Restoration of the Mucous Phenotype by Retinoic Acid in
Retinoid-Deficient Human Bronchial Cell Cultures:
Changes in Mucin Gene Expression
Ja Seok Koo, Joo-Heon Yoon,* Thomas Gray, Derek Norford, Anton M. Jetten, and Paul Nettesheim
Laboratory of Pulmonary Pathobiology, National Institute of Environmental Health Sciences,
Research Triangle Park, North Carolina
Retinoid-deficient cultures of airway epithelial cells undergo squamous differentiation. Treatment of such
cultures with retinoic acid (RA) leads to restoration of the mucous phenotype. The purpose of our study
was to characterize the cellular and molecular changes following RA treatment of retinoid-deficient human
tracheobronchial epithelial cell cultures. Of particular interest was to determine when during the conversion of the squamous to the mucous phenotype the mucin genes MUC2, MUC5AC, and MUC5B were expressed. We used cornifin a and secreted mucin as markers to monitor the squamous and mucous phenotypes, respectively. Our studies showed that the RA responsiveness of the cultures progressively decreased
with protracted retinoid deficiency, requiring higher RA concentrations to restore the mucous phenotype.
Within 12 h after the start of RA treatment, cornifin a expression decreased, signaling the beginning of a
change in cellular phenotype. At 24 h after addition of RA to the cultures, a significant number of mucous
cells appeared, and at 72 h mucin was secreted in measurable amounts. Induction of mucin gene expression occurred sequentially: MUC2, MUC5AC, and MUC5B mRNAs were upregulated at 24, 48, and 72 h,
respectively. When cultures maintained in 1028 M RA were treated with 1026 M RA, MUC2 but not
MUC5AC and MUC5B mRNA levels were upregulated within 6 h. Our study indicates that MUC2 mRNA
is an early marker of mucous differentiation, whereas MUC5AC and MUC5B mRNAs are expressed during more advanced stages of mucous differentiation. Our studies further suggest that each of the mucin
genes is regulated by distinct mechanisms. Koo, J. S., J.-H. Yoon, T. Gray, D. Norford, A. M. Jetten,
and P. Nettesheim. 1999. Restoration of the mucous phenotype by retinoic acid in retinoid-deficient
human bronchial cell cultures: changes in mucin gene expression. Am. J. Respir. Cell Mol. Biol. 20:
43–52.
Vitamin A (retinol) and related compounds, summarily
called retinoids, are known to regulate cell proliferation,
differentiation, and morphogenesis. Their effects are mediated through nuclear retinoid receptors, retinoic acid receptors (RARs) and retinoid X receptors (RXRs), which
can activate or repress transcription of a multitude of target genes (1). Numerous studies have shown that retinoids
(Received in original form January 20, 1998 and in revised form June 12,
1998)
Address correspondence to: Paul Nettesheim, M.D., Laboratory of Pulmonary Pathobiology, MD D2-01, P.O. Box 12233, National Institute of Environmental Health Sciences, Research Triangle Park, NC 27709. E-mail:
[email protected]
* Current address: Department of Otorhinolaryngology, Yonsei University
College of Medicine, Seoul, Korea.
Abbreviations: dimethyl sulfoxide, DMSO; normal human tracheobronchial epithelial, NHTBE; retinoic acid, RA; reverse transcriptase–polymerase chain reaction, RT–PCR; sodium dodecyl sulfate, SDS; saline sodium citrate, SSC.
Am. J. Respir. Cell Mol. Biol. Vol. 20, pp. 43–52, 1999
Internet address: www.atsjournals.org
are important for normal lung development and maturation, and since the pioneering studies of Mori (2) and Wolbach and Howe (3), it is known that vitamin A is essential
for the maintenance of the mucociliary epithelium in the
conducting airways (4–7). Chronic vitamin A deficiency
leads to replacement of the mucociliary epithelium by a
stratified squamous epithelium, and reversal of this abnormal form of differentiation has been reported to occur
within a few days after vitamin A was restored to the diet
(8). In vitro studies with tracheal explants and primary epithelial cells have shown that in the absence of retinoids
the cultures undergo squamous differentiation and that
addition of vitamin A to the medium restores mucous differentiation (5, 9–11).
We have been interested in elucidating the role of retinoids in mucous cell differentiation and mucin gene expression (12–14). Using cultured tracheobronchial epithelial cells from rats and humans, we previously showed that
the development of the mucous phenotype as well as expression of MUC2 and MUC5AC depend on the presence
of retinoids in the culture media. In the absence of retinoic
44
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 20 1999
acid (RA), MUC2 and MUC5AC mRNA levels were very
low or undetectable (13, 14) and the cultures contained
very few mucous cells (15). The purpose of our study was
to examine changes in expression of mucin genes and the
squamous differentiation marker cornifin a (16) during
the restoration of the mucous phenotype following RA
treatment of RA-deficient cultures.
We concentrated on the mucin genes MUC2, MUC5AC,
and MUC5B for the following reasons: MUC2, even though
it is only weakly expressed in normal bronchial epithelium
and may not be a major secreted mucin (17), has been
shown to be markedly upregulated in cystic fibrosis patients (18); MUC5AC and MUC5B mucins, on the other
hand, have been shown to be major components of human
airway secretions (19–21). Furthermore, several investigators have demonstrated that the expression of MUC2 and
MUC5AC mucin genes is regulated by cytokines and bacterial products (18, 22–24). Athough it is not clear at the
moment whether other mucins are also significant components of human airway secretions and to what extent the
mixture of mucins changes in disease, it appears that
MUC5AC and MUC5B mucins are important constituents
of tracheobronchial secretions and that MUC2 mRNA
levels are regulated by factors causing hypersecretion.
Our studies showed that the vitamin A-deficient cultures became progressively less responsive to RA. Tenfold higher concentrations were required to reestablish
the mucous phenotype in late compared with early vitamin
A-deficient cultures. After RA treatment, cornifin a mRNA
was suppressed at 12 h; MUC2, MUC5AC, and MUC5B
mRNAs were induced at 24, 48, and 72 h, respectively; and
mucous cells reappeared at 24 h, but mucin was not secreted until 72 h. Our study delineates a progressive chain
of events during remodeling of the RA-deficient squamous epithelium triggered by RA treatment, starting with
repression of the squamous differentiation program, followed by sequential expression of mucin genes and culminating in mucin secretion.
We also compared the kinetics of mucin gene induction
in RA-deficient and RA-sufficient cultures following treatment with high doses of RA. We found that MUC2, but
not MUC5AC and MUC5B, was much more rapidly upregulated in RA-sufficient than in RA-deficient cultures,
that is, within 6 h rather than 24 h. This suggests perhaps
that in RA-sufficient cultures RA affects MUC2 transcription more directly than in RA-deficient cultures, in which
induction of mucin genes occurs as part of complex changes
in the differentiation program.
Materials and Methods
Air–Liquid Interface Cultures and RA
Normal human tracheobronchial epithelial (NHTBE)
cells (strains 2002 and 17684; Clonetics Co., San Diego,
CA) from passage 2 were seeded at a density of 1 3 105
cells onto 24-mm permeable membranes in serum-free,
hormone- and growth-factor-supplemented medium as described previously (15) except for the following changes.
Instead of using Transwell-COL, we used Transwell-clear
membranes (Costar Co., Cambridge, MA). Because we
found that NHTBE cells grew and differentiated equally
well on these membranes regardless of the presence or absence of collagen gel (14), we seeded the cells directly onto
the membranes. All trans-retinoic acid (Sigma, St. Louis,
MO) was dissolved in dimethyl sulfoxide (DMSO) and
stored at 2208C until needed. The final DMSO concentration in the medium was never greater than 0.1%. Control
cultures were treated with DMSO only.
Histology and Cytology
For histologic analysis, cultures were fixed in 10% neutral
buffered formalin, embedded in paraffin, sectioned, and
stained with hematoxylin and eosin (H&E). Cytospin
slides and cell pellets were prepared from cell suspensions
obtained following enzymatic dissociation of cultures and
were fixed and stained as described previously using the
17Q2 antibody (15).
Immunoassays
The dot-blotting assay for quantitation of secreted mucin
and the immunocytochemical methods for quantitating
mucous cells using antibodies 17Q2 and H6C5 have been
described previously (15). In short, apical secretions accumulating during a 24-h period were collected using 1.5 ml
phosphate-buffered saline. Dilutions of apical secretions
and mucin standards were applied to nitrocellulose membranes, reacted with the antimucin antibody, and followed
by reaction with horseradish peroxidase-conjugated secondary antibody. The signal was detected by chemiluminescence (ECL kit; Amersham, Arlington Heights, IL),
and a standard curve was generated by linear regression
analysis from which the concentration of mucin of individual samples could be determined. Antibody 17Q2 (a generous gift from Dr. J. St. George, Genzyme, Framingham,
MA) was prepared against rhesus monkey tracheal secretions and shown to react with human goblet cells (25, 26).
Antibody H6C5 (a generous gift from Dr. C. William
Davis, University of North Carolina at Chapel Hill) was
prepared against high molecular weight components of
human cystic fibrosis sputum obtained according to procedures described by Thornton and colleagues (27). Antibodies H6C5 and 17Q2 have previously been shown to
give similar results in both dot-blotting analysis and quantitation of immunopositive cells (15). Because of the limited supply of 17Q2 antibody, we used it for immunostaining of cytospins at a dilution of 1:250. This antibody
gives a better staining of secretory granules. The H6C5 antibody was mostly used for dot blotting of secretions (dilution 1:250). Human mucin purified from cystic fibrosis
sputum (0.85 mg dry weight of purified protein per milliliter) was used as mucin standard. The amount of secreted
mucin was presented as micrograms of mucin per 106 cells.
To determine the number of cells per culture, the cultures
were enzymatically dissociated and the cells were visually
counted by using a hemacytometer. The data are represented as the means 6 SD of triplicate cultures. Statistical
comparisons were made using Student’s t test. Immunolabeling of mucous cells were quantitated by scoring 1,500
cells from 10 random fields on each slide as described previously (15).
Koo, Yoon, Gray, et al.: Regulation of the Mucous Phenotype and Mucin Genes by RA
Isolation of Total RNA and Northern Blot Analysis
Total RNA was isolated from pools of triplicate cultures
using Tri-Reagent (Molecular Research Center, Cincinnati, OH). Five micrograms of RNA samples were separated in 1.4% agarose-formaldehyde gel electrophoresis
and transferred overnight onto Nytran membrane (Schleicher & Schuell, Keene, NH) by capillary action (28). The
filters were dried, baked for 2 h at 808C, and prehybridized
for 30 min at 688C in 15 ml of QuickHyb Hybridization solution (Stratagene, La Jolla, CA). Radiolabeling of 20 to
50 ng of cornifin a cDNA (16) fragment was performed
using redi-Prime random primer labeling kit (Amersham)
and [a-32P] deoxycitidine triphosphate (DuPont NEN, Wilmington, DE) according to manufacturer’s recommendations. The filters were hybridized with the radiolabeled
probes (specific activity of z 109 cpm/mg DNA) and 1 mg
of boiled salmon sperm DNA for 1 h at 688C. Membranes
were washed three times with 2 3 saline sodium citrate
(SSC)/0.1% sodium dodecyl sulfate (SDS) for 15 min at
room temperature, once for 30 min at 608C in 0.13 SSC/
0.1% SDS, and once for 5 min at room temperature in
0.13 SSC/0.1% SDS. Membranes were placed in contact
with Hyperfilm-MP autoradiography film (Amersham)
with an intensifying screen at room temperature for 30 to
60 min. Equivalent loading of the RNA samples was determined by ethidium bromide staining of 28S ribosomal
RNA of the agarose-formaldehyde gel.
Quantitative Reverse Transcriptase–Polymerase
Chain Reaction
Methods to detect MUC2 and MUC5AC mRNA levels using quantitative reverse transcriptase–polymerase chain
reaction (RT–PCR) have been reported elsewhere in detail (13). Oligonucleotide primers for RT–PCR were designed according to the published sequences for human
MUC2 (29; Genbank accession no. L21998, 59 primer: TGC
CTG GCC CTG TCT TTG; 39 primer: CAG CTC CAG
CAT GAG TGC); and human MUC5AC (30; Genbank
accession no. U06711, 5 9 primer: TCC GGC CTC ATC
TTC TCC; 39 primer: ACT TGG GCA CTG GTG CTG);
and human MUC5B (31; Genbank accession no. Z72496,
59 primer: ACT CCA GAG ACT GTC CAC AC; 39 primer:
TAC CAC TGG TCT GTG TGC TA). The MUC2 oligonucleotides generated the expected 440-bp fragment,
MUC5AC oligonucleotides generated a 680-bp fragment,
and MUC5B amplimers produced a 388-bp fragment. All
of the amplimers for mucin mRNA were designed to bind
nontandem repeat regions of the respective genes. Oligonucleotide amplimers for b2 microglobulin (b2M, which
was used as a control gene for RT–PCR) were purchased
from Clontech (Palo Alto, CA) and generated a 335-bp
PCR fragment.
PCRs for MUC2 and MUC5AC were performed as described previously (13). For PCR of MUC5B, 27 cycles of
denaturation (958C/1 min), annealing (55 8C/1 min), and
extension (728C/1 min) were used in the presence of 1.5
mM MgCl2 followed by 20 min extension at 72 8C. PCRs
were done in the presence of an internal standard—the socalled MIMIC (32, 33)—which is double-stranded DNA
with ends that are complementary to the specific MUC2,
45
MUC5AC or MUC5B primers but contain different lengths
of stuffer sequences with no homology to the MUC2,
MUC5AC, and MUC5B target sequences. The gene-specific primers hybridize to both the target genes and the
MIMICs in a competitive manner, and the PCR amplification products are identified by the length of DNA between
two primers. A similar amplification efficiency and mutual
competition of target cDNA and MIMIC were demonstrated by the titration of PCR cycle number and amount
of MIMIC cDNA, respectively.
To calculate changes in the levels of MUC2 and
MUC5AC mRNAs, RT–PCRs were performed in the presence of 1023 amol MIMIC per PCR, except for the experiment summarized in Figure 5 in which 1022 amol MIMIC
was used for MUC5AC because the MUC5AC signal was
very strong. For MUC5B, 1021 amol MIMIC per PCR was
used. PCR products were separated by electrophoresis on
a 2% Seakem agarose gel (FMC, Rockland, ME) containing ethidium bromide (50 ng/ml), and the resulting bands
were captured by an IS-1000 digital imaging system (Alpha Innotech Co., San Leandro, CA) or were photographed with Polaroid type 55 film. The intensity of the
bands was analyzed using IA-200 image analysis software
provided by the manufacturer (Alpha Innotech). The ratio
of the signal intensity of the target DNA to the MIMIC
was used to estimate the relative abundance of the target
mRNA among the various samples. When mucin mRNA
was undetectable, the lower limit of detection (2.5 3 1025
amol/mg total RNA) (13) was used to estimate the relative
changes in mRNA levels. Specific amplification of MUC2,
MUC5AC, and MUC5B mRNAs was confirmed by sequencing (dsDNA Cycle Sequencing System; GIBCO BRL,
Gaithersburg, MD) the PCR fragments.
Results
Changes in RA Responsiveness of NHTBE
Cells with Time in Culture
We previously showed that in the absence of retinoid
NHTBE cell cultures differentiate into a stratified squamous epithelium instead of a columnar mucous epithelium
(15). We found that 10 28 to 1027 M RA is required to
maintain mucous differentiation. To study the reversibility
of squamous differentiation by RA, cells were grown for
various lengths of time in RA-free medium, and at either
Day 7, 8, 9, 10, or 11 the cultures were treated for 4 d with
1027 M RA. Control cultures were sham treated. To monitor changes in squamous and mucous differentiation, we
measured cornifin a mRNA as a marker of squamous differentiation (16) and apically secreted mucin as a marker
of mucous differentiation (15), respectively. As seen in
Figure 1A, RA-deficient cultures secreted little or no mucin. Treatment of Day 7, RA-deprived cultures with RA
over a 4-d period resulted in production of large amounts
of mucin. When RA treatment was started at later times
(Days 8 to 11), the ability of RA to induce mucin steadily
decreased. In all RA-treated cultures, cornifin a mRNA
was strongly suppressed (Figure 1B) regardless of when
RA treatment was initiated. These data suggest that the
cultures remained RA responsive, even though restoration
46
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 20 1999
ter the start of treatment, the mRNA levels of MUC2 and
MUC5AC, and of cornifin a were determined using quantitative RT–PCR (13) and Northern blotting, respectively.
As is shown in Figure 2, MUC2 mRNA expression was
clearly upregulated in cultures treated on Day 7 with 10 28
M RA for 4 d, and on the basis of densitometric analysis at
least a 40-fold induction was observed. Maximal induction,
approximately 110-fold above control cultures, occurred
with 1026 M RA. With 10-d-old cultures, 1027 M RA was
required to obtain induction of MUC2 mRNA (z 35-fold
induction). Maximal expression of MUC2 mRNA occurred with 1026 M RA (at least 45-fold induction). To induce MUC5AC expression in 7-d-old cultures required
1027 M RA (z 90-fold induction); maximal induction was
obtained with 1026 M RA (z 140-fold induction). In comparison, cultures treated initially on Day 10 required 10 26
M RA (z 60-fold induction) to induce significant levels of
MUC5AC. In all cultures treated with 1025 M RA, expression of both mucin genes was less than that observed with
1026 M RA, suggesting that 1025 M RA may be toxic. We
also monitored the effect of RA treatment on cornifin a
gene expression. Cornifin a mRNA levels were markedly
reduced by 10 28 M RA in cultures treated on Day 7,
whereas in cultures treated on Day 10, 1027 M RA was required to cause a significant reduction. These studies
showed that the responsiveness of NHTBE cultures to RA
treatment decreased with time. The longer the cells were
maintained in RA-free medium, the more RA was required to induce the two mucin genes and to inhibit cornifin a mRNA expression. The studies also showed that
MUC5AC, in comparison with MUC2, required higher
RA concentrations for induction of gene expression.
Figure 1. Effect of RA treatment on RA-deficient NHTBE cells
at different times of culture. Cells were grown in RA-free medium. At either Day 7, 8, 9, 10, or 11, 10 27 M RA was added to
the cultures (black bars). Control cultures were sham treated
(gray bars). (A) Four days after the start of treatment, apically secreted mucin was measured by immunoblotting. Values are the
means 6 SD of triplicate cultures. (B) Cornifin a mRNA levels
were determined by Northern blotting (top panel), and RNA
loading was verified by staining of 28S rRNA with ethidium bromide (bottom panel). The findings were confirmed in two similar
experiments.
of the mucous phenotype appeared to be less effective
when the onset of treatment was delayed.
RA Dose-Dependent Effects on Mucin and
Cornifin a Gene Expression
To define more precisely the changes of RA responsiveness of early and late RA-depleted cultures, NHTBE cells
grown for either 7 or 10 d in RA-free medium were
treated with different concentrations of RA. Four days af-
Figure 2. Dose-response effects of RA treatment on mucin gene
and cornifin a expression. NHTBE cells were cultured for either
7 or 10 d in RA-deficient medium and were then treated with various concentrations of RA for 4 d. The levels of MUC2 and
MUC5AC mRNA expression were determined by quantitative
RT–PCR from total RNA isolated on Days 11 and 14, respectively. Cornifin a mRNA levels were measured by Northern blotting, and equal RNA loading was verified by staining of 28S
rRNA ethidium bromide. Lane c represents sham-treated controls. Similar results were obtained in a separate experiment.
Koo, Yoon, Gray, et al.: Regulation of the Mucous Phenotype and Mucin Genes by RA
Histologic and Immunocytochemical Analysis
To monitor the morphologic conversion of the squamous
to the mucous phenotype, histologic and cytologic preparations were obtained from RA-deficient cultures and
from cultures treated for 4 d with 1026 M RA. Figures 3A
and 3B show the progression in RA-deficient cultures from
a thin layer of squamous epithelium on Day 7 to a multilayered squamous epithelium on Day 11. Cultures treated
from Day 7 to Day 11 with RA (Figure 3C) exhibited a
cuboidal to low columnar epithelium with only a thin sheet
of flattened cells remaining at the apical surface. RA-deficient cultures on Day 10 (Figure 3D) revealed a stratified
squamous epithelium that became extensively cornified by
Day 14 (Figure 3E). After 4 d of RA treatment beginning
on Day 10 (Figure 3F), the squamous epithelium was converted to a polarized columnar epithelium. Only a thin
layer of flattened cells remained on the apical surface and
desquamation of entire sheets of cells was evident.
Cells obtained from enzymatically dissociated cultures
were stained with antimucin antibody. Between 1% and
3% of the cells from Day 10 and Day 14 RA-deficient cultures (Figures 3G and 3H) showed a weak, diffuse cytoplasmic staining with antimucin antibody. In contrast,
many cells (z 40%) obtained from cultures treated from
Day 10 to Day 14 with 10 26 M RA reacted strongly with
antimucin antibody (Figure 3I), with many cells containing
positively stained cytoplasmic granules. Thus the biochemical changes described in Figures 1 and 2 accompanied the
morphologic remodeling in the cultures that led to the reestablishment of a mucous epithelium.
Kinetics of Induction of the Mucous Phenotype by RA
We wanted to establish the temporal relationship between
the induction of mucin genes, mucin secretion, the appearance of mucous cells, and suppression of the squamous cell
marker cornifin a, following treatment of RA-deficient cultures with a maximally effective concentration of RA. Therefore, RA-deficient cultures were treated on Day 10 with
1026 M RA, and at different time intervals the levels of cornifin a and mucin mRNAs, as well as the number of mucous cells and the amount of secreted mucin, were determined.
As shown in Figure 4A, cornifin a mRNA levels were
decreased within 12 h of RA treatment and remained suppressed thereafter. MUC2 mRNA was upregulated within
24 h and the level of expression continued to increase
thereafter. MUC5AC mRNA was not detectable until 48 h
after the start of treatment, and its level of expression continued to increase at subsequent time points. We also included MUC5B mRNA measurements in this study because this mucin was recently shown to be an important
component of human airway secretions (21). Interestingly,
MUC5B mRNA was not increased until 72 h after RA
treatment. Low levels of MUC5B mRNA were sometimes
seen in RA-free cultures. The control gene b2M was not
significantly affected by RA treatment. Our data show a
distinct time course of events induced by RA treatment:
The squamous marker cornifin a was suppressed first, followed by sequential expression of the three mucin genes.
This time-dependent change in cornifin a and mucin gene
expression was observed in two independent experiments.
47
As shown in Figure 4B, mucin secretion did not increase significantly above control levels until 72 h after addition of RA to the cultures. However, the number of cells
staining with the mucin antibody (Figure 4C) was markedly increased above that in control cultures within 24 h,
and more than 15% of the cells reacted with the antibody.
Thereafter the number of mucous cells continued to increase to nearly 40%. These results suggest that mucin accumulated in cells before significant amounts of mucin
were secreted onto the apical surface.
Enhancement of Mucin Gene Expression in
RA-Supplemented Cultures by High
Concentrations of RA
The foregoing experiments on the time-dependent effects
of RA treatment of RA-deficient cultures showed that between 24 and 72 h elapsed before the mRNA levels of the
three mucin genes were measurably increased. We wondered whether mucin gene expression could be more rapidly upregulated in RA-sufficient cultures maintained in
1028 M by raising the levels of RA to 1026 M.
NHTBE cells were grown in medium containing a basal
level of 1028 M RA. On Day 10 the RA concentration was
increased to 1026 M, and MUC2, MUC5AC, and MUC5B
mRNA levels were determined 6, 12, 24, 48, and 96 h later.
The results of the experiment are shown in Figure 5. Increasing the RA concentration from 1028 M to 1026 M increased MUC2 mRNA levels within 6 h (6-fold above untreated control cultures). Expression of MUC5AC and
MUC5B mRNA was not significantly increased by raising
the RA level 100-fold; mucin secretion was not increased
(data not shown). The control gene b2M was not affected.
The same results were obtained in a repeat experiment.
These data indicate that in cultures in which the cells are
already committed to express the mucous phenotype,
MUC2 mRNA levels are much more rapidly increased
than in RA-deficient cultures treated with the same concentration of RA.
Discussion
The importance of retinoids as regulators of cell differentiation has been recognized for many years (for review, see
Reference 34). Depending on the cell or tissue type, retinoids can either induce or inhibit the differentiation process. For instance, in the F9 embryonal teratocarcinoma
cells, RA induces differentiation to parietal endoderm
(35). In contrast, in epidermal keratinocytes, RA inhibits
differentiation (36).
Previously we showed that two mucin genes that are expressed in RA-sufficient NHTBE cultures, namely, MUC2
and MUC5AC, are not expressed when RA is omitted
from the medium (12). Such cultures are uniformly squamous metaplastic and secrete no mucin (15). In our study,
we describe the molecular and cellular changes during epithelial remodeling leading to the restoration of the mucous phenotype following RA treatment of RA-depleted
squamous cultures.
We used two markers to monitor changes in NHTBE
cell differentiation. Cornifin a was employed as a marker
of squamous differentiation and secreted mucin to moni-
48
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 20 1999
Figure 3. Histologic and immunocytologic effects of RA treatment on RA-deficient cultures. NHTBE cells cultured in RA-free medium
were treated either from Day 7 to Day 11 or from Day 10 to Day 14 with 1026 M RA. Cross-sections of intact cultures were stained with
H & E (A–F). For immunocytochemical analysis, the cultures were enzymatically dissociated and resulting cell suspensions were transferred to cytospin slides and reacted with mucin antibody (G–I). Histology of RA-deficient cultures on Days 7, 11, 10, and 14, respectively (A, B, D, and E). Histology of cultures treated with RA from Day 7 to Day 11 (C) and from Day 10 to Day 14 (F). Immunocytochemistry of cultures treated with RA from Day 10 to Day 14 (I). Bar 5 10 mm.
Koo, Yoon, Gray, et al.: Regulation of the Mucous Phenotype and Mucin Genes by RA
49
Figure 4. Time-dependent effects of RA treatment on RA-deficient cultures. NHTBE cells were cultured for 10 d in RA-deficient
medium and treated with 10 26 M RA for up to 96 h. (A) Total
RNA was isolated from the cultures at different time intervals, and
MUC2, MUC5AC, and MUC5B mRNA levels were determined by
quantitative RT–PCR. A control gene, b2M, was not affected by
RA treatment. Cornifin a mRNA levels by Northern blotting and
equal RNA loading were verified by staining of 28S rRNA with
ethidium bromide. (B) Apically secreted mucin was assayed by immunoblotting. Sham-treated cultures, gray bars; RA-treated cultures, black bars. Values are the means 6 SD of triplicate cultures.
(C) Percentage of mucin-positive cells was determined from
cytospin slides stained with mucin antibody. Sham-treated cultures, gray bars; RA-treated cultures, black bars. Statistical comparisons
were made by Student’s t test. This experiment was performed three times with similar results; one representative experiment is shown.
tor mucous cell differentiation. The antibodies used in this
study were previously shown to react with human mucin
but are not known to be directed against specific mucin
subtypes. Such mucin subtype-specific antibodies have
only recently been produced (19) but are not yet readily
available. To what extent changes in expression of specific
mucin genes will result in altered production of the corresponding gene products remains to be determined in the
future.
Our studies showed that RA-deficient cultures in advanced stages of squamous differentiation were much less
responsive to RA treatment than were cultures in early
stages of squamous differentiation. In late squamous cultures, 10 times as much RA was required to induce similar
levels of MUC2 and MUC5AC gene expression and to
suppress cornifin a than in early squamous cultures. However, the late squamous cultures were by no means completely RA resistant. Treatment with 1026 M RA resulted
in morphologic conversion to the mucous phenotype and
in induction of mucin mRNAs. The reasons for this difference in RA responsiveness between early and late fully
differentiated squamous cultures are presently not clear.
A comparison of the histomorphology of the early and late
RA-deficient cultures (compare Figure 3A with Figure
3D) suggests that the thick, stratified epithelium of the Day
10 cultures contains many more cells committed to squamous differentiation than the Day 7 cultures, in which only
a thin sheet of squamous cells overlays a row of seemingly
undifferentiated basal cells. A much more extensive process of epithelial remodeling, and thus a more intensive
RA treatment, is required to convert the late squamous
cultures to the mucous phenotype than to redirect differentiation in the early squamous cultures. It is also conceivable that retinoid receptors or other transcription factors
essential for mucin gene transcription are declining with
protracted retinoid deficiency, or that RA metabolism and
transport are altered in such a way in the late squamous
cultures that RA treatment is less efficient.
The kinetics of restoration of the mucous phenotype
triggered by RA treatment of RA-deficient cultures are
depicted in Figure 6. As shown, the squamous differentiation marker cornifin a was suppressed before any of the
50
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 20 1999
Figure 5. Effect of high-dose RA treatment on mucin gene expression in RA-sufficient cultures. NHTBE cells cultured for 10 d
in 1028 M RA (RA-sufficient cultures) were treated with 10 26 M
RA for up to 96 h. Total RNA was isolated from the cultures at
different time intervals, and MUC2, MUC5AC, and MUC5B
mRNA levels were determined by RT–PCR. A control gene,
b2M, was not affected by RA treatment. Data are representative
of three independent experiments. (Lane C) RNA from control
cultures maintained in 1028 M RA; (lane T) RNA from cultures
maintained in 1028 M RA for 10 d and treated with 10 26 M RA
for up to 4 d.
three mucin genes studied were induced. This may be an
indication that the first step in the conversion of the squamous to the mucous phenotype is downregulation of the
squamous differentiation program. The earliest sign that
mucous differentiation was being induced was the expression of the mucin gene MUC2 at 24 h and, simultaneously,
the appearance of a significant number of mucous cells.
Our studies support earlier in vivo experiments in vitamin
A-deprived hamsters, which showed that within 24 h after
the start of vitamin A treatment small PAS-positive vacuoles, presumed to be mucous granules, appeared in the
squamous lesions of the airways (5). Thus, in vivo as well
as in vitro remodeling of the vitamin A deficient metaplastic epithelium appears to commence soon after the start
of vitamin A treatment. MUC5AC mRNA was first expressed at 48 h and MUC5B at 72 h after addition of RA
to the cultures. Expression of the latter coincided with the
Figure 6. Schematic presentation of the sequence of events during restoration of mucous differentiation following RA treatment
of RA-deficient cultures. ↑, induction of gene expression; ↓, suppression of gene expression.
onset of apical secretion of mucin. We propose that this sequence of events following RA treatment of the NHTBE
cultures indicates progressive stages of mucous cell differentiation and maturation.
Upregulation of the mucin genes in RA-deficient cultures by RA treatment is not a typical “early” response
comparable to RARb receptor or Hox a1 induction, which
occurs within 2 or 8 h following RA treatment of hepatoma-derived cells or F9 teratocarcinoma cells (37, 38).
Rather, the induction of these mucin genes resembles the
delayed response of laminin b1, a high molecular weight
glycoprotein secreted by F9 cells treated with RA (39).
Upregulation of laminin b1 transcription by RA may involve not only binding of RARs to retinoic acid response
elements (RARE), but also activation of a homeodomain
gene (39), thus perhaps explaining the delayed response.
Because the MUC2 promoter as published by Velcich and
colleagues (40) and Gum and coworkers (41) does not
contain canonical RAREs, it is likely that the transcriptional activation of MUC2 expression also involves complex
regulatory mechanisms. Promoter sequences of MUC5AC
and MUC5B have not been published at the time of submission of this report. The RA time course studies suggest that
the induction of MUC2, MUC5AC, and MUC5B expression
may involve different mechanisms because the different
mucin genes are induced at different times after RA addition to the cultures. Previous studies in which we showed
that thyroid hormone preferentially inhibits MUC5AC expression (14) also suggest that MUC2 and MUC5AC are
differentially regulated.
In the RA-deficient cultures, the epithelium underwent
a massive change in its normal differentiation program, resulting in expression of the squamous phenotype with little
or no mucin gene expression. Under these conditions, at
least 24 h of RA treatment were required to induce any
mucin genes. We wanted to determine whether mucin
gene expression can be more rapidly upregulated when
cultures, in which the normal mucous phenotype is maintained by relatively low levels of RA (1028 M), are acutely
treated with high concentrations of RA. We found that
within 6 h of treatment with 1026 M RA, MUC2 transcript
levels were elevated. Because the epithelium in these cultures was already committed to mucous differentiation, cytoplasmic and nuclear factors needed for upregulation of
MUC2 mRNA must have been present before the RA
level was increased. Under these experimental conditions
MUC5AC and MUC5B mRNA levels were increased only
marginally by the high-dose RA treatment, possibly because expression levels were already near maximum in the
control cultures maintained in 1028 M RA. Other factors,
including tumor necrosis factor a and lipopolysaccharides
(LPS) have been shown to upregulate MUC2 mRNA rapidly in human lung carcinoma cells (18, 22). Our studies
showed that RA treatment can also rapidly upregulate
MUC2 mRNA in NHTBE cultures expressing the mucous
phenotype. In contrast, in vitamin A-deprived cultures,
the effect of RA treatment on mucin gene expression is a
delayed response. This raises the possibility that RA acts
via two different mechanisms: one that regulates MUC2
expression more directly through regulation of transcription or mRNA stability and one that involves changing a
Koo, Yoon, Gray, et al.: Regulation of the Mucous Phenotype and Mucin Genes by RA
differentiation program, as a result of which mucin genes
become expressed. Such a dual mechanism of action of
RA was also proposed for the effects of retinoids on keratinocyte differentiation and keratin gene expression (42).
Recent studies suggest that MUC5AC and MUC5B
mucins are major components of human airway secretions
(17, 19). MUC2, on the other hand, does not appear to be
a prominent airway mucin (17, 19). However, it has been
shown that MUC2 mRNA is expressed in goblet cells of
human bronchial (43) and nasal epithelium (44) and that
expression levels are upregulated in cultured airway epithelial cells by cytokines (22) and LPS (18). It is intriguing
that MUC2 is the first of the three mucin genes expressed
during restoration of the mucous phenotype and thus may
serve as an early marker of mucous differentiation. What
role it might play in the differentiation process of mucous
cells is presently unknown. It will be interesting to determine, when mucin subtype-specific antibodies become
available, whether intracellular MUC2 mucin can be detected in the early mucous cells appearing within the first
24 h after RA treatment. On the basis of immunohistochemistry or in situ hybridization studies, at least two
other mucins, MUC4 and MUC8, may also be components
of human airway secretions (45–47). We plan to include
these mucin genes in future studies of mucous differentiation and mucin gene expression.
Our studies showed that in the restoration of the mucous phenotype by RA treatment of RA-deficient NHTBE
cultures, induction of the mucin gene MUC2 and intracellular accumulation of mucin occurred first, and MUC5AC
and MUC5B were expressed at later stages of mucous differentiation. To further elucidate mechanisms by which retinoids regulate mucous differentiation and mucin gene
expression, studies are underway to determine which retinoid receptors are critical in the RA signaling pathway.
Acknowledgments: The authors thank Dr. J. St. George for the monoclonal antibody 17Q2 and Dr. C. William Davis for the monoclonal antibody H6C5.
They also gratefully acknowledge Dr. Lawrence Ostrowski and Dr. Kaya Andrews for critical reading of this manuscript. Part of this study was funded by
grant HL 36982 from the National Institutes of Health.
References
1. Mangelsdorf, D. J., K. Umesono, and R. M. Evans. 1994. The retinoid receptors. In The Retinoids: Biology, Chemistry, and Medicine. M. B. Sporn,
A. B. Roberts, and D. S. Goodman, editors. Raven Press, New York. 319–
349.
2. Mori, S. 1922. The changes in the para-ocular glands which follow the administration of diets low in fat-soluble A; with notes of the effect of the
same diets on the salivary glands and the mucosa of the larynx and trachea.
J. Hopkins Hosp. Bull. 33:357–359.
3. Wolbach, S. B., and P. R. Howe. 1926. Tissue changes following deprivation
of fat-soluble A vitamin. J. Exp. Med. 42:753–781.
4. McDowell, E. M., K. P. Keenan, and M. Huang. 1984. Effects of vitamin
A-deprivation on hamster tracheal epithelium. A quantitative morphologic study. Virchows Arch. B Cell Pathol. 45:197–219.
5. McDowell, E. M., K. P. Keenan, and M. Huang. 1984. Restoration of mucociliary tracheal epithelium following deprivation of vitamin A. A quantitative morphologic study. Virchows Arch. B Cell Pathol. 45:221–240.
6. Wong, Y. C., and R. C. Buck. 1971. An electron microscopic study of metaplasia of the rat tracheal epithelium in vitamin A deficiency. Lab. Invest.
24:55–66.
7. Wilhelm, D. L. 1954. Regeneration of the tracheal epithelium in the vitamin
A-deficient rat. J. Pathol. Bact. 67:361–365.
8. Wolbach, S. B., and P. R. Howe. 1933. Epithelial repair in recovery from vitamin A deficiency: an experimental study. J. Exp. Med. 57:511–526.
9. Clamon, G. H., M. B. Sporn, J. M. Smith, and U. Saffiotti. 1974. a- and bretinyl acetate reverse metaplasias of vitamin A deficiency in hamster tra-
51
chea in organ culture. Nature 250:64–66.
10. Clark, J. N., A. J. Klein-Szanto, A. H. Pine, K. B. Stephenson, and A. C.
Marchok. 1980. Reestablishment of a mucociliary epithelium in tracheal
organ cultures exposed to retinyl acetate: a biochemical and morphometric
study. Eur. J. Cell Biol. 21:261–268.
11. Sporn, M. B., G. H. Clamon, N. M. Dunlop, D. L. Newton, J. M. Smith, and
U. Saffiotti. 1975. Activity of vitamin A analogues in cell cultures of mouse
epidermis and organ cultures of hamster trachea. Nature 253:47–50.
12. Guzman, K., T. Bader, and P. Nettesheim. 1996. Regulation of MUC5 and
MUC1 gene expression: correlation with airway mucous differentiation.
Am. J. Physiol. 270(Lung Cell. Mol. Physiol. 14):L846–L853.
13. Guzman, K., T. E. Gray, J. H. Yoon, and P. Nettesheim. 1996. Quantitation
of mucin RNA by PCR reveals induction of both MUC2 and MUC5AC
mRNA levels by retinoids. Am. J. Physiol. 271(Lung Cell. Mol. Physiol.
15):L1023–L1028.
14. Yoon, J. H., T. Gray, K. Guzman, J. S. Koo, and P. Nettesheim. 1997. Regulation of the secretory phenotype of human airway epithelium by retinoic
acid, triiodothyronine, and extracellular matrix. Am. J. Respir. Cell Mol.
Biol. 16:724–731.
15. Gray, T. E., K. Guzman, C. W. Davis, L. H. Abdullah, and P. Nettesheim.
1996. Mucociliary differentiation of serially passaged normal human tracheobronchial epithelial cells. Am. J. Respir. Cell Mol. Biol. 14:104–112.
16. Marvin, K. W., M. D. George, W. Fujimoto, N. A. Saunders, S. H. Bernacki,
and A. M. Jetten. 1992. Cornifin, a cross-linked envelope precursor in keratinocytes that is down-regulated by retinoids. Proc. Natl. Acad. Sci. USA
89:11026–11030.
17. Hovenberg, H. W., J. R. Davies, A. Herrmann, C. J. Linden, and I. Carlstedt. 1996. MUC5AC, but not MUC2, is a prominent mucin in respiratory
secretions. Glycoconj. J. 13:839–847.
18. Li, J. D., A. F. Dohrman, M. Gallup, S. Miyata, J. R. Gum, Y. S. Kim, J. A.
Nadel, A. Prince, and C. B. Basbaum. 1997. Transcriptional activation of
mucin by Pseudomonas aeruginosa lipopolysaccharide in the pathogenesis
of cystic fibrosis lung disease. Proc. Natl. Acad. Sci. USA 94:967–972.
19. Thornton, D. J., I. Carlstedt, M. Howard, P. L. Devine, M. R. Price, and
J. K. Sheehan. 1996. Respiratory mucins: identification of core proteins
and glycoforms. Biochem. J. 316:967–975.
20. Hovenberg, H. W., J. R. Davies, and I. Carlstedt. 1996. Different mucins are
produced by the surface epithelium and the submucosa in human trachea:
identification of MUC5AC as a major mucin from the goblet cells. Biochem. J. 318:319–324.
21. Thornton, D. J., M. Howard, N. Khan, and J. K. Sheehan. 1997. Identification of two glycoforms of the MUC5B mucin in human respiratory mucus.
Evidence for a cysteine-rich sequence repeated within the molecule. J.
Biol. Chem. 272:9561–9566.
22. Levine, S. J., P. Larivee, C. Logun, C. W. Angus, F. P. Ognibene, and J. H.
Shelhamer. 1995. Tumor necrosis factor-a induces mucin hypersecretion
and MUC-2 gene expression by human airway epithelial cells. Am. J.
Respir. Cell Mol. Biol. 12:196–204.
23. Temann, U. A., B. Prasad, M. W. Gallup, C. Basbaum, S. B. Ho, R. A. Flavell, and J. A. Rankin. 1997. A novel role for murine IL-4 in vivo: induction of MUC5AC gene expression and mucin hypersecretion. Am. J.
Respir. Cell Mol. Biol. 16:471–478.
24. Steiger, D., J. Hotchkiss, L. Bajaj, J. Harkema, and C. Basbaum. 1995. Concurrent increases in the storage and release of mucin-like molecules by rat
airway epithelial cells in response to bacterial endotoxin. Am. J. Respir.
Cell Mol. Biol. 12:307–314.
25. Lin, H., D. M. Carlson, J. A. St. George, C. G. Plopper, and R. Wu. 1989.
An ELISA method for the quantitation of tracheal mucins from human
and nonhuman primates. Am. J. Respir. Cell Mol. Biol. 1:41–48.
26. St. George, J. A., D. L. Cranz, S. C. Zicker, J. R. Etchison, D. L. Dungworth, and C. G. Plopper. 1985. An immunohistochemical characterization
of rhesus monkey respiratory secretions using monoclonal antibodies. Am.
Rev. Respir. Dis. 132:556–563.
27. Thornton, D. J., J. R. Davies, M. Kraayenbrink, P. S. Richardson, J. K.
Sheehan, and I. Carlstedt. 1990. Mucus glycoproteins from “normal” human tracheobronchial secretion. Biochem. J. 265:179–186.
28. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Extraction, purification,
and analysis of messenger RNA from eukaryotic cells. In Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring
Harbor, NY. 7.1–7.52.
29. Gum, J. R., Jr., J. W. Hicks, N. W. Toribara, B. Siddiki, and Y. S. Kim. 1994.
Molecular cloning of human intestinal mucin (MUC2) cDNA. Identification of the amino terminus and overall sequence similarity to prepro-von
Willebrand factor. J. Biol. Chem. 269:2440–2446.
30. Meezaman, D., P. Charles, E. Daskal, M. H. Polymeropoulos, B. M. Martin,
and M. C. Rose. 1994. Cloning and analysis of cDNA encoding a major airway glycoprotein, human tracheobronchial mucin (MUC5). J. Biol. Chem.
269:12932–12939.
31. Desseyn, J. L., V. Guyonnet Duperat, N. Porchet, J. P. Aubert, and A.
Laine. 1997. Human mucin gene MUC5B, the 10.7-kb large central exon
encodes various alternate subdomains resulting in a super-repeat. Structural evidence for a 11p15.5 gene family. J. Biol. Chem. 272:3168–3178.
32. Siebert, P. D., and J. W. Larrick. 1993. PCR MIMICS: competitive DNA
52
33.
34.
35.
36.
37.
38.
39.
40.
AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 20 1999
fragments for use as internal standards in quantitative PCR. Biotechniques
14:244–249.
Siebert, P. D., and J. W. Larrick. 1992. Competitive PCR. Nature 359:557–558.
Gudas, L. J., M. B. Sporn, and A. B. Roberts. 1994. Cellular biology and
biochemistry of the retinoids. In The Retinoids: Biology, Chemistry, and
Medicine. M. B. Sporn, A. B. Roberts, and D. S. Goodman, editors. Raven
Press, New York. 443–520.
Strickland, S., and V. Mahdavi. 1978. The induction of differentiation in teratocarcinoma stem cells by retinoic acid. Cell 15:393–403.
Jetten, A. M., C. Nervi, and T. M. Vollberg. 1992. Control of squamous differentiation in tracheobronchial and epidermal epithelial cells: role of retinoids. Monogr. Natl. Cancer Inst. 13:93–100.
de The, H., A. Marchio, P. Tiollais, and A. Dejean. 1989. Differential expression and ligand regulation of the retinoic acid receptor a and b genes.
EMBO J. 8:429–433.
LaRosa, G. J., and L. J. Gudas. 1988. An early effect of retinoic acid: cloning of an mRNA (Era-1) exhibiting rapid and protein synthesis-independent induction during teratocarcinoma stem cell differentiation. Proc.
Natl. Acad. Sci. USA 85:329–333.
Li, C., and L. J. Gudas. 1996. Murine laminin B1 gene regulation during the
retinoic acid- and dibutyryl cyclic AMP-induced differentiation of embryonic F9 teratocarcinoma stem cells. J. Biol. Chem. 271:6810–6818.
Velcich, A., L. Palumbo, L. Selleri, G. Evans, and L. Augenlicht. 1997. Organization and regulatory aspects of the human intestinal mucin gene
(MUC2) locus. J. Biol. Chem. 272:7968–7976.
41. Gum, J. R., J. W. Hicks, and Y. S. Kim. 1997. Identification and characterization of the MUC2 (human intestinal mucin) gene 59-flanking region:
promoter activity in cultured cells. Biochem. J. 325:259–267.
42. Tomic, M., C. K. Jiang, H. S. Epstein, I. M. Freedberg, H. H. Samuels, and
M. Blumenberg. 1990. Nuclear receptors for retinoic acid and thyroid hormone regulate transcription of keratin genes. Cell Regulation 1:965–973.
43. Dohrman, A., T. Tsuda, E. Escudier, M. Cardone, B. Jany, J. Gum, Y. Kim,
and C. Basbaum. 1994. Distribution of lysozyme and mucin (MUC2 and
MUC3) mRNA in human bronchus. Exp. Lung Res. 20:367–380.
44. Aust, M. R., C. S. Madsen, A. Jennings, J. L. Kasperbauer, and S. J. Gendler. 1997. Mucin mRNA expression in normal and vasomotor inferior turbinates. Am. J. Rhinol. 11:293–302.
45. Porchet, N., V. C. Nguyen, J. Dufosse, J. P. Audie, V. Guyonnet Duperat,
M. S. Gross, C. Denis, P. Degand, A. Bernheim, and J. P. Aubert. 1991.
Molecular cloning and chromosomal localization of a novel human tracheobronchial mucin cDNA containing tandemly repeated sequences of
48 base pairs. Biochem. Biophys. Res. Commun. 175:414–422.
46. Shankar, V., P. Pichan, R. L. Eddy, Jr., V. Tonk, N. Nowak, S. N. Sait, T. B.
Shows, R. E. Schultz, G. Gotway, R. C. Elkins, M. S. Gilmore, and G. P.
Sachdev. 1997. Chromosomal localization of a human mucin gene
(MUC8) and cloning of the cDNA corresponding to the carboxy terminus.
Am. J. Respir. Cell Mol. Biol. 16:232–241.
47. Reid, C. J., S. Gould, and A. Harris. 1997. Developmental expression of mucin genes in the human respiratory tract. Am. J. Respir. Cell Mol. Biol.
17:592–598.