REGULATION OF RARb EXPRESSION BY RAR- AND RXR-SELECTIVE
RETINOIDS IN HUMAN LUNG CANCER CELL LINES: EFFECT
ON GROWTH INHIBITION AND APOPTOSIS INDUCTION
Yin LI1, Marcia I. DAWSON2**, Anissa AGADIR1, Mi-Ock LEE1, Ling JONG2, Peter D. HOBBS2 and Xiao-kun ZHANG1*
1The Burnham Institute, La Jolla Cancer Research Center, La Jolla, CA, USA
2Retinoid Program, SRI International, Menlo Park, CA, USA
Retinoids regulate the growth and differentiation of human
tracheobronchial epithelial cells. In this study, we investi-
gated the effects of all-trans-retinoic acid (trans-RA) and
receptor class-selective retinoids on the growth and apopto-
sis of human lung cancer cell lines. Trans-RA significantly
inhibited the growth of Calu-6 and H460 cells, accompanied
by induction of RA receptor (RAR)b expression. In contrast,
it had little effect on the growth of H292, SK-MES-1 and H661
lung cancer cell lines, in which RARb expression was not
induced. Stable expression of RARb in RARb-negative, trans-
RA-resistant SK-MES-1 and H661 lung cancer cells led to
recovery of trans-RA-induced growth inhibition, which oc-
curred, however, only at low serum concentration. Using
fluorescent microscopy and the terminal deoxyribonucleo-
tidyl transferase (TdT) assay, we demonstrated that induc-
tion of apoptosis by trans-RA contributed to its growth-
inhibitory effect in trans-RA-sensitive lung cancer cell lines.
Analysis of RAR-selective and retinoid X receptor (RXR)-
selective retinoids showed that activation of both RARs and
RXRs could induce growth inhibition in trans-RA-sensitive
lung cancer cells. Also, an additive synergistic effect on
growth inhibition and RARb induction was observed when
cells were treated with combinations of RAR-selective and
RXR-selective retinoids. Together, our results show that
expression of RARb plays a role in mediating retinoid re-
sponse in lung cancer cells and that activation of RARs or
RXRs contributes to induction of RARb, growth inhibition
and apoptosis by retinoids. Int. J. Cancer 75:88–95, 1998.
r
1998 Wiley-Liss, Inc.
Retinoids, the natural vitamin A derivatives and synthetic
analogs, regulate a broad range of biological processes, including
growth, differentiation and development (Lotan, 1981). They are
currently used in the treatment of epithelial cancer and promyelo-
cytic leukemia and are being evaluated as preventive and therapeu-
tic agents for a variety of other human cancers (Lotan, 1981). The
natural retinoids, trans-retinoic acid (trans-RA) and 9-cis-RA,
interact with 2 nuclear receptor classes: the retinoic acid receptors
(RARs) and the retinoid X receptors (RXRs) (Zhang and Pfahl,
1993), which function as ligand-inducible transcription factors.
Trans-RA binds and activates the RARs, whereas 9-cis-RA binds
and activates both receptor classes. Both RARs and RXRs have 3
subtypes, a, b and g, which have distinct patterns of expression in
development and differentiation, suggesting that each may have
discrete functions. RARs and RXRs modulate the expression of
their target genes, including those for the RARs, by interacting as
heterodimers with RA response elements (RAREs) (Zhang and
Pfahl, 1993). A RARE (bRARE) in the RARb promoter mediates
RA-induced RARb gene expression in many cell types (Hoffmann
et al., 1990). Thus, autoregulation of the RARb gene may play a
role in amplifying RA responses, which is often lost as cells
become neoplastic.
Retinoids regulate the proliferation and differentiation of tracheo-
bronchial epithelial cells in vitro and in vivo. In lung cancer cells,
normal muco-ciliary differentiation of tracheobronchial epithelial
cells is either absent or blocked or occurs inappropriately and
incompletely, suggesting that the normal retinoid pathway may be
perturbed in lung cancer cells. The mechanism by which retinoid
response is impaired in lung cancer cells remains largely unknown.
It has been reported that expression of RARb is lost or reduced in
human primary lung cancers and human lung cancer cell lines
(Gebert et al., 1991; Nervi et al., 1991; Houle et al., 1993; Geradts
et al., 1993; Zhang et al., 1994). Trans-RA is able to induce RARb
expression in breast cancer cells (Liu et al., 1996). This effect is
associated with the growth inhibition observed in these cells after
trans-RA treatment (Liu et al., 1996). In addition, an increase of
RARb levels is correlated with the clinical response of oral
leukoplakia patients to 13-cis-RA (Lotan et al., 1995). Further-
more, stable expression of RARb reduced the tumorigenicity of
RARb-negative Calu-1 lung cancer cells (Houle et al., 1993) and
allowed growth inhibition by trans-RA in RARb-negative breast
cancer cell lines (Liu et al., 1996; Li et al., 1995; Seewaldt et al.,
1995). Unfortunately, trans-RA does not induce RARb expression
in most lung cancer cell lines (Zhang et al., 1994). The failure of
trans-RA to up-regulate RARb expression suggests that lung
cancer cells may have lost most, if not all, of their responsiveness to
trans-RA. Loss of RARb gene up-regulation may represent a
mechanism by which lung cancer cells escape normal growth
control by retinoids.
Response of lung cancer cells to retinoids is modulated by
additional factors. We previously observed that trans-RA did not
induce RARb expression in several lung cancer cell lines despite
functional retinoid receptor expression (Zhang et al., 1994). In a
lung carcinogenesis model, human bronchial epithelial cells were
refractory to trans-RA, though retinoid receptor DNA binding and
transcriptional activation were intact (Kim et al., 1995). We have
shown that the orphan receptors nur77 and COUP-TF are involved
in regulating retinoid response in lung cancer cells by modulating
RARE-binding activity (Wu et al., 1997). Because nur77 expres-
sion is highly induced by growth factors (Hazel et al., 1988) and is
associated with retinoid resistance in lung cancer cells (Wu et al.,
1997), growth-factor signaling may have a role in regulating
retinoid activity.
In this study, we investigated the effect of trans-RA and receptor
class-selective retinoids on the growth of human non-small-cell
lung cancer (NSCLC) cell lines which express different levels of
RARb. Our results demonstrate a correlation between trans-RA-
induced RARb expression and growth inhibition in several human
lung cancer cell lines. Furthermore, expression of RARb in
RARb-negative, trans-RA-resistant lung cancer cell lines allowed
growth inhibition by trans-RA. Using receptor-selective retinoids,
we found that both RAR- and RXR-signaling pathways play a
critical role in modulating retinoid response in lung cancer cells
through apoptosis induction. Both RAR and RXR class-selective
retinoids inhibit the growth of trans-RA-sensitive NSCLC cell
Contract grant sponsor: National Cancer Institute; Contract grant num-
bers: R01 CA60988; P01 CA51993; Contract grant sponsor: University of
California.
*Correspondence to: The Burnham Institute, La Jolla Cancer Research
Center, 10901 N. Torrey Pines Road, La Jolla, CA 92037, USA. Fax: (619)
646-3195. E-mail: xzhang@ljcrf.edu **or to: SRI, 333 Ravenswood
Avenue, Menlo Park, CA 94025, USA. Fax: (415) 859-3153. E-mail:
marcia_dawson@qm.sri.com
Received 13 May 1997; Revised 23 August 1997
Int. J. Cancer: 75, 88–95 (1998)
r
1998 Wiley-Liss, Inc.
Publication of the International Union Against Cancer
Publication de l’Union Internationale Contre le Cancer

lines, and growth inhibition by RAR-selective retinoids is associ-
ated with their ability to up-regulate RARb expression. In addition,
RXR-selective retinoid, which had no effect on RARb expression
by itself, enhanced lung cancer cell responsiveness to RAR-
selective retinoids.
MATERIAL AND METHODS
Retinoids
Trans-RA was obtained from Sigma (St. Louis, MO). 2-(4-
Carboxyphenyl)-2-(5,6,7,8-tetrahydro-5,5,8,8-tetramethylnaphtha-
len-2-yl)- 1,3-dioxalane (SR11237) and 4-[(5,6,7,8-tetrahydro-
5,5,8,8-tetramethylnaphthalen-2-yl)-cyclopropyl]benzoic acid
(SR11246) were prepared as described by Dawson et al. (1995).
(E)-4-[3-(5,6,7,8-Tetrahydro-5,5,8,8-tetramethylnaphthalen-2-yl)-
3-o xopropenyl] benzoic acid (SR11383) was prepared by the
method of Kagechika et al. (1988).
4-(5,6,7,8-Tetrahydro-3-hydroxy-5,5,8,8-tetramethylnaphthalen-
2-ylcarboxamido) benzoic acid (SR11281) was synthesized as
follows: (i) 1-(5,6,7,8-tetrahydro-3-hydroxy-5,5,8,8-tetramethyl-
naphthalen-2-yl)ethanone was prepared by the method of Kagechika
et al. (1988) by Fries re-arrangement (AlCl3, 130°C; 90% yield) of
2-acetoxy-5,6,7,8-tetrahydro-5,5,8,8-tetramethylnaphthalene. (ii)
Protection of the phenolic group as the benzyl ether (benzyl
bromide, K2CO3, acetone, reflux; 92% yield) and oxidation of the
acetyl group (NaOCl, EtOH, reflux; 30% yield) gave 3-benzyloxy-
5,6,7,8-tetrahydro-5,5,8,8-tetramethylnaphthalene-2-carboxylic
acid. (iii) The carboxylic acid was converted (oxalyl chloride,
CH2Cl2) to the acyl chloride and treated (pyridine/benzene) with
ethyl 4-aminobenzoate to yield (95%) the benzamide. (iv) Ester
hydrolysis (NaOH, aqueous EtOH, 25°C, aqueous HCl; 97% yield)
and hydrogenolysis (H2, Pd[C], EtOH, 25°C) of the benzyl
ether-protecting group afforded SR11281 (93% yield): m.p. 275°–
278°C; 1H NMR (300 MHz, CDHCL3) d 1.29 (s, 6, CMe2), 1.32 (s,
6, CMe3), 1.69 (s, 4, CH2CH2), 6.93 (s, 1, ArH), 7.69 (s, 1, ArH),
7.74 (d, 2, J 5 8.8 Hz, ArH), 8.08 (d, 2, J 5 8.8 Hz, ArH); IR (KBr)
3330, 1686, 1530, 1419, 1174 cm21.
Cell culture
NSCLS cell lines Calu-6, H292, H661, H460 and SK-MES-1
were obtained from the ATCC (Rockville, MD). Calu-6 and
SK-MES-1 cells were grown in MEM supplemented with 10%
FCS; H292, H661 and H460 cells were grown in RPMI 1640
supplemented with 10% FCS.
Growth inhibition
For growth-inhibition studies, cells were seeded at 1,000 to
2,000 cells per well in 96-well plates at the indicated concentra-
tions of delipidized FCS and treated 24 hr later with 1028 to 1026 M
retinoids for 7 days. Control cells received ethanol alone. Media
and retinoids were changed every 48 hr. Viable cell number was
determined using the MTT cell proliferation/cytotoxicity assay
(Promega, Madison, WI), in which the capacity of cells to convert a
tetrazolium salt to blue formazan was measured (Liu et al., 1996).
Results obtained were confirmed by cell counting with a hemocy-
tometer.
Apoptosis analyses
Analyses for DNA fragmentation, nuclear morphological change
and DNA end-labeling were as previously described (Liu et al.,
1996). For ELISA, about 2 3 104 cells were treated with 1026 M
trans-RA for 2 days. DNA fragmentation was measured using a cell
death detection ELISA kit (Boehringer-Mannheim, Mannheim,
Germany) following the manufacturer’s protocol. Results were
expressed relative to untreated cells. For nuclear morphological
change analysis, 1026 M trans-RA-treated or untreated cells were
trypsinized, washed with PBS, fixed with 3.7% paraformaldehyde
and stained with 50 µg/ml of DAPI (4,6-diamidino-2-phenylindole)
containing 100 µg/ml DNase-free RNase A to visualize the nuclei.
Stained cells were examined by fluorescent microscopy. For DNA
end-labeling, cells were treated with or without 1026 M trans-RA,
trypsinized 24 hr later, washed with PBS, fixed in 1% formalde-
hyde in PBS for 15 min, washed with PBS, resuspended in 70%
ice-cold ethanol and immediately stored at 220°C overnight. Cells
were labeled with biotin-16-dUTP by terminal transferase and
stained with avidin-FITC (Boehringer-Mannheim). Fluorescently
labeled cells were analyzed using a FACScater-Plus (Becton
Dickinsons, Mountain View, CA). Representative histograms are
shown.
RNA preparation and Northern blot analysis
Total RNA samples were prepared by the guanidine hydrochlo-
ride/ultracentrifugation method (Liu et al., 1996). About 30 µg of
total RNA from each cell line were fractionated on a 1% agarose
gel, transferred to nylon filters and probed with the 32P-labeled
ligand-binding domain of RARb cDNA, as previously described
(Liu et al., 1996). To normalize the amount of RNA used, filters
were probed with b-actin.
Stable transfection
cDNA for the RARb gene was cloned into the pRc/CMV
expression vector (InVitrogen, San Diego, CA) as described (Liu et
al., 1996). The resulting recombinant expression vector was stably
transfected into lung cancer cells using the calcium phosphate
precipitation method and screened using G418 (GIBCO BRL,
Grand Island, NY). Integration and expression of RARb were
determined by Southern and Northern blots, respectively.
RESULTS
RARb up-regulation correlates with trans-RA-induced growth
inhibition in human lung cancer cell lines
Because NSCLC cell lines and tumor tissues have abnormally
low levels of RARb (Gebert et al., 1991; Houle et al., 1993; Nervi
et al., 1991; Geradts et al., 1993; Zhang et al., 1994), we examined
whether RARb expression is related to the sensitivity of lung
cancer cell lines to trans-RA-induced growth inhibition. Calu-6,
H460, SK-MES-1, H292 and H661 lung cancer cell lines were
treated with trans-RA (1028 to 1026 M) for 7 days and analyzed for
cell viability by the MTT assay. Our results demonstrated that
Calu-6 cells were the most sensitive to the trans-RA growth-
inhibitory effect, with 84% inhibition when they were treated with
1026 M trans-RA, whereas H460 cells displayed a moderate
response, with 50% inhibition (Fig. 1a). The other cell lines were
relatively insensitive, with only about 20% inhibition.
We next determined RARb expression in these lung cancer cell
lines by Northern blot analysis. The lung cancer cell lines were
treated with and without 1026 M trans-RA for 24 hr and evaluated
for RARb expression (Fig. 1b). In the absence of trans-RA, RARb
was highly expressed in trans-RA-sensitive H460 and trans-RA-
resistant H292 cells, suggesting that RARb expression alone is not
sufficient to confer growth inhibition by trans-RA. RARb tran-
scripts were hardly detected in the other untreated cell lines.
However, they were induced by trans-RA in trans-RA-sensitive
Calu-6 and H460 cell lines but not in the trans-RA-resistant ones
(H292, SK-MES-1 and H661). Thus, up-regulation of RARb by
trans-RA correlates with its growth-inhibitory effect in retinoid-
sensitive lung cancer cell lines.
Recovery of trans-RA sensitivity by stable expression of RARb
in trans-RA-resistant lung cancer cells
Trans-RA at 1026 M was a poor inhibitor of H661 (20%) and
SK-MES-1 (17%) cell growth and did not induce RARb expres-
sion (Fig. 1). To investigate whether the resistance of these cell
lines to trans-RA was due to their inability to express RARb, we
stably expressed RARb in these cell lines. Two H661-stable clones
(H661/RARb1 and H661/RARb4) and 2 SK-MES-1-stable clones
(SK-MES-1/RARb6 and SK-MES-1/RARb7), which expressed
89RARb EXPRESSION, REGULATION AND APOPTOSIS