Introduction

Top Summary Introduction Experimental Procedures Results Discussion References

Retinoids, structural analogs of vitamin A, are involved in the regulation of differentiation, morphogenesis, proliferation, and apoptosis. Many but not all of these effects appear to be mediated by the nuclear retinoic acid receptors (RARs) and retinoid X receptors (RXRs) (Giguère, 1994). Retinoid receptors regulate transcription by binding to specific DNA sequences (retinoid response elements, RAREs, or RXREs) in the regulatory region of target genes. Retinoids can also exert anti-AP1 activity, and although this action is mediated by retinoid receptors, it does not involve RAREs or RXREs (Fanjul et al., 1994). However, a number of effects induced by retinoids occur independently of RARs or RXRs. For example, the growth-stimulatory effects of retroretinols on lymphocytes and fibroblasts are not mediated by retinoid receptors but through a novel yet unidentified signaling pathway (Buck et al., 1991; O'Connell et al., 1996). Several of the effects of the synthetic retinoid 6-[3-(1-adamantyl)-4-hydroxyphenyl]-2-naphthalene carboxylic acid (AHPN, also referred to as CD437) have been found to inhibit the growth and induce apoptosis of a variety of carcinoma cell lines (Shao et al., 1995; Schadendorf et al., 1996; Chao et al., 1997; Hsu et al., 1997a,b; Lu et al., 1997; Oridate et al., 1997; Sun et al., 1997a,b; Adachi et al., 1998a,b; Li et al., 1998; Meister et al., 1998). Although this retinoid binds RAR selectively, the molecular mechanism by which AHPN induces apoptosis in carcinoma cells does not involve this receptor but is mediated by a yet unknown mechanism.

Lung cancer represents one of the leading causes of death (Mannino et al., 1998). Although many lung cancers are sensitive to initial chemotherapy, in many cases they become rapidly resistant. In contrast to normal cells, many lung carcinoma cell lines have been reported to be rather resistant to the growth-inhibitory effects of retinoic acid (Geradts et al., 1993; Kalemkerian et al., 1994, 1995; Sun et al., 1997a,b; Zhang and Jetten, 1997; Adachi et al., 1998b). However, the synthetic retinoids 4-(hydroxyphenyl)retinamide and AHPN have been shown to inhibit cell proliferation and induce apoptosis in several lung cancer cell lines (Kalemkerian et al., 1995; Adachi et al., 1998b; Li et al., 1998; Zou et al., 1998). Recently, AHPN has been shown to very effectively inhibit the growth of lung tumors in mice, indicating their potential usefulness in chemotherapy (Lu et al., 1997).

In this study, we used the differential mRNA display method (Liang et al., 1994) to identify changes in gene expression that are related to AHPN-induced growth arrest and apoptosis in human lung carcinoma H460 cells. Several AHPN-inducible genes were identified and cloned by this method. The sequence of one of these genes exhibited high homology to the murine myeloid differentiation primary response gene MyD118 (Abdollahi et al., 1991) and likely represents the human homolog, which has not been described previously. MyD118 is closely related to the growth arrest and the DNA damage gene GADD45 (Fornace, 1992). mMyD118 and GADD45 encode nuclear proteins that interact with proliferating cell nuclear antigen and the cyclin-dependent kinase inhibitor p21^WAF1/Cip1 (Smith et al., 1994; Vairapandi et al., 1996; Prosperi, 1997). Increased expression of MyD118 and GADD45 have been shown to suppress cell growth (Zhan et al., 1994). These studies suggest that these proteins have an important function in negative growth control and apoptosis.

We demonstrate that AHPN enhanced the level of both hMyD118 and GADD45 mRNA in a variety of carcinoma cell lines. It is likely that the increase in these growth-suppressing genes contribute to the growth-inhibitory effects and the induction of apoptosis by AHPN. Our results also indicate that the increase in the level of hMyD118 and GADD45 mRNA by AHPN is regulated at the post-transcriptional level and is largely due to an increased RNA stability. The induction of hMyD118 and GADD45 expression is highly specific for AHPN, in agreement with the concept that this action is not mediated by the nuclear retinoid receptors RARs and RXRs but via a novel signaling pathway.

	Experimental Procedures

Top Summary Introduction Experimental Procedures Results Discussion References

Materials. The retinoid AHPN (also referred to as CD437) was described previously (Delescluse et al., 1991) and obtained from Dr. U. Reichert (CIRD Galderma, Valbonne, France). The RAR- (SRI-6751-84/TTAB, 4-(5,6,7,8-tetrahydro-5,5,8,8-tetramethyl-2-anthracenyl)-benzoic acid), RXR- (SR11217, 4-[1-(5,6,7,8-tetrahydro-5,5,8,8,-tetramethyl-2-naphthalenyl)-2-methylpropenyl]benzoic acid), and anti-AP1-selective [SR11302, (E)-3-methyl-9-(2,6,6-trimethylcyclohexenyl)-7-(4-methyl-phenyl)-2,4,6,8-nonatetraenoic acid] retinoids (Lehmann et al., 1992; Fanjul et al., 1994) were provided by Dr. M. Dawson (SRI, Menlo Park, CA). All-trans-retinoic acid was obtained from Hoffmann-La Roche (Nutley, NJ). Retinoids were dissolved in dimethyl sulfoxide (DMSO). Control cells received DMSO only. Cycloheximide and actinomycin D were purchased from Sigma Chemical Co. (St. Louis, MO). The cell-permeable, irreversible caspase inhibitor benzyloxycarbonyl-Val-Ala-Asp fluoro-methylketone (ZVAD.fmk) and the selective p38 mitogen-activated protein (MAP) kinase inhibitor PD169316 ([4-(4-Fluorophenyl)-2-(4-nitrophenyl)-5-(4-pyridyl)-11+-imidazole]) were purchased from Calbiochem (San Diego, CA).

Cell Culture. The human lung carcinoma cell lines H441, H460, A549, H1355, and Calu-6 and the human mammary carcinoma cell line MCF-7 were obtained from American Type Culture Collection (Rockville, MD). The human fibrosarcoma Ht1080 cell lines containing either wild-type or mutant p53, bladder carcinoma T24, and colorectal carcinoma cell line RKO were obtained from Dr. Gloria Preston (National Institute of Environmental Health Sciences). All cell lines except for Ht1080 and RKO were maintained in RPMI 1640 supplemented with 10% fetal bovine serum, penicillin, and streptomycin. The Ht1080 and the RKO cell lines were grown in Dulbecco's modified Eagle's Medium (DMEM) supplemented with 10% fetal bovine serum. All cell lines were Mycoplasma free.

Differential Display. Differential mRNA display was performed essentially as described by Liang et al. (1994). H460 cells were treated with 2.5 µM AHPN or vehicle (DMSO). After 8 h of treatment, total RNA was isolated. RNA was reverse-transcribed to cDNA with random hexamers. Polymerase chain reaction (PCR) was performed with a 5'-end primer containing a HindIII site, an one-base anchored oligo-(dT)₁₁ primer (Liang et al., 1994) including a BamHI site, [-³³P]dCTP (1000-3000 Ci/mmole; Amersham Life Sciences, Arlington Heights, IL), and Expand High Fidelity polymerase (Boehringer Mannheim, Indianapolis, IN). PCR products were separated on a 6% polyacrylamide/8 M urea sequencing gel. Differentially displayed bands were recovered from the polyacrylamide gel and then reamplified with the same primer pair. Reamplified products were used as probes to confirm differential expression by Northern blot analysis and then cloned into pGEM-3Z (Promega Corp., Madison, WI).

cDNA Library Screening. To obtain full-length cDNA clones, a human heart cDNA library (Stratagene, La Jolla, CA) was screened with a radiolabeled, reamplified band as a probe. Sequence analysis was performed in both directions by the dideoxynucleotide chain-termination method using a T7 Sequenase II kit (Amersham, Cleveland, OH). Nucleotide sequences were compared with those in Genbank using BLAST (National Center for Biotechnology Information, Bethesda, MD).

Northern Blot Analysis. RNA from cultured cells was isolated using Tri-Reagent (Sigma Chemical Co., St. Louis, MO) according to the manufacturer's protocol. Poly(A)⁺ RNA was isolated from total RNA with oligo(dT) columns (Qiagen, Chatsworth, CA). Total RNA (30 µg) was electrophoresed through a 1.2% denaturing agarose-formaldehyde gel as described (Adachi et al., 1998b), blotted to a Nytran Plus membrane (Schleicher & Schuell, Keene, NH), and UV-cross-linked. Hybridizations were performed for 1-2 h at 68°C using QuikHyb reagent (Stratagene, La Jolla, CA); blots were washed twice with 2× SSC, 0.05% SDS at room temperature for 15 min. The final wash was with 0.5× standard saline citrate, 0.1% SDS for 30 min at 65°C. Autoradiography was carried out with Hyperfilm-MP (Amersham) at 70°C using double intensifying screens. The cDNA probe for hMyD118 that was isolated by this differential display experiment was used as a probe for Northern analysis. The cDNA for GADD45 was obtained from Dr. A.J. Fornace (National Institutes of Health, Bethesda, MD). Tissue distribution was determined with a human multiple tissues RNA blot (Clontech, Palo Alto, CA).

Transfection. Reporter gene constructs containing different regions of the upstream GADD45promoter region in the vector p CAT-Basic (Promega) were obtained from Dr. A.J. Fornace. The construct pdc45-40 contains the region from 2578 to 149, pdc45-20 contains the region from 848 to 149, and pdc45-24 contains the region from 41 to 149 (Hollander et al., 1993). A549 cells (2 × 10⁵/well) were plated into six-well dishes 16 h before transfection. Reporter plasmids and a reference -actin-luciferase plasmid were transfected into cells by 1,3-di-oleoyloxy-2-(6-carboxy-spermyl)-propylamid (DOSPER, Boehringer Mannheim) according to the manufacturer's protocol. Twenty h after transfection, cells were treated with AHPN or DMSO, and 24 h later, cells were assayed for CAT and luciferase. CAT protein was measured using a CAT-ELISA kit (Boehringer Mannheim). Luciferase activity was assayed with a luciferase assay kit (Promega) in a Lumat LB9501 luminometer (Berthold). The relative level of CAT protein was normalized for luciferase activity.

RNA Stability Assay. H460 cells grown for 8 h in the presence or absence of 2.5 µM AHPN were treated with actinomycin D (5 µg/ml), and at different time intervals, cells were collected, and RNA was isolated. For RNA stability assay, total RNA from AHPN-treated cells and poly(A)⁺ RNA from untreated H460 cells were examined by Northern blot analysis using radiolabeled probes for hMyD118, GADD45, or glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Hybridization signals were quantitated with a PhosphorImage analyzer (STROM 860; Molecular Dynamics, Sunnyvale, CA) using ImageQuant software (Molecular Dynamics). mRNA half-life (T_1/2) was calculated using the equation C = C₀e^Kdt, where C is the remaining mRNA level at time point t, C₀ is the amount of RNA at zero time of actinomycin D addition, and Kd is the mRNA decay constant.

	Results

Top Summary Introduction Experimental Procedures Results Discussion References

Differential mRNA Display Analysis. To obtain additional insight into the mechanism of action of AHPN, differential mRNA display analysis was used to identify specific response genes induced by AHPN in human lung carcinoma cells. For this purpose, exponentially growing cultures of human lung carcinoma H460 cells were treated with or without AHPN, and after 8 h of treatment, total RNA was extracted. cDNA was synthesized by reverse transcriptase reaction and amplified with one-base anchored oligo(dT)₁₁ and 5'-end primers. Radiolabeled PCR products were separated on a sequencing gel and analyzed by autoradiography. Differentially expressed cDNA fragments were recovered from the polyacrylamide gel and reamplified with the same pair of primers; their differential expression was confirmed by Northern blot analysis. The fragments, for which differential expression was confirmed, were subjected to sequence analysis. The obtained nucleotide sequences were compared with those in Genbank. One sequence, 0.4 kb in size, had high homology to the sequence in the 3'-untranslated region of the mMyD118 gene (Abdollahi et al., 1991). To clone the full-length cDNA, a human heart cDNA library was screened with the radiolabeled 0.4 kb probe, and the inserts of the obtained clones were sequenced. The nucleotide sequence of the full-length cDNA is shown in Fig. 1. The nucleotide sequence (Genbank accession no. AF090950) was 86% identical with that of the mMyD118 gene. hMyD118 encodes a M_r 18,000 protein that exhibited a 90.6% identity with mMyD118 (Fig. 2). hMyD118 protein exhibited a 56.3% identity with the closely related gene GADD45 (Fig. 2).

View larger version (76K): [in this window] [in a new window]   Fig. 1.   The nucleotide and deduced amino acid sequence of hMyD118 (Genbank accession no. AF090950). The amino acid sequence is noted below the nucleotide sequence, and the translation start and stop codons are shown in bold. The polyadenylation signal is underlined.

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Comparison of the amino acid sequences of hMyD118, mMyD118, and GADD45. Amino acids that are identical with those in hMyD118 sequence are shown as dots, whereas sequence differences are indicated.

Tissue-Specific Expression of hMyD118. We next analyzed the expression of hMyD118 in several normal human tissues. Northern blot analysis showed that hMyD118 was expressed in many tissues including heart, placenta, liver skeletal muscle, kidney, pancreas, spleen, thymus, prostate, uterus, and peripheral blood leukocytes (Fig. 3). MyD118 mRNA expression was very low in brain, lung, pancreas, and testis. Our results show that MyD118 is much more widely expressed than originally reported (Abdollahi et al., 1991).

View larger version (30K): [in this window] [in a new window]   Fig. 3.   Tissue-specific expression of hMyD118. Northern blot analysis was performed on total RNA isolated from different human tissues using a radiolabeled probe for hMyD118. The source of the RNA is indicated above each lane. The size of MyD118 mRNA is shown on the left.

View larger version (48K): [in this window] [in a new window]   Fig. 4.   Comparison of the induction of hMyD118 and GADD45 by AHPN in several cell lines. A, human lung carcinoma H441, H460, A549, H1355, and Calu-6 cells were treated for 24 h with 1 µM AHPN. Total RNA was isolated and examined by Northern blot analysis using radiolabeled probes for hMyD118, GADD45, and GAPDH. B, human fibrosarcoma MutHt 1080 and WtHt 1080 cells, mammary carcinoma MCF-7, colorectal carcinoma RKO, and bladder carcinoma T24. Cells were treated for 16 h with AHPN.

Specificity of AHPN Action. To determine the specificity of the action of AHPN, we compared the effect of various retinoids on hMyD118 expression in H460 and A549 cells. Cells were treated with the indicated retinoid; after 8 h, cells were collected, and RNA was isolated and examined by Northern analysis. Figure 5 shows that treatment of H460 and A549 with AHPN greatly increases the level of MyD118 and GADD45 mRNA expression. Two retinoids, TTAB and SR11217, which bind and activate selectively RAR or RXR receptors, respectively, did not increase MyD118 or GADD45 expression, suggesting that activation of RARs but not that of RXRs is sufficient for this action. Many retinoids have been reported to exhibit anti-AP1 activity, including TTAB and SR11217. The fact that SR11217 did not increase the level of MyD118 and GADD45 mRNA indicates that this action does not require the anti-AP-1 activity of retinoids. The latter is supported by the observation showing that the retinoid SR11302 with reported selective anti-AP-1 activity (Fanjul et al., 1994) also did not have any effect on the expression of these genes. These results demonstrate that the induction of hMyD118 and GADD45 mRNA is highly specific for AHPN, as has been reported for the growth-inhibitory activity of these retinoids (Adachi et al., 1998b). These findings support the concept that these actions of AHPN do not involve retinoid receptors but are mediated by a novel mechanism.

View larger version (35K): [in this window] [in a new window]   Fig. 5.   Effect of various retinoids on the expression of hMyD118 and GADD45 mRNA in lung carcinoma H460 and A549 cells. Cells were treated for 8 h with 1 µM for the indicated retinoid. Total RNA was isolated and examined by Northern blot analysis using radiolabeled probes for hMyD118, GADD45, and GAPDH. A, RNA from H460 cells treated with DMSO (0.1%; NA), RAR-selective retinoid (RARsel), RXR-selective retinoid (RXRsel), AHPN, and anti AP-1 selective retinoid (anti AP-1sel). B, RNA from A549 cells treated with the same retinoids.

View larger version (34K): [in this window] [in a new window]   Fig. 6.   Time course of the induction of hMyD118 and GADD45 mRNA by AHPN in lung carcinoma H460 and A549 cells. Cells were treated with 1 µM AHPN for the time period indicated. Total RNA was isolated and examined by Northern blot analysis using radiolabeled probes for hMyD118, GADD45, and GAPDH. Hybridization signals were quantitated with a PhosphorImage analyzer using ImageQuant software as described in Experimental Procedures. The results are plotted as the relative level of hMyD118 (open bars) or GADD45 mRNA (solid bars). H460 (A and C) and A549 (B and D) cells treated with AHPN.

Induction of MyD118 and GADD45 mRNA Is Independent of Activation of Caspases or p38 MAP Kinase. Previous studies have shown that the induction of apoptosis by AHPN is associated with an increase in activity of specific cysteine proteases, named caspases (Piedrafita and Pfahl, 1997; Adachi et al., 1998a). To determine whether activation of caspases is required for the induction of MyD118 by AHPN, we examined the effect of the caspase inhibitor ZVAD.fmk on MyD118 expression. Concentrations of ZVAD. fmk that totally inhibited caspase activity (Adachi et al., 1998a) had no effect on the AHPN-dependent increase of MyD118 mRNA (Fig. 7A). These results indicate that the increase in MyD118 does not require the activation of caspases. Piedrafita and Pfahl (1997) have shown that the induction of caspase is independent of protein synthesis. It is therefore likely that the increase in MyD118 and caspase activation occur independently from each other.

View larger version (24K): [in this window] [in a new window]   Fig. 7.   The induction of MyD118 and GADD45 mRNA by AHPN is independent of the activation of caspases or p38 MAP kinase. A, H460 cells were treated with AHPN in the presence or absence of the caspase inhibitor ZVAD.fmk (30 µM). After 8 h of treatment, RNA was isolated and examined by Northern analysis for hMyD118 mRNA expression. The level of 18S RNA shows equal loading of RNA. B, H460 cells were treated with AHPN in the presence or absence of the MAP kinase inhibitor PD169316 at the concentrations indicated. After 8 h of treatment, RNA was isolated and examined by Northern analysis for hMyD118 and GADD45 mRNA expression. The inhibition of p38 MAP kinase phosphorylation at 1 µM PD169316 was 82%, and at all other concentrations, >99%.

Regulation of MyD118 and GADD45 by AHPN. To determine at what level AHPN regulates the expression of these genes, we examined the effect of AHPN on the transcriptional activation from a GADD45 promoter regulatory region. Transcription of the reporter form a 2578/149 or a 848/149 GADD45 promoter region was very efficient (not shown). However, AHPN had little effect on this transcriptional activation, suggesting that it does not control the transcription of this gene through elements contained in this region.

View larger version (35K): [in this window] [in a new window]   Fig. 8.   Induction of hMyD118 and GADD45 mRNA expression by AHPN is related to increased RNA stability. A, after 8 h of treatment with AHPN or vehicle, H460 cells were treated subsequently with actinomycin D (5 µg/ml). At the indicated times, RNA was isolated and examined by Northern blot analysis with radiolabeled probes for hMyD118, GADD45, and GAPDH. Poly(A)+ RNA was used in the case of control H460 cells, whereas total RNA was used in the case of AHPN-treated cells. Hybridization signals were quantitated with a PhosphorImage analyzer using ImageQuant software as described in Experimental Procedures. The results are plotted as the relative level of hMyD118 (B) or GADD45 mRNA (C) as the percentage of the amount at time 0 of actinomycin D addition.

	Discussion

Top Summary Introduction Experimental Procedures Results Discussion References

The synthetic retinoid AHPN has been demonstrated to induce growth arrest and apoptosis in many carcinoma cells, including human lung, cervical, and mammary carcinoma and lymphoma cell lines (Shao et al., 1995; Chao et al., 1997; Lu et al., 1997; Oridate et al., 1997; Sun et al., 1997a, b; Li et al., 1998; Adachi et al., 1998a, b). Although AHPN has been demonstrated to bind RAR selectively and to weakly induce transcriptional activation via this receptor, several studies have demonstrated that RAR or RXR nuclear retinoid receptor pathways are not involved in AHPN-induced growth arrest and apoptosis (Shao et al., 1995; Hsu et al., 1997b; Piedrafita and Pfahl, 1997; Sun et al., 1997a; Adachi et al., 1998a, b). In an effort to understand the mechanism by which AHPN induces growth arrest and apoptosis, we began to identify genes that are induced after treatment with AHPN using differential display. One of the differentially expressed clones isolated encodes the human MyD118, based on its high sequence homology with murine MyD118 (Abdollahi et al., 1991). The mouse homolog was cloned previously as a novel primary response gene from terminally differentiating murine myeloid M1 cells after stimulation with multiple cytokines, including interleukin 1 and leukemia inhibitory factor. Both murine and human MyD118 encode a 1.3-kb mRNA and a M_r 18,000 nuclear protein. In contrast to previous reports on mMyD118 (Abdollahi et al., 1991), hMyD118 was expressed in all tissues tested, including liver, spleen, kidney, and heart. In addition, we show that hMyD118 mRNA is expressed at low levels in many carcinoma cell lines and is dramatically induced by AHPN. These results suggest a much wider role for MyD118 than originally believed.

MyD118 is closely related to the growth arrest- and DNA damage-inducible gene GADD45. GADD45 is induced in mammalian cells by a wide range of stimuli, including environmental stress, genotoxic drugs, and ionizing irradiation (Fornace et al., 1989, 1992). Although these genes are regulated by different mechanisms, they are often coordinately expressed and can function cooperatively in inhibiting cell growth (Zhan et al., 1994). In this study, we demonstrate that growth inhibition and induction of apoptosis by AHPN is associated with increased levels of MyD118 and GADD45 mRNA. Induction of GADD45 was recently also demonstrated in AHPN-treated myoblastic leukemia HL-60 cells (Hsu et al., 1997b). In most carcinoma cell lines tested, the increase in MyD118 and GADD45 mRNA by AHPN occurs in parallel. However, in mammary carcinoma MCF-7 cells, AHPN enhanced GADD45 but not MyD118 mRNA, whereas in lung carcinoma Calu-6 cells, GADD45 mRNA levels increased only weakly. In H460 cells, the increase in MyD118 mRNA is transient, whereas that of GADD45 is not, suggesting different mechanisms of regulation. The enhancement in both hMyD118 and GADD45 is highly specific for AHPN because an RAR-panagonist and RXR-panagonist were unable to induce GADD45 and hMyD118 mRNA. This action of AHPN also does not involve the anti-AP-1 activity of retinoids because an anti-AP-1-selective retinoid and the RXR-selective retinoid, which also exhibits anti-AP-1 activity, did not increase GADD45 or hMyD118 mRNA. These results support previous conclusions that changes associated with AHPN growth arrest and apoptosis are not mediated by nuclear retinoid RAR and RXR receptors but through a novel signaling pathway.

Several reports have provided evidence indicating that GADD45 and MyD118 play a critical role in the control of cell cycle progression and apoptosis (Abdollahi et al., 1991; Fornace, 1992; Zhan et al., 1994, Smith et al., 1994; Selvakumaran et al., 1994; Vairapandi et al., 1996). This is illustrated by studies showing that terminal differentiation induced by various cytokines in murine myeloblastic M1 cells and the induction of growth arrest and apoptosis by transformin growth factor  is associated with a marked increase in MyD118 expression (Abdollahi et al., 1991; Selvakumaran et al., 1994). Moreover, overexpression of these proteins has been shown to induce growth arrest and apoptosis (Zhan et al., 1994). Further support for a negative role of MyD118 and GADD45 in growth control comes from studies showing that these proteins can interact with cyclin-dependent kinase inhibitor p21^Cip1/Waf1 and proliferating cell nuclear antigen (Smith et al., 1994; Vairapandi et al., 1996; Prosperi, 1997). The association of increased MyD118 and GADD45 expression with AHPN-induced growth arrest and apoptosis in some of the cell lines (Adachi et al., 1998b) provides additional support for a negative growth-regulatory role of these proteins.

Previous studies showed that treatment of several lung and mammary carcinoma cells with AHPN causes growth arrest in the G₁ phase of the cell cycle and is associated with a decrease in Rb hyperphosphorylation and cyclin-dependent kinase activity and increases in the level of p53 and p21 (Shao et al., 1995; Li et al., 1996; Adachi et al., 1998b). However, the induction of growth arrest and apoptosis by AHPN occurs independently of the increase in p53 and p21 because AHPN is able to induce these genes very effectively in lung carcinoma H441, H1355, and Calu-6 cells, which either do not express p53 or p21 or express mutant p53 (Adachi et al., 1998b). Previous studies have shown that GADD45 can be regulated in a p53-dependent and -independent manner, but that MyD118 is not controlled by p53 (Zhan et al., 1994). As was reported previously for p21 (Shao et al., 1995; Adachi et al., 1998b), the increase in MyD118 and GADD45 mRNA by AHPN is independent of p53 expression.

Little is known about the p53-independent mechanisms of MyD118 and GADD45 regulation. We demonstrate that the induction of MyD118 and GADD45 mRNA is unrelated to activation of caspases or p38 MAP kinase, two cellular responses that are rapidly induced by AHPN (Adachi et al., 1998a; Sakaue et al., manuscript in preparation). We also found no difference in the promoter activity of the upstream regulatory region of the GADD45 gene in H460 cells treated with and without AHPN. However, AHPN caused a large increase in the half-life of both MyD118 and GADD45 mRNAs; their half-life was extended almost 9-fold. These results suggest that the increase in MyD118 and GADD45 mRNA levels by AHPN is largely regulated by a post-transcriptional mechanism and is due to increased stability of their mRNAs. RNA stability can be controlled by different mechanisms (Chen and Shyu, 1995; Ross, 1996). The family of early response genes, to which GADD45 and mMyD118 belong, encode highly unstable RNAs that often contain one or more adenylate/uridylate-rich instability elements in their 3' UTR. Alternatively, mRNA-binding proteins, which bind to RNA stem-loops, may protect RNA from degradation by endonucleases and stabilize mRNAs (Ross, 1996). Future studies have to determine whether binding of specific RNA-binding protein to adenylate/uridylate-rich elements or to RNA stem-loop motifs are involved in the increased RNA stability induced by AHPN. Recently, Koonin (1997) has found thorough computer protein alignments that GADD45 and MyD118 are homologues of the eukaryotic ribosomal (r) protein S12 and belong to an ancient superfamily of ribosome-associated proteins. It has been suggested that these proteins may also possess RNA binding properties and could themselves have a role in regulating transcription and translation.

In summary, we demonstrate that the induction of growth arrest by AHPN, and in certain cell lines apoptosis, is associated with an increase in the level of MyD118 and GADD45 mRNA. This association is in agreement with a negative growth regulatory function of these genes. The up-regulation of these mRNAs occurs at a post-transcriptional level and is due to increased stability of the respective mRNAs. This increase appears not to be mediated by RAR or RXR receptors but by a novel signaling pathway.