Differential Expression of the B′β Regulatory Subunit of Protein Phosphatase 2A Modulates Tyrosine Hydroxylase Phosphorylation and Catecholamine Synthesis*

Tyrosine hydroxylase (TH), the rate-limiting enzyme in catecholamine synthesis, is stimulated by N-terminal phosphorylation by several kinases and inhibited by protein serine/threonine phosphatase 2A (PP2A). PP2A is a family of heterotrimeric holoenzymes containing one of more than a dozen different regulatory subunits. In comparison with rat forebrain extracts, adrenal gland extracts exhibited TH hyperphosphorylation at Ser19, Ser31, and Ser40, as well as reduced phosphatase activity selectively toward phosphorylated TH. Because the B′β regulatory subunit of PP2A is expressed in brain but not in adrenal glands, we tested the hypothesis that PP2A/B′β is a specific TH phosphatase. In catecholamine-secreting PC12 cells, inducible expression of B′β decreased both N-terminal Ser phosphorylation and in situ TH activity, whereas inducible silencing of endogenous B′β had the opposite effect. Furthermore, PP2A/B′β directly dephosphorylated TH in vitro. As to specificity, other PP2A regulatory subunits had negligible effects on TH activity and phosphorylation in situ and in vitro. Whereas B′β was highly expressed in dopaminergic cell bodies in the substantia nigra, the PP2A regulatory subunit was excluded from TH-positive terminal fields in the striatum and failed to colocalize with presynaptic markers in general. Consistent with a model in which B′β enrichment in neuronal cell bodies helps confine catecholamine synthesis to axon terminals, TH phosphorylation was higher in processes than in somata of dopaminergic neurons. In summary, we show that B′β recruits PP2A to modulate TH activity in a tissue- and cell compartment specific fashion.

Tyrosine hydroxylase (TH) 2 catalyzes the rate-limiting step in the biosynthesis of catecholamines (dopamine, norepinephrine, and epinephrine) from the amino acid precursor L-tyro-sine (1). The enzyme consists of an N-terminal regulatory domain, a central catalytic domain, and a C-terminal association domain, which mediates tetrameric assembly (2,3). Several kinases regulate TH activity by phosphorylating key serines in the regulatory domain (Ser 8 , Ser 19 , Ser 31 , and Ser 40 ). Best characterized is Ser 40 phosphorylation by cyclic AMP-dependent protein kinase (PKA), which markedly enhances TH catalytic activity both in vitro and in vivo by relieving feedback inhibition by the catecholamines (4 -9). Phosphorylation at Ser 19 by calcium/calmodulin-dependent kinase II (CaMKII) and at Ser 31 by proline-directed kinases has comparatively moderate effects on TH activity but can facilitate subsequent Ser 40 phosphorylation (10,11).
Although the kinases that regulate catecholamine synthesis have been studied extensively, little is know about inactivation of TH by protein Ser/Thr phosphatases. Previous studies used the phosphatase inhibitor okadaic acid to implicate PP2A as a major negative regulator of TH (12)(13)(14). However, okadaic acid also inhibits the related PP4, PP5, and PP6 catalytic subunits with high affinity (15,16), and no information is available regarding the subunit composition of the presumptive TH phosphatase(s).
As one of four major groups of serine/threonine phosphatases, PP2A exists predominantly as a heterotrimer of a 36-kDa catalytic or C subunit, a 65-kDa scaffolding or A subunit, and a variable regulatory subunit (17,18). Three families of regulatory subunits referred to as B, BЈ, and BЉ associate with the core dimer of A and C subunits. The B family (or PR55) of regulatory subunits consists of four genes (␣-␦). The BЈ family (or B56/ PR61) is encoded by five genes (␣-⑀), whereas the third BЉ family consists of three genes (PR72/130, PR59, and PR48). Alternative splicing of several genes further diversifies the variable subunit repertoire. Regulatory subunits determine subcellular localization and substrate specificity and enable PP2A holoenzymes to respond to specific second messengers (19 -27).
The identification of the phosphatase holoenzyme that gates catecholamine synthesis may offer insights into the etiology of neurological disorders such as Parkinson and Huntington diseases and could suggest novel interventions (28,29). Here, we show that the neuron-enriched PP2A/BЈ␤ heterotrimer dephosphorylates TH and inhibits its activity. We also provide evidence that tissue-and cell compartment-specific expression of BЈ␤ modulates TH phosphorylation locally, thereby defining sites of elevated catecholamine synthesis.

MATERIALS AND METHODS
BЈ␤ Antibody Generation-To generate antisera specific to BЈ␤, the synthetic peptide CPLQRLTPQVAASGGQS from the extreme C terminus of BЈ␤ was coupled to maleimide-activated keyhole limpet hemocyanin (Pierce). Antiserum was raised by Covance Research Products (Princeton, NJ) and was affinity purified by coupling the same peptide to Affi-Gel 10 (Bio-Rad). Specificity was confirmed in heterologous expression experiments.
Tissue and Cell Extract Preparation, Purification, and Immunoblotting-Six weeks or older male Sprague-Dawley rats were euthanized by swift decapitation according to protocols approved by the Institutional Animal Care and Use Committee at the University of Iowa. Tissues dissected within 5 min of euthanasia were flash frozen on dry ice and stored at Ϫ70°C until use. For total extract preparation, frozen tissues were pulverized and homogenized in ϳ10 volumes of SDS sample buffer supplemented with 0.5 M microcystin-LR and 2 mM EDTA to inhibit phosphatases. Homogenates were sonicated to shear DNA and cleared by centrifugation (15 min at 20,000 ϫ g). For microcystin-agarose affinity purification and in vitro phosphatase assays, the tissues were homogenized in 3-4 volumes of buffer containing 1% Triton X-100, and 150 mM NaCl, 20 mM Tris, pH 7.5, 1 mM EDTA, 1 mM EGTA, 1 mM phenylmethylsulfonyl fluoride, 1 mM benzamidine, and 1 g/ml leupeptin. Insoluble debris was removed by centrifugation. Affinity purification of PP1 and PP2A holoenzymes on microcystin-agarose was carried out as described (36). For immunoblot analysis, PC6-3 cells were harvested by adding SDS sample buffer plus 0.5 M microcystin-LR, 2 mM EDTA directly to the culture plate. Homogenates were sonicated and cleared as above. Protein concentrations were determined by a dot blot assay, and quantitative immunoblotting was performed as described (35), normalizing phospho-TH to total TH signals. Control experiments showed that signal intensities scaled linearly with the amount of lysate loaded.
TH Activity Assay-An in situ dopamine biosynthesis assay, which monitors 14 CO 2 production following the conversion of L-[1-14 C] tyrosine to dopamine, was carried out as described (37) with modifications. Briefly, PC6-3 cells growing in 24-well plates (4 -6 wells/condition) were treated with doxycycline for 3 days to induce either overexpression or RNAi-mediated silencing of PP2A regulatory subunits. After replacing the medium with Hanks' balanced salt solution containing 1 Ci/ml of L-[1-14 C]tyrosine, the wells were immediately sealed with a small chimney capped with a rubber stopper and fitted with a plastic cup containing 300 l of 1 M NaOH to absorb the 14 CO 2 released by the cells. After 1 h, NaOH was subjected to liquid scintillation counting, subtracting blank values of wells without cells.
Total and phospho-Ser 19 TH immunofluorescence signals were digitally quantified after outlining cell bodies and processes using ImageJ software.

Correlation of TH Dephosphorylation and PP2A/BЈ␤ Expression in Tissues-
We examined expression and phosphorylation of TH in a panel of rat tissues by immunoblotting with total and phospho-specific TH antibodies. As expected, TH was detected only in forebrain and adrenal gland. Unexpectedly, Ser 40 -phosphorylated TH was strikingly enriched in adrenal gland compared with forebrain ( Fig.  1A, top panel). In addition to the PKA site, TH phosphorylation of the CaMKII site Ser 19 and the ERK/ Cdk5 site Ser 31 was also substantially higher in adrenal gland (Fig.  1B). The simplest explanation for this result is that adrenal gland has less TH-directed phosphatase activity than brain tissue.
Speculating that a brain-specific PP2A regulatory subunit could be responsible for the comparative hypophosphorylation of TH in this tissue, we explored the expression pattern of different PP2A subunits by immunoblotting tissue extract after affinity purification with microcystin-agarose. Microcystin binds with high affinity to catalytic subunits of PP1 and PP2A and microcystin affinity resins have been used to document the specificity of PP2A regulatory subunit antibodies (36,39). In agreement with previous results (36,39), the core scaffolding and catalytic (A and C) subunits and B␣/␦ regulatory subunits (detected with a pan B antibody) were ubiquitously expressed (Fig. 1A, bottom panel). In contrast, one of the five BЈ subunit family members, BЈ␤ was almost exclusively localized to forebrain and cerebellum. The enrichment of BЈ␤ protein in the brain is in accordance with earlier mRNA expression studies (40,41).
To test the hypothesis that increased TH phosphorylation in adrenal gland is due to the absence of a specific TH phosphatase such as PP2A/BЈ␤, we compared phosphatase activities in forebrain and adrenal gland extracts, using either a PKA-phosphorylated GST-TH fusion protein (regulatory domain residues 31-164 (32)) or myelin basic protein (MBP) as substrates. Forebrain extracts were able to dephosphorylate 32 P-labeled TH much more readily than adrenal gland extracts, with an activity ratio of 3.2 Ϯ 0.7 (n ϭ 4, p Ͻ 0.01) in extract dilutions matched for PP2A catalytic subunit content (Fig. 1C). The higher phosphatase activity in brain was specific for the [ 32 P]TH substrate, because forebrain and adrenal gland extracts dephosphorylated [ 32 P]MBP to the same extent. In combination, these data suggest that tissue-specific expression of PP2A/BЈ␤ inhibits TH phosphorylation.
PP2A/BЈ␤ Mediates TH Dephosphorylation in Cells-To directly examine the modulation of TH by different PP2A heterotrimeric complexes, we turned to PC12 pheochromocytoma cells as a model system, specifically the PC6-3 subline (42). Derived from rat adrenal chromaffin cells, PC12 and PC6-3 cells abundantly express TH and synthesize and secrete catecholamines (43,44). Treatment of PC6-3 cells with the PP2A inhibitor okadaic acid promoted dose-dependent TH hyperphosphorylation at Ser 40 ( Fig. 2A). Shown for a single concen-FIGURE 1. PP2A/B␤ tissue expression correlates with TH dephosphorylation. A, total lysates (top, 12 g of adrenal gland (gl.) protein, 20 g of protein from other tissues) and microcystin-agarose precipitates (bottom, from ϳ100 g of protein) from the indicated rat tissues were immunoblotted for Ser 40 -phosphorylated TH (pTH(S40)), total TH, and PP2A subunits. The negative (neg.) control is a microcystin-agarose precipitate from a forebrain lysate containing 10 M free microcystin-LR. B, forebrain and adrenal gland extracts were probed for Ser 19 , Ser 31 , and Ser 40 phosphorylated and total TH. C, forebrain and adrenal gland homogenates (inset, immunoblot of PP2A/C subunit) were used to dephosphorylate GST-TH  and MBP that had been 32 P-labeled with PKA. 32 PO 4 2Ϫ release was quantified by liquid scintillation counting and is expressed as the means Ϯ S.D. of triplicate determinations. The data are representative of three or more independent experiments. *, p Ͻ 0.01 by Student's t test. tration, 250 nM, which inhibits PP2A-like phosphatases, but not PP1, PP2B, or PP2C (15), okadaic acid increased TH phosphorylation at all three physiologically relevant sites (Ser 19 , Ser 31 , and Ser 40 ; Fig. 2B), confirming PP2A or a related enzyme as the predominant TH phosphatase (7,(12)(13)(14).
To study the regulation of TH by PP2A in a homogenous cell population, we generated stable PC6-3 cell lines expressing FLAG epitope-tagged regulatory subunits under the control of a tetracycline/doxycycline-inducible cytomegalovirus promoter. Overexpressed regulatory subunits appear to associate with a pool of free PP2A core dimer and do not cause compensatory changes in the levels of other PP2A subunits (19,34). In particular, BЈ␤ and B␥, a neuron-specific member of the B family of PP2A regulatory subunits, immunoprecipitated with equivalent amounts of the catalytic subunit (see Fig. 5A), indicating efficient incorporation into the PP2A holoenzyme. Inducible expression of the BЈ␤ regulatory subunit significantly reduced TH phosphorylation at Ser 40 , Ser 31 , and Ser 19 (Fig. 3, A  and C), whereas phosphorylation of ERK1/2 was unaffected. In contrast, induction of B␥ did not alter TH phosphorylation at any site but increased ERK1/2 phosphorylation as previously reported (34). Neither total TH nor total ERK levels were altered by inducible PP2A regulatory subunit expression.
To complement these forced expression experiments, clonal PC6-3 cell lines were generated that inducibly express BЈ␤-directed short hairpin RNA to silence the endogenous protein (35,45). Doxycycline treatment for 3-4 days decreased BЈ␤ subunit levels to 56 Ϯ 1% (n ϭ 4) but had no effect on TH, ERK1/2, and PP2A/B pan levels ( Fig. 3B and data not shown). Silencing of endogenous BЈ␤ was associated with a significant increase in TH phosphorylation at Ser 40 and Ser 19 (Fig. 3, B and C). Remarkably, even though BЈ␤ overexpression decreased Ser 31 phosphorylation, inducible BЈ␤ knock-down had no effect on phosphorylation at that site (Fig. 3, A, and C). It is possible that PP2A/BЈ␤ dephosphorylates Ser 31 with higher efficiency, so that the ϳ50% of BЈ␤ remaining after doxycycline treatment (Fig.  3B) is enough to modulate Ser 31 but not Ser 19 and Ser 40 phosphorylation. In aggregate, these experiments establish PP2A/BЈ␤ as the major phosphatase that mediates TH dephosphorylation in PC6-3 cells.
PP2A/BЈ␤ Modulates TH Activity-TH phosphorylation by PKA at Ser 40 stimulates TH activity and catecholamine synthesis (46). To demonstrate that PP2A/BЈ␤ can also modulate TH activity, we performed TH activity assays, measuring the release of 14 CO 2 from cells incubated with L-[1-14 C]tyrosine (Fig. 4A). The amount of CO 2 produced during the conversion of L-3,4-dihydroxyphenylalanine (DOPA) to dopamine by DOPA decarboxylase reflects TH activity, because TH is the rate-limiting enzyme in this biosynthetic cascade (37). The contribution of other metabolic pathways involving tyrosine was negligible during the 1-h assay, because preincubation of PC6-3 cells with 100 M of the DOPA decarboxylase inhibitor carbidopa completely blocked 14 CO 2 production (data not shown).
Inducible overexpression of BЈ␤ but not B␥ resulted in a robust (Ͼ50%) inhibition of 14 CO 2 release, closely paralleling the decrease in TH phosphorylation at Ser 40 (compare Figs. 4B and 3C). Conversely, inducible RNAi of BЈ␤ significantly increased 14 CO 2 release (Ͼ60%; compare Figs. 4C and 3C). Global inhibition of PP2A-like phosphatases with okadaic acid (300 nM) increased TH activity to levels not statistically different from cells in which BЈ␤ was knocked down (Fig. 4C). These results indicate that PP2A/BЈ␤ opposes kinase-mediated stimulation of catecholamine synthesis. PP2A/BЈ␤ Directly Dephosphorylates TH-The PP2A family of Ser/Thr phosphatases are critical regulators of many kinase signal transduction cascades (17,47). It is therefore conceivable that PP2A could indirectly mediate TH dephosphorylation, by e.g. inactivating a kinase (48). To investigate whether BЈ␤ can dephosphorylate TH directly, in vitro phosphatase assays were performed. PP2A holoenzymes containing FLAG-tagged BЈ␤ and B␥ subunits were immunoprecipitated from overexpressing cell lines, and aliquots containing equal PP2A catalytic subunit amounts were incubated with GST-TH 31-164 and MBP substrates that had been 32 P-phosphorylated with PKA in vitro. PP2A holoenzyme containing BЈ␤ dephosphorylated [ 32 P]TH more avidly than B␥. The increased activity was specific for the TH substrate, because [ 32 P]MBP was dephosphorylated to the same extent by PP2A/BЈ␤ and B␥ (Fig. 5A). We calculated TH/MBP dephosphorylation ratios as a measure of specificity. The PP2A/BЈ␤ holoenzyme showed more than 2-fold greater specificity for TH than PP2A/B␥ (Fig. 5B). Thus, not only is PP2A/BЈ␤ capable of dephosphorylating TH directly, but it actually maintains substrate specificity in vitro.

PP2A/BЈ␤ and TH Colocalize in
Catecholaminergic Neurons-In the brain, TH is found in dopaminergic neurons in the substantia nigra and in noradrenergic neurons in the locus coeruleus and other brain stem nuclei. Dopaminergic cell bodies present in the substantia nigra project to corpus striatum, and dopamine released in striatum plays an important role in movement control. Loss of these dopaminergic neurons is the cause of Parkinson disease (49). Immunofluorescence labeling of rat brain sections with a polyclonal antibody raised against a peptide antigen derived from the unique C terminus of BЈ␤ demonstrated that the protein is widely expressed in neurons throughout all brain regions but virtually absent from glia (not shown). Double labeling with antibodies to BЈ␤ and TH showed colocalization in cell bodies of rat substantia nigra pars compacta and in various hindbrain regions ( Fig. 6A and data not shown). In fact, virtually every TH-positive cell body also contained BЈ␤. Interestingly, upon examining BЈ␤ expression in the striatum, we found the PP2A subunit to be present in striatal cell bodies but absent from TH-positive terminals arising from the substantia nigra (Fig. 6, B and C). To confirm this finding, brain sections were double-stained for synaptotagmin (a marker for presynaptic terminals) and BЈ␤. No colocalization was found between synaptotagmin and BЈ␤ in striatum or any other brain region (Fig. 6D). We substantiated the exclusively somatodendritic distribution of BЈ␤ by colabeling dissociated hippocampal cultures for the presynaptic protein bassoon. BЈ␤ was found in the soma and dendritic shafts and spines apposed to presynaptic terminals but did not overlap with bassoon, except in some intracellular patches in the cell body (Fig. 6E). These results indicate a somatodendritic site of action for PP2A/BЈ␤, perhaps involving suppression of catecholamine synthesis far away from presynaptic release sites.
To test the hypothesis that BЈ␤ controls TH phosphorylation locally, we costained brain sections with total and phospho-specific TH antibodies. Consistent with higher phosphorylation stoichiometry of brain TH at Ser 19 than at Ser 31 and Ser 40 (50), the phospho-Ser 19 antibody resulted in the most robust and specific staining (as judged by overlap with total TH) of the three phosphorylation site-specific TH antibodies. Images containing both dopaminergic cell bodies and their processes were captured at the edge of the substantia nigra pars compacta. Compared with total TH, phospho-Ser 19 TH immunoreactivity was evidently enriched in processes (both axons and dendrites; Fig. 6F). Quantification of process and soma staining intensities indicated that Ser 19 TH phosphorylation is ϳ2-fold elevated in processes compared with cell bodies of substantia nigra neurons (Fig. 6G). These data are consistent with a model in which TH phosphorylation and catecholamine synthesis are spatially restricted by the PP2A/BЈ␤ holoenzyme (Fig. 6H).  Tyrosine Hydroxylase Regulation by Neuronal PP2A/B␤ JANUARY 5, 2007 • VOLUME 282 • NUMBER 1 JOURNAL OF BIOLOGICAL CHEMISTRY 577 DISCUSSION TH controls the synthesis rate of dopamine, norepinephrine, and epinephrine. Consequently, TH activity is subject to multiple levels of regulation, involving both long term (transcription, RNA stability, translation, and protein stability) and short term mechanisms (end product inhibition, allosteric regulation, and phosphorylation). In particular, PKA-mediated phosphorylation at Ser 40 can stimulate TH activity up to 20-fold (4), allowing for rapid adaptation of catecholamine synthesis to changes in release rates. Ser 19 and Ser 31 phosphorylation are thought to enhance TH activity primarily by facilitating Ser 40 phosphorylation (10,11).
Previous studies on TH dephosphorylation relied on divalent cation requirements and inhibitors to distinguish between activities of the four major Ser/Thr phosphatase families (PP1, PP2A, PP2B/calcineurin, and PP2C) in crude or partially purified extracts (7,(12)(13)(14). The conclusions from these studies, pointing to PP2A as the major TH phosphatase, were qualified by the subsequent discovery of PP4, PP5, and PP6, which have inhibition profiles similar to the PP2A catalytic subunit (15,16). With this study, we provide unequivocal gain-and loss-of-function evidence that a specific PP2A heterotrimer modulates TH phosphorylation and activity in vitro and in intact cells. In PC12 cells, RNAi-mediated silencing of BЈ␤ and pharmacological inhibition of all PP2A-like enzymes raise TH activity to similar levels (Fig. 4C), suggesting that PP2A/BЈ␤ is the predominant phosphatase that inactivates TH.
What initially cued us in to the existence of a brain-enriched TH phosphatase was the pronounced TH hyperphosphorylation and reduced TH phosphatase activity in adrenal gland extracts (Fig. 1). The high levels of Ser 19 , Ser 31 , and Ser 40 phosphorylation of TH in these rapidly prepared adrenal gland extracts contrast with low basal TH phosphorylation stoichiometries reported for isolated bovine adrenal chromaffin cells (10,11), of which Ser 31 phosphorylation was in fact lower than in brain TH (50). The most reasonable explanation is that tissue integrity and sympathetic innervation of the adrenal gland are critical for physiological levels of TH phosphorylation (51).
Splanchnic nerve activity in response to a variety of stressors triggers massive release of epinephrine and norepinephrine from the adrenal medulla into in the blood stream. By lowering the affinity of TH for catecholamines and thus relieving feedback inhibition, high basal Ser 40 phosphorylation of TH may permit intracellular catecholamines to accumulate to levels sufficient for the "fight-or-flight" response (52). As a therapeutic implication of our findings, future interventions into cardiovascular and central nervous system damage resulting from pheochromocytomas or chronic stress may involve up-regulation of endogenous BЈ␤ or targeted delivery of BЈ␤ to the adrenal medulla.
The neuromodulatory role of dopamine and norephinephrine in the central nervous system in conjunction with the morphological complexity of neurons apparently requires an additional level of control in the form of a TH-directed phosphatase. We found that PP2A/BЈ␤ was colocalized with TH in cells bodies of most if not all dopaminergic and noradrenergic neurons but was undetectable in axons and presynaptic terminals (Fig.  6). As a probable consequence of the restricted expression of BЈ␤, we detected significantly elevated TH phosphorylation in processes compared with cell bodies in the substantia nigra (Fig. 6, F and G). These results are consistent with a report showing that the dopamine D2 receptor agonist haloperidol elicits a greater increase in Ser 40 TH phosphorylation in dopaminergic terminals in the striatum than in their cell bodies of origin in the substantia nigra (50).
What is the functional significance of the somatodendritic localization of PP2A/BЈ␤ in regards to TH regulation? Curiously, none of the enzymes in the catecholamine synthesis pathway are enriched in synaptic terminals, where the majority of dopamine and norepinephrine release occurs (53,54). Catecholamine synthesis in the absence of release would not only place an unnecessary metabolic demand on the cell but could actually induce neurotoxicity because of build-up of monoamine oxidase metabolites (55). The axon and terminal-exclusive distribution of PP2A/BЈ␤ may thus provide a mechanism to curb wasteful and potentially harmful "ectopic" production of catecholamines (Fig. 6H).
Despite high homologies within each gene family, the three families of PP2A regulatory subunits, B, BЈ, and BЉ, display little sequence and predicted structural similarity (but see Ref. 56). We show here that BЈ␤, a protein with predicted predominantly ␣-helical secondary structure, imparts TH specificity to the PP2A heterotrimer, as compared with B␥, a predicted ␤-propeller (57). The fact that this substrate specificity is maintained in vitro implies that either regulatory subunits alter the conformation of the catalytic subunit or, more likely, participate in substrate docking (58). The inferred interaction between TH and BЈ␤ may have the fast off-rate necessary for high catalytic turnover, because we did not detect a stable complex between the proteins by standard immunoprecipitation methods.
BЈ␤ is one of the most abundant PP2A regulatory subunits in the rodent brain (24). Our immunolocalization results reveal the protein to be widely expressed in most neuron populations, consistent with the notion that BЈ␤ targets PP2A to a variety of neuronal substrates. A role for PP2A/ BЈ␤ in synaptic plasticity that is consistent with its postsynaptic localization (Fig. 6E) was suggested by Fukunaga et al. (24). BЈ␤ (referred to as BЈ␣ in the cited work following older nomenclature) was found to be phosphorylated by CaMKII FIGURE 6. B␤ localization in the soma of dopaminergic neurons is associated with local TH dephosphorylation. A, immunofluorescence localization of TH (red) and BЈ␤ (green) in rat substantia nigra sections (pc, pars compacta; pr, pars reticulata). Yellow in the merged image indicates colocalization; nuclear DNA is stained blue. B-D, BЈ␤ colocalized neither with TH-positive terminals (B and C) nor with the presynaptic protein synaptotagmin (syntg, D) in the striatum (S, soma). E, dissociated hippocampal neuron (17 days in culture) showing BЈ␤ immunolabeling in dendrites and dendritic spines (open arrowheads) adjacent to but not overlapping with the presynaptic marker bassoon (bass, filled arrowheads, magnified in inset). F, immunofluorescence localization of TH (red) and phospho-Ser 19 TH (pTH(S19), green) in rat substantia nigra sections. Arrowheads point to presumptive axons with enriched phospho-TH. G, quantification of total TH and phospho-Ser 19  during long term potentiation in the hippocampus. Phosphorylated PP2A/BЈ␤ displayed reduced activity toward autophosphorylated, autonomously active CaMKII, suggesting a positive feedback mechanism important for learning and memory. How BЈ␤ phosphorylation by CaMKII and other kinases affects catecholamine synthesis is an interesting question that deserves further study.