complications associated with bovine corneal endothelial cell native endothelium had been...

Download Complications associated with bovine corneal endothelial cell native endothelium had been mechanically…

Post on 30-Nov-2018




0 download

Embed Size (px)


  • Complications associated with bovinecorneal endothelial cell-lined homografts

    in the catCharles F. Bahn, Donald K. MacCallum, John H. Lillie,

    Roger F. Meyer, and Csaba L. Martonyi

    Cultured bovine corneal endothelial cells were subcultured onto feline corneas from which thenative endothelium had been mechanically removed, and, transplanted into cats via penetratingkeratoplasty. Although the transplants remained thin and. clear in the immediate postoperativeperiod, correlative clinical and morphologic analysis disclosed evidence of a host responsedirected against the heterologous endothelium by the ninth postoperative day. Eyes withrotational autografts or transplanted homografts did, not disclose evidence of a similar hostresponse. (INVEST OPHTHALMOL VIS SCI 22:73-90, 1982.)

    Key words: corneal endothelium, cell culture, corneal transplantation,transplant rejection, retrocorneal membrane

    The maintenance of corneal transparency isdependent on a population of corneal endo-thelial cells sufficient to affect deturgescenceof the corneal stroma.1 Reduction in thenumber of corneal endothelial cells in thehuman,2 rhesus monkey,3 or the domesticcat4species with a limited endothelial cellregenerative capacityresults in a compen-satory hypertrophy of the remaining cells andretention of an intact endothelial cell mono-

    From the Department of Ophthalmology, University ofMichigan Medical School, Ann Arbor (C.F.B.,C.L.M., and R.F.M.), and Department of Anatomyand Dental Research Institute, University of MichiganMedical School and School of Dentistry, Ann Arbor(D.K.M. and J.H.L.).

    This research was supported by grants from the NationalInstitutes of Health (DE-02731), Research to PreventBlindness, the Michigan Eye Bank Research Fund,and the Office of the Vice President for Research Uni-versity of Michigan.

    Submitted for publication July 2, 1980.Reprint requests: Charles F. Bahn, M.D., M.S., De-

    partment of Ophthalmology, Box 012, University ofMichigan Hospital, Ann Arbor, Mich. 48109.

    layer. If the endothelial cell population isgreatly reduced, the effectiveness of the de-turgescent mechanism is lost and stromalthickening and opacification occur as the cor-nea hydrates.5 Epithelial bullous formationwith recurrent epithelial erosions, stromalscarring, and blindness are the final result ifan adequate population of endothelial cellscannot be restored. Clinically, endothelialcell replacement is achieved by homotrans-plantation of a healthy donor cornea.

    Recent experiments have demonstratedthat corneal transplants bearing culturedhomologous6' 7 or heterologous8 corneal en-dothelial cells remain clear and appear tofunction normally. These experiments haveimportant clinical implications because theability to use cultured corneal endotheliumto repopulate corneal surfaces prior to penet-rating keratoplasty would increase the num-ber of corneas suitable for transplantation aswell as provide a means of standardizing theendothelial population of donor preparations.The purpose of this article is to report the

    0146-0404/82/010073+18$01.80/0 1982 Assoc. for Res. in Vis. and Ophthal., Inc. 73

    Downloaded From: on 11/29/2018

  • 74 Bahn et al.Invest. Ophthalmol. Vis. Sci.

    January 1982


    Fresh Feline Corneo Donor Fresh Bovine Endothelium Donor

    Endothelium ,0 \ .Denuded and \ j h J/Oiscorded ^ = = * 5 '

    8 mm Button

    PenetratingKerotoplasty(Living FelineRecipient)

    Fig. 1. Schematic flow diagram illustrating thetechniques used to produce bovine corneal endo-thelial cell-lined corneal transplants.

    results of a series of feline corneal transplantsrepopulated with bovine corneal endothe-lium that had been grown in monolayer cellculture. Clinical and morphologic evaluationof the transplants demonstrated a host re-sponse directed against the grafted heterolo-gous endotheliuma complication that hasnot been previously reported.

    Materials and methods

    A schematic illustration of the sequence used toproduce heterologous cell-lined feline homograftsis presented in Fig. 1.

    Bovine corneal endothelial cell culture. Primaryendothelial cell cultures were established frombovine eyes that were obtained within 6 hr afterdeath. The eyes were trimned of excess tissue andwashed with running tap water prior to sequentialrinsing of the corneas and adjacent scleras withantibiotics (Neosporin Ophthalmic Solution; Bur-roughs Wellcome Co., Research Triangle Park,N.C., and Aureomycin [50 jug/ml]; Lederle Lab-oratories, Pearl River, N.Y.). Corneas were ex-cised with a 2 mm scleral rim and placed in plastic

    cups fabricated to conform to the corneal contour.Endothelial cells were dislodged with a siliconerubber spatula after a 5 to 7 min incubation at 37C in sterile, pH 7.5, Ca++- and Mg++-free Earle'sBalanced Salt Solution (EBSS) (Gibco, Grand Is-land, N.Y.) containing 5 mM ethylene diamino-tetraacetate (EDTA) and 0.25% trypsin (1:250)(Difco, Detroit, Mich.). Cells from one set of eyeswere collected in Minimum Essential Medium(MEM) (Gibco) with 10% calf serum containing 50yug/ml gentamicin (Garamycin; Schering Corp.,Kenilworth, NJ.) and 160 U/ml nystatin (Myco-statin; E. R. Squibb, Princeton, N.J.). The approx-imately 0.5 x 106 cells, when cultured in 25 cm2

    flasks maintained in 5% CO2 and air, reachedconfluence in 6 to 8 days. Cultures were fed every3 days with the same medium except the nystatinwas omitted.

    At confluence the endothelial cells form ahighly ordered monolayer (Figs. 2 and 3) and con-tinue to produce a basement membrane (Fig. 4)that contains Type IV collagen9 and one or moreglycoproteins common to basement membranes.10

    A comprehensive report of bovine corneal endo-thelial cells in culture and the deposition of an invitro basement membrane has been submitted forpublication."

    Graft preparation and heterologous endothe-lial cell repopulation. Fresh donor cat eyes wereobtained from animals sacrificed by an overdose ofintravenous sodium pentobarbital. The eyes wererinsed in sterile saline and the corneas and adja-cent scleras were flushed with antibiotic (Neospo-rin Ophthalmic Solution). Scleral rim preparationsof the cornea were placed in plastic cups moldedto the contour of the cornea. The endothelial sur-face was covered with MEM containing 5% de-xtran (D-MEM; Sigma Chemical Co., St. Louis,Mo.) (M.W. 40,000) and buffered to pH 7.3 with15 mM HEPES, 10 mM TES, and 10 mM BES(Calbiochem, La Jolla, CA). The native endothe-lial cells were removed by gently swabbing theendothelial surface with a moistened cotton-tipped applicator. After a thorough rinse inD-MEM the denuded corneas were kept in theplastic cups during the initial stages of repopula-tion with cultured bovine corneal endothelium.

    First or second subcultures of bovine cornealendothelial cells were removed from culture flaskswith 5 mM EDTA and 0.125% trypsin as de-scribed below. The action of the trypsin was stop-ped by adding the cell suspension to an equal vol-ume of D-MEM containing 10% calf serum. Analiquot was taken for cell counting, and the re-

    Downloaded From: on 11/29/2018

  • Volume 22Number 1 Heterologous cell-lined corneal homografts 75

    main ing cells were centrifuged and subsequentlywashed once in serum-free D-MEM. The cellswere diluted with serum-free D-MEM to yieldapproximately 1 X 105 cells/60 fx\. The 60 /A in-oculum was added to the denuded corneas and thepreparations were incubated in a humidified at-mosphere for 2 hr. After four gentle rinses inD-MEM to wash off nonadherent cells the corneaswere incubated for an additional 12 to 18 hr in 10ml of D-MEM containing gentamicin (100 /Ug/ml)and amphotericin B (4 /ig/ml). The addition ofthe dextran to the MEM effectively minimizedstromal swelling.

    The optimal protocol for subculturing bovinecorneal endothelial cells onto feline homograftswas experimentally determined and found to bedependent on the status of the particular mono-layer culture. In most instances endothelial cellswere removed from flasks with 0.125% trypsin and5 mM EDTA 24 to 48 hr after subculturing at a 1:2split ratio, a process that both facilitated detachingcells from the flask bottom (2 min at 20 C) andenhanced early spreading of the cells onto Desce-met's membrane. Cultures that had been conflu-ent for 1 to 2 weeks required incubation at 37 C in0.25% trypsin and EDTA for 5 to 7 min to achieveefficient cell detachment. We believe the 1 to 2week postconfluent cells attached with almostequal efficiency as the recently subcultured cellsdescribed above, but these cells required a muchlonger period of time (10 to 12 hr) to spread ontothe feline Descemet's membrane. The signi-ficantly lower amount of trypsin (0.005%) and 10mM EDTA used successfully by Gospodarowiczet al.12 required a 60 to 120 min incubation at37 C before cells began to detach in our systemand therefore were not used for the subculturingof endothelial cells onto feline corneas.

    In vitro specular microscopy, alizarin red stain-ing1 (Fig. 8, a), and scanning electron microscopyof representative feline corneas from which theendothelium had been removed by swabbingdemonstrated an infrequent remaining brokencell; however, the membranes appeared remark-ably free of cells considering the nature of the de-nuding technique. Similar studies of the repopu-lated grafts indicated approximately 50% to 60%(Figs. 5 and 8, b) of the attached cells had begun tospread onto the surface of Descemets membraneat 2 hr and that a complete, although irregularlyarranged (when compared with the native en-dothelium), monolayer was formed by 12 to 18 hrafter adding the cultured bovine cells (Figs. 6, 7,and 8, c).



View more >