Lectin binding sites on human endocervixA comparison with secretory and proliferative endometrium.

THE ANATOMICAL RECORD 215:262-266 (1986)
Lectin Binding Sites on Human Endocervix: A
Comparison With Secretory and Proliferative
Endometrium
YUN YEN, MAO-CHI LEE, MARTINA SALZMANN, AND N A N DAMJANOV
Department of Pathology and Laboratory Medicine, Hahnemann University School of
Medicine, Philadelphia, PA 19102-1192
ABSTRACT
Endocervix and corresponding endometrium of women of reproductive age were studied histochemically with 13 fluorescein isothiocyanate-labeled
lectins to delineate the differences between the epithelial cells in two anatomical
sites. Lectin from Maclura pornifera (MPA), Ulex europaeus (UEA-I), Glycine max
(SBA), and Vicia villosa (VVA) bound only to endocervical epithelium and were the
only four lectins that distinguished endocervical from endometrial epithelium. These
differences were independent of menstrual cyclic changes and blood group antigen
secretion. These data show that lectins can be used to histochemically distinguish
endocervical from endometrial glands.
In our previous work Gee and Damjanov, 1985) we
have shown that fluorescein isothiocyanate-labeled lectins can be used as histochemical reagents to specify
pregnancy-associated changes in the uterus. Several lectins reacted differentially with pregnant and nonpregnant endometrial cells and were useful for histochemical
distinction of pregnant and nonpregnant endometrium.
Similar differential binding of lectins was found in the
pregnant and nonpregnant mouse endometrium and oviduct (Lee et al., 1983).
In the present study we have extended these lectin
histochemical investigations to include human endocervix. We have compared lectin binding to endocervical
and endometrial glands and show that some lectins can
be used to distinguish cells in these two anatomical
locations. We also show that lectin binding sites expressed during the proliferative phase of the menstrual
cycIe remain on both endocervical and endometrial
glands in the secretory phase and that the lectin binding
to these cells does not depend on the blood group antigen
secretion.
MATERIALS AND METHODS
from these uteri were taken from the fundus, body, and
isthmus to determine whether there are any differences
between various sites.
Since uterine glands express blood group antigens in
secretors (Szulman, 1960, 19621, the secretor status of
individuals was determined by indirect immunofluorescence on parafin sections of endometrial tissue using
monoclonal antibodies to these antigens (Dako, Santa
Barbara, CA). Specimens showing blood group reactivity in blood vessels but not in glands were considered to
be from nonsecretors, and those showing reactivity of
both glands and vessels were secretors. There were 8
specimens from blood group A, 3 from blood group B,
and 5 from blood group 0 individuals; 7 were from
secretors and 9 from nonsecretors. These data corresponded to the serologically determined blood groups of
these patients. All tissues were fixed in 4% formaldehyde solution €or 12 h r and embedded in parafin. Paraffh sections (5 pm) were prepared routinely and
deparaffinized in xylene and rehydrated in graded
ethanol. Following a brief rinse in distilled water, the
sections were treated with 0.4% pepsin in 0.01 N hydrochloric acid for 2 hr at 37°C (Ekblom et al., 1982) and
washed twice with phosphate-buffered saline (PBS). Sections were then incubated with various FITC-labeled
lectins, listed in Table 1, in a moist chamber a t room
temperature for 30 min. Before incubation, FITC-labeled lectins were diluted in PBS to a concentration of
50 pg/ml. A11 FITC-labeled lectins and sugars were purchased from E.Y. Laboratory (San Mateo, CA). The con-
Endometrial and endocervical tissues from women of
reproductive age were obtained by curettage or hysterectomy. The phase of the endometrial cycle was determined by the last menstrual period and was corroborated
by histology by means of standard histologic criteria for
dating of human endometrium. Fourteen specimens
were collected by curettage: Seven specimens were collected during the midproliferative phase (days 7-10),
and seven specimens during the secretory phase. Two
Fig. 1. Tissues reacted with FITC-lectin from Maclura pornifera
specimens corresponded to day 16, two to day 20, two to
day 22-23 and one to day 26-27 of a 28-day cycle. Each (MPA). a. Endocervix, secretory phase, shows strong cytoplasmic reacin most cells. Some cells are nonreactive (arrows) ~ 2 2 0 .b.
specimen was composed of endocervical and endometrial tivity
Endometrium in secretory phase is nonreactive x 220.
tissues. Two additional hysterectomy specimens (one in
proliferative and one in day 18 secretory phase) removed
for leiomyoma were systematically examined. Samples Received August 19, 1985; accepted February 19, 1986.
0 1986 ALAN R. LISS, INC.
263
LECTIN BINDING TO ENDOCERVIX
TABLE 1. Lectins used and corresponding inhibiting sugars
Carbohydrate specificities
Source
Lectin
SUC.
ConA
WGA
STA
SBA
VVA
PNA
GSA-I
MPA
RCA-I
RCA-I1
UEA-I
PHA-E
PHA-L
Canaualia ensiformis
Triticum uulgare
Solanum tuberosum
Glycine max
Vicia uillosa
Arachis hypogaea
Griffonia simplicifolia
Maclura pomifera
Ricinus communis
Ricinus communis
Ulex eurouaeus
Phaseolug uulgaris
Phaseolus uulgaris
Inhibiting sugar
a-D-Man> a-D-Glc
(@-(1
+~)-D-G~CNAC)~,
sialic acid
(P-(1+4)-D-GlcNAc)'
a-D-GalNAc=P-D-GalNAc
a-D-GalNAc
0-D-Gal-(1- 3)-D-GalNAc
a-D-Gal
a-D-GalNAc> D-Gal
6-D-Gal> a-D-Gal
a- or 0-D-Gal=D-GalNAc
a-L-Fuc
Complex
Complex
D-GlcNAc = N-acetyl-D-glucosamine;D-GalNAc
=
Methyl a-mannopyranoside
D-GlcNAc
D-GIcNAc
D-GalNAc
D-GalNAc
D-Gal
D-Gal
D-GalNAc
D-Gal
D-Gal
L-FUC
D-GalNAc
D-GalNAc
N-acetyl-D-galactosamine; D-Gal = D-galactose; L-Fuc = L-fucose.
TABLE 2. Lectin binding sites on human endometrium and endocewix'
Lectin
SUC.
ConA
WGA
STA
SBA
VVA
PNA
GSA-I
MPA
RCA-I
RCA-I1
UEA-I
PHA-E
PHA-L
Proliferative phase
Endometrium
Endocervix
gland epithelium gland epithelium
++
++
++
-
+or
-
+
++
+
-'
+
++
+
+
+
+-+
++
+
++
++
+
+
Secretory phase
Endometrium
Endocervix
gland epithelium gland epithelium
++
++
++
-
+ or- -'
-
+
+-
+ or -'
+ or -'
+
++
+
+
+
++
++
+
++
++
+
+
'Symbols: The intensity of Lectin binding was subjectively marked as negative (-), faintly visible (+),
weak (+), and strong (+ +).
2PNA stains some glands but not others in both the secretory-phase and proliferative-phase endometrium.
PHA-E and PHA-L stain some glands but not others in the secretory endometrium.
264
Y.YEN, M.-C. LEE, M. SALZMANN, AND I. DAMJANOV
RESULTS
The findings are summarized in Table 2 and illustrated in part in Figures 1-4.
Endocervix
All 13 lectins used in this study, except GSA-I, reacted
with endocervical epithelium (Figs. 1-4). The reactivities of the lectins MPA (from Maclura pornifera) (Fig. 1)
and UEA-I (from Ulex europaeus)(Fig. 2 ) were estimated
as strong, whereas SBA (from Glycine rnax) (Fig. 3) and
VVA (from Vicia uillosa) (Fig. 4) reacted less intensely.
These lectins reacted with most but not all endocervical
cells. The unreactive cells were interspersed individually with reactive cells (Figs. l a , 2b) or formed groups
covering short segments of the mucosal surface (Fig. 3).
In some areas essentially all cells reacted with the lectins (Fig. 2a). In most instances the cytoplasmic lectin
binding sites were both apical and basal, although in
some instances the apical reactivity appeared more intense (Fig. 4). Mucus within the lumen of some glands
reacted also with all four lectins (Figs. l a , 3).
Endometrium
Epithelium of endometrial glands in proliferative
phase reacted with 8 lectins: SUC.ConA, WGA, STA,
PNA, RCA-I, RCA-11, PHA-E, PHA-L. The corresponding luminal epithelium reacted with all those lectins.
SBA reacted with luminal epithelial cells but was unreactive with glandular cells, as reported previously (Lee
and Damjanov, 1985). No difference was noted in the
reactivity of glands in the basalis and functionalis. Tissues sampled from fundus, body, and isthmus all reacted
in a n identical manner,
Epithelium obtained from secretory endometrium
showed the same lectin binding pattern as the proliferative phase samples (Figs. 1,2).
Comparison of Endometrial and Endocervical Epithelium
Three lectins (MPA, UEA-I, and VVA) reacted exclusively with endocervical glands (Figs. 1-4). SBA reacted
with endocervical glands and luminal epithelium of the
endometrium but was unreactive with endometrial
glands. The blood group and the secretor status of individual tissue donors did not influence the reactivity of
either endocervical or endometrial epithelium.
DISCUSSION
Fig. 2. Tissues reacted with FITC-lectin from Ulex europaeus WEAI). a. Endocervical glands in secretory phase show strong reactivity
uniformly in all cells. x350. b. Another area of cervical glandular
epithelium from the same specimen shows uneven reactivity of cells,
accounting for the so-called mosaic pattern. X 350. c. Endometrial
glands, secretory phase, are unreactive. x220.
trol sections were incubated in a solution containing
FITC-labeled lectins and 0.2 mmoles of the specific inhibitory sugar(s) listed in Table 1.Following incubation,
the slides were washed in PBS, coverslipped with glycerin, and examined with a U V microscope.
The present study shows that 12 lectins of different
carbohydrate specificities react with endocervical glands.
However, all except four reacted also with endometrial
glands, as reported previously (Lee and Damjanov, 1985).
Since MPA, UEA-I, SBA, and W A show differential
reactivity, these four lectins could be used as markers to
distinguish endometrial from endocervical glands.
Three cell types were identified in the endocervix by
electron microscopy: mucus cells, cells similar to the
epithelial cells of the endometrium, and ciliated cells
(Nilsson and Westman, 1961). Since mucus-secreting
cells of endocervix are different from either luminal or
glandular epithelial cells in the endometrium, most of
the differential binding of the four lectins could be due
to their reactivity with endocervical mucus-secreting
cells. These cells form the majority of endocervical cells.
The endocervical cells that are unreactive with lectins
LECTIN BINDING TO ENDOCERVIX
265
Fig. 3. Tissues reacted with FITC-lectin from Glycine mat: (SBA)
show reactivity of secretory phase endocervical epithelium in one area
and no reactivity in the adjacent segment (arrows). Intraluminal mucus is strongly reactive. ~ 2 2 0 .
Fig. 4. Tissue reacted with FITC-lectin from Vicia uillosa W A ) . In
the endocervical glandular epithelium in secretory phase, cells of strong
reactivity are intermixed with nonreactive cells. ~ 2 2 0 .
probably correspond to ciliated cells or to cells similar to
epithelial cells of the uterine body (Nilsson and Westman, 1961).
The endocervical glands undergo cyclic morphological
changes (Bradburn and Webb, 1951). The physical properties of endocervical mucus also change during the
menstrual cycle (Odeblad, 1962, 1968; Davajan and
Kunitake, 1969). However, there is no evidence that the
biochemical composition of the mucus undergoes cyclic
changes (Schumacher, 1970). Our findings disclose no
menstrual period-related differences in the lectin binding to intracellular or extracellular mucin and are thus
consistent with previous studies indicating that the biochemical composition of mucus does not change during
the menstrual cycle.
Odeblad (1969) and Schumacher and Pearl (1969) have
shown qualitative differences in the secretion of different glands from different portions of the endocervix that
account for the so-called “mosaic pattern” of epithelial
secretions. The present lectin histochemical study clearly
outlines this mosaic pattern within individual glands as
well as within groups of glands. Thus lectins are a good
probe for studying the heterogeneity of cell populations
as well as glands in the cervix.
Szulman (1960,1962) showed blood group antigens on
the endocervical and endometrial glands of secretors.
We have examined tissues from both secretor and nonsecretor donors and see no difference in lectin binding to
their endocervical glands. Thus, the reactivity of these
12 lectins to endocervical glands does not depend on the
secretor status or the blood group of the donor.
In summary, the present study shows that FITC-lectins can be used to distinguish normal endocervical from
endometrial glands. These changes probably reflect the
histological and functional differences between these
two anatomically and developmentally related epithelia. Next, we plan to use these histochemical data in the
study of uterine tumors in order to distinguish endometrial from endocervical neoplasms.
ACKNOWLEDGMENTS
This work was supported by National Institutes of
Health grants CA-23097 and CA-38405 and a fellowship
from the Lalor Foundation, Wilmington, Delaware. The
skillful secretarial help of Ms. Jacklyn Powell is gratefully acknowledged.
LITERATURE CITED
Bradburn, G.B., and C.F. Webb (1951) Cyclic variations in the endocervix. Am. J. Obstet. Gynecol., 62997-1008.
Davajan, V., and G. Kunitake (1969) Fractional in vivo and in vitro
examination of postcoital cervical mucus in the human. Fertil.
Steril., 20: 197-210.
Ekblom, P., M. Miettinen, J. Rapola, and J.M. Foidart (1982) Demonstration of laminin, a basement membrane glycoprotein in routinely processed formalin fixed human tissues. Histochemistry,
75t301-307.
Lee, M.C., and I. Damjanov (1985) Pregnancy-related changes in the
human endometrium revealed by lectin histochemistry. Histochemistry, 82275-280.
266
Y. YEN, M.-C. LEE, M. SALZMANN, AND I. DAMJANOV
Lee, M.-C., T.-C. Wu, Y.-J. Wan, and I. Damjanov (1983) Pregnancyrelated changes in the mouse oviduct and uterus revealed by differential binding of fluoresceinated lectins. Histochemistry, 79:365375.
Nilsson, O., and A. Westman (1961)The ultrastructure of the epithelial
cells of the endocervix during the menstrual cycIe. Acta Obstet.
Gynecol. Scand., 40:223-233.
Odeblad, E. (1962) Undulations of macromolecules in cervical mucus.
Int. J. Fertil., 7:313-319.
Odeblad, E. (1968) The functional structure of human cervical mucus.
Acta Obstet. Gynecol. Scand., 47(Suppl. 1):57-79.
Odeblad, E. (1969) Types of human cervical secretions. Acta Europ.
Fertil. It99-104.
Schumacher, G.F.B. (1970) Biochemistry of cervical mucus. Fertil. Steril., 21.697-705.
Schumacher, G.F.B., and M.J. Pearl (1969) Cyclic changes of muramidase (lysozyme) in cervical mucus. J. Reprod. Med., 3~171-178.
Seulman, A.E. (1960) The histological distribution of blood group substances A and B in man. J. Exp. Med., 111:785-799.
Szulman, A.E. (1962) The histological distribution of the blood group
substances in man as disclosed by immunofluorescence. II. The H
antigen and its relation to A and B antigen. J. Exp. Med., 115:977995.