Experimental Eye Research 84 (2007) 412e422www.elsevier.com/locate/yexer

Alterations in the choroid in hypercholesterolemicrabbits: Reversibility after normalization of cholesterol levels

Juan J. Salazar a, Ana I. Ramırez a, Rosa de Hoz a, Blanca Rojas a, Emilio Ruiz b,Teresa Tejerina b, Alberto Trivino a,*, Jose M. Ramırez a,*

a Instituto de Investigaciones Oftalmologicas Ramon Castroviejo, School of Medicine, Complutense University, 28040 Madrid, Spainb Department of Pharmacology, School of Medicine, Complutense University, 28040 Madrid, Spain

Received 24 July 2006; accepted in revised form 19 October 2006

Available online 18 December 2006

Abstract

Endothelial damage in atherosclerosis is characterized by abnormal vascular functionality. Hyperlipidemic patients show alterations in ocularvascularization. However, it is not known whether these alterations are reversible after the lipid profile returns to normal. This study evaluatesa rabbit model of hypercholesterolemia, examining the ultrastructural changes in the choroid, and the changes in it after a period of normalblood-cholesterol values induced by a standard diet. Rabbits were divided into three groups: G0, fed a standard diet; G1A, fed a 0.5% choles-terol-enriched diet for 8 months; and G1B, fed a 0.5% cholesterol-enriched diet for 8 months followed by a standard diet for a further 6 months.Eyes were processed for transmission electron microscopy. G1A had a buildup of lipids at the suprachoroidea that compressed the vascularlayers, and hypertrophy of endothelial and vascular smooth muscle cells. In G1B there was less lipid accumulation than in G1A, but thiswas not followed by reversal of the choroidal damage. The suprachoroidea thickness of G1B was still greater than in G0 due to abundant col-lagen fibers. The intervascular spaces of the choroid had fewer lipids than G1A but more collagen fibers than G0. The large- and medium-sizedvessel layers and choriocapillaris were less compressed than in G1A but exhibited basal membrane and endothelial changes similar to those inG1A. Normalization of serum cholesterol levels is not enough to reverse cholesterol-induced vascular damage to the choroid. These choroidalchanges could be compatible with a chronic ischemia that could produce retinal degeneration.� 2006 Elsevier Ltd. All rights reserved.

Keywords: cholestrol-fed rabbit; choroid; choriocapillaris; electron microscopy; ischemia

1. Introduction

The importance of endothelial dysfunction as an early eventin the atherosclerotic process was summarized by Ross in1999. Endothelial dysfunction is characterized by the expres-sion of adhesion molecules such as ICAM-1 and VCAM-1,which favors migration of monocytes to the subendothelialspace, where they transform into macrophages (Ross, 1999;Crispin, 2002).

* Corresponding authors. Tel.: þ34 91 3941358; fax: þ34 91 3941359.

E-mail addresses: atrivino@med.ucm.es (A. Trivino), ramirezs@med.

ucm.es (J.M. Ramırez).

0014-4835/$ - see front matter � 2006 Elsevier Ltd. All rights reserved.

doi:10.1016/j.exer.2006.10.012

A deficient endothelial function in association with a highplasmatic lipid concentration encourage increased endothelialpermeability for low density lipoproteins (LDL). When LDLspass through an impaired endothelium, LDLs become oxi-dized and induce over-expression of adhesion molecules onthe cell surface. Upon interaction with scavenger receptors,oxidized LDLs penetrate macrophages, which then becomefoam cells (Rong et al., 1999). Overcharged foam cells deliveresterified and oxidized cholesterol to the extracellular space,inducing several effects. Among them are increased endothe-lial damage, progression of the atherosclerotic lesion second-ary to changes in the extracellular matrix (Oorni et al., 2000;Ross, 1993), and major alteration of the vascular tone dueto the production of more vasoconstrictive factors than

413J.J. Salazar et al. / Experimental Eye Research 84 (2007) 412e422

vasodilator factors by endothelial cells (Fuster et al., 1992;Jimenez et al., 2002).

The mechanisms cited are known features of the cardiovas-cular system and may be present in what has been called thesmall vessel disease of cerebral atherosclerosis (Sanchez-Perez et al., 1999). In this pathology, microatheromas in 90-mm vessels and arteriolar lipohyalinosis have been reported(Sanchez-Perez et al., 1999). However, few studies have exam-ined ultrastructural changes in choroidal and retinal vascular-ization during hyperlipidemia, despite epidemiological (Kleinet al., 2000; Wong et al., 2001, 2002a,b) and clinical evidencesof alterations in patients with this condition. Hyperlipidemicpatients are found to have a generalized decrease of retinalsensitivity in the visual field (Hayreh, 1970, 1971; Hayrehet al., 1970; Greve and Geijssen, 1983; Phelps et al., 1983;Caprioli et al., 1987; Schulzer et al., 1990; Terres et al.,1992) and/or areas of peripapillary atrophy on fundus exami-nation (Jonas et al., 1990, 1992; Tuulonen and Airaksinen,1992) and histopathology (Kubota et al., 1996). These findingsare probably related to choroidal ischemia and alteration of theretinal pigment epithelium (RPE)ephotoreceptor complex.

Laboratory rabbits have been widely used for the study ofatherosclerosis for several reasons: they exhibit hypercholes-terolemia within a few days of administration of a choles-terol-enriched diet and high sensitivity to the inducement ofatheromatic lesions similar to human atherosclerosis (Finkingand Hankeh, 1997; Yanni, 2004). Because of their similarity tohuman fatty streaks, the cholesterol-fed rabbit model of aorticatherosclerosis has been adopted as a ‘‘universal’’ test modelfor determination of the anti-atherogenic activity of a large va-riety of drugs, such as calcium antagonist, hypolipidemicagents and assorted natural and synthetic compounds (Daugh-erty et al., 1998; Del Rio et al., 1995; Huff and Carrol, 1980;Zauberman and Livni, 1981).

Table 1

Serum concentration of lipids

G0

(mg/100 ml)aG1A

(mg/100 ml)aG1B

(mg/100 ml)a

Cholesterol (c) 43�15.8 1753�303 33.8�6.21

Triglycerides (Tg) 167�46.4 352�109.6 195.2�44.4

Phospholipids (Ph) 68�18 576�84.6 66.4�9.12

c-VLDL 16�18 1130�310 19�5

Tg-VLDL 97�23.1 274�104.1 105�24

Ph-VLDL 20.3�7 332�84.9 22�5

c-HDL 10�2.1 28.6�4.8 11.4�3.1

Ph-HDL 30�2 42�4.8 37.6�5.1

c-LDL 6.3�2 259�27.2 3.8�1.4

Ph-LDL 7.3�1.4 100�8.3 5.4�0.92

c-IDL 10.6�9.2 350�80.1 33.8�4.8

Tg-IDL 24.7�19.2 57�28.7 195.2�40.1

Ph-IDL 9.3�8.3 114.7�25.5 66.4�8.3

Total cholesterol at the end of the experiments: very low density lipoproteins,

c-VLDL; high density lipoproteins, c-HDL; low density lipoproteins, c-LDL;

intermediate density lipoproteins, c-IDL. G0, Control group, standard diet;

G1A, hypercholesterolemic rabbits, standard diet enriched with 0.5% choles-

terol for 8 months; and G1B, reverted rabbits, standard diet enriched with

0.5% cholesterol for 8 months plus standard diet for another 6 months.a Mean standard deviation.

To the best of our knowledge, chorioretinal changes after nor-malization of plasmatic lipid levels has not been reported. In thepresent work we analyze ultrastructural changes in the choroidin a rabbit hypercholesterolemia model and compare themwith changes induced by normalization of serum cholesterol

Fig. 1. Atherosclerotic changes in thoracic aorta. (A) G1A the foam-cell-streak

reduced the vascular lumen. (B) The vascular lumen in G1B was less reduced

due to less infiltration of the intima by foam cells. (C) G0, Histological sec-

tions, Hematoxylineeosin. 1, Intima; 2, media; 3, adventitia [Bar¼ 100 mm].

414 J.J. Salazar et al. / Experimental Eye Research 84 (2007) 412e422

values. The aim of this study is to determine whether normaliza-tion of serum cholesterol values reverse such alterations.

2. Methods

2.1. Experimental design

Thirty adult male New Zealand White rabbits weighing2.5� 0.5 kg were housed separately in cages in an air-condi-tioned room with a 12-h light/dark cycle. All animals were feda rabbit standard diet (carbohydrates 50%, fibre 15.5%, protein13.5%, moisture 11%, minerals 7%, and lipids 3%; Panlab S.L.Barcelona, Spain) at least 7 days before the beginning of the ex-periment and allowed ad libitum access to water. The animalswere divided into three groups: (1) Control group (G0;n¼ 10), rabbits fed the standard diet; five animals were killedat 8 months (control for the age of the hypercholesterolemic rab-bits, G0A) and the other five at 14 months (control for the age ofthe reverted group, G0B). (2) Hypercholesterolemic rabbits(G1A; n¼ 10), rabbits fed the standard diet enriched with0.5% cholesterol (U.A.R., Paris, France) for 8 months. (3) Re-verted group (G1B; n¼ 10) rabbits fed a rabbit standard diet

Fig. 2. Thickening of the scleraechoroid complex (double arrow) in the study

groups. Semithin sections (toluidine blue). C, vascular layers of the choroid ;

SCS, suprachoroidea-sclera; G0, control group, standard diet; G1A, hypercho-

lesterolemic rabbits, standard diet enriched with 0.5% cholesterol for 8

months; and G1B, reverted rabbits, standard diet enriched with 0.5% choles-

terol for 8 months plus standard diet for another 6 months [Bar¼ 100 mm].

enriched with 0.5% cholesterol (U.A.R., Paris, France) for 8months, and then with the standard diet for another 6 months.

Weight was recorded immediately before the experimentand monitored monthly thereafter. The same schedule wasused to track serum values of total cholesterol, triglycerides,phospholipids, high density lipoproteins (HDL), low densitylipoproteins (LDL) and very low density lipoproteins(VLDL) (Table 1).

Blood samples were taken from the marginal vein of the earand analyzed with commercially available enzymatic kits fol-lowing the manufacturer’s instructions. Briefly, phospholipids,triglycerides and total cholesterol ester were hydrolyzed byenzymes, giving hydrogen peroxide, which was detectableby colorimetry. Lipoproteins were selectively separated byprecipitation with polymers (Bio Merieux, France) followedby detection by colorimetry. Comparison between groupswas performed by two-way ANOVA (Table 1).

The animals were killed with an overdose of sodium pento-barbital. Care and use of animals conformed to the ARVOguidelines for the use of animals in ophthalmic and visionresearch.

The thoracic aorta was removed for the histopathologystudy. Segments of this artery were fixed with 4% paraformal-dehyde in 0.1 M phosphate buffer (PB) pH 7.4 for 4 h at 4 �C.Then the tissues were dehydrated and embedded in paraffin.Tissue sections ranging from 6 to 10 mm were then stainedwith hematoxylineeosin.

Eyes (n¼ 60) were enucleated immediately after death andslit behind the limbus with a razor blade in order to facilitatepenetration of the fixative. For each animal, one of the eyeswas used for electron microscopy (n¼ 30).

2.2. Transmission electron microscopy (TEM)

Eyes were immersed in 2% glutaraldehyde in 0.1 M PB, pH7.4 at 4 �C for 5 h. After washing in 0.1 M PB, the wall of theposterior segment of the eyes was cut into small pieces compris-ing the sclera, choroid and all layers of the retina. These piecescame from either the medullated nerve fibre region (MNFR) orthe periphery of the retina and were post-fixed in 1% osmiumtetroxide in 0.1 M PB for 2 h at 4 �C. The tissues were thendehydrated in graded acetone and embedded in araldite. Thesemithin sections were stained with toluidine blue, and afterselection the blocks were further trimmed for ultramicrotomy(Reichert OM-V3 ultramicrotome, Leica, Germany). The thinsections were contrasted with 2% uranyl acetate in water andlead citrate (Reynolds, 1963), and examined by transmissionelectron microscopy (TEM; Zeiss 902 electron microscope).

3. Results

3.1. General results

Acceptance of the diet supplemented with cholesterol wasrapid and there were no significant differences among thegroups regarding daily consumption of the diet.

415J.J. Salazar et al. / Experimental Eye Research 84 (2007) 412e422

Fig. 3. Choroidal tissue. Semithin sections (toluidine blue). (A) G0: Control. (B) In hypercholesterolemic rabbits (G1A), the buildup of lipids in the suprachoroidea

compresses the choroidal vessels (arrows). (C) In reverted animals (G1B), there are far fewer lipids in the suprachoroidea and less compression of choroidal ves-

sels; however, there are abundant collagen fibers in the suprachoroidal spaces. B, Bruch’s membrane; C, collagen; CC, choriocapillaris; S, choroidal stroma; SC,

suprachoroidea; SCL, sclera; F, fibroblast; RPE, retinal pigment epithelium; (*), foam cells [AeC Bar¼ 12.5 mm].

Serum concentration of total cholesterol was43.3� 15.8 mg/100 ml at the beginning of the study. Therewas a statistical significant increase ( p< 0.001) in total serumcholesterol levels in G1A compared to G0. In G1B, the levelof total cholesterol (33.8� 6.2 mg/100 ml) was lower than inG0 ( p< 0.05) and G1A ( p< 0.01). The rest of the parametersstudied (triglycerides, phospholipids, VLDL, HDL, LDL andIDL) increased in G1A compared to G0. These values returnedto normal when cholesterol was withdrawn from the diet(G1B) (Table 1).

3.2. Thoracic aorta

In G1A there were abundant foam cells in the intima andendothelial proliferation, which reduced the vascular lumenwith respect to G0 (Fig. 1C); a thickening of the media wasalso observed (Fig. 1A). In comparison with G1A, the vascularlumen in G1B was less reduced because there was less infiltra-tion of the intima by foam cells (Fig. 1B).

3.3. Ocular studies

3.3.1. Optic microscopyIn the area corresponding to the MNFR, the scleraechoroid

complex registered, respectively, a threefold and a twofold in-crease in thickness among the rabbits in G1A and G1B

compared with G0 (no differences were found between G0Aand G0B). This feature was even more notable in the choroidalperiphery (Fig. 2).

In G1A a buildup of lipids was detected in the suprachoroi-dea, large- and medium-sized vessel layers. As a result of thelipids accumulated in the suprachoroidea, the thickness of thislayer increased and the lumens of the choroidal vessels werecompressed (Fig. 3B). The above-mentioned lipid buildupcompressed the choriocapillaris against the Bruch’s membraneand reduced the capillary lumens to the point of collapse insome instances (Fig. 3B).

Animals from G1B had far fewer lipids in the suprachoroi-dea, so that the thickness of this choroidal layer was dimin-ished and less compression was exerted on choroidal vesselas compared to G1A. However, the suprachoroidea thicknessof G1B (Fig. 3C) still increased with respect to G0(Fig. 3A) due to the abundance of collagen fibers.

In the suprachoroidea, the rabbits from G1A were found tohave clusters of lipid-charged macrophages (foam cells) whichwere surrounded by collagen fibers. In some instances, theseclusters of foam cells were encircled by cholesterol clefts(Fig. 4A). These features were also found to a lesser extent inanimals from G1B (Fig. 4B). In both groups, the intensity ofthese findings varied depending on the region of the choroid an-alyzed. In the peripheral choroid, they extended beyond thesuprachoroidea and were also present in the large- and medium

416 J.J. Salazar et al. / Experimental Eye Research 84 (2007) 412e422

sized vessel layers. However, in the choroid corresponding tothe MNFR, they were not present throughout the entire thick-ness of the choroid.

3.3.2. Transmission electron microscopy (TEM)In control groups no differences were found between G0A

and G0B. Overall, the rabbits from G1A and G1B exhibitedchanges with respect to G0 (Fig. 5). The intensity of choroidalchanges in G1B was not as uniformly distributed throughoutthe choroid as in G1A. In the former, modifications rangedfrom areas where the choroidal structure could be identified(Fig. 5C) to areas where the tissue was highly disorganized,making it difficult to differentiate the choroidal layers(Fig. 5D).

Fig. 4. Suprachoroidea in G1A and G1B. Semithin sections (toluidine blue).

(A) Foam cells (*) surrounded by collagen fibers (C) and encircled by cho-

lesterol clefts (arrows) in G1A. (B) Foam cells, collagen and cholesterol

clefts are also found in G1B. G1A, hypercholesterolemic rabbits, 0.5%

cholesterol-enriched diet; G1B, reverted rabbits, hypercholesterolemic dietþstandard diet.

As it was observed with light microscopy, the lumens of thechoroidal vessels of animals from G1A and G1B were reducedor collapsed (Fig. 6A, B) with respect to G0. This feature wasmore pronounced in G1A (Fig. 6A).

The suprachoroidea of G1A and G1B had foam cells, col-lagen and electron-lucent fusiform structures surrounded byor inside foam cells. The intensity of these changes varied de-pending on the group studied. Thus, foam cells and electron-lucent fusiform structures were numerous in G1A (Fig. 6C),and collagen predominated in G1B (Fig. 6D).

The most striking changes in the large-sized vessel layer inG1A were reduction of the vascular lumens (Fig. 7A, C and F)and hypertrophy of the muscle cells, which contained drops oflipids in their cytoplasm (Fig. 7B). Lumen reduction was dueto both compression caused by the buildup of lipids in the supra-choroidea and hypertrophy of endothelial cells, which also ex-tended cytoplasmic projections to the vascular lumen (Fig. 7A,C). In both groups G1A and G1B, the endothelial cells in thisvascular layer had electron-dense particles in the cytoplasm(Fig. 7C). Also, in G1B the intima and the media had foam cellsand collagen (Fig. 7D, E); in some instances, there was only col-lagen at the intima and among muscle cells (Fig. 7E).

In the medium-sized vessel layer of rabbits from G1A(Fig. 7F) and G1B, the basal membrane (BM) of endothelialand muscle cells was thicker than in the control. Additionally,the BM of endothelial cells contained electron-dense and elec-tron-lucent particles (Fig. 7F). The cytoplasm of the musclecells and the adventitia had drops of lipids in G1A (Fig. 7F)and electron-dense material in G1B. Also, the adventitia con-tained increased collagen with respect to controls and G1A.

The intervascular spaces at the large- and medium-sizedvessel layers in G1A and G1B contained more fibroblastsand collagen fibers than G0. However, in G1B the collagenfibers were more numerous and densely packed (Fig. 7E).

Endothelial cells of the choriocapillaris of the rabbits fromG1A and G1B presented changes with respect to the controls:some were hypertrophic, some presented irregularity in cellmorphology and others were necrotic (Fig. 8). On occasion,the endothelial cells of G1B bulged to the vascular lumen,considerably reducing its diameter (Fig. 8C) and extended cy-toplasmic projections into the outer collagen layer of Bruch’smembrane (Fig. 8C). In some cases these cytoplasmic projec-tions extended to the vicinity of the elastic layer (Fig. 8C). Inaddition, the BM of the endothelial cells of the choriocapillarisexhibited more collagen in G1A and G1B than in G0. This col-lagen increment was greater in G1B than in G1A (Fig. 8B).

Erythrocytes were generally found in the vascular walls inG1A and G1B (Fig. 5C, D; Fig. 8B).

4. Discussion

Rabbits fed with a 0.5% cholesterol-enriched diet for 8months (G1A) showed a statistical increase in total serum cho-lesterol levels compared to G0, and also an atheroscleroticlesion in the thoracic aorta. This study shows that when highlevels of cholesterol are withdrawn from the diet of rabbits(G1B), they recover some of the biochemical and histological

417J.J. Salazar et al. / Experimental Eye Research 84 (2007) 412e422

Fig. 5. Choroidal tissue, TEM. (A) G0: Control group. (B) In G1A, the choroidal layers can be identified. (CeD) In G1B, modifications in the choroid range from

areas where the choroidal structure can be identified (C) to areas where the tissue is highly disorganized (D), making it difficult to differentiate the choroidal layers.

Red blood cells (*) outside the vascular lumens. G1A, hypercholesterolemic rabbits, 0.5% cholesterol-enriched diet; G1B, reverted rabbits, hypercholesterolemic

dietþ standard diet; B, Bruch’s membrane; BM, basal membrane; CC, choriocapillaris; E, endothelial cell; M, muscle cell; mac, macrophage; RPE, retinal

pigment epithelium; S, choroidal stroma; TEM, transmission electron microscopy [A, bar¼ 5 mm; B, bar¼ 2.2 mm; CeD, bar¼ 5 mm].

parameters that are altered in cholesterol-fed animals (Sasoet al., 1992). That is, the biochemical parameters studiedwere augmented in cholesterol-fed rabbits, but when a standarddiet was administered to the G1B group, these values returnedto normal. However, histological parameters did not totallyrecover (choroid and thoracic aorta).

There have been few experimental studies on the effects ofhypercholesterolemia on the posterior segment of the eye(Francois and Neetens, 1966; Yamakawa et al., 2001; Bhuttoand Amemiya, 2001; Ong et al., 2001; Li et al., 2002; Kouchiet al., 2006; Trivino et al., 2006; Ramırez et al., 2006). Someof the alterations described in the choroidal vessels in the pres-ent work have been reported in rats (Yamakawa et al., 2001;

Bhutto and Amemiya, 2001) and in Watanabe heritable hyper-lipidemic rabbits (Kouchi et al., 2006). However, to the best ofour knowledge, no report to date provides a detailed descrip-tion of ultrastructural changes induced by hypercholesterol-emia in the choroid of New Zealand White rabbit, as well aschanges after normalization of serum lipid levels.

One of the most striking histological findings in hypercho-lesterolemic rabbits (G1A) was the buildup of lipids in thesuprachoroidea, a fact reported in the literature (Francoisand Neetens, 1966; Vinger and Sachs, 1970) and which causeda thickening of this ocular layer. It is thought that the vascularnature of this tissue make it the preferred site for lipid deposi-tion (Francois and Neetens, 1966). This thickening could

418 J.J. Salazar et al. / Experimental Eye Research 84 (2007) 412e422

Fig. 6. Choroidal tissue, TEM. Lumen reduction of choroidal vessels in G1A (A) and G1B (B). The adventitia of rabbits from G1B are thick (B). Foam cells (*),

electron-dense particles (arrow head), collagen and electron-lucent fusiform structures (arrows) in the suprachoroidea of G1A (C) and G1B (D). In G1B (D), the

collagen predominates over the foam cells and electron-lucent fusiform structures. G1A, hypercholesterolemic rabbits, 0.5% cholesterol-enriched diet; G1B, re-

verted rabbits, hypercholesterolemic dietþ standard diet; Ad, adventitia; C, collagen; E, endothelial cell; M, muscle cell; L, vascular lumen; arrow heads, lipids;

TEM, transmission electron microscopy [A, bar¼ 2 mm; B, bar¼ 2 mm; C, bar¼ 5 mm; D, bar¼ 5 mm].

cause a straightening of the vessels passing through it and a re-duction of their lumens. Additionally, the thickened suprachor-oid exerts mechanical compression on the choroidal vascularlayers which could reduce the choroidal blood flow. The lawof Poiseulle states that tangential tension is directly propor-tional to the blood flow viscosity and inversely proportionalto the third power of the inside radius (Corti et al., 2002). Ac-cordingly, reduction of the choroidal vascular lumens in hyper-cholesterolemic rabbits could induce an increase in the localtangential force exerted by the choroidal blood flow on thevessel walls. With time, these changes in blood flow couldcause a reorganization of the endothelial phenotype, resultingin endothelial dysfunction (Topper and Gimbrone, 1999). Thiscould explain, at least in part, the endothelial alterations ob-served in the present study.

Endothelial changes and BM thickening of ocular vessels inhypercholesterolemic rabbits may precede the atheroscleroticprocess. In this process, the adhesion molecules VCAM-1and ICAM-1 favor monocyte binding and migration to thesubendothelial space, where they transform into macrophages(Ross, 1999; Crispin, 2002). The internalization of lipidsby macrophages results in the formation of foam cells, asdescribed by Wu et al. (1995) in iris arterioles of

hypercholesterolemic rabbits. We observed foam cells at thesuprachoroid and large- and medium-sized vessel layers ofrabbits from G1A and G1B. Before and after death, foam cellsmay release products such as cholesterol (oxidized and ester-ified) that increase endothelial damage, thus favoring the ath-erosclerotic lesion (Galceran and Martinez, 2000; Paramoet al., 2001). In our results, the foam cells underneath the basalmembrane of the large-sized vessel layer, where the vesselshave a diameter of up to 90 mm (Hogan et al., 1971), couldrepresent the microatheromas described in brain circulation(Sanchez-Perez et al., 1999). In the medium-sized vessel layer,arterioles with diameters ranging from 20 to 40 mm (Hoganet al., 1971), combined with the increased collagen observedby us in the adventitia and the deposits of lipids (electron-dense vesicles) inside muscle cells, could produce arteriolarhyalinosis (Sanchez-Perez et al., 1999).

The cholesterol clefts that surrounded the clusters of foamcells observed in G1A and G1B were similar to cholesterolclefts described in the anterior segment of hypercholesterol-emic rabbits (Sebesteny et al., 1985). The thickening of the in-tima and adventitia observed in the choroidal vessels wasmainly due to an increase of collagen fibers and could havecontributed to a reduction of the vascular tone.

419J.J. Salazar et al. / Experimental Eye Research 84 (2007) 412e422

Fig. 7. Large- and medium-sized vessel layers of the choroids, TEM. The reduction of the vascular lumens in the large-sized vessel layer is more intense in G1A

(A, C, F) than in G1B (D, E). In G1A, the muscle cells of this choroidal layer are hypertrophic and contain drops of lipids (black arrowheads) in the cytoplasm (B).

Additionally, the endothelial cells are hypertrophic and extend cytoplasmic projections to the vascular lumen, contributing to its reduction (A, C). Electron-dense

particles (black arrowheads), foam cells (*) (D) and collagen (E) in the intima and media of rabbits from G1B (insert: collagen fibres at higher magnification).

Medium-sized vessel layer of G1A. Thickening of the BM of the muscle cells (white arrowheads) and electron-dense and electron-lucent particles (black arrow-

heads) in the BM of the endothelial cells in G1A (F). G1A, hypercholesterolemic rabbits, 0.5% cholesterol-enriched diet; G1B, reverted rabbits, hypercholester-

olemic dietþ standard diet Ad, adventitia; BM, basal membrane; C, collagen; E, endothelial cell; (*), foam cells; M, muscle cell; L, vascular lumen; black arrow

heads, lipids; TEM, transmission electron microscopy [AeC, bar¼ 5 mm; D, bar¼ 2.5 mm; E, bar¼ 2 mm (insert, bar¼ 0.5 mm); F, bar¼ 2.2 mm].

Changes in the vessel walls of the large- and medium-sizedvessel layers and increased thickening of the BM of the retinalvessels have been described in hypercholesterolemic rats(Yamakawa et al., 2001; Bhutto and Amemiya, 2001). Addition-ally, increasing of the BM of the retinal vessels has also beendescribed in hypercholesterolemic rabbits (Trivino et al., 2006).

In the choroid, the main difference between G1B andG1A was a marked reduction of suprachoroidal buildup of

lipids, implying less compression of the vessels. This findinghas also been reported in iris and cornea of hypercholester-olemic reverted rabbits (Rodger, 1972). However, althoughin most cases no lipids were visible in the suprachoroidalspaces in G1B, there were still numerous changes detectableby TEM in comparison with G1A and G0. These changescould contribute to chronic ischemia, due to the followingsituations.

420 J.J. Salazar et al. / Experimental Eye Research 84 (2007) 412e422

(a) The intervascular spaces of the choroid contained fewerlipids than in G1A but more collagen fibers than in thecontrols, which could contribute to the increased choroidalthickness observed. It is well known that the reabsorptionof both esterified and free cholesterol is followed by in-tense sclerogenic activity (Abdulla et al., 1967), due tothe capacity of cholesterols and their esters to induce in-flammation. Some inflammatory cytokines (TNFa andIL-1) control the remodeling activity of macrophages

Fig. 8. Endothelial cells of the choriocapillaris, TEM. Irregularities in cellular

morphology (arrow) (A); necrosis (arrowhead) (A, B); and hypertrophy, H (C).

The endothelial cells of G1B bulge to the vascular lumen reducing their diam-

eter (H) and extend cytoplasmic projections (open arrow) to the outer collagen

layer of Bruch’s membrane (C). Red blood cells (*) outside the vascular lumen

(B). (A) G1A, hypercholesterolemic rabbits, 0.5% cholesterol-enriched diet.

(B, C) G1B, reverted rabbits, hypercholesterolemic dietþ standard diet.

B, Bruch’s membrane, C, collagen; e, elastic layer of the Bruch’s membrane;

L, vascular lumen; oc, outer collagen layer of the Bruch’s membrane;

RPE, retinal pigment epithelium; TEM, transmission electron microscopy

[A, bar¼ 3.4 mm; B, bar¼ 1 mm; C, bar¼ 50 mm].

and smooth muscle cells, which have the capacity to pro-duce several matrix metalloproteinases (MMPs) (Henneyet al., 1991; Galis et al., 1994). It has been observedthat such MMPs are mainly synthesized where the concen-tration of foam cells is greater. A relationship betweenlipid uptake and MMP activity has also been reported(Galis et al., 1995). The observations cited underlinethe important connection between inflammation, tissueremodeling and lipids metabolism.

(b) In both groups (G1A and G1B), greater buildup of lipidswere found on the periphery of the choroid than in thechoroid corresponding to the medullary region of the ret-ina, possibly because the vascular density is anatomicallyless on the choroidal periphery. In this region in G1B, al-though there were less buildup of lipids than in G1A, therewere groups of foam cells surrounded by collagen fibersand fibroblasts which could represent atherosclerotic mi-croplaques similar to those described in the sclerocorneallimbus of hypercholesterolemic dogs affected by lipidickeratopathy (Linton et al., 1993)

(c) The large- and medium-sized vessel layers were lesscompressed than in G1A but exhibited BM and endothe-lial changes similar to those in hypercholesterolemicanimals.

(d) The endothelial cells of the choriocapillaris extended cyto-plasmic projections that reduced the vascular lumen.

In conclusion, in New Zealand White Rabbits the replace-ment of a hyperlipidemic diet by a standard one normalizedserum lipid levels and eliminated most buildup of lipids atthe posterior segment of the eye. However, this lipid reductionwas not followed by a reversal of the changes in choroidal ves-sels, where persisting ultrastructural changes were compatiblewith a chronic ischemia. These changes, together with the al-terations found in the retinal vessels and Bruch’s membrane(Ramırez et al., 2006), could account for the progressivedegeneration found in the retinal tissue in reverted rabbits(Ramırez et al., 2006).

Acknowledgments

The research was supported by Comunidad de MadridGrant 08.4/0017.1/99, Universidad Complutense de MadridGrant PR48/01-9905, Instituto Salud Carlos III Grant RedesTematicas de Investigacion C03/01 and C03/13, FISS- PI-041018. We wish to thank Agustın Fernandez and the Centrode Microscopia Electronica ‘‘Luis Bru’’ (Universidad Com-plutense de Madrid) for technical assistance in electron mi-croscopy; Alistair Ross and David Nesbitt for their linguisticassistance.

References

Abdulla, Y.H., Adams, C.W.M., Morgan, R.S., 1967. Connective tissue reac-

tions to implantation of purified sterol, sterol esters, phosphoglycerides

and free fatty acids. J. Pathol. Bacteriol. 94, 63e71.

421J.J. Salazar et al. / Experimental Eye Research 84 (2007) 412e422

Bhutto, I.A., Amemiya, T., 2001. Corrosion casts and scanning electron mi-

croscopy of choroidal vasculature in rats with inherited hypercholesterol-

emia. Curr. Eye Res. 23, 240e247.

Caprioli, J., Sears, M., Miller, J.M., 1987. Patterns of early visual field loss in

open angle glaucoma. Am. J. Ophthalmol. 103, 512e517.

Corti, R., Badimon, L., Fuster, V., Badimon, J.J., 2002. Endotelio, flujo y

aterotrombosis. In: Fuster, V. (Ed.), Valoracion y modificacion de la

placa aterosclerotica vulnerable. Futura Publishing Company Inc.,

Malden, pp. 135e148.

Crispin, S., 2002. Ocular lipid deposition and hyperlipoproteinaemia. Progr.

Ret. Eye Res. 21, 169e224.

Daugherty, A., Zweifel, B.S., Schonfeld, G., Daugherty, A., Zweifel, B.S.,

1998. Probucol attenuates the development of aortic atherosclerosis in cho-

lesterol-fed rabbits. Br. J. Pharmacol. 2, 612e618.

Del Rio, M., Chulia, T., Merchan-Perez, A., Remezal, M., Valor, S.,

Gonzalez, J., Gutierrez, J.A., Lasuncion, M.A., Tejerina, T., 1995. Effects

of indapamide on Atherosclerosis development in cholesterol-fed rabbits.

J. Cardiol. Pharmacol. 25, 973e978.

Finking, G., Hankeh, H., 1997. Nikolaj Nikolajewitsch Anitschkow (1885e1964) established the cholesterol-fed rabbit as a model for atherosclerosis

research. Atherosclerosis 135, 1e7.

Francois, J., Neetens, A., 1966. Vascular manifestations of experimental hy-

percholesteraemia in rabbits. Angiologica 3, 1e20.

Fuster, V., Badimon, L., Badimon, J.J., Chesebro, J.H., 1992. The pathogenesis

of coronary artery disease and the acute coronary syndromes. N. Engl. J.

Med. 326, 310e318.

Galceran, J.M., Martinez, A., 2000. Fenomenos oxidativos en la fisiopatologıa

vascular. Hipertension 17, 17e21.

Galis, Z.S., Sukhova, G.K., Lark, M.W., Libby, P., 1994. Increased expres-

sion of matrix metalloproteinases and matrix degrading activities in

vulnerable regions of human atherosclerotic plaques. J. Clin. Investig.

94, 2493e2503.

Galis, Z.S., Sukhova, G.K., Kranzhofer, R., 1995. Macrophage foam cells from

experimental atheroma constitutively produce matrix degrading protein-

ases. Proc. Natl. Acad. Sci. U.S.A. 92, 402e440.

Greve, E.C., Geijssen, H.C., 1983. The relation between excavation and visual

field in glaucoma patients with high and with low intraocular pressures.

Doc. Ophthalmol. 35, 35e40.

Hayreh, S.S., 1970. Pathogenesis of visual field defects. Role of the ciliary cir-

culation. Brit. J. Ophthalmol. 54, 289e311.

Hayreh, S.S., Revie, I.H.S., Edwards, J., 1970. Vasogenic origin of visual field

defects and optic nerve changes in glaucoma. Brit. J. Ophthalmol. 54,

461e472.

Hayreh, S.S., 1971. Posterior ciliary arterial occlusive disorders. Trans.

Ophthalmol. Soc. U.K. 91, 291e303.

Henney, A., Wakeley, P., Davies, M.J., 1991. Localization of stromelysin gene

expression in atherosclerotic plaques by in situ hybridization. Proc. Natl.

Acad. Sci. U.S.A. 88, 8154e8158.

Hogan, M.J., Alvarado, J.A., Weddell, J.E., 1971. Histology of the Human

Eye. WB Saunders Company, Philadelphia (PH), pp. 320e392.

Huff, M.W., Carrol, K.K., 1980. Effects of dietary protein on turnover, oxida-

tion, and absorption of cholesterol, and on steroid excretion in rabbits.

J. Lipid Res. 21, 546e548.

Jimenez, A.M., Millas, I., Farre, J., Garcıa-Mendez, A., Jimenez, P.,

Arriero, M.M., Garcıa-Colis, E., Andres, R., Gomez, J., Casado, S.,

Lopez-Farre, A., 2002. Efecto de la inhibicion de la HMG-CoA reductasa

sobre la proteına inductora de disfuncion edotelial en conejos hipercoles-

terolemicos. Rev. Esp. Cardiol. 55, 1151e1158.

Jonas, J.B., Fernandez, M.C., Naumann, G.O.H., 1990. Glaucomatous optic

nerve atrophy in small discs with low cup-to-disc ratios. Ophthalmology

97, 1211e1215.

Jonas, J.B., Fernandez, M.C., Naumann, G.O.H., 1992. Glaucomatous peripa-

pillary atrophy. Occurrence and correlations. Arch. Ophthalmol. 110,

214e222.

Klein, R.A., Sharrett, A.R., Klein, B.E.K., Chambless, L.E., Cooper, L.S.,

Hubbard, L.D., Evans, G., 2000. Are retinal arteriolar abnormalities relates

to atherosclerosis? The atherosclerosis risk in communities study. Aterios-

cler. Thromb. Vasc. Biol. 20, 1644e1650.

Kouchi, M., Ueda, Y., Horie, H., Tanaka, K., 2006. Ocular lesions in Watanabe

heritable hyperlipidemic rabbits. Vet. Ophthalmol. 9, 145e148.

Kubota, T., Schlotzer-Schrebardt, U.M., Naumann, G.O.H., Kohno, T.,

Inomata, H., 1996. The ultrastructure of parapapillary chorioretinal atro-

phy in eyes with secondary angle-closure glaucoma. Graefe’s. Arch.

Clin. Exp. Ophthalmol. 234, 351e358.

Li, C.M., Bradley, K., Chung, B.H., Millican, C.L., Curcio, C.A., 2002. Absence

of Bruch’s Membrane (BrM) cholesterol deposition in rabbits consuming

atherogenic diets. Investig. Ophthalmol. Vis. Sci. 43. E-Abstract 2813.

Linton, L.L., Moore, C.P., Collier, L.L., 1993. Bilateral lipid keratopathy in

a boxer dog: cholesterol analysis and dietary management. Prog. Vet.

Comp. Ophthalmol. 3, 9e14.

Oorni, K., Pentikainen, M.O., Ala-Korpela, M., Kovanen, P.T., 2000. Aggrega-

tion, fusion, and vesicle formation of modified low density lipoprotein par-

ticles: molecular mechanisms and effects on matrix interactions. J. Lipid

Res. 41, 1703e1714.

Ong, J.M., Zorapapel, N.C., Rich, K.A., Wagstaff, R.E., Lambert, R.W.,

Rosenberg, S.E., Moghaddas, F., Pirouzmanesh, A., Aoki, A.M.,

Kenney, C., 2001. Effects of cholesterol and apolipoprotein E on retinal

abnormalities in ApoE-deficient mice. Investig. Ophthalmol. Vis. Sci.

42, 1891e1900.

Paramo, J.A., Orbe, M.J., Rodriguez, J.A., 2001. Papel de los antioxidantes

en la prevencion de la enfermedad cardiovascular. Med. Clin. 116,

629e635.

Phelps, C.D., Hayreh, S.S., Montague, P.R., 1983. Visual field in low tension

glaucoma, primary open angle glaucoma, and anterior ischemic optic neu-

ropathy. Doc. Ophthalmol. 35, 113e128.

Ramırez, A.I., Salazar, J.J., De Hoz, R., Rojas, B., Ruiz, E., Tejerina, T.,

Ramırez, J.M., Trivino, A., 2006. Macroglial and retinal changes in hyper-

cholesterolemic rabbits after normalization of cholesterol levels. Exp. Eye

Res. 83, 1423e1438.

Reynolds, E.S., 1963. The use of lead citrate at high pH as an electron-opaque

stain in electron microscopy. J. Cell Biol. 17, 208e212.

Rodger, F.C., 1972. A new preparation for the study of experimental athero-

sclerosis progressive-regressive changes in albino rabbit iris. Exp. Eye

Res. 14, 1e6.

Rong, J.X., Shen, L., Chang, Y.H., Richters, A., Hodis, H.N., Sevanian, A.,

1999. Cholesterol oxidation products induce vascular foam cell lesion for-

mation in hypercholesterolemic New Zealand white rabbits. Arterioscler.

Thromb. Vasc. Biol. 19, 2179e2187.

Ross, R., 1993. The pathogenesis of atherosclerosis: a perspective for the

1990s. Nature 362, 801e808.

Ross, R., 1999. Atherosclerosis-an inflammatory disease. N. Engl. J. Med. 340,

115e126.

Sanchez-Perez, R.Ma., Molto, J.M., Medrano, V., Beltran, I., Dıaz-Marın, C.,

1999. Aterosclerosis y circulacion cerebral. Rev. Neurol. 28, 1109e1115.

Saso, Y., Kitamura, K., Yasoshima, A., Iwasaki, H.O., Takashima, K., Doi, K.,

Morita, T., 1992. Rapid induction of atherosclerosis in rabbits. Histol. His-

topathol. 7, 315e320.

Schulzer, M., Drance, S.M., Carter, C.J., Brooks, D.E., Douglas, G.R.,

Lau, W., 1990. Biostatistical evidence for two distinct chronic open angle

glaucoma populations. Brit. J. Ophthalmol. 74, 196e200.

Sebesteny, A., Sheraidah, G.A.K., Trevan, D.J., Alexander, R.A., Ahmed, A.I.,

1985. Lipid keratopathy and atheromatosis in an SPF laboratory rabbit col-

ony attributable to diet. Lab. Anim. 19, 180e188.

Terres, G., Ramırez, J.M., Trivino, A., Gutierrez, J.A., Rueda, A.,

Avellaneda, A., Cascio, C., 1992. Early visual field defects in hypercholes-

terolaemic patients. In: Halpern, M.J. (Ed.), Molecular Biology of Athero-

sclerosis. John Libbey & Company Ltd., London, pp. 353e362.

Topper, J.N., Gimbrone Jr., M.A., 1999. Blood flow and vascular gene expres-

sion: fluid shear stress as a modulator of endothelial phenotype. Mol. Med.

Today 5, 40e46.

Trivino, A., Ramırez, A.I., Salazar, J.J., De Hoz, R., Rojas, B., Padilla, E.,

Tejerina, T., Ramırez, J.M., 2006. A cholesterol-enriched diet induces

ultrastructural changes in retinal and macroglial rabbit cells. Exp. Eye

Res. 83, 357e366.

Tuulonen, A., Airaksinen, J., 1992. Optic disc size in exfoliative, primary open

angle and low tension glaucoma. Arch. Ophthalmol. 110, 211e213.

422 J.J. Salazar et al. / Experimental Eye Research 84 (2007) 412e422

Vinger, P.F., Sachs, B.A., 1970. Ocular manifestations of hiperlipoproteinemia.

Am. J. Ophthalmol. 70, 563e573.

Wong, T.Y., Klein, R., Couper, D.J., Cooper, L.S., Shahar, E., Hubbard, L.D.,

Wofford, M.R., Sharrett, A.R., 2001. Retinal microvascular abnormalities

and incident stroke: the atherosclerosis risk in communities study. Lancet

358, 1134e1140.

Wong, T.Y., Klein, R., Sharrett, A.R., Couper, D.J., Klein, B.E.K.,

Lao, D.P., Hubbard, L.D., Mosley, T.H., 2002a. Cerebral white

matter lesions, retinopathy, and incident clinical stroke. JAMA 288,

67e74.

Wong, T.Y., Klein, R., Sharrett, A.R., Duncan, B.B., Couper, D.J., Tielsh, J.M.,

Klein, B.E.K., Hubbard, L.D., 2002b. Retinal arteriolar narrowing and risk

of coronary heart disease in men and women. The atherosclerosis risk in

communities study. JAMA 287, 1153e1159.

Wu, Ch., Chang, S., Chen, M., Lee, Y., 1995. Early change of vascular perme-

ability in hypercholesterolemic rabbits. Arterioscler. Thromb. Vasc. Biol.

15, 529e533.

Yamakawa, K., Bhutto, I.A., Lu, Z., Watanabe, Y., Amemiya, T., 2001. Retinal

vascular changes in rats with inherited hypercholesterolemia e Corrosion

cast demonstration. Curr. Eye Res. 22, 258e265.

Yanni, A.E., 2004. The laboratory rabbit: an animal model of atherosclerosis

research. Lab. Anim. 38, 246e256.

Zauberman, H., Livni, N., 1981. Experimental vascular occlusion in hypercho-

lesterolemic rabbits. Investig. Ophthalmol. Vis. Sci. 2, 248e255.