[CANCER RESEARCH 54, 13X1-1387. March 1. 1W4)
Expression of CD44 in Human Lung liimors1
Margaret B. Penno,2 J. Thomas August, Stephen B. Baylin, Mack Mabry, R. Ilona I,¡minila,Valerie S. Lee,
IK IHMali C'roleau,'
Xiuu Ling Yung, and Cecilia
Rosada
Dfixinmfiils of Medicine /M. B. I'., V. S. L., S. B. B.j ami I'liarmamlogy ami Molecular Sciences ¡J.T. A., I). C, X. L. Y., C. R./, Johns Hopkins University School of Medicine.
and The Oncology Center /S. B. B., M. M./, Baltimore, Maryland 21205, and Biomarkers and Prevention Research Branch, National Cancer Institute, NIH,
Rockvilh: Maryland 2ÕW.W/«. /. /../
on a number of biological and clinical observations, lung cancers can
be divided into two main groups. SCLCs compose approximately
one-fourth of the cases, express neuroendocrine markers, and are
defined as "classic" by the high level of endocrine differentiation
ABSTRACT
CD44 is an integral membrane glycoprotein that functions as a receptor
for the extracellular matrix glycan, hyaluronan. Here we report that (1)44
is a novel biomarker for non-small cell lung tumors, squamous metaplasia
of the lung, and activated type II pneumocytes. We have examined the
expression of CD44 in 12 human lung tumor cell lines and 23 fixed,
paraffin-embedded lung cancers. CD44 transcription and translation is
consistently high among non-small cell tumors (5 of 5 cell lines, 10 of 14
tumors) but rare in small cell tumors (1 of 6 cell lines, 0 of 9 tumors). In
normal lung, (1)44 was confined to the surface of bronchial basal cells and
alveolar macrophages. Squamous metaplasia of the lung showed strong
CD44 immunoreactivity. Resting type II pneumocytes were largely CD44
negative but rows of active, surfactant-secreting type II cells had signifi
cant amounts of CD44 located on lateral surfaces of adjacent cells. The
correlation between (1)44 and the non-small cell phenotype was further
demonstrated in studies of a cultured small cell lung cancer line induced
to exhibit characteristics of a non-small lung cancer by infection with
v-Ha-ras. Following ras gene insertion, these cells showed a 40-fold in
crease in CD44 expression. The CD44 detected in lung cancer cells
throughout these studies was predominantly the "standard" rather than
the "variant" species. Taken together, these results suggest that CD44 is a
protein expressed on non-small cell lung tumors, squamous metaplasia,
and activated type II cells. In addition, CD44 in cultured small cell lung
cancer cells is transcriptionally activated following differentiation by the
rif, oncogene. The fact that immunohistochemistry
can be used to dis
criminate among the cell types makes CD44 a valuable new marker for
lung neoplasia.
INTRODUCTION
CD444 is an integral membrane glycoprotein (hat functions as a
receptor for the extracellular matrix glycan, hyaluronan ( 1-6). CD44
is expressed on the surface of specific glandular, fibroblastoid, and
hematopoietic cells and is up-regulated following activation of lym
phocytes and monocytes (7). In tumor cells, two general alterations of
CD44 expression have been observed: (a) the enhanced transcription
of an alternately spliced, "variant" CD44 mRNA (CD44v) which
encodes a protein 10 or more kilodaltons larger than the 85-90-kDa
"standard" (CD44s) product (2, 8, 9); (b) the loss of standard CD44
by repression of transcription (10). Although quantitative and quali
tative changes in CD44 can be used to discriminate primary tumors
from corresponding normal tissue in a number of systems, there is no
consensus as to the role of CD44 in the progression of tumors.
Lung neoplasms represent an ideal opportunity to match the ex
pression of CD44 with well characterized parameters associated with
cellular activation, differentiation, and metastatic predilection. Based
Received 9/30/93; accepted 12/28/93.
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.
1This work was supported by NIH Grants CA58184 and CA09243 and American
Cancer Society Grant PDT430.
- To whom requests for reprints should be addressed.
3 D. C. was supported by National Cancer Institute Training Grant T32 CA 09243.
4 The abbreviations used are: CD44s, standard hyaluronic acid binding form of CD44
(also known as CD44H); CD44v. CD44 with any of the alternate splice variants; SCLC.
small cell lung cancer; NSCLC, non-small cell lung cancer; FACS. microfluorimetry;
PCR, polymcrase chain reaction; NCAM. neural cell adhesion molecule; PBS. phosphatebuffered saline; cDNA. complementary DNA; nt. nucleotide.
molecules including i.-dopa decarboxylase, neuron-specific enolase,
creatine kinase-BB isoenzyme, and gastrin-releasing peptide. "Vari
ant" SCLC has variable loss of endocrine differentiation with reten
tion of creatine kinase-BB isoenzyme (11). N-wyc and L-myc overexpression in classic SCLCs and c-myc overexpression in variant
forms are relatively frequent. SCLC generally metastasize early to
lymph nodes, brain, bones, lung, and liver. NSCLCs comprise the
majority of the remaining lung tumor types and include adenocarcinoma, squamous cell carcinoma, and large cell carcinoma. NSCLCs,
characterized by epithelial-like growth factors and receptors, are often
locally invasive (12, 13). Mutations in the Ki-ras oncogene, predomi
nantly at codon 12, have been found in a subset of NSCLCs but not
SCLCs (14—16).It has recently been shown that v-Ha-ras infection of
SCLCs in coordination with amplification of the N-wyc or c-myc
genes induces a number of phenotypic characteristics typical of
NSCLCs (17-20).
In the present study, we have examined CD44 expression in normal
lung tissue, SCLC and NSCLC tumors, and ras-transduced SCLCs.
Our results demonstrate that CD44 is associated with activated type II
pneumocytes, squamous metaplasia, and the NSCLCs. In addition to
its value as a new marker for lung neoplasia, CD44 expression may
elucidate the cell lineages and stages involved in lung tumorigenesis.
MATERIALS
AND METHODS
Overview. CD44 transcription, splicing, and translation were measured in
cells from seven established lines of SCLC and five NSCLC lines, including
squamous cell, udeno-, and large cell carcinomas. Cells were tested for CD44
expression by flow cytometry (FACS), Western blot. Northern blot, and PCR
analyses. Also, 23 solid lung tumors which were formalin fixed and paraffin
embedded were subjected to CD44-specific immunohistochemistry. For these
experiments, antibodies and probes recognizing epitopes in the CD44 constant
region were chosen to identify both standard and variant forms of CD44.
Cell Culture and Antibodies. The human lung carcinoma cell cultures
used in this study include the NSCLC lines NCI-H460 and NCI-H157 (large
cell carcinomas), lines NCI-HI25, and A549 (adenocarcinoma), and SCLC
lines NCI-H249, NCI-H69, NCI-H82, NCI-H209, and NCI-HI28 (12). SCLC
line NCI-H249 was clonally derived from line NCI-H69.5 Two SCLC lines,
DMS53 and OH3, and one squamous cell line, U1752, were established in
dependently as described (21-23). All NSCLC lines grew as adherent cultures
in RPMI 1640 supplemented with 10% (v/v) heat-inactivated fetal calf serum
(Gibco, Grand Island, NY). All of the SCLC lines grow in RPMI/10% fetal calf
serum as suspension cultures except DMS53, which adheres to the flask
surface in this medium. Cultured cells within a given type were in the same
passage for FACS and Western blot experiments. Dislodging adherent cells
from tissue culture flasks with brief exposure to trypsin-EDTA did not affect
experimental results.
Anti-CD44 monoclonal antibodies, used in the form of hybridoma supernatants, were derived from hybridoma line H4C4 (24). The polypcptide epitope
recognized by this antibody is present in both CD44s and CD44v molecules.
Polyclonal antibodies to human CD44 were raised in rabbits following injec5 A. F. Gazdar. personal communication.
1381
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1994 American Association for Cancer Research.
CD44 AND
LUNG CANCER
Table 1 Expression of CD44 on the surface of cultured SCLC and NSCLC tumors
Relative amounts of CD44 on the surface of cells from 12 lung carcinomas were determined by flow microfluorometry. Cells (2 X 10s) were mixed with antibody H4C4, then stained
with fluorescein isothiocyanate-goat anti-mouse serum and analyzed (5000 cells/point) on an automated flow cytometer. The fluorescence of cells treated with all but primary antibody
was used to set the negative gale. The percentage of cells expressing CD44 is indicated. Mean ±SD was calculated from Iriplicale samples repeated on 3 separate days.
SubtypeNon-small
Type
cell
SquamousLarge
CellAdenoLarge
1752H460A549H157H
cellAdenoSmall
125H249fH69DMS53OH3H82H209H128Mean90.0091.7995.6295.6898.9377.6660.903.9915.230.520.690.94CD44»(%)(1)"(3.
cell
ClassicClassicClassicClassicVariantClassicClassicDesignationU
" Fluorescent cells. Mean of three trials in triplicate.
h 1, log fluorescence intensity.
' H249 is a derivative of H69. Both have N-myr amplification.
tion of affinity-purified
CD44 (25). Anti-NCAM
antibody (NCL-CD56;
No-
vocastra Labs, Newcastle upon Tyne, United Kingdom) was diluted in accord
with the manufacturer's recommendation. Fluoresceinylated goat anti-mouse
antibody [20 ¿ti/mi;F(ab')2; Boehringer Mannheim, Indianapolis, IN] was
used to detect primary murine monoclonals.
Flow Microfluorimetry. Cells were washed in PBS, triturated to disrupt
aggregates, and suspended in 10% normal goat serum in PBS at a final
concentration of 2 X IO6 cells/ml. One hundred (j.1of cells were mixed with
100 p.1 of antibody (H4C4 hybridoma culture supernatant or RPMI) and
incubated for 45 min on ice according to published procedures (26). Cells were
washed with 2 ml cold PBS/0.1% bovine serum albumin/0.1% sodium azide
and then stained with 100 /xl fluorescein isothiocyanate-goat anti-mouse serum
(Fab fragment; Boehringer Mannheim) at 20 ftg/ml in 10% normal goat serum
in PBS. After 45 min on ice, cells were washed with 2 ml cold PBS/0.1%
bovine serum albumin/0.1% sodium azide and resuspended in 0.5 ml PBS and
analyzed (5000 cells/run) on an automated flow cytometer (FAScan; Beckman
Instruments, Inc., Fullerton, CA). The fluorescence of cells treated with all but
primary antibody was used to set the negative gate. Means and standard
deviations of CD44 surface fluorescence values were calculated from triplicate
samples repealed on three separate days. Average fluorescence intensity is
expressed as log intensity.
Western Blot Analysis. Cells (5 x 10'') were washed twice in Ix PBS,
lysed in 50 /xl of buffer (50 mm Tris-HCl, pH 8-150 IHMNaCl-0.2% sodium
a/ide-KM) /xg/ml phenylmclhylsullonyl
fluoride-1 /-ig/ml aprotinin-1% Triton
X-100), and incubated for 30 min on ice. The protein concentration was
determined for each lysale using a commercial kit (Bio-Rad, Melville, NY).
Protein clectrophoresis was performed using 9% acrylamide running and 3%
acrylamide stacking gels. Protein (10 ng) was loaded into each well after
reduction with a ß-mercaptoclhanol-containing sample buffer at 100°C.Pro
tein was transferred via electrophoretic
transfer cell (Bio-Rad) onto a polyvinyl
dilluoridc membrane [Immohilon; Millipore, Bedford, MA (27)]. Membranes
were blocked with 5%' dry milk in PBS at 0°Cfor 18 h and then incubated for
l h (room temperature) with rabbit polyclonal antibodies to CD44 (1:10(10
PBS/1% dry milk). Following a wash in 1% dry milk/0.1% Twecn 20 in PBS,
the blot was incubated (I h) with horseradish peroxidase-IgG anti-rabbit (en
hanced chemiluminescence kit; Amersham, Arlington Heights, IL), washed,
and detected according to the manufacturer's instructions (Amersham).
Northern Blot Analysis. mRNA was isolated from cells using a commer
cial oligodeoxythymidylate
kit (FastTrack version 3.2; Invitrogen, San Diego,
CA). mRNA (5 fig) was subjected to electrophoresis and transferred onto a
nylon membrane [Genescreen Plus; Dupont, Wilmington, DE (28)]. The oligolabeled 32P-CD44 cDNA probe contained exons 1-15 and was isolated from a
human breast cancer line (ZR751'; Boehringer Mannheim). Following hybrid
ization and washing autoradiography was performed (20 h). The filters were
then stripped and rehybridized with a l:P-labeled ß-actinfragment.
* J. T. August, unpublished data.
PCR Analysis. Total RNA, prepared from cultured cells by standard meth
ods (29), was reverse transcribed using specific 3' antisense primers for CD44
(39-mer) or actin (21-mer), respectively. The cDNA (5 /xg) was then amplified
for 30 cycles (90°Cfor 1 min, 55 °Cfor 1 min, and 72°Cfor 2 min) in a
thermocycler (Ericomp, San Diego, CA). The 21-base pair 5' sense and 3'
antisense primers for actin generate a 228-base pair fragment (30). CD44s,
from exon 2 (nt 114) to exon 17 (nt 982) was amplified as a 900-base pair
fragment. The sequences of the sense and antisense CD44 primers were CGTCAGCTCCGAGAAGCTTCCATGGACAAGTTTTGGTGGC
and GGACTATCTAGAGAATTCGACTGCAATGCAAACTGCCAAG
respectively. Aliquots of the PCR reaction were electrophoresed in 2% agarose (Nusieve; FMC,
New Brunswick, NJ), passively transferred to nylon filters (Genescreen Plus),
and fixed by UV cross-linking. Blots were probed with an internal digoxygeninlabeled CD44 probe (nt 934—954,representing the transmembrane region) and
detected according to manufacturer's instructions (Boehringer Mannheim).
Immunohistochemistry.
Immunohistochemical
staining used the avidin-
biotinylated peroxidase complex technique (Vector Laboratories, Burlingame,
CA), with previously reported modifications (31, 32). Incubations with rabbit
polyclonal CD44 antibodies were performed at 4°Cfor 18 h. A 1:1000 dilution
was considered optimal following tllration experiments using a panel of control
tumors. The same positive control tumors were included in each assay. Nega
tive controls for immunoreactivity consisted of the serial lung sections treated
with either nonimnume rabbit serum or phosphate-buffered saline.
Formalin-fixed, paraffin-embedded specimens were obtained from the sur
gical pathology files of the Department of Pathology, Bethesda Naval Hospital.
Only well fixed, surgically removed material adequate for providing multiple
serial sections for immunohistochemical studies was used. Fivc-fim sections
were cut and mounted on gelatin-coated slides. Hemaloxylin and eosin sections
were reviewed lo classify each specimen. The immunohistochemical staining
results were scored for the percentage of tumor cells positive (0; <!()%;
10-50%; >50%) and for the intensity of staining (0 = no staining; weak;
moderate; strong). Any number of tumor cells positive for CD44 were called
a positive tumor.
Viral Infection. Viral infection was done as described previously (18).
Cells were tested for CD44 surface expression after 12 h, 24 h, and 4 weeks.
Expression was greatest at 4 weeks. Cells infected with the helper virus alone
were tested after 4 weeks.
RESULTS
Expression of CD44 on the Surface of Cultured SCLC and
NSCLC Tumors. Relative amounts of CD44 on the surface of cells
from 12 lung carcinomas were determined by flow microfluorometry
and are summarized in Table 1. Five of the cell lines, all in the NSCLC
category, bind high levels of CD44-specific antibodies. Ninety % or
more of the cells in these lines have a fluorescence intensity at least
10 times greater than background (same cells treated with all but
1382
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1994 American Association for Cancer Research.
CD44 AND LUNG
CANCER
primary antibody). In contrast, 5 cultured lines of the SCLC phenotype have low CD44 expression (<15% of the cells were positive), and
one SCLC line, NCI-H69, and its clonal derivative, NCI-H249, are
intermediate (50-80% of the cells were positive). Signal intensity
generally ranged between IO3 and IO4 for NSCLCs. Fluorescence
intensity of most SCLCs was very low (IO1). Although SCLC line
H249 and its clone H69 were positive for CD44, fluorescence inten
sity was an order of magnitude lower than that of most NSCLCs.
To confirm that the differences in CD44 expression in NSCLC did
not reflect some general overexpression of all proteins, another sur
face protein, NCAM, was measured by immunofluorescence in the
same way. The percentage of cells that expressed this marker was, in
general, low in NSCLCs (U1752, 3%; H460, 9%; A549, 2%; H157,
3%; and H125, 3%) and moderately high but variable in SCLCs
(H249, 67%; H69, 32%; DMS 53, 6%; OH3, 13%; H82, 36%; H209,
4%; and H128, 1%). Thus, the expression of CD44 and NCAM are
entirely different and in some cases inversely related in the cultured
human lung carcinomas examined here.
The low CD44 expression in SCLC lines was generally associated
with a highly aggregated, non-anchorage-dependent growth pattern. In
order to determine whether substrate adherence was a factor in CD44
expression, we prevented the cells from adhering to the flask in a
spinner culture. Within 4 days, the percentage of CD44-positive
NSCLCs growing in suspension was markedly reduced in all five lines
examined (U1752, 72%, J, 18%; H460, 72%, | 20%; A549, 52%,
i 44%; H157, 50%, 1 46%; and H125, 72%, | 26%) compared to
Non-Small
900 bp
Fig. 2. PCR amplification and detection of potential CD44 splice variants. Sequences
spanning the splice region were PCR amplified, subjected to Southern blot analysis, and
detected with an internal digoxygenin-Iabeled probe as described in "Materials and Meth
ods." Lane 1, CD44s-positive control (human monocyte line U937); Lane 2, CD44v-
Cell
positive control (endothelial cell line HMEV); Lane 3, SCLC line H249; Lane 4, SCLC
line H82; Lane 5, NSCLC line H157. The 900-base pair (bp) PCR product identified in
all three lines is indicative of CD44s. Insel, equivalent actin amplification for each sample.
12345
Small Cell
8
9
10
the same cells grown as adhesion cultures (compared to values in
Table 1). In contrast, the expression of NCAM in cultured NSCLCs
remained unchanged when cells were grown in suspension with the
exception of line A549 in which NCAM increased from 2% to 28%.
No differences in cell viability were evident during the suspension
experiment.
Total Cellular CD44 Content Correlates with Surface Levels.
Since six of the seven CD44-negative grew as aggregates, our finding
that CD44 was absent from their cell surface could have been ex
plained by an inaccessibility of the antibody for the antigen. To rule
out this possibility, Western blots of total cellular proteins were pre
pared and screened with anti-CD44 polyclonal antibodies. These
blots, shown in Fig. 1, demonstrate that CD44 is present in all five of
the NSCLCs (Fig. I, Lanes 1-5) and SCLC H249 (Fig. 1, Lane 6) but
is low or absent in the other SCLC lines (Fig. 1, Lanes 7-10; CD44negative line H128 is not shown). The predominant protein species
detected was the 85-90-kDa "standard" form. Bands corresponding to
higher molecular weights likely represent alternate glycosylation since
reverse transcription-PCR detects only one predominant mRNA tran
script species (see below). Western blot results confirm the FACS data
that little or no CD44 protein expression occurs in the majority of
SCLC cell cultures.
High Levels of CD44 mRNA and Lack of CD44 Splicing Vari
ants in NSCLC Cultures. To determine the NSCLC pattern of CD44
RNA splicing, PCR analysis of cDNA was done by reverse transcrip
tion and PCR amplification. Fig. 2 shows representative CD44 PCR
products probed with a labeled internal oligonucleotide that recog
nizes both standard (CD44s) and variant species (CD44v). For com
Fig. 1. Western blot analysis of human lung carcinoma cell lines using polyclonal
CD44 antibodies. Top, NSCLC lines 1 to 5 (U1752, H460, A549, H157, H125); bottom
parison, U937 and HMEC, cells negative and positive for CD44
SCLC lines 6 to 10 (H249, DMS 53, OH3, H82, and H209). H249 (Lane 6) contains
splicing, respectively, were amplified in the same way. CD44 expres
SCLC cells with N-mvc amplification. Cellular proteins (10 ng) were resolved on 9%
polyacrylamide gels. Chemiluminescence detection by autoradiography was for 1 min.
sion was found only in NSCLC cell lines and SCLC clones of H249
1383
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1994 American Association for Cancer Research.
CD44
AND
LUNG CANCER
(examples in Fig. 2, Lanes 3 and 5). The size of the predominant
amplified product, 900 nt, was identical to that expected for CD44s.
Although we cannot rule out the presence of low levels of splice
variants, they are clearly in the minority. This result was true whether
the PCR products were examined by UV illumination or after South
ern blotting. Northern blots using polyadenylated RNA from the lung
carcinoma cells and probed with a Ã-2P-labeledCD44 cDNA contain
Table 2 Expression of CD44 in normal bronchial epithelium, bronchial metaplasia,
and noncultured lung carcinomas
Immunohistochcmistry of CD44 was carried out on fixed, paraffin-embedded resected
tumor sections (5 ¿IM)
from 9 SCLC and 14 NSCLC cancer patients. Immunohistochcmical staining used the avidin-biotinylated peroxidase complex technique following incu
bations with rabbit polyclonal CD44 antibodies. Negative controls for immunoreactivity
consisted of the serial lung sections treated with either nonimmune rabbit serum or
phosphate-buffered saline.
positiveTumor
ing exons 1-17 revealed the expected three alternatively polyadenyl
ated mRNAs (4, 1.7, and 1.2 kilobases) only in the NSCLC lines and
in SCLC H249 (not shown). These data indicate that CD44 mRNA is
not transcribed in SCLCs and that the NSCLC CD44 seen by Western
blot analysis is predominantly encoded by the CD44s message.
Expression of ( '1)44 in Normal Bronchial Epithelium, Bron
chial Metaplasia, and Noncultured Lung Carcinomas. Inumino
histochemistry of CD44 was carried out with fixed, paraffin-embed
ded specimens from nine SCLC and 14 NSCLC tumors. Ten of 14
resected NSCLCs were positive by immunochemistry for CD44, in
cluding 6 of 8 squamous cell carcinomas, 3 of 3 adenocarcinomas, and
1 of 3 large cell carcinomas (Fig. 3; Table 2). The immunoreactivity
of CD44 in tumors was cell surface, mostly focal, and detected in
subpopulations of tumor cells. None of the nine SCLCs tested were
positive, while one carcinoid tumor, a neoplasm with more differen
tiated neuroendocrine features than SCLC, had distinct CD44 immu
noreactivity. Three squamous cell carcinomas had 10-50% of tumor
cells positive for CD44 immunoreactivity. All three adenocarcinomas
had <10% tumor cells positive while one of three large cell carcino
mas had 10-50% cells positive.
In nonneoplastic lung, moderate to strong staining for CD44 was
seen in scattered lymphocytes and macrophages throughout the tissue.
However, the most intense immunoreactivity was detected in hyper-
Tumors
typeNon-small
cell("
14)Small
=
cell
CellClassic
Large
39No.36
10%10075
330
cell(n
and variantn38
= 9)SubtypeAdenocarcinomaSquamous
plastic type II pneumocytes (secretory alveolar cells) occurring in
rows and in groups lining tumors, fibrotic septae, and pulmonary scars
(Fig. 4). Serial sections and double immunostaining revealed that most
of the CD44-positive reactive type II pneumocytes were also positive
for SP-A, a specific marker for type II pneumocytes. Only occasion
ally were solitary type II cells in the alveoli positive for CD44. In the
epithelium of larger bronchi, basal cells, and cells participating in the
injury-response by squamous metaplasia were positive for CD44.
Selected cells in bronchial glands also demonstrated CD44 immuno
reactivity. Staining was most intense on the plasma membrane in all
of these cell types.
v-Ha-roy Induces CD44 Expression in a Variant SCLC Line.
Since v-Ha-ras transduction has been shown to cause a phenotypic
switch from SCLC to NSCLC (18-20), we chose this means to
:•¿,
VFig. 3. Photomicrographs of CD44 expression in nonneoplastic and neoplastic human lung. Paraffin sections were stained for CD44 by an immunoperoxidase technique. In/t, sections
of bronchial epithelium revealed intensely stained basal cells, covered by negative ciliated cells. Lu, bronchial lumen. X 270. In B, sections of squamous cell metaplasia in bronchus
showed CD44 immunoreactivity on the cell membrane in all the epithelial cells. X 410. In C, the cells of a SCLC remained negative for CD44. Note the positive stromal lymphocytes
in the middle and upper right hand corner of the specimen. X 270. In D. moderately differentiated pulmonary adenocarcinoma (papillary subtype) revealed intense staining for CD44.
X 410.
1384
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1994 American Association for Cancer Research.
CD44 AND LUNG CANCER
\
.
'.
•¿-.
-
n sv K
•¿
l
r'
Fig. 4. Photomicrographs of CD44 expression in activated type II pneumocytes. Paraffin sections were stained for CD44 by an immunoperoxidase technique. In the alveolar region,
CD44 immunoreactivity was detected mainly in the reactive type II cells lining fibrotic areas. Note the membrane staining on lateral, intercellular junctions (arrows). X 460. The same
section was subsequently stained for SP-A by an alkaline phosphatase immunohistochemistry technique to distinguish the type II pneumocytes (grayish cytoplasmic stain).
confirm that CD44 expression is associated with the NSCLC type
tumor. Following v-Ha-ras infection, SCLC cells of line H82 lost the
typical aggregated growth pattern of SCLC (Fig. 5, top left) and grew
in monolayers, typical of NSCLC (Fig. 5, top right). Flow cytometry
(Fig. 5, bottom left), Western blot analysis (Fig. 5, bottom right), and
Northern blot analysis (not shown) showed that transcription and
translation of CD44 were markedly increased in the ras-treated
SCLCs. These results further confirm that CD44 is associated with a
NSCLC-like phenotype and that CD44 may be subject to regulation
by a signal transduction pathway involving ras or ras-related proteins.
No CD44 expression was observed following infection with helper
virus alone (not shown).
to speculate that CD44-positive cells may flank these gelatinous re
gions and that CD44 expression may signify hyaluronate-rich areas.
The fact that CD44 was found to be located in between activated type
II pneumocytes that were arranged in distinct rows, one cell deep,
could indicate a simple cell-cell adhesion role for CD44. On the other
hand, this unusual positioning of CD44 may reflect the presence of an
intercellular hyaluronan matrix through which soluble molecules can
diffuse. Future studies will be required to evaluate these hypotheses.
A number of studies have demonstrated that "variant" CD44
(CD44v) is associated with tumor progression (7). The hallmark of
CD44v is the inclusion of additional extracellular domains which have
been shown to decrease the ability of CD44 to serve as a hyaluronan
receptor (35, 36). The virtual absence of CD44 on the highly meta-
DISCUSSION
static SCLCs raises the question as to whether lack of standard CD44
is the functional equivalent of variant CD44. If CD44-hyaluronan
In this study of CD44 expression in human lung carcinomas, the
interactions are used to anchor the cells in one area, then it is not
presence of CD44 correlates with the NSCLC phenotype and marks
unreasonable
to speculate that lack of functional CD44 would relieve
reactive changes in pulmonary cells. Both cultured NSCLC tumor
the cells of any restraint to hyaluronate-rich tissues and enable the
lines and resected primary NSCLC tumors were immunoreactive for
process of métastasesto begin. Precedence for this inverse relation
CD44. In contrast, none of the nine SCLC tumors examined or the
ship between high metastatic potential and low CD44s is found in
majority of cultured lines of this tumor type showed CD44 positivity.
cases
of neuroblastoma (10). The functional consequences of CD44
In normal lung tissue, CD44 was found on basal cells of the upper
expression,
hyaluronate binding, and metastatic potential of lung
bronchi, pulmonary macrophages, alveolar lymphocytes, and acti
cancer cells is presently under investigation.
vated type II alveolar pneumocytes. The functional relevance of CD44
The finding that, among lung cancer cells, CD44 is associated with
in these tissues most likely relates to the fact that CD44 possesses a
specific differentiation phenotypes (NSCLC, activated type II cells,
high affinity binding site for hyaluronan, the repeating disaccharide
found in extracellular matrices (1-6). In the upper bronchi of the basal cells, and squamous metaplasia), raises the possibility that this
hamster lung, hyaluronan is an abundant component of the perivasmarker could be used to elucidate the histogenesis and progression of
this disease. Squamous metaplasia of the lung is indicative of the type
cular matrix surrounding large blood vessels and basement membrane
regions of epithelium (33, 34). Studies in animal tissues suggest that
of cellular insult associated with cigarette smoking (37). Although
hyaluronan-rich matrices function as gelatinous channels through
opinions differ as to its origin, the prevailing view is that squamous
which water soluble molecules can diffuse (37). It is not unreasonable
metaplasia arises from the replacement of ciliated and mucous cells by
1385
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1994 American Association for Cancer Research.
CD44
H82
AND
LUNG CANCER
A549, and U1752 have mutations or overexpression of the K-ra.y gene
(14, 15, 41, 42). Further studies will be required to fully define the
effect of ras and/or c-myc gene expression on CD44 appearance.
SCLC H69 and its clonal derivative H249 differ from other SCLCs
in that they have extensive N-myc overexpression (43). The idea that
N-myc amplification, alone is sufficient to cause transcriptional acti
vation of CD44 was ruled out by examination of other N-myc overexpressing SCLC lines, H526 and H187, that were found to be CD44
negative (not shown). Whether H249 and H69 have additional genetic-
H82-ras
o'ù
1
.(L*
•¿â€¢'*
•¿
>*• "• -Jf
t».."»
.
O
changes that could contribute to CD44 expression or transition to
NSCLCs remains to be determined.
Lung cancers can take 10-20 years to develop but, once detected,
are often well advanced. Thus, the need for markers to identify early
forms of lung cancer is essential. Nearly all NSCLC examined here
expressed CD44. Among this group, the adenocarcinoma had the most
frequent representation. Historically, a number of markers have been
used to define the adenomatous class of lung cancer. These include
surfactant-associated protein, carcinoembryonic antigen, the ras on
cogene, and the Clara 10-kDa protein (44). Other subclasses of
NSCLCs also expressed CD44. Although confirmation of CD44 as a
general marker for NSCLC or a specific marker for a subset of
NSCLC awaits results of larger studies, the advantages of CD44 in
terms of cell surface localization, expression in early stages of squamous metaplasia, and association with specific histological differen
tiation phenotypes make it a promising candidate for diagnostic pan
els. Since therapeutic strategies differ dramatically depending on
histological differentiation, results of discriminatory markers such as
the one described here are especially valuable.
li
SW
FACS
H82
H82-ras
WESTERN
H82
H82-ras
CD44 EXPRESSION
ACKNOWLEDGMENTS
Fig. 5. Effect of v-Ha-ras infection on CD44 levels in small cell lung carcinomas.
Aggregation status of H82 and H82 ra.s-infccted cells (¡op).Cells were assessed in all
experiments 4 weeks following v-Ha-rw.v infection. Flow microfluorimetry of CD44 on
nonpermeabilized cells using monoclonal antibody H4C4 before (H82) and after v-Ha-ra.v
infection (H82-ra.s) (bonom left). Western blot analysis of H82 and HH2-ras cells with
polyclonal CD44-specific antibodies (prepared as described in Fig. /. bottom righi). No
significant quantitative differences were seen in levels of actin following rehybridization
with anti-actin monoclonal antibody clone C4 (insel).
The authors acknowledge Dr. James E. K. Hildreth for his kind help with the
FACS protocol.
REFERENCES
hyperplastic basal cells (38, 39). Our observation that both basal cells
and sections of squamous metaplasia are CD44 positive is fully con
sistent with a common origin. That SCLCs are, in general, CD44
negative could mean that they are derived from one of the many
CD44-negative lung cells. The CD44-positive NSCLCs, on the other
hand, could arise either via a CD44-positive progenitor (activated type
II pneumocytes or basal cells) or by the differentiation of SCLCs.
Evidence for the latter possibility, that some NSCLCs can emerge
from populations of SCLCs, is provided by clinical observations and
by in vitro studies in which v-Ha-ras causes a transition from the
SCLC to NSCLC phenotype (19, 20). Our results, that CD44-negative
SCLCs induced to take on the differentiation status of NSCLC in
culture become CD44 positive, are also consistent with the latter
hypothesis. Although CD44 alone cannot discriminate among the
progenitor cell possibilities, as a member of a multimarker panel it
promises to contribute to our understanding of lung tumor histogenesis and progression.
In these studies, v-Ha-ras gene insertion was used simply as a
means to induce a NSCLC phenotype. Although we do not intend this
to be a study of ras gene effects on CD44 expression, our results are
consistent with the ras-induced CD44 expression seen in model sys
tems (40). The fact that the recipient cultured SCLC line H82 had
c-myc amplification prompted us to examine SCLC line H209 that had
been transduced with c-myc, v-Ha-ras, or both. Preliminary results
(not shown) suggest that both ras and myc are required for CD44
up-regulation. It is noteworthy that NSCLCs lines H460, H157, H125,
1. Goldstein, L. A., Zhou, D. F., Picker. L. J.. el al. A human lymphocyte homing
receptor, the hermes antigen, is related to cartilage proteoglycan core and link pro
teins. Cell, 56: 1063-1072, 1989.
2. Stamenkovic. I., Amiot. M., Pesando. J. M., and Seed, B. A lymphocyte molecule
implicated in lymph node homing is a member of the cartilage link protein family.
Cell, 56: 1057-1062. 1989.
3. Culty, M., Miyake, K.. Kincade, P. W., el al. The hyaluronate receptor is a member
of the CD44 (H-CAM) family of cell surface glycoproteins. J. Cell Biol., Ill:
2765-2774, 1990.
4. Aruffo, A., Stamenkovic, I.. Mclnick, M., Underbill. C. B., and Seed, B. CD44 is the
principal cell surface receptor for hyaluronate. Cell, 61: 13U3-1313, 19<)().
5. Miyake, K., Underbill, C. B., Lesley, J.. and Kincade, P. W. Hyaluronate can function
as a cell adhesion molecule and CD44 participates in hyaluronate recognition. J. Exp.
Med., 772: 69-75, 1990.
6. Lesley, J., Schulte, R., and Hyman, R. Binding of hyaluronic acid to Ivmphoid cell
lines is inhibited by monoclonal antibodies against Pgp-1. Exp. Cell Res., 1K7:
224-233, 1990.
7. Lesley, J., Hyman, R., and Kincade, P. W. CD44 and its interactions with the extra
cellular matrix. Adv. Immunol., 54: 271-335, 1993.
8. Heider, K. H-, Hofmann, M., Hors, E., et al. A human homologue of the rat metastasisassociated variant of CD44. J. Cell Biol., 120: 227-233, 1993.
9. Birch. M.. Mitchell, S., and Hart, I. R. Isolation and characterization of human
melanoma cell variants expressing high and low levels of CD44. Cancer Res., 51:
6660-6667, 1991.
10. Shtivelman, E.. and Bishop. J. M. Expression of CD44 is repressed in neuroblastoma
cells. Mol. Cell. Biol.. //: 5446-5453, 1991.
11. Kalemkerian, G. P., and Mabry, M. Cellular and molecular biology of small-cell lung
cancer. In: i. A. Roth. J. D. Cox, and W. K. Hong (eds.), Lung Cancer, pp. 57-84.
Boston: Blackwell Scientific Publications, 1993.
12. Carney, D. N., Gazdar, A. F., Beplcr, G., el at. Establishment and identification of
small cell lung cancer cell lines having classic and variant features. Cancer Res.. 45:
2913-2923, 1985.
13. Gazdar, A. F., Carney, D. N., Nau, M. M., and Minna. J. D. Characterization of variant
subclasses of cell lines derived from small cell lung cancer having distinctive bio
chemical, morphological, and growth properties. Cancer Res., 45: 2924-2930. 1985.
14. Mitsudomi. T.. Viallet, J., Mulshine, J. L, Linnoila, R. 1., Minna, J. D., and Gazdar,
A. F. Mutations of ras genes distinguish a subset of non-small-cell lung cancer cell
lines from small-cell lung cancer cell lines. Oncogene, 6: 1353-1362, 1991.
1386
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1994 American Association for Cancer Research.
CD44 AND
LUNG CANCER
15. Kurzrock, R., Gallick, G. E., and Gutterman, J. U. Differential expression of p21 ras
gene products among histológica! subtypes of fresh primary human lung tumors.
Cancer Res., 46: 1530-1534, 1986.
16. van de Rodenhuis, S., et al. Mutational activation of the K-ras oncogene. A possible
pathogenetic factor in adenocarcinoma of the lung. N. Engl. J. Med., 317: 929—935,
1987.
17. Doyle, L. A., Mabry, M., Stahel, R. A., Waibel, R.. and Goldstein, L. H. Modulation
of neuroendocrine surface antigens in oncogene-activated small cell lung cancer lines.
Br. J. Cancer (Suppl.), 14: 39-42, 1991.
18. Mabry, M., Nelkin, B. D., Falco, J. P., Barr, L. F., and Baylin, S. B. Transitions
between lung cancer phenotypes—implications for tumor progression. Cancer Cells,
3: 53-58, 1991.
19. Mabry, M., Nakagawa, T., Nelkin, B. D., et al. v-Ha-ras oncogene insertion: a model
20.
21.
22.
23.
for tumor progression of human small cell lung cancer. Proc. Nati. Acad. Sci. USA,
85: 6523-6527, 1988.
Falco, J. P., Baylin, S. B., Lupu, R., et al. v-rasH induces non-small cell phenotype,
with associated growth factors and receptors, in a small cell lung cancer cell line. J.
Clin. Invest., «5:1740-1745, 1990.
Sorenson, 0. D., Pettengill, 0. S., Brinck, J. T., Cate, C. C., and Maurer, L. H.
Hormone production by cultures of small-cell carcinoma of the lung. Cancer (Phila.),
47: 1289-1296, 1981.
Griffin, C. A., and Baylin, S. B. Expression of the c-myb oncogene in human small
cell lung carcinoma. Cancer Res., 45: 272-275, 1985.
Bergh, J., Nilsson, K., Zech, L., and Giovanella, B. Establishment and characteriza
tion of a continuous lung squamous cell carcinoma cell line (U-1752). Anticancer
Res., /.- 317-322, 1981.
24. Belitsos, P. C., Hildreth, J. E., and August, J. T. Homotypic cell aggregation induced
by anti-CD44 (Pgp-1) monoclonal antibodies and related to CD44 (Pgp-1) expression.
J. Immunol., 144: 1661-1670, 1990.
25. Hughes, E. N., and August, J. T. Characterization of plasma membrane proteins
identified by monoclonal antibodies. J. Biol. Chem., 256: 664-671, 1981.
26. Lin, S. L., Derr, D., and Hildreth, J. E. A monoclonal antibody against a novel 20-kDa
protein induces cell adhesion and cytoskeleton-dependent morphologic changes. J.
Immunol., 149: 2549-2559, 1992.
27. Scott, J. E. Secondary structures in hyaluronan solutions. In: The Biology of Hyaluronan, pp. 6-15. Ciba Foundation Symposium, Vol. 143, United Kingdom:
Chichester, John Wiley & Sons, Inc., 1989.
28. Ausubel, F. M.. Brent, R., Kingston, R. E., et al. (eds.). Current Protocols in Mo
lecular Biology. New York: Green Publishing Associated and Wiley-Interscience,
1988.
29. Chirgwin, J. M., Przybyla, A. E., MacDonald, R. J., and Rutter, W. J. Isolation of
30.
31.
32.
33.
biologically active ribonucleic acid from sources enriched in ribonuclease. Biochem
istry, 18: 5294-5299, 1979.
Ponte, P.. Ng, S. Y, Engel, J., Gunning, P., and Kedes, L. Evolutionary conservation
in the untranslated regions of actin mRNAs: DNA sequence of a human ß-actin
cDNA. Nucleic Acids Res., 12: 1687-1696, 1984.
Linnoila, R. 1., Mulshine, J. L., Steinberg, S. M., et al. Neuroendocrine differentiation
in endocrine and nonendocrine lung carcinomas. Am. J. Clin. Pathol., 90: 641-652,
1988.
Linnoila, R. I., Mulshine, J. L., Steinberg, S. M., and Gazdar, A. F. Expression of
surfactant-associated protein in non-small-cell lung cancer: a discriminant between
biologic subsets. Nail. Cancer Inst. Monogr., 1992: 61-66, 1992.
Green, S. J., Tarone, G., and Underhill, C. B. Distribution of hyaluronate and hyaluronate receptors in the adult lung. J. Cell Sci., 89: 145-156, 1988.
34. Underhill,C. B., Nguyen,H.A., Shizari,M.,and Culty,M. CD44positivemacro
phagestake up hyaluronanduringlung development.Dev.Biol., 755: 324-336,
1993.
35. Sy, M. S., Guo, Y. J., and Stamenkovic, I. Distinct effects of two CD44 isoforms on
tumor growth in vivo. S. Exp. Med., 174: 859-866, 1991.
36. Thomas, L., Byers, H. R., Vmk, J., and Stamenkovic, I. CD44H regulates tumor cell
migration on hyaluronate-coated substrate. J. Cell Biol., 118: 971-977, 1992.
37. Lee, S. J., and Hong, W. K. Biology of preneoplastic lesions. In: J. A. Roth, J. D. Cox,
and W. K. Hong (eds.). Lung Cancer, pp. 34-56. Boston: Blackwell Scientific Pub
lications, 1993.
38. Black. H., and Ackerman, L. Importance of epidermoid carcinoma in situ in histogenesis of carcinoma of the lung. Ann. Surg., 136: 44—55,1952.
39. Weller, R. Metaplasia of bronchial epithelium: a postmortem study. Am. J. Clin.
Pathol., 23: 768-774, 1953.
40. Hofmann, M., Rudy. W., Gunthert, U., et al. A link between ras and metastatic
behavior of tumor cells: ras induces CD44 promoter activity and leads to low-level
expression of metastasis-specific variants of CD44 in CREF cells. Cancer Res., 55:
1516-1521, 1993.
41. Kiefer, P. E., Wegmann, B., Bacher. M., Erbil, C., Heidtmann, H., and Havemann, K.
Different pattern of expression of cellular oncogenes in human non-small-cell lung
cancer cell lines. J. Cancer Res. Clin. Oncol., 116: 29-37, 1990.
42. Bos, J. L. The ras gene family and human carcinogenesis. Mutât.Res., 195: 255-271,
1988.
43. Kiefer, P. E., Bepler, G.. Kubasch, M., and Havemann, K. Amplification and expres
sion of protooncogenes in human small cell lung cancer cell lines. Cancer Res., 47:
6236-6242, 1987.
44. Mulshine, J. L., Linnoila. R. I., Treston, A. M., et al. Candidate biomarkers for
application as intermediate end points of lung carcinogenesis. J. Cell. Biochem.
(Suppl.), 16C: 183-186, 1992.
1387
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1994 American Association for Cancer Research.
Expression of CD44 in Human Lung Tumors
Margaret B. Penno, J. Thomas August, Stephen B. Baylin, et al.
Cancer Res 1994;54:1381-1387.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/54/5/1381
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. © 1994 American Association for Cancer Research.