Induction of calmodulin kinase IV by the thyroid hormone during the development of rat brain.

This communication reports the specific induction of calmodulin kinase IV by the thyroid hormone 3,3',5-triiodo-L-thyronine (T3) in a time- and concentration-dependent manner at a very early stage of brain differentiation using a fetal rat telencephalon primary cell culture system, which can grow and differentiate under chemically defined conditions. The induction of the enzyme that can be observed both on the mRNA and on the protein level is T3-specific, i.e. it cannot be induced by retinoic acid or reverse T3, and can be inhibited on both the transcriptional and the translational level by adding to the culture medium actinomycin D or cycloheximide, respectively. The earliest detection of calmodulin kinase IV in the fetal brain tissue of the rat is at days E16/E17, both on the mRNA as well as on the protein level. This is the first report in which a second messenger-dependent kinase involved in the control of cell regulatory processes is itself controlled by a primary messenger, the thyroid hormone.

Since thyroid hormones have been shown to be required for normal growth and differentiation of the mammalian brain (26 -28), we initiated studies to investigate the influence of the thyroid hormone 3,3Ј,5-triiodo-L-thyronine (T 3 ) on the expression of a number of neuronal and glial membrane markers during development in rat brain cell cultures (29 -31). Here we report the specific induction of CaMKIV by T 3 in a time-and concentration-dependent manner at a very early stage of brain differentiation using a fetal rat telencephalon primary cell culture system, which can grow and differentiate under chemically defined conditions (32). The induction is T 3 -specific, i.e. the expression of the enzyme cannot be induced by either reverse T 3 or retinoic acid. The expression of CaMKIV is regulated on both the transcriptional and the translational level, since both the addition of actinomycin D as well as cycloheximide to the cultural medium can prevent the T 3 -dependent induction of the enzyme. In addition, the T 3 -specific induction can be observed both on the mRNA and on the protein level. This is the first report in which a second messenger-dependent kinase involved in the control of cell regulatory processes is itself controlled by a primary messenger, the thyroid hormone. Preliminary accounts of part of the data presented here have been given elsewhere (33,34).

EXPERIMENTAL PROCEDURES
Cell Culture-Serum-free, rotation-mediated aggregating cell cultures were prepared from fetal (15 days of gestation) rat (OFA/Ico/Ibm strain, Biological Research Laboratories Ltd., Basel, Switzerland) telencephalon, as described in detail previously (32). Aliquots of 6 ϫ 10 7 cells were transferred to DeLong flasks and maintained under constant gyratory agitation at 37°C, in an atmosphere of 10% CO 2 and 90% humidified air. The culture medium used was Dulbecco's modified Eagle's medium (Life Technologies, Inc.), supplemented with nutritional factors, vitamins, trace elements, transferrin (1 mg/liter), insulin (800 nM), and hydrocortisone 21-phosphate (20 nM). Gentamicin sulfate (25 * This work was supported by the Swiss National Science Foundation Grants 31-30858.91 and 31-37292.93 (to J. K. and P. H.), and by a National Institutes of Health Grant HD-07503 (to A. R. Means). 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.
SDS-PAGE and Electrophoretic Blotting-To 40 l of supernatants of cellular extracts, 20 l of loading buffer (35) was added and the proteins were separated on a 10% SDS-PAGE (35). After electrophoretic transfer onto nitrocellulose sheets for 2 h using 240 mM glycine, 30 mM Tris, 20% methanol, and 0.02% SDS as transfer buffer, the blots were incubated overnight at 4°C with 2% milk powder in PBS as blocking buffer, afterward incubated at room temperature for 3 h with 125 I-CaM (36) (2 ϫ 10 5 cpm/ml) in 2% milk powder/PBS containing either 1 mM CaCl 2 or 5 mM EGTA, washed once with 2% milk powder/PBS containing either CaCl 2 or EGTA, then with PBS containing either CaCl 2 or EGTA, and third with PBS. Each washing was done for 15 min. After drying the blots under cold air, the sheets were exposed either to x-ray sensitive films (Kodak XAR-5) at Ϫ70°C or to PhosphorImager screens (Molecular Dynamics, Sunnyvale, CA) at room temperature and developed. Quantitation was obtained using the PhosphorImager software as recommended by the manufacturers. CaM was purified from bovine brain as described previously (37).
Preparation and Characterization of Total RNA-For RNA preparation, cell culture aggregates (6 -8 ϫ 10 8 cells) were grown for 5 days either in the absence or in the presence of 3 ϫ 10 Ϫ8 M T 3 , washed three times with PBS, and quickly frozen in liquid nitrogen. Total RNA was isolated by using the Ultraspec RNA kit (Biotecx Laboratories, Inc., Houston, TX) as described by the manufacturer. For the identification of CaMKIV-specific mRNA, 40 g of total RNA was denatured and size-fractionated on a 1.5% agarose, 2.2 M formaldehyde gel. Electrophoresis was carried out at 30 V for 16 h with circulating Na 3 PO 4 buffer at 4°C. Prior to blotting the gels were stained with ethidium bromide to identify 18 and 28 S rRNA as markers. The RNA was transferred onto Biotrans nylon membranes (ICN Biomedicals, Inc., Costa Mesa, CA) and stabilized by using UV cross-linking. Northern blots were prehybridized and hybridized at 37°C in 50% formamide as described previously (38). The probe was a 1-kilobase pair (BamHI-KpnI) fragment of the rat CaMKIV cDNA (4) and was labeled with [␣-32 P]dCTP according to the random oligonucleotide priming method of Feinberg and Vogelstein (39).
Purification and Characterization of p64 -Aggregates of a total of 20 culture flasks (ϳ10 9 cells), grown for 5 days in the presence of T 3 (3 ϫ 10 Ϫ8 M), have been homogenized at 4°C in three volumes of ice-cold buffer A, basically following the purification protocol for CaMKIV described in Ref. 40. Briefly, the lysates were centrifuged at 100,000 ϫ g for 30 min, and the supernatants were pooled and applied to a 15-ml DEAE-cellulose column equilibrated with 25 mM Hepes, pH 7.5, 0.5 mM EGTA, 0.5 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, and 1 mM DTT (buffer B). The column was washed with 10 bed volumes of buffer B, and bound proteins were eluted with a stepwise gradient using buffer B and increments of 100 mM NaCl at each step. p64 eluted between 150 and 300 mM NaCl. The fractions containing p64 were pooled and applied onto a prepacked 1-ml hydroxylapatite column from Bio-Rad equilibrated with buffer B containing 200 mM NaCl (buffer C). The column was washed with five bed volumes of buffer C, and bound proteins were eluted with a linear 0 -200 mM NaH 2 PO 4 gradient. p64containing fractions eluting between 75 and 100 mM NaH 2 PO 4 were pooled, dialyzed overnight against 5 liters of 25 mM Hepes, pH 7.5, 2 mM CaCl 2 , 2 mM MgCl 2 , 1 mM DTT (buffer D), and applied to a 5-ml CaM column (36) equilibrated with buffer D. After washing the column with 10 bed volumes of buffer D containing 10% glycerol (buffer E), followed by a wash of five bed volumes with buffer D containing 0.5 M NaCl, p64 was eluted by a buffer containing 25 mM Hepes, pH 7.5, 2 mM EDTA, 5 mM EGTA, 1 mM DTT, 10% glycerol. Purified p64 was identified as a calmodulin-dependent kinase by autophosphorylation using the assay as described by Cruzalegui and Means (11). Autophosphorylated CaMKII ␣ and ␤ were identified by specific antibodies, kindly provided by Dr. H. Schulman. Identification of p64 as CaMKIV by specific antibodies was carried out by Western blotting using 2% milk powder/PBS as blocking buffer.

RESULTS AND DISCUSSION
Aggregating cell cultures prepared from 15-day fetal rat telencephalon and grown for 5 days in the presence or absence of the thyroid hormone 3,3Ј,5-triiodo-L-thyronine (T 3 ) in a chemically defined medium were collected and extracted with an EGTA-containing buffer. The extracted proteins were separated by SDS-polyacrylamide gel electrophoresis and electrophoretically transferred onto nitrocellulose, and CaM-binding proteins were identified by incubation with 125 I-labeled CaM in the presence or absence of Ca 2ϩ . As shown in Fig. 1A, a number of different CaM-binding proteins were present in the M r range between 40,000 and 160,000, which could not be detected in the presence of EGTA (data not shown). By comparing cultures that were grown for 5 days either in the presence or absence of 3 ϫ 10 Ϫ8 M T 3 , it was obvious that one CaM-binding protein with a M r of 64,000 (later identified as CaMKIV; see below) appeared only in cultures grown in the presence of T 3 , suggesting that its expression depended on this hormone. This result could be corroborated by the observation that in Northern blots of total RNA isolated from similar cell cultures a mRNA of 2 kilobases could be clearly identified by a probe specific for CaMKIV (Fig. 1B). This band could only be observed in the presence of T 3 , but not in its absence (even at higher RNA concentrations; data not shown). On the other hand, a band of 3.5 kilobases, which has been described before as a much less abundant mRNA of CaMKIV in adult rat brain (3,4), could be observed as a weak band even in the absence of T 3 -induced cultures. The difference between the two mRNAs could derive from differences in the 3Ј-untranslated region by use of alternate poly(A) sites as discussed before (4), but the reason for the apparent difference in the T 3 -dependent expression between the two mRNAs needs further investigation.
Since the apparent molecular weight and autophosphorylation kinetics showed similarities between p64 and CaMKIV we attempted to purify p64 from a pool of aggregate cultures  Further purification was obtained by using a hydroxylapatite column, and finally a CaM affinity column. After extensive washing with a calcium-containing buffer, p64 was eluted using an EGTA-containing buffer (Fig. 2). The purified protein was identified as a CaM-dependent kinase by CaM-dependent autophosphorylation (Fig. 2B) and as CaMKIV by specific antibodies ( Fig. 2A). Using immunoprecipitation CaMKIV could be identified only in cells that had been grown in the presence of T 3 (data not shown). With respect to its M r of 64,000 -65,000 and its prenatal appearance, p64 most likely represents the ␣-isoform of CaMKIV since the ␤-polypeptide has a slightly higher M r , i.e. 67,000, is specifically expressed in brain only in cerebellum, and can be observed only postnatally (6).
It was noted that autophosphorylation of pure p64 was rather slow, as described by Cruzalegui and Means (11) for a recombinant CaMKIV. On the other hand, slow phosphorylation of p64 could also be indicative for the presence of contaminating amounts of the recently described CaMKIV kinase (13)(14)(15)(16). The other CaM-binding proteins exhibiting M r values between 60,000 and 63,000 were identified as CaMKII␤ and calcineurin A, respectively, by using monoclonal antibodies against CaMKII␤ and by comparison with purified calcineurin. It appeared that in contrast to CaMKIV, the expression of both CaMKII and calcineurin was independent of T 3 (data not shown).
The influence of T 3 was studied further in a dose-and timedependent expression of CaMKIV. As shown in Fig. 3, the enzyme was detectable already at very low concentrations of T 3 (3 ϫ 10 Ϫ10 M; Fig. 3) and increased in intensity with increasing concentrations of T 3 in the culture medium, indicating that the induction of CaMKIV by T 3 was dose-dependent. In addition, when cultures received T 3 (3 ϫ 10 Ϫ8 M) for various lengths of time, CaMKIV was already detectable after 6 h of stimulation (Fig. 4), and the amount of CaMKIV increased as a function of the duration of the stimulus until maximal expression was reached between 24 and 48 h, suggesting that protein synthesis played a role in the induction of this gene. This interpretation was corroborated by incubating the cell cultures with either actinomycin D or cycloheximide, respectively, to prevent transcription or protein synthesis of the inducible gene. As can be seen from Fig. 5, CaMKIV was clearly induced after exposure to 3 ϫ 10 Ϫ8 M T 3 for 24 h (Fig. 5, lane 3; see also Fig. 4, lane 4), but the protein was not detectable if the cultures had been preincubated for 1 h with either 1 M actinomycin D (Fig. 5, lane 5) or 5 M cycloheximide (Fig. 5, lane 4), respectively, before T 3 was added. This is in contrast to the other detectable calmodulin-binding proteins, which during the time period of observation (i.e. 24 h) were independent of transcription or translation.
A further observation supporting the view that the expression of CaMKIV was a T 3 -specific process was the finding that next to T 3 only T 4 was able to induce CaMKIV (Fig. 6, lanes 2  and 3). On the other hand, neither reverse T 3 nor retinoic acid could induce the expression of CaMKIV (Fig. 6, lanes 4 and 5). In addition, neither nerve growth factor nor epidermal growth factor were able to induce the expression of the kinase (data not shown), although both growth factors have been demonstrated to regulate developmental processes in these cultures (41,42). Also, CaMKIV may be specifically expressed in neurons, since it was found at high levels in neuron-enriched aggregate cultures (data not shown), in which highly proliferating glial cells have been suppressed by the addition of 1-␤-D-arabino-furanosylcytosine (Ara-C) (31).
Recent reports suggested that CaMKIV is responsible for the Ca 2ϩ -dependent regulation of expression of a number of immediate early genes such as c-fos, due to the phosphorylation of the cAMP-responsive element-binding protein (CREB) (12,18,23,25) or the serum response factor SRF (19). Our results indicate that during rat brain development the expression of CaMKIV, not detectable at the early stages of ontogenesis (i.e. at E15; data not shown) is regulated by the thyroid hormone in a time-and concentration-dependent manner. Whether this T 3 -dependent regulation is due to a direct interaction of the T 3 -receptor with a responsive element of the CaMKIV gene (28) or whether the effect is indirect remains to be determined, but since the T 3 -specific induction of CaMKIV could also be observed on the mRNA level, this observation could be indicative for a T 3 -receptor-dependent regulation. In this respect it is of interest that in a recent abstract (43), it was reported that in a mouse embryonic stem cell-derived neuronal culture system the expression of CaMKIV was strictly dependent on the presence of the thyroid hormone receptor. In addition, it should be noted that Shakagami et al.  1 and 5, control)   These findings strongly corroborate our observation that the presence of the thyroid hormone is essential for the expression of CaMKIV. It will be now of interest whether the expression of the CaMKIV kinase is also regulated during brain development and how it correlates with the expression of CaMKIV. In addition, it will be important to know whether in the immune system and in testis CaMKIV is also under the control of the thyroid hormone or whether different signals regulate the expression. Preliminary results indicate that the synthesis of CaMKIV in the rat embryonal thymus is induced following the same time course as in the embryonal brain.