Evidence that Retinoic Acid Receptor β Induction by Retinoids is Important for Tumor Cell Growth Inhibition

Retinoic acid receptor beta (RARbeta) is thought to be involved in suppressing cell growth and tumorigenicity. Many premalignant and malignant cells exhibit a reduced RARbeta expression. However, in some of these cells (e.g. H157 human squamous cell carcinoma cells), RARbeta can be induced by retinoids (e.g. all-trans-retinoic acid, ATRA) because its promoter contains a retinoic acid response element. To examine the hypothesis that RARbeta induction is important for inhibition of cell proliferation by retinoids, we blocked ATRA-induced RARbeta expression in H157 cells using a retroviral vector harboring multiple copies of antisense RARbeta2 sequences. Antisense RARbeta-transfected cells showed not only decreased expression of ATRA-induced RARbeta protein but also reduced ATRA-induced RARE binding activity and transactivation. Importantly, all antisense RARbeta transfectants of H157 cells were less responsive than vector-transfected cells to the growth inhibitory effects of the retinoids ATRA and Ch55 in vitro. These results demonstrate that RARbeta induction may play an important role in mediating growth inhibitory effects of retinoids in cancer cells.


5
RARβ induction by retinoids has been demonstrated in vivo (16,27,28). However, it is not clear to what extent if any the induced RARβ contributes to the response to growth inhibitory effects of retinoids, or whether it plays no role in the overall response to retinoids.
To further understand the importance of RARβ induction, we constructed and transfected a retroviral expression vector harboring antisense RARβ2 into the H157 human lung squamous cell carcinoma cell line, which expresses RARβ only after ATRA treatment and found that blocking RARβ induction decreases cell sensitivity to retinoids.

EXPERIMENTAL PROCEDURES
Retinoids-ATRA and Ch55 were kindly provided by Dr. Werner Bollag (F. Hoffmann-La Roche, Basel, Switzerland) and Dr. Koichi Shudo (Faculty of Pharmaceutical Sciences, University of Tokyo, Tokyo, Japan), respectively. They were dissolved in dimethylsuloxide at a concentration of 10 mM under N 2 and stored in the dark at -80•C. Stock solutions were diluted to the desired concentrations with growth medium just prior to use. 6 from Dr. D. Miller (Fred Hutchinson Cancer Research Center, Seattle, WA) (29), was used in this study. We introduced into LNSX vector multiple copies of hRARβ2 cDNA fragments corresponding to the initial site of RARβ2 translation in an antisense orientation by ligating 385-bp Bam H I/Sph I cDNA fragments released from pSG5-RARβ plasmid (30) into unique Hind III, Cla I, and BamH I sites through blunt end ligation with T4 DNA ligase, respectively. We obtained a series of vectors harboring different number of copies of RARβ2 cDNA inserts in an antisense orientation as identified by sequencing and enzymatic digestions. Vectors harboring 1, 2, 3, 4, 5, and 6 antisense RARβ2 inserts were designated as LNASβ, LNASβΙΙ, LNASβΙΙΙ, LNASβIV, LNASβV and LNASβVI, respectively. The LNASβVI shown in Fig. 1 was used in all subsequent experiments. Retroviral Transduction-LNASβVI or LNSX as a vector control were transfected directly into PA317 amphotropic packaging cell line by the calcium phospahte precipitation method (31).
The transfected cells were cultured in G418 (500 µg/ml), and individual clones of resistant cells were picked up using cloning cylinders after 14 days and expanded. The supernatants from these retroviral producers were titered on thymidine kinase-negative NIH3T3 target cells as described previously (31). Supernatant from the highest titer producer clone for either LNASβVI or LNSX (up to 5 x 10 5 /ml) was used in all subsequent experiments. H157 cells were plated at 1:10 split in 6 cm diameter tissue culture dishes (Corning Incorporated, Corning, NY). On the second day, the medium was replaced with 2 ml of fresh medium containing different amounts of viral supernatant and 8 µg/ml of polybrene, and 2 h later, 7 15 days using cloning cylinders. The rest of the clones from different dishes were pooled as pool transfectants. Transfectants were expanded and maintained under continuous G418 selection at 500 µg/ml.

RNA Purification and Northern
Blotting-Total cellular RNA purification and Northern blotting were performed as previously described (32). Protein Extraction and Western Blotting-Nuclear extracts were prepared from H157 transfectants by a method described by Pollock et al (33). Eighty µg of protein was electrophoresed through a 10% polyacryamide slab gel and transferred to nitrocellulose membranes (Bio-Rad, The blot was then washed four more times with blocking solution containing 0.1% Tween 20, and exposed to X-ray film at -80•C for 2 to 5 days. plasmid purification, transfection, and luciferase activity assay procedures were the same as described previously (34).
Cell Growth Assay-The effects of retinoids on the growth of different transfectants were evaluated by the sulforhodamine B assay as previously described (32). Colony formation assay was performed as follows: cells were plated at a density of 2 000 cells in 6 cm diameter tissue culture dishes (Corning Incorporated) and treated on the next day with retinoids. The medium was replaced with fresh medium containing retinoids every 3 days. After a 12-day treatment, colonies were stained with 0.5% methylene blue in 70% ethanol and counted.

Establishment of H157 Cells Stably Expressing Antisense RARβ by Retroviral-mediated
Transduction-H157 cells express undetectable levels of RARβ (by Northern blotting) but high level of induced RARβ after treatment with ATRA or other retinoids. This cell line was infected with the LNASβVI retroviral vector (Fig. 1), and individual G418-resistant clones were isolated.
These clones were designated H157-LNASβ. Vector only control clones were obtained by infecting the same cell line with the retroviral vector LNSX, which does not contain the antisense RARβ sequence (Fig. 1). These control clones were designated H157-LNSX.
Northern blotting was performed on transfectants to determine the expression level of antisense RARβ mRNA. The 4.3 kb antisense mRNA controlled by the SV40 promoter and the antisense mRNA larger than 5 kb driven by the LTR promoter were detected using a 385 bp BamH I/Sph I RARβ2 cDNA fragment as a probe in all the LNASβVI-infected clones but not in the LNSX-infected control cells ( Fig. 2A). RXRs. Fig. 3B shows that RARγ and at least one RXR protein and traces of RARβ can be supershifted from complexes with RARE in LNASβ-9 cells after ATRA treatment. It is plausible to suggest that these receptors are responsible for the activation of RARE-Luc reporter in the antisense RARβ transfectants The partial suppression of RARE transactivation in the LNASβ transfected cells appeared to be specific for RARE because the transactivation of another reporter construct mediated by AP-1 binding was not suppressed in these transfectants (Fig. 3C).

Antisense RARβ Transfection Decreases Cell Responsiveness to Retinoid Treatment-We
next compared the responsiveness of LNASβVI-transfected clones to retinoid treatment with that of LNSX-transfected cells. Fig. 4A shows the effects of different retinoids on the population growth of LNASβVI-transfected and LNSX-transfected cells. Four LNASβVI-transfected clones exhibited much lower sensitivity to ATRA treatment than 3 LNSX-transfected cells did. ATRA at concentrations of 1 µM and 2.5 µM caused 20-40% growth inhibition in LNSX-transfected cells but less than 20% growth inhibition in all LNASβ VI-transfected clones (Fig. 4A). The synthetic retinoid Ch55 is a pan RAR-selective agonist and has better receptor binding affinities, especially to RARβ than ATRA but is much more active that ATRA in inhibiting the growth of H157 and other lung cancer cells (32). To better demonstrate the low responsiveness of antisense RARβ- respectively, in LNSX-transfected cells but by less than 20% and 35%, respectively, in LNASβVItransfected cells after 6-day treatment (Fig. 4A). In addition, we analyzed the effects of ATRA and Ch55 on the colony fomation of different transfectants. Similar to the population growth inhibition results, all LNASβVI-transfected cells exhibited less sensitivity (about 50%) than LNSXtransfected cells to inhibition of colony formation by ATRA and Ch55 (Fig. 4B). These results clearly show that blockage of RARβ induction decreased the cell responsiveness to retinoids.

DISCUSSION
It has been suggested that RARβ may play a role as a tumor suppressor. This hypothesis was based on the observation that RARβ levels decreased in a variety of tumor cell lines including lung carcinomas (9-13) as well is in premalignant and malignant epithelial tissues in vivo (14-17, 27, 28). It has been shown that RARβ expression, which is suppressed at early stages of head and neck and lung carcinogenesis, can be induced by retinoid treatment (27,28,36). The induction of RARβ by retinoids is not surprising because the RARβ gene promoter contains a RARE (5-8). However, it was not clear whether this induction is important for the overall effect of ATRA on cell phenotype, namely whether RARβ increase plays a role in growth inhibition. Our study has addressed this question by blocking RARβ induction by ATRA using a retroviral vector harboring antisense RARβ. The results show clearly that the antisense vector was effective in decreasing the 13 level of the RARβ protein and that this was accompanied by a decrease in the response of the transfected cells to two RAR-selective retinoids (ATRA and Ch55). Thus, our data support the conclusion that the induction of RARβ by retinoids may be an early step in the cascade of events leading to growth inhibition.
We used the H157 lung carcinoma cells, which do not express RARβ constitutively but can be induced to express this receptor after ATRA treatment. These cells were, therefor, a good system to explore the role of RARβ induction in mediating the growth inhibitory effects of retinoids. We used for the first time antisense RARβ to block induction of RARβ by retinoids.   Nuclear extracts were prepared and subjected to Western blot anaylsis 24 h after the last feeding as described under "Experimental Procedures". NS, non-specific band.