evaluation of toxicity due to vital stains in isolated rat retinas

Evaluation of toxicity due to vitalstains in isolated rat retinas

Kazuhiro Tokuda,1 Teruhisa Tsukamoto,2 Shigeki Fujisawa2 and

Masao Matsubara1

1Department of Ophthalmology, Daini Hospital, Tokyo Women’s Medical University,

Tokyo, Japan2Ako Research Institute, Division of Dermatologicals and Ophthalmologicals, Otsuka

Pharmaceutical Company Ltd, Hyogo, Japan

ABSTRACT.

Purpose: We investigated whether artificial aqueous humours are adequate

incubation media compared with artificial cerebrospinal fluid (aCSF), and eval-

uated the retinal toxicity of two vital stains � trypan blue (TB) and indocyanine

green (ICG) � and triamcinolone acetonide (TA) using isolated rat retinas

incubated in artificial aqueous humours.

Methods: In experiment 1, retinal segments were isolated and incubated in aCSF,

BSS plus1, Opeguard1Neo Kit, or phosphate-buffered saline (PBS). In experiment

2, retinal tissues were exposed to one of the agents and incubated in BSS plus1.

Retinal damage was assessed by morphological examination and biochemical assay,

which measured lactate dehydrogenase (LDH) in the medium once every hour.

Results: In experiment 1, BSS plus1 was confirmed as a suitable incubation

medium. In experiment 2, there were no significant changes in the retinas exposed

to TB or TA. Tissues exposed to ICG showed damage in every retinal layer and

significantly higher release of LDH.

Conclusion: Exposure to ICG caused retinal damage in isolated rat retina tissue

in our experimental model (in vitro).

Key words: vital stain – indocyanine green – trypan blue – triamcinoline acetonide – lactate

dehydrogenase – histology

Acta Ophthalmol. Scand. 2004: 82: 189–194Copyright # Acta Ophthalmol Scand 2004.

doi: 10.1111/j.1395-3907.2004.00224.x

Introduction

Vital stains have been widely used invitro in retinal disease during vitrect-omy. These stains allow us to visualizethe transparent vitreous body or retina(Da Mata et al. 2001; Feron et al. 2002;Sakamoto et al. 2002). Their retinaltoxicity has been evaluated mainly interms of changes in the retinal structureand by electroretinogram (Enaida et al.2002). However, there have been no

studies determining the retinal toxicityof different stains from a histopatholo-gical or biochemical viewpoint.

We evaluated the retinal toxicity oftwo vital stains – indocyanine green(ICG) and trypan blue (TB) – andtriamcinolone acetonide (TA), cur-rently used to detect the internal limit-ing membrane (ILM), vitreous body orpreretinal membrane, using isolated ratretinas. As rat retinal tissues were incu-bated in an artificial cerebrospinal fluid

(aCSF) medium for the study of neuraldamage in previous reports (Izumi et al.1995; Romano et al. 1995), we firstcompared artificial aqueous humourswith aCSF to evaluate them as incuba-tion media before determining theinfluence of the stains. Using artificialaqueous humours as the incubationmedia, we performed staining at themaximal concentration for clinical usein the second experiment. Retinaldamage was assessed not only by mor-phological (light microscopy) examina-tion but also by biochemical assay,measuring the amount of lactate dehy-drogenase (LDH) released from thedamaged cells.

Material and Methods

Retinal isolation and incubation conditions

The retinas were prepared from 8-week-old male Sprague-Dawley ratsbased on the methods described byIzumi et al. (1995). The rats were anaes-thetized with pentobarbital sodium at adose of 60mg/kg body weight anddecapitated. The left eyes were removedfrom the orbits and placed in ice-coldaCSF containing 124mM NaCl, 5mMKCl, 2mM MgSO4, 2mM CaCl2,1.25mM NaH2PO4, 22mM NaHCO3

and 10mM glucose bubbled with 95%O2 : 5% CO2 to adjust it to pH7.4 inadvance. The sclera was cut circumfer-entially with microscissors to removethe cornea, lens, iris and vitreous. Theempty eyeball was divided into four

ACTA OPHTHALMOLOGICA SCANDINAVICA 2004

189

segments. The sensory retinal tissuewas then gently detached from the ret-inal pigment epithelium with smooth-tipped forceps. These segments weredivided into groups so that each groupincluded four segments from four rats.These isolation techniques wereemployed under a microscope in aPetri dish filled with aCSF maintainedat an ice-cold temperature to avoidneurodegenerative change in the retinaltissues. These retinal tissues were incu-bated in bubbled aCSF at 37 � for a 1-hour recovery period. After recovery,the buffer was washed twice with aCSF.Experiment 1

After recovery, the retinal piece wasplaced in a 1.5ml solution of variousmedia, which was oxygenated with95% O2 : 5% CO2 for 2min beforeuse, in a sealed 6.0ml glass vial. Themedia included aCSF (287mOsm/kg,pH7.6), BSS plus1 (305mOsm/kg,pH7.6) (Alcon Laboratories, FortWorth, Texas, USA), Opeguard1 NeoKit 294mOsm/kg, pH7.7 (OtsukaPharmaceutical Co. Ltd, Tokyo,Japan), and phosphate-buffered saline(PBS) 244mOsm/kg, pH7.1.

Experiment 2

After recovery, the retinal tissues wereplaced in one of the three vital stains(100ml) and were exposed to room tem-perature (25 �) for 1min, then rinsedtwice with BSS plus1 (25 �) (AlconLaboratories), and incubated in 1.5mlfresh BSS plus1. Each vital stain wasprepared as follows. The ICG solution(25 �, pH7.8) was prepared by

dissolving 25mg ICG powder (Dai-ichiPharmaceutical Co. Ltd, Tokyo, Japan)in 1ml sterile distilled water and adding4.0ml BSS plus1 to make a final ICGconcentration of 0.5%. The TB solu-tion (25 �, pH 7.5), consisting of0.6mg/ml (0.06%) TB (Wako PureChemical Industries Ltd, Osaka,Japan), was prepared with BSS plus1.The TA solution (25 �, pH 8.0), consist-ing of A-triamcinolone (Kenacort-A1;Bristol Pharmaceuticals KK, Tokyo,Japan) was kept upright for 30min, andthe triamcinolone vehicle was discardedas previously described (Peyman et al.2000). The remaining triamcinolonewas mixed with 5ml BSS plus1.

During retinal incubation in experi-ments 1 and 2, 95% O2 :5% CO2 wasprovided in the atmosphere in a sealedvial, and the vial was incubated in awater bath maintained at 37 �. Samples(25ml) for LDH measurement weretaken from each medium every 60minfor 5hours. The retinal pieces werecrushed by sonication (Model 250Dsonifier, Branson Ultrasonics Corpor-ation, Danbury, Connecticut, USA) atpostincubation to obtain a total LDHvalue. The light level during the studywas 700–800 lux.

The research was carried out in accor-dance with the Principles of LaboratoryAnimal Care (NIH publication no.85–23, revised 1985), the Office forProtection from Research Risks PublicHealth Service Policy on the HumaneCare and Use of Laboratory Animals(revised 1986), and the US AnimalWelfare Act, as amended.

Lactate dehydrogenase assay

Lactate dehydrogenase activity wasmeasured spectrophotometricallyusing an LDH-Cytotoxic Test (WakoPure Chemical Industries Ltd) (mea-sured at 570 nm). The assay was per-formed with 50 ml (dilated twice) inindividual wells of a 96-well plateusing a kinetic plate reader (MultiscanMS, Labsystems, Helsinki, Finland).The total amount of LDH was calcu-lated with the addition of:

(1) the sum of LDH from the sam-ples each hour, and

(2) LDH obtained from the crushedtissue at postincubation.

To control the difference in size ofindividual retinal segments, LDHrelease into the medium at each pointwas normalized to a percentage of thetotal LDH.

Histology

The retinal tissues were fixed in PBScontaining 10% formalin neutral buffersolution (Wako Pure Chemical Indus-tries Ltd) overnight at room tempera-ture. The fixed tissue was rinsed in tapwater, dehydrated with alcohol andxylene, embedded in paraffin, sectionedat 2mm and stained with haematoxylin-eosin.

Statistics

Statistical analysis between the groupsover the period was performed by com-puter. This process was repeated forsignificance of LDH release. The statis-tical significance of LDH release

Fig. 1. Lactate dehydrogenase (LDH)

release from rat retinal tissues incubated

in various media – artificial cerebrospinal

fluid (aCSF), BSS plus1, Opeguard1 Neo

Kit, and phosphate-buffered saline (PBS)

– after recovery. Lactate dehydrogenase

in the medium was measured spectro-

photometrically. There was a significant

difference between aCSF, BSS plus1,

Opeguard1 Neo Kit, and PBS at all hours

after incubation up to 5hours. However,

no differences were found between the first

three (n¼ 3–4) (þ 1hour, þ2hours,

þ3 hours, þ4 hours p< 0.05; þ5 hours,

p< 0.01, t-test compared with PBS).


190

between the groups at every hour wasperformed using a t-test. A p-value of<0.05 was considered significant.

Results

In experiment 1, there was a significantdifference in LDH release into the med-ium between PBS and aCSF, BSSplus1, and Opeguard1 Neo Kit at anyhour after the incubation of the isolatedrat retina up to 5 hours. Figure 1 showsresults at þ1 hour,þ 2hours,þ 3hours,þ4hours (all p< 0.05) and at þ5 hours(p< 0.01) (t-test compared with PBS).However, no statistical difference inLDH release was found between aCSFand BSS plus1 or Opeguard1 Neo Kitas a result of incubation. Histologicalfindings showed mild somatic and neuri-tic swelling in the retinal layers incu-bated in aCSF for 3hours afterrecovery. Although there was little dif-ference in BSS plus1 or in Opeguard1

Neo Kit compared to aCSF, somaticswelling in the inner nuclear layer andspongiform changes in the inner plexi-form layer were significant in PBS(Fig. 2, A�E). Based on both biochem-ical and histological results, BSS plus1

was used as an incubation media in thesubsequent experimental study.

In experiment 2, the retinal segmentswere exposed to one of the three stainsand then incubated in BSS plus1. Fig-ure 3A�C shows that the retinasexposed to 0.06% TB solution and TAsolution indicated slight vacuolizationin the ganglion cell layer and innernuclear layer compared to the controlsamples. However, there was no signifi-

cant change in LDH release incubatedwith the TB solution and the TA solu-tion compared to BSS plus1 as a con-trol (Fig. 4 ). By contrast, the tissuesexposed to a 0.5% ICG solutionshowed damage in every retinal layer(Fig. 5). The retinal segments exposedto ICG showed a significantly higherrelease of LDH compared to thoseexposed to 4ml BSS plus1 and 1mldistilled water as a control (Fig. 6)(p< 0.01, t-test compared with BSSplus1 and distilled water).

Discussion

Experimental model and incubation

medium

Isolated chick embryo retinas (Zeek-valk & Nicklas 1990) and rat retinas(Izumi et al. 1995) have been used toevaluate neurodegeneration. In pre-vious reports, retinal tissues were incu-bated in aCSF. As artificial aqueoushumours are used clinically in intraocu-lar surgery, we compared two products,BSS plus1 (Alcon Laboratories, FortWorth, Texas, USA) and Opeguard1

Neo Kit (Otsuka Pharmaceutical Co.Ltd, Tokyo, Japan), which are com-monly used as artificial aqueoushumours in Japan, with aCSF inexperiment 1 before evaluating the ret-inal toxicity of several kinds of vitalstain.

Measuring LDH from damaged reti-nas has been used as an index of neuraldamage (Romano et al. 1995). We mea-sured the LDH present in the mediumat various times for an objective analy-sis of retinal damage. A percentage of

the total LDH in the medium at eachpoint was described to normalize thedifferent size of the divided retinalpieces. In experiment 1, these two pro-ducts showed no difference comparedto each other or as compared toaCSF. Our study confirmed the suit-ability of artificial aqueous humoursas incubation media for rat retinas.Based on these results, a subsequentstudy (experiment 2) was conductedwith BSS plus1 under biochemical con-ditions as close to in vivo as possible.

Aim of the vital stain and its ocular

toxicity

In experiment 2, we used BSS plus1 asan incubation medium and evaluatedthe toxicity of vital stains. They areused to identify transparent retinas orvitreous bodies to safely and effectivelyperform vitrectomy. Indocyanine greenis a dye with an absorbency of 790�815 nm (peak 805 nm) used for fun-dus angiography as well as for stainingthe anterior capsule of the lens and theILM of the retina (Burk et al. 2000;Kadonosono et al. 2000) duringintraocular surgery. Trypan blue isreported to be effective for anteriorlens capsule staining in patients withan absence of red fundus reflex (Melleset al. 1999; Feron et al. 2002). The latterreport introduced the usefulness of TBin visualizing the epiretinal membrane.Triamcinolone acetonide is a corticoidwith an anti-inflammatory effect.Peyman et al. (2000) reportedly usedintraocular injection of TA to visualizethe vitreous. Triamcinolone acetonideis not a dye, however, and we evaluatedit in this study because TA is used for

Fig. 2. Light micrographs of isolated rat

retinas stained with haematoxylin-eosin

incubated in various media for 3 hours

after recovery: (A) immediately after the

60-min recovery; (B) aCSF; (C) BSS

plus1; (D) Opeguard1 Neo Kit; (E)

PBS. Retinas incubated with BSS plus1

and Opeguard1 Neo Kit showed little

difference compared to those incubated

with aCSF except for the vacuolization in

the ganglion cell layer. Neural damage

was significant in every layer of the tissue

with PBS. GCL¼ ganglion cell layer;

IPL¼ inner plexiform layer; INL¼ inner

nuclear layer; OPL¼outer plexiform

layer; ONL¼outer nuclear layer; POS¼photoreceptor outer segment. Original

magnification �280.


191


retinas followed by a 3-hour incubation

after exposure to (A) BSS plus1; (B) BSS

plus1 and distilled water (1 : 4); (C) 0.5%

ICG. The retinal tissue exposed to 0.5%

ICG showed remarkable somatic changes

and vacuolization in the nuclear layers

and severe histological damage to the

plexiform layers. GCL¼ ganglion cell

layer; IPL ¼ inner plexiform layer;

INL¼ inner nuclear layer; OPL¼ outer

plexiform layer; ONL¼outer nuclear

layer; POS¼photoreceptor outer segment.

Original magnification �280.


retinas incubated in BSS plus1 for 3 hours

after 1min exposure to the vital stain: (A)

exposure to BSS plus1 as a control; (B)

triamcinolone acetonide; (C) trypan blue.

Retinas exposed to TB showed slight

vacuolization in the ganglion cell layer

and inner nuclear layer. GCL¼ ganglion

cell layer; IPL¼ inner plexiform layer;

INL¼ inner nuclear layer; OPL¼ outer

plexiform layer; ONL¼outer nuclear

layer; POS¼photoreceptor outer segment.

Original magnification �280.

Fig. 4. Lactate dehydrogenase (LDH)

release from rat retinas incubated in BSS

plus1 after 1min exposure to 0.06%

trypan blue solution or triamcinolone

acetonide solution showed no difference

to that of the control samples (n¼ 3–4)

(p< 0.05, t-test compared with BSS plus1).


192

assisting vitrectomy with the same pur-pose as the two dyes.

Veckeneer et al. (2001) reported noocular toxicity of TB 0.06% in vivo (inrabbits). Studies with rabbits haveproved that TA is non-toxic to theretina (McCuen et al. 1981; Scholeset al. 1985). Intraocular use of ICGhas also been reported to have noadverse effects in the same way asabove. In cadaveric human eyes, nohistological damage of the retina dueto 0.3ml of 0.5% ICG injection intothe vitreous has been described (Burket al. 2000). In clinical studies, it hasbeen suggested that staining the ILMusing ICG allows safer and easierremoval of the ILM (Kadonosonoet al. 2000). On the other hand, damageto the sensory retina and retinal pig-ment epithelium using ICG has beenreported (Gandorfer et al. 2001; Sippyet al. 2001). As with previous reports,we found no significant retinal changewith TB and TA solution exposurefrom a biochemical and morphologicalviewpoint. In our model, ICG solutionat the maximal concentration for clin-ical use caused severe structural changein isolated rat retinas.

Da Mata et al. (2001) described amethod of staining the ILM where thedispersion of ICG is limited by mini-mizing intraocular fluid currents.Under these conditions, it is assumedthat ICG is not dilated equally in thevitreous cavity, and most of the pre-pared concentration of ICG attachesdirectly to the retina, especially overthe macula, throughout the stainingperiod. For this point, we evaluated

the toxicity of stains at the maximalreported concentration in this study.

In clinical use, ICG powder was firstdissolved in distilled water and thendiluted with an intraocular irrigatingsolution. Stalmans et al. (2002)reported the osmotic effects of the sol-vent on the retinal pigment epithelium.The change in the osmolarity of thesolution may influence the effect onthe sensory retina. However, there wasno difference between BSS plus1 only(305mOsm/kg) and BSS plus1 withdistilled water (4 : 1) (244mOsm/kg) inthis study (Figs 5 and 6). The higherrelease of LDH from the tissues exposedto ICG confirmed a biochemical impacton the retinas (p< 0.01, t-test com-pared with BSS plus1 and distilledwater). Weinberger et al. (2001) reportedpersistent ICG fluorescence 6weeksafter intraocular ICG administrationfor macular hole surgery. Indocyaninegreen can adhere to residual vitreousafter vitrectomy assisted with ICGsolution in the vitreous’ bradytrophicenvironment, and can cause longtermdamage to the human retina.

In our experimental model, the iso-lated rat retinas were exposed to thevital stains and triamcinolone aceto-nide from all directions, which differsfrom actual clinical conditions. How-ever, 0.06% TB and TA solutionsshowed no significant effect on rat ret-inal specimens compared to the controlsolutions, but ICG did have a toxiceffect on the rat retina in this study.The mechanism of ICG toxicity andthe influence of concentration requireinvestigation.

ReferencesBurk SE, Da Mata AP, Snyder ME, Rosa RH

Jr & Foster RE (2000): Indocyanine green-

assisted peeling of the retinal internal limiting

membrane. Ophthalmology 107: 2010–2014.

DaMataAP,BurkSE,RiemannCD,RosaRHJr,

Snyder ME, Petersen MR & Foster RE

(2001): Indocyanine green-assisted peeling

of the retinal internal limiting membrane

during vitrectomy surgery for macular hole

repair. Ophthalmology 108: 1187–1192.

Enaida H, Sakamoto T, Hisatomi T, Goto Y &

Ishibashi T (2002): Morphological and func-

tional damage of the retina caused by intra-

vitreous indocyanine green in rat eyes.

Graefes Arch Clin Exp Ophthalmol 240:

209–213.

Feron EJ, Veckeneer M, Parys-Van

Ginderdeuren R, Van Lommel A,

Melles GRJ & Stalmans P (2002): Trypan

blue staining of epiretinal membranes in pro-

liferative vitreoretinopathy. Arch Ophthal-

mol 120: 141–144.

Gandorfer A,Haritoglou C, Gass CA,UlbigMW

&KampikA (2001): Indocyanine green-assisted

peeling of the internal limiting membrane may

cause retinal damage. Am J Ophthalmol 132:

431–433.

Izumi Y, Benz AM, Kirby CO, Labruyere J,

Zorumski CF, Price MT & Olney JW (1995):

An ex vivo rat retinal preparation for excito-

toxicity studies. J Neurosci Methods 60:

219–225.

Kadonosono K, Itoh N, Uchio E, Nakamura S

& Ohno S (2000): Staining of internal limiting

membrane in macular hole surgery. Arch

Ophthalmol 118: 1116–1118.

McCuen BW 2nd, Bessler M, Tano Y,

Chandler D & Machemer R (1981): The

lack of toxicity of intravitreally administered

triamcinolone acetonide. Am J Ophthalmol

91: 785–788.

Melles GRJ, de Waard PWT, Pameyer JH &

Houdijn Beekhuis W (1999): Trypan

blue capsule staining to visualize the

Fig. 6. The lactate dehydrogenase (LDH)

release from rat retinas incubated in BSS

plus1 after 0.5% ICG exposure for 1min

was significantly higher than that of the

control samples (n¼ 3–4) (p< 0.01, t-test

compared with BSS plus1 and distilled

water).


193

capsulorhexis in cataract surgery. J Cataract

Refract Surg 25: 7–9.

Peyman GA, Cheema R, Conway MD &

Fang T (2000): Triamcinolone acetonide as

an aid to visualization of the vitreous and

the posterior hyaloid during pars plana

vitrectomy. Retina 20: 554–555.

Romano C, Price MT & Olney JW (1995):

Delayed excitotoxic neurodegeneration

induced by excitatory amino acid agonists

in isolated retina. J Neurochem 65: 59–67.

Sakamoto T, Miyazaki M, Hisatomi T,

Nakamura T, Ueno A, Itaya K &

Ishibashi T (2002): Triamcinolone-assisted

pars plana vitrectomy improves the surgical

procedures and decreases the postoperative

blood–ocular barrier breakdown. Graefes

Arch Clin Exp Ophthalmol 240: 423–429.

Scholes GN, O’Brien WJ, Abrams GW &

Kubicek MF (1985): Clearance of triamcino-

lone from vitreous. Arch Ophthalmol 103:

1567–1569.

Sippy BD, Engelbrecht NE, Hubbard GB et al.

(2001): Indocyanine green effect on cultured

human retinal pigment epithelial cells: impli-

cation for macular hole surgery. Am J

Ophthalmol 132: 433–435.

Stalmans P, Van Aken EH, Veckeneer M,

Feron EJ & Stalmans I (2002): Toxic effect

of indocyanine green on retinal pigment

epithelium related to osmotic effects of the

solvent. Am J Ophthalmol 134: 282–285.

Veckeneer M, van Overdem K, Monzer J,

Kobuch K, van Marle W, Spekreijse H &

van Meurs J (2001): Ocular toxicity study of

trypan blue injected into the vitreous cavity

of rabbit eyes. Graefes Arch Clin Exp


Weinberger AWA, Kirchhof B, Babak EM &

Norbert FS (2001): Persistent indocyanine

green (ICG) fluorescence 6 weeks after

intraocular ICG administration for macular

hole surgery. Graefes Arch Clin Exp


Zeekvalk GD & Nicklas WJ (1990): Chemi-

cally induced hypoglycaemia and anoxia:

relationship to glutamate receptor-mediated

toxicity in retina. J Pharmacol Exp Ther 253:

1285–1292.

Received on June 26th, 2003.

Accepted on November 28th, 2003.

Correspondence:

Kazuhiro Tokuda MD

Department of Ophthalmology

Daini Hospital

Tokyo Women’s Medical University

2-1-10 Nishi-Ogu, Arakawa

Tokyo 116-8567

Japan

Tel:þ 81 3 3810 1111

Fax:þ 81 3 3810 9817

Email: [email protected]


194

evaluation of toxicity due to vital stains in isolated rat retinas

Documents