Cell Cultures of Human Ciliary Muscle: Growth

Exp. Eye Res. (lY91)
53, 375-387
Cell Cultures
of Human Ciliary Muscle: Growth,
and lmmunocytochemical
of Anatomy,
19 September
II, University of Erlangen-Niirnberg,
Erlangen, Germany
1990 and accepted in revised form 13 December
9, 8520
Primary ciliary muscle cell cultures derived from human donors (16-9 1 years) were established and
characterized by comparing them with ciliary muscle in tissue sections using immunocytochemical
ultrastructural methods. Monoclonal antibodies against desmin, vimentin, a-actinin, smooth muscle (sm)
specific a-actin and von Willebrand factor were used. In tissue sections of the ciliary body, ciliary muscle
cells, vascular muscle cells, pericytes, endothelial cells and fibroblasts stain for vimentin. Both types of
muscle cells and the pericytes stain for cr-sm-actin, but only ciliary muscle cells stain for desmin. For tissue
cultures, explants of the meridional and partly the reticular portion of the ciliary muscle were dissected
and grown directly or after digestion of the explant with collagenase. Ten primary cell cultures with a
typical hill-and-valley growth pattern similar to smooth muscle cells and two with a growth pattern
similar to fibroblasts were established. All cultures could be subcultured up to the fifth passage. In
fibroblast-like cultures S-10% of the cells stained for z-sm-actin. Staining for desmin was not observed.
In smooth muscle-like cultures, all cells stained positive for a-sm-actin. Desmin staining was not seen in
growing non-confluent smooth muscle-like cultures. In confluent cultures, about 10% of the cells stained
positive for desmin, preferentially in areas where the cells had formed hills. No culture stained for von
Willebrand factor. Staining for a-actinin in smooth muscle-like cultures showed that the dense bands of
the myofilaments were arranged in register, similar to the typical ciliary muscle cell morphology seen in
tissue sections. Ultrastructurally,
the smooth muscle-like cultures showed the typical morphology of
cultured smooth muscle cells. We conclude that the smooth muscle-like cultures consist of ciliary muscle
Key words : ciliary muscle ; smooth muscle ; tissue culture ; a-sm-actin ; desmin ; vimentin ; a-actinin :
; ultrastructure : human eye.
1. Introduction
It is well established that human ciliary muscle
undergoes marked structural changes with age (Kerschbaumer, 1888; Fuchs, 1928; Stieve, 1949; van
der Zypen, 19 70). There is an increase in extracellular
material and a pronounced sclerosis and hyalinization
of the connective tissue between the ciliary muscle
bundles. The age-related changes of the extracellular
well affect the function
of the muscle in
accommodation and influence uveoscleral outflow.
Local treatment of monkey eyes with prostaglandin F,,
causes lysis of connective tissue elements between the
ciliary muscle bundles (Liitjen-Drecoll and Tamm,
1988, 1989: Tamm et al., 1989). This might explain
why prostaglandin F,, increases uveoscleral outflow in
this species (Nilsson et al., 1989).
Extracellular ‘plaque-material’ is found in increased
amounts in the trabecular meshwork of human eyes
with primary open angle glaucoma (Rohen and
Wittmer. 1972 ; Liitjen-Drecoll et al., 1986a). This
material is also increased in the anterior portion of the
ciliary muscle of glaucomatous eyes (Liitjen-Drecoll
et al., 1986b). The reasons for this increase in extra*
cellular material in old and glaucomatous eyes are still
Cell culture techniques have become an important
tool in the study of cellular physiology and the
synthesis of extracellular material by smooth muscle
cells (Burke and Ross, 19 79). To obtain further insight
into cellular changes associated with age or glaucoma,
we established primary cultures of ciliary muscle cells
from donors of different ages. The cultures were
characterized by comparing the cells with ciliary
muscle cells in tissue sections using immunocytochemical and ultrastructural methods.
2. Materials and Methods
Cell Culture
Ciliary muscle cell cultures were established from
post-mortem human eyes, donor ages 16-91 years.
The eyes were enucleated within 248 hr of death.
Immediately after enucleation, they were placed in
medium 199 containing
50 U ml-’ penicillin and
50 pugml-’ streptomycin (all Gibco Ltd, Paisley, Scotland) for 10 min prior to dissection. Each eye was
bisected along the equator and the posterior half
removed. The anterior half was placed on a sterile
0 199 1 Academic Press Limited
Primary cell cultures derived from ciliary muscle
~. ~~
Type A cells
(smooth muscle-like-cells)
Type B cells
(fibroblast-like cells)
time (hr)
Cell suspensiont
Whole explant*
Whole explant*
Whole explant*
Whole explant*
Whole explant*
Cell suspensiont
Whole explant*
Cell suspension?
Cell suspensiont
Whole explant*
Whole explant*
Explants from ciliary muscle were placed as a whole in tissue culture* or digested with collagenase to obtain a suspension of cellst which
were allowed to settle in tissue culture dishes.
Petri-dish, cornea1 side down. The zonule was cut with
fine scissors and the Iens removed. A small segment of
the anterior half with a width of approximately 5 mm
was cut meridionally
and further processed for
(see ’ Antibody staining ‘). The
remaining part of the anterior half of the eye was
processed for tissue culture of ciliary muscle. One
branch of a fine forceps was inserted in the supraciliary
or suprachoroidal space of the specimen The ciliary
body was grasped and the ciliary tissue gently pulled
away from the underlying sclera and thereby torn
from its attachment to the scleral spur. The uveal
explant, containing ciliary processes, ciliary muscle
and iris was placed in a new sterile Petri dish with the
muscle on top. Against the dark background of the
pigmented epithelium, the muscle could be easily
distinguished, as a broad. pale, circular band. Under
the operating microscope ( x 40), the outermost parts
of the ciiiary muscle were cut free with fine scissors
and placed in medium 199. Before continuing, two
samples approximately 2 mm in length were cut from
the muscle to be explanted. One sample was deep
frozen and immediately sectioned on a cryostat. The
other sample was fixed in Ito’s solution (2.5%
glutaraldehyde, 2.5 % paraformaldehyde and picrinic
acid : Ito and Karnovsky 1968) and embedded in Epon.
After polymerization, semi-thin and ultra-thin sections
were cut and used as additional controls.
If the frozen sections confirmed that only uncontaminated muscular tissue had been removed, the
muscle explants were further processed. Two methods
were used (Table I):
Method 7, The samples were incubated for 30 min at
37°C in Hank’s balanced salt solution (HBSS), supple-
mented with 1 mg ml-’ collagenase (collagenase type
V, Sigma Chemical Co., St Louis, MO). The samples
were then placed in a 3 5-mm sterile, uncoated, plastic
Petri dish and covered with a coverslip. One milliliter
of medium 199 was added, supplemented with 20%
fetal calf serum (Gibco Ltd) penicillin and streptomycin. The explants were incubated at 37°C in
humidified air enriched with 5 % CO,. After 2 days the
medium was renewed, thereafter twice weekly. The
fresh medium contained the same antibiotics and 10 %
fetal calf serum. Primary cultures were examined
routinely using positive phase contrast optics. Care
was taken to make sure that only cells morphologically
characteristic of smooth muscle cells in viva (type A
cells, see Results) were kept in culture. If the
impression was that fibroblast-like cells (type B) grew
out from the explant, the area covered by these cells
was marked with a pencil on the underside of the Petri
dish. Under sterile conditions, the coverslip was iifted.
Then, the marked part of the explant was cut out and
removed and the outgrowing
fibroblast-like cells
scratched away with a sterile needle. The procedure
was controlled with phase contrast optics and repeated
if necessary. Two cell lines were established by
cuIturing the excised parts of the explants separately.
They contained the type B or fibroblast-like cells
Method 2. The explants were placed in collagenase
(1 mg ml-‘, collagenase type V, Sigma) for 2 hr with
occasional gentle agitation. After this time the major
part of the explants had been dispersed into single
cells. The suspension was centrifuged at 900 rpm for
10 min. resuspended in 2 ml medium 199 with 20%
fetal calf serum. penicillin and streptomycin and
injected into a 3 S-mm sterile, fibronectin-coated,
plastic Petri dish. After 2 days, the medium was
changed as described above.
Confluent cultures were passagedat a ratio of about
1: 4 using trypsin-EDTA.
Antibody Staining
The specimens were divided meridionally in two
parts of equal size. One half was rapidly frozen in
isopentane cooled with liquid nitrogen. The other half
was fixed in paraformaldehyde-lysin-periodate solution (McLean and Nakane, 1974) for 4 hr at 4°C
washed for 24 hr in phosphate-buffered saline (PBS),
then dehydrated in graded alcohols and embeddedin
paraffin. Ten-micrometer sections were cut from the
frozen specimen at a temperature of -20°C and
mounted on chrome-alum-gelatine coated glass slides.
Fixation of the frozen sections was performed with
acetone for 10 min at -20°C. After drying at room
temperature, the frozen sections were pre-incubated
for 45 min in Blotto’s non-fat dry milk solution
(Duhamel and Johnson, 1985).
From the paraffin-embedded specimen, meridional
sections were cut at 5 pm and mounted on 0.1% polyI-lysin-coated slides. The sections were deparaffinized
and pre-incubated for 20 min in normal rabbit serum
(5 o/oin PBS).
After pre-incubation, both frozen and paraffin
sections were incubated with the primary antibody for
90 min. For demonstration of the intermediate filament desmin, monoclonal mouse antibodies (Dakopatts, Hamburg, Germany) were used [clone DE-R-l 1,
anti-swine IgGl (Debus, Weber and Osborn, 1983)
and clone D33, anti-human IgGl) both at a dilution of
1: 50. For vimentin, a monoclonal mouse anti-swine
antibody from Dakopatts (clone V9, IGGl: Osborn,
Debus and Weber, 1984) at a dilution of 1: 5 was
applied to the slides. The antibodies were diluted in
0.1 M PBS, pH 7.2-7.4. Demonstration of a-smooth
muscle (sm)-actin was performed with a monoclonal
mouse anti-bovine antibody to smooth muscle specific
a-sm-actin from Sigma (clone no. lA4, IgG2a: Skalli
et al., 1986) diluted in PBS (1:lSO) with 1% bovine
serum albumin (Sigma). Each incubation was performed in a moist chamber at room temperature. The
sections were washed three times with PBS, each for
10 min and incubated with fluorescein or rhodamineladelled rabbit-anti-mouse IgG (Dakopatts) diluted
with PBS (1:20). The sections were washed three
times in PBS and mounted in Entellan (Merck,
Darmstadt, Germany) containing 2.5 % 1,4-diazobicycle-octane (Merck: Johnson et al., 1982). Control
experiments were performed using either PBS or a
mouse pre-immune serum instead of the primary
For staining of cell cultures, cells were seeded in
tissue culture chambers mounted on Permanoxw
microslides (Lab Tek*, Nunc Inc. Naperville, IL). Cells
from the first to fifth passages of all established
cultures were stained 1, 4, 7 and 21 days after
seeding. Culture medium was removed by rinsing
three times with PBS. The cells were fixed with
methanol for 3 min at -20% then incubated for
90 min with the same primary antibodies at the same
dilution as used for tissue staining. Additionally, the
cells were stained with monoclonal mouse antibodies
against human von Willebrand-factor (factor VIII
associated antigen) from Dakopatts, clone F8/86,
IgGl, (Naiem et al., 1982) at a dilution of 1:50 and
bovine a-actinin from Sigma, clone no. BM-75.2, IgM,
at a dilution of 1: 200. The cells were rinsed three
times with PBS then incubated for another 60 min
with a fluorescein-conjugated rabbit anti-mouse IgG
(Dakopatts) at a dilution of 1:20. The cells were
washed again three times with PBS, then the tissue
chambers were removed from the glass slides. The
slides were mounted in Entellan containing 2.5 % 1,4diazobicyclo-octane. Control experiments were performed using either PBS or a mouse pre-immune
serum instead of the primary antibody.
Stained cells and tissue sections were viewed with a
L&z Aristoplan photomicroscope (Ernst Leitz GmbH.
Wetzlar, Germany) equipped with epifluorescence
optics and appropriate filters. A Kodak T-max 400
black and white lilm or Ektachrome 400 colour slide
film (Kodak Limited, Hempstead, U.K.) were used for
The ultrastructure of cells from pre-confluent cultures (1 and 4 days), confluent (7 days) and highly
confluent cultures (3 and 4 weeks) of second to fifth
passages was investigated from all established cell
cultures. The cells were grown in uncoated, plastic
Petri dishes or in tissue culture flasks and fixed with
Ito’s solution for 4 hr. The flxed cells, still in the plastic
dish, were post-fixed with 1% osmium tetroxide,
dehydrated with graded alcohols and embedded in
Epon. Polymerization was done at 60°C. Tangential
and perpendicular sections of the cells were cut on an
ultramicrotome. The sections were treated with lead
citrate and uranyl acetate. For electron microscopic
examination a JEOL(JEM 100 B) and a Zeiss(EM 9021
electronmicroscope were used.
3. Results
Tissue sections. Ciliary muscle cells in tissue sections
showed a pronounced immunostaining with antibodies against smooth muscle (s.m.) specific a-actin
[Fig. l(A)]. This was true for all parts of the ciliary
muscle. Similar staining was also observed in the
vascular smooth muscle cells of the media of the major
arterial circle of the iris [Fig. 1(A)] and in pericytes.
which were rarely observed around the capillaries of
FIG. 1. Immunocytochemistry
of ciliary muscle in tissue sections. A, Ciliary muscle cells and vascular muscle cells both stain
positive for a-sm-actin. The vascular muscle cells belong to the vessels of the major arterial circle of the iris (arrows). which
pass near the muscle’s inner circular portion, (paraffin section, x 230). B, Positive staining for a-sm-actin is also seen in the
pericytes around some of the ciliary muscle capillaries (arrows:paraffi
section. x 1000). C. Stain for desmin is seen in ciliary
muscle cells only. The vascular muscle cells (arrow) are not stained (frozen section, x 300). D, Endothelial cells, vascular muscle
cells around the ciliary body arteries (asterisk) and the fibroblasts (arrows) in the ciliary muscle stain for vimentin. Faint
immunostaining for vimentin is also seen in all ciliary muscle cells (arrowhead: frozen section, x 300).
FIG. 2. Histological control section of a ciliary muscle
explant. The explant consists of muscle bundles from the
meridional and partly the reticular portion. No ciliary
epithelium. or stroma from the ciliary processesor from the
choroid is present. Vessels with a typical thick media of
smooth muscle ceils are not seen (paraffin section, Crossman’s strain, x 150).
the ciliary muscle [Fig. l(B)]. The fibroblasts of the
ciliary muscle and of the stroma of the ciliary processes
did not stain. There were no differences in intensity or
distribution of the stain, when frozen sections were
compared with paraffin sections. In contrast, only
ciliary muscle cells stained for desmin, not the
pericytes, vascular smooth muscle cells or fibroblasts
of the ciliary body [Fig. l(C)]. The intensity of the
staining was much more pronounced in frozen
sections than in paraffin sections. Endothelial cells,
pericytes, vascular smooth muscle cells and fibroblasts
stained for vimentin on paraffin sections. On frozen
sections, however, faint immunostaining for vimentin
was also observed in all ciliary muscle cells [Fig. l(D)].
Cell Culture
Both light and electron microscopy confirmed that
the explants from the ciliary muscle of all donors
consisted only of the meridional and partly the
reticular portion of the ciliary muscle. No ciliary
epithelium, scleral spur, trabecular meshwork and
tissue from the stroma of the ciliary processesor from
the choroid were present (Fig. 2). Vesselswith smooth
muscle cells in the media, which are located in vivo
near the circular portion of the ciliary muscle were not
observed. The size of the explants varied with the age
of the donors. Explants, which were harvested from
older donors (> 60 year old) were about half the size
of those from younger donors (< 35 year old).
If the explants were cultured as a whole (method l),
initial outgrowth of cells appeared within 24 weeks
after placing the explants in culture [Fig. 3(A)]. Explants from all dissected younger donors ( < 35 year
old) grew out even when the post-mortem time
extended up to 48 hr (Table I). In contrast, outgrowth
of explants from donors more than 35 year old could
only be established, if the post mortem time was
shorter than 5 hr (Table I). Cells grew from all sidesof
the explant and were characteristically bipolar, long
and cylindric or ribbon-like with oval nuclei. The cells
showed long tapering fusiform ends, which usually
formed smaller lateral filopodia. When the cells grew
to higher densities, they were still elongated but
became thinner. At this stage, the ends of the cells
usually showed no branching, the cells were closely
attached to each other and arranged themselves in a
parallel fashion [Fig. 3(B)]. Thus, the culture became
organized into a monolayer consisting of longitudinal
and slightly curved bands of parallel cells. This growth
pattern was seenin all primary cell cultures regardless
of the age of the donors. The cells of the primary
cultures grew to confluency within 2 months. When
the primary cultures were subcultured at a split ratio
of 1: 4, they usually reached confluency within 7 days.
At low densities, subcultured cells were bipolar but
broader than the outgrowing cells from the primary
cultures. A characteristic feature of the cells was the
presence of longitudinal ridges in the cytoplasm. At
confluency, the cells showed the same shape and
spatial organization as seen in confluent primary
cultures. If cells were maintained in the same flask
after reaching confluency for up to 4 weeks, the
longitudinal arrays of parallel cells displayed a curved
pattern terminating in whorls, which consisted of
hillocks of multilayered cells and extracellular matrix
[Fig. 3(C)]. This growth pattern is similar to the ‘hill
and valley’ growth pattern of smooth muscle cells in
culture. Cell cultures showing this growth pattern
were named type A cells (smooth muscle-like). All
different type A cultures could be subcultured up to
the fifth passage, without showing any changes in
their growth pattern.
In about one-third of the cases, however, another
cell type (type B) grew out from the explants. These
cells were of variable sizeand shape, usually polygonal
and broader than those cells previously described. As
soon as this cell type grew to higher densities, the cells
formed irregular multiple layers, similar to those seen
in cultures of skin fibroblasts [Fig. 3(E)]. If both cell
types, namely type A (smooth muscle-like) and type B
(fibroblast-like) grew out from the same explant, the
type B cells grew out 7-10 days later than the type A
cells. If they grew out from different parts of the
explants, it was possible to remove the type B cells
from the culture as described in Materials and
Methods. Without removal, the type B cells overgrew
the type A cells quite easily within l-2 weeks.
In suspensions of ciliary muscle cells (method 2)
approximately 10 % of the cells settled within 48 hr on
the culture substrate. In about 50% of the casesthese
cells started to divide 1 week after initiation of the
culture [Fig. 3(D)], preferably in areas where the cells
had settled as cell clumps. Initially, small cell clones of
bipolar ribbon or spindle-shaped cells were formed.
After several days, these clones also arranged themselves in longitudinal bands of parallel cells and were
not distinguishable from the type A cells grown
directly from the explants. Clones consisting of the
type B cell type could not be observed.
With both methods, ten cell cultures showing the
type A (smooth muscle-like) growth characteristics
could be established from donors with an age range of
16-9 1 yr. The growth pattern was the samein all type
A cultures, regardless of the age of the donors.
cultures, grown for 4 days, did not stain for desmin. In
confluent cultures, however, a small amount of cells
stained positive for desmin [Fig. 4(E)]. This was most
pronounced in highly confluent cultures, which were
grown for 3-4 weeks. In these cultures, the desmin
was mainly expressedin areas where the cells grew in
multilayers and formed hillocks [Fig. 4(F)]. Becauseof
their arrangement in multilayers, quantitation of
desmin-positive cells in these cultures was difficult.
When the cultures were passagedand the cells allowed
to settle on the substratum for 6 hr, approximately
10% of the cells showed perinuclear aggregates of
desmin-positive filaments.
(b) Type B (fibroblust-Zike cells). In both of the two
cultures, intracellular fibers stained positive for aactinin. Staining for smooth-muscle specific a-smactin, however, was only observed in 5-10x of the
cells. In contrast to the type A cultures, the amount of
fibers was markedly reduced in most of the cells, when
the fibroblast cultures grew to high density. While all
cells in the fibroblast cultures stained positive for
vimentin, no staining was observedfor desmin, neither
in pre-confluent nor in confluent cultures.
Cell Cultures
(a) Type A (smooth muscle-like cells). Cells from pre-
confluent cultures, which were grown for 4 days,
invariably stained positive for a-sm-actin. The staining
revealed typical straight, non-interrupted, cable-like
fibers running parallel to each other along the long
axis of the cells. When stained with antibodies against
n-actinin, the fibers were decorated in an interrupted
pattern with closely spaced beads of ol-actinin. The
beadson neighbouring stressfibers were often aligned,
giving the cell a striated appearance [Fig. 4(A)]. In
regions where the stress fibers terminated near the
cell’s margin, staining for a-actinin was seen in
uninterrupted lines or bands. These lines extended
from the cell margin for some distance towards the
interior, running roughly parallel to the straight or
gently curved edges of the cells.
At confluency. the whole cytoplasm of the cells was
filled with densely arranged parallel fibers, staining
brightly both for ol-sm-actin and cc-actinin [Figs 4(B)
and (C)l. With antibodies against intermediate filaments, all cells stained positive for vimentin [Fig.
4(D)]. The vimentin filaments formed typical baskets
around the nucleus and extended in gently curved
arrays towards the cells’ periphery. Pre-confluent
(a) Type A culture (smooth muscle-like cells). In preconfluent cultures, the perinuclear cytoplasm of the
cells was filled with extensive aggregates of rough
endoplasmic reticulum and a prominent Golgi apparatus. The rough endoplasmic reticulum was often
dilated and filled with amorphous material. Many of
the cells had lysosomeswith myelin figures or electrondensematerial. The cells contained numerous parallel
fibers, consisting of bundles of thin (actin) filaments
with associated dense bodies [Fig. 5(A)]. The bundles
passedthrough the cytoplasm near the cell membrane
and were mainly oriented along the longitudinal axis
of the cells. Interspersed between the bundles were
numerous elongated mitochondria, free ribosomes,
coated vesicles, accumulations of glycogen and lipid
droplets. Near the periphery, long parallel rows of
membrane bound caveolae were seen. These caveolae
were aligned along the bundles of thin filaments. The
thin filaments ended in membrane-bound densebands,
which often extended for some distance into the cells.
Occasionally, the cells were connected by lateral
filopodia. which formed adherens-type or intermediate
junctions and gap junctions at their ends. After 4 days
FIG. 3. Cell cultures derived from ciliary muscle. A, Outgrowing type A cells from a ciliary muscle explant (asterisk). The cells
are characteristically
bipolar, with long tapering fusiform ends (phase-contrast micrograph, x 125). B, After confluency.
smooth muscle-like (type A) cultures become organized into a monolayer consisting of bands of parallel cells (phase-contrast
micrograph. x 125). C. Confluent type A cultures. The cells grow in longitudinal arrays of parallel cells, which bend and
terminate in whorls, which consist of hillocks of multilayered cells (light micrograph, PLP-fixed cells, stained with methyleneblue. x 70). D. After digestion of ciliary muscle explants, small cell clones of ribbon or spindle-shaped cells are formed. preferably
in areas where the cells have settled in clumps (asterisk) (phase-contrast micrograph, x 260). E. Fibroblast-like cells (type B)
derived from the ciliary muscle grow in irregular multilayers (phase-contrast micrograph. x 125).
FIG. 4. Immunofluorescence
of smooth muscle-like (type A) cultures. A, Stained for a-actinin, intracellular fibers are
decorated with closely spaced beads of a-actinin. The beads of neighbouring fibers are aligned, giving the cell a striated
appearance (arrows). In addition the focal contacts of the muscle cells are stained (arrowheads) ( x 1100). B and C. At
confluency. the whole cytoplasm of the cells is filled with densely arranged parallel fibers, staining brightly both for x-sm-actin
(B. x 320) and a-actinin (C. x 900). D. All cultured muscle cells stain for vimentin. The vimentin filaments display typical
in culture, a small and incomplete basal lamina
surrounded the cells. In confluent cultures, the
elongated cells were connected by intermediate junctions while gap junctions were rare. The rough
endoplasmic reticulum consisted of a few juxtanuclear
cisternae associated with the nuclear envelope and a
small Golgi apparatus. The peripheral cytoplasm was
filled with thick bundles of actin filaments densely
aggregated with no other cell organelles between
them. The bundles were not confined to the superficial
region of the cell as in non-confluent cells, but were
present throughout the whole cytoplasm.
(b) Type B cds ~broblast-like ce2ls). The ultrastructure of non-confluent type B cells was remarkably
similar to that of non-confluent type A cells. Bundles
of thin (actin) filaments, however, were much rarer. In
confluent cultures, type B cells showed, in contrast to
confluent type A cells, many aggregates of dilated
rough endoplasmic reticulum and a pronounced Golgi
apparatus, while bundles of thin filaments were still
only rarely observed.
4. Discussion
In the present study we report on methods to
establish homogenous primary cell cultures derived
from human ciliary muscle. Donors of different ages
were used. After a careful and histologically controlled
dissection procedure, two different cell types, a
smooth-muscle like (type A) and a fibroblast-like (type
B) can be grown from explants or isolated by enzymatic
digestion, These cells retain typical structural characteristics for at least five passages in culture. We
conclude that the smooth muscle-like cultures consist
of ciliary muscle cells in culture. The fibroblast-like
cells are probably derived from fibroblasts in the ciliary
Ciliary muscle cells from all established cell cultures
show a distinct growth pattern seen in phase contrast
optics. In confluent cultures, the muscle cells grow in
a ’ hill and valley ’ pattern, which is regarded as typical
for vascular and visceral smooth muscle cells in
culture (Gimbrone and Cotran, 1975 ; ChamleyCampbell, Campbell and Ross, 1979).
The characteristic growth pattern of ciliary muscle
cultures facilitates the discovery of contamination by
fibroblasts derived from the ciliary body. The fibroblasts show the same disorganized multilayered
growth pattern seen with scleral fibroblasts or keratocytes (Polansky et al., 1979 ; Hernandez, Igoe and
Neufeld, 1988).
The purity and homogeneity of the cell cultures was
further confirmed by immunocytochemistry. The
presence of actin isoforms has been proposed as a
useful marker for distinguishing between smooth
muscle cells and flbroblasts both in vivo and in vitro
(Gown et al., 1985; Skalli et al., 1986: Skalli,
Vandekerckhove and Gabbiani, 1987). We used an
antibody against the a-actin isoform, which is specific
for smooth muscle (Vandekerckhove and Weber,
1979). This antibody has been shown to stain
specifically smooth muscle cells of different origins as
well as myofibroblasts in tissue sections (Skalli et al.
1986, 1987; Czernobilsky et al., 1989. Darby, Skalli
and Gabbiani, 1990; Fliigel, Tamm and LiitjenDrecoll, 1991). Also under culture conditions, smooth
muscle cells still express their n-isoactin (Skalli et al..
1986; Owens et al., 1986), while fibroblasts in culture
usually continue to express the non-muscle actin
isoforms only (Garrels and Gibson 1976 ; Gown et al.,
1986; Skalli et al., 1986). Our study shows that both
the ciliary muscle cells in tissue sections as well as the
cell lines derived from them stain uniformly for a-smactin, while the fibroblasts of the ciliary body in situ do
not. Additionally, in cultures which are grown from
ciliary body fibroblasts, less than 10% of the cells
express ol-actin-positive filaments, possibly because of
a contamination with muscle cells.
The presenceof endothelial cells in our cultures can
be excluded by the lack of a typical endothelial cell
growth pattern (Gimbrone, Cotran and Folkman.
1974) as well as by the lack of staining for von
Willebrand-factor (factor VIII associatedantigen : Jaffe
et al., 1973). In tissue sections, positive staining for Zsm-actin is not confined to the ciliary muscle cells.
Staining is also seen in the smooth muscle cells of the
media of the major arterial circle of the iris as well as,
similar to retinal pericytes (Herman and D’Amore.
1985), in the pericytes of some of the ciliary muscle
capillaries. The major arterial circle, however, was
avoided in the dissection procedure. Compared with
our cell lines, retinal pericytes in culture show a very
different growth pattern (Buzney et al., 1983). The
pericytes of the ciliary muscle capillaries are just a few
in number when compared to the vast majority of the
ciliary muscle cells. Therefore, it is not very probable
that ol-actin-positive cells in our cultures are derived
from other sources than ciliary muscle cells.
Cells can further be characterized by their content of
intermediate filaments, which are cell-type-specific
and differentiation-dependent (for review see Osborn
and Weber, 1983). The present study shows that
human ciliary muscle not only stains for the musclespecific intermediate filament desmin, but also for
vimentin. The same is the case for skeletal myoblasts
in culture (Gard and Lazarides, 1980) and in tissue
sections (Van Muijen, Ruiter and Warnaar, 1987). as
basketsaround the nucleus and extend in gently curved arrays towardsthe cells’ periphery ( x 900). E and F, In confluent
cultures,about loo/, of the cellsstain for desmin(E, x 800). Desmin-positivecellsare predominatlylocatedin areas,wherethe
ceils grow in m&layers (F. x 240).
well as for vascular smooth muscle in certain arterial
vessels (Travo, Weber and Osborn, 1982; Schmid
et al., 1982 : Kocher et al., 1984; Fujimoto, Tokuyasu
and Singer, 1987). In contrast to the ciliary muscle,
the muscle cells of the media of the major arterial
circle of the iris belong to that kind of vascular smooth
muscle cells, which express vimentin only (Frank and
Warrne, 198 1; Gabbiani et al., 198 1: Osborn, Caselitz
and Weber. 1981). The fact that in vitro desminpositive cells were stainable in confluent ciliary muscle
cultures only, might be due to a dedifferentiation of the
cells in pre-confluent cultures. It is also known for
other cultured smooth muscle cells (Skalli et al., 1986 :
Palmberg and Thyberg, 1986; Ricciardelli et al.,
1989) that loss of desmin-positive cells occurs in vitro.
Interestingly, in highly confluent ciliary muscle cultures staining for desmin is mostly found in the
multilayered hills, where the cells have produced a lot
of extracellular
matrix. It might be that some
components of the extracellular matrix promote the
desmin expression of the muscle cells. It has been
reported that skeletal myoblasts. grown on laminin
substrata differentiate and express more desmin (von
der Mark and ijcalan, 1989).
Smooth muscle cells in sub-confluent growing
cultures also show other signs of dedifferentiation, e.g.
a relative loss of myofilaments together with an
increase in rough endoplasmic reticulum and the
presence of a large Golgi apparatus (for review see
Charnley-Campbell et al., 1979). The same ultrastructural changes are observed in ciliary muscle
cultures. But even pre-confluent cultures stain for asm-actin and all cells in confluent cultures recover the
typical characteristics of cultured smooth muscle cells
(for review see Charnley-Campbell et al., 1979). The
grade of differentiation
was much
higher in these cultures derived from adult ciliary
muscle than in a cell line derived from a l-day-old
infant, which was established recently (Korbmacher
et al., 1990). This difference might be due to the fact
that ciliary muscle of newborn primates is not fully
differentiated even in situ and shows a more mesenchymal appearance (Liitjen-Drecoll,
Tamm and
Kaufman., 1988).
Thus, cultured ciliary muscle cells do not fully
preserve their typical in vivo ultrastructure, which has
been reported to be very characteristic and different
from other smooth muscle cells (Ishikawa, 1962;
Liitjen-Drecoll et al., 1988 ; Fliigel, B&-tiny and LiitjenDrecoll, 1990). Several structural characteristics,
however. seem to be conserved in cultures. Staining
for z-actinin shows that the dense bands of the
myofilaments in cultured cells are aligned in register,
a feature, which has also been reported to be
characteristic for ciliary muscle cells in tissue sections
(Van der Zypen. 1967; Liitjen-Drecoll et al., 1988).
We would like to thank Jutta Gehr and LJte Maurer for
their expert assistance with tissue cultures and immunocytochemistry. We would also like to thank Simone Klein for
her excellent help in electronmicroscopy and Marco G613wein for his excellent preparation of the photographs. The
study was supported by grants from the Deutsche Forschungsgemeinschaft (Dre 124/6- 1 and Ro 8 1/ 18-4 ).
Burke, J. M and Ross, R. (1979). Syntbesis of connective
tissue macromolecules by smooth muscle. Int. Rev.
Connect. Tissue Res. 8, 119-57.
Buzney. S. M.. Massicotte, S. J., Hetu, N. and Zetter, B. R.
(198 3). Retinal vascular endothelial cells and pericytes.
Differential growth characteristics in vitro. Invest.
Ophthnlmol. Vis. Sci. 24. 470-80.
Charnley-Campbell, J.. Campbell, G. R. an.d Ross. R. (1979).
The smooth muscle in cell culture. Physiol. Rev. 59,
Czernobilsky. B.. Shezen, E., Lifschitz-Mercer, B.. Fogel, M.,
Luzon, A., Jacob, N., Skalli, 0. and Gabbiani, G. (1989).
Alpha smooth muscle actin (a-SM actin) in normal
human ovaries, in ovarian stromal hyperplasia and in
ovarian neoplasms. Virchows Arckiv B Cell Pathol. 57.
Darby, I., Skalli, 0. and Gabbiani, G. (1990). a-Smooth
muscle actin is transiently expressed by myofibroblasts
during experimental wound healing. Lab. Invest. 63,
Debus, E.. Weber, K. and Osborn, M. (1983). Monoclonal
antibodies to desmin, the muscle-specific intermediate
filament protein. EMBO 1. 2, 2305-12.
Duhamel, R. C. and Johnson, D. A. (1985). Use of nonfat dry
milk to block nonspecific nuclear and membrane
staining by avidin conjugates. 1. Histochem. Cytochem.
33. 711-4.
Fliigel. C., Bira’ny. E. H. and Liitjen-Drecoll, E. (1990).
Histochemical differences within the ciliary muscle and
its function in accommodation. E?cp. Eye Res. 50.
Fliigel, C., Tamm, E. and Liitjen-Drecoll, E. ( 199 1). Different
cell population in trabecular meshwork: An ultrastructural and immunohistochemical
study. Exp. Eye
Res. 52, 681-90.
Frank, E. D. and Warren. L. (1981). Aortic smooth muscle
contain vimentin instead of desmin. hoc. Natl. Acad.
Sci. U.S.A. 78, 3020-4.
Fuchs, E. ( 1928). ijber den Ziliarmuskel. Albrecht von Graefes
Arch. Ophthalmol. 120, 733-41.
Fujimoto. T., Tokuyasu, K. T. and Singer, S. J. ( 198 7). Direct
morphological demonstration of the coexistence of
vimentin and desmin in the same intermediate filaments
of vascular smooth muscle cells. 1. Suhmicroscop. Cytol.
19. l-9.
FIG. 5. Electromnicrographs of cultured smooth muscle-like (type A) cells. A, In pre-confluent cultures. the cytoplasm of the
cells is filled with numerous parallel bundles of thin (actin) filaments (arrows) with associated, regular arranged dense bodies
(arrowheads). Between the filaments numerous elongated mitochondria are seen (white arrows) ( x 6600). B. At confluency.
the filaments (asterisks) are densly aggregated. N, Nucleus. ( x 3000).
Gabbiani, G., Schmid. E., Winter, S., Chaponnier. C.,
dechastonay, C., Vandekerckhove, J.. Weber, K. and
Franke, W. W. (1981). Vascular smooth muscles differ
from other smooth muscle cells: Predominance of
vimentin filament and a specific a-type actin. Proc. Nat].
Acad. Sci, U.S.A. 78, 298.
Gard, D. L. and Lazarides. E. (1980). The synthesis and
distribution of desmin and vimentin during myogenesis
in vitro. CeZI19. 263-75.
Garrels, J. I. and Gibson. W. (1976). Identification and
of multiple forms of actin. Cell 9.
Gimbrone, M. A., Cotran, R. S. and Folkman. J. ( 1974).
Human vascular endothelial cells in culture. Growth
and DNA synthesis. 1. Cell Biol. 60, 673-84.
Gimbrone. M. A. and Cotran, R., Z. (1975 ). Human vascular
smooth muscle in culture. Growth and ultrastructure.
Lab. Invest. 33, 16-27.
Gown, A. M.. Vogel, A. M.. Gordon, D. and Lu. P. L. ( 198 5 ).
A smooth muscle-specific antibody recognizes smooth
muscle actin isoenzymes. 1. Cell Biol. 100, 807-l 3.
Herman, I. M. and D’Amore, P. A. (1985). Microvascular
pericytes contain muscle and nonmuscle actins. 1. Cell
Biol. 101. 43-52.
Hernandez. M. R.. Igoe, F. and Neufeld, A. H. ( 1988 ). Cell
culture of the human lamina cribrosa. Invest. Ophthalmol. Vis. Sci. 29. 78-89.
Ishikawa. T. (1962). Fine structure of the human ciliary
muscle. Invest. Ophthalmol. 1, 587-608.
Ito, S. and Karnovsky, M. J. (1968). Formalaldehydeglutaraldehyde fixatives containing trinitro compounds.
1. Cell. Biof. 39, 168A-9A.
Jaffe. E. A., Nachman, R. I,., Becker. C. G. and Minick. C. R.
( 19 7 3 ). Culture of human endothelial cells derived from
umbilical veins. Identification by morphologic and
immunologic criteria. 1. Clin. Invest. 52, 2745-56.
Johnson, G. D.. Davidson, R. S., McNamee, K. C.. Russel. G..
Goodwin, D. and Holborow, E. J. (1982). Fading of
immunofluorescence during microscopy : a study of the
phenomenon and its remedy. 1. Immunol. Meth. 55,
Kerschbaumer. K. (1888). Ueber Altersvergnderungen
Uvea. Albrecht von Graefes Arch. Ophthalmol. 34, 16-34.
Kocher. O., Skalli. O., Bloom, W. S. and Gabbiani, G. (1984).
Cytoskeleton of rat aortic smooth muscle cells. Normal
conditions and experimental thickening. Lab. Invest. 50.
Korbmacher. C., Helbig. H., Coroneo. M., Erickson-Lamy
K. A.. Stiemer, B., Tamm. E.. Liitjen-Drecoll. E. and
Wiederholt. M. (1990). Membrane voitage recordings
in a cell line derived from human ciliary muscle. Irnrest.
Ophthalmol. Vis. Sci., 31, 2420-24 30.
Liitjen-Drecoll. E., Shimizu. T.. Rohrbach. M. and Rohen, J.
W. (1986a). Quantitative analysis of “plaque-material”
in the inner- and outer wall of Schlemm’s canal in
normal and glaucomatous eyes. E.rp. Eye Res. 42.
Liitjen-Drecoll. E.. Shimizu. T.. Kohrbach. M. and Rohen.
J. W. ( 1986b). Quantitative
analysis of “plaquematerial” between ciliary muscle tips in normal and
glaucomatous eyes. Exp. Eye Res. 42, 457-65.
Liitjen-Drecoll. E. and Tamm. E. (1988). Morphological
study of the anterior segment of cynomolgus monkey
eyes following treatment with prostaglandin F,,. Exp.
Eye Res. 47, 761-9.
Liitjen-Drecoll, E., Tamm, E. and Kaufman, P. L. ( 1988). Age
changes in Rhesus monkey ciliary muscle: Light and
electronmicroscopy. Exp. Eye Res. 47, 885-99.
Liitjen-Drecoll. E. and Tamm. E. (1989). The effects of ocular
hypotensive doses of PGF,,-isopropylesther on anterior
segment morphology. In The Ocular EfSrcts of Prost-
aglandins and Other Eicosanoids. Progress in Clinical and
Biological Research, Vol. 312. (Eds Bito. 1.. Z. and
Stjernschantz, J.) Pp. 437-46. Alan R. Liss. Inc: New
McLean, I. W. and Nakane, P. K. (19 74). Periodate-lysineparaformaldehyde fixative : a new fixative for immunoelectron microscopy. I. Histochem. Cytochem. 22, 107783.
Naiem, M., Gerdes, J., Abdulaziz, Z.. Sunderland, C. A..
Allington. M. 1.. Stein, H. and Mason. D. Y. ( 1982). The
value of immunohistological
screening in the production of monoclonal antibodies. 1. Immunol. Meth. 50.
Nilsson, S. F. E.. Samuelsson, M., Bill, A. and Stjernschantz.
J. (1989). Increased uveoscleral outflow as a possible
mechanism of ocular hypotension caused by prostaglandin F,,-1-isopropylester in the cynomolgus monkey.
Erp. Eye Res. 48, 707-16.
Osborn. M., Caselitz, J. and Weber. K. (198 1). Heterogenity
of intermediate filament expression in vascular smooth
muscle: A gradient in desmin positive cells from the rat
aortic arch to the level of the arteria illiaca communis.
Diflerentiafion 20. 196-202.
Osborn, M.. Debus, E. and Weber. K. (1984). Monoclonal
antibodies specific for vimentin. Eur. 1. Ceil Biol. 34.
Osborn, M. and Weber. K. (I 983). Biology of disease. Tumor
diagnosis by intermediate filament typing : A novel tool
for surgical pathology. Lab. Invest. 48, 372-94.
Owens, G. K.. Loeb, A., Gordon, D. and Thompson, M. M.
(1986). Expression of smooth muscle n-isoactin in
cultured vascular smooth muscle cells: relationship
between growth and cytodifferentiation. I. Cell Biol.
102. 343-52.
Palmberg. L. and Thyberg. J. ( 19 86). IJterine smooth muscle
cells in primary culture. Alterations in fine structure.
cytoskeletal organization and growth characteristics.
Cell Tissue Res. 246. 253-62.
Polansky. J. R., Weinreb, R. N.. Baxter, J. I). and Alvarado. J,
(1979). Human trabecular cells. I. Establishment in
tissue culture and growth characteristics. lnvrsl. Opizthalmol. Vis. Sci. 18. 1043-Y.
Ricciardelli. C.. Horsfall, D. 1.. Skinner. j. M.. Henderson,
D. W.. Marshall, V. R. and Tilley. W. D. ( 1989). Development and characterization of primary cultures of
smooth muscle cells from the fibromuscular stroma of
the guinea pig prostate. In Vitro Cell. Tkvel. Biol. 25.
Rohen, J. W. and Wittmer, K. (1972). Electron microscopic
studies on the trabecular meshwork in glaucoma
simplex. Albrecht von Graefes Arch. Klin. E.xp. Ophthalmol.
183. 251-66.
Schmid, E., &born. M.. Rungger-Brgndle. E.. Gabbiani. G..
Weber. K. and Franke, W. W. (1982). Distribution of
vimentin and desmin filaments in smooth muscle tissue
of mammalian and avian aorta. Exp. Cell Res. 137.
Skalli. 0.. Bloom, W. S.. Ropraz. P.. Azzarone, B. and
Gabbiani, G. (1986). Cytoskeletai remodeling of rat
aortic smooth muscle cells in vitro: Relationship to
culture conditions and analogies to in vitro situations. 1,
Submicrosc. Cytol. 18. 48 l-9 3.
Skalli. 0.. Ropraz. P.. Trzeciak. A., Benzonana. G.. Gillessen.
D. and Gabbiani. G. (1986). A monoclonal antibody
against a-smooth muscle actin : a new probe for smooth
muscle differentiation. 1. Cell Biol. 103. 7787-96.
Skalli. 0.. Vandekerckhove. J. and Gabbiani, G. ( 1987).
Actin-isoform pattern as a marker of normal or
pathological smooth-muscle and fibroblastic tissues.
Diflerentiation 33. 2 32-8.
Stieve. R. (1949). tiber den Bau des menschlichen Ciliar-
musk&. seine physiologischen Vergnderungen wghrend des Lebens und seine Bedeutung fiir die Akkommodation. Z. mikro.-anatom. Forsch. 55, 3-88.
Tamm, E.. Liitjen-Drecoll. E.. Rittig, M. and Rohen, J. W.
(1989). Connective tissue changes in the uveoscleral
pathways of primate eyes after treatment with prostaglandin F,,. Invest. Ophthalmol. I/is. Sci. 30 (Suppl.).
Travo. P.. Weber, K. and Osborn, M. (1982). Co-existence of
vimentin and desmin type intermediate filaments in a
subpopulation of adult rat vascular smooth muscle cells
growing in primary culture. Exp. Cell Res. 139, 87-94.
Vandekerckhove. J. and Weber, K. (1979). The complete
amino acid sequence of actins from bovine aorta.
bovine heart, bovine fast skeletal muscle and rabbit
slow skeletal muscle. Diflerentiation 14. 123-33.
van der Zypen, E. (1967). Licht-und elektronenmikroskopische Untersuchungen
iiber den Bau und die
Innervation des Ziliarmuskels bei Mensch und Affe
(Cercopithecus aethiops). Albrecht v. Graefes Arch. Clin.
Exp. Opthalmol. 174, 143-68.
van der Zypen. E. (1970). Licht-und elektronenmikroskopische Untersuchungen
iiber die Altersvergnderungen am M. ciliaris im menschlichen Auge. Albrecht
von Graefes Arch. Klin. Exp. Ophthalmol. 179. 332-57.
von der Mark, K. and acalan. M. (1989). Antagonistic
effects of laminin and fibronectin on the expression of
the myogenic phenotype. Differentiation 40, 150-7.
Van Muijen. G. N. P.. Ruiter, D. J. and Warnaar, S. 0.
(1987). Coexpression of intermediate filament polypeptides in human fetal and adult tjssues. tab. Invest.
57. 359-69.