INTRODUCTION

Top Abstract Introduction Procedures Results & Discussion References

Protein kinase A (PKA)¹ mediates actions of hormones and neurotransmitters that activate adenylate cyclase (1-4). Signals carried by cAMP are often directed to effectors that accumulate at discrete intracellular sites (5-7). Targeting of signals to these sites can be achieved by generating a non-uniform distribution of PKA molecules. This occurs when PKAII and II isoforms are attached to cytoskeleton or organelles by A kinase anchor proteins (AKAPs) (5, 6). Prototypic anchor proteins (AKAPs 75, 79, and 150) have a binding site for regulatory (RII) subunits of PKAII isoforms and distinct domains that mediate non-covalent coupling of AKAP·PKAII complexes to the microtubule-based dendritic cytoskeleton of neurons and the cortical actin cytoskeleton of non-neuronal cells (5-13). Both cytoskeletal locations are closely apposed to the plasma membrane. Thus, anchored PKAII is positioned near a signal generator (adenylate cyclase) and multiple PKA substrate/effector proteins (e.g. myosin light chain kinase, microtubule-associated protein-2, ion channels, serpentine receptors that couple with G_s) (5, 6, 14). Effector proteins in cAMP signaling pathways can also be associated with organelles, specialized regions of plasma membrane, or sites in cytoskeleton that are separated from adenylate cyclase by substantial distances. Novel RII-binding proteins apparently participate in the assembly of "distal signaling modules" that are associated with mitochondria, Golgi membranes, peroxisomes, and centrioles (5, 6, 15, 16).

Anchored PKAII isoforms may be essential for dissemination of cAMP signals in highly polarized epithelium. Epithelial cells of lung and kidney provide examples of maximally polarized signal transduction, in which hormone-activated adenylate cyclase and downstream target/effectors can be located at opposite ends of the cell. Tight junctions between adjacent cells block communication by diffusible molecules and create two functionally distinct regions of plasma membrane. In epithelial layers of nephrons the basolateral surface is accessible to hormones in plasma and contain receptors and adenylate cyclase; the apical surface is enriched in channel and transporter proteins that mediate absorption/elimination of ions, nutrients, metabolites, and water (17, 18). Activities of many channels and transporters that traverse the apical membrane are regulated by hormones that stimulate adenylate cyclase, thereby promoting PKA activation (17-21). Several considerations suggest that C subunits derived from cytoplasmic PKAs may not mediate trans-epithelial signaling. Signals (cAMP) generated by modest, physiological levels of hormone at the basolateral surface can become weak, diffuse, and insufficient to activate target PKA molecules dispersed (at a relatively low concentration) in cytoplasm. This is due to the transient and intermittent nature of both hormone-release and hormone-mediated stimulation of adenylate cyclase in situ, rapid desensitization of receptor-G protein interactions, dilution of cAMP into the large volume of the cytoplasm, and degradation mediated by cAMP phosphodiesterases. Moreover, studies on knock-out mice and cultured neurons indicate that the ability of PKA isoforms to respond to small changes in cAMP content is often crucial for regulation in vivo (14, 22, 23). Another constraint is that channels/transporters cluster into patches that constitute only a small portion of the apical membrane surface (24, 25). A mechanism that confers increased sensitivity in a specific microenvironment involves concentrating and colocalizing PKAII isoforms with substrate effectors via AKAPs. AKAP·PKAII complexes could serve as sensor/transducers that capture highly diffuse and relatively small trans-cellular signals (cAMP generated at the basolateral surface) with enhanced sensitivity (because of the elevation in local PKAII concentration). Signals can then be rapidly amplified and focused directly on the co-localized effector channel (also at high local concentration) to achieve targeted physiological regulation.

The observation that PKAII is associated with fragments of membrane/cytoskeleton derived from the apical portion of kidney epithelium (21) suggests that the preceding model may be operative in polarized tissue. To gain insights into precise physiological roles for anchored PKAII in establishing and mediating polarized signal transduction, it is essential to identify and characterize AKAPs that are (a) expressed in highly polarized epithelial cells, (b) targeted to the vicinity of the apical surface, and (c) able to avidly bind RII (PKAII) in the context of intact cells. We now describe the discovery and characterization of a novel AKAP, which is expressed in lung and kidney (AKAP-KL) and fulfills the listed criteria.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results & Discussion References

Isolation of cDNAs Encoding AKAP-KL-- A full-length cDNA clone encoding murine RII was obtained as described previously for RII (26). The cDNA was subcloned into the yeast bait plasmid pAS1 (CLONTECH) to generate a fusion gene. The gene encodes a hybrid protein in which the yeast GAL4 DNA-binding domain is appended at the N terminus of RII. A rat brain cDNA library in the yeast expression plasmid (library vector) pGAD10 was obtained from CLONTECH and screened via the two-hybrid interaction assay as described by Fields (27) and Durfee et al. (28). Growth of auxotrophic yeast on medium lacking Trp, His, and Leu and -galactosidase assays were used to isolate and characterize cDNAs encoding candidate AKAPs. A fragment of cDNA (1.4 kilobase pairs) encoding a portion of a novel RII-binding protein was excised from recombinant pGAD10 (by digestion with EcoRI) and was used as a template to generate a random-primed, ³²P-labeled probe. This probe was used to screen a 5'-stretched mouse brain cDNA library in bacteriophage gt11 (CLONTECH) as indicated in previous papers (29, 30). Six positive recombinant phage clones were plaque purified and the cDNAs (0.9-3.8 kilobase pairs) were subcloned in plasmids pGEM7Z (Promega) and pBluescript (Stratagene) and sequenced.

DNA Sequence Analysis-- cDNA inserts were sequenced by a dideoxynucleotide chain termination procedure (31) using T3, T7, and custom synthetic oligonucleotide primers as described previously (30).

Computer Analysis-- Analysis of sequence data, sequence comparisons, and data base searches were performed using PCGENE-Intelligenetics software (Intelligenetics, Mountainview, CA) and BLAST programs (32, 33) provided by the NCBI server at the National Institutes of Health.

Electrophoresis of Proteins-- Proteins were denatured in gel loading buffer and subjected to electrophoresis in 9 or 10% polyacrylamide gels containing 0.1% SDS as described previously (8). Myosin (M_r = 210,000) phosphorylase b (97,000), transferrin (77,000), albumin (68,000), ovalbumin (45,000), and carbonic anhydrase (29,000) were used as standards for the estimation of M_r values.

Western Immunoblot Assays-- Size-fractionated proteins were transferred from denaturing polyacrylamide gels to an Immobilon P membrane (Millipore Corp.) (34). Blots were blocked, incubated with antiserum directed against AKAP-KL (1:2000), and washed as described previously (34, 35). Antigen-IgG complexes were visualized by an indirect chemiluminescence procedure (34, 35). Signals were recorded on Kodak XAR-5 x-ray film.

Overlay Assay for RII Binding-- Overlay binding assays have been described previously (8, 9). In brief, a Western blot is probed with ³²P-labeled RII (using a subunit concentration of 0.3 nM and 1 × 10⁵ cpm of ³²P radioactivity/ml). Complexes of ³²P-RII and AKAPs are visualized by autoradiography. Results were quantified by scanning densitometry (Pharmacia-LKB Ultroscan XL laser densitometer) or PhosphorImager analysis (Molecular Dynamics) (35).

Equilibrium Binding Assay-- Aliquots (80 ng) of highly purified partial AKAP-KL protein (residues 354-741, Fig. 2A) were used for binding assays. Residues 586-605 constitute the high-affinity RII-binding site in AKAP-KL (see "Results"). RII was expressed, purified, and labeled with ³²P as described previously (8, 26). Assays were performed in 250 µl of buffer A (10 mM Tris-HCl, 50 mM sodium phosphate, pH 8.0, 0.1 M NaCl). ³²P-Labeled RII was varied from 0.1 to 75 nM. Incubations were carried out at 0 °C for 90 min. Subsequently, 40 µl of a 50% suspension of Ni²⁺-chelate Sepharose 4B beads (Talon resin, CLONTECH), which was pre-equilibrated with Buffer A, was added and the incubation continued for 30 min. Next, samples were diluted to 1 ml with buffer A and the beads were pelleted by centrifugation at 2,000 × g for 5 min at 4 °C. The beads were washed an 5 additional times by resuspension in 1 ml of buffer A and centrifugation at 2,000 × g at 4 °C. Bound ³²P-labeled RII was eluted from the beads in 0.2 ml of 1% SDS and radioactivity was determined in a scintillation counter. ³²P radioactivity in an aliquot of the first supernatant solution was determined to measure the amount of free RII.

Growth and Transfection of HEK-293 and AV-12 Cells-- Human embryonic kidney fibroblasts (HEK293 cells) and a cell line derived from a hamster subcutaneous tumor (AV-12) were obtained from the American Type Culture Collection. Cells were grown in Dulbecco's modified Eagle's medium containing 10% fetal calf serum. Full-length AKAP-KL2 and AKAP-KL3 cDNA inserts were excised from recombinant pBluescript plasmids by digestion with SpeI and ClaI and ligated into the pEBG mammalian expression vector (36), which was cleaved with the same enzymes. This placed the cDNAs downstream from a powerful EF₁ promoter and a GST gene and upstream from a poly(A) addition signal. Thus, the vector promotes high level expression of GST-AKAP-KL in mammalian cells. Transfections were performed via calcium phosphate precipitation as described previously (13, 35). Stable transformants were obtained by selection with 1 mg/ml G418 for 14 days (34).

Expression and Purification of a Fusion Protein That Contains the RII-binding Domain of AKAP-KL-- AKAP-KL cDNA (nucleotides 1573 to 2736, Fig. 2A) that encodes amino acids 354-741 in the anchor protein was cloned into expression plasmid pET14b as described in previous papers (26, 37). This enabled abundant synthesis of a His-tagged partial AKAP-KL fusion protein in Escherichia coli BL21 (DE3) that was transformed with recombinant plasmid and induced with isopropyl-1-thio--D-galactopyranoside (26, 37). Induced bacteria were disrupted in a French press and the soluble His-tagged AKAP-KL protein was purified to near homogeneity by affinity chromatography on Ni²⁺-chelate-Sepharose 4B beads (Pharmacia) as described previously (26). Two mg of protein were purified from a 600-ml culture.

Production of IgGs Directed against AKAP-KL-- Samples of AKAP-KL fusion protein were injected into rabbits (0.4-mg initial injection; 0.2 mg for each of four booster injections) at Covance Laboratories (Vienna, VA) at 3-week intervals. Serum was collected at 3-week intervals after the first injection.

Deletion and Site-directed Mutagenesis of AKAP-KL-- Deletion mutagenesis was performed via polymerase chain reaction as described for AKAP75 and S-AKAP84 (11, 15). Amino acid substitutions were introduced into the RII-binding domain of AKAP-KL via site-directed mutagenesis, as described previously (11).

Determination of AKAP-KL Localization by Immunofluorescence Analysis and Immunoperoxidase Histochemistry-- HEK293 cells that were stably transfected with an AKAP-KL transgene were fixed and incubated with anti-AKAP-KL antibodies using procedures described by Li et al. (13). AKAP-KL·IgG complexes were visualized by incubation with fluorescein isothiocyanate-coupled secondary antibodies and the utilization of a laser scanning confocal microscope system as described previously (13). F-Actin was visualized by its interaction with rhodamine-tagged phalloidin, as previously reported (13). Mice were sacrificed by in situ fixation (12). Sections of kidney and lung were prepared, probed with antibodies directed against AKAP-KL, and stained via an indirect peroxidase procedure as previously reported (12). A black precipitate is formed at sites containing AKAP-KL·IgG complexes.

	RESULTS AND DISCUSSION

Top Abstract Introduction Procedures Results & Discussion References

Discovery of a Novel Anchor Protein, AKAP-KL-- A cDNA encoding full-length RII was cloned into the yeast plasmid pAS1-CYH2. This recombinant "bait" vector was used to screen a rat brain cDNA library inserted into the yeast "library" vector pGAD10. Interaction of the GAL4 DNA-binding domain-RII fusion protein encoded by the bait vector with RII-binding proteins fused to the activation domain of GAL4 (encoded by library vector) reconstitutes a transcription factor that stimulates GAL4-dependent promoters to transcribe HIS3 and lacZ-reporter genes. The combination of chimeric genes in plasmids and interaction-dependent HIS3 gene expression enable selection of candidate AKAPs by growth of auxotrophic yeast colonies on TrpLeuHis medium. Yeast carrying four different cDNAs were obtained from 10⁶ transformants. Each yeast clone produced positive results in -galactosidase assays. Library plasmids were recovered from yeast and amplified in E. coli. Reconstitution assays showed that plasmids encoding candidate AKAPs did not activate HIS3 or lacZ alone, or in combination with irrelevant bait plasmids. In contrast, co-transformation of yeast with selected library plasmids and the original bait plasmid enabled growth on "triple minus" medium and -galactosidase expression. RII-binding assays were performed on samples of total protein from yeast carrying positive, recombinant plasmids. One fusion protein bound ~10-fold more ³²P-labeled RII than the others (Fig. 1, lane 8). When the corresponding cDNA was sequenced it yielded a derived protein comprising 415 residues. A translation termination codon was evident, but the 5' end of the cDNA corresponded to an open reading frame. Therefore the cDNA insert was used to screen a 5'-stretched mouse brain cDNA library in bacteriophage gt11 via standard DNA hybridization procedures. Six overlapping cDNA clones (including one near the full-length clone) were obtained and sequenced (Fig. 2A). A translation initiation codon was detected and the amino acid sequence of the murine RII-binding protein (named AKAP-KL) was elucidated. The name AKAP-KL was selected because of the expression and highly asymmetric targeting of the anchor protein in epithelial cells of kidney and lung (see below).

View larger version (47K): [in this window] [in a new window]   Fig. 1.   Expression of RII-binding proteins in yeast. Samples of total protein (35 µg) from colonies of transformed yeast were subjected to electrophoresis in a denaturing polyacrylamide gel (10%). Subsequently, size-fractionated proteins were transferred to an Immobilon P (Millipore) membrane and probed with 32P-labeled RII (overlay assay, see "Experimental Procedures") to determine the binding activities of the fusion proteins. Lanes 1, 2, and 8 received fusion proteins from yeast transformed with recombinant pGAD10 plasmids that encode three distinct RII-binding proteins, which were identified in the 2-hybrid interaction screen. Lanes 4 and 5 contained proteins from non-transformed yeast and yeast containing the RII cDNA in the bait vector, respectively. Lanes 6 and 7 received proteins from yeast transformed with pGAD10 plasmids encoding proteins that do not interact with RII. The sample in lane 8 corresponds to a fragment AKAP-KL. An autoradiogram is shown. Only the relevant part of the blot is presented.

View larger version (65K): [in this window] [in a new window]   View larger version (69K): [in this window] [in a new window]   Fig. 2.   Sequence of murine AKAP-KL cDNA and protein isoforms. Panel A presents the nucleotide sequence for AKAP-KL cDNA. The derived amino acid sequence is given below the corresponding codons. Panel B shows an alignment of C-terminal amino acid sequences of AKAPs KL1A, KL2A, and KL3A (see Table I). Residues 1-780 are identical in each of the anchor proteins. The unique C-terminal hexapeptide in AKAP-KL3 isoforms is underlined.

Structure-function Relationships in AKAP-KL-- AKAP-KL is composed of 885 amino acids and has a calculated M_r of 98,000 (Fig. 2A). The sequence of the acidic polypeptide (pI ~ 5.0) is not homologous with sequences of previously characterized proteins. However, several domains that could potentially contribute to functional roles for AKAP-KL are evident. The sequence between residues 586 and 605 (Fig. 2A) of AKAP-KL aligns with the RII-binding site of AKAP75 (11) to yield 45% overall identity (Fig. 3A). A central core of 7 amino acids (Leu⁵⁹³ to Gln⁵⁹⁹ in AKAP-KL) is nearly invariant (86% identity) in the two proteins. Furthermore, five amino acids with large aliphatic side chains, which coordinately regulate RII binding affinity in AKAP75 (11), are conserved in AKAP-KL (Leu⁵⁸⁶, Leu⁵⁹³, Val⁵⁹⁴, Ile⁵⁹⁸, and Ile⁶⁰²). A partial AKAP-KL protein (residues 354-741) that includes the putative tethering site binds both RII (Fig. 3B, lane 1) and RII (not shown). Two approaches verified the functionality of the potential RII-binding site in AKAP-KL. First, a hydrophobic residue (Ile⁵⁹⁸) predicted to be essential for binding RII (by analogy with AKAP75) was mutated to Ala. This substitution diminished RII binding activity of partial AKAP-KL by >95% (Fig. 3B, lane 3). A similar effect was observed when the corresponding Ile in AKAP75 was replaced with Ala (11). Mutation of an amino acid predicted to be non-essential (Ala⁵⁹⁰ to Ser) had little effect on the tethering of RII (Fig. 3B, lane 2). Partial AKAP-KL proteins that lack residues 586-605 have no RII binding activity.²

View larger version (28K): [in this window] [in a new window]   Fig. 3.   Characterization of the RII-binding site in AKAP-KL. A, the RII tethering site of AKAP-KL is aligned with the RII-binding site of AKAP75 (11). Asterisks indicate the positions of conserved residues. B, amino acids in the RII-binding site of AKAP-KL were altered by site-directed mutagenesis as indicated under "Experimental Procedures." Partial AKAP-KL proteins (residues 354-741) containing wild-type and mutant binding domains were expressed and assayed for RII binding activity (overlay assay) as described under "Experimental Procedures." Equal amounts of fusion protein (0.5 µg) were applied to each lane. An autoradiogram is shown. The AKAP-KL fusion protein exhibits an apparent Mr of 50,000. Lane 1 received fusion protein with the wild type binding region. Two other fusion proteins contained RII-binding domains with mutations: lane 2, Ala590  Ser; lane 3, Ile598  Ala. Only the relevant portion of the gel is shown. No other bands were observed. C, equilibrium binding of 32P-labeled RII with the binding domain of AKAP-KL was determined as described under "Experimental Procedures" and "Results and Discussion." The data are plotted according to the method of Scatchard (39) and the Kd was determined from the slope.

Six Isoforms of AKAP-KL Are Produced in Mammalian Cells-- Sequencing of 3' ends of AKAP-KL cDNAs revealed two alternative modes of AKAP-KL mRNA splicing. In the first instance, a splice donor sequence at nucleotide 2878 (Fig. 2A) is joined to an acceptor site at position 3097, thereby deleting 218 nucleotides in cDNA. This deletion shifts the reading frame after codon 789 so that the succeeding 21 nucleotides encode a novel, C-terminal hexapeptide (Fig. 2B) and a translation termination signal. As a result, the C terminus of the anchor protein is truncated by 90 residues. AKAP-KL proteins that lack 90 C-terminal residues are designated AKAP-KL3; AKAP-KL proteins that retain this region are designated AKAP-KL1. A second mode of RNA splicing joins a donor sequence at nucleotide 3058 (Fig. 2A) with the acceptor site at nucleotide 3097. This splicing reaction excises 13 codons (for amino acids 849-861) and does not alter the reading frame (Fig. 2B). Anchor proteins that lack residues 849-861 are named AKAP-KL2.

View larger version (48K): [in this window] [in a new window]   Fig. 4.   Production and characterization of antibodies directed against AKAP-KL. A His-tagged fusion protein that includes amino acids 354-741 of AKAP-KL (Fig. 2A) was expressed in E. coli and purified by affinity chromatography as indicated under "Experimental Procedures" and "Results and Discussion." A, samples of total protein from induced E. coli (30 µg, lane 2), proteins in the column flow-through (30 µg, lane 1), and protein from the pooled peak fractions eluted with 1 M imidazole (1 µg, lane 3) were size-fractionated by denaturing electrophoresis. A 10% polyacrylamide gel stained with Coomassie Blue is shown. The partial AKAP-KL protein has an apparent Mr of 50,000. B, cytosol (100,000 × g supernatant solution) and two particulate fractions (P1, 10,000 × g pellet; P2, 100,000 × g pellet) were prepared from mouse kidney as described previously (8, 12). Samples (35 µg) of cytosolic, P1 and P2 proteins were size-fractionated in a denaturing polyacrylamide gel (9%) and transferred to an Immobilon P membrane. Lanes 1 and 5 received cytosolic proteins; lanes 2 and 6 contained P2 proteins; lanes 3, 4, and 7-9 were loaded with proteins from the P1 pellet. Lanes 1-4 from the Western blot were incubated with preimmune serum (1:1000 dilution) and lanes 5-8 were probed with antiserum directed against AKAP-KL (1:2000 dilution) as indicated under "Experimental Procedures." Excess purified antigen (3 µg) was present when lanes 4 and 8 were probed with preimmune serum and antiserum. The immunoblots were developed by an enhanced chemiluminescence procedure and signals were recorded on x-ray film (see "Experimental Procedures"). RII-binding proteins in the P1 pellet were identified and characterized by the overlay binding assay (lane 9). An autoradiogram is shown.

Preparation and Specificity of Antibodies Directed against AKAP-KL-- AKAP-KL cDNAs were cloned from a brain cDNA library. Since a high proportion of the complete constellation of mammalian genes is expressed in the central nervous system, the efficiency of screening is maximized. However, isolation of reverse transcripts of AKAP-KL mRNAs from this library does not imply that brain is the principal site of accumulation of the anchor protein. To determine the tissue and cellular distributions, as well as functions of AKAP-KL it is essential to generate highly specific anti-AKAP-KL IgGs.

Expression of AKAP-KL Is Tissue-specific and Isoform-selective-- Analysis of P1 fractions from various tissues disclosed that AKAP-KL is abundantly expressed in lung (Fig. 5A, lane 1). In contrast, the anchor protein is either absent or produced at very low levels in multiple tissues, including liver, heart, and cerebral cortex (Fig. 5A, lanes 3-5). Moderate levels of AKAP-KL were detected in thymus and cerebellum. AKAP-KL isoforms observed in cerebellum (Fig. 5A, lane 6) do not co-migrate with anchor proteins that accumulate in lung. Thus, expression of the AKAP-KL gene is tightly regulated. Accumulation of AKAP-KL polypeptides is restricted to certain cell/tissue types and anchor protein isoforms that predominate in a given tissue are evidently determined by a combination of regulated AKAP-KL mRNA splicing and utilization of two translation initiation codons.

View larger version (76K): [in this window] [in a new window]   Fig. 5.   Tissue-specific and isoform-selective expression of AKAP-KL isoforms. A, P1 pellet fractions were prepared from various tissues as indicated in the legend for Fig. 4. AKAP-KL polypeptides were detected by Western immunoblot analysis (see "Experimental Procedures" and Fig. 4), using anti-AKAP-KL serum at a dilution of 1:2000. P1 proteins (30 µg/lane) from the following mouse tissues were analyzed: lung (lane 1), thymus (lane 2), heart (lane 3), cerebral cortex (lane 4), liver (lane 5), and cerebellum (lane 6). B, Western immunoblot analysis was performed using anti-AKAP-KL serum at 1:2000. Samples (30 µg) of total particulate proteins were isolated from AV-12 cells transfected with AKAP-KL transgenes that encode: only the 3B isoform of AKAP-KL (lane 1), both the 3A and 3B isoforms of AKAP-KL (lane 2), or both the 2A and 2B isoforms of AKAP-KL (lane 3). (See Table I for nomenclature and sizes of the anchor protein isoforms.) Lane 4 received P1 proteins isolated from a lung homogenate.

Properties of AKAP-KL in Vitro and in Cells-- Extraction with buffers containing 0.5% Triton X-100 failed to solubilize the anchor protein, indicating that is not embedded in a lipid bilayer (Fig. 6A). Buffers containing 0.5% sodium deoxycholate (which disrupts association of proteins in cytoskeleton) efficiently solubilize AKAP-KL. Confocal immunofluorescence microscopy of HEK293 cells that were stably transfected with cDNA encoding full-length AKAPs KL2A and 2B revealed the intracellular distribution of the anchor protein. AKAP-KL accumulates in regions of cortical cytoskeleton that appear as projections or large clusters of antigen (Fig. 6B). These structures are not present in control HEK293 cells or cells transfected with unrelated transgenes (13). In contrast, AKAP75 is dispersed throughout the cortical cytoskeleton in the same cells (13). Thus, AKAP-KL appears to be targeted to specific microenvironments in cytoskeleton. It is possible that AKAP-KL actively promotes assembly/organization of specialized structures via interactions with other proteins in cytoskeleton. The differential clustering of AKAP-KL in subdomains of cytoskeleton in HEK293 cells may also reflect properties involved in targeting of the anchor protein to the vicinity of the apical surface of polarized epithelial cells (see below).

View larger version (40K): [in this window] [in a new window]   Fig. 6.   AKAP-KL associates with actin cytoskeleton. Proteins were solubilized from the P1 particulate fraction of kidney homogenates as described previously (29). Supernatant fractions (40 µg of protein) derived from the P1 pellet were collected and assayed for AKAP-KL polypeptides by a Western immunoblot assay. Lane 2 received proteins solubilized with 0.5% sodium deoxycholate; lane 3 contained proteins solubilized with 1% Triton X-100; lane 4 was loaded with proteins extracted with 1 M NaCl; lane 1 contained 30 µg of total P1 particulate proteins. B, confocal immunofluorescence microscopy of HEK293 cells stably transfected with an AKAP-KL transgene. Signals were obtained with anti-AKAP-KL antibodies and secondary antibodies tagged with fluorescein isothiocyanate. No signals were seen after blocking with excess antigen or with preimmune serum. AKAP-KL accumulates in clusters and projections. C, cells shown in B were incubated with 5 µM cytochalasin D for 1 h. Immunofluorescence microscopy shows that AKAP-KL re-distributes uniformly at the periphery and in large internal aggregates. D, double immmunostaining for AKAP-KL and F-actin is shown after treating cells described in B with cytochalasin D. The distribution of AKAP-KL (shown on the left) was determined by incubating with anti-AKAP-KL serum and fluorescein isothiocyanate-coupled secondary antibodies. The location of F-actin (right) was established by probing with rhodamine-phalloidin.

AKAP-KL Is Targeted to the Apical Surface of Polarized Epithelial Cells-- Anti-AKAP-KL IgGs were used in an indirect immunoperoxidase staining procedure to determine the location of AKAP-KL in sections of rat kidney. The anchor protein is expressed in epithelial cells and accumulates exclusively at the apical surface, which abuts the lumen (e.g. see epithelial cells in proximal tubules, Figs. 7, A and B). RII is also concentrated at the apical surface of epithelial cells in renal proximal tubules (21). In lung, AKAP-KL accumulates in long, linear structures just below the lumenal surface of alveolar epithelial cells (Fig. 7C). In parallel, a substantial amount of RII is also concentrated in the elongated structures near the apical surface of alveolar epithelial cells (Fig. 7D). The preceding results suggest that AKAP-KL and RII are co-localized and co-enriched in two types of polarized epithelium. Confirmation of this working hypothesis will ultimately require high-resolution, immunoelectron microscopy and manipulation of the system by molecular genetics (e.g. disrupting AKAP-KL·RII complexes by expressing AKAP-KL mutants that bind RII, but are mis-targeted, or by ablating AKAP-KL gene expression) in future studies. However, the present observations support the idea that AKAP-KL incorporates PKAII into distal signaling modules at sites near junctions of cytoskeleton and apical plasma membrane. Such modules might include PKA substrate/effector proteins (e.g. transporters) which span the apical membrane. This mode of organization can greatly facilitate cAMP-mediated trans-epithelial signal transduction.

View larger version (101K): [in this window] [in a new window]   Fig. 7.   Distribution of AKAP-KL in kidney and lung. A, immunoperoxidase staining of kidney sections that contain proximal tubules. The stain for AKAP-KL (heavy black precipitate) is located at the lumenal (apical) surface of the tubules. No staining was detected with preimmune serum or with immune serum preincubated with excess AKAP-KL antigen. Magnification × 400. B, examination of proximal tubule staining at lower magnification confirms that AKAP-KL is selectively targeted to the apical surface (as) of polarized epithelial cells. The lumen is marked as lu; the (unstained) glomerulus as g. Magnification × 200. Panels C and D show the immunoperoxidase staining of alveolar epithelial cells for AKAP-KL and total RII, respectively. The anchor protein accumulates in long strand-like structures (e.g. see arrowhead in C) that lie just under the plasma membranes of highly flattened and polarized epithelial cells that mediate gas exchange between blood and air. A substantial portion of RII in alveolar epithelial cells (arrowhead in D) appears to be co-localized in the same structures as AKAP-KL. Counterstained nuclei appear as dark spheroids in the lung sections. Preincubation of anti-AKAP-KL serum with excess purified partial AKAP-KL antigen (panel E) eliminates the anchor protein signal; likewise, panel F shows that excess purified antigens (3 µg of RII and RII) abrogate signals evident in panel D.

AKAP-KL Binds RII with High Affinity in Intact Cells-- The pEBG expression vector (36) was used to assess the ability of the tethering domain AKAP-KL to sequester RII in intact cells. pEBG contains a powerful EF₁ promoter that lies upstream from the glutathione S-transferase (GST) gene and a 3' multiple cloning region. AKAP-KL cDNA was inserted, in-frame with the GST gene, and the vector was co-transfected with R_C/CMV (an expression plasmid which contains the neo^R gene under control of an SV40 promoter) into hamster AV-12 cells. Stable cell lines expressing GST-AKAP-KL were isolated by selection with G418. Typically, ~80% of GST-AKAP-KL appeared to be anchored, whereas 20% was present in cytosol. Cytosolic GST·AKAP-KL complexes were purified to near homogeneity by affinity chromatography on GSH-Sepharose 4B. Western blot analysis revealed that a large proportion of available cytosolic RII was complexed with the anchor protein (Fig. 8). Two conditions were varied to exclude the possibility that AKAP-KL·RII complexes assembled after cell homogenization. Since AKAP-KL·RII complexes dissociate very slowly at 4 °C,² cells were disrupted: (a) in a volume of lysis buffer that was increased 20-fold to reduce concentrations of free RII (PKAII) and anchor protein or (b) in a standard volume of buffer that contained a vast excess (15 µg/ml) of His-tagged AKAP-KL partial protein (see Fig. 4A). The partial anchor protein avidly binds RII (Fig. 3C), but is not sequestered by GSH-Sepharose 4B resin. Neither dilution nor post-lysis competition with an RII-binding protein altered the results obtained in Fig. 8 (data not shown). Thus, stable AKAP-KL·RII complexes are efficiently produced in the environment of the internal milieu of intact cells.

View larger version (79K): [in this window] [in a new window]   Fig. 8.   Isolation of AKAP-KL·RII complexes from intact cells. Cytosol (5 ml) was prepared from five 10-cm plates of AV-12 cells (13, 34) that were stably transfected with a chimeric GST-AKAP-KL gene (see "Experimental Procedures"). GST-AKAP-KL and associated proteins were isolated and purified by affinity chromatography on GSH-Sepharose 4B using previously described standard procedures (29, 47). Samples of proteins (50 µg) from non-fractionated cytosol (lane 1) and cytosol depleted of GST-AKAP-KL (lane 3), as well as an aliquot (20 µl) of proteins eluted from the affinity column with 20 mM GSH (lane 2) were analyzed for RII content by a Western immunoblot assay. RII has an apparent Mr of 54,000.

Conclusions and Implications-- AKAP-KL isoforms are differentially expressed and selectively targeted to apical regions of polarized lung and kidney epithelial cells. This family of novel anchor proteins efficiently sequesters RII (PKAII) in vitro and in situ. AKAP-KL proteins may be involved in (a) establishing polarity in trans-epithelial signaling systems and (b) creating/organizing distal PKA·effector complexes that receive, amplify, and focus trans-cellular signals carried by cAMP. Generation of six AKAP-KL isoforms by post-transcriptional processes creates substantial anchor protein diversity and suggests possible mechanisms by which intracellular localization of PKAII isoforms may be modulated in different cell types and in response to environmental stimuli.