Identification and characterization of a retinoid-induced class II tumor suppressor/growth regulatory gene

Abstract

Retinoids, synthetic and natural analogs of retinoic acid, exhibit potent growth inhibitory and cell differentiation activities that account for their beneficial effects in treating hyperproliferative diseases such as psoriasis, actinic keratosis, and certain neoplasias. Tazarotene is a synthetic retinoid that is used in the clinic for the treatment of psoriasis. To better understand the mechanism of retinoid action in the treatment of hyperproliferative diseases, we used a long-range differential display-PCR to isolate retinoid-responsive genes from primary human keratinocytes. We have identified a cDNA, tazarotene-induced gene 3 (TIG3; Retinoic Acid Receptor Responder 3) showing significant homology to the class II tumor suppressor gene, H-rev 107. Tazarotene treatment increases TIG3 expression in primary human keratinocytes and in vivo in psoriatic lesions. Increased TIG3 expression is correlated with decreased proliferation. TIG3 is expressed in a number of tissues, and expression is reduced in cancer cell lines and some primary tumors. In breast cancer cell lines, retinoid-dependent TIG3 induction is observed in lines that are growth suppressed by retinoids but not in nonresponsive lines. Transient over-expression of TIG3 in T47D or Chinese hamster ovary cells inhibits colony expansion. Finally, studies in 293 cells expressing TIG3 linked to an inducible promoter demonstrated decreased proliferation with increased TIG3 levels. These studies suggest that TIG3 may be a growth regulator that mediates some of the growth suppressive effects of retinoids.

Figure 1

(a) Long-range DD-PCR of mock-treated keratinocyte RNA (lanes 1 and 2), tazarotene-treated keratinocyte RNA (lanes 3 and 4), and normal HeLa cell RNA as a control (lanes 5 and 6). The cDNA fragment identified as TIG3 (arrow) appears in lanes 3 and 4 but not in the other lanes. (b) Northern blot analysis of TIG3 expression in mock-treated (lane 1) or tazarotene-treated (lane 2) primary human foreskin keratinocytes. Cells were cultured in keratinocyte growth medium without serum until confluent, then were mock-treated or treated with 10⁻⁶ M tazarotene for 3 days. Cells were harvested, and 15 μg of total RNA was analyzed. The blot then was stripped and reprobed with GAPDH. (c) Semiquantitative reverse transcription–PCR analysis of week 2 biopsies from patients treated with topical tazarotene gel. The TIG3 message was detected initially at cycle 24 in treated biopsies but was not detected until cycle 26 in control untreated biopsies. (d) Multiple sequence alignment showing homology between TIG3, H-rev 107, and rat H-rev 107 (R-rev 107). (e) Western blot analysis of 15 μg of protein extract from mock-treated T47D cells (lane 1), T47D cells treated for 2 days with 1 μM tazarotene (lane 2), and ECR-293-TIG3 cells induced for 24 hours with 1 μM muristerone A (lane 3). Standards indicate the TIG3 protein runs at an apparent molecular mass of 18 kDa.

Figure 2

(a) Multiple tissue Northern blot analysis revealed TIG3 expression in a variety of tissues, including lung, liver, kidney, spleen, thymus, prostate, ovary, small intestine, colon, and peripheral blood leukocytes. (b) TIG3 was not expressed in HL-60 peripheral blood acute promyelocytic leukemia cells, HeLa S3 cervical adenocarcinoma cells, K-562 chronic myelogenous leukemia cells, SW480 colon carcinoma, A549 lung carcinoma, or G361 melanoma cells. Low levels of expression were detected in MOLT-4 T-cell acute lymphoblastic leukemia cells and Raji B-cell Burkitt’s lymphoma cells. (c) TIG3 was not expressed in MCF-7 or T47D breast cancer cells, nor in HaCaT cells in the absence of tazarotene (−). TIG3 could be induced in T47D cells with the addition of 1 μM tazarotene (+) but was not induced in the MCF-7 or HaCaT cell lines. This in part may be because of a general lack of retinoid response in these cells as indicated by the lack of induction of RARβ2. (d) Expression of TIG3 was analyzed by Northern hybridization in biopsies from 25 primary human tumors of various tissues and was compared with expression levels in the adjacent normal tissues. Six tumors (lymphoma, ureter, two kidney, rectal, and uterine) showed significantly reduced or absent expression of TIG3 when GAPDH levels were used to control for variation in RNA levels. In no case was the expression of TIG3 increased in the tumor compared with normal tissue. (e) BrdUrd assay of T47D (closed triangle) and MCF-7 (open triangle) after 1 week of treatment with tazarotene shows that the growth of the T47D cell line but not the MCF-7 cell line is inhibited in a dose-responsive manner. (f) Time course induction of TIG3 message by tazarotene (1 μM) in T47D cells and analysis by Northern blot hybridization.

Figure 3

Colony formation in T47D (a) and CHO (b) cells containing the control expression vector (pcDNA3), the TIG3 expression vector (pcTIG3), the antisense TIG3 vector [pcTIG3(AS)], or the inducible vector in the absence of muristerone A (pIND-TIG3). Colony numbers represent the number of colonies per well (T47D) or per 50-cm² dish (CHO) after 3 weeks in G418 selection medium. TIG3 expression significantly inhibited the colony formation in both cell types by 60–70%. Experiments were done in triplicate and were repeated at least two times. (c) Representative plates from the CHO cell colony formation experiments stained with hematoxylin. (d1 and d2) Several T47D cell lines from the control pcDNA3 and the pcTIG3 plates were isolated and expanded into cell lines, and their growth rates were assayed by BrdUrd incorporation. The pcTIG3-containing cell lines grew at a significantly slower rate than the control pcDNA3-containing cells. (e) Western blot analysis showing TIG3 expression in the 10 isolated EcR-293-TIG3 cell lines after 24 hours induction with 1 μM muristerone A. Cell lines 2, 3, 4, 8, and 10 showed some level of TIG3 expression, with 4 and 8 exhibiting the highest levels. (f) The growth rates of clones 4 and 8 were inhibited when cultured for 1 week in the presence of 1 μM muristerone A (+) when compared with those cells cultured in the vehicle alone (−). The pIND-LacZ cell line 2 was not inhibited significantly by 1 μM muristerone A.

Figure 4

Schematic representation of the genomic organization of the first two exons of TIG3 and the location of the putative RA response elements in the TIG3 promoter as determined by homology with human chromosome 11 cosmid 187d6, localized to 11q23.