Current Eye Research

A primary culture model of rabbit conjunctival epithelial cells exhibiting tight barrier properties

Pratik Sahal, Kwang-Jin Kim2.3J.5 and Vincent H. L. Lee1

Departments of I Pharmaceutical Sciences and Ophthalmology, ’Medicine, ’Physiology and Biophysics and 4Bioinedical Engineering, and ‘Will Rogers Institute Pulmonary Research Center, University of Southern California, Los Angeles, CA, USA

Abstract Purposr. The present study was conducted to develop and characterize a functional primary culture of pigmented rabbit conjunctival epithelial cells on permeable support exhibiting tight barrier properties.

Metlzoris. Conjunctival epithelial cells were isolated by 0.2% protease treatment, cultured at 0.5-1.8 X 1 0 6 cells/cm’ onto collagen-treated Transwell filters, and were maintained either in the presence of 1 % fetal bovine serum throughout or serum- free media from day 3 onwards. Transepithelial potential dif- ference (PD) and transepithelial electrical resistance (TEER) were measured and equivalent short-circuit current (Ieq = PD/ TEER) estimated.

Results. There appears to be a critical plating density of 1.5 X 1 0 6 celldcni? for functional development of tight epithelial cell cultures. The culture conditions as noted above did not affect either the time when peak bioelectric parameters were attained (days 8-10) or the magnitude of these parameters at a plating density of 1.5 X lo6 cells/cni2. Specifically, cells grown in a serum-free media showed a peak TEER of 1.9 f 0.2 kil.cm2, a PD of 14.2 f 1.6 mV (apical side negative), and a n I,, of 8.0 k 0.4 pA/cm2 (mean f SEM, n = 45). Electron microscopy of serum-weaned cultures revealed a multilayered epithelium with numerous microvilli on the outermost layer of cells, while sporadic positive Periodic Acid Schiff (PAS) stain- ing under light microscopy suggested the presence of mucin- secretory goblet cells.

Conclusions. A functional, tight, epithelial barrier of the pig- mented rabbit conjunctiva on a permeable support has been developed, which may be useful for mechanistic studies of ion and drug transport at the cellular level. Curr. Eye Res. 15: I 163-1 169, 1996.

Corrc..\~~jrrik.ric.r.: Dr. Vincent H. L. Lee, University of Southern California, School of Phannacy, Department of Pharmaceutical Sciences, I985 Zonal Avenue, LOS Angrler, CA 90033, USA

Key words: rabbit conjunctival epithelial primary culture; permeable support; transepithelial electrical resistance; poten- tial difference

Introduction The bulbar conjunctiva is the first tissue across which a topi- cally applied drug must pass in order to reach the underlying tissues in the uveal tract via the non-corneal route (1). Previous work in this laboratory has revealed that this tissue is permeable to a variety of molecules differing in size and polarity (2, 3) and that this tissue contributes substantially to the absorption of topically applied drugs into the bloodstream (4). The next important step in designing strategies to modulate systemic drug absorption would be facilitated by information on the capacity and mechanisms of drug transport in the conjunctiva. In this regard, conjunctival cell cultures grown on a permeable support with replicate barrier properties might be a useful in vitro model. This model would offer a better controlled envi- ronment for drug transport studies and allow rapid screening of how structural and formulation changes of a particular drug may alter conjunctival drug transport properties.

To date, there exists no functional conjunctival epithelial cell line, except for ‘Chang conjunctival cells’ that were subse- quently found to be contaminated with HeLa cells ( 5 ) . Primary cultures of rabbit conjunctiva have been developed on plastic substrata by Patton et af. (6) for evaluating the growth of Clzlurn-ydia trachornatis and by Tsai et ul. (7) for evaluating cell morphology, differentiation, and transdifferentiation of the conjunctiva. However, neither system, by its very nature, is suitable for transepithelial drug transport studies. Toward that end, the present study was conducted to develop, on a perme- able support, a functional primary culture of pigmented rabbit conjunctival epithelial cells that exhibits a high TEER and an equivalent I,, comparable to that observed in the intact tissue (8).

~~ ~

Received o n November 7, 1995; revi\ed on July 8 and accepted on July 26, 1996 1) Oxford University Press

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Mel

bour

ne o

n 11

/09/

14Fo

r pe

rson

al u

se o

nly.

I164 P. Saha, K.-J. Kim und V. H . L. Lee

Materials and methods

Materials

Male Dutch-belted pigmented rabbits, weighing 3.0-3.5 kg, were purchased from American Rabbitry (Los Angeles, CA). The studies utilizing rabbits described in this report con- formed to the Guiding Principles in the Care and Use of Ani- mals (DHEW Publication, NIH 80=23). Protease (Type XIV from Srt-c.ptomycrs griseus) and deoxyribonuclease I (Type IV froni bovine pancreas) were obtained from Sigma Chemical Co. (St. Louis, MO). Ca2+-free Hanks' balanced salt solution (HBSS), Ca'+-free minimum essential medium (S-MEM) and certitied fetal bovine serum (FBS) were purchased from GlBCO BRL (Grand Island, NY). Cell strainer (40 prn) was obtained froin Falcon (Becton Dickinson Labware, NJ). PC- I medium (a serum-free, low protein, defined medium) was pur- chased from Hycor Biomedical (Portland, ME). Collagen (types 1 and 111)-treated Transwells (0.45 p m pore size, 12 m m diameter) were obtained from Costar (Cambridge, MA). Biocoat inserts (6.5 mm diameter) were obtained from Collab- orative Biomedical Products (Bedford, MA).

Isolation of rabbit conjunctival epithelia1 cells

Rabbits were sacrificed with a rapid overdose of sodium pento- barbital solution (85 mg/kg) administered via a marginal ear vein. The eyeball was aseptically and gently removed from the orbit and placed in ice-cold Hanks' balanced salt solution (HBSS). The entire conjunctiva was trimmed off to within approximately 4 mm from the limbus, freed of extraneous tis- sues, placed in a sterile petri dish containing ice-cold 0.2% protease in S-MEM (9), and incubated for 90 min at 37°C in 95% air/5% COz. Epithelial cells were then teased off with sterile fine forceps and a #10 scalpel blade.

The isolated cells were inimediately transferred to preequili- brated S-MEM containing 10% FBS and 0.5 mg/ml DNase I to stop protease action (9), mixed, and centrifuged twice at 2 10 X g for 10 min at room temperature. The S-MEM had been pre- equilibrated in 5% C02 for 1 hr at 37°C. The resulting pellet was suspended in S-MEM containing 10% FBS, filtered through a 40 p m cell strainer, and centrifuged with the same setting for a third time. The final pellet was resuspended in pre- defined PC-1 growth medium supplemented with 1% FBS. 2 mM L-glutaniinc, 100 units/ml penicillin-streptomycin, 05% gentamicin and 0.4% fungizone.

Primary culture of conjunctival epithelial cells

After ascertaining cell viability based on exclusion of 0.1% tryplan blue, the cells were plated at a density of 1.5 X lo6 cells/cm2 (0.5 ml) onto Costar 12-mm Transwell-COL inserts (precoated with collagen type I and 111) in a 12-well plate and bathed on the bottom-side of inserts with 1.5 ml of supple- mented PC-I medium. Cells were cultured in a humidified incubator at 37°C in 95% air/5% C02. On day 2, cells were washed with defined PC-1 medium containing either 1% or 0. I % FBS. Thereafter, cells were fed regularly with 1% FBS in PC-I medium (throughout), or with the serum-free medium (PC- 1 ) from day 3 onwards.

The TEER in kfl.cm2 and PD of the monolayers in mV were monitored daily with an EVOM epithelial voltohmmeter (World Precision Instruments, Sarasota, FL), and corrected for back- ground values contributed by the blank filter (coated with sub- stratum) and culture medium. I,, in p " m * was estimated from the ratio of corrected PD over TEER. In addition to bioelectric parameters, the plating efficiency was also determined. Cells were detached from the collagen substratum following 5 min of treatment with 0.1% trypsin-EDTA at 37°C and counted; the plating efficiency was estimated from the percentage of cells attached to the collagen substratum from day 4 onwards.

(u) Effect of pluting density

The effect of plating density on the development of the con- junctival epithelial cell monolayers, as gauged through TEER and PD values, was determined by plating conjunctival epithe- lial cells onto Costar 12-mm Transwell-COL inserts at 0.5, 0.75, 1.25, 1.5, and 1.8 ( X lo6) cells/cm*.

(b) Effect ofsubstrata

This was studied by plating conjunctival epithelial cells at a density of 1.5 X 10fVcm2 on Biocoat 6.5-mm inserts (0.4 p m pore size), precoated with either collagen type I, laminin, fibronectin, or matrigel. Matrigel comprises laminin, type 1V collagen, heparan sulfate proteo-glycans, entactin, TGF p, and basic FGF. The effect of substrata on conjunctival cell layer developrnent was ascertained by TEER and PD measurements.

Histological examination of cultured conjunctival epithelial cells

(u) Transmission electron microscopy

Cultured conjunctival epithelial cells were examined by trans- mission electron microscopy (TEM) on days 3 and 10. The cul- tures were washed three times with chilled (4°C) phosphate- buffered saline (PBS, pH 7.2) and fixed in 2.5% glutaraldehyde in PBS at 4°C for 2 hr. Fixed cells were rinsed in chilled PBS five times, and further fixed in 1.5% Os04 in PBS at 4°C over- night. Osmicated cells were rinsed in PBS twice, further rinsed with distilled water three times and stained with 1.5% uranyl acetate in 50% ethyl alcohol for 30 min. Stained cells were dehydrated in graded ethyl alcohol of 70%, 85%, 95% and 100% for 5 , 5, 5 and 10 min, respectively. These cells were then embedded with Spurr resin (Polysciences, Warrington, PA) and sectioned for TEM with a JEOL l0OC (Peabody, MA) at 80 keV.

( b ) Periodic Acid Sch@(PAS) staining

Cultured conjunctival epithelial cells of day 0, 4, 6, 8, 10 and 14 were fixed in 10% formaldehyde in buffered PBS for stain- ing with the PAS reaction set and counterstaining with Harris hematoxylin for identification of epithelial secretory (e.g. gob- let) cells.

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Mel

bour

ne o

n 11

/09/

14Fo

r pe

rson

al u

se o

nly.

Rabbit coizjunctivd epithelid primary culture I165

2.0.

1.5-

1.0-

0.5 -

Data analysis

Data are represented as mean f SEM (n), where n = the num- ber of observations. Statistical significance was tested by the Student's T-test or one-way ANOVA, using the Statview I1 sta- tistical software package (Abacus Concepts, Inc., Berkeley, CA). When ANOVA was used, group means were contrasted for significant differences using a post hoc (Fisher PLSD) method. Statistical significance was set at p < 0.05.

Results Our cell isolation procedure yielded a heterogenous population of four or five different cell types (10) (as observed under the light microscope), comprising approximately 15-20 X 10" cells per animal with a viability exceeding 9.596, as assessed by tryptan blue dye exclusion. Epithelial cells from both eyes of each rabbit generally yielded one plate of twelve Transwell units (1.13 cni'). At a plating density of 1.8 X loh cells/cm*, approximately 40% of the cells were found to be attached to the collagen-treated substratum by day 4. The cell count remained fairly constant averaging 0.72 X ]Oh cells/cm2 through day 10 (Table 1, p 0.05 by one-way ANOVA), representing a conflu- ent culturc, at a n approximate plating efficiency of 40%. How- ever, by day 14 there was a 22% decrease in the cell count (Table I . p < 0.05 by one-way ANOVA). Cell viability also remained constant at approximately 98% throughout this period (Table 1 ).

(a) Rolil of'culture tnedia

With and without 1% FBS, cultured conjunctival cell layers showed peak bioelectric parameters around days 8-10. Cell cultures maintained in 1% FBS throughout exhibited, at the peak, TEER o f I .7 f 0.3 k R cm? (n = 10) (day 9) (Fig. la), PD of9.3 k 2.2 mV (day lo), apical side negative (Fig. lb); and I,, of 6.4 f 0.5 pA/cm* (day 10) (Fig. lc). Cell layers cultured under the condition where the FBS concentration was weaned from 1 % on day 0 to 0.1 % on day 2 and finally to 0% from day 3 onwards exhibited a peak TEER of 1.9 k 0.2 kfZ cm? (n = 45) (day 9) (Fig. 1 a), PD of 14.2 k 1.6 mV (day 8) (Fig. I b), and I,, of 8.0 f 0.4 pA/cmZ (day 8) (Fig. lc). While statistically sig- nificant differences (p < 0.05 by the Student's T-test) were

Table /. cultures grown in serum-free media from day 3 onwards

Cell number and viability as a function of age for

n N

E u

a w W I-

- > E 0 n

Y

n N

3 2 - - 4

Cell count" Cell viability Day (X I 0 6 cells/cni') (97.)"

0 4 6 8

10 14

1.80b 0.74 f 0.04 0.81 k 0.07 0.70 f 0.05 0.64 k 0.04 0.56 f 0.04

97.1 98.1 f 0.6 98.1 k 0.4 96.3 f 0.9 97.9 f 0.3 95.0 f 0.9

.'Mean f S E M (11 = 6) . p > 0.05 by one-way ANOVA across days 4-10. t Initial planting density

I A

0.0 4 0 3 6 9 12 15

Days

LU

B

t T

0 3 6 9 12 1

Days 5

I 1

0 0 3 6 9 12 15

Days

Figure 1. Mean TEER (A), PD (B) and I,, ( C ) of cultures grown in the presence of 1 % FBS throughout (0, n = 10 cell layer preparations), and in serum-free media from day 3 onwards (0. n = 45 cell layer prepara- tions). Error bars represent SEM. Where not shown, the error bar is smaller than the size of the symbol. Asterisks indicate ;I statistically sig- nificant difference (p < 0.05) by Student's T-test on a given day.

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Mel

bour

ne o

n 11

/09/

14Fo

r pe

rson

al u

se o

nly.

I166 P. S<thri . K.-J. Kiiri avid V. H . L. Lrr

observed in each bioelectric parameter around days 5-7, for cell layers cultured under serum-replete and sciurn-free condi- tions, no significant differences were observed in the bioelec- tric parameters at the peak (days 8-10). To avoid possible com- plications that may arise from the serum components in future drug transport studies, the serum-Cree condition was selected as [he culture condition for all subscquent cultures.

A rnultilayer appearance with numerous microvilli o n the outermost layer, similar to those found in the in vivo, was apparent in cultures on day 3 (Fig. 2a) and 10 (Fig. 2b) by TEM at 39,000X magnification. In addition, tight junctional complexes and lateral intercellular spaces between adjacent epithelial cells were also observed (day 10) (Fig. 2c). A spo- radic PAS positive staining pattern (5 +_ 0.5%, n = 4) from day 0 through 10, suggesting the presence of mucin secretory cells (c.g. goblet cells), was observed using light microscopy at 300X (original) magnitication (Fig. 3). A diffuse positive staining was also apparent, similar to the staining pattern observed for human conjunctival epithelium in culture ( 1 I ) .

Figuvr 2. Transmission electron photo micrographs of day 3 (A) (magnitication 6,OOOX) and day 10 (B) (magnification 6,OOOX) cultures grown in serurn-free media froin day 3 onwards, displaying a inultilayered epithelium with microvilli o n the outermost layer. The multilayercd appearance was more prominent in day 10 cultures than in day 3 cultures. (C) (magni- fication 35,OOOX) shows the tight junctions ( 7 ) and lateral intercellular splice (L) for cells grown for 10 days.

(c) E&cf qf'suhtrrrtrr

Cells plated o n collagen(-I)-precoated inserts led to high peak TEER (1.8 kSZ. cmz) and PD (7.3 mV) values (Fig. 4). In con- trast, cells cultured on fibronectin and matrigel yielded lower TEER (< 0.5 kf l . cm?) and PD values (< 4 mV) (Fig. 4). Lami- nin performed the best of the three on the basis of a peak TEER of 1 .0 k f l . a n 2 and PD of 4.7 mV (day 7), which was 57% and 65%, respectively, of that afforded by the commercially avail- able collagen substratum.

(11) Ljfrct of'pluting clcnsity

A plating density of 0.5 X I 0 6 cellskm' failed to yield appre- ciable TEER or PD throughout the 14-day period (Table 2 , p > 0.05). All other plating densities yielded TEER values of at least I kbl. cm2 between day 8 and 10 (Table 2). A maximum Peak TEEK of 1.7 f 0.3 kfl. cm' and a PD of 5.9 * 0.7 m v werc attained in the cell cultures at a plating density of 1.5 X 10" cells/cm2 (Table 2). cells.

FiRl,rr -{, Periodic Acid Schiff staining light photomicrograph (mag- nification 3OOX) of day 4 culture grown in serum-free media from day 3 onwards. Thc prolninent black dots correspond to PAS positive

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Mel

bour

ne o

n 11

/09/

14Fo

r pe

rson

al u

se o

nly.

Rabbit conjunctival epithelial primaty culture I167

Tuhle 2. onwards'

Values of peak TEER, PD, and I,, as a function of plating density for cultures grown in serum-free media from day 3

Plating density Day of peak TEER PD I,,, Group ( X I Oh cells/cm2) bio-electric performance (kfl.cm2) (mV) ( kA/cm2)

I 0.50 7.8 k 0.8 0.2 f 0.0 0.3 f 0.1 1.5 f 0.9 I1 0.75 8.5 f 0.6 0.9 f 0.1 2.1 f 0.9 2.3 k 0.3

111 1.25 8.3 rt 0.2 0.8 f 0.2 1.6 f 0.4 2.8 f 0.6 IV I .so 8.2 f 0.2 1.7 f 0.3 5.9 f 0.7 4.6 ? 1.0

1.2 f 0.2 4.2 f 0.8 4.0 k 0.8 V 1.80 8.3 f 0.2

"Mean f SI!M (n = 6 ) . TEER was statistically significant (p < 0.05) by one-way ANOVA (Fisher PLSD) between groups 1 and 11. 1 and IV, I and V, I1 and IV, 111 and IV; PD was statistically significant (p < 0.05) by one-way ANOVA (Fi\her PLSD) between groups I and IV, 1 and V, 11 and IV, 11 and V, IT1 and IV, 111 and V; whereas I,, was not statistically significant ( p > 0.05) by one-way ANOVA

Discussion Our early attempts to adapt the procedures reported by Patton cr ul. (6) and Tsai et ul. (7) to culture conjunctival epithelial cell monolayers on collagen-treated pcrineable supports were unsuccessful. The resulting epithelial cell layers yielded a very low peak TEER (<I00 f1.cm2) and an equally low PD ( < I mV) throughout 14 days. These procedures used either dispase 11 (7) or 5 mM EDTA plus 0.1 % trypsin (6). The culture condition reported herein was modified from that which has been used to successfully culture guinea pig tracheal epithelial cell mono- layers (9). Thus, the PC-I medium used was formulated in a specially modified DMEM-F12 base. However, in the present study, cells were maintained in the presence of only 1% FBS or serum-free conditions, as opposed to 2-5% serum used throughout by Patton et ul. (6) and Tsai et ill. (7).

Our culture methods led to the successful development, from a heterogenous population of five epithelial cell types (lo), a functional, tight epithelial culture of the pigmented rabbit con- junctiva on a permeable support. Since the focus of this study was not to develop a conjunctival barrier comprised of a single cell type, neither detailed histological studies with respect to individual cell type lineage nor gradient separation of different cell types were performed. The high TEER (21 .O k 0 . cmz, day 7-10), PD (27.3 mV, day 8-10), and I,, values (26 FA/cm?, day 8-12) (Fig. 1) were qualitatively comparable to those exhibited by the excised tissue (TEER, 1.3 kfZ.cm2; PD, 17.7 mV; and I,,, 14.5 pA/cm2) (8). The peak TEER and PD values were, respectively, 8 and 4 times higher than those reported across cultured fetal human retinal pigmented epithelium (TEER = 227 f1.crn2, PD = 3.2 mV) (12), while being - 1 I times higher than the TEER exhibited by bovine corneal endo- thelial monolayers (TEER = 180 a. cm2) (13). Compared to conjunctival epithelial cells, rabbit non-pigmented ciliary body epithelial cells in culture generated an 8 times lower TEER of 230 R. cmZ( 14).

After about 10 days, the functional characteristics of the con- junctival epithelial cultures change, as suggested by a rapid decrease in the TEER (and PD) of the culture. Assuming that the high electrical resistance in our culture system is attributable to the characteristic tight junctional complexes, such a fall in

T

A

0 4 8 12 16 Days

10 B

8 I n 0 4 8 12 16

Days

Figure 4. Effect of substrata on TEER and PD of cultures grown in serum-free media from day 3 onwards. Error bars represent SEM for n = 6 Transwells. Where not shown, the error bar is smaller than the size of the symbol. Key: collagen; 0 tibronectin; A laminin; 0 matrigel.

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Mel

bour

ne o

n 11

/09/

14Fo

r pe

rson

al u

se o

nly.

1168 P. Suhti. K.-J. Kirn tirid V. H. L. Lee

Figure 5 . inetlia from day 3 onwards.

Transmission electron photo micrographs (magnification 6,OOOX) of day 8 (A), 10 (B), and 14 (C) cultures grown in serum-free

TEEK may then be ascribed to changes in either junctional ultr:iytructure or number ofjunctional strands, or both. However, at present, this is not fully discernible by TEM (15) (Fig. 5).

Morphologically, the culture has a multilayer appearance with numerous microvilli on the outermost layer (Fig. 2a and 2b) and well-developed tight junctions (Fig. 2c). The cell sur- lice PAS staining pattern, as shown in Figure 3 , was consistent with Y 5% population of secretory (e.g. goblet) cells. This pro- portion retnained remarkably constant throughout the 14 day culture period, and was at the low end of the 4-22% of goblet cells in the rabbit conjunctival epithelium a s reported by Steuhl 1'1 ( I / . (10). This finding may suggest conservation of the rela- tive proportion of secretory and non-secretory epithelial cell<.

An important tinding in this study is that after two days of initial exposure to 1 ?k FBS, a serum-free medium can be used to culture conjunctival epithelial cell layers that develop tight barrier properties akin to the excised tissue. This is indicated by the lack of statistical difference by Student's T-test in the values of peak bioelectric parameters between cells cultured in the presence of 1 % FBS throughout and those cultured in scrum-free media I'rorn day 3 onwards (Fig. I ) . This finding is, however, at variance with that of Mortell ('1 NI. (16), where a progressive fall in the TEER of MDCK cell monolayers over the much higher 10-50% FBS concentration range was seen. Given that cultures grown in completely serum-free media throughout the culture period yielded a very low peak TEER (< I00 0. cm?) and ;in equally low PD ( < I mV) in our hands, i t is very likely that FBS is required for the initial attachment of con.junctival epithelial cells to the substratum.

Another important finding i n this study is that the collagen substratum is superior to the fibronectin, laminin, and matrigel substrata in promoting conjunctival cell attachment and the development of a functional epithelial barrier. The poor perfor- mance of the matrigel substratum is surprising, in light of the finding of Tsai ('1 ti/. (7) that conjunctival epithelial cells, at a

plating density of 1 X 10s cellskmz, were able to reach conflu- ence around days 7-9 on matrigel-coated plastic plates. How- ever, since no bioelectric measurements were made in those studies, it is not possible to judge how well the tight junctions were developed. No explanation for the apparent discrepancy in the matrigel substratum effect is immediately forthcoming.

To the best of our knowledge, the present study marks the first successful development of a functional, tight, multilay- ered, epithelial barrier of the pigmented rabbit conjunctiva on a permeable support. Mechanistic ion and drug transport studies at the cellular level can be easily performed using this in vitro culture model. The cultured epithelium maintains its barrier properties for four days, a practical time-frame for transepithe- lial drug and ion transport studies.

Acknowledgements This work was supported in part by NIH grants EY10421 (VHLL) and HL38658 (KJK), an unrestricted gift from the Bausch and Lomb Personal Products Division, and the United States Pharmacoepial (USP) Fellowship (PS).

References Ahmed, I . and Patton, T. F. (1985) Importance of the non- corneal absorption route in topical ophthalmic drug deliv- ery. Invest. Ophthalmol. Vis. S c i 26,584-587. Wang, W., Sasaki, H., Chien, D. S. and Lee, V. H. L. ( I99 I ) Lipophilicity influence on conjunctival drug pene- tration in the pigmented rabbit: a comparison with corneal penetration. Curr: Eye Res. 10,57 1-579. Hayakawa, E., Chien, D. S., Inagaki, K., Yamamoto, A,, Wang, W. and Lee, V. H. L. (1992) Conjunctival penetra- tion of insulin and peptide drugs in the albino rabbit. Pharm. Res. 9,769-775.

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Mel

bour

ne o

n 11

/09/

14Fo

r pe

rson

al u

se o

nly.

Rabbit conjunctival epithelial primary culture 1169

4. Lee, Y. H. , Kompella, U. K. and Lee, V. H. L. (1993) Sys- temic absorption pathways of topically applied beta adren- ergic antagonists in the pigmented rabbit. Exp. Eye Res. 57,34 1-349.

5 . ATCC CLitalog ($Cell Lines and Hybridomus, 7th Edition (1992). P. IS.

6. Patton, D. L., Chan, K. Y., Kuo, C.-C., Cosgrove, Y. T. and Langley, L. ( 1 988) In vitro growth of Chlamydia tracho- matis in conjunctival and corneal epithelium. Invest. Oph- thalmol. Vis. Sci. 29, 1087- 1095.

7. Tsai, R. J.-F. and Tseng, S. C. G. (1988) Substrate modula- tion of cultured rabbit conjunctival epithelial cell differen- tiation and morphology. Invest. Ophthalmol. vis. Sci. 29,

8. Kompella, U. K., Kim, K.-J. and Lee, V. H. L. (1993) Active chloride transport in the pigmented rabbit conjunc- tiva. Cum Eye Rex 12, 1041-1048.

9. Robison, T. W., Dorio, R. J. and Kim, K.-J. (1993) Forma- tion of tight monolayers of guinea pig airway epithelial cells cultured in an air-interface: bioelectric properties. Bio-Tpchrziques, 15, 468-473.

10. Steuhl, P. and Rohen, J. W. (1984) Cellular structure of the conjunctival epithelium of rabbits. Graefes Arch. Clin. Exp. Ophthalmol. 221,265-271.

11 . Kosaku. K . , Okamoto, S. and Ubels, J. L. (1994) Human conjunctival epithelium in culture expresses goblet cells. (Abstract). Invest. Ophthalrnol. Vis. Sci. 35 (Suppl.), 1560.

12. Gallemore, R. P., Hu, J. G., Frambach, D. A. and Bok, D. (1995) lon transport in cultured fetal human retinal pig- ment epithelium. (Abstract). Invest. Ophthalmol. Vis. Sci. 36 (Suppl.), 2 16.

13. Narula, P., Xu, M., Kuang, K., Akiyama, R. and Fischbarg, J. ( 1992) Fluid transport across cultured bovine corneal endothelial monolayers. Amer. J. Physiol. 262, C98-C103.

14. Cilluffo, M. C., Fain, M. J. and Fain, G. L. (1993) Tissue culture of rabbit ciliary body epithelial cells on permeable supports. Exp. Eye Res. 57,5 13-526.

15. Reinach, P. S., Tarvin, J. T. and Hirsch, M. (1991) Changes in cellular membrane and paracellular conductances by amphotericin B in the epithelium of the bullfrog cornea. Biochim. Biophys. Acta, 1066, 115-1 23.

16. Mortell, K. H., Marmorstein, A. D. and Cramer, E. B. (1993) Fetal bovine serum and other sera used in tissue culture increase epithelial permeability. In Vitro Cell. Dev. Biol. 29A, 235-238.

17. Weiss, N. A. and Hassett, M. J. (1987) Analysis of vari- ance. In Introductory Statistics, 2nd Ed. Pp. 565-592. Addison-Wesley, Massachusetts.

1565- 1576.

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Mel

bour

ne o

n 11

/09/

14Fo

r pe

rson

al u

se o

nly.