Expression of A Kinase Anchor Protein 121 Is Regulated by Hormones in Thyroid and Testicular Germ Cells*

Distinct A KinaseAnchor Proteins (AKAPs) immobilize and concentrate protein kinase A II (PKAII) isoforms at specific intracellular locations. AKAP121 binds and targets PKAIIα to the cytoplasmic surface of mitochondria. Mechanisms that control expression of this mitochondrial AKAP are unknown. We have cloned cDNA for rat AKAP121 and show that AKAP121 protein expression is regulated by thyroid stimulating hormone (TSH) and cAMP. Differentiated thyroid cells (TL5) accumulate AKAP121 upon incubation with TSH or a cAMP analog. Levels of total and newly synthesized AKAP121 mRNA also increased after treatment. AKAP121 mRNA accumulated in the presence of cycloheximide, suggesting that transcription of the anchor protein gene is directly controlled by cAMP and PKA. AKAP121 is induced with similar kinetics when an unrelated, spermatocyte-derived cell line (GC-2) is incubated with 8-chlorophenylthio-cAMP. Thus, AKAP121 concentration may be controlled by hormones that activate adenylate cyclase. This mode of regulation could provide a general mechanism for (a) enhancing the sensitivity of distal organelles to cAMP and (b) shifting the focus of cAMP-mediated signaling from cytoplasm to organelles.

Signaling mediated by cAMP and activation of protein kinase A (PKA) 1 plays an essential role in the regulation of many important cellular activities, including motility, metabolism, differentiation, synaptic transmission, ion channel activities, growth, and coordination of gene transcription (1)(2)(3). The distinctive characteristics of the PKA holoenzymes are largely determined by the structure and properties of their R subunits. Unlike PKAI, which is typically cytosolic, PKAII (␣ and ␤) are often targeted to certain subcellular locations by specific anchor proteins (AKAPs) (4 -20). Localized PKAII isoforms are thought to perform specialized aspects of cAMP-activated signal transduction (4,5,21). The tethering of PKAII holoenzymes in specific intracellular microenvironments also influences accessibility to upstream and downstream effector molecules of cAMP action (22)(23)(24). Intracellular targeting of PKAII isoforms has been linked to cAMP-dependent gene transcription in differentiated and non-differentiated cells. In thyroid cells, displacement of immobilized PKAII from perinuclear sites to the cytoplasm impaired cAMP-regulated transcription of the thyroglobulin gene (25). Overexpression of AKAP75 and C subunit in variant PC12 cells defective in cAMP signaling restored cAMP-activated gene transcription, whereas overexpression of either protein alone was ineffective (26). We recently demonstrated that the assembly of AKAP75-PKAII complexes in the cortical cytoskeleton of HEK293 cells enhanced the propagation of cAMP signals to the nucleus, as shown by increased CREB/CRE-controlled gene transcription (27). These data suggest the possibility that cells could control their sensitivity to hormones and/or developmental factors that activate adenylate cyclase by regulating the content and distribution of specific AKAPs. S-AKAP84 and its splice variant AKAP121, anchor PKAII␣ to the cytoplasmic surface of mitochondria. The anchoring of PKAII␣ may be involved in the translocation of mitochondria to the site of cytoskeleton assembly in male germ cells (28,29).
Enrichment and localization of C subunits of PKAII␣ in proximity of target molecules on mitochondria/cytoskeleton may regulate the phosphorylation of regulator/effector proteins involved in the dynamic reorganization of the flagellar cytoskeleton. This ultimately may modulate the motility and/or the fertilization capacity of spermatozoa (28,29). Disruption of the targeting of PKAII in mammalian sperm may result in a reduction in sperm motility (30). Accumulation of S-AKAP84 and its cognate mRNA are developmentally regulated during sperm development. The anchor protein is expressed de novo during late spermiogenesis, and this is coincident with the maximal expression and subsequent anchoring of RII␣ and RII␤ subunits (28).
An important, but still open, question is how hormones and growth or developmental factors promote or regulate expression of AKAPs. Are cAMP, PKAs, and PKA-dependent protein phosphorylation involved in the regulation of the location and * This work was supported by National Institutes of Health Grant DK50702 (to M. E. G.), by Associazione Italiana Ricerca sul Cancro (AIRC), Progetti Finalizzati CNR "Biotecnologie" (to E. V. A.), and by National Institutes of Health Grant GM 22792 (to C. S. R.). 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 1 The abbreviations used are: PKA, protein kinase A (cAMP-dependent protein kinase); R, regulatory subunit of cAMP-dependent protein kinase; C, catalytic subunit of cAMP-dependent protein kinase; AKAP, protein kinase A anchor protein; TSH, thyroid stimulating hormone; TPA, 12-O-tetradecanoylphorbol-13-acetate; CPT-cAMP, 8-chlorophenylthio-cAMP. amounts of AKAP⅐PKAII complexes? In the present study, we describe the cloning and hormonal regulation of rat AKAP121. The expression and accumulation of AKAP121 is induced by a cAMP analog in rat thyroid cells (TL5) and mouse GC2 germ cells. TSH, which stimulates adenyl cyclase, also induces AKAP121 in thyroid cells. The cAMP-stimulated induction of AKAP121 mRNA is resistant to cycloheximide. These results suggest a mechanism whereby hormones promote signaling aimed at specific intracellular target sites.
GC2 (GC2 spd (ts)) cells were generously provided by J. L. Millan and M. C. Hofmann, La Jolla Cancer Research Foundation, La Jolla, CA. GC2 cells were derived from primary mouse preleptotene spermatocytes by stable co-transfection with transgenes encoding SV40 large T antigen and a temperature-sensitive variant of the p53 transcriptional regulator protein (39). Immortalized GC2 cells exhibit morphological and biochemical properties characteristic of developing spermatids when they are grown at temperatures that permit low (37°C) or high (32°C) level expression of p53. Growth at 37°C enables continuous proliferation of GC2 cells, whereas incubation at 32°C results in withdrawal from the cell cycle and cell death after ϳ10 passages (39). For studies described in this paper, GC2 cells were grown at 37°C in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum.
Screening of cDNA Libraries-A rat thyroid expression cDNA library was prepared in the bacteriophage Hybri ZAP (Stratagene) using poly(A) RNA from cultured thyroid PC-C13 cells as a template. The library was screened using 32 P-labeled RII␤ as a probe by the procedure of Bregman et al. (6) to identify ␤-galactosidase fusion proteins that bind RII. Two recombinant phage clones containing overlapping cDNA sequences were isolated from 7 ϫ 10 5 plaques. The extreme 5Ј end of rat AKAP121 was isolated by employing a Taq DNA polymerase chain reaction as described under "Results." Full-length AKAP121 cDNA was sequenced by the dideoxynucleotide chain termination procedure of Sanger et al. (37).
Northern Gel Analysis-Cells were grown to semiconfluence, deprived of serum and hormones for three days, and then treated for the FIG. 1. Identification of an RII binding protein in rat thyroid (thy) and TL5 cells. An RII-overlay binding analysis on 120 g of "particulate" and "cytosolic" fractions of TL5 thyroid cells is shown in panel A. In the left panel, the filter was incubated with 32 P-labeled RII␤ probe. In the right panel, the probe was added to the hybridization buffer in presence of large excess of Ht-31 peptide (20 M). An autoradiogram is presented. In panel B, the predicted RII binding domain of rat AKAP121 is aligned with the binding domain of mouse S-AKAP84 and AKAP121. Positions of variable residues are shown in bold.

FIG. 2. Induction and anchoring of AKAP121 in thyroid cells (TL5).
TL5 cells were maintained in the absence of serum and hormones for 3 days (0), treated with a mixture containing TSH (1 milliunits/ml), insulin (0.5 g/ml), and 5% calf serum, and then harvested at the indicated times (h). Representative sets of autoradiograms are shown. In panel A, a 32 P-RII␤ overlay binding assay was performed on 120 g of either "particulate" (P) or "soluble" (S) proteins. In panel B, immunoblot analysis was performed on the same samples using antibodies directed against AKAP121 (see "Experimental Procedures"). Filters were exposed for 35 s. An AKAP121 signal in control cells (0) is visible with longer exposure times. Only relevant portions of the immunoblots are shown.
indicated times with medium alone or with medium containing the indicated drugs. Northern gel analysis was performed as described (27). Relative amounts of AKAP121 mRNA were determined by scanning densitometry (Molecular Dynamics).
Nuclear Run-on Assay-Elongation rates of nascent transcripts were assayed as described by Quinn et al. (38). Briefly, 2 ϫ 10 8 TL5 cells were deprived of serum and hormones for three days in F-12 medium containing 0.1% bovine serum albumin and then treated with 1 mM 8-chlorophenylthio-cAMP (CPT-cAMP) for 2 h. Cells were harvested in phosphate-buffered saline and lysed at 4°C for 5 min in lysis buffer (10 mM NaCl, 3 mM MgCl 2 , 0.5% Nonidet P-40, 10 mM Tris-Cl, pH 7.4). Nuclei were pelleted (500 ϫ g, 5 min) and washed once in cold reaction buffer (50 mM Tris-HCl, pH 8.3, 5 mM MgCl 2 , 300 mM KCl, 40% glycerol, and 0.5 mM each UTP, GTP, and CTP). The nuclear pellet was resuspended in reaction buffer (cold reaction buffer plus 150 Ci [␣-32 P]ATP, 3000 Ci/mmol) and incubated for 30 min at 30°C. Reactions were stopped by adding 40 units of RNase-free DNase and incubation at room temperature for 30 min. The reaction mixture was diluted 5-fold with TE (10 mM Tris, 0.5 mM EDTA) and incubated with proteinase K (25 g/ml) at 37°C for 30 min. RNA was extracted with phenol:CHCl 3 :isoamyl alcohol (25:24:1, v/v) and precipitated with isopropanol. The pellet was washed with 70% ethanol, air dried, and resuspended in TE. Plasmids containing or lacking AKAP121 cDNA were applied on N-Hybond membranes using a Life Technologies, Inc. slot blot manifold. Each plasmid (20 g of DNA) was resuspended in 6ϫ SSC and applied to presoaked filters (6ϫ SSC) by vacuum. DNAs were UV-cross-linked to the filters (Stratalinker TM , Stratagene). Filters were prehybridized for 4 h and hybridized with 32 P-labeled transcripts for 3 days at 42°C in 40% formamide, 5ϫ SSC, 2ϫ Denhardt's solution, 0.1% SDS, 100 g/ml salmon sperm DNA, 25 g/ml yeast tRNA. Next, filters were washed once for 15 min at room temperature in 2ϫ SSC, 0.5% SDS, and twice for 15 min at 55°C in 1ϫ SSC, 0.1% SDS. Filters were air dried and analyzed by autoradiography.

RESULTS
Identification and Cloning of Rat AKAP121 cDNA-In rat TL5 thyroid cells, targeting of PKA regulates accumulation of the catalytic subunit of PKA (C) in the nucleus, CREB phosphorylation, and cAMP-induced gene transcription (25). To identify thyroid AKAPs, we used 32 P-labeled RII␤ as a probe in an overlay binding assay (6). A 120-kDa RII-binding protein was identified in particulate fractions from rat thyroid and TL5 cells (Fig. 1A, left panel). Treatment with a peptide, Ht-31, which contains a domain that binds RII (24), competitively blocked RII␤ binding to the thyroid AKAP (Fig. 1A, right panel). We then screened a rat thyroid expression cDNA library with 32 P-labeled RII␤ (see "Experimental Procedures"). Several cDNAs that encode a homolog of a previously identified AKAP, mouse AKAP121 (24), were isolated. Two overlapping cDNA clones coding for 3Ј region of rat AKAP121 were further characterized by sequencing. By using oligonucleotide primers for the 5Ј end of mouse AKAP121 (29) and the 3Ј end of the rat cDNA, we employed a polymerase chain reaction on a thyroid cDNA template. The resulting full-length rat AKAP121 cDNA encodes a protein composed of 854 amino acids. Alignment of the sequence of rat AKAP121 with mouse AKAP121 and S-AKAP84, and human S-AKAP84 (28, 29) revealed extensive regions of homology. The RII tethering domains of rat (residues 303-322) and mouse (residues 306 -325) AKAP121 are 80% identical. Moreover, large aliphatic amino acids that govern the high affinity binding of RII (4, 5) appear at identical positions in both proteins. The N-terminal mitochondrial targeting do-FIG. 3. TSH and cAMP promote the rapid accumulation of AKAP121 in TL5 cells. TSH, CPT-cAMP, insulin, or TPA were added to the medium of serum-deprived TL5 cells at indicated concentrations. Cells were harvested at indicated times after stimulation. In panel A, total cell-proteins (120 g) (see "Experimental Procedures") were fractionated on a denaturing gel (8% polyacrylamide) and then immunoblotted with anti-mouse AKAP121 serum. The lower part of the gel was probed with anti-MAPK antibodies as a control. A representative set of autoradiograms is shown. In panel B, the relative abundance of AKAP121 was quantified by scanning densitometry. Levels of AKAP121 are presented as arbitrary densitometric units (ADU) and represent a mean Ϯ S.E. from three independent experiments that yielded similar results. In panel C, TL5 cells were deprived of TSH for three days and then stimulated with the indicated concentration of CPT-cAMP. AKAP121 expression was assayed by immunoblotting as described above. Panel D shows an immunoblot analysis of extracts (100 g protein) from control cells and cells treated with insulin or TPA. Cells were harvested at indicated times (h). The autoradiogram was overexposed (1 min) relative to panels A and C. main (residues 1-30) and KH domains (residues 565-613), previously identified in the mouse homolog, are also highly conserved in rat AKAP121 (residues 1-30 and residues 562-610), respectively (29). Coimmunoprecipitation experiments with antibodies directed against mouse AKAP121 and RIIbinding assays confirmed that the major thyroid RII-binding protein is AKAP121 (data not shown). Northern blot analysis detected expression of a 4.3-kbp AKAP121 mRNA in several tissues including: thyroid, brain, liver, and heart (data not shown).
TSH and cAMP Elicit the Accumulation of AKAP121 in TL5 Cells-Thyroid cell growth and differentiation are dependent upon TSH. Binding of ligand to the plasma membrane TSH receptor activates Gs-coupled adenylase cyclase and increases the intracellular levels of cAMP. Physiological effects of cAMP are mediated by activation of PKA (31)(32)(33)(34)(35). To determine whether AKAP121 is regulated by hormone, we performed RII-binding and Western blot analyses on "particulate" (P) and "cytosolic" (S) proteins extracted from TL5 cells. Cells were incubated in standard, serum-free medium containing 0.1% bovine serum albumin for three days and then were treated with a mixture of TSH, insulin, and serum. Serum-starved TL5 cells contain low levels of AKAP121 polypeptide and RII␤ binding activity (Fig. 2, A and B) (29). Incubation with serum, insulin, and TSH increased AKAP121 levels in a time-dependent fashion. Accumulation of the anchor protein was first evident 12 h post-stimulus, and the level increased over the next 24 h. AKAP121 was recovered in the particulate fraction. To

FIG. 4. Northern blot analyses of AKAP121 mRNA in TSH-or CPT-cAMP-treated TL5 cells.
A, kinetics of accumulation of AKAP121 mRNA in control and CPT-cAMP treated cells. 18 S RNA content was used as internal control. A representative set of autoradiograms is shown. B, AKAP121 mRNA content was quantified by scanning densitometry at the indicated times after treatment with TSH or cAMP analog. The data represent a mean Ϯ S.E. from three independent experiments and are expressed as arbitrary densitometric units (ADU). C, the cells were exposed to the indicated concentrations of agonist and harvested at 6 h (CPT-cAMP) or 12 h (TSH) post-stimulus. AKAP121 mRNA was assayed by a Northern blot. E, AKAP121 mRNA and 18 S RNA content were assayed in TL5 cells treated with 0.5 mM CPT-cAMP in the presence or absence of cycloheximide (20 g/ml).
further define the signaling pathway governing induction of AKAP121, TL5 cells were treated with TSH or with CPT-cAMP, a stable and potent cAMP analog. TSH or CPT-cAMP elicited substantial increases in AKAP121 content (Fig. 3, A  and B). CPT-cAMP induced AKAP121 12-15-fold, whereas a 5-6-fold increase in anchor protein ensued after treatment with TSH. AKAP121 accumulation was first detected 3 h poststimulus and reached a maximal value at 6 h with both CPT-cAMP and TSH. Induction of AKAP121 by CPT-cAMP was also observed in TL5 cells incubated for 3 days in medium containing 5% calf serum and 5H (no TSH; see "Experimental Procedures"). Thus, the presence of serum growth factors and hormones had little effect on CPT-cAMP-mediated accumulation of AKAP121 in TL5 cells (Fig. 3C). Treatment with insulin or the phorbol ester TPA, failed to elevate AKAP121 content in TL5 cells (Fig. 3D). Because TSH and a cAMP analog induce AKAP121 accumulation, it appears that PKA activation regulates the concentration of AKAP121.
TSH/cAMP Increases AKAP121 mRNA Levels and Gene Transcription in TL5 Cells-Quiescent TL5 cells contain a very low level of AKAP121 mRNA (Fig. 4A). TSH treatment increased AKAP121 mRNA abundance within 6 h; the AKAP121 mRNA concentration peaked 12 h post-stimulus, reaching a concentration of 6 -7-fold above the initial level. The level of the AKAP121 transcript then declined over the next 12 h. CPT-cAMP provoked a more rapid and robust response. A substantial increase in AKAP121 mRNA was detected 3 h after exposure of cells to CPT-cAMP. AKAP121 mRNA concentration peaked at 6 h (10 -15-fold over basal levels) and then declined slowly during the succeeding 18 h. Accumulation of AKAP121 mRNA was evident over a wide range of TSH or CPT-cAMP concentrations (Fig. 4, C and D). To determine whether accumulation of AKAP121 transcripts by cAMP requires new protein synthesis, we added CPT-cAMP in the presence of the protein translation inhibitor, cycloheximide. Treatment with 20 g/ml cycloheximide had no effect on accumulation of AKAP121 transcripts (Fig. 4E). Analysis of protein extracts prepared from TL5 cells labeled with [ 35 S]Met indicated that cycloheximide inhibited protein synthesis by 90% (data not shown). Thus, induction of AKAP121 mRNA is a direct, primary response to cAMP. To define further the mechanism of PKA-dependent induction of AKAP121 mRNA, we analyzed the effect of cAMP on the transcription of the AKAP121 gene. Run-on experiments were performed on nuclei isolated from TL5 cells (Fig. 5). Addition of 1 mM CPT-cAMP to the medium increased transcription of the AKAP121 gene severalfold over the basal level. The increase in the rate of AKAP121 gene transcription was consistent with the CPT-cAMP induced increase in AKAP121 mRNA abundance determined by Northern analysis. Thus, cAMP-dependent accumulation of mRNA and protein is regulated, at least in part, at the level of AKAP121 gene transcription.
Induction of AKAP121 by cAMP in Spermatocyte-derived GC2 Cells-Thyroid cells depend on TSH for both growth and differentiation. To investigate the generality of the effects of PKA/cAMP on AKAP121 expression and separate the induction process from growth and differentiation, we assayed spermatocytes (GC2 cells) that were immortalized by SV40 T antigen (39). Basal expression of AKAP121 in GC2 cells is higher than in TL5 cells (Fig. 6). Nevertheless, CPT-cAMP promoted a 3-4-fold increase in AKAP121 protein accumulation in GC2 (Fig. 6, A and B). AKAP121 levels increased 3 h post-stimulus and reached a maximal value 6 -9 h after exposure to CPT-cAMP. No further increase was detected at 24 h post-stimulus (data not shown). Thus, PKA activation increases the abundance of AKAP121 in distinct cell systems. This suggests a potential general mechanism whereby hormonal regulation of specific anchor proteins may modulate the sensitivity of cell compartments to the cAMP signal transduction pathway. DISCUSSION PKA isoenzymes are tethered and targeted to discrete intracellular microenvironments by AKAPs (4,5). Co-localization of PKA with adenylate cyclase may reduce effects of cAMP diffusion and augment PKA signaling when modest physiological levels of hormones encounter target cells (27). The proximity of released catalytic PKA subunits to certain target/effectors also influences the propagation of cAMP signals in cells and focuses signals on specific effector molecules (22)(23)(24)(25)(26)(27). A centrally important aspect of cAMP-PKA signaling is that spatially regulated accumulation of AKAPs, achieved by altering the intra-  6. cAMP induces accumulation of AKAP121 protein in mouse germ cells. Cells (5 ϫ 10 5 ) were plated and grown for 24 h in medium supplemented with 10% fetal calf serum. Cells were stimulated with CPT-cAMP for the indicated time periods. Total cell proteins (50 g) were immunoblotted with antibodies directed against mouse AKAP121 after denaturing electrophoresis. In panel A, representative autoradiograms are shown. Filters were exposed for 8 s. In panel B, the relative abundance of AKAP121 was quantified by scanning densitometry. Levels of AKAP121 are presented as arbitrary densitometric units (ADU) and represent a mean Ϯ S.E. from four independent experiments that yielded similar results. cellular distribution of PKAII isoenzymes, could influence cAMP signal dissemination to distal organelles. In granulosa cells, follicle-stimulating hormone induces a cytosolic translocation of PKAII coincident with an increase in the levels of an uncharacterized cytosolic RII-binding protein (36). In spermatids, S-AKAP84 accumulates at the outer membrane of mitochondria at a late phase of development, the beginning of nuclear condensation and tail elongation (28). At present, molecular mechanisms underlying developmentally or hormoneregulated accumulation of anchor proteins are poorly understood.
In differentiated thyroid cells, the binding of TSH to its cognate transmembrane receptor activates adenylate cyclase, which, in turn, increases the level of intracellular cAMP. Reversible phosphorylation of nuclear transacting factors by PKA is linked to transcriptional regulation of several genes (33)(34)(35). We have discovered that AKAP121 expression is regulated by the cAMP signal transduction pathway in thyroid-derived TL5 cells. TSH and/or cAMP induce AKAP121 gene transcription and accumulation of AKAP121 mRNA and anchor protein. The transcriptional induction of AKAP121 mRNA is unaffected by cycloheximide, suggesting the presence of target sequences for PKA-dependent transcription factors in the promoter region of the AKAP121 gene (35). These data do not exclude the possibility that post-transcriptional mechanisms could also increase the stability of AKAP121 mRNA, thereby further facilitating accumulation of the transcript in response to cAMP.
cAMP also markedly increases AKAP121 in GC2 spermatocytes with kinetics similar to that observed in TL5 cells. Thus, cAMP-induced expression of AKAP121 may represent a regulatory mechanism available in a variety of mammalian cells. The developmentally controlled biogenesis of S-AKAP84 in sperm cells is also regulated at the levels of transcription and/or mRNA stabilization in vivo (28). However, it is not known whether cAMP plays a role in this induction process. The observation that AKAP121 can be regulated by hormones (via cAMP signaling) in different cell systems suggests the occurrence of a positive feed-back loop between membranegenerated signals and organelle-specific targeting of AKAP121⅐PKAII complexes. This, ultimately, may modulate the sensitivity and/or the activity of mitochondrial target sites to cAMP action.