The Human Limbus A
Scanning
Electron
Jack V. Greiner, PhD;
Microscopic Study
Henry I. Covington, MS;
human limbal biopsy speciobtained from seven normal subjects. The epithelial surface was examined by scanning electron microscopy. Epithelial cell surfaces varied greatly in shape and size, and mosaics of three- to six-sided irregular polygons were formed. Microvillar borders of cells were distinct. Cell sizes ranged from 3 to 20 \g=m\ across. Light and dark cells were present and randomly distributed. A few cells were covered with microplicae; the remainder of the cells were covered with microvilli. Intercellular crypt openings measuring 1 to 10 \g=m\ in diameter were distributed randomly over all specimens. These openings were believed to be related to goblet cells. Many openings were plugged with what appeared to be mucus. Many surface-level intercellular structures were present; they corresponded in diameter to the surface openings. The human limbal epithelium varies from both the upper tarsal conjunctiva and the \s=b\ Fourteen
mens were
cornea.
(Arch Ophthalmol 97:1159-1165, 1979)
Accepted
for publication Sept 22, 1978. From the Department of Ophthalmology, Harvard Medical School, and the Department of Cornea Research, Eye Research Institute of Retina Foundation, Boston (Drs Greiner and Allansmith), and the Ernest E. Just Laboratory of Cellular Biology, Department of Anatomy, Howard University College of Medicine, Washington, DC (Mr Covington). Reprints not available.
Mathea R.
Allansmith,
MD
'"Phe human limbus is affected by
a
variety of diseases, including su¬ perior limbic keratoconjunctivitis, tra¬ choma, and vernal conjunctivitis. Knowledge of the electron microscopic
surface appearance of the normal limbal conjunctiva is necessary for study and interpretation of conjuncti¬ val disorders that have limbal surface changes. No scanning electron micro¬ scopic study of the human limbal conjunctival surface cells and surface openings has been reported, to our
knowledge. This scanning electron microscopic study of the limbal conjunctival epithelium provides standards for comparison so that changes in dis¬ eased tissues may be recognized.
cardboard. Mounting the biopsy specimens cardboard prevented the tissue from
on
folding during processing (Fig 1). Fixation
biopsy specimens were immedi¬ ately immersed in 3% glutaraldehyde solu¬ Mounted
tion in 0.15M of sodium cacodylate buffer (pH, 7.2), at room temperature, for three to four hours. Specimens were rinsed in buffer three times and were refrigerated overnight. They were dehydrated in graded ethyl alcohols, treated with a tran¬ sitional fluid (Freon 113), and dried in a critical point drying apparatus using a critical point transitional fluid (Freon 13). Specimens were mounted with silver conducting paint on aluminum stubs and coated with thin layers of carbon and goldpalladium in a sputter coater. Tissues were examined with a scanning electron micro¬ scope operated at 20 kV.
MATERIALS AND METHODS
RESULTS
Biopsy Specimens
Surface Cells
A total of 14 biopsy specimens were obtained from seven subjects (Table). In this study, the limbal area is defined as that area located 1 to 3 mm from the cornea. Sites of biopsy were at one or more of the four limbal areas shown in Fig 1. Anesthesia was done by one of the follow¬ ing methods: (1) subconjunctivally by infil¬ tration with a 2% solution of lidocaine (Xylocaine), (2) topically by instillation of
proparacaine hydrochloride drops,
local
nerve
block for
an
or
(3)
unrelated eye
problem (Table). No diagnostic or irrigat¬ ing solutions were instilled in the conjunc¬ tiva before biopsy. After anesthesia, the conjunctiva was grasped with a fine-toothed forceps and the specimen, measuring 1x2 mm, was
was cut with iris scissors. The tissue mounted flat on a supportive piece of
In general, the limbal surfaces were flat (Fig 1 and 2). However, in some areas the cells formed a rough surface (Fig 3). Occasionally some wrinkled areas were observed (Fig 2). This wrinkling may have been artifactual, caused by pulling on the tissue with forceps during biopsy. At low magni¬ fication, randomly distributed light and dark cells were seen. Light cells were usually smaller than dark cells, which appeared slightly depressed from the surface (Fig 2). Surface cell boundaries were outlined by microvillar borders (Fig 2, inset). Surface cells formed a mosaic of three- to six-sided irregular polygons on the limbal
Downloaded From: http://archopht.jamanetwork.com/ by a West Virginia University Library User on 06/21/2015
Biopsy
of Limbal
Conjunctival
Area Limbal Area
Age, yr 29 31 38 46
Biopsy Site, Anesthesia Local nerve block Subconjunctival infiltration Subconjunctival infiltration Nerve block and proparacaine
Clock Position 12 12 12 12
hydrochloride 56 58 64
Retrobulbar and local Retrobulbar and local Retrobulbar and local
nerve
block
nerve
block block
nerve
12, 3, 6, 9 12, 3, 6, 9 12, 9
Location of biopsy specimens at 3-, 6-, 9-, and 12-o'clock positions (squares). Right, Biopsy specimen mounted flat on supporting piece of cardboard (scale indicates 1 mm)
Fig 1.—Left,
(x23).
Fig 2.—Low magnification of limbal surface at 12-o'clock position with light cells (Ic), dark cells (dc), and epithelial surface crypts (small arrows). Some surface wrinkling is present. Large arrow is area of inset (scale indicates 10 µ) ( 790). Inset, High magnification shows irregular size and shape of light (Ic) and dark (dc) cell surfaces and surface crypts. Some mucous debris is present (scale indicates 1 µ) ( 4,070).
Downloaded From: http://archopht.jamanetwork.com/ by a West Virginia University Library User on 06/21/2015
Fig
3—At
high magnification,
smooth-surfaced cells
(large arrows)
Fig 4.—Limbal epithelium at 12o'clock position showing cell covered with microplicae (small arrows). Other cells have microvillar-tufted structures (large ar¬ rows). Cell boundaries are prom¬ inent (scale indicates 5 µ)
have matted
microvilli, whereas rough-surfaced cells have microvilli that vary in length; longer
(small arrows) stand out tiom surface (scale indicates 1 µ) ( 3,320). Inset, Low magnification of limbal surface at 6-o'clock position with some raised cells. Some cells (arrows) have smoother surfaces than others (scale indicates 10 µ) ( x 727). ones
surface. Cells varied greatly in shape and size. Size ranged from 3 to 20 µ across. In general, cell surfaces were covered with microvilli. At high mag¬ nification, microvilli showed some variation in length but were generally similar in diameter (0.25 µ). Microvilli of light cells appeared longer than those of dark cells. There were areas with light cells showing substantial variation in length of microvilli (Fig 3). Microvilli of dark cells appeared matted. A few cells among the light cells showed tuft-like aggregations of microvilli (Fig 4). Microvilli were densely packed in both light and dark cells. Although most cells were cov¬ ered with microvilli, a few cell
surfaces
were
found to be covered
microplicae.
by
Mucous debris was seen in some as white particulate masses on the cell surfaces. Occasionally spheroi¬ dal granules measuring 0.05 to 0.2 µ in diameter were seen (Fig 5).
places
Surface
Crypts
Round or elliptical intercellular sur¬ face crypt openings, measuring 1 to 10 µ in diameter, were randomly distrib¬ uted over the surface, usually at the junction of three or four surface cells. Some areas showed no crypt openings (Fig 3) but others showed groupings of them (Fig 6). Crypt openings were often surrounded by a microvillar
(x 4,620).
border. Some crypts appeared empty, and cytoplasmic processes were seen at their bases (Fig 7). Many openings were plugged with what appeared to be mucus (Fig 6). Mucus was common¬ ly observed extruding from some crypt openings onto the epithelial surface (Fig 6). Other, more shallow
openings were bridged by cytoplasmic processes resembling microvilli (Fig
7). Some round intercellular surface
structures had diameters similar to
those of the crypt openings and were covered with a plasma membrane with some microvillar-like processes. Some of these structures appeared almost level with the general surface and others appeared to be either shallow
Downloaded From: http://archopht.jamanetwork.com/ by a West Virginia University Library User on 06/21/2015
Fig 5.—Limbal epithelial surface at 12-o'clock position showing small spheroidal structures adherent to microvilli covering surface. Large arrow indicates area of inset; microvillar borders are shown by small arrows (scale indicates 5 µ) ( 4,660). Inset, High magnification of spheroidal structures and microvillar surface. Note mucous material between microvilli, indicated by arrows (scale indicates 1 µ) ( 19,330).
Downloaded From: http://archopht.jamanetwork.com/ by a West Virginia University Library User on 06/21/2015
Fig 6.—Top, Low magnification of limbal biopsy specimen at 12-o'clock position with numerous crypt openings with mucus plugs (arrows). Surface wrinkling (upper right) and artifact of tissue cracking (upper left and along top) are present. Black square outlines region shown in bottom micrograph (scale indicates 50 µ) ( 400). Bottom, High magnification of crypt opening showing mucus appearing to be extruding onto limbal surface (scale indicates 5 µ) ( 7,550).
Downloaded From: http://archopht.jamanetwork.com/ by a West Virginia University Library User on 06/21/2015
from 6-o'clock position showing empty crypt opening (large arrow). Surrounding of various lengths. Long microvillus (small arrow) is present (scale indicates 1 µ) ( 11,540). , Specimen from 12-o'clock position with microvilli (arrow) inside of empty crypt opening (scale indicates 1 µ) ( 13,240). C, Specimen from 12 o'clock position. Circular surface structure is covered with microvilli showing mucus-like forms centrally (scale indicates 1 µ) ( 10,520). D, Specimen from 12-o'clock position. Round surface structure shows microvilli centrally (scale indicates 1 µ) ( x 13,240). E, Specimen from 12-o'clock position. Mucus (arrows) is from round surface structure sparsely populated with microvilli (scale indicates 1µ) ( 9,200). F, Specimen from 6-o'clock position. Microvilli show through mucus In mucus-filled crypt, and mucus particle is located centrally (scale indicates 1 µ) ( 6,000). G, Specimen from 12-o'clock position showing mucus-filled crypt (scale indicates 1 µ) ( 9,200). , Specimen from 12-o'clock position. Mucus-filled crypt opening is filled with smooth mucus, appearing to be extruding mucus onto epithelial surface. Particle of mucous debris is present (scale indicates 1 µ) ( 5,400).
Fig 7.—A, Specimen microvilli
are
Downloaded From: http://archopht.jamanetwork.com/ by a West Virginia University Library User on 06/21/2015
depressions or raised elevations of the epithelial surface (Fig 7). These struc¬ tures usually appeared to have micro¬ villar borders.
COMMENT Surface Cells
A comparison of the limbal conjunc¬ tival surface with the upper tarsal conjunctival surface1 shows that the epithelial cells over the limbal con¬ junctiva are much more irregular in shape and larger in size. The presence of light and dark cells has been seen previously on rabbit conjunctiva4 and human conjunctiva.' Although most cells were covered with microvilli, a few cells on the limbal area had curvi¬ linear microplicae, similar to those reported by Pfister and Burstein4 on the human corneal surface. Surface
Crypts
As in the rabbit conjunctiva4 and the human upper tarsal conjunctiva,'·' small intercellular crypt openings
found at the junction of three or four cells. These were thought to be the openings of underlying crypts. Since we know that goblet cells are present in the limbal conjunctiva47 and since the crypt openings appear identical to crypts that are formed by goblet cells in other tissues,"" it seems reasonable that the small crypts we found in this study are the collapsed surfaces of empty goblet cells. Fur¬ ther evidence for the relationship between these surface openings and goblet cells is provided by scanning electron microscopic studies by Pfis¬ ter- and Greiner and associates,' who have shown mucus extruding from conjunctival crypt openings. Many of the surface crypt openings in the limbal areas examined in this study were larger in diameter than those found over the upper tarsal conjuncti¬ va.' This enlarged diameter is ex¬ pected since goblet cells in the limbal area are generally slightly wider in diameter and shorter in height than were
those in the upper tarsal conjunctiva.7 This is due to the difference in the type of epithelium; on the tarsal conjunctiva, the epithelium is strati¬ fied columnar epithelium, and on the globe the epithelium is flattened. This results in a change in goblet cell
shape.7
In this study the appearance of the human limbal conjunctival epithelial surface showed variations from that of the upper tarsal conjunctival epi¬ thelium and that of the corneal epithe¬ lium. This study was supported by Institutional National Research Service Award EY-07018 and grants EY-00208 and EY-01552 from the Nation¬ al Eye Institute, National Institutes of Health; and by a grant from Bausch and Lomb. Larry Cain provided photographic assistance.
Name and Trademarks of Drug
Nonproprietary
Proparacaine hydrochloride—Alcaine, Ophthaine, Ophthetic.
References 1. Greiner JV, Covington HI, Allansmith MR: Surface morphology of the human upper tarsal conjunctiva. Am J Ophthalmol 83:892-905, 1977. 2. Pfister RR: The normal surface of conjunctiva epithelium: A scanning electron microscopic study. Invest Ophthalmol 14:267-279, 1975. 3. Pfister RR, Burstein NL: The normal and abnormal human corneal epithelial surface: A scanning electron microscope study. Invest Ophthalmol Visual Sci 16:614-622, 1977.
4. Dark AJ, Durrant TE, McGinty F, et al: Tarsal conjunctiva of the upper eyelid. Am J Ophthalmol 77:555-564, 1974. 5. Virchow H: Conjunctiva, in Graefe A, Saemisch T (eds): Handbuch der Augenheilkunde, ed 2. Leipzig, Germany, W. Engelmann, 1910, vol 1, pt 1, chap 3, p 431. 6. Podhoranyi G: \l=U"\berdie Becherzellen der Bindehaut. Albrecht von Graefes Arch Ophthalmol 169:285-293, 1966.
7. Kessing SV: Mucus gland system of the conjunctiva: A quantitative normal anatomical study. Acta Ophthalmol 95(suppl):61, 1968. 8. Marsh MN, Swift TA: A study of the small intestinal mucosa using the scanning electron microscope. Gut 10:940, 1969. 9. Mebus CA, Newman LE, Stare EL: Scanning electron, light, and transmission electron microscopy of intestine of gnotobiotic calf. Am J
Vet Res 36:985, 1975.
Downloaded From: http://archopht.jamanetwork.com/ by a West Virginia University Library User on 06/21/2015
Scanning
Electron
Jack V. Greiner, PhD;
Microscopic Study
Henry I. Covington, MS;
human limbal biopsy speciobtained from seven normal subjects. The epithelial surface was examined by scanning electron microscopy. Epithelial cell surfaces varied greatly in shape and size, and mosaics of three- to six-sided irregular polygons were formed. Microvillar borders of cells were distinct. Cell sizes ranged from 3 to 20 \g=m\ across. Light and dark cells were present and randomly distributed. A few cells were covered with microplicae; the remainder of the cells were covered with microvilli. Intercellular crypt openings measuring 1 to 10 \g=m\ in diameter were distributed randomly over all specimens. These openings were believed to be related to goblet cells. Many openings were plugged with what appeared to be mucus. Many surface-level intercellular structures were present; they corresponded in diameter to the surface openings. The human limbal epithelium varies from both the upper tarsal conjunctiva and the \s=b\ Fourteen
mens were
cornea.
(Arch Ophthalmol 97:1159-1165, 1979)
Accepted
for publication Sept 22, 1978. From the Department of Ophthalmology, Harvard Medical School, and the Department of Cornea Research, Eye Research Institute of Retina Foundation, Boston (Drs Greiner and Allansmith), and the Ernest E. Just Laboratory of Cellular Biology, Department of Anatomy, Howard University College of Medicine, Washington, DC (Mr Covington). Reprints not available.
Mathea R.
Allansmith,
MD
'"Phe human limbus is affected by
a
variety of diseases, including su¬ perior limbic keratoconjunctivitis, tra¬ choma, and vernal conjunctivitis. Knowledge of the electron microscopic
surface appearance of the normal limbal conjunctiva is necessary for study and interpretation of conjuncti¬ val disorders that have limbal surface changes. No scanning electron micro¬ scopic study of the human limbal conjunctival surface cells and surface openings has been reported, to our
knowledge. This scanning electron microscopic study of the limbal conjunctival epithelium provides standards for comparison so that changes in dis¬ eased tissues may be recognized.
cardboard. Mounting the biopsy specimens cardboard prevented the tissue from
on
folding during processing (Fig 1). Fixation
biopsy specimens were immedi¬ ately immersed in 3% glutaraldehyde solu¬ Mounted
tion in 0.15M of sodium cacodylate buffer (pH, 7.2), at room temperature, for three to four hours. Specimens were rinsed in buffer three times and were refrigerated overnight. They were dehydrated in graded ethyl alcohols, treated with a tran¬ sitional fluid (Freon 113), and dried in a critical point drying apparatus using a critical point transitional fluid (Freon 13). Specimens were mounted with silver conducting paint on aluminum stubs and coated with thin layers of carbon and goldpalladium in a sputter coater. Tissues were examined with a scanning electron micro¬ scope operated at 20 kV.
MATERIALS AND METHODS
RESULTS
Biopsy Specimens
Surface Cells
A total of 14 biopsy specimens were obtained from seven subjects (Table). In this study, the limbal area is defined as that area located 1 to 3 mm from the cornea. Sites of biopsy were at one or more of the four limbal areas shown in Fig 1. Anesthesia was done by one of the follow¬ ing methods: (1) subconjunctivally by infil¬ tration with a 2% solution of lidocaine (Xylocaine), (2) topically by instillation of
proparacaine hydrochloride drops,
local
nerve
block for
an
or
(3)
unrelated eye
problem (Table). No diagnostic or irrigat¬ ing solutions were instilled in the conjunc¬ tiva before biopsy. After anesthesia, the conjunctiva was grasped with a fine-toothed forceps and the specimen, measuring 1x2 mm, was
was cut with iris scissors. The tissue mounted flat on a supportive piece of
In general, the limbal surfaces were flat (Fig 1 and 2). However, in some areas the cells formed a rough surface (Fig 3). Occasionally some wrinkled areas were observed (Fig 2). This wrinkling may have been artifactual, caused by pulling on the tissue with forceps during biopsy. At low magni¬ fication, randomly distributed light and dark cells were seen. Light cells were usually smaller than dark cells, which appeared slightly depressed from the surface (Fig 2). Surface cell boundaries were outlined by microvillar borders (Fig 2, inset). Surface cells formed a mosaic of three- to six-sided irregular polygons on the limbal
Downloaded From: http://archopht.jamanetwork.com/ by a West Virginia University Library User on 06/21/2015
Biopsy
of Limbal
Conjunctival
Area Limbal Area
Age, yr 29 31 38 46
Biopsy Site, Anesthesia Local nerve block Subconjunctival infiltration Subconjunctival infiltration Nerve block and proparacaine
Clock Position 12 12 12 12
hydrochloride 56 58 64
Retrobulbar and local Retrobulbar and local Retrobulbar and local
nerve
block
nerve
block block
nerve
12, 3, 6, 9 12, 3, 6, 9 12, 9
Location of biopsy specimens at 3-, 6-, 9-, and 12-o'clock positions (squares). Right, Biopsy specimen mounted flat on supporting piece of cardboard (scale indicates 1 mm)
Fig 1.—Left,
(x23).
Fig 2.—Low magnification of limbal surface at 12-o'clock position with light cells (Ic), dark cells (dc), and epithelial surface crypts (small arrows). Some surface wrinkling is present. Large arrow is area of inset (scale indicates 10 µ) ( 790). Inset, High magnification shows irregular size and shape of light (Ic) and dark (dc) cell surfaces and surface crypts. Some mucous debris is present (scale indicates 1 µ) ( 4,070).
Downloaded From: http://archopht.jamanetwork.com/ by a West Virginia University Library User on 06/21/2015
Fig
3—At
high magnification,
smooth-surfaced cells
(large arrows)
Fig 4.—Limbal epithelium at 12o'clock position showing cell covered with microplicae (small arrows). Other cells have microvillar-tufted structures (large ar¬ rows). Cell boundaries are prom¬ inent (scale indicates 5 µ)
have matted
microvilli, whereas rough-surfaced cells have microvilli that vary in length; longer
(small arrows) stand out tiom surface (scale indicates 1 µ) ( 3,320). Inset, Low magnification of limbal surface at 6-o'clock position with some raised cells. Some cells (arrows) have smoother surfaces than others (scale indicates 10 µ) ( x 727). ones
surface. Cells varied greatly in shape and size. Size ranged from 3 to 20 µ across. In general, cell surfaces were covered with microvilli. At high mag¬ nification, microvilli showed some variation in length but were generally similar in diameter (0.25 µ). Microvilli of light cells appeared longer than those of dark cells. There were areas with light cells showing substantial variation in length of microvilli (Fig 3). Microvilli of dark cells appeared matted. A few cells among the light cells showed tuft-like aggregations of microvilli (Fig 4). Microvilli were densely packed in both light and dark cells. Although most cells were cov¬ ered with microvilli, a few cell
surfaces
were
found to be covered
microplicae.
by
Mucous debris was seen in some as white particulate masses on the cell surfaces. Occasionally spheroi¬ dal granules measuring 0.05 to 0.2 µ in diameter were seen (Fig 5).
places
Surface
Crypts
Round or elliptical intercellular sur¬ face crypt openings, measuring 1 to 10 µ in diameter, were randomly distrib¬ uted over the surface, usually at the junction of three or four surface cells. Some areas showed no crypt openings (Fig 3) but others showed groupings of them (Fig 6). Crypt openings were often surrounded by a microvillar
(x 4,620).
border. Some crypts appeared empty, and cytoplasmic processes were seen at their bases (Fig 7). Many openings were plugged with what appeared to be mucus (Fig 6). Mucus was common¬ ly observed extruding from some crypt openings onto the epithelial surface (Fig 6). Other, more shallow
openings were bridged by cytoplasmic processes resembling microvilli (Fig
7). Some round intercellular surface
structures had diameters similar to
those of the crypt openings and were covered with a plasma membrane with some microvillar-like processes. Some of these structures appeared almost level with the general surface and others appeared to be either shallow
Downloaded From: http://archopht.jamanetwork.com/ by a West Virginia University Library User on 06/21/2015
Fig 5.—Limbal epithelial surface at 12-o'clock position showing small spheroidal structures adherent to microvilli covering surface. Large arrow indicates area of inset; microvillar borders are shown by small arrows (scale indicates 5 µ) ( 4,660). Inset, High magnification of spheroidal structures and microvillar surface. Note mucous material between microvilli, indicated by arrows (scale indicates 1 µ) ( 19,330).
Downloaded From: http://archopht.jamanetwork.com/ by a West Virginia University Library User on 06/21/2015
Fig 6.—Top, Low magnification of limbal biopsy specimen at 12-o'clock position with numerous crypt openings with mucus plugs (arrows). Surface wrinkling (upper right) and artifact of tissue cracking (upper left and along top) are present. Black square outlines region shown in bottom micrograph (scale indicates 50 µ) ( 400). Bottom, High magnification of crypt opening showing mucus appearing to be extruding onto limbal surface (scale indicates 5 µ) ( 7,550).
Downloaded From: http://archopht.jamanetwork.com/ by a West Virginia University Library User on 06/21/2015
from 6-o'clock position showing empty crypt opening (large arrow). Surrounding of various lengths. Long microvillus (small arrow) is present (scale indicates 1 µ) ( 11,540). , Specimen from 12-o'clock position with microvilli (arrow) inside of empty crypt opening (scale indicates 1 µ) ( 13,240). C, Specimen from 12 o'clock position. Circular surface structure is covered with microvilli showing mucus-like forms centrally (scale indicates 1 µ) ( 10,520). D, Specimen from 12-o'clock position. Round surface structure shows microvilli centrally (scale indicates 1 µ) ( x 13,240). E, Specimen from 12-o'clock position. Mucus (arrows) is from round surface structure sparsely populated with microvilli (scale indicates 1µ) ( 9,200). F, Specimen from 6-o'clock position. Microvilli show through mucus In mucus-filled crypt, and mucus particle is located centrally (scale indicates 1 µ) ( 6,000). G, Specimen from 12-o'clock position showing mucus-filled crypt (scale indicates 1 µ) ( 9,200). , Specimen from 12-o'clock position. Mucus-filled crypt opening is filled with smooth mucus, appearing to be extruding mucus onto epithelial surface. Particle of mucous debris is present (scale indicates 1 µ) ( 5,400).
Fig 7.—A, Specimen microvilli
are
Downloaded From: http://archopht.jamanetwork.com/ by a West Virginia University Library User on 06/21/2015
depressions or raised elevations of the epithelial surface (Fig 7). These struc¬ tures usually appeared to have micro¬ villar borders.
COMMENT Surface Cells
A comparison of the limbal conjunc¬ tival surface with the upper tarsal conjunctival surface1 shows that the epithelial cells over the limbal con¬ junctiva are much more irregular in shape and larger in size. The presence of light and dark cells has been seen previously on rabbit conjunctiva4 and human conjunctiva.' Although most cells were covered with microvilli, a few cells on the limbal area had curvi¬ linear microplicae, similar to those reported by Pfister and Burstein4 on the human corneal surface. Surface
Crypts
As in the rabbit conjunctiva4 and the human upper tarsal conjunctiva,'·' small intercellular crypt openings
found at the junction of three or four cells. These were thought to be the openings of underlying crypts. Since we know that goblet cells are present in the limbal conjunctiva47 and since the crypt openings appear identical to crypts that are formed by goblet cells in other tissues,"" it seems reasonable that the small crypts we found in this study are the collapsed surfaces of empty goblet cells. Fur¬ ther evidence for the relationship between these surface openings and goblet cells is provided by scanning electron microscopic studies by Pfis¬ ter- and Greiner and associates,' who have shown mucus extruding from conjunctival crypt openings. Many of the surface crypt openings in the limbal areas examined in this study were larger in diameter than those found over the upper tarsal conjuncti¬ va.' This enlarged diameter is ex¬ pected since goblet cells in the limbal area are generally slightly wider in diameter and shorter in height than were
those in the upper tarsal conjunctiva.7 This is due to the difference in the type of epithelium; on the tarsal conjunctiva, the epithelium is strati¬ fied columnar epithelium, and on the globe the epithelium is flattened. This results in a change in goblet cell
shape.7
In this study the appearance of the human limbal conjunctival epithelial surface showed variations from that of the upper tarsal conjunctival epi¬ thelium and that of the corneal epithe¬ lium. This study was supported by Institutional National Research Service Award EY-07018 and grants EY-00208 and EY-01552 from the Nation¬ al Eye Institute, National Institutes of Health; and by a grant from Bausch and Lomb. Larry Cain provided photographic assistance.
Name and Trademarks of Drug
Nonproprietary
Proparacaine hydrochloride—Alcaine, Ophthaine, Ophthetic.
References 1. Greiner JV, Covington HI, Allansmith MR: Surface morphology of the human upper tarsal conjunctiva. Am J Ophthalmol 83:892-905, 1977. 2. Pfister RR: The normal surface of conjunctiva epithelium: A scanning electron microscopic study. Invest Ophthalmol 14:267-279, 1975. 3. Pfister RR, Burstein NL: The normal and abnormal human corneal epithelial surface: A scanning electron microscope study. Invest Ophthalmol Visual Sci 16:614-622, 1977.
4. Dark AJ, Durrant TE, McGinty F, et al: Tarsal conjunctiva of the upper eyelid. Am J Ophthalmol 77:555-564, 1974. 5. Virchow H: Conjunctiva, in Graefe A, Saemisch T (eds): Handbuch der Augenheilkunde, ed 2. Leipzig, Germany, W. Engelmann, 1910, vol 1, pt 1, chap 3, p 431. 6. Podhoranyi G: \l=U"\berdie Becherzellen der Bindehaut. Albrecht von Graefes Arch Ophthalmol 169:285-293, 1966.
7. Kessing SV: Mucus gland system of the conjunctiva: A quantitative normal anatomical study. Acta Ophthalmol 95(suppl):61, 1968. 8. Marsh MN, Swift TA: A study of the small intestinal mucosa using the scanning electron microscope. Gut 10:940, 1969. 9. Mebus CA, Newman LE, Stare EL: Scanning electron, light, and transmission electron microscopy of intestine of gnotobiotic calf. Am J
Vet Res 36:985, 1975.
Downloaded From: http://archopht.jamanetwork.com/ by a West Virginia University Library User on 06/21/2015
