Thyroid hormones stabilize acetylcholinesterase mRNA in neuro-2A cells that overexpress the beta 1 thyroid receptor.

We investigated the intracellular events involved in the 3,3',5-triiodo-L-thyronine (T3)-induced accumulation in acetylcholinesterase (AChE) activity in neuroblastoma cells (neuro-2a) that overexpress the human thyroid receptor beta 1 (hTR beta 1). Treatment of these cells with T3 increased AChE activity and its mRNAs after a lag period of 24-48 h, and these levels increased through stabilization of the transcripts by T3. T3 had no effect on the transcriptional rate or processing of AChE transcripts. The protein kinase inhibitor H7 inhibited T3-induced accumulation in AChE activity and its mRNAs, whereas okadaic acid (a potent inhibitor of phosphatases 1 and 2A) potentiated the effect of T3. Okadaic acid and H7 have no effect on the binding of hTR beta 1 to T3 or the transcriptional rate of the AChE gene. Finally, treatment of cells with T3 stimulated cytosolic serine/threonine, but not tyrosine kinase, activities. The time course analysis reveals that the increase in serine/threonine activity precedes the effect of T3 on AChE mRNAs. These results suggest that activation of a serine/threonine protein kinase pathway might be a link between nuclear thyroid hormone receptor activation and stabilization of AChE mRNA.

We investigated the intracellular events involved in the 3,3,5-triiodo-L-thyronine (T 3 )-induced accumulation in acetylcholinesterase (AChE) activity in neuroblastoma cells (neuro-2a) that overexpress the human thyroid receptor ␤1 (hTR␤1). Treatment of these cells with T 3 increased AChE activity and its mRNAs after a lag period of 24 -48 h, and these levels increased through stabilization of the transcripts by T 3 . T 3 had no effect on the transcriptional rate or processing of AChE transcripts. The protein kinase inhibitor H7 inhibited T 3induced accumulation in AChE activity and its mRNAs, whereas okadaic acid (a potent inhibitor of phosphatases 1 and 2A) potentiated the effect of T 3 . Okadaic acid and H7 have no effect on the binding of hTR␤1 to T 3 or the transcriptional rate of the AChE gene. Finally, treatment of cells with T 3 stimulated cytosolic serine/threonine, but not tyrosine kinase, activities. The time course analysis reveals that the increase in serine/threonine activity precedes the effect of T 3 on AChE mRNAs. These results suggest that activation of a serine/threonine protein kinase pathway might be a link between nuclear thyroid hormone receptor activation and stabilization of AChE mRNA.
Acetylcholinesterase (AChE; EC 3.1.1.7) 1 is responsible for the hydrolysis of acetylcholine at peripheral and central cholinergic synapses. This enzyme is encoded by a single gene in those vertebrate species examined to date, and the structural diversity in gene products arises from alternative mRNA processing and post-translational associations of catalytic and structural subunits (1)(2)(3)(4)(5).
Several lines of evidence suggest that thyroid hormones are involved in regulating AChE activity in neural cells: AChE activity is decreased in the brains of neonate thyroidectomized rats (6 -8), and triiodo-L-thyronine (T 3 ) increases AChE activity in primary neuronal cultures initiated from fetal rat brains (9). Much less is known about the molecular basis for hormone control. It is widely accepted that thyroid hormones exert their actions via nuclear receptors, which regulate the transcription of thyroid hormone-responsive genes. Three forms of thyroid receptors have been identified in neural cells: TR␣1, TR␤1, and TR␤2 (for review, see Ref. 10). Recently, we showed that T 3 increases the activity of AChE and its mRNAs in neuro-2a cells, which specifically overexpress human thyroid receptor ␤1 (hTR␤1), indicating that the control of AChE gene expression by T 3 occurs at a pretranslational step (11). Furthermore, the lag time of several hours following hormone administration suggests that this effect is likely to be indirect, mediated by unknown factors.
Phosphorylation events regulate AChE activity. 1-(5-Isoquinolinylsulfonyl)-2-methylpiperazine (H7), a potent inhibitor of protein kinase, stimulates AChE activity in neuro-2a cells (12). Phosphorylation events also participate in thyroid hormone action. Continued protein phosphorylation is required for stimulation of transcription of the lipogenic genes by T 3 (13), and transient transfection assays showed that phosphorylation of TR isoforms increases their transcriptional activity (14,15). Phosphorylation of the TRs could lead to a change in either hormone binding or DNA binding of the receptor (16). Another possibility is that the target for the phosphorylation events may be one or more of the auxiliary proteins or coactivators that are important in T 3 -dependent transcription (17,18). Alternatively, phosphorylation could mediate the post-transcriptional effects of thyroid hormones. Thyroid hormones regulate the turnover of numerous mRNAs (19 -23). Most of these effects are typically blocked by inhibitors of protein synthesis, suggesting the possible involvement of one or more components that influence the degradative pathway for specific mRNAs. Phosphorylation events also participate in the control of mRNA turnover. The turnover of several mRNAs is regulated by agents that activate the protein kinase C or A pathway either directly (such as phorbol esters or cAMP) or indirectly (such as via receptor activation) (24 -31). It is therefore possible that activation of a protein kinase pathway might be the link between thyroid receptor activation and mRNA turnover. However, there is no example of such a mechanism for thyroid hormones.
In the present study, we investigated the mechanism of regulation of AChE gene expression by T 3 using neuro-2a cells that overexpress hTR␤1 as a model. Our results suggest that activation of a serine/threonine protein kinase pathway may be the link between thyroid receptor activation and the stabilization of AChE mRNA induced by T 3 .

EXPERIMENTAL PROCEDURES
Tissue Culture-Neuro-2a cells that overexpress hTR␤1 were cultured in Dulbecco's modified Eagle's medium and F-12 in a 1:1 (v/v) ratio, supplemented with 10% fetal calf serum as described previously (11). The serum was depleted of thyroid hormones according to the procedure of Samuels et al. (32). Cells were grown at 37°C in a humidified atmosphere of 5% CO 2 , 95% air.
RNA Preparation and Northern Blot Hybridization-Total RNA was prepared by the guanidinium thiocyanate procedure (33), and poly(A) ϩ mRNA was prepared as described by Hartmann et al. (34). RNA concentrations were determined spectrophotometrically by measuring the A 260 . For Northern blot, 6 g of poly(A) ϩ mRNA or 15 g of total RNA was separated on 1.5% agarose gels containing 0.66 M formaldehyde and transferred to nitrocellulose membranes as described (11). * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Specific oligonucleotides (nucleotides 1779 -1824 and 1936 -1979 of mouse AChE cDNA) were synthesized, 3Ј end-labeled, and used to probe endogenous AChE mRNAs as described (11). The recovery of RNAs and poly(A) ϩ was determined by hybridization with either a 3Ј end-labeled 18 S rRNA-specific oligonucleotide (35) or a 32 P-labeled cDNA cyclophilin probe, an unrelated gene constitutively expressed in cells (36), respectively. All values were normalized to cyclophilin or 18 S rRNA signal. Quantitation of relative RNA levels of autoradiograms was determined by densitometry (BioImage-Visage 110S from Millipore Corp., Ann Arbor, MI).
Transcription assays were performed essentially as described (38). Nuclei were resuspended in buffer containing 250 Ci of [␣ 32 P]UTP (Amersham Corp.) and 1 mM each of ATP, CTP, and GTP and 5 mM dithiothreitol. For experiments with ␣-amanitin (from Sigma), ␣-amanitin (2 g/ml) was added to medium prior initiating synthesis. After incubation at 30°C for 30 min, the radiolabeled RNAs were isolated by ethanol precipitation after DNase I and proteinase K treatment. Total incorporation of [␣ 32 P]UTP into RNAs was 3-6 ϫ 10 6 cpm/10 7 nuclei.
The radiolabeled RNAs were hybridized to dot blots containing 5 g of plasmid DNA. Plasmids containing the 2.1-kb AChE insert and a control plasmid (pcDNA3) were linearized with HindIII. In previous experiments, we showed that the transcription factor NGFI-A (39) (also known as zif268 (40), Krox-24 (41), Erg-1 (42), TIS28 (43), and CEF-5 (44)) is transcriptionally regulated by T 3 in these cells. 2 Thus, a plasmid containing the 3.2-kb NGFI-A insert (from ATCC) was used as positive control. 2 ϫ 10 6 cpm were hybridized to the blots in 10 mM Tris-HCl, pH 7.4, 0.2% SDS, 10 mM EDTA, 300 mM NaCl, 1 ϫ Denhardt's solution, and 200 mg/ml Escherichia coli tRNA. Blots were incubated for 36 h at 65°C, washed in 2 ϫ SSC and 0.1% SDS at room temperature, and then washed at high stringency (0.2 ϫ SSC and 0.1% SDS at 65°C) for 1 h. 5,6-Dichloro-1-␤-D-ribofuranosylbenzimidazole (DRB) Inhibition of Transcription-Neuro-2a cells were maintained for 3 days in cultured media with and without T 3 (30 nM). On day 3, cells were extensively washed with fresh media and treated for the specified times with 20 g/ml DRB, which was dissolved (4 mg/ml) in ethanol and diluted 1/200 in cultured media. The same volume of ethanol was added to controls. Total RNA was isolated and used for Northern blot. Blots were normalized to the 18 S rRNA.
Determination of AChE Activity-AChE activity was determined by colorimetrically measuring the rate of hydrolysis of acetylthiocholine (45). Protein determinations were performed by the method of Lowry et al. (46).
Determination of T 3 Binding Capacity-Cells were grown for 24 h in the absence or presence of either H7 (25 M) or okadaic acid (OA) (10 nM). Then 1 nM 125 I-labeled T 3 was added to the cultures, and cells were further incubated for 3 h at 37°C. After incubation, the nuclei were isolated, and the radioactivity was determined as described (37). Nonspecific binding was determined by incubating cells with a 1000-fold excess of nonradioactive T 3 . DNA was assayed by the method of Burton (47).
Measurement of Kinase Activity-Cells grown in the presence or absence of T 3 (30 nM) were homogenized in 10 mM Tris-HCl, pH 7.4, 15 mM MgCl 2 , 10 mM EDTA, 1 mM dithiothreitol, 17 g/ml phenylmethylsulfonyl fluoride, 5 g/ml leupeptin, 5 g/ml antipain, and 5 g/ml pepstatin and centrifuged at 10,000 ϫ g for 10 min at 4°C. The soluble fraction was centrifuged at 100,000 ϫ g for 60 min, and the resulting supernatants were stored at Ϫ80°C. Aliquots (5 g of protein) were mixed with 50 mM Hepes buffer, pH 7.4, containing 40 mM MgCl 2 and 30 g of myelin basic protein (MBP) or 75 g of a copolymer glutamic acid and tyrosine (4:1), poly(Glu/Tyr) (48). The reaction was initiated by adding [␥-32 P]ATP (final concentration 25 M, 32 Ci/nmol). After 30 min, the reaction was stopped by spotting aliquots on chromatography paper squares (Whatman No. 3MM, 2 ϫ 2), which were then placed in a 10% trichloroacetic acid bath containing 10 mM of sodium pyrophosphate for 60 min at 4°C. The squares were washed twice for 30 min at room temperature, rinsed with 100% ethanol, and air-dried. Radioactivity was determined by liquid scintillation counting. Incubation in the absence of substrate was used as control. Control values were subtracted from the total counts.

RESULTS
AChE Activity and mRNA Levels Are Induced by T 3 -Neuroblastoma cells that overexpress the thyroid receptor ␤1 were used to analyze the effect of T 3 on AChE gene expression. The time course analysis of the effect of T 3 on AChE activity is shown in Fig. 1 (upper panel). There was a significant increase in AChE activity after 24 h of treatment with 30 nM T 3 to 180% of the control value, with a maximal effect occurring after 48 h of treatment. The increase in AChE activity was paralleled by an increase in AChE mRNA, as determined by Northern analysis (Fig. 1, lower panel). Site-specific oligonucleotides recognized a major 2.4-kb mRNA species and a minor 3.2-kb mRNA species, which are the expected size of mouse AChE mRNAs (11,49). Treatment of cells with T 3 ( Tris buffer, pH 7.0, containing 1% Triton X-100 and 1 M NaCl, and AChE activity was determined. Lower panel, poly(A) ϩ -selected mRNA isolated from parallel culture plates was electrophoresed, blotted, and hybridized to AChE-specific oligonucleotides. Blots were exposed to Kodak XAR-5 film and quantified by densitometry. Cyclophilin cDNA probe was used to standardize for sample loading. Results represent the mean Ϯ S.E. of four and two experiments for AChE activity and mRNA, respectively.
3.2-kb AChE mRNA was not significantly changed by T 3 treatment (not shown). Because the 2.4-and 3.2-kb mRNAs are alternatively spliced AChE mRNAs (49,50), our finding indicates that T 3 had no effect on the process that altered the size of AChE mRNAs.
T 3 Stabilizes AChE mRNA-To determine whether the T 3induced accumulation in AChE mRNA resulted from an increased rate of transcription, we performed nuclear run-on measurements with nuclei isolated from neuro-2a cells maintained for 1 and 24 h in the presence of T 3 . Fig. 2 shows densitometric analysis of the data from two different experiments normalized to the cyclophilin signal. The transcription rate for NGFI-A gene showed a 2-fold increase (633 Ϯ 41 and 1384 Ϯ 260 arbitrary units, for control and T 3 -treated cells, respectively). The enhancement of the rate of transcription for NGFI-A gene was in accord with the changes in mRNA levels detected by Northern blot (not shown). No change was detected in the transcriptional rate for c-jun gene or the AChE gene in cells treated for 24 h with T 3 . The AChE signal is specific since it was completely abolished by ␣-amanitin at a concentration (2 g/ml) that specifically inhibited RNA polymerase II (38) (Fig.  2, upper panel). Similar results were obtained with cells treated for 1 h with T 3 (not shown), indicating that the effect of T 3 on AChE mRNAs is exerted at a post-transcriptional step.
The adenosine analog DRB specifically inhibits RNA polymerase II (51,52), and its capacity to block the transcription of AChE gene has rendered it useful in determining the half-life of AChE mRNAs (53). In the absence of T 3 , the AChE mRNA level decayed significantly after cells were treated with DRB ( Fig. 3). The calculated AChE mRNA half-life is about 4 -5 h, which is in agreement with the AChE mRNA half-life reported in muscle cells (53). In cells pretreated with T 3 for 72 h, no decrease in AChE mRNA levels was evident over the next 24 h, suggesting the presence of a very stable mRNA following a global block of transcription (Fig. 3). Furthermore, the addition of OA (10 nM) prevented the decrease in AChE mRNA levels observed 48 h after DRB treatment, suggesting that OA further stabilizes the mRNAs. The addition of DRB 15 min before treatment with T 3 completely inhibited T 3 's ability to increase AChE mRNA levels (Fig. 4).
Phosphorylation Events Mediate the T 3 -induced Accumulation of AChE Activity and Its mRNA-OA, a selective inhibitor of serine/threonine protein phosphatases 1 and 2A (54), was used to determine whether phosphorylation on serine/threonine residues is involved in T 3 -induced AChE activity. Treatment of cells with OA for 24 h increased the activity of AChE in a dose-dependent manner in cultures grown in the absence of T 3 (Fig. 5, upper panel). The simultaneous addition of T 3 (30 nM) and OA (5-20 nM) to the culture caused a synergistic effect. At 5 nM, OA had no significant effect on AChE activity in the absence of T 3 , whereas it potentiated the effect of T 3 on AChE activity. These results suggest that T 3 regulated a protein kinase that increases phosphorylation on serine/threonine res- idues, thereby increasing AChE activity. To test this possibility, we examined the effect of protein kinase inhibitors (H7, calphostin C, and N-(2-guanidinoethyl)-5-isoquinolinesulfonamide (HA1004)) on the T 3 -induced accumulation in AChE activity. H7 (5-50 M) had no effect on AChE activity in cells grown in the absence of T 3 (not shown). The increase in AChE activity caused by the addition of T 3 for 24 h was inhibited by 90% when H7 (50 M) was added for the same period of time as T 3 (Fig. 5, lower panel). Fifty percent of maximal inhibition was achieved at about 10 -15 M H7. On the other hand, calphostin C (100 -400 nM, a specific inhibitor of protein kinase C) and HA1004 (a specific inhibitor of protein kinase A) had no effect on the T 3 -induced accumulation of AChE activity (Table I). Furthermore, treatment of cells with various activators (8bromo-cAMP, phorbol 12-myristate 13-acetate) did not affect basal AChE activity or T 3 -induced accumulation in AChE activity (Table I).
To determine whether phosphorylation mediates the effect of T 3 on AChE mRNA, we examined the effect of H7 and OA on the T 3 -induced accumulation in AChE mRNA. H7 (20 M), added coincidentally with T 3 , caused an almost 90% inhibition of the T 3 -induced accumulation of AChE mRNA after 24 h, whereas treatment of cells with OA for 24 h potentiated the effect of T 3 on AChE mRNA levels (not shown). A transcriptional effect of H7 and OA on AChE gene was excluded by run-on transcriptional analysis (not shown).
T 3 Stimulates Serine/Threonine Protein Kinase Activities-To determine whether T 3 activates a protein kinase pathway, serine/threonine and tyrosine kinase activities were measured in the cytosolic fractions prepared from treated and untreated cells using MBP and poly(Glu/Tyr) as substrates, respectively. As shown in Table II, treatment of cells with T 3 (30 nM) induced a 1.5-and 2.1-fold increase in serine/threonine protein kinase activities after 16 and 24 h of treatment, respectively. Phosphoamino acid analysis showed that MBP was phosphorylated predominantly on serine and, to a lesser extent, on threonine residues (not shown). No significant effect of T 3 was observed on phosphorylation of poly(Glu/Tyr) (183 Ϯ 23 and 190 Ϯ 15 for untreated and cells treated for 24 h with T 3 , respectively). Furthermore, treatment of cells with genistein, a specific inhibitor of tyrosine kinase, had no effect on the T 3induced serine/threonine protein kinase activity (not shown), indicating that the effect of T 3 on serine/threonine protein activity is not mediated by a tyrosine kinase.
T 3 Binding Capacity-To exclude an effect of H7 and OA on the binding of T 3 to its receptors, H7 and OA were added for 24 h to neuro-2a cells and then T 3 binding capacity was measured in nuclei isolated from intact cells. H7 (25 M) and OA (10 nM) had no effect on T 3 binding capacity (control, 1967 Ϯ 221; H7, 1966 Ϯ 548; OA, 1742 Ϯ 200 fmol/mg DNA; mean Ϯ S.D., n ϭ 3), indicating that the effect of H7 and OA was not caused by a change in T 3 binding to its receptors.

DISCUSSION
In neuro-2a cells that overexpress hTR␤1, we previously showed that the effect of T 3 on AChE gene expression is independent of its antiproliferative effect, indicating that both effects are regulated by distinct mechanisms (11).
We show here that the T 3 -induced accumulation in AChE activity and its mRNA result from stabilization of the mRNAs. No effect of T 3 was observed on the transcriptional rate or processing of AChE mRNA. Although we previously reported that the increase in 3.2-kb mRNAs was much greater than the increase of 2.4-kb mRNAs after T 3 treatment (11), this result was not confirmed in the present study, in which we analyzed a larger number of Northern blots. It is likely that nuclear thyroid receptors indirectly participate in regulated AChE mRNA turnover, because they are required for the transcriptional regulation of one or more components that influence the degradative pathway of AChE mRNAs. Consistent with this point is the suppression of the T 3 -induced accumulation in AChE mRNAs after a global block of transcription.
The protein kinase inhibitor H7 blocks the ability of T 3 to increase AChE activity and its mRNAs. H7 had no effect on AChE activity, AChE mRNA levels, or the rate of AChE gene transcription in the absence of T 3 which suggests that H7 selectively inhibits the increase in AChE mRNA induced by T 3 . At 500 M, however, H7 increases AChE activity in neuro-2a cells (12). Isoquinoline sulfonamide derivatives vary with respect to the affinity with which they bind to different protein kinases (55). In our study, half-maximal inhibition of the activity of AChE enzyme occurs at about 10 -15 M H7. This concentration of H7 is slightly higher than the K i values of protein kinase C (6 M). It is likely that at higher concentration, H7 may inhibit other protein kinases that are also involved in the regulation of AChE enzyme activity. In contrast, OA potentiated the effect of T 3 on AChE activity and its mRNA. The finding that OA had no effect on the transcription of the AChE gene indicates that OA selectively potentiates the effect of T 3 at a post-transcriptional step. This effect was observed at a concentration of 5 nM OA, which is in the range of the concentration necessary to inhibit protein phosphatases 1 and 2A (0.1-10 nM) (54). At this concentration, OA does not inhibit protein phosphatase 2B and 2C, protein-tyrosine phosphatases, acid and alkaline phosphatase (54), indicating that OA selectively inhibits serine/threonine phosphatases 1 and 2A. Furthermore, we showed that T 3 stimulated cytosolic serine/ threonine but not protein tyrosine kinase activities. The time course analysis revealed that the effect of T 3 on serine/threonine protein kinase activity precedes its effect on AChE mRNAs, supporting the hypothesis that activation of a serine/ threonine protein kinase pathway might be the intermediate between thyroid receptor activation and stabilization of the AChE mRNA. Attempts were made to identify the serine/threonine kinase(s). However, our study suggests that the endogenous serine/threonine kinase (s) involved in the stabilization of AChE mRNAs did not have specificities similar to those of cyclic AMP-dependent protein kinase, protein kinase C, phosphorylase kinase, glycogen synthetase kinase 3, and casein kinases I and II.
Our working hypothesis is that T 3 activates a serine/threonine protein kinase, which in turn causes the stabilization of AChE mRNAs. The mechanisms by which protein phosphorylation controls the stability of AChE mRNA are unknown. The binding activity of an inducible stabilizing factor that recognizes the AU-rich motif in the 3Ј-untranslated region of granulocyte-macrophage colony-stimulating factor is dependent on its phosphorylation state (56). It is therefore possible that phosphorylation of cytosolic turnover factor(s) could influence their binding activity to specific sequence in AChE mRNAs, which in turn causes the stabilization of AChE transcripts.
In conclusion, the present study provides the first evidence that activation of the protein kinase pathway might be the linkage between nuclear thyroid receptor activation and a posttranscriptional effect of thyroid hormones.  (30 nM) in the presence or absence of various activators or inhibitors of protein kinases. AChE activity was determined as described in the legend to in Fig. 1