HuD, a neuronal-specific RNA-binding protein, increases the in vivo stability of MYCN RNA.

MYCN amplification and consequent deregulated expression plays a crucial role in determining the clinical behavior of neuroblastoma. Enhanced expression of MYCN confers growth potential to neuroblastoma cells, and a direct link between MYCN expression and the development of neuroblastoma has been demonstrated in transgenic mice studies. Although the molecular pathways underlying the regulation of MYCN have not been fully elucidated, post-transcriptional mechanisms appear to be important. Previously, we reported that an embryonic lethal abnormal vision-like (ELAV) protein binds with high specificity to at least two AU-rich elements within the MYCN 3'-untranslated region. In this study, we characterized the ability of cis-acting elements within the MYCN 3'-untranslated region to destabilize mRNA in cells and examined the functional consequences of its interactions with the ELAV protein HuD. We show that at least 4 cis-acting elements within the MYCN 3'-untranslated region are able to signal the degradation of stable heterologous mRNA. Ectopic overexpression of HuD dramatically inhibits RNA decay mediated by the full-length MYCN 3'-untranslated region and cis-acting destabilizing elements that harbor HuD binding sites in vivo. HuD may contribute to the malignant phenotype of neuroblastoma cells by stabilizing MYCN mRNA, thereby enhancing steady-state levels of expression of this oncogene.

tumors (5). Although the clinical significance of MYCN expression in NB tumors that lack MYCN amplification remains controversial (6 -10), a strong correlation between high levels of MYCN expression consequent to genomic amplification of this oncogene and aggressive disease is well established (3)(4)(5)11). Laboratory experiments have further supported an important role for MYCN in determining NB phenotype. Ectopic overexpression of MYCN results in enhanced malignant growth (12,13), whereas MYCN antisense studies performed by our laboratory and others have shown that down-regulation of MYCN in human NB is associated with a decrease in cellular proliferation and inhibited tumor cell growth in vitro (14 -16). Furthermore, a direct link between MYCN expression and the development of NB has been demonstrated in transgenic mice studies (17).
Previously we examined MYCN regulation in N-(neuroblastic, tumorigenic) and S-type (substrate-adherent, non-tumorigenic) subclones (W-N and W-S) of the MYCN-amplified NB cell line NBL-W (18). In addition to distinct morphology and growth characteristics, the subclones have differential levels of MYCN expression despite having the same genomic MYCN copy number. We found that the disparity in steady-state levels of MYCN mRNA in the W-N and W-S cells was largely determined by differences in MYCN mRNA stability (19). Lazarova and co-workers (20) examined MYCN expression in other MYCN-amplified NB cell lines, and similarly found that tumorigenic N-type NB cells express higher levels of MYCN than S-type cells. Their studies indicated that nuclear processing and/or stabilization of MYCN pre-mRNA species were important events in the regulation of MYCN gene expression (20).
Modulation of mRNA stability provides a powerful means for controlling gene expression (for a review see Ref. 21). Although, the molecular mechanisms responsible for regulating mRNA turnover are not completely understood, it is known that many labile mRNAs, including MYCN (22), contain AU-rich elements (ARE) within their 3Ј-untranslated regions (UTRs) (23,24 and reviewed in Ref. 25). At least 14 apparently distinct AREbinding proteins have been identified in eukaryotic cell extracts by UV-cross-linking and gel shift assays (reviewed in Ref. 26), and there is evidence indicating that binding of these trans-acting factors to AREs influences mRNA degradation (27)(28)(29)(30)(31)(32).
Members of the embryonic lethal abnormal vision (ELAV) family of proteins have been shown to bind with high affinity to AREs within the 3Ј-UTR of mRNAs of genes involved in cellular growth and neuronal differentiation (20,(33)(34)(35)(36)(37). The functional significance of the interactions between AREs and ELAV proteins has been investigated, and a number of studies have demonstrated that ELAV proteins inhibit ARE-mediated mRNA turnover (28, 31, 33, 38 -41). Previous work from our laboratory has shown that N-type NB cell lines and a subset of primary NB tumors contain abundant quantities of an ELAV family protein that bind in a specific manner to AU-rich sequence elements within the 3Ј-UTR of MYCN (19,42). Similarly, others have reported that the ELAV protein HuD binds to elements within the MYCN 3Ј-UTR (20). Taken together, these studies suggest that HuD is involved in the post-transcriptional regulation of MYCN.
The present study was undertaken to functionally characterize cis-acting elements within the MYCN 3Ј-UTR and to further define the role HuD plays in the post-transcriptional regulation of MYCN expression in NB. Using a series of heterologous mRNAs that span the entire MYCN 3Ј-UTR, we demonstrate that at least four destabilizing cis-acting elements are present in the MYCN 3Ј-UTR, and that fragments containing the HuD binding sites are able to signal the cellular degradation of stable heterologous mRNA. We further show that ectopic overexpression of HuD dramatically inhibits MYCN ARE-mediated mRNA decay in vivo and that down-regulation of HuD in NB cells results in a significant decrease in MYCN expression.

EXPERIMENTAL PROCEDURES
Construction of Recombinant Plasmids-Genomic DNA was isolated from W-N NB cells and used in a PCR to amplify the 3Ј-untranslated region of the MYCN gene. The forward primer 5Ј-GGGAGATCT-CACGCTCGGACTTGCTAG-3Ј with a BglII site and reverse primer 5Ј-GGGGAGCTCTTTAATTTTAAGCTATTTATTTTCATAAAC-3Ј with a SacI site were used with Pwo polymerase (Roche Molecular Biochemicals, Indianapolis, IN) in a total reaction volume of 100 l for 31 cycles in a DNA Thermal Cycler (MJ Research, Waltham, MA). Each cycle consisted of denaturation at 94°C for 1 min, annealing at 45°C for 1 min, and extension at 72°C for 1.5 min. The PCR product was bluntended by T4 DNA polymerase (Invitrogen, Rockville, MD) in a final volume of 20 l at 37°C for 5 min. The amplified MYCN 3Ј-UTR DNA fragment was cloned into the unique BglII and SacI sites of the pBBB plasmid that contains the rabbit ␤-globin gene (43) (a kind gift of Dr. A. B. Shyu, M.D. Anderson, Houston, TX) immediately downstream of the translation termination codon of the ␤-globin coding region. The resulting construct, pBBN, contained the ␤-globin coding region and the MYCN 3Ј-UTR.
Two series of chimeric constructs containing MYCN 3Ј-UTR fragments were also generated for this study. Short MYCN 3Ј-UTR fragments (100 -150 bp in length) were generated by PCR using forward and reverse primers at the following positions: F1, 2015-2136 bp; F2, 2115-2244 bp; F3, 2225-2356 bp; F4, 2337-2474 bp; F5, 2455-2595 bp; F6, 2575-2712 bp; F7, 2693-2824 bp; and F8, 2805-2915 bp. Both the forward and reverse primers used to amplify the fragments for series I contained BglII sites. In the first series of constructs, pBB(F1-F8)B, the MYCN 3Ј-UTR fragments were cloned into the unique BglII site of pBBB between the ␤-globin coding region and the ␤-globin 3Ј-UTR. Thus, the ␤-globin 3Ј-UTR was retained in the first series of constructs. The second series of constructs, pBB(F1-F8) contained the same set of MYCN 3Ј-UTR DNA fragments. The forward primers used to generate the series II fragments contained a BglII site and the reverse primer a SacI site to facilitate cloning into the BglII/SacI sites immediately downstream of the ␤-globin coding region of pBBB. In the second series of constructs, the MYCN fragments replaced the entire ␤-globin UTR.
A FLAG epitope was synthesized by using a forward primer with an NheI site (5Ј-CTAGCATGGACTACAAGGACGACGATGACA-3Ј) and a complementary reverse primer with a HindIII site (5Ј-AGCTTGT-CATCGTCGTCCTTGTAGTCCATG-3Ј). After annealing, the doublestranded fragment was cloned into the NheI/HindIII sites of pcDNA3.1 (Invitrogen, San Diego, CA). HuD cDNA was generated by PCR using a forward primer with a HindIII site (5Ј-AGCTTATGGAGCCTCAGGT-GTCAAATGG-3Ј) and a reverse primer with an XbaI site (5Ј-CTA-GAATCGATTCAGGACTTGTGGGCTTTGTTGG-3Ј) and cloned into the HindIII/XbaI sites of pcDNA3.1 downstream of the FLAG epitope to generate an NЈ FLAG-HuD fusion cDNA. Upstream, a viral ribosomal binding site was cloned into the EcoRI site to enhance the translation of the FLAG-tagged HuD protein. The orientation and sequence of the cloned cDNAs were confirmed by sequencing.
Generation of ␤-Globin and GAPDH Probes-Labeled ␤-globin and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) probes for the RNase protection assay were prepared by in vitro transcription. The ␤-globin template was generated from pBBB by PCR using the forward primer 5Ј-TGCCAGGTGCTGACTTCTCTC-3Ј and the reverse primer 5Ј-GGATCCTAATACGACTCACTATAGGGAGGAGGTCCATGGTGAT-ACAAGGGAC-3Ј. To yield the 154-nt GAPDH template, the commercially available pTRI-GAPDH-mouse antisense template (Ambion, Austin, TX) was digested with DdeI for 1 h at 37°C. The 380-nt ␤-globin and 154-nt GAPDH probes were excised and eluted at room temperature overnight in 1ϫ elution buffer (0.5 M ammonium acetate, 1 mM EDTA, 2.5% SDS). Specific activity was measured in a scintillation counter.
RNase Protection Assays-NIH 3T3 cells stably transfected with pBBB (3T3 BBB), pBBN (3T3 BBN), pBB(F1-F8)B (3T3 BBFB), or pBB(F1-F8) (3T3 BBF) were grown to 60 -70% confluence in six-well plates. The pcDNA3.1-HuD (4 g) plasmid or empty vector (pcDNA3.1) was complexed for 10 min at room temperature with 20 l of SuperFect transfection reagent in 150 l of serum and antibiotic-free DMEM media, and transfections were done according to manufacturer's instructions. Cells were incubated for 3 h at 37°C in 5% CO 2 , washed with phosphate-buffered saline, and allowed to recover overnight in DMEM with 10% FBS. The cells were next serum-starved in 0.5% FBS-DMEM for 24 h and stimulated with 20% FBS-DMEM as described previously (44). Following stimulation, cells were harvested at various time points, and total RNA was isolated using the TRIzol reagent (Invitrogen) and then treated with RQ1 DNase. The degradation or stabilization of ␤-globin mRNA was analyzed by T1 RNase protection as described (45) with the following modifications: 50,000 cpm of labeled antisense probe (␤-globin and GAPDH) was added to 10 g of total isolated RNA and hybridized at 40°C for 16 h. The RNA⅐DNA complex was digested with an RNase A (1 mg/ml)/RNase T1 (20,000 units/ml) mixture (Ambion) at 37°C for 45 min. The complex was then incubated at 37°C for 45 min with 200 g of Proteinase K and 2% SDS. Nucleic acids were precipitated and protected fragments were resolved on a 6% denaturing polyacrylamide gel (8 M urea). The RNA bands were quantitated by densitometry (Molecular Dynamics, Sunnyvale, CA).
Antisense Oligodeoxynucleotide Treatment of NB Cells-Sense and antisense phosphorothioated oligodeoxynucleotides complementary to a sequence in the 5Ј region of the HuD cDNA were synthesized by an automated solid-phase synthesizer (Biotech Facility, Northwestern University). Both the sense (5Ј-TAGCACCATGGAGCCTCAGGTGTC-3Ј) and antisense (5Ј-GACACCTGAGGCTCCATGGTGCTA-3Ј) sequences were specific for HuD. The oligodeoxynucleotides were added to the medium of cultured W-N cells daily at a concentration of 10 M. Under these experimental conditions the cells remained viable and showed no evidence of toxicity. Medium was changed every 2 days. After 6 days, total RNA was extracted from sense-and antisense-treated cells, and the level of HuD expression was examined by RT-PCR.
Real-time RT-PCR Analysis of MYCN Expression-Total RNA was isolated from W-N cells treated with sense or antisense oligodeoxynucleotides. The RNA (25 ng) was subjected to real-time RT-PCR analyses using the TaqMan PCR reagent kit (Applied Biosystems, Foster City, CA) in a total volume of 50 l. Reverse transcription reaction was performed as follows: 30 min at 48°C, 15 min at 95°C, and then 40 cycles of PCR were done with DNA denaturation for 15 s at 95°C and primer annealing and extension for 1 min at 58°C. The gene-specific primers and probes for the MYCN target gene sequence and ␤ 2 -microglobulin endogenous control gene for use in real-time RT-PCR have been described previously (7). Quantitation of the sample was based on the cycle when the amplicon was first detected. A parameter, threshold cycle (C T ) was defined for each PCR reaction as the cycle number at which the reporter fluorescence generated by the cleavage of the sequence-specific probes passed above a fixed baseline. The expression of MYCN was normalized to ␤ 2 -microglobulin message levels, and the data are expressed as the C T value relative to ␤ 2 -microglobulin (␤ 2 -microglobulin C T Ϫ MYCN C T ).

RESULTS
The MYCN 3Ј-UTR Mediates mRNA Decay in Vivo-To determine if cis-acting elements within the MYCN 3Ј-UTR mediate mRNA decay, the half-life of chimeric ␤-globin-MYCN 3Ј-UTR transcripts was measured. To produce the chimeric transcripts, the human MYCN 3Ј-UTR was subcloned into the pBBB vector in a unique BglII site within the 3Ј-UTR of the rabbit ␤-globin gene to generate pBBN (Fig. 1A). pBBN was co-transfected with pSV2neo into NIH 3T3 cells and neomycinresistant colonies were selected. Control experiments were performed with pBBB. All constructs contained the c-fos promoter, which drives transient induction of the cloned sequences with serum stimulation. Clones that expressed the chimeric ␤-globin-MYCN-3Ј-UTR transcript or the control ␤-globin message following serum stimulation were expanded. To measure mRNA turnover, transcription of ␤-globin and the chimeric ␤-globin-MYCN 3Ј-UTR mRNAs was transiently induced with 20% serum. Total RNA was isolated from cells harvested at defined time intervals, and the quantity of exogenously introduced ␤-globin and chimeric ␤-globin-MYCN 3Ј-UTR transcripts at each time point was analyzed by RNase protection assays using a ␤-globin probe. The endogenous level of GAPDH mRNA was also measured to control for amounts of RNA analyzed. As shown in Fig. 1 (B and C), the chimeric ␤-globin MYCN-3ЈUTR transcript underwent rapid decay with a halflife of 80 Ϯ 15 min, whereas no evidence of ␤-globin mRNA degradation was seen at 300 min. Thus, elements within the MYCN 3Ј-UTR are able to enhance the degradation of the stable ␤-globin mRNA.
Identification of RNA-destabilizing Cis-acting Elements within the MYCN 3Ј-UTR-To identify the destabilizing cisacting elements within the MYCN 3Ј-UTR, fragments (110-to 150-nt long) spanning the entire length of the 900-nt MYCN 3Ј-UTR were generated by PCR and cloned downstream of the ␤-globin coding region in pBBB to create two series of plasmid constructs for this study (Fig. 2). Fragments 1 and 8 (F1 and F8) contained the previously identified HuD binding sites (42). In the first series of constructs (pBB(F1-F8)B), the MYCN 3Ј-UTR fragments were cloned 3Ј of the ␤-globin coding region and the ␤-globin 3Ј-UTR was retained. In the second series of constructs (pBB(F1-F8)) the ␤-globin 3Ј-UTR was replaced by the MYCN 3Ј-UTR fragments. To investigate if the MYCN 3Ј-UTR fragments contained functional destabilizing elements, individual chimeric constructs were co-transfected into NIH 3T3 cells with pSV2neo. Neomycin-resistant clones were established and characterized to confirm expression of the exogenously introduced chimeric transcripts following serum stimulation. Total RNA was isolated from the transfected cells at defined time points, and the half-life of each of the chimeric transcripts was determined using RNase protection assays. In the experiments performed with the first series of chimeric transcripts, no significant change in the rate of mRNA degradation was seen compared with the ␤-globin control, indicating that the MYCN AREs were not sufficient to destabilize the chimeric message when the ␤-globin 3Ј-UTR was retained (Table I). However, in the absence of the ␤-globin 3Ј-UTR sequences, the turnover of four of the eight chimeric transcripts was enhanced (Table I). Interestingly, both chimeric transcripts (BBF1 and BBF8) that contained the previously identified ELAV protein binding sites (42) underwent rapid degradation ( Fig. 3 and Table I). In addition, a pronounced destabilization was seen with BBF5 and BBF6. In contrast, chimeric transcripts in which MYCN fragments F2, F3, F4, or F7 replaced the ␤-globin 3Ј-UTR remained stable. Thus, in addition to the previously identified ELAV-protein binding sites, the MYCN 3Ј-UTR contains additional destabilizing cisacting elements within nt sites 2455-2595 (F5) and 2575-2712 (F6).
Antisense HuD Mediated Down-regulation of MYCN Expression-To further evaluate the role HuD plays in regulating MYCN turnover in NB cells, experiments were performed wherein the high MYCN mRNA-expressing W-N cells were treated with sense or antisense oligodeoxynucleotides specific to HuD. HuD antisense oligomer treatment resulted in a more than 3-fold down-regulation of HuD expression in W-N NB cells (Fig. 4A). MYCN expression was assessed by quantitative realtime RT-PCR in the treated cells. The level of MYCN RNA was normalized to ␤ 2 -microglobulin, and the data are expressed as ␤ 2 -microglobulin C T -MYCN C T . As shown in the bar graph in Fig. 4B, the level of MYCN expression in the W-N cells treated with the HuD antisense oligomers was almost 3-fold less than that measured in the control W-N cells treated with sense HuD oligomers. Thus, down-regulation of HuD in W-N NB cells resulted in a coordinate, significant decrease in MYCN expression.

RNA Stabilization by the ELAV RNA-binding Protein
HuD-To test our hypothesis that MYCN mRNA stabilization is the functional consequence of interactions between HuD and cis-acting elements within the MYCN 3Ј-UTR, we overexpressed HuD in NIH 3T3 cells stably transfected with pBBN (3T3 BBN). A plasmid that constitutively expresses high levels of HuD was transiently transfected into 3T3 BBN cells, and ectopic expression of HuD was confirmed by Western blot analysis (data not shown). Control experiments were performed with an empty expression vector. Expression of the chimeric ␤-globin-MYCN 3Ј-UTR transcript in the control and test cells was induced with serum, and total RNA was isolated from cells harvested at defined times. The RNA was subjected to RNase protection assays, and the half-life of the chimeric ␤-globin-MYCN 3Ј-UTR transcript was determined. As shown in Fig. 5, overexpression of HuD protein dramatically stabilized the chimeric ␤-globin MYCN 3Ј-UTR transcript, whereas rapid mRNA turnover was seen in control experiments. Additional studies were performed with NIH 3T3 cells stably transfected with constructs in which the ␤-globin 3Ј-UTR was replaced by each of the four cis-acting destabilizing MYCN elements (pBBF1, pBBF5, pBBF6, and pBBF8). In the presence of HuD, the chimeric transcripts containing the HuD binding sites (pBBF1 and pBBF8) were stabilized, whereas no significant change in half-life was seen in the other two chimeric transcripts (pBBF5 and pBBF6) ( Fig. 6 and Table I). The most pronounced effect was seen in experiments performed with constructs containing the 3Ј-most HuD binding site (pBBF8). The half-life of the BBF8 mRNA in the absence of HuD was 82 Ϯ 18 min, whereas in the presence of HuD the half-life of the chimeric mRNA was Ͼ 3 h. DISCUSSION Previously, we have conducted an extensive analysis of ELAV-like RNA-binding protein activity in human NB. We were not able to detect HuC in the W-N and W-S cells and found that HuD was the only ELAV-like RNA-binding protein family member that was differentially expressed in these NB subclones (42). Furthermore, UV-cross-linking studies, gel mobility band shift assays, and immunoprecipitation experiments indicated that HuD protein binds directly to two MYCN 3Ј-UTR sequences (42). Others have similarly reported that HuD binds with high specificity to AREs within the MYCN 3Ј-UTR (20). In this study, we characterized the ability of cis-acting elements within the MYCN 3Ј-UTR to destabilize mRNA in cells, and examined the functional significance of interactions between HuD and these elements in vivo. We demonstrate that the MYCN 3Ј-UTR contains at least four cis-acting regulatory elements that accelerate the cellular degradation of the stable ␤-globin mRNA. Two of these four cis-acting destabilizing elements harbor HuD binding sites. We also provide the first in vivo evidence that HuD participates in the regulation of AREmediated MYCN mRNA turnover as ectopic overexpression of exogenous HuD dramatically inhibits MYCN ARE-directed mRNA decay. Furthermore, HuD stabilizes heterologous mRNA containing MYCN 3Ј-UTR fragments that harbor the HuD binding sites. In contrast, HuD does not inhibit the degradation of heterologous mRNA containing the other two MYCN 3Ј-UTR instability elements. Similar to a previous report examining MYCN expression in NB cells stably transfected with antisense-HuD (20), our studies show that treatment of NB cells with HuD antisense oligomers results in a coordinate, significant decrease in MYCN expression. These results, together with data presented here and in other studies (20,42), strongly support a role for HuD in the post-transcriptional regulation of MYCN in NB.
Although relatively little is known about MYCN mRNA metabolism, a plethora of information exists about the decay of other labile mRNAs, including granulocyte macrophage-colony-stimulating factor and the c-fos and c-myc oncogenes (24, 43, The level of expression of the transcripts at each time point was determined by RNase protection assays. Quantitation of data was obtained through densitometry. Representative RNase protection assays demonstrating the decay of mRNA from the reporter constructs pBBF1 and pBBF8 are shown. 46 -50). Control of mRNA degradation for these and other unstable transcripts has been shown to be mediated largely by AREs in the 3Ј-UTR (51). There have been many reports identifying ARE-binding activity in crude cell extracts, and it is thought that the binding activity of these proteins plays a role in regulating the stabilities of ARE-containing mRNAs (21,26). However, the functional consequences and physiologic significance of the in vitro interactions for most ARE-binding proteins remain unknown.
The ELAV proteins are among the best characterized RNAbinding proteins. ELAV, the founder member of this family, is a Drosophila RNA-binding protein required for neuronal differentiation (52). This family of RNA-binding proteins is highly conserved in vertebrates, and to date, four human homologues have been cloned. HuR is ubiquitously expressed in proliferat-ing cells (53), whereas Hel-N1, HuC, and HuD are expressed exclusively in postmitotic neurons and in specific neuroendocrine tumors (33,34). The unique structural feature of the ELAV proteins is the organization of their three ribonucleoprotein-2/ribonucleoprotein-1 type RNA recognition motifs (54). The first and second of these RNA recognition motifs are in tandem and are separated from the third by a segment rich in basic amino acids. Members of this family are able to bind simultaneously to the ARE and the poly(A) tail in vitro (55), and several laboratories have provided evidence indicating that the ELAV proteins play a critical role in mediating mRNA stabilization (27,31,32,38,56). Jain and co-workers (31) reported that overexpression of Hel-N1 leads to an enhancement of cytoplasmic expression of glucose transporter (GLUT1) mRNA, which bears a U-rich region in its 3Ј-UTR. HuD has been shown to be required for GAP-43 mRNA stability (56), and overexpression of HuR results in enhanced stability of VEGF mRNA (27). Furthermore, recent studies conducted in the laboratories of Shyu and Steitz (28,38) using two different approaches have demonstrated that HuR increases the in vivo stability of ARE-containing mRNAs.
An intimate association between the ELAV family of proteins and the neuronal phenotype has been clearly established in developing and mature neurons (33,52,57). Therefore, it is not surprising that all three neuronal-specific ELAV proteins (HuD, Hel-N1, and HuC) are expressed in many NB cell lines and primary NB tumors, which are derived from the neural crest (20,42,58). Using a sensitive and quantitative RNase protection assay, Ball and King (58) detected HuD and Hel-N1 mRNA expression in 35 of 36 (97%) primary NB tumors. HuD is thought to function in the initiation of neurite outgrowth, in part, by increasing GAP-43 mRNA expression. In support of this hypothesis, overexpression of HuD in PC12 cells and cortical cultures has been shown to result in neuronal differentiation with the extension of neurites (56,59). In NB, an association between high levels of HuD and Hel-N1 expression and prognostically favorable features, including single copy MYCN and localized stage of disease has been reported (58). However, because tumor histology was not evaluated in that study, it is not possible to determine if the tumors that expressed high levels of HuD had a more differentiated phenotype. Although less HuD was detected in the MYCN-amplified samples, it is possible that the lower level of expression, which was defined as below the mean for the entire group of 36 tumors, was sufficient to enhance MYCN mRNA stability.
In conclusion, our studies indicate that HuD stabilizes MYCN mRNA in MYCN-amplified NB cells, resulting in increased levels of steady-state MYCN. It is well established that tumors with MYCN amplification and consequent high levels of expression are clinically aggressive (3,4,11). Furthermore, numerous laboratory studies have shown that NB phenotype is strongly correlated with the steady-state levels of MYCN expression (12-16, 18, 60). Modulation of mRNA stability provides a powerful means for controlling gene expression. Thus, further studies investigating the mechanisms by which HuD mediates MYCN mRNA turnover may lead to new strategies to down-regulate MYCN expression and thereby inhibit NB tumor growth.