JPhysiology (1993) 87, 153-157 153 © Elsevier, Paris

Tyrosine hydroxylase regulation in neurotransmission and neuroplasticity

C Icard-Liepkalns, S Berrard, N Faucon Biguet, B Lebourdelles, P Ravassard, JJ Robert, J Mallet

Laboratoire de Ggndtique Moldculaire de la Neurotransmission et des Processus Neurodggdngratifs, Centre National de la Recherche Scientifique, 91198 Gif-sur-Yvette, France.

Summary - The modulation of neurotransmitter synthesis is a fundamental mechanism influencing neurotransmission and neuronal plasticity during development. The regulation of the tyrosine hydroxylase (TH) has been used to elucidate specific adaptative responses in neurons. Trans-synaptic impulse activity elicits sort- and long-term changes in the activity of TH. Acute regulation involves the activation of preexisting TH molecules via phosphorylation and possibly through alternative splicing events in humans, whereas long-term regulation results from an increased synthesis of the enzyme due in part to the transcriptional stimulation of the TH gene. The long-term increase of TH activity was addressed using the drug reserpine known to modify the secretion of neurotransmitters and the tetradecanoyl phorbol acetate (TPA). Inductions of TH expression by reserpine in vivo as well as by TPA in vitro seem to be mediated by an AP-1 complex acting on a TPA responsive element (TRE) of the rat TH promoter indicating that the TRE-TH site plays a critical role in trans-synaptic induction. Our results also demonstrate a degree of adaption by sympathetic neurons to their environment by conversion from adrenergic to cholinergic phenotype.

neurotransmission / phosphorylation / transcriptional regulation / plasticity / tyrosine hydroxylase / choline acetyltransferase

I n t r o d u c t i o n

Neurotransmit ter (NT) expression can be regu- lated during development by the cellular environ- ment of the neuroblast, by the targest tissue being innervated (Le Douarin, 1980) and throughout life by neuronal activity (Joh et aI, 1973). We discuss here how studies on the regulation of the expres- sion of the tyrosine hydroxylase (TH) gene, which encodes an enzyme synthesizing neurotransmit- ters, have been used to elucidate specific adapta- tive responses in neurons. TH is the key enzyme in the biosynthesis of catecholamine which con- verts L-tyrosine into L-dopa and is a specific marker for adrenergic neurons. This enzyme has been widely used to study ontogenesis as well as the neuronal activity of the nervous system in the adult. The expression of TH in response to various stimuli can be regulated by short-term mecha-

nisms (0-60 s) at the post-translational level, and by long-term mechanisms (up to several weeks) at the transcriptional level. The maintenance of a quantitative regulation in the adult allows the nervous system and the tissues with which it in- teracts to respond and adapt to environmental fac- tors.

The peripheral nervous system can be used to study qualitative regulation of neuronal develop- ment because the neurons in the superior cervical ganglion (SCG) become either cholinergic or adrenergic. Both in v i vo and in v i t ro studies have shown that the NT choice is labile during a pro- longed period of life. In order to elucidate the molecular basis of this neurotransmitter-related phenotypic plasticity, the expression of the cho- line acetyltransferase, ChAT (the biosynthesis enzyme of acetylcholine) with that of TH was compared in sympathetic neurons from SCG grown under various culture conditions.

154

Short-term regulation of TH

Phosphorylation appears to be the most important short-term mechanism for controlling TH activity. In the rat, for example, TH is phosphorylated in vitro by a number of different protein kinases in- cluding adenosine 3",5'-monophosphate (cyclic AMP)-dependent protein kinase, protein kinase C and CaZ+/calmodulin-dependent protein kinase. Phosphorylation modifies the enzymatic charac- teristics of TH and in some cases increases its activity. Analysis of the amino acid sequence of TH indicates that the enzyme has two domains: the carboxyl-terminal region contains the catalytic site and the amino-terminal region acts as an in- hibitory regulatory domain. Phosphorylation of species residues in the amino-terminal regulatory domain has been shown to modulate TH activity. These phosphorylations have been shown to be mediated by second messengers such as Ca 2+ and cAME and involve multiple kinases, such as the cyclic AMP-dependent protein kinase (PKA) or the calmodulin-dependent protein kinase (CAM- PKII) and several phosphorylated serine residues have been identified in vitro and in v ivo (Camp- bell et al, 1986; Vulliet et al, 1989; Mitchell et al, 1990; Haycock, 1990).

An additional mode of regulation of the human enzyme could occur through differential splicing. In humans, four different TH mRNA (TH-I to TH-4) have been characterized in pheochromocy- toma tumors (Grima et al, 1987). TH-I and TH-2 forms, which are the predominant forms in the locus ceruleus and the substantia nigra of the human brain, differ by the insertion in the amino- terminal regulatory domain of an exon coding for four amino acid residues. This insertion in the TH-2 generates a consensus sequence for CaM- PKII (-R-X-X-S) (Pearson et al, 1985), and the adjacent serine (Ser-31) represents a putative site of phosphorylation by CaM-PKII, When injected in X e n o p u s oocytes, each of the four human mRNA species yielded functional TH proteins. These proteins show different specific activities, TH-1 being the most active and TH-2 the least, having approximately 70% of the activity of TH- 1 (Horellou et al, 1988). A bacterial expression sys- tem was used to study large amounts of TH-1 and TH-2. Total bacterial protein electrophoretic pat- terns revealed the presence of additional bands of 62 and 64 kDa corresponding to TH-I and TH-2 constructs respectively (LeBourdellbs et al, 1991). The schematic representation of the N-ter- minal regions of two TH isoforms with their

phosphorylation sites for PKA and CaM-PKII is shown in figure 1. Phosphorylation by PKA decreased the Km of the enzymes for the pyridine cofactor from 94.8 to 51.5 ~M for TH-I and 72.9 to 48.4 txM for TH-2, and thus increased enzyme- specific activity. No changes in kinetic parameters of the isoenzymes were observed after phospho- rylation by CaM-PKII since the Km of the phosphorylated TH-1 was 98.1 pM and 74.4 pM for the phosphorylated TH-2 vs 94.8 and 72.9 re- spectively for the non-phosphorylated forms.

Dopamine was found to inhibit both TH isoenzymes with similar Ki values. Unlike TH-1, phosphorylation of TH-2 by CaM-PKII results in an increase of the Ki value for dopamine (from 0.5 ~tM to 0.7 I.tM for TH-2). This property may be related to the presence of the additional phosphorylated residue in TH-2 form. Thus, al- ternative mRNA splicing may regulate both the basal activity of the enzyme and the modulation of activity by phosphorylation. It is therefore at- tractive to hypothesize that differential splicing provides an additional mode of control of cate- cholamine synthesis.

Long-term regulation of the TH gene

It is now well established that the increase in TH activity reflects higher concentration of TH mRNA (Faucon Biguet et al, 1986; Labatut et al, 1988). Therefore, this long-term increase in TH activity is due to de novo synthesis of enzyme molecules. This phenomenon, which may last up to several weeks, is termed trans-synaptic induc- tion. The time course of the increase in both TH

PKA

hTH1

hTH2

PKA ~.,,...

, m . i ~ - . s . ~ , , ~ s ~ . . , , , , . , , - -

PKA

C a M - P K I I

hTH1

hTH2

CaM-PKII CaM-PKl l A~,.. - ~ I - - - - - s ~ - - - - s ~ s - , , , , ..... - -

CaM-PKII CaM-PKIL CaM-PKIt

Fig 1. Phosphorylation sites by protein kinase A and Ca/cal modulin-dependent protein kinase of human TH-1 and TH-2 isoforms.

155

activity and the amount of TH mRNA caused by a single injection of reserpine has been analysed in rat brain locus coeruleus and adrenal medulla (AM). In addition, we have shown by nuclear run- on experiments that this increase of TH mRNA levels in PC12 cells following TPA treatment and in AM of reserpine injected rats is achieved by increased transcription of the gene (Icard-Liep- kalns et al, 1992). Analysis of the promoter of the TH gene has provided some information about the events that are responsible for the long-term induction of TH activity through gene activation. Several regulatory binding sites have been located in the promoter region of the rat TH gene (fig 2). Both rat and human TH gene promoters contain among other elements a POU element and an E box (binding site of basic helix-loop-helix pro- teins), which may be important determinants of the tissue specificity of gene expression; they also contain a cAMP-responsive element and a TPA re-

immmmm | ~

TH Consensus TRE TGATTCA TGACTCA

E box CAGGTG CANNTG

POU ATGCAATT ATGCAAAT

ATGCTAAT CRE TGACGTCA TGACGTCA

Fig 2, Transcription factor binding sites of the rat TH promoter.

PCI2 + T P A ~

P C I 2

sponsive element (TRE). We are presently inves- tigating the regulatory role of the E box and the POU element. We have studied the involement of the TRE motif in the TH gene expression. We first tested the binding properties of the TRE-TH site by gel shift experiments. As shown in figure 3, nuclear extracts of PC12 cells that have been ex- posed to TPA yield an additional band that is not present with control cells. The binding activities of the TRE site were also investigated using nu- clear extracts of adrenals from reserpine treated and control rats. Here also the retarded TRE nucleoprotein complexes were significantly in- creased following reserpine induction. The re- sponse of the TH gene promoter to TPA has also been analysed in transactivation experiments, in which the TH promoter ( -754/-19) was linked to the coding sequence of the CAT gene. CAT ac- tivity was increased two-fold in TPA-treated cells (fig 4), which correlates with the increase in tran- scriptional activity of the TH gene caused by TPA (data not shown). Furthermore, of the various nu- clear proteins that bind to the TRE consensus site,

e ' ,

c -

A v

U

m

18

16

14

12

I(I

"I'PA

m

f i

c-jun +

¢ - fos

m

AM

AM + Rcs.

Fig 3. Mobility shift assays of nuclear extracts derived from TPA treated and control PC I2 cells, and adrenals from control and reserpine injected rats. 5 lag of nuclear proteins were in- cubated with 32p-labelled double-stranded TRE-TH oligonu- cleotide. The nucleoprotein complexes were resolved on a 4% non-denaturing polyacrylamide gel.

e e Fig 4. Transcriptional activation of the rat TH promoter by TPA and c-fos, c-jun expression vectors in PC12 cells. Cells were treated with TPA (100 nM) 48 h post-transfection for 1 h prior harvesting. 10 pg of total DNA were used in each assay. CAT activities were determined in three independent CaPo preci- pitates in four sets of experiments. Values are expressed _+ SEM.

156

c-Jun and c-Fos were found to activate CAT gene transcription driven by the TH promoter (fig 4). But, surprisingly, there was no potentiation of c- Jun by c-Fos as would have been expected for a typical TRE site (data not shown). This may be explained by the fact that the TRE-TH differs from the consensus site at the central base, which is a thymidine instead of the usual cytosine (TGATTCA v s TGACTCA) (fig 2). The TRE binding activity comprises c-Jun and is likely to involve c-Fos, since it was found to be expressed following reserpine induction (Icard-Liepkalns e t

a l , 1992). However, its expression was only transient suggesting that additional Fos-related proteins are likely to participate in the trans-syn- aptic regulation of TH in an orderly time sequence. Such a process would account for the maintenance of trans-synaptic induction for a pro- longed period.

Phenotypic regulation

Tracing TH anatomically and biochemically led to the observation that NT expression is regulated during development by cellular environment of the neuroblast. A well established model to study the mechanisms that control the choice of neu- rotransmitter phenotype is constituted by primary cultures of sympathetic neurons of superior cer- vical ganglion (SCG) (Hawrot and Patterson, 1979). In this system, the expression of NT can be modified by several external signals. Studies performed on cultured sympathetic neurons have led to the identification of several factors that me- diate the effects of the target and the non-neuronal cells on transmitter expression, and so regulate phenotypic expression. The adrenergic-choliner- gic plasticity was determined at the TH and ChAT mRNA levels by Northern blots as shown in figure 5. TH mRNA were detected in all three culture conditions, although its level decreased in the presence of muscle-conditioned medium while it increased in the high K + culture condition (fig 5, lanes 2 and 3 respectively). In contrast, ChAT mRNA was only detectable in neurons grown in the presence of muscle-conditioned me- dium (fig 5, lane 2). While muscle-conditioned medium depresses the expression of TH, it stimu- lates that of CHAT.

The modulation of the neurotransmitter pheno- type of sympathetic neurons is therefore accom- panied by variations of the levels of TH and ChAT mRNA, in opposite ways. Sympathetic neurons of

1 2 3

~!!!!i

--,--CHAT

T I-i

Fig 5. Effect of cell culture conditions on TH and ChAT mRNA expression in rat SCG. Northern blot analysis was performed on 4 ~tg of total RNA extracted from SCG grown for 18 days, Hybridization was performed with 32p-labelled rChAT-1 and pTHI cDNA probes. Lane 1, control medium; lane 2, muscle- conditioned medium; lane 3, high K + medium (40 mM KC1).

SCG are able to modify their neurotransmitter phenotype according to environmental signals. Studies on the promoters of both TH and ChAT genes are in progress to investigate the molecular basis of adrenergic-cholinergic plasticity.

Conclusion

In this review we described several mechanisms by which neurotransmitter expression is in- fluenced by neuronal activity through short-term and long-term regulatory systems und neuronal environment. Short-term regulation involves phosphorylation of the TH enzyme by a number of kinases both in rat and human. The multiple human TH enzymes that can be generated through alternative splicing have different phosphoryla- tion sites for specific kinases (such as Cam-PKII) , which suggests an additional mode of short-term regulation. The molecular basis of the long-term increase of TH activity was approached using the

model of reserp ine drug admin is t ra t ion to the rat. This phenomenon , also ca l led t rans -synap t ic in- duct ion, involves the b ind ing o f r egu la to ry tran- scr ipt ion factors on the TRE site o f the TH gene. Final ly , the role of ce l lu la r env i ronment in the in- duct ion of NT pheno type was inves t iga ted demon- strat ing p las t ic i ty by sympa the t i c neurons in culture.

Acknowledgments

This work was supported by the Centre National de la Recherche Scientifique, the Institut National de la Santfi et de la Recherche M6dicale, Rh6ne-Poulenc Rorer, and the Minist~re de la Recherche et de la Technologie.

References

Campbell DG, Hardie DG, Vulliet PR (1986) Identifi- cation of four phosphorylation sites in the N-terminal region of tyrosine hydroxylase, J Biol Chem 261, 10489-10492

Faucon Biguet N, Buda M, Lamouroux A, Samolyk D, Mallet J (1986) Time course of the change of TH mRNA in rat brain and adrenal medullae after single injection of reserpine. EMBO J 5, 287-291

Grima B, Lamouroux A, Boni C, Julien JF, Javoy-Agid F, Mallet J (1987) A single human gene encoding multiple tyrosine hydroxylases with different pre- dicted functional characteristics. Nature 326, 707-- 711

Hawrot E, Patterson PH (1979) Long-term culture of dissociated sympathetic neurons. Methods Enzymol 53, 574-584

Haycock JW (1990) Phosphorylation of tyrosine hy- droxylase in situ at Serine 8, 19, 31 and 40. J Biol Chem 265, 11682-11691

157

Horellou P, Le Bourdellbs B, Clot-Humbert J, Guibert B, Leviel V, Mallet J (1988) Multiple human tyrosine hydroxylases enzymes, generated through alternative splicing in Xenopus oocytes. J Neurochem 51, 652- 655

Icard-Liepkalns C, Faucon Biguet N, Vyas S, Robert JJ, Sassone-Corsi P, Mallet J (1992) AP-1 complex and c-fos transcription are involved in TPA provoked and transsynaptic inductions of the tyrosine hydroxy- lase gene: insights into long-term regulatory mech- anisms. J Neurosci Res 32, 290-298

Joh TH, Geghman C, Reis D (1973) Immunochemical demonstration of increased accumulation of tyrosine hydroxylase protein in sympathetic ganglia and adrenal medulla elicited by reserpine. Proc Natl Acad Sci USA 70, 2767-2771

Labatut R, Buda M, Berod A (1988) Long term changes in rat brain tyrosine hydroxylase following reserpine treatment: a qualitative immunochemical analysis. J Neurochem 50, 1375-1380

Le Bourdellbs B, Horellou R Le Caer JR Denbfle R Latta M, Haavik J, Guibert B, Mayaux JF, Mallet J (1991) Phosphorylation of human recombinant ty- rosine hydroxylase isoforms 1 and 2: an additional phosphorylated residue in isoform 2, generated through alternative splicing. J Biol Chem 266, 17124-17130

Le Douarin N (1980) The ontogeny of the neural crest in avian embryo chimaeras. Nature 286, 663-669

Mitchell JR Hardie DG, Vulliet PR (1990) Site specific phosphorylation of tyrosine hydroxylase after KCI depolarization and nerve growth factor treatment of PCI2 celles. J Biol Chem 265, 22358-22364

Pearson RB, Woodgett JR, Cohen P, Kemp BE (1985) Substrate specificity of a multifunctional cal- modulin-dependent protein kinase. J Biol Chem 260, 14471 - 14476

Vulliet PR, Hall FL, Mitchell JP, Hardie DG (1989) Identification of a novel proline-directed Serine/Threonine protein kinase in rat pheochro- mocytoma. J Biol Chem 264, 16292-16298