comparison of central and peripheral human corneal ... epithelium in tissue culture ... * time of...

Comparison of Central and Peripheral Human CornealEpithelium in Tissue Culture

Bunshu Ebato, Judith Friend, and Richard A. Thofr

Past attempts to grow human corneal epithelium in culture had limited success, with confluence rarelyattained. This work is to determine whether different areas of human corneal epithelium grow better intissue culture. We compared the extent, the mitotic rates, and morphology of outgrowths and histologyof explants from central and peripheral human corneas in culture. Explants, 2 mm in diameter,removed from eye bank eyes, were placed epithelial side up on a culture dish with modified SHEMtissue culture medium (Jumblatt et al, 1983). After 7 days, the tissues were fixed, stained and the areaof outgrowths from explants measured using an image processor. For eight eyes from donors averaging66 yr old, the average area of central outgrowths was 7.8 ± 1 . 1 mm2, while that of peripheraloutgrowths was 52.8 ± 5.2 mm2 (P < 0.001). The mitotic rate of outgrowths of central epithelium wassignificantly less than that of peripheral epithelium (1.1 ± 0.5% vs 18.8 ± 0.8%) (P < 0.001). After 14days, central outgrowths had not attained confluence and consisted of large cells. Peripheral out-growths had attained confluence and consisted of small polygonal cells. Histology of explants showedthat only one layer of epithelium remained on the stroma in central explants, but several layers werepresent on the peripheral explants. Thus, peripheral human corneal epithelium grows better in culturethan does central human corneal epithelium. Invest Ophthalmol Vis Sci 28:1450-1456,1987

Culture of human corneal epithelium has been dif-ficult using the methods that work well with rabbittissue.1 Attempts to establish and maintain humancorneal epithelial tissue cultures have been reportedonly sporadically and seem to have met with littlesuccess.2"8 In those studies, a variety of methods suchas simple culture with tissue culture medium supple-mented with fetal calf serum and antibiotics,2 growthof epithelial cells in the presence of feeder layers ofirradiated 3T3 fibroblasts3"7 or with monoclonal anti-bodies against fibroblasts8 were used. Recently, usinga tissue culture medium supplemented with epider-mal growth factor, insulin, cholera toxin, glutamine,and dimethylsulfoxide in addition to fetal calf serumand antibiotics (supplemented hormonal epithelialmedium [SHEM]), Jumblatt and Neufeld havegreatly improved the culture of rabbit corneal epithe-lium.1 Prior to development of this medium, primarycultures of rabbit epithelium had grown to con-fluence, but it had not been possible to passage thecells; use of the SHEM medium permitted multiplepassaging of the cells.

From the Department of Ophthalmology, The Eye and Ear Insti-tute of Pittsburgh, and University of Pittsburgh, Pittsburgh, Penn-sylvania.

Supported by NEI grant EY-06186 and an unrestricted grant tothe Department from Research to Prevent Blindness, Inc.

Submitted for publication: August 4, 1986.Reprint requests: Richard A. Thoft, MD, The Eye and Ear Insti-

tute, 203 Lothrop Street, Pittsburgh, PA 15213.

It has been proposed that peripheral human cor-neal epithelium is important in the normal mainte-nance of the cell mass covering the cornea.910 Theimplication of this hypothesis is that peripheral epi-thelium, as a more actively growing area, may growbetter in tissue culture than central epithelium. Thissuggests that the culture of peripheral corneal epithe-lium in SHEM may optimize the growth of humancells and permit sufficient production of the epithe-lium to allow clinical ocular surface replacementusing cultured cells.

The purpose of this work is to (1) establish culturesof pure human corneal epithelium using enrichedmedium, and (2) test the hypothesis that differentareas of human corneal epithelium grow better intissue culture than others. To do so, primary culturesof central or peripheral epithelium were established,and (1) the area of outgrowth from the explants, (2)the mitotic rates of outgrowth, (3) the histology ofexplants, and (4) the morphology of the outgrowthsfrom the two areas were compared.

The work clearly demonstrates that peripheralhuman corneal epithelial cells grow better in vitrothan central ones when grown under the same condi-tions.

Materials and MethodsHuman eyes deemed unsuitable for grafting were

provided for this study by the Medical Eye Bank ofWestern Pennsylvania (Pittsburgh, PA).

1450

Downloaded From: https://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933134/ on 08/27/2018

No. 9 CULTURE OF HUMAN CORNEAL EPITHELIUM / Eboro er ol. 1451

Primary Cultures of Human Corneal Epithelium

Primary cultures of human corneal epitheliumwere started using endothelium-free explants."Under a tissue culture hood and using a dissectingmicroscope and sterile techniques, 2 mm diameterexplants were prepared from the central or peripheralareas of corneas using a trephine and spring scissors.Six to eight 2 mm diameter explants were removedfrom the central area of each donor eye, and 10 to 122 mm diameter explants were removed from the pe-ripheral cornea. Four central or peripheral explantswere placed epithelial side up on a 35 mm tissueculture dish (Corning [New York, NY] 25000 or on a22 X 22 mm sterile coverslip [Gold Seal, BectonDickinson and Co., Rutherford, NJ] placed in a 35mm tissue culture dish) and left covered for approxi-mately 15 min in the tissue culture hood. Then, 1 mlenriched supplemented tissue culture medium wasadded to the culture dishes. The next day, another 1ml medium was added to the culture dishes. Explantsthat did not remain attached or that floated off thedish were removed from the dish. Thus, approxi-mately two dishes using central and three dishesusing peripheral explants were prepared from eachdonor eye.

The medium was modified SHEM which con-tained Hams F12 and Dulbecco's Modified EaglesMedium (DMEM) (1:1), mouse-epidermal growthfactor (m-EGF, 10 ng/ml), insulin (5 Mg/ml), choleratoxin (0.1 Mg/ml), 1-glutamine (1 Mg/ml), fetal calfserum (to 15%), donor horse serum (to 5%), dimeth-ylsulfoxide (0.5%) and gentamycin (40 Mg/ml). Thecultures were incubated in a 37°C, 5% CO2-95% air,water-jacketed incubator for 7 to 14 days with themedium changed twice a week. Explants were left inthe culture dish for the duration of the incubation.

Measurement of the Extent of Outgrowths

After 7 days of incubation, the epithelial out-growths and explants were fixed in absolute ethanolfor 15 min and stained with hematoxylin (Gill #3,Sigma Chemical Co., St. Louis, MO) for 5 min fol-lowed by rinsing in tap water and air drying. Photo-graphs of the outgrowths were taken and the area ofeach outgrowth measured using computerized imageprocessing on an IS 2000 Image Processor (NewHartford, NY). The outgrowths from eight eyes (twosingle, three paired) were measured (Table 1).

Mitotic Rate

Uptake of 3H-thymidine determined by analysis ofautoradiographs was used as a measure of the mitoticrate of the tissues. After 7 days of incubation, cultures

Table 1. Donor eyes used for measurement of areaof outgrowths

Donoreye#

12345678

AverageTotal

Donor

ABBCCDDE

Age(years)

1765657070737396

66 ±8

TOD-

TOC*(hours)

963232.52020.52020.534.5

34.5 ±9.1

Number of explants

Center

44484444

36

Periphery

1144

127888

62

* Time of death to time of culture.

were prepared for autoradiography as follows"12: tri-tiated thymidine (1.0 ^Ci, New England Nuclear,Boston, MA, 20 Ci/nMol) in 1.0 ml modified SHEMwas placed over the cultures growing on coverslips in35 mm culture dishes, and the tissues were incubatedat 37°C for 30 min. After incubation, the cultureswere rinsed in phosphate buffered saline (PBS), fixedin absolute ethanol for 15 min, washed three times inice cold trichloroacetic acid (TCA), washed two timeswith distilled water and then air dried. The dried ex-plants were removed and discarded. The coverslipswith epithelial cell outgrowths were mounted cell sideup on glass histology slides using Permount® (FisherScientific, Fair Lawn, NJ). The slides were dipped inKodak NTB-2 emulsion (Kodak, Rochester, NY),stored for 14 days at —20°C, developed in KodakD-19 developer, fixed, and stained with hematoxylin(Gill #3, 5 min). After washing and dehydration, theslides were soaked in xylene, and the coverslips wereremoved from the slides and remounted cell sidedown on new glass slides.

The number of labeled cells (those which incorpo-rated tritiated thymidine) and the total number ofcells in various fields were counted. The fields werelocated in the center (adjacent to the site of the ex-plant), middle or leading edge of the outgrowths. Ap-proximately 20 fields at X400 magnification werecounted for each sample. Since there were no statisti-cally significant differences in mitotic rate betweendifferent areas of the outgrowths on the culturedishes, the results are averages of all the areas countedfor any one eye.

The mitotic rate is expressed as the labeling index:

Labeling index =Labeled cells

Total cellsX 100

Mitotic rates were measured in outgrowths from four(two single, one pair) eyes (Table 2).


14SI INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / September 1987 Vol. 28

Table 2. Donor eyes used for autoradiography of outgrowths

' Time of death to time of culture.

Donoreye #

1234

Average

Donor

FGGH

Age(years)

69797981

77 ± 5

TOD-TOC*(hours)

26181922

21 ± 4

Labeling index t

Center

1.2 ±0.1 (3)1.6 ± 1.0(8)1.2 ±0.4 (5)0.5 ± 0.4 (4)

1.1 ±0.5(20)

Periphery

17.6 ±4.7 (4)19.3 ±4.0 (10)19.4 ± 3.4(7)18.7 ±4.0 (4)

18.8 + 0.8(26)

f Mean ± SD (number of samples).

Histology of Explants

Explants removed from cultures after 7 days of in-cubation were fixed in 10% buffered formalin andsectioned. The sections were stained with hematoxy-lin and eosin using standard procedures. Explantsfrom four cultures were studied.

Morphology of Outgrowths

Cultures were examined every day and photo-graphed after 14 days incubation using a phase con-trast microscope (Nikon AFX-II, Tokyo, Japan).

B

Fig. 1. Typical epithelial outgrowths from central (A) and periph-eral (B) explants after 1 week in culture. The explants are the rounddiscs, and the dark areas around them are the outgrowths (Hema-toxylin stain, original magnification X4.5, grids = 1 mm).

Results

Area of Outgrowths

The areas of outgrowths of epithelium from centraland peripheral explants of eight donor eyes were ex-amined after 7 days in culture. The average age of thedonors was 66 ± 8 yr and the average time of death totime of culture was 34.5 ±9 .1 hr (Table 1). In alleight donor eyes from patients, the epithelial out-growths from peripheral explants were significantlylarger than those from central explants. The averagearea of outgrowths for all eight eyes was 52.8 ± 5.2mm2 for growth from peripheral explants and 7.8±1.1 mm2 for growth from central explants (P< 0.001) (Figs. 1 and 2).

Mitotic Rate of Outgrowths

The mitotic rate of outgrowths from four donoreyes was measured after 7 days in culture, expressedas the labeling index. The average age of donors was77 ± 5 yr and the average time of death to time ofculture was 21 ± 4 hr (Table 2). In all cases, thelabeling index of peripheral outgrowths was muchhigher than that of central outgrowths. No cultureshad reached confluence, nor did outgrowth from any

AREA OF OUTGROWTHS FROM EXPLANTS

• CENTER

E3 PERIPHERY

_r\

Fig. 2. Bar graph of areas of outgrowths from central and periph-eral explants of eight donor eyes.


No. 9 CULTURE OF HUMAN COP.NEAL EPITHELIUM / Ebaro er ol. 1450

Fig. 3. Tritiated thymi-dine uptake by 1 week oldcentral (A) and peripheral(B) epithelial outgrowths.Only two labeled cells arepresent in the central out-growth while approximately20 labeled cells are presentin the peripheral outgrowth(Hematoxylin stain, origi-nal magnification X800).

B

explant touch adjacent outgrowth. The average label-ing index was 18.8 ± 0.8 percent (n = 4 donor eyes)labeled cells for peripheral outgrowths and 1.1 ±0.5percent (n = 4 donor eyes) for central outgrowths {P< 0.001) (Table 2, Fig. 3).

Histology of Explants

After 7 days in culture, several layers of epitheliumwere present on peripheral explants, but only one

thin layer of epithelium remained on the stroma ofcentral explants (Fig. 4).

Morphology of Outgrowths

Cultures from four donor eyes (four dishes eachfrom peripheral and central explants) were culturedfor a total of 14 days (Table 1, #5-8). By that time, allthe outgrowths from peripheral epithelium were con-fluent in the 35 mm culture dishes and consisted of


INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / Seprember 1987 Vol. 28

f

Fig. 4. Central (A) andperipheral (B) explants after1 week in culture. Severallayers of epithelium arepresent on the peripheralexplant, but only one thinlayer of epithelium re-mained on the stroma of thecentral explant (H-E stain,original magnificationX720).

^ftw»d

small polygonal cells. Central outgrowths had not at-tained confluence, consisted of large cells, had manyholes, and floating dead cells were present in the me-dium (Fig. 5).

Discussion

These results show that it is possible to create andmaintain viable cultures of human corneal epithe-lium if certain criteria are met. The enriched SHEMmedium used in these studies contains growth factorsand components which have been shown to be useful

in culture of rabbit corneal epithelium11314 or ofuveal melanoma and other tissues.15 In these studies,the medium was arrived at empirically, and it is notclear whether all of the ingredients are essential forgrowth.

Most importantly, it appears that the location ofdonor cornea from which explants are taken is criti-cal. In these studies, at 3 to 4 days the cultures did notappear to be very different, but the differences be-came quite apparent soon thereafter. Cultures de-rived from peripheral epithelium had larger out-growths and higher mitotic rates after 1 week than


No. 9 CULTURE OF HUMAN CORNEAL EPITHELIUM / Eboro er ol.

\-J " V

Fig. 5. Phase contrast mi-crographs of 2 week oldcentral (A) and peripheral(B) epithelial outgrowths.Central outgrowths neverattained confluence andconsisted of large cells. Pe-ripheral outgrowths at-tained confluence and con-sisted of small polygonalcells (original magnificationX400).

-mm

comparable cultures derived from central epithelium.Moreover, cultures of peripheral epithelium becameconfluent within 2 weeks while cultures from centralepithelium died within 2 weeks. Studies of the platingefficiency and doubling time of passaged cells fromthe periphery are underway.

It appears that the human central epithelial cellsare not capable of replication in culture, at leastunder the conditions studied in these experiments,and that the cells spread out as much as they can fromthe explant to cover the surface of the culture dish butthe cultures do not thrive. This poor replication of

central corneal epithelium supports the hypothesisthat there may be a centripetal migration of cornealepithelial cells from the periphery to the center of thecornea to maintain the cell mass.910 It has recentlybeen reported that there are cells located in thelimbus that are stem cells for the corneal epithe-lium,16 further demonstrating that there are differ-ences between ocular surface epithelial cells from dif-ferent areas. These studies suggest that the centralepithelial cells may be more highly differentiatedthan peripheral epithelial cells. If that is the case, it isperhaps not unexpected that peripheral cells should


1456 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / September 1987 Vol. 28

be capable of more division, and, therefore, be morelikely to survive in culture than central epithelialcells.

This discovery of the different growth characteris-tics of central and peripheral corneal epithelial cellsmay explain some of the previous inconsistencies inthe reports of cultures of human corneal epithelium.It is possible that some investigators used explants ofperipheral tissue while others used explants of centralepithelium.

There are important differences between rabbitand human corneal epithelium.17 Therefore, having areadily available, reproducible system with which togrow cultures of human corneal epithelium is impor-tant for studies of epithelial biochemistry and physi-ology. Since only peripheral explants are needed, itappears that the corneal rim left after penetratingkeratoplasty can be used as a source of tissue for cul-ture (as is the case for human corneal endothelium,18

which will greatly expand the availability of tissueand therefore of cultures. For example, use of thesecultures will permit us to determine whether the ad-hesion characteristics of human tissue are similar tothose found in rabbit.19-20

There are several important clinical implications ofthis work. First, it indicates that the lenticules whichare used as an epithelial source in keratoepithelio-plasty21 should be oriented so that the peripheral cor-neal epithelium is directed toward the center of thecornea. This will permit the most active peripheralcells to grow onto and populate the corneal surface,rather than the sluggish central epithelial cells. Sec-ond, disturbance of peripheral corneal epithelium,such as inhibition of mitosis, may lead to a reductionof the central epithelial mass, leading to central epi-thelial defects. Third, central penetrating keratoplastywith epithelial retention is not an efficient way totreat diseases of epithelial origin, since the host epi-thelium will eventually repopulate the surface of thedonor tissue, as has already been shown in rabbits.22

Finding that peripheral human corneal epitheliumgrows well in culture encourages us to explore the useof such cultured sheets for ocular surface replacementin patients. Studies to characterize the adhesive prop-erties of these sheets to human corneal stroma areunderway to explore the possibility of cultured epi-thelial cell transplantation.Key words: human, corneal epithelium, tissue culture

AcknowledgmentsWe thank Paul Rehkopf and Joseph Warnicki for helping

us measure areas of outgrowth on the IS 2000 Image Pro-cessor.

References1. Jumblatt M and Neufeld AH: /8-adrenergic and serotonergic

responsiveness of rabbit corneal epithelial cells in culture. In-vest Ophthalmol Vis Sci 24:1139, 1983.

2. Newsome DA, Takasugi M, Kenyon KR, Stark WF, and OpelzG: Human corneal cells in vitro: Morphology and histocom-patibility (HL-A) antigens of pure cell populations. InvestOphthalmol 13:23, 1974.

3. Sun TT and Green H: Cultured epithelial cells of cornea, con-junctiva and skin: Absence of marked intrinsic divergence oftheir differentiated states. Nature 269:489, 1977.

4. Sun TT and Green H: Immunofluorescent staining of keratinfibers in cultured cells. Cell 14:469, 1978.

5. Sun TT, Shih C, and Green H: Keratin cytoskeletons in epithe-lial cells of internal organs. Proc Natl Acad Sci USA 76:2813,1979.

6. Doran TI, Vidrich A, and Sun TT: Intrinsic and extrinsicregulation of the differentiation of skin, cornea and esophagealepithelial cells. Cell 22:17, 1980.

7. Nelson WG and Sun TT: The 50- and 58-kdalton keratinclasses as molecular markers for stratified squamous epithelia:Cell culture studies. J Cell Biology 97:244, 1983.

8. SundarRaj N, Sundar-Raj CV, and Martin J: Selective cyto-toxicity of monoclonal antibodies against fibroblasts: Applica-tion in growth of human corneal epithelial cells in culture.Curr Eye Res 3:637, 1984.

9. Thoft RA and Friend J: The X, Y, Z hypothesis of cornealepithelial maintenance. Invest Ophthalmol Vis Sci 24:1442,1983.

10. Buck RC: Measurement of centripetal migration of normalcorneal epithelial cells in the mouse. Invest Ophthalmol Vis Sci26:1296, 1985.

11. Friend J, Kinoshita S, and Thoft RA: Corneal epithelial cellcultures on stromal carriers. Invest Ophthalmol Vis Sci 23:41,1982.

12. Kinoshita S, Friend J, and Thoft RA: Biphasic cell prolifera-tion in transdifferentiation of conjunctival to corneal epithe-lium in rabbits. Invest Ophthalmol Vis Sci 24:1008, 1983.

13. Jumblatt MM, Fogle JA, and Neufeld AH: Cholera toxin stim-ulates adenosine 3',5'-monophosphate synthesis and epithelialwound closure in the rabbit cornea. Invest Ophthalmol Vis Sci19:1321, 1980.

14. Jumblatt MM and Neufeld AH: A tissue culture assay of cor-neal epithelial wound closure. Invest Ophthalmol Vis Sci 27:8,1986.

15. Albert DM, Ruzzo MA, McLaughlin MA, Robinson NL, CraftJL, and Epstein J: Establishment of cell lines of uveal mela-noma. Invest Ophthalmol Vis Sci 25:1284, 1984.

16. Schermer A, Galvin S, and Sun TT: Differentiation-relatedexpression of a major 64K corneal keratin in vivo and in cul-ture suggests limbal location of corneal epithelial stem cells. JCell Biol 103:49, 1986.

17. Friend J: Biochemistry of ocular surface epithelium. In TheOcular Surface, Thoft RA and Friend J, editors. InternationalOphthalmology Clinics. Boston, Little, Brown and Co., 1979,pp. 73-91.

18. Nayak SK and Binder PS: The growth of endothelium fromhuman corneal rims in tissue culture. Invest Ophthalmol VisSci 25:1213, 1984.

19. Gipson IK, Grill SM, Spurr SJ, and Brennan SJ: Hemidesmo-some formation in vitro. J Cell Biol 98:1565, 1984.

20. Trinkhaus-Randall V and Gipson IK: Role of calcium andcalmodulin in hemidesmosome formation in vitro. J Cell Biol98:1565, 1984.

21. Thoft RA: Keratoepithelioplasty. Am J Ophthalmol 97:1,1984.

22. Kinoshita S, Friend J, and Thoft RA: Sex chromatin of donorcorneal epithelium in rabbits. Invest Ophthalmol Vis Sci21:434, 1981.


comparison of central and peripheral human corneal ... epithelium in tissue culture ... * time of...

Documents