Label-Retaining Cells Are Preferentially Located
in Fornical Epithelium: Implications on
Conjunctival Epithelial Homeostasis
Zhi-Gang Wei,* George Cotsarelis* Tung-Tien Sun,^ and Robert M. Lavkerf
Purpose. To determine the cell kinetic properties of epithelial cells from various zones of the
conjunctiva.
Methods. The morphology and cell kinetics of bulbar, fornical, and palpebral conjunctival
epithelium were studied in neonatal and adult SENCAR mice. To examine the proliferative
rate of the conjunctival epithelium, a single administration of tritiated thymidine (3H-TdR)
was used to detect cells in " S " phase. Proliferative rates were also assessed by determining
mitotic activity after an intraperitoneal injection of colchicine to arrest cells in mitosis. To
detect slow-cycling cells, mice received 3H-TdR continuously for 1 week. After a 4-week chase,
animals were sacrificed and eyes were surgically removed. All tissues were immediately fixed
in formalin and processed for histology and autoradiography.
Results. Slow-cycling cells, detected as label-retaining cells (LRCs), were identified in bulbar,
fornical, and palpebral epithelia, as well as in limbal epithelium. The greatest number of
LRCs was found in fornical epithelium. In addition, we found a number of label-retaining
goblet cells. This cell population was shown to incorporate 3H-TdR after a single pulse administration, and mitotic figures were seen in goblet cells after colchicine treatment, indicating
that conjunctival goblet cells have proliferative capabilities. .
Conclusions. These findings are consistent with earlier in vitro data that the fornical epithelium
may be a zone enriched in conjunctival epithelial stem cells. This has important implications
in conjunctival epithelial development and is relevant in wound repair. Furthermore, the
concept that goblet cells are slow-cycling cells with proliferative capabilities provides new
insights into the area of conjunctival homeostasis. Invest Ophthalmol Vis Sci. 1995;36:236246.
•
Supported by National Institutes of Health grants HY6769 (RAIL) and EY4722
('ITS).
Submitted for publication March 10,'1994; reuised June 27, 1994; accepted July
21, 1994.
Proprietary interest category: N.
Reprint requests: Robert M. Ijivker, Department of Dermatology, University of
Pennsylvania School of Medicine, Clinical Research Building, Room 235A, 422
Curie Boulevard, Philadelphia, PA 19104.
thelia, undergoes a series of changes in which epithelial cell loss due to terminal differentiation is balanced
by new cell formation. Conjunctival epithelium can
also serve as a source of epithelial cells that migrate
rapidly and resurface on the corneal stroma after a
loss of corneal epithelium due to wounding.5"12 One
of the responses to corneal epithelial depletion is increased proliferation of conjunctival epithelial cells.13
For example, Thoft and coworkers observed that one
day after the removal of corneal epithelia and a portion of bulbar conjunctival epithelia, there was a 10fold increase in the tritiated thymidine (3H-TdR) labeling index of the remaining conjunctival epithelium.13 This kind of self-renewing epithelium, by definition, must contain stem cells.14 Such cells are
thought to play a major role in normal proliferative
homeostasis, as well as in a tissue's responsiveness to
population depletion, resulting from wounding and
236
Investigative Ophthalmology & Visual St
Copyright © Association for Research in
L^onjunctival epithelium can be divided into three
morphologically distinct zones: bulbar, fornical, and
palpebral.'" 4 This epithelium forms a physical protective barrier and, through goblet cell secretions, contributes to the formation a n d maintenance of a "tear
film" on the ocular surface. In performing these functions, conjunctival epithelium, like other stratified epi-
Fnm the *Depanmenl of Dermatology, University of Pennsylvania School of
Medicine, Philadelphia, Pennsylvania, and the ^Epithelial Biology Unit, The
Ronald 0. Perelman Department of Dermatology and Pharmacology, New York
University Medical Center, Kaplan Comprehensive Cancer Center, New Ymk, Neiu
York.
e, January 1995, Vol. 36, No.
on and Ophthalmology
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933181/ on 06/17/2017
Enrichment of Conjunctival Epithelial Stem Cells in Fornix
237
other extrinsic perturbations. However, despite the
obvious functional importance of conjunctival epithelium in ocular physiology, relatively little is known
about conjunctival epithelial stem cells.
Our understanding about the location and biologic properties of stem cells has come primarily from
investigations on the hematopoietic system, as well as
the palmar epithelium, follicular epithelium, corneallimbal epithelia, and small intestinal epithelium.15"26
From these studies, there has emerged an awareness
that stem cells in general, and those of epithelia in
particular, share certain features. These cells are relatively undifferentiated, are slow cycling but can be
preferentially stimulated to proliferate by wounding,15
and can give rise to transient amplifying cells that have
limited proliferative potential. Furthermore, stem
cells have great proliferative potential, are usually located in well-protected and well-vascularized areas,
and are in contact with a unique stromal environment
that provides a "niche" for the maintenance of these
cells. Finally, they may be a target for carcinogens and,
as such, they play a central role in tumor development"" 8 8
Vision Research and were approved by the University
of Pennsylvania Animal Care and Ethics Committee.
We have recently shown that under identical culture conditions, keratinocytes of rabbit fornix had a
much greater in vitro proliferative potential than bulbar and palpebral keratinocytes.2 Inasmuch as high
proliferative potential is an important characteristic
of stem cells, it raises the interesting possibility that
fornix might be the preferential site of conjunctival
epithelial stem cells. Because stem cells are known to
be slow cycling, one way to "tag" these cells involves
repeated administration of 3H-TdR for a prolonged
period of time, followed by a long chase period so
that only cells that cycle slowly will retain the isotope
(the label-retaining cells, or LRCs). Using this approach, we identified LRCs in mouse conjunctival epithelium, and we report here that the fornix contains
many more LRCs than other regions of the conjunctival epithelium. This is consistent with our earlier data,
indicating that rabbit fornix keratinocytes have a
much greater in vitro proliferative potential than bulbar and palpebral epithelial cells.2 Interestingly, we
found that some of the LRCs had a goblet cell phenotype. On further investigation, using mitotic arrest and
thymidine autoradiographic techniques, we observed
that goblet cells are capable of proliferating.20 These
results have important implications concerning the
homeostasis and regenerative properties of mammalian conjunctival epithelium.
To detect slow-cycling cells, neonatal SENCAR mice
were injected subcutaneously in the back with 5 y^Ci of
methyl-3H-TdR (specific activity, 82.7 Ci/mmol; New
England Nuclear, Boston, MA) per gram of body
weight twice daily (8:00 AM and 4:00 PM) for the first
7 days of life (5 to 15 /j,d per animal per day, depending on the weight of the mouse). After a 4-week
chase when mice received only food and water ad
libitum, animals were killed and eyes were removed
with the eyelids still attached. Specimens were fixed
with 10% formalin in phosphate-buffered saline and
embedded in JB4 plastic embedding medium, and sections were cut at 3 //m as previously described. Sections were processed for autoradiography (see Autoradiography), and those cells with silver grains retained
in their nuclei (LRCs) were detected.
MATERIALS AND METHODS
AJ1 experimental procedures conformed to the ARVO
Statement for the Use of Animals in Ophthalmic and
Conjunctival Development
To gain further insight into the relationship between
bulbar, fornical, and palpebral conjunctival epithelia
with respect to proliferation, we analyzed the morphology and cell kinetics of these three zones during
development. Late gestational age (15 to 17 days)
SENCAR mice obtained from Harlan Sprague provided newborns used in the studies on conjunctival
development. Three mice were killed at 1, 2, 4, 6, 8,
10, 12, and 14 days after birth by cervical dislocation,
and eyes and eyelids were surgically removed and processed for histology or autoradiography as described
below. To determine the proliferative rate of the developing conjunctival epithelium, one group of neonatal mice with eyelids fused (1, 2, 4, 6, 8, and 10 days
old) received a subcutaneous injection of 5 fiCi of SHTdR (specific activity, 82.7 Ci/mmol) per gram of
body weight 1 hour before sacrifice.
Detection of Slow-Cycling Cells
Determination of Proliferative Activity of Adult
Conjunctiva
To examine the proliferative rate of the adult conjunctival epithelium, 3H-TdR was used to detect cells in
" S " phase. One hour before sampling, 4-week-old female SENCAR mice (Harlan Sprague) received intraperitoneal injections of 40 /LtCi 3H-TdR (specific activity, 82.7 Ci/mmol; New England Nuclear, Boston,
MA) in 0.1 ml of sterile phosphate-buffered saline.
Eyes were removed and prepared for histology as described.
We also assessed the proliferative activity of the
conjunctival epithelium by determining mitotic activity. Because mitosis is a relatively rare event, we used
mitotic arrest techniques to enhance the number of
mitotic figures present in tissue sections. Mice were
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933181/ on 06/17/2017
238
Investigative Ophthalmology & Visual Science, January 1995, Vol. 36, No. 1
treated with intraperitoneal colchicine (10 mg in 0.1
ml phosphate-buffered saline per animal; Sigma, St.
Louis, MO) that arrests cells in mitosis, and eyes were
removed 8 hours later and prepared for histologic
examination as described. In an earlier experiment,
mice were sacrificed hourly for 12 hours after an intraperitoneal injection of colchicine, and a mitotic index
(number of mitotic figures per 1000 basal cells) was
determined for each time point. The maximum number of mitotic figures occurred 8 hours after colchicine injection.
Autoradiography
Tissue sections 2-//m to 3-//m thick were picked up on
glass slides and air dried; the slides were then dipped
in a nuclear track emulsion (Ilford, London, UK) and
diluted 1:1 with distilled water at 40°C. These coated
slides were air dried and exposed for 10 to 14 days at
4°C in light-tight boxes. The autoradiographs were
developed in Kodak D-19 developer (Eastman Kodak,
Rochester, NY) for 5 minutes at 20°C, rinsed in distilled water, and fixed for 10 minutes. After a 30-minute wash, the slides were lightly stained with hematoxylin and eosin.
A labeling index was determined for the different
conjunctival zones by counting at least 1000 nuclei
from three different sections for each determination,
and the results were expressed as the percentage of
labeled nuclei.
RESULTS
Morphology of Murine Conjunctival Epithelium
Neonatal. Although the organization and development of rabbit and rat conjunctival epithelia have
been described in detail, relatively little attention has
been directed to the murine conjunctival epithelium.
At birth, mouse conjunctival epithelium consists of
one to two layers of relatively flattened and undifferentiated cells. As development proceeds (2 to 4 days after
birth), the bulbar and palpebral basal keratinocytes
maintain a flattened appearance, whereas the fornical
basal keratinocytes become cuboidal (Fig. 1A). During
the next several days, the fornical keratinocytes assume a columnar shape typical of adult tissue. Palpebral keratinocytes become cuboidal, whereas die bulbar conjunctival cells remain relatively flattened. Overall, cell density is greatest in the fornical region. This
increase in epithelial cell density in the fornix is paralleled in the underlying stroma, which is also more
cellular in the fornix compared with either the bulbar
or palpebral regions. In addition, the fornical stroma
is the most heavily vascularized (Figs. IB, ID). Finally,
before eyelid opening, goblet cells were first seen in
the fornix approximately 10 days after birth (Fig. ID).
Goblet cell density remained highest in the fornix
throughout subsequent stages of development.
Adult. Based on its anatomic location, degree of
stratification, and density of goblet cells, adult conjunctival epithelium can be divided into three zones (Fig. 2A):
bulbar (Fig. 2D), fornix (Fig. 2E), and palpebral (Fig.
2F). The bulbar zone is covered by a stratified squamous
epithelium and is contiguous with die limbal zone (Fig.
2C) of the cornea (Fig. 2B). The bulbar epithelium
consists of a single basal cell layer, two to three cuboidal
wing cell layers, and one slightly flattened superficial cell
layer. Bulbar epithelium contains goblet cells and is,
therefore, easily distinguished from the adjacent limbal
epidielium. The fornical epithelium (Fig. 2E) contains
one to two cell layers and appears much less stratified
than eidier the bulbar or die palpebral region. Two
types of cells are present within the fornical zone: small,
round epidielial cells with a relatively large nucleancytoplasmic ratio, and numerous goblet cells. Goblet cells
are columnar and have small, foot-like structures at the
base in contact with die basement membrane. The nucleus is most often positioned beneath a prominent mass
of mucous droplets, which bulges die cell into its characteristic "wine goblet" appearance. The apical portion
of the goblet cell is often exposed to the surface. Goblet
cells usually appear in clusters, each of which can contain as many as five goblet cells. The palpebral epithelium (Fig. 2F) is die most highly stratified of die diree
zones, consists of four to seven cell layers, and is contiguous with die epidermis of die eyelid. In contrast to the
fomical epidielium, palpebral goblet cells appear singly
and increase in number toward die fornix.
Proliferative Rates of the Conjunctival
Epithelium
Neonatal. To study die proliferative status of die developing conjunctival epidielium, we used tritiated tfiymidine-labeling to detect cells in die "S" phase of die
cell cycle. When 3H-TdR was administered as a single
pulse immediately after birdi, labeled nuclei were found
to distribute randomly in die basal cell layer of die neonatal conjunctival epithelium. During die ensuing developmental period (2 to 10 days after birdi), labeled nuclei
were preferentially observed in die fornical zone and in
die eyelid region (Figs. 1A, 1C).
Adult. Labeled nuclei were observed in the basal cell
layer of the diree conjunctival zones, as well as in limbal
and corneal epidielia (Figs. 2B to 2F). As in previous
studies, die labeling index for limbal epidielium
(2.58%) was equivalent to, if not slighdy lower dian, diat
of die corneal epidielium (2.77%; Table 1). Bulbar and
palpebral epidielia had similar labeling indices (1.02%
and 0.90%, respectively), diough diey were somewhat
lower dian limbal and corneal epidielia. Interestingly,
fornical epidielium had a lower proliferative rate
(0.23%) dian the odier conjunctival zones.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933181/ on 06/17/2017
Enrichment of Conjunctival Epithelial Stem Cells in Fornix
1.
239
\
nGURE 1. Morphology and cell kinetics of developing SENCAR mouse ocular epithelium.
(A) Autoradiogram showing the distribution of 3H-thymidine-labeled nuclei in a transverse
tissue section of ocular epithelium from a 4-day-old SENCAR mouse, showing the corneal
(C), limbal (L), bulbar (B), fornical (F), and palpebral (P) zones. (B) High magnification
histologic section from a 4-day-old SENCAR mouse showing a portion of the fornical (F)
and limbal (L) epithelia. Fornical epithelium consists of a single layer of round to columnar
cells resting on a highly vascularized stroma (arrows). Limbal epithelium is more stratified,
consisting of one- to two-cell flattened layers. Similar to the fornical epithelium, limbal
epithelium rests on a highly vascularized stroma (arrows). (C) High magnification autoradiogram from a 4-day-old SENCAR mouse showing a greater distribution of 3H-thymidinelabeled nuclei in the fornical epithelium (F, arrows) compared to bulbar epithelium (B,
single arrow). Fornical epithelial consists of a single layer of round to oval keratinocytes,
whereas keratinocytes of bulbar epithelium are primarily flattened. (D) High magnification
histologic section of a portion of the fornical epithelium from a 10-day-old SENCAR mouse.
Single and paired goblet cells (G, arrow) are first seen in fornical epithelium at this time.
Note prominent vascular profiles in fornical stroma (targe arrows).
Label-Retaining Cells Are Preferentially
Located in Fornix
Pulse-labeling techniques detect those cells that are regularly cycling and yield information on the proliferative
rate of a tissue. However, this methodology provides little
insight into the slow-cycling cells. To identify the slowcycling cells, we provided a continuous supply of 3HTdR and then detected the LRCs.15'25 As in previous
studies, no LRCs were observed in corneal epithelium
(Fig. 3A, Table 1), whereas a subpopulation of LRCs was
routinely detected in the basal layer of limbal epithelium
(Fig. 3B; Table 1). In conjunctival epithelium, LRCs
were observed in the basal layers and the bulbar (Fig.
3C), fornical (Fig. 3D), and palpebral epithelia (Fig.
3E). However, LRCs were not evenly distributed within
the conjunctival epidielium (Table 1). The greatest density of LRCs was found in the fomical zone (14% of
epithelial cells). Within bulbar epithelium, 5% of epithelial cells were detected as LRCs, and slighdy fewer than
1 % of palpebral keratinocytes were LRCs.
Goblet Cells Have Proliferative Capabilities
In previous investigations, we detected relatively undifferentiated LRCs in the basal layer of murine epidermis, hair follicle, and limbal epithelium. Consistent results were obtained in the present study with
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933181/ on 06/17/2017
840
Investigative Ophthalmology & Visual Science, J a n u a r y 1995, Vol. 36, N o . 1
'••M-
6.
FIGURE 2. Morphology a n d cell kinetics of a d u l t SENCAR ocular epithelium. (A) Histologic
section of ocular epithelium from adult SENCAR m o u s e showing the relationship of the
corneal (C) a n d limbal (L) epithelium with bulbar (B), fornical ( F ) , a n d palpebral (P)
conjunctival epithelia. Scale bar = 50 fim. ( B - F ) Autoradiograms showing distribution of
3
H-thymidine-labeled nuclei in corneal (B, arrowheads), limbal (C, arrowheads), bulbar (D,
arrowheads), a n d palpebral (F, arrowheads) epithelia after a pulse administration of nucleotide. 3 H-thymidine-labeled nuclei were rarely observed in fornical (E) epithelium. G = goblet
cells. Scale b a r = 10 /xm.
respect to the morphology of LRCs from limbal, palpebral, and bulbar epithelia (Fig. 3D). Unexpectedly, however, we found a number of label-retaining
goblet cells within the fornical epithelium. To see
whether these label-retaining goblet cells were capable of replicating, we arrested mitosis during an 8hour period using colchicine. Indeed, some mitotically arrested goblet cells were observed, proving
that these cells were capable of proliferating (Figs.
4A to 4D). In addition, occasional goblet cell nuclei
were found to be labeled after a single pulse administration of 3H-TdR (Figs. 4E to 4F).
DISCUSSION
Fornix Is the Stem Cell-Rich Zone of the
Conjunctival Epithelium
In an in vitro investigation of the growth characteristics of rabbit bulbar, fornical, and palpebral keratino-
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933181/ on 06/17/2017
Enrichment of Conjunctival Epithelial Stem Cells in Fornix
TABLE l. Comparison of the Distribution of
Pulse-Labeled Keratinocytes (LI) With
Label-Retaining Keratinocytes (LRC) in
Various Ocular Zones
U%
(*= 6)
Corneal
Limbal
Bulbar
Fornical
Palpebral
2.77
2.58
1.02
0.23
0.90
± 1.2
± 1.2
± 0.32
± 0.2
± 0.62
LRC %
(n = 10)
0.10
5.38
4.96
13.63
0.90
±
±
±
±
±
0.01
1.76
1.76
3.0
0.45
All values are means ± SD.
LI = Pulse-labeled keratinocytes; LRC = label-retaining
keratinocytes.
cytes, we found earlier that fornical keratinocytes
could be subcultured three times, whereas under identical experimental conditions bulbar and palpebral
cells could be subcultured only once, or not at all.2
This provides strong evidence that fornical epithelial
cells have a much greater proliferative potential than
241
bulbar and palpebral cells. Because one of the defining features of stem cells is a great proliferative potential, this suggests that the fornix may contain more
stem cells than the other regions of the conjunctiva.
Here we demonstrate that the mouse fornical epidielium indeed contains many more slow-cycling cells,
detected as LRCs, than bulbar and palpebral epithelia.
Because slow-cycling is one of the most distinguishing
features of the stem cells, our kinetic data are entirely
consistent with the in vitro growth data and strongly
support the idea that the fornical region is the stem
cell-rich zone of the conjunctival epithelium.
Because stem cells are slow cycling, they rarely
incorporate tritiated thymidine after single pulse administration. This explains why, as we have shown earlier, stem cell-rich regions (e.g., limbal epithelium,
bulge region of the hair follicle, tips of deep rete
ridges of palmar epithelium) usually do not incorporate large amounts of pulse-administered tritiated thymidine.15'17'2125 In the adult mouse, the fornical epithelium had a lower labeling index than either palpebral or bulbar epithelia. This labeling pattern is
JfeM*
LRF
4%
FIGURE 3. Detection of label-retaining cells in SENCAR mouse
ocular epithelium. Label-retaining cells (LRC) are rarely, if
ever, detected in corneal epithelium (A) but are routinely
present in limbal epithelium (B). Occasional LRCs are detected in bulbar (C) and palpebral (E) epithelia, whereas
numerous LRCs are observed in fornical epithelium both in
basal cells (arrowheads) and goblet cells (D, LRG). Label-retaining fibroblasts (LRF) are present in the stroma of all
zones. G = goblet cell. Scale bar = 10 fim.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933181/ on 06/17/2017
Investigative Ophthalmology & Visual Science, January 1995, Vol. 36, No. 1
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933181/ on 06/17/2017
243
Enrichment of Conjunctival Epithelial Stem Cells in Fornix
FIGURE 4. Conjunctival goblet cells have proliferative capabilities. Using mitotic arrest by
intraperitoneal injection of colchicine (panels A-D), mitotic figures are frequently observed
in goblet cells (MG). Occasional 3H-thymidine-labeled nuclei can be detected in goblet cells
(arrowheads) after a single pulse administration of radioactive nucleotide (panels E-F).
Sections in panels A-D are stained with hematoxylin and alcian blue and PAS. Autoradiograms in panels E-F are stained with hematoxylin and eosin. Scale bar = 10 /xm.
entirely consistent with previously described
for other epithelial stem cell systems.
findings
Fornical Epithelium Is the Major Proliferative
Zone During Conjunctival Growth and
Development
After pulse administration of tritiated thymidine, numerous labeled nuclei were routinely detected in the
basal cells of fornical epithelium during the first 10
days of life. If the fornix were enriched in conjunctival
epithelial stem cells, proliferation of these cells would
be expected during development, because theoretically this would result in the generation of transient
amplifying cells needed for continued conjunctival
growth and expansion. Proliferation of the normally
quiescent fornical epithelial cells during a specific
stage of conjunctival growth and development is analogous to what occurs in the bulge region of mouse
hair follicle (site of follicular epithelial stem cells) during the early growing phase of the hair.
Clustering of Conjunctival Stem Cells Is
Consistent With Stem Cell Distribution in
Other Stratified Squamous Epithelia
That conjunctival epithelial stem cells are not randomly distributed throughout the tissue, but show a
preferential fornical location, is consistent with the
distribution patterns of putative stem cells in other
stratified squamous epithelia. For example, corneal
epithelial stem cells are clustered in the limbal region1519'21; interfollicular epidermal stem cells have
been postulated to reside at the bottom of the deep
rete ridges' 718 ; stem cells of follicular structures, such
as hair, eyelash, and vibrissae, have been demonstrated recently to concentrate in the upper portion
of the outer root sheath known as the bulge27'28'80-31;
in murine epidermis, stem cells are thought to reside
in the center of a column of keratinocytes known as
the epidermal proliferative unit26'32(EPU); and pluripotent stem cells, which give rise to epithelial goblet
cells, are thought to reside at the bottom of the small
intestinal crypts.32'33
The apparent nonrandom and clustered location
of epithelial stem cells supports the concept of a stem
cell "niche"—perhaps maintained by the underlying
stroma—which provides a microenvironment for the
maintenance of the "sternness." *b Careful analysis of
the microanatomic compartments of stem cell systems
reveals many common features: With respect to the
external environment, epithelial stem cells are positioned deep in the tissue, presumably to provide maximal protection from environmental insults; the stroma
underlying these regions is organized into a network
of randomly arranged collagen and elastin fibers; a
high degree of vascularity and a high density of fibroblasts and other mesenchymal cells (e.g., mast cells)
are characteristically interspersed within the collagen-elastin network.26 Consistent with these attributes, fornical epithelium is located further from the
external environment than either the bulbar or palpebral epithelia. In addition, fornical stroma is the most
heavily vascularized and most cellular of the three conjunctival zones. Thus, the preferential distribution of
a subpopulation of cells with stem cell characteristics
in the fornix is consistent with the anatomic distribution of stem cells in other epithelia.
The Fornix Plays a Major Role in Conjunctival
Development
Fornical epithelial proliferation during the early
stages of neonatal life is indicative that this region
plays an important role during conjunctival epithelial
maturation. Consistent with this idea is the observation that goblet cells are first seen in the fornical region of the 8-day-old mouse; with time, they appear
in bulbar and palpebral zones. In a recent study on
the development of human conjunctival goblet cells,
alcian blue and PAS-positive staining cells (histochemical markers for goblet cells) first appeared in the
fornix region at 8 weeks of gestational age. In 9-week
specimens, goblet cells were most numerous in the
fornix.34 Some goblet cells were also present in the
palpebral conjunctiva but were rarely seen in the bulbar region. Because one of the hallmarks of conjunctival maturation is the presence of goblet cells, these
findings are further evidence of the central role that
the fornix plays in conjunctival epithelial development.
Goblet Cells Have Proliferative Capabilities
Conjunctival epithelium constandy renews itself and
has a remarkable regenerative capacity, as evidenced
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933181/ on 06/17/2017
244
Investigative Ophthalmology & Visual Science, January 1995, Vol. 36, No. 1
by the conjunctiva's ability to reepithelialize a denuded cornea and limbus.23'35 Furthermore, conjunctival epithelial proliferation has been shown to increase in response to wounds that involve only the
central corneal epithelium.13 Traditionally, the (nongoblet) basal epithelial cell has been considered the
major proliferative cell responsible for both conjunctival epithelial maintenance and corneal epithelial replacement. However, our finding that a subpopulation
of morphologically mature goblet cells can be labeled
after exposure to continuous and pulse administrations of tritiated thymidine, as well as the detection of
mitotically arrested goblet cells, are indicative that this
cell type also has a significant proliferative capability.
The ability of goblet cells to proliferate has been observed in several other epithelial tissues.36~39 For example, 0.8% of the "pit" cells (surface mucous cells with
a goblet cell phenotype) located in the glandular epithelium of mouse stomach can be labeled after a single exposure to 3H-TdR.3S A similar situation exists in
nonstimulated mammalian airway epithelium where
basal, Clara, serous, and mucous (goblet) cells have
all been observed to divide.31'3740 In hamster tracheal
epithelium, small numbers of basal (0.3%) and mucous (0.1%) cells were observed to be either resting
or proliferating more slowly than other labeled cells.
However, after perturbation (e.g., cigarette smoke or
mechanical trauma), an increased percentage of tracheal epithelial goblet cells was observed to proliferate,36'37 suggesting that goblet cells could play a major
role in tracheal epithelial repair.
That conjunctival goblet cells can actually divide,
as our data clearly show, raises the question as to where
these cells are positioned in the scheme of "stem cell
-* 'transient amplifying cell' —• postmitotic cell." Their
ability to proliferate, at least in mice, does not support
the concept that goblet cells are terminally differentiated. What remains uncertain, though, is whether goblet cells are transient amplifying cells or stem cells.
Evidence favoring a stem cell role comes from our
finding that some goblet cells can be identified as
LRCs and, hence are, slow cycling, an important criterion for stem cells. An alternative interpretation of
the label-retaining data would be that a goblet cell
precursor incorporated 3H-TdR before a final round
of division, entered a postmitotic state (still having a
labeled nucleus), and subsequently differentiated into
a goblet cell. In this regard, our observation that at
least some of the label-retaining goblet cells could
undergo mitosis (labeled mitotic figures after longterm labeling; Fig. 4A) is of major importance, because this proves that these label-retaining goblet cells
still have proliferative capabilities. Additional indirect
evidence for the slow-cycling nature of goblet cells
comes from pulse labeling data, which revealed that
only a small number of goblet cells incorporate 3H-
TdR after a single exposure. As previously mentioned,
regions enriched with slow-cycling cells (e.g., limbal
epithelium, bulge region of the hair follicle, tips of
deep rete ridges of palmar epithelium) usually do not
incorporate large amounts of pulse-administered tritiated thymidine. Thus, the combination of slow-cycling
kinetics (LRCs), proliferative capability, and infrequent thymidine incorporation after pulse labeling
protocols are features consistent with stem cells. In
normal situations, goblet cells are relatively quiescent,
although it is well established that they can rapidly
proliferate in response to wounding.41"48 Indeed, the
suggestion has been made that numerous goblet cells
play the major regenerative role in lung epithelium.3637 Thus, the ability of conjunctival goblet cells
to enter the proliferative pool in response to tissue
depletion, a typical feature of stem cells, is further
evidence of their stem cell nature.
The one inconsistent feature of goblet cells with
respect to stem cells concerns their state of differentiation. In all other systems studied so far, the slow-cycling cells (stem cells) are relatively undifferentiated
or primitive (e.g., limbal basal cells, nonserrated basal
keratinocytes). However, goblet cells have the phenotype of a highly differentiated or specialized cell. A
possible explanation for this paradoxic observation
centers around the secretory physiology of these cells.
Under normal situations, goblet cells function to synthesize and secrete mucin, which serves as a protective
barrier for the underlying epithelium. As has been
shown in intestinal, tracheal, and lung epithelia, as
well as in conjunctival epithelium, secretion occurs
either by slow, periodic exocytosis of the contents of
single mucin granules (baseline secretion) or through
the fusion of many adjacent granules followed by expulsion of the entire stored mass of granules at one
time (accelerated or stimulated secretion) . M4 - 46 After
accelerated secretion, the goblet cell resynthesizes
new granules and continues its secretory function.1'44"
46
Because these secretory events are independent of
proliferation and can apparently occur multiple times
during the life of a goblet cell, it explains how a goblet
cell can be proliferatively slow cycling and detected as
an LRC.
CONCLUSIONS
We have provided evidence to support our earlier suggestion that the Comical epithelium may be a zone
enriched in conjunctival epithelial stem cells. This notion has important implications in conjunctival epithelial development and can be extremely relevant in
wound repair. The finding that goblet cells are relatively quiescent, slow-cycling cells with certain stem
cell attributes provides new insights into the area of
conjunctival homeostasis. Although these studies es-
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933181/ on 06/17/2017
Enrichment of Conjunctival Epithelial Stem Cells in Fornix
245
tablish the proliferative nature of the various conjunctival zones and cell types within these zones, they do
not provide information on the lineage of the heterogeneous proliferative cell types. Are there two separate
stem cell populations that give rise to goblet cells and
stratified squamous epithelial cells, or is there a single
pluripotent conjunctival epithelial stem cell capable
of giving rise to both cell types? Ongoing studies comparing the growth and differentiation of isolated goblet and epithelial cell fractions should provide insight
into some of these issues.
basis of their proliferative behavior. NCI Mongraph.
1963;14:19-145.
Cotsarelis G, Cheng S-Z, Dong G, Sun T-T, Lavker
RM. Existence of slow-cycling limbal epithelial basal
cells that can be preferentially stimulated to proliferate: Implications of epithelial stem cells. Cell.
1989;57:201-209.
Doran TI, Vidrich A, Sun T-T. Intrinsic and extrinsic
regulation of the differentiation of skin, corneal and
esophageal epithelial cells. Cell. 1980;22:17-25.
Lavker RM, Sun T-T. Heterogeneity in epidermal
basal keratinocytes: Morphological and functional
correlations. Science. 1982;215:1239-1241.
Lavker RM, Sun T-T. Epidermal stem cells. J Invest
Dermatol. 1983;80:121-127.
Tseng SCG. Concept and application of limbal stem
cells. Eye. 1989;3:141-157.
Tsai RJF, Sun T-T, Tseng SCG. Comparison of limbal
and conjunctival autograft transplantation in corneal
surface reconstruction in rabbits. Ophthalmology.
1990; 97:446-455.
Lavker RM, Dong G, Cheng SZ, Cotsarelis G, Sun TT. Relative proliferative rates of limbal and corneal
epithelia: Implications on corneal epithelial migration, circadian rhythm, and suprabasally located DNAsynthesizing keratinocytes. Invest Ophthalmol Vis Sd.
1991;32:1864-1875.
Kinoshita S, Klorpes TC, Friend J, Thoft RA. Limbal
epithelium in ocular surface wound healing. Invest
Ophthalmol Vis Sri. 1982;23:73-80.
Kruse FE, ChengJJY, Tsai RJF, Tseng SCG. Conjunctival transdifferentiation is due to incomplete removal
of limbal basal epithelium. ARVO Abstracts. Invest
Ophthalmol Vis Sri. 1989; 30:520.
Tseng SCG, Chen JJY, Huang AJW, Kruse FE, Maskin
SL, Tsai RJF. Classification of conjunctival surgeries
for corneal diseases based on stem cell concept. Ophthalmic Clin North Am. 1990;3:595-610.
Cotsarelis G, Sun T-T, Lavker RM. Label-retaining
cells reside in the bulge of the pilosebaceous unit:
Implications for follicular stem cells, hair cycle and
skin carcinogenesis. Cell 1990;61:1329-1337.
Miller SJ, Lavker RM, Sun T-T. Keratinocyte stem cells
of cornea, skin and hair follicle: Common and distinguishing features. Semin Dev Biol. 1993;4:2l7-240.
Lavker RM, Miller S, Wilson C, et al. Hair follicle stem
cells: Their location, role in hair cycle and involvement in tumor formation. J Invest Dermatol.
1993; 101 (suppl):16S-25S.
Miller SJ, Lavker RM, Sun T-T. Hair follicles, stem
cells, and skin cancer. J Invest Dermatol. 1993; 100
(suppl):288S-294S.
Wei Z-G, Sun T-T, Lavker RM. Conjunctival goblet
cells have proliferative capabilities. ARVO Abstracts.
Invest Ophlhalmol Vis Sci. 1991;32:734.
Sun T-T, Cotsarelis, and Lavker RM: Hair follicular
stem cells: The bulge-activation hypothesis./Twwtf/Ztermatol. 1991;96(suppl):77S-78S.
Lavker RM, Margolis-Fryer E, Ostad M, Cotsarelis G,
Sun T-T. Cells in the bulge region of mouse hair follicle undergo transient proliferation during onset of
15.
16.
17.
Key Words
18.
conjunctival epithelium, stem cells, development, goblet
cells, proliferation
19.
References
1. Kessing SV. Mucous gland system of the conjunctiva.
AGPA Ophlhalmol SuppL 1968;95.
2. Wei Z-G, Wu R-L, Lavker RM, Sun T-T. In vitro growth
and differentiation of rabbit bulbar, fornix, and palpebral conjunctival epithelia. Invest Ophthalmol Vis Sri.
1993; 34:1814-1828.
3. Bron AJ, Mengher LS, Davey CC. The normal conjunctiva and its responses to inflammation. Trans Ophthalmol Soc UK. 1985; 104:424
4. Setzer PY, Nichols BA, Dawson CR. Unusual structure
of rat conjunctival epithelium. Invest Ophthalmol Vis
Sri. 1987; 27:531-537.
5. Maumenee AE, Scholz RO III. The histopathology of
the ocular lesions produce by the sulfur and nitrogen
mustards. Bull Johns Hopkins Hosp. 1948;82:121-147.
6. Friedenwald JS. Growth pressure and metaplasia of
conjunctival and corneal epithelium. Doc Ophthalmol.
1951;5-6:184-187.
7. Thoft RA, Friend J. Biochemical transformation of
regenerating ocular surface epithelium. Invest Ophthalmol. 1977; 16:14-20.
8. Thoft RA, Friend J. The X, Y, and Z hypothesis of
corneal epithelial maintenance. Invest Ophthalmol Vis
Sri. 1983;24:1442-1443.
9. Shapiro MS, Friend J, Thoft RA. Corneal re-epithelialization from the conjunctiva. Invest Ophthalmol Vis Sri.
1981; 21:135-142.
10. Kinoshita S, FriendJ, Thoft RA. Biphasic cell proliferation in transdifferentiation of conjunctival to corneal
epithelium in rabbits. Invest Ophthalmol Vis Sd.
1983;24:1008-1014.
11. Tseng SCG, Hirst LW, Farazdaghi M, and Green WR.
Goblet cell density and visualization during conjunctival transdifferentiation. Invest Ophthalmol Vis Sd.
1984; 25:1168-1176.
12. Banjo S, FriendJ, Thoft RA. Conjunctival epithelium
in healing of corneal epithelial wounds. Invest Ophthalmol Vis Sri. 1987;28:1445-1449.
13. Danjo S, FriendJ, Thoft RA. Conjunctival epithelium
in healing of corneal epithelial wounds. Invest Ophthalmol Vis Sri. 1987; 28:1445.
14. Leblond CP. Classification of cell populations on the
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933181/ on 06/17/2017
246
32.
33.
34.
35.
36.
37.
38.
Investigative Ophthalmology & Visual Science, January 1995, Vol. 36, No. 1
anagen and in response to physical and chemical stimu l i . / Invest Dermatol. 1992;98:581. Abstract.
Sun T-T, Cotsarelis, Lavker RM. Hair follicular stem
cells: The bulge-activation hypothesis. JInvest Dermatol.
1991;96(suppl):77S-78S.
Ebato B, Friend J, Thoft RA. Comparison of limbal
and peripheral human corneal epithelium in tissue
culture. Invest Ophthalmol Vis Sd. 1987; 28:1450-1456.
Lavker RM, Cotsarelis G, Wei ZG, Sun T-T. Stem cells
of pelage, vibrissae and eyelash follicles: The hair cycle
and tumor formation. Ann NYAcad Sd. 1991;642:214225.
Thrasher JD, Greulich RC. The duodenal progenitor
population: III: The progenitor cell cycle of principal,
goblet and paneth cells. J Exp ZooL 1966; 161:9-20.
Pptten CS. Regeneration in epithelial proliferative
units as exemplified by small intestinal crypts. Wile,
Chichester: Ciba Foundation Symposium 160;
1991; 54-76.
Miyashita K, Azuma N, Hida T. Morphological and
histochemical studies of goblet cells in developing human conjunctiva. JpnJ Ophthalmol. 1992;36:169-174.
Huang AJW, Tseng SCG. Corneal epithelial wound
39.
40.
41.
42.
43.
44.
45.
46.
healing in the absence of limbal epithelium. Invest,
Ophthalmol Vis Sd. 1991;32:96-105.
Ayers MM, Jeffery PK. Proliferation and differentiation in mammalian airway epithelium. Eur Respir J.
1988; 1:58-80.
Jeffery PK, Gaillard D, Moret S. Human airway secretory cells during development and in mature airway
epithelium. Eur Respir J. 1992;5:93-104.
MerzelJ, Lcblond CP. Origin and renewal of goblet
cells in the epithelium of the mouse small intestine.
Am J Anal 1969; 124:281-306.
Leblond CP, Cheng H. Identification of stem cells in
the small intestine of the mouse. In: Cairnie AB, Lala
PK, Osmond PG, eds. Stem Cells of Renewing Cell Population. New York: Academic Press; 1976:7-31.
Nettesheim P, Jetten AM, Inayama Y, et al. Pathways of
differentiation of airway epithelial cells. Environ Health
Perspect. 1990;85:3l7-329.
Verdugo P. Goblet cells secretion and mucogenesis.
Annu Rev Physiol. 1990;52:157-176.
Specian RD, Oliver MG. Functional biology of intestinal goblet cells. Am Physiol Soc. 1991;260:183-193.
Wagner U, von Wichert P. Control of mucus secretion
in airways. Respiration. 1991;58:l-8.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933181/ on 06/17/2017