Full text of DOI:10.1007/s004170100354

Graefe’s Arch Clin Exp Ophthalmol
(2001) 239:778–782
L A B O R AT O RY I N V E S T I G AT I O N
DOI 10.1007/s004170100354
Torben Møller-Pedersen
Ulrike Hartmann
Niels Ehlers
Evaluation of potential organ culture media
for eye banking using a human corneal
endothelial cell growth assay
Katrin Engelmann
Abstract Background: To evaluate
the ability of different commercially were evaluated by phase-contrast mi-
available cell culture media to in- croscopy. Results: The morphologi-
duce proliferation and morphological cal characteristics of HCEC were
whereas changes in cell morphology
Received: 11 October 2000
Revised: 4 April 2000
Accepted: 13 July 2001
Published online: 25 September 2001
© Springer-Verlag 2001
changes in primary cultures of hu-
man corneal endothelial cells
best maintained in the group of F99-
based media, which also induced the
highest level of cell proliferation un-
der serum-reduced conditions. Spe-
(HCEC). This screening model was
used in an attempt to establish a
rational basis for the development of cifically, the medium F99-Sr (F99
well-defined, serum-free preserva-
tion media for long-term organ
culture of human donor corneas.
enriched with ascorbic acid, insulin,
bFGF, transferrin, selenium, and lip-
ids) induced a two- to three-fold
Methods: A total of 11 different cul- higher HCEC density at both 0% and
ture media enriched with 0%, 2%,
5%, and 10% fetal calf serum (FCS)
were compared. The test media
were divided into three groups:
Group 1: Media based on minimal
essential medium (MEM), currently
used for long-term corneal organ
culture in European eye banks;
2% FCS when compared to all other
test media, and it also maintained the
most endothelial cell-like morpholo-
gy. Also, at higher serum concentra-
tions (5% and 10% FCS), the cell
growth was most prominent in
This study was performed at the Cornea
Bank of Hamburg and was supported by
the German Research Council (Deutsche
Forschungsgemeinschaft; Ha 2484/1-1)
T. Møller-Pedersen ( ) · N. Ehlers
✉
Department of Ophthalmology,
Århus University Hospital,
8000 Århus C, Denmark
e-mail: tmp@dadlnet.dk
F99-Sr, as well as in the medium
SFM that originally was designed for
Group 2: F99-based media, enriched serum-free growth of vascular endo-
U. Hartmann
for growth of corneal endothelial
cells at serum-reduced conditions;
and Group 3: Media designed for
growth of special cell types or for
short-term corneal organ culture.
The growth-promoting capacity of
each test medium was quantified
using an HCEC proliferation assay,
thelial cells. Conclusion: This study
suggests that the media F99-Sr and
SFM should be further tested and
refined as potential new storage
solutions for long-term corneal
organ culture at physiological tem-
peratures.
Augenklinik der
Medizinischen Hochschule Hannover,
Carl-Neuberg-Straße 1,
30625 Hannover, Germany
K. Engelmann
Universitäts-Augenklinik Hamburg,
Martinistraße 52, 20246 Hamburg,
Germany
ture media ingredients and storage conditions are being
used with no well-established consensus among the
Introduction
In European eye banks, organ culture is the preferred European eye banks [15]. Both the basic media composi-
technique for storage of human donor corneas at physio- tion and the level of bovine serum supplementation vary
logical temperatures for up to 4–7 weeks [1, 2, 4, 5, 11, considerably, and no systematic comparison and optimi-
14, 15]. Annually, more than 14,000 corneas are pro- zation of the present culture solutions has been per-
cessed using this method. Currently, a wide range of cul- formed. However, it has been demonstrated that differ-

779
ences in the specific organ culture conditions can induce
substantial variation in postoperative graft quality and
performance [7, 10,]. Thus, because of an increasing
clinical demand for donor tissue of high quality, as well
as an increasing legal demand for standardized proce-
dures, there is a fundamental need to compare and con-
trast the current preservation media. Furthermore, it is
desirable to search for serum-free culture solutions to
Materials and Methods
Test media
A total of 11 commercially available cell culture solutions were
tested (see company information sheets for details on the specific
ingredients). Due to previously identified variations in preparation
and purification procedures, some of the test media had the same
general biochemical composition but were provided by different
manufacturers. For descriptive purposes, the media were divided
rule out the potential risk of infections and other unwant- into three groups (Table 1).
ed side-effects from unknown and variable bovine or hu-
man serum supplements.
Group 1 consisted of minimal essential medium-based solu-
tions (MEM). This type of medium contains relatively low num-
bers of supplements and nutrients and was never intended for use
without serum supplementation due to the lack of several essential
components. However, MEM-based media supplemented with
2%–10% FCS are generally used for corneal organ culture in
European eye banks [15]. In this study, we tested both regular
MEM solutions (MEM-S and MEM-H), as well as supplementa-
tion with 2 mM stable glutamine (MEM-G) and growth factors
thymidine and hypoxanthine (OPTI-MEM) (Table 1).
In an attempt to establish a rational basis for the de-
velopment of well-characterized corneal organ culture
conditions, we have initiated experimental studies to
quantitatively evaluate the impact of a wide range of
commercially available basal culture media (that may
work as starting points for subsequent refinements). Tra-
ditionally, the endothelial cell density and morphology is
used as the most critical parameter defining the quality
and suitability of a donor cornea for transplantation [2,
15]. In the search for useful screening parameters to test
Group 2 consisted of F99-based media (1:1 mixture of M199
and Ham’s F12) designed for serum-reduced growth of HCEC [8,
10]. This type of medium is highly enriched with multiple essen-
tial nutrients and supplements including vitamins, coenzymes,
amino acids, trace elements, and purines. To test the effects of fur-
possible new organ culture solutions, it therefore seems ther enrichment, more supplements (ascorbic acid, insulin, bFGF,
reasonable to use the growth of human corneal endothe-
lial cells (HCEC) as an initial screening assay. This mod-
transferrin, selenium, and lipids) were added to the medium
F99-Sr, as described previously [8].
Group 3 was a heterogeneous group consisting of media devel-
el has been previously shown to be valid and sensitive in
the assessment of storage media quality [10, 7]. There-
fore, in this study, we compared the ability of different
preservation media to induce proliferation and morpho-
logical changes in primary cultures of HCEC. This paper
summarizes our initial evaluation of 11 different cell cul-
ture solutions supplemented with varying concentrations
of fetal calf serum (FCS) ranging from 0% to 10%.
oped for serum-free growth of special cell lines including hybri-
doma cells (DIF-1000) and vascular endothelial cells (SFM)
(Table 1). Furthermore, the basal medium M199, intended for
short-term corneal preservation at 4°C [2], was included – togeth-
er with a Chemically-Defined-Medium (CDM) [16].
Human corneal endothelial cell growth assay
Primary cultures of HCEC were established from four adult do-
nors (age range 50–70 years) and subcultured in L-valine-free me-
Table 1 Specifications of the 11 test media
Name
Basic Component
Comments
Manufacturer
Group 1 MEM-S
MEM-H
MEM-Earle
MEM-Earle
Prepared by Seromed
Prepared by Hamburg University Pharmacy
Seromed
Hamburg
University Pharmacy
MEM-G
MEM-Earle
MEM supplemented with 2 mM L-analyl-L-glutamine,
a stabilized form (dipeptide) of the amino acid L-glutamine
Seromed
OPTI-MEM MEM
Enriched with growth factors, thymidine, and hypoxanthine
Gibco
Group 2 F99-H
F99-G
M199 + Ham’s F12 (1:1) Designed for corneal endothelial cell growth [8]
M199 + Ham’s F12 (1:1) Designed for corneal endothelial cell growth [8]
Hyclone
Gibco
F99-Sr
M199 + Ham’s F12 (1:1) F99 supplemented with ascorbic acid,
insulin, bFGF, transferrin, selenium, and lipids
(see the specific concentrations in [8])
Gibco
Group 3 DIF-1000
Company formula
Company formula
M199-Earle
Initially used for serum-free growth of hybridoma cells
Developed for serum-free growth of vascular endothelial cells Gibco
Intended for short-term corneal preservation at 4°C
(cold storage for up to ten days)
Seromed
SFM
M199
Seromed
CDM
“Mixture”
M199, Ham’s F-10, Dulbecco’s modified-Eagle medium
(1:1:1) (ref. 16)
Gibco

780
Table 2 Predominant cell morphology following 18 days of incu-
bation in 11 test media
Test media
HCEC morphology
Group 1
MEM-S
Elongated
MEM-H
MEM-G
OPTI-MEM
Elongated
Elongated
Spindle shaped
Group 2
Group 3
F99-H
F99-G
F99-Sr
Elongated
Rounded
Rounded
DIF-1000
SFM
M199
CDM
Spindle shaped
Spindle shaped
Elongated
Elongated
dium (for four passages) to avoid contaminating fibroblasts as de-
scribed previously [9, 6]. At this time point, all established cell
lines expressed the typical rounded endothelial cell-like morpholo-
gy. Then, onto 6 cm2 wells coated with a mixture of laminin and
chondroitin sulfate, 800 cells per well were seeded in medium
F99. This low cell-seeding density allows for a sufficient number
of population doublings in order to identify the ability of each test
medium to promote long-term cellular growth. The HCEC were
incubated at 37°C and allowed to attach for 18–20 hours. Subse-
quently, F99 was replaced by one of the 11 test media supplement-
ed with 0%, 2%, 5%, and 10% FCS (batch B: 425C, Seromed,
Germany) and 10 µg/ml gentamicin. All media were changed ev-
ery third day and tested in parallel. After 18 days, when the HCEC
covered approximately two-thirds of the surface, the experiment
was terminated and the final cell number was counted using a
Coulter Counter ZM (Counter Electronics, Luton, England) as pre-
viously reported [6, 9]. Thus, the cells were not grown into a con-
fluent monolayer, in order to avoid contact inhibition. The assay
was repeated four times for all 11 media and four serum concen-
trations, giving a total of 176 cell cultures.
Fig. 1 Proliferation of primary HCEC cultures induced by 11 test
media without serum or supplemented with 2% FCS. At 0% FCS,
(*) indicates P<0.05 for F99-Sr versus all other media. At 2%
FCS, (*) indicates P<0.05 for F99-Sr versus all other media ex-
cept for F99-H and SFM
Results
Serum-free incubation
Among the 11 test solutions, the medium F99-Sr induced
the highest level of cell proliferation (three population
doublings) under serum-free conditions (Fig. 1); the final
HCEC density was approximately two- to three-fold
higher than all other media (p<0.05). Compared to the
two nonsupplemented F99 media (F99-H and F99-G),
the medium F99-Sr was specifically enriched with: as-
corbic acid, insulin, bFGF, transferrin, selenium, and lip-
ids (see ref. 8 for the specific concentrations). Thus, it
appeared that the addition of these six components
played an important role in HCEC growth regulation,
which is in agreement with previous observations [8].
Moreover, the cells kept in F99-Sr maintained a rounded
endothelial-like morphology with a normal nucleus/cyto-
plasm ratio and few cytoplasmatic vacuoles during the
18-day incubation (Fig. 2a, Table 2).
Morphological evaluation
At day 18, the predominant cell morphology was graded using
phase contrast microscopy and the following criteria: rounded,
elongated, or spindle-shaped (Table 2). This subjective evaluation
was done according to previously published photographs [6, 8, 9,
10] by an experienced observer, although not in a masked design.
Since the HCEC were not grown into confluence, they were not
expected to express the cobblestone-like morphology of contact-
inhibited endothelial cells as seen in vivo or in primary culture [6,
8, 9, 10].
Statistics
Serum supplementation
Data are given as mean ± SEM (n=4) and represent increase in
cell numbers per well. Using SigmaStat (SPSS Inc., Chicago), dif-
ferences in mean values among the groups were compared by
One-Way Analysis of Variance, and Student-Newman-Keuls
method for all pairwise comparisons.
Addition of 2% FCS generally promoted HCEC growth
in all test media, particularly in the group of F99-based
solutions (Fig. 1). Again, the highest cellularity was in-
duced by the medium F99-Sr (Fig. 1), which also main-
tained the morphological characteristics of typical
rounded endothelial cells (Fig. 2b, Table 2). Addition of
5% FCS further stimulated cellular division in all test so-
lutions, again with the highest proliferative capacity pro-
vided by F99-Sr (Fig. 3). Cells kept in F99-Sr continued

781
to show a rounded morphology, whereas cells incubated
in several other media (including MEM-based solutions)
presented more elongated and spindle-shaped features
(Table 2). Also, at 10% FCS, the cellular growth was
most prominent in F99-Sr as well as in the medium
SFM; however, cells kept in SFM generally appeared
spindle-shaped, independent of the serum-concentration.
This SFM medium was originally designed for serum-
free growth of vascular endothelial cells but the exact
composition is currently not available from the manufac-
turer.
Discussion
Development of optimal storage conditions for long-term
corneal organ culture requires a valid screening assay to
evaluate systematically the ingredients needed to simu-
late the complex physiological environment. Currently,
there are no specific methods available for monitoring
the nutritive requirements of the entire donor cornea. For
this reason, we have used the response of primary cul-
tures of HCEC as an initial screening model. This ap-
proach previously has shown to be valid in the assess-
ment of storage media quality [7, 10], although the nutri-
tive requirements of proliferating cells may not mirror
exactly those needed for quiescence. This limitation
must be kept in mind when the present data are inter-
preted.
Nevertheless, it might be beneficial to choose a cul-
ture medium with a high HCEC growth-promoting ca-
pacity for corneal organ culture. Thus, it is well estab-
lished that the intact human donor endothelium has a dy-
namic potential for cell division and migration in vitro
and is able to repair even large defects during preserva-
tion [3, 13, 16, 17]. Although this in vitro regenerative
capacity is limited, it may be possible to promote prolif-
eration of the intact donor endothelium during storage in
order to prevent and reverse the usual loss of 10%–25%
endothelial cells during 4–5 weeks of organ culture [11,
12]. Thus, by applying certain growth factors and addi-
tives to the storage medium, endothelial cell division
might be appropriately stimulated without inducing cel-
lular dedifferentiation [16]. Obviously, this perspective
needs to be studied further, but it could be a useful strat-
egy to avoid exclusion of many donor corneas for trans-
plantation due to low endothelial cell counts.
Fig. 2A, B Phase-contrast microscopy of HCEC morphology fol-
lowing 18 days of incubation in medium F99-Sr without serum
(A) or with 2% FCS (B). Note that the rounded endothelial cell-
like morphology is maintained independent of the level of serum,
although the cellularity is higher at 2% FCS
The present HCEC growth assay clearly demonstrates
that the specific composition of the basic culture solu-
tions influences both cell proliferation and morphology,
independent of the level of serum supplementation
(Figs. 1 and 3, Table 2). Among the 11 media tested, the
medium F99-Sr revealed the highest growth-promoting
capacity under both serum-free conditions and following
enrichment with 2%–10% FCS. This is an important
finding, especially since the proliferating cells in F99-Sr
Fig. 3 Proliferation of primary HCEC cultures induced by 11 test
media supplemented with 5% or 10% FCS. At 5% FCS, (*) indi-
cates P<0.05 for F99-Sr versus all other media, except for F99-G
and SFM. At 10% FCS, (*) indicates P<0.05 for F99-Sr versus all
other media, except for SFM and M199; and P<0.05 for SFM ver-
sus all other media, except for F99-Sr

782
maintained their normal rounded endothelial cell mor- proliferation under serum-free conditions, especially
phology, regardless of the serum concentration (Fig. 2, following serum supplementation (Figs. 1 and 3). How-
Table 2). Thus, it appears that this special mixture of ever, cells kept in SFM usually appeared spindle-shaped
F99, ascorbic acid, insulin, bFGF, transferrin, selenium, (Table 2).
and lipids [8] provides an excellent environment for hu-
In conclusion, the present study provides an initial
man corneal endothelial cells that may have an important evaluation and comparison of 11 commercially available
potential for corneal organ culture. It should be noted cell culture media. Assuming the HCEC growth assay
that the MEM-based solutions currently used in Europe- can be used as a relevant screening model, it could be
an eye banks [15] revealed only a moderate proliferative advantageous to replace the current MEM-based solu-
capacity, even following enrichment with growth factors tions with F99-Sr-based media for long-term corneal or-
and other supplements (Figs. 1 and 3, Table 1). More- gan culture. Also, the medium SFM may have an impor-
over, MEM-based solutions generally induced a more tant potential for corneal preservation. Obviously, both
elongated morphology (Table 2). Another potential can- these media need to be evaluated further in an organ cul-
didate for a new organ culture solution appeared to be ture test system using intact human donor corneas. Such
the medium SFM that also induced a high level of HCEC experiments are underway in our laboratories.
References
1. Andersen J, Ehlers N (1987) Corneal
transplantation using 4–week banked
donor material. Long-term results.
Acta Ophthalmol Copenh 65:293–299
2. Böhnke M (1991) Corneal preservation
in organ culture. Curr Opin Ophthalmol
2:432–442
3. Doughman DJ, Van Horn D, Rodman
WP, Byrnes P, Lindstrom RL (1976)
Human corneal endothelial layer repair 10. Engelmann K, Sobottka Ventura A,
8. Engelmann K, Friedl P (1995) Growth
of human corneal endothelial cells in a
serum–reduced medium. Cornea
14:62–70
9. Engelmann K, Böhnke M, Friedl P
(1988) Isolation and long–term
cultivation of human corneal endotheli-
al cells. Invest Ophthalmol Vis Sci
29:1656–1662
14. Møller–Pedersen T, Møller HJ (1996)
Viability of human corneal keratocytes
during organ culture. Acta Ophthalmol
Copenh 74:449–455
15. Pels L, Mass H, Tullo A (2000) Euro-
pean Eye Bank Association Directory.
Eighth Edition. Available from the
Netherlands Ophthalmic Research In-
stitute, Amsterdam
16. Schultz G, Cipolla L, Whitehouse A,
Eiferman R, Woost P, Jumblatt M
(1992) Growth factors and corneal en-
dothelial cells. III. Stimulation of adult
human corneal endothelial cell mitosis
in vitro by defined mitogenic agents.
Cornea 11:20–27
17. Simonsen AH, Sørensen KE, Sperling
S (1981) Thymidine incorporation by
human corneal endothelium during or-
gan culture. Acta Ophthalmol Copenh
59:110–118
during organ culture. Arch Ophthalmol
94:1791–1796
Drexler D, Staude HJ (1998) A sensi-
tive method for testing the quality of
organ culture media and of individual
medium components in a cornea bank.
Graefes Arch Clin Exp Ophthalmol
236:312–319
4. Ehlers H, Ehlers N, Hjortdal JØ (1999)
Corneal transplantation with donor tis-
sue kept in organ culture for 7 weeks.
Acta Ophthalmol Scand 77:277–278
5. Ehlers N, Hjortdal JØ, Møller–
Pedersen T (1994) Corneal storage and
complications related to grafting.
Curr Opin Ophthalmol 5, IV:75–80
6. Engelmann K, Friedl P (1989) Optimi-
zation of culture conditions for human
corneal endothelial cells. In Vitro Cell
Dev Biol 25:1065–1072
11. Frueh BE, Böhnke M (1995) Corneal
grafting of donor tissue preserved for
longer than 4 weeks in organ-culture
medium. Cornea 14:463–466
12. Hjortdal JØ, Ehlers N, Andersen CU
(1989) Some metabolic changes during
human corneal organ culture. Acta
Ophthalmol Copenh 67:295–300
7. Engelmann K, Winter R (1993) Quality 13. Hoppenreijs VP, Pels E, Vrensen GF,
control in the corneal bank–a necessary
measure? Klin Monatsbl Augenheilkd
203:262–268
Treffers WF (1994) Basic fibroblast
growth factor stimulates corneal endo-
thelial cell growth and endothelial
wound healing of human corneas.
Invest Ophthalmol Vis Sci 35:931–944