these findings suggest that they may regulate some complications of diabetic retinopathy. Gene expression levels of fibronectin and laminin-β1 changed between 1 and 3 weeks. These data are supported by electron microscopy, which showed the increase in basement membrane thickness in vivo. Loss of sympathetic innervation to the eye also caused a decrease in the number of pericytes. Steady state mRNA expression of PDGF-BB was reduced, suggesting a mechanism for the loss of pericytes in the sympathectomized retina. Overall, these results suggest that sympathetic nerve alterations may function in some complications observed in diabetic retinopathy, and this may be a suitable model to investigate therapies for this disorder.
The complications of diabetes are the number one cause of blindness, end-stage renal disease, and nontraumatic lower limb amputation.
1 Diabetic retinopathy is the number one cause of blindness in working-age people (ages 25–60 years).
1 2 The principal characteristics of diabetic retinopathy are thickening of the basement membrane in the retina, loss of pericytes, increased proliferation of endothelial cells, and formation of microaneurysms, which can lead to neovascularization.
3 However, it is still unclear how these alterations are caused by chronic diabetes.
Alterations to sympathetic innervation in the eye could contribute to diabetes-induced change. Sympathetic nerves are significantly altered in diabetes.
4 5 We have shown that sympathectomy results in increased density of capillaries in the outer nuclear layer of the retina.
6 These changes are mediated by β-adrenergic receptors, since administration of propranolol causes changes similar to those noted in sympathectomy.
7 Even though the retina is autoregulated, we and others
8 9 have shown that β-adrenergic receptors are present on endothelial cells of retinal blood vessels and thus may influence retinal physiology. Furthermore, stimulation of β-adrenergic receptors decreases basement membrane thickness,
10 and so we hypothesized that ganglionectomy would increase basement membrane thickening by removing the inhibitory effect of the sympathetic nervous system. In addition to changes in the basement membrane, we sought to assess the ability of sympathectomy to produce another common complication noted in retinopathy: loss of pericytes.
To test this hypothesis, female Sprague-Dawley rats underwent surgical removal of the superior cervical ganglion, followed by assessment of gene and protein expression of two key basement membrane components (laminin-β1 and fibronectin), electron microscopy to determine basement membrane thickness, and evaluation of the number of pericytes. Steady state mRNA expression was also assessed for the growth factor PDGF-BB.
Twenty-six 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 previously described methods.
11 Right eye ptosis was used to confirm denervation, and only rats displaying good ptosis were used in the experiments. Retinal samples were taken 1, 3, and 6 weeks after sympathectomy. The contralateral or left eye served as an intra-animal control. All surgical procedures were approved by the Institutional Animal Care and Use Committee at Southern Illinois University-Carbondale and conform to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and NIH guidelines.
RNA was isolated from retinal samples of six rats at each time point (TriReagent; Molecular Research Center, Inc., Cincinnati, OH), by using chloroform and isopropanol. RNA purity was detected by agarose gel electrophoresis, and RNA concentration was measured spectrophotometrically. Reverse transcription of 1 μg RNA for cDNA synthesis was performed (Improm II Kit; Promega, Madison, WI). The reaction mixture consisted of diethyl pyrocarbonate (DEPC) water, 5× reaction buffer (Improm II; Promega), 25 mM MgCl2, 10 mM dNTP, and 20 U 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 PCR primers to detect rat fibronectin, laminin-β1, and PDGF-BB were designed using computer software (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′ to 3′) that were used for each gene are as follows: rat fibronectin, 5′-GGGATCAAAGGGAAACACAG-3′ (forward) and 5′-AGACGGCAAAAGAAAGCAG-3′ (reverse); laminin-β1, 5′-TGTAGATGGCAAGGTCTTATTTCA-3′ (forward) and 5′-CTCAGGCAGTTCTGTTTGATGT-3′ (reverse); and PDGF-BB, 5′-GGGAATACTGCTCACAACG-3′ (forward) and 5′-GCATACAAAATAGCACTTCCG-3′ (reverse). Real-time PCR reactions were then performed with a PCR mix (iQ SYBR Green Supermix, containing 100 mM KCl, 40 mM Tris-HCl [pH 8.4], 0.4 mM of each dNTP, 50 U/mL DNA polymerase [iTaq], and 6 mM MgCl2, SYBR Green I, 20 nM fluorescein, and stabilizers; Bio-Rad, Hercules, CA). Thermocycling was performed in a final volume of 25 μL (8 μL DEPC H2O, 2 μL cDNA, 1.25 μL = 500 nM of each primer, and 12.5 μL of 2× iQ SYBR Green Supermix; Bio-Rad) using the PCR conditions of initial heating at 95°C for 300 seconds to denature cDNA and activate the Taq DNA polymerase, followed by 45 cycles consisting of denaturation at 95°C for 20 seconds, annealing at 58°C for 20 seconds, and extension at 72°C for 20 seconds using thermocycler (Smart Cycler; Cepheid, Sunnyvale, CA). One additional step, a melting curve, was added to determine specificity. The melting curve was constructed by increasing the temperature from 60°C to 95°C with a temperature transition rate of 0.2°C/s. To ensure that the correct product was amplified, all samples were separated by 1.2% agarose gel electrophoresis.
To correct for differences in both RNA quality and quantity between samples, data were normalized by using the ratio of the target cDNA concentration to that of GAPDH. To calculate the increases (
x-fold) in steady state RNA levels, the change in cycle threshold Δ
C T for control rat retina expression of fibronectin and laminin-β1 was calculated by subtracting from their threshold (
C T) the corresponding GAPDH Δ
C T (internal control). Then Δ
C T was calculated by subtracting the average of the Δ
C T in the control rat retina from the Δ
C T in sympathectomized rat retinal expression of the gene of interest. Changes in steady state gene expression are reported as
x-fold increases (2
−ΔΔCT) relative to control rat retina. The statistical analysis of steady state RNA levels is based on previous work by other groups.
12 13 Significance in the 2
−ΔΔCT was set at
P < 0.05, determined by computer (Prism Software; GraphPad, San Diego, CA).
Four sympathectomized animals were used at 6 weeks after surgery. The blood vessels were cleared by perfusion with Millonig phosphate buffer (2.26% monobasic sodium phosphate, 2.52% NaOH, 5.4% glucose) followed by perfusion fixation with 0.5% glutaraldehyde and 0.54% dextrose in Millonig phosphate buffer. Surgically extracted retina was also fixed in phosphate buffer containing 2% glutaraldehyde. After postfixation with 1% buffered osmium tetroxide, samples were dehydrated, en bloc stained with alcoholic uranyl acetate, and embedded (Polybed-Araldite; Polysciences, Warrington, PA) according to previously described procedures.
14 15 Sections of 70- to 80-nm nominal thickness were stained with 2% aqueous uranyl acetate and Reynold lead citrate. Sections were visualized on an electron microscope (H71-FA; Hitachi, Ltd., Tokyo, Japan) located in the Southern Illinois University-Carbondale Image Center. Scale bars were generated by the microscope and imaged onto each negative. Images were recorded on standard transmission electron microscopy (TEM) film and subsequently digitized with a digital scanner (T-2500; AGFA, Orangeburg, NY). Basement membrane thickness of capillaries in the retinal ganglion cell layer was determined by morphometric analysis of the digitized images. At least 25 images were recorded and analyzed from each independent preparation. Because of the oblique sectioning of the capillaries, four measurements in each image were taken of the basement membrane to address the question of varied membrane thickness along a capillary.
16 17 The mean of the four measurements of each image was calculated. The means of all capillaries assessed in each treatment group were combined to give a total mean thickness of basement membranes for that eye. The means of four sympathectomized and four contralateral eyes were assessed with a paired
t-test, with
P < 0.05 considered significant.
To assess pericyte dropout after surgical sympathectomy (SNX), five sympathectomized and five contralateral eyes were fixed in 4% paraformaldehyde overnight at 4°C. After they were rinsed in PBS, each retina was dissected from the uvea by small incisions, without cutting the optic nerve. The eyes were then placed in blocking solution (3% horse serum, 3% goat serum, 0.1% Triton X-100, 0.05% saponin, 50 mM NaCl, and PBS) overnight at 4°C. The following day, the eyes were incubated in rabbit anti-NG2 primary antibody (Chemicon, Temecula, CA) diluted 1:200 in blocking solution without saponin and Triton X-100 for 24 hours at 4°C. Eyes were then washed with PBS and incubated in secondary antibody conjugated to Cy3 for 2 hours. After they were washed, the retinal sections were placed flat on a slide and coverslipped (Fluoromount G; Southern Biotechnology, Birmingham, AL). Micrographs of the retinal flatmounts were taken at 40× with a fluorescence microscope (Olympus, Tokyo, Japan) connected to a computer (Macintosh G4; Apple Computer, Cupertino, CA, running OpenLab; Improvision, Inc., Lexington, MA). Ten images per retina were taken randomly throughout the peripheral retina. A 2.13-mm2 grid was placed over the images and the number of pericytes determined. A mean was determined for each retina for five contralateral and four sympathectomized eyes.
All data represent the mean ± SEM. Statistics were determined on computer (Prism Software, Graph Pad). One-tailed, paired t-tests were used to compare the sympathectomized retinal data with the contralateral retina data.
Effect of Loss of Sympathetic Innervation on Time-Dependent Changes in Gene Expression of Fibronectin and Laminin-β1
Fibronectin steady state mRNA levels were significantly increased 3 weeks after sympathectomy compared with that of the contralateral eye (
P = 0.04,
Fig. 1B ). The gene was upregulated 39% at 1 week after surgery; however, this change was not statistically significant because of animal variability
(Fig. 1A) . Six weeks after SNX, endogenous mRNA expression had returned to basal levels
(Fig. 1C) . Steady state mRNA levels for the gene encoding laminin-β1 were significantly decreased at both 1 and 6 weeks after SNX (
P = 0.02,
Fig. 1D , and
P = 0.04,
Fig. 1F , respectively). There was a 15% increase in endogenous mRNA levels at 3 weeks after surgery
(Fig. 1E) . These results suggest that sympathectomy affects regulation of prominent genes that are likely to be involved in basement membrane thickening in the rat retina.
Time-Dependent Changes in Protein Expression of Fibronectin and Laminin-β1 after Sympathectomy
Protein expression of fibronectin and laminin-β1 displayed trends similar to those observed for steady state mRNA levels, but was not significant (data not shown). Although there were significant changes in gene expression of fibronectin and laminin-β1, the changes in protein expression were minor. However, it appears that slight changes in protein expression could have important physiological implications in the retina.
Basement Membrane Thickness of Capillaries in Retinal Ganglion Cell Layer after Sympathetic Denervation
The average width of capillary basement membrane thickness in sympathectomized retinas increased significantly (
P = 0.04,
Fig. 2 ) compared with the contralateral eyes 6 weeks after ganglionectomy. These results support those showing increases in steady state mRNA and protein expression of fibronectin and laminin-β1, providing further evidence that sympathetic nerves alter basement membrane changes in the retina.
Fewer pericytes were present 6 weeks after sympathectomy. Quantitative analysis of the number of pericytes provided statistically significant results, in that sympathectomy decreased the number of pericytes by 50% (
P = 0.03,
Fig. 3 ), relative to the number of pericytes in the contralateral eye. These data indicate that the sympathectomy rat model causes pericyte loss, a hallmark of proliferative diabetic retinopathy.
Effect of Ganglionectomy on Steady State mRNA Expression of PDGF-BB 6 Weeks after Surgery
Steady state mRNA levels for the growth factor PDGF-BB were significantly decreased (
P = 0.03,
Fig. 4 ) 6 weeks after sympathectomy. Hypothesized to be a stabilizing factor for pericytes, decreased PDGF-BB mRNA could be a mechanism of pericyte loss in sympathectomized animals.
The focus of the current studies was to investigate a potential role of sympathetic nerves in inducing complications that are common in diabetic retinopathy, such as increased basement membrane thickening and pericyte loss. Fibronectin and laminin-β1 are two key extracellular matrix (ECM) components, making them essential in the regulation of basement membranes. Basement membranes of retinal vessels in diabetic rats contain increased amounts of the β1 chain of laminin, as well as that of fibronectin, as early as 8 weeks after induction of type I diabetes.
18 In addition, O’Callaghan and Williams
10 have shown that stimulation of β
1-adrenoceptors on ECM proteins on human cardiovascular cells has a significant inhibitory effect on ECM protein synthesis in vitro. Therefore, removing the innervation to these receptors and the inhibitory regulation they provide may cause a subsequent increase in the thickening of the basement membrane in the retina. Because ECM components are altered and sympathetic nerves may undergo significant deterioration in a diabetic state, we questioned whether fibronectin and laminin-β1 expression are altered by sympathetic denervation. In the present study, sympathectomy induced increases in steady state mRNA levels of the fibronectin and laminin-β1 genes. In addition, protein levels of fibronectin and laminin-β1 were upregulated, coinciding well with mRNA data. Furthermore, electron microscopic analysis of basement membrane thickness of capillaries in the retinal ganglion cell layer revealed a significant increase in sympathectomized retina compared with the contralateral retina. Thus, genetic changes were producing substantial alterations of retinal ECM. Based on these results, it appears that sympathetic nerves regulate basement membrane components, and loss of sympathetic regulation leads to increases in basement membrane thickness.
In addition to basement membrane thickening, pericyte loss is another common complication of diabetic retinopathy. Pericytes are supporting cells of the microvasculature, located within the basement membrane of capillaries and venules.
19 After induction of diabetes in rodents, reduction of the number of pericytes in retinal capillaries is one of the earliest morphologic changes, followed by the formation of an increased number of acellular-occluded capillaries, occasional microaneurysms, and thickening of the vascular basement membrane.
20 Past investigations
21 22 have shown that retinal pericytes possess functional adrenergic receptors, suggesting that they are regulated by the sympathetic branch of the autonomic nervous system. Removal of the right superior cervical ganglion induced a significant decrease in pericytes within the retinal vasculature.
A potential mechanism for a sympathectomy-induced decrease in pericytes is lowered expression of steady state mRNA for PDGF-B. PDGF has major effects on pericyte activation, survival, and growth.
23 Furthermore, adrenoceptor stimulation increases gene expression of PDGF-A, an isoform and member of the PDGF gene family.
24 In the present study, sympathetic denervation was shown to decrease steady state mRNA expression of PDGF-BB, indicating that sympathetic nerves probably play a role in regulation of the number of pericytes.
Whereas sympathectomy produces increases in both basement membrane thickness and pericyte loss common in retinopathy, these rats were not diabetic. Furthermore, these changes occurred within 6 weeks of sympathetic nerve loss. However, the sympathectomy model also produced increased capillary density in the outer nuclear layer of the retina, suggesting it may mimic complications of both nonproliferative and proliferative diabetic retinopathy. In addition, sympathetic nerve innervation can be altered ipsilaterally to allow for an intra-animal control. Therefore, although the sympathectomy model is not truly diabetic, it may provide novel information regarding some factors common in diabetic retinopathy.
The findings in the present study indicated that sympathetic innervation plays a role in the rat retina. Previous data have also shown that destruction of sympathetic nerves to the eye causes increased vascular density of capillaries in the outer nuclear layer of the retina.
6 This study built on previous findings to demonstrate that other common changes noted in diabetic retinopathy are produced by sympathetic denervation, in addition to vascular changes. Thus, since sympathetic nerves are negatively affected by diabetes and appear to influence basement membrane thickness and pericyte loss, treatments intended to restore sympathetic activity in the eye may lead to reduction of these complications of diabetic retinopathy.