[CANCER RESEARCH 52, 2923-2930, May 15, 1992]
Mucin Gel Formed by Tumorigenic Squamous Lung Carcinoma Cells Has Lea-X
Oligosaccharides and Excludes Antibodies from Underlying Cells1
Patricia L. Stranahan, Randy B. Howard,2 Oswald Pfenninger, Michael E. Cowen,3 Michael R. Johnston,2 and
David E. Pettijohn4
University of Colorado Cancer Center, Department of Biochemistry, Biophysics and Genetics [D. E. P., P. L. S., O. P.], and Department of Surgery [R. B. H., M. E. C\,
M. R. J.J, University of Colorado Health Sciences Center, Denver, Colorado 80262
ABSTRACT
Cells of cloned lines of human squamous lung carcinomas elaborate
large glycoproteins that are associated with their tumorigenic potential.
Two groups of clones (called Le"-X-positive and Le*-X-negative) were
studied that either do or do not express the Le'-X oligosaccharide
associated with large glycoproteins and mucins secreted by these clones.
Le"-X-positive cells elaborate a mucin gel complex associated with their
apical surfaces, which appears as a mosaic of extracellular plates. Clones
of this type are tumorigenic in nude rodents when injected s.c. or when
introduced into the lungs via intrabronchial aerosol. By contrast, the I.e"X-negative clones do not form extracellular plates and are not tumorigenic
in the lungs or subcutaneously. We demonstrate that the extracellular
plates of Le*-X-positive cells exclude antibodies from interacting with
the underlying squamous lung carcinoma cells
an immunoprotective effect. In support of this
that: (a) There is a substantial inflammatory
with regressing nodules of Le*-X-negative cells
and may therefore exert
possibility it was found
cell infiltrate associated
in nude rodent lung and
subcutaneous nodules, while there is no observable infiltration associated
with progressing Le'-X-positive tumors, (b) In the brain (an immunoprivileged site) tumors develop and progress when either Le"-X-negative or
-positive cells are introduced.
INTRODUCTION
One model for carcinogenesis, unregulated growth, is a triad
(1). The components which are believed to be necessary for
autonomous cell behavior are (a) the synthesis, storage, and
release of products which alter (b) interaction with ECM5 as
well as (c) immune surveillance escape. In normal tissue, the
positions of cells are organized by the ECM, which also serves
as a source of specific signals regulating cell proliferation and
differentiation (1). Epithelial cells, like those lining the bron
chial tree, are not completely surrounded by ECM but are
bounded on the basal surface by a specialized ECM, the basal
lamina. The ECM and basal lamina are complex lattices made
up of several specialized proteins such as collagen, laminin, and
elastin, and the cells are also associated with specific proteoglycans and mucins. Epithelial cell glycoconjugates are normally
synthesized, stored, and secreted onto a luminal or ductal
surface. Upon transformation, the secretion of these glycocon
jugates no longer is directional or orderly; thus, altered molec
ular interaction with extracellular matrix occurs. It has been
Received 10/22/91; accepted 3/10/92.
The costs of publication of this article were defrayed in part by the payment
of page charges. This article must therefore be hereby marked advertisement in
accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1Supported by Grant BC-685 from the American Cancer Society (D. E. P.)
and by NIH National Cooperative Drug Discovery Group Grant CA46088 (M.
R. J.).
2 Present address: Samuel Lunenfeld Research Institute, Division of Thoracic
Surgery, Mt. Sinai Hospital, Room 643, 600 University Avenue, Toronto, On
tario, Canada M5G 1X5.
3 Present address: Cardiothoracic Centre, Thomas Drive, Liverpool, England
L143LB.
4 To whom requests for reprints should be addressed, at Department of
Biochemistry. Biophysics and Genetics, Box B188, University of Colorado Health
Sciences Center, 4200 E. 9th Avenue, Denver, CO 80262.
5The abbreviations used are: ECM, extracellular matrix: Mab, monoclonal
antibody; SLC, squamous lung carcinoma; FITC, fluorescein isothiocyanate.
established that certain of these large glycoconjugates are inti
mately associated with the ability of transformed cells to invade
and metastasize (2-5).
We have developed over 30 cloned variants of a human SLC
of the lung, which affords the opportunity to determine a variety
of cellular and molecular phenotypes for tumorigenic cells (6,
7). This report focuses on clones that either react strongly with
43-9F Mab or do not react detectably. The epitope recognized
by this antibody is an oligosaccharide called Lea-X with the
sequence Gal/33(Fuc«4)GlcNAc/:îGal/34(Fuc«3)
GlcNAc/îGal—
(8), and this epitope is found to be predominantly associated
with large glycoproteins and mucins synthesized by the SLC
cells (6). The two sets of clones are called Lea-X-positive and
Lea-X-negative, respectively. Previously (7) we established the
ability of Lea-X-positive cells to grow and invade when intro
duced s.c. into nude rodents and the failure of Lea-X-negative
clones to grow in this manner. Recently it was demonstrated
that Le"-X is a sensitive marker of carcinoma in situ of human
testes (9) and that the presence or absence of this oligosacchar
ide epitope also has significant prognostic value in human
squamous lung carcinomas.6
Considering the importance (10, 11) of establishing whether
or not orthotopic growth would parallel our previous studies,
we describe in the present report the effects of intrabronchial
aerosol in Rowett nude rats. We also describe how Lea-Xpositive cells elaborate a mucin gel that prevents antibody from
interacting with the underlying cells. It is proposed that the gel
can provide an immunoprotective effect on the SLC cells, and
supporting evidence for this is provided.
MATERIALS AND METHODS
Cells and Cell Growth. Different clones of human squamous lung
carcinoma line RH-SLC-L11 were grown as described (6, 7) and
attached in plastic vessels or, in alternate experiments, grown attached
to Millipore CM or Falcon membranes, using RPMI 1640 plus 5% calf
serum supplemented with iron. Most experiments used the Le'-Xpositive clone NU6-1 or the Le"-X-negative clone NE-18 (7). Over
many passes it has been found that the NU6-1 cell line can be main
tained with a supplement of 5% calf serum, while the NE-18 cell line
grows maximally with 15% supplementation. An iron-calf serum com
bination also allowed the cells to grow at the same rate as 5% or 15%
fetal calf serum. Single cell suspensions were prepared from confluent
monolayers by a 5-10-min incubation at room temperature in 0.05%
trypsin plus 2 HIM EDTA or in 2 HIM EDTA alone as previously
described (6, 7). None of the various trypsin or EDTA concentrations
or the supplementation with calf serum affected the described relative
tumorigenicities.
Monoclonal Antibodies. The 43-9F antibody (IgM) was purified from
serum-free culture medium of the 43-9F hybridoma as described (6, 7).
In some experiments the antibody-containing medium was used without
purification. The monoclonal antibodies were applied for immunoflu"H. Battifora, H. Sorensen, P. Mehta, C. Ahn, J. Niland, E. Hage, D.
Pettijohn, and L. Olsson. Tumor-associated antigen 43-9F is of prognostic value
in squamous cell carcinoma of the lung. A retrospective immunohistochemical
study. Cancer, in press, 1992.
2923
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.
MUCIN GEL OF SQUAMOUS LUNG CARCINOMA CELLS
orescence or immunohistochemistry as previously described (6, 7) to
define the cell surface carbohydrates of the SLC clones and to further
evaluate the mucins which are elaborated by NU6-1. The 2E4 Mab
specific for human NuMA protein was routinely used as described (7)
to distinguish human from rat or mouse cells in tissue or tumor sections.
The Mab against proteoglycan 4S (IgG) was purchased from ICN
Immunobiochemicals and Mab CO-514 recognizing Le" was purchased
from the American Type Culture Collection. Centocor CA 19-9 anti
body specific for sialosyl Le" oligosaccharide was a gift from B. Delvelona (Centocor, Malvern, PA).
Animals. Female Nu/Nu mice were purchased from Charles River
Breeding Laboratories (Kingston, NY) and were maintained on a stand
ard laboratory rodent Purina diet in the reverse isolation animal facil
ities at the University of Colorado Health Sciences Center. Surgical
implantation of the cell lines was carried out s.c. (6, 7) and intracranially
(12) as described under sterile conditions in the surgical suites of the
same facilities. The laboratory animals were followed on a daily basis
for consideration of health status and growth of the heterotransplanted
cells. In those situations where tumors grew rapidly the animals were
sacrificed to conform to animal-care requirements. Male or female nude
(CR:NIH-RNU) rats (National Cancer Institute Research Facility) were
received at 4 weeks of age and acclimated for 1 week before entering
the protocol. Whole-body irradiation was carried out, cells were im
planted intrabronchially, and animals were euthanized as previously
described (13). Tumor development in nude mice and nude rats was
evaluated using described criteria (7, 13).
Fixation. Tumors removed from Nu/Nu mice and Rowett nude rats
were routinely fixed overnight in 10% phosphate-buffered formalin (6,
7). Experiments involving cultured cells utilized either fresh cells or
cells which were fixed in 10% phosphate-buffered formalin (6, 7). Many
fixation methods were evaluated to elucidate the mucin layer overlying
Lea-X-positive clones. When Le'-X-positive cells were grown on plastic,
glass, or transparent membrane surfaces, the observed extracellular
plates indicative of the mucin gel were easily and reproducibly preserved
and analyzed, en fos, using a variety of fixation procedures including
the routine formalin procedure. However, the mucin gel was easily
dissociated from these cells during transverse sectioning. Formalin
fixation under vacuum with membranes wrapped in filter paper ulti
mately resolved this problem.
Microscopy. Heterotransplanted tumors and monolayer cultures
were prepared for incident and fluorescent light microscopy as previ
ously described (6, 7). The microscope was a Leitz Dialux 20 with
UFX-II photographic attachment. Filter cube N for FITC incident light
microscopy was used.
Histology. On a routine basis, hematoxylin and eosin sections were
obtained adjacent to those sections which were evaluated by immuno
histochemistry or immunofluorescence. The hematoxylin and eosin
sections were prepared as previously described (6, 7). Occasionally,
special stains such as periodic acid-Schiff, Alcian blue, and mucicarmine
were used to evaluate glycoconjugate concentrations.
RESULTS
Characteristics of Le"-X-positive and Le'-X-negative Clones.
During culture, Lea-X-positive cells such as clone NU6-1 plate
down more firmly than Lea-X-negative cells and are more
difficult to disperse from the surfaces of culture flasks. More
than 95% of Lea-X-negative clones are released from the sur
faces of flasks in less than 30 s by a 50 ng/m\ trypsin, 3 mM
EDTA solution, while Lea-X-positive clones release slowly over
a 10-12-min period under the same conditions. Moreover,
confluent cultures of NU6-1 cells and other Lea-X-positive
clones elaborate an extensive mucin gel matrix that appears as
a mosaic of plate-like structures (Fig. 1). These structures are
stained by antibodies that have epitopes on mucin, such as Mab
43-9F, having affinity for the Lea-X oligosaccharide. Com
parable structures are not seen in association with NE-18 cells
or other Lea-X-negative clones at any stage of confluency.
Incubation of the NE-18 cells with neuraminidase is required
to expose the Lea-X oligosaccharide and allow reaction with
Fig. 1. Extracellular plate-like structures in mucin gel. A, confluent culture of NU6-1 cells reacted with Mab 43-9F plus FITC conjugated secondary antibody and
visualized by fluorescent microscopy. B, phase microscopy of the same field as in /(, revealing cells underlying the plates. C, confluent culture of NE-18 cells similarly
reacted with Mab 43-9F after incubation with neuraminidase to expose Le'-X epitope (7) and visualized by fluorescent microscopy. D, phase microscopy of the same
field revealing NE-18 cells. Bar. 200 ^m. x 200.
2924
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.
MUCIN GEL OF SQUAMOUS LUNG CARCINOMA CELLS
Mah 43-9F. As shown previously (6, 7) the NE-18 cells are LeaX-negative because Le"-X is sialated. Incubation of Le"-Xpositive cells with neuraminidase has no effect on the appear
ance of the extracellular plates (not shown).
The application of FITC-labeled 43-9F Mab to confluent
Le"-X-positive cells reveals 4+ positivity of many of the plates
(Fig. 1). The plates occur over the apical surface of the Le"-Xpositive cells as they grow on solid surfaces; this is indicated by
the need to adjust the plane of focus of the fluoresceinated
image when focusing the plates relative to the underlying cells.
Adjacent plates seem to be continuous with each other in the
mucin gel, but neighboring plates can be distinguished because
the fluorescence intensity varies from plate to plate. Some of
the plates appear to be negative for Lea-X and exhibit no
previous studies (6, 7) that SLC Lea-X-positive cells have large
amounts of 43-9F epitope on their cell surface, yet the cell
surfaces are not stained by the Mab when the cells are confluent
and have formed plates. Instead, only the plates of the mucin
gel are stained (Figs. 1 and 2). This suggests that the mucin gel
elaborated in the plate-like structures lays down a protective
layer which is not penetrable by the IgM or IgG molecules.
This gel may perform an immunoprotective function similar to
that attributed to mucin secretory products of the apical surface
of many normal epithelial cells. Anti-Lea antibodies (IgG) and
observable fluorescence. Individual plates defined by the criteria
of constant fluorescence seem to cover groups of cells number
ing from 1 to 20 cells/plate. Other antibodies such as Mab 199, with affinity for fucosylated Lea, and Mab CO-514, with
affinity for Le", also stain these plates. Moreover, the mucin
antibodies directed against the 4S proteoglycans (IgG) were
also tested with regard to the above observation. In neither case
did antibodies penetrate the mucin gel of the NU6-1 cells,
although they bind to the NU6-1 cells before they are confluent
and have therefore not yet elaborated a mucin gel layer (results
not shown).
To confirm both the location of the mucin gel relative to the
associated SLC cells and the effect the gel has in excluding
antibodies, transverse sections of confluent Lea-X-positive SLC
layer can be scraped from tissue culture flasks and stained by
the histochemical stains mucicarmine, periodic acid-Schiff, and
Alcian blue (results not shown).
Antibodies Do Not Penetrate the Mucin Gel. It is known from
cells were examined (Fig. 2). When extracellular plates were
stained with 43-9F antibody and then sectioned through a plane
perpendicular to the solid surface on which the SLC cells grew,
the sectioned fluorescent plates were clearly seen at positions
Fig. 2. Visualization of transverse sections of
NU6-1 cells and associated mucin gel. NU6-1 cells
were grown on semipermeable membranes and
were incubated with Mab 43-9F and FITC-conjugated secondary antibody either prior to fixation
and transverse sectioning or afterward. Left, cells
grown on Millipore CM filters, which permits
stratification; right, cells grown on Falcon mem
branes. In A, NU6-1 cells stained with hematoxylin
and eosin reveal intact tumor cells with no degen
erative change. In B, fluorescent microscopy of a
transverse section adjacent to that in A incubated
with Mab 43-9F prior to fixation and sectioning
reveals Le'-X-positive material overlying the clone
but not associated with NU6-1 cells. C, phase
microscopy of the same field as in B. D, fluorescent
microscopy of a transverse section adjacent to that
in A where 43-9F Mab was applied after the sec
tioning, showing that the Mab now binds to the
NU6-1 cells. Bar. 150 ?m. x 200.
2925
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.
MUCIN GEL OF SQUAMOUS LUNG CARCINOMA CELLS
their surface contact with the plastic flasks, the fluorescent
plate-like structures were removed, and over a period of 10 to
30 min the 43-9F antibody with the FITC secondary antibody
shifted to the NU6-1 cell surfaces which had previously not
been stained (Fig. 3). Thus it appears that the disruption of the
mucin gel reorients the binding of the 43-9F Mab to the SLC
cells. When the overlying gel was disrupted by light trypsin
digestion prior to adding the 43-9F antibody, the Mab bound
exclusively to the surface of the remaining attached NU6-1 cells
(results not shown). These experiments confirm that an intact
overlying gel prevents attachment of antibodies to the cell
surface and suggest that parts of the disrupted gel can attach to
the surface of NU6-1 cells.
Le"-X-positive SLC Cells Form Invasive Intrabronchial Tu
mors. Previously we reported that Le'-X-positive clones devel
oped invasive subcutaneous tumors in nude mice while Le"-Xnegative clones failed to develop tumors (7). Some of the LeaX-negative clones were unstable and acquired Le°-X-positive
revenants, which, after long lag periods were able to develop
tumors composed predominantly of Le"-X-positive cells. Clones
NU6-1 and NE-18 are relatively stable in their Le"-X phenotype
and were therefore used in the following studies of orthotopic
implantation into nude rats.
Suspensions of 5 x IO4 to 6 x IO6 cells of both clones were
introduced intrabronchially or s.c. The subcutaneous heterotransplants in both rodent models reproduced previously re
ported experiments. NE-18 and other Le"-X-negative clones
developed only transient subcutaneous nodules, which began
regressing 9 to 20 days after injection. NU6-1 and other Le'X-positive clones developed into invasive tumors (Table 1).
Only a single rat receiving the smallest number of NU6-1 cells
(5 x IO4)failed to develop a subcutaneous tumor. Intrabronchial
tumor development was followed via chest roentgenographic
studies as described (13), and after 6-8 weeks the rats were
sacrificed and necropsies were performed (e.g., see Figs. 4 and
5). Results are also summarized in Table 1. All surviving rats
receiving NU6-1 cells had invasive intrabronchial tumors, while
none of the NE-18-injected animals revealed gross tumor
growth. As described below in more detail, microscopic clusters
of NE-18 cells could be identified within the bronchi at their
sites of injection, but these cells had not invaded the bronchial
wall. The identity of the human NE-18 and NU6-1 cells in
frozen sections of rat lung tissue was confirmed using the
Fig. 3. Dispersion of mucin gel with trypsin. Confluent cultures of live NU61 cells were incubated with Mah 43-9F and FITC-conjugated secondary antibody
as described above. While viewing these cells under an inverted microscope, the
cells were treated with 0.05% trypsin at time 0. In A, immediately after trypsinization plate-like structures are observed overlying cells, as noted in Fig. 2. B, 2
min after application of trypsin. Plate-like structures begin to regress. In C, 10
min after application of trypsin the mucin gel is almost completely removed, and
SLC cell surfaces are stained. Many of the stained cells have started to round up,
as surface attachments are altered. Bar, 200 um. x 200.
Table 1 Correlation between Le°-Xoligosaccharide marker, formation of
extracellular plates, and tumorigenicity in nude rodents
bronchial/s.c.++++
Le'-X*
plate formation''YesYesYesYesNoNoN
Clone"NU6-1NU6-3NU6-4NU6-7NE-1NE-13NE-6NE-18Tumorigenicity'
11/11++++
ND'++++
ND++++
NDNDNDND0/9'19/226/66/66/60/24ND0/70/19Extracellular
over the confluent cells. In agreement with the prior study none
of the underlying cells were stained. However, when the trans
verse sections were made before applying the antibody, the
exposed underlying cells were readily stained. This result con
firms that the Lea-X oligosaccharide is available for interaction
with the 43-9F antibody on the cell surfaces under the mucin
gel, but the antibody cannot penetrate the gel.
In further experiments immunofluorescence was performed
on living confluent NU6-1 cultures. While viewing the living
cells under an inverted microscope, 0.05% trypsin was applied.
During digestion, as the NU6-1 cells rounded up and reduced
"Different
clones of SIX1 I.I I cells were implanted bronchiali}' or s.o. as
previously described (7, 12).
* Mab 43-9F was applied to paraffin sections as previously described (6, 7) to
identify Le'-X-positive or -negative fields. —¿,
not detected.
' Gross observation of tumor using previously described criteria (7, 13).
'' Samples of cells prior to injection or from tumors in experimental animal
were cultured, and the presence of extracellular plates was evaluated as described
in Fig. 2.
' ND, not done.
'Microscopie nodules were located within the lungs of S of 5 rats bronchiali}'
implanted with NE-18 cells and histologically examined (Fig. 6).
2926
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.
MUCIN GEL OF SQUAMOUS LUNG CARCINOMA CELLS
Fig. 4. Morphology of NU6-1 bronchial tumors in Rowett nude rats. Hematoxylin and eosin sections of tumor:. I,
observation at the growth front, revealing no appreciable
immune cell infiltrate. B, organoid growth pattern with
numerous mitotic figures and nuclear material dispersed in
a vesicular pattern. Bar, 320 *im. x 100.
human-specific Mah 2E4 as previously described (7). Thus,
these studies of orthotopic propagation of human SLC cells
extend the previous findings and show that only the Lea-Xpositive cells form progressive tumors in the lung. It should
also be emphasized that there is a correlation between the
formation of mucin gel and tumorigenicity in the nude rodents.
Only the Le"-X-positive clones that form observable extracel
lular plates are tumorigenic (Table 1).
Intracranial Growth of Heterotransplanted SLC Cells. Holer
otransplantation of tumor cells into immunoprivileged sites has
aided the study of cell lines which have low tumorigenicity even
in immunodeficient animals (see, e.g., Ref. 11). We examined
tumor development by the NU6-1 and NE-18 SLC clones in
the brain of nude mice to investigate the possible role of the
host immune system in rejecting the NE-18 tumors. Eleven
Nu/Nu mice were injected intraerán¡allywith 6 x IO6cells, and
the mice were observed daily for neurological changes and after
8 weeks were sacrificed. Whole brains were fixed in buffered
formalin, the tissue was sectioned, the cryosections were stained
with 2E4 Mab to identify human cells, and paraffin sections
were stained with the standard histochemical stains and Mab
43-9F to further characterize tumors (see "Materials and Meth
ods"). Tumors developed in all 11 nude mice at similar rates, 4
of 4 for those receiving NU6-1 cells and 7 of 7 for those
receiving NE-18 cells (see Fig. 6 for examples). The cells of the
NU6-1 tumors retained their strong Lea-X positivity, and each
NE-18 tumor was uniformly Le"-X-negative. Thus in this priv
ileged site tumor development of the Le"-X-positive line paral
lels that of the Le"-X-negative line.
Tumor Histology. Of interest is the difference seen in the
heterotransplanted SLC Le°-X-positive and -negative cells as
tumors develop and the human cells interact with host tissues.
The patterns seen with each of the Le"-X-positive clones did
not differ when either the location of implantation or the mouse
or rat host was changed. NU6-1 cells, at the light microscopic
level, examined after routine hematoxylin and eosin staining,
form an organoid pattern. Islands of tumor cells proliferate
between ribbons and festoons of dense fibrous connective tissue
(Figs. 4 and 6). Examination at low power suggests a monoto
nous cell population, but at higher power numerous mitotic
figures are seen. Many cells are quite pleomorphic, with irreg
ular nuclear membranes. The nuclear material is dispersed in a
vesicular pattern. In contrast to the situation with NE-18 cells
described below, no appreciable lymphocytic or plasma cell
infiltrate is noted (Fig. 4). NE-18 cells consistently simulate the
usual morphology seen with well-differentiated squamous cell
2927
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.
MUCIN GEL OF SQUAMOUS LUNG CARCINOMA CELLS
Fig. 5. Morphology of NE-18 bronchial micronodules in
Rowett nude rats. Hematoxylin and eosin section. In A, a
single micronodule of NE-18 cells is seen in the center of
the field, confined within a bronchiole lumen. Numerous
polymorphonuclear leukocytes and plasma cells, as well as
occasional lymphocytes, are present within the micronodule
with increased concentration at the periphery (examples of
inflammatory cells within the microncdule are indicated by
thin arrows and by inset, lower left; horizontal black line of
inset, 450 firn) and circumscribe the bronchiole at the site of
the micronodule (concentrations indicated by broad arrows).
Bands of fibrous tissue also circumscribe the bronchiole.
Horizontal black lines of small inset box, 400 pm. x 40. B,
magnification of a section of the same micronodule showing
squamous cell differentiation. Individual cells contain orangophilic cytoplasm and hyperchromatic, irregular nuclei.
Intercellular zona occludens are prominent. Bar, 320 urn. x
100.
*v "<..*"F£
,^
.***
*/" . ï
P "m «fe -
A.f-;
. *
jBr
.. ^wl^'
¿S^ ^.-¥
[?¡*'<
'
y
fi-,
"T
.
î'«.
^';V*XÕ
'
B
•¿
v«*carcinomas of the lung. Irregular large cells are interspersed
with small cells with hard eosinophilic cytoplasm. Nuclear
membranes are irregular, and nuclei are hyperchromatic. Mitotic figures abound. Easily noted on high power are the zona
occludens between cells. Associated with the NE-18 tumors are
thick bands of dense fibrous connective tissue. These bands
usually circumscribe the intrabronchial and subcutaneous micronodules, and within the fibrous tissue are numerous inflam
matory cells (Fig. 5). Two pathologists independently reviewed
these sections. They were in agreement that the majority of
cells are plasma cells. However, polymorphonuclear leukocytes
are not uncommon. The small lymphocyte is rarely noted (Fig.
5). In neither NU6-1- or NE-18-induced tumors is appreciable
necrosis present. Tumors within brain tissue did not mount an
inflammatory response with either NU6-1 or NE-18 cell lines
(Fig. 6).
DISCUSSION
Malignant tumors have been characterized as having atypical
cell surface glycoproteins, altered extracellular matrix, and cellcell adhesion properties which are abnormal (14-16). In hu
mans these changes have been verified in breast, colon, and
pancreatic carcinomas (17-19) as well as in malignant mela
nomas (20). Monoclonal antibody technology has played a vital
role in elucidating aberrant carbohydrate synthesis of associated
mucin molecules (21) as well as atypia of extracellular matrix
protein and proteoglycan composition (22).
In the present study we utilized variant clones of human SLC
cells to explore differences that correlate with the ability of the
cells to grow autonomously and invade neighboring tissues. The
results confirmed our earlier observations (6, 7) of the growth
and development of Le"-X-positive and -negative cell lines when
implanted s.c. into nude mice. The findings were extended to
show that in the lung only the Lea-X-positive cell lines progress
into invasive tumors. The presence of large numbers of cells of
immune origin associated with the regressing NE-18 microtumors but not with the progressing NU6-1 tumors suggested
that the host immune system may play a role in the failure of
the Le"-X-negative clones to form progressing tumors. This
possibility was supported by the finding that the NE-18 cells
develop tumors at the same rate as the NU6-1 clones in the
nude mouse brain, where presumably the NE-18 cells are pro
tected from the immune system of the host. Results from the
intracranial injections suggest that the B-cell system may play
a vital role in working in association with ECM to limit cells
2928
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.
MUCIN GEL OF SQUAMOUS LUNG CARCINOMA CELLS
Fig. 6. Morphology of inlracranial tumors in nu tin
mice. Hematoxylin and eosin sections. A, tumor developed
by NU6-1 cells which is similar to those seen in subcuta
neous and pulmonary lesions. An organoid/glandular pat
tern is present with numerous mitotic figures and vesicular
nuclei. B, tumor developed by NE-18 cells, suggesting squamous differentiation, as seen in residual subcutaneous and
pulmonary lesions. However, reorganization of adjacent
host tissue is not present, nor are plasma cells, lymphocytes,
or pol\ mm illuni urli-m leukocytes. Bar, 320 um. X 100.
which are Le"-X-negative. It is also important to recall that the
extracellular matrix within brain tissue is different from matrix
which can be elaborated in extracranial sites. Whichever ele
ments are determined to be critical at the human tumor-host
matrix level, the interactions are different intracranially. Both
the immunodeficient mice and rats mount responses within the
subcutis and within the bronchial tree which do not allow for
growth and eventual invasion by Le"-X-negative cells. Because
these immunodeficient rodents contain B-cells but are deficient
in mature T-cells (15), the failure of NE-18 to form large tumors
extracranially correlates with B-cell function and/or T-cellinitiated B-cell response, particularly with regard to natural
killer cell and/or macrophage-mediated tumor cell lysis.
The demonstration of a geometric plate-like matrix overlying
Le"-X-positive cells in culture may indicate one way by which
these cells evade host immune surveillance. When the mucin
gel overlies the Lea-X-positive cells, antibody cannot penetrate
the mucin layer to interact with receptors which are known to
be present on NU6-1 cell surface membranes. Therefore this
mucinous material forms a protective barrier similar to that
which is normally present within normal airway lumina (23,
24). How the mucin gel allows cells to escape detection must
be rigorously evaluated in future studies. Mechanisms which
have been suggested are modification of cell membrane recep
tors (18); alteration of the extracellular matrix itself, such that
cytolytic immune cells cannot "reach" these tumorigenic cells
(19); or an alteration in soluble signals which either promote
or inhibit B-cell activation (20-22). Furthermore, recent evi
dence suggests that cell-to-cell surface interactions also alter Bcell function via B-cell surface molecules (25, 26).
The results presented here expand upon our earlier studies
and further demonstrate that variant cell lines expressing var
ious tumorigenic phenotypes are useful in elucidating cellular
and molecular mechanisms which underlie malignant transfor
mation. The utilization of immunodeficient rodents allows for
in vivo evaluation as well as confirmation of in vitro observations
regarding the possible role of the Le"-X oligosaccharide se
quence in the evolution of non-small cell carcinomas of the
lung.
ACKNOWLEDGMENTS
We wish to thank Mark Overland, Joseph Stranahan, Terry Marceli,
Audrey Handsel, Janet Hattenbach, and David Davis for their excellent
technical assistance. The clerical assistance of Phyllis Schroeder and
Elsie Vacano was invaluable.
2929
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.
MUCIN GEL OF SQUAMOUS LUNG CARCINOMA CELLS
REFERENCES
1. lozzo. K. Biology of disease. Lab. Invest., 53: 373-396, 1985.
2. Esko, J. D., Rostand. K. S., and Weinke, J. L. Tumor formation dependent
on proteoglycan biosynthesis. Science (Washington DC), 241: 1092-1096,
1988.
3. Irimura, T.. Carlson, D., and Ota, D. Electrophoretic analysis of four high
molecular weight sialoglycoproteins produced by metastatic human colon
carcinoma cells. J. Cell Biol., 37: 1-9, 1988.
4. Schirrmacher, V., Altevogt, P., and Lang, E. Adhesive variants of a highly
metastatic tumor show distinct changes in expression and glycosylation of
cell surface molecules: association with observed changes in organotropism
and malignancy? In: C. L. Reading, S. Hakamori, and D. M. Marcus (eds.).
Altered Glycosylation in Tumor Cells, 79, UCLA Symposia on Molecular
and Cellular Biology, pp. 251-256. New York: Alan R. Liss, Inc., 1988.
5. Dennis. J.. and Laferte, S. Asn-linked oligosaccharides and the metastatic
phenotype. In: C. L. Reading. S. Hakamori, and D. M. Marcus (eds.). Altered
Glycosylation in Tumor Cells, 79, UCLA Symposia on Molecular and
Cellular Biology, pp. 256-267. New York: Alan R. Liss, Inc.. 1988.
6. Pettijohn, D. E., Stranahan, P. L., Due, C., Ronne, E., Sorensen, R., and
Olsson, L. Glycoproteins distinguishing non-small cell from small cell human
lung carcinoma recognized by monoclonal antibody 43-9F. Cancer Res., 47:
1161-1169,1987.
7. Pettijohn. D. E., Pfenninger. O.. Brown, J., Duke, R., and Olsson, L.
Tumorigenic human squamous lung cancer cells have defined cell surface
carbohydrates that are absent from non tumorigenic cells. Proc. Nati. Acad.
Sci., 85: 802-806, 1988.
8. Martensson. S., Due, C., Pahlsson, P., Nilsson. B., Eriksson, H., Zopf, D.,
Olsson. L., and Lundblad. A. A carbohydrate epitope associated with human
squamous lung cancer. Cancer Res., 48: 2125-2131, 1988.
9. Giwercman, A., Lindenberg, S., Kimber, S., Andersson, T., Muller, J., and
Skakkebaek, N. Monoclonal antibody 43-9F as a sensitive immunohistochemical marker of carcinoma in situ of human testis. Cancer (Phila.), 65:
1135-1142, 1990.
10. Fidler, I. J. Rationale and methods for the use of nude mice to study the
biology and therapy of human cancer metastasis. Cancer Metastasis Rev., 5:
29-49. 1986.
11. Mclemore, T. L., Liu, M., Blacker, P., Gregg, M., Alley, M., Abbott, B. J.,
and Shoemaker. R. A novel intrapulmonary model for the orthotopic prop
agation of human lung cancers in athymic nude mice. Cancer Res. 47: 51325140, 1987.
12. Epstein, A. L.. Herman. M. M., Kim, H., Dorfman. R., Path, M., and
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
Kaplan, H. Biology of human malignant lymphomas. III. Intracranial heterotransplantation in the nude, athymic mouse. Cancer (Phila.), 37: 21582176, 1976.
Howard, R., Chu, H., Zeligman, B., Marceli, T., Bunn, P., Mclemore, T.,
Mulvin, D., Cowen, M., and Johnston, M. Irradiated nude rat model for
orthotopic human lung cancers. Cancer Res., 51: 3274-3280,1991.
Turley, E. Proteoglycans and cell adhesion: their putative role during tumorigenesis. Cancer Metastasis Rev., 3: 325-339, 1984.
Raz, A., and Ben-Ze'ev, A. Cell contact and architecture of malignant cells
and their relationship to metastasis. Cancer Metastasis Rev., 6: 3-21, 1987.
Cella, M. Intercellular glycosaminoglycans in human cancer. Tumori, 65:
677-686, 1979.
Burchell, J., Durbin, H., and Taylor-Papadamitriou, J. Complexity of expres
sion of antigenic determinants recognized by monoclonal antibodies HMFG1 and HMFG-2 in normal and mammary malignant epithelial cells. J.
Immunol., 131: 508-513, 1983.
Sorokin, J., Sugarbaker, P., and Zamcheck, N. Serial carcinoembryonic
antigen assays: use in detection of cancer recurrence. JAMA, 228: 49-53,
1974.
Pleskow, D., Berger, H„and Gyves, J. Evaluation of a serologie marker,
CAI 9-9, in the diagnosis of pancreatic cancer. Ann. Intern. Med., 110: 704709. 1989.
Harper, J., and Reisfeld, R. Cell-associated proteoglycans in human malig
nant melanoma. In: T. Wright and R. Mecham (eds.). Biology of Proteogly
cans, pp. 345-366. New York: Academic Press, 1987.
Dillman, R. Monoclonal antibodies for treating cancers. Ann. Intern. Med.,
///: 592-603, 1989.
Raz, A., Zoller, M., and Ben-Ze'ev, A. Cell configuration and adhesive
properties of metastasizing and non-metastasizing BSp73 rat adenocarcinoma cells. Exp. Cell Res., 162:127-141, 1986.
Kühn,C. Normal anatomy and histology. In: W. Thurlbeck (ed.). Pathology
of the Lung, pp. 11-50. New York: Thieme Medical Publishers. 1988.
Gallagher, J. T., and Richardson, P. S. Respiratory mucus: structure, mettabolism and control of secretion. In: E. Chantier, J. Elder, and M. Elstein
(eds.), Mucus in health and disease. Adv. Exp. Med. Biol., 335-351, 1982.
Clark, E. Structure, function and genetics of human B cell associated surface
molecules. Adv. Cancer Res., 52: 81-149, 1989.
Pohajdak, B., Lee, K., Sugawara, I., Miller, V., Wright, J., and Greenberg,
A. Comparative analysis of natural killer cell and macrophage recognition of
con A-resistant CHOC: role of membrane oligosaccharides. J. Nati. Cancer
Inst., 76:257-267,1986.
2930
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.
Mucin Gel Formed by Tumorigenic Squamous Lung Carcinoma
Cells Has Le a-X Oligosaccharides and Excludes Antibodies
from Underlying Cells
Patricia L. Stranahan, Randy B. Howard, Oswald Pfenninger, et al.
Cancer Res 1992;52:2923-2930.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/52/10/2923
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.