BIOLOGICAL OVERVIEW

Protein kinase A (PKA) holoenzyme is anchored to specific subcellular regions by interactions between regulatory subunits (Pka-R) and A-kinase anchoring proteins (AKAPs). The functional importance of PKA anchoring during Drosophila oogenesis has been examined by analyzing membrane integrity and actin structures in mutants with disruptions in an AKAP known as Akap200. In wild-type ovaries, cAMP-dependent protein kinase R2 (Pka-RII) the regulatory subunit of cAMP-dependent protein kinase 1 (Pka-C1, also known as Protein kinase A) and Akap200 localize to membranes and to the outer rim of ring canals (actin-rich structures that connect germline cells). In Akap200 mutant ovaries, Pka-RII membrane localization decreases, leading to a destabilization of membrane structures and the formation of binucleate nurse cells. Defects in membrane integrity could be mimicked by expressing a constitutively active PKA catalytic subunit (Pka-C) throughout germline cells. Unexpectedly, nurse cells in Akap200 mutant ovaries also have enlarged, thin ring canals. In contrast, overexpressing Akap200 in the germline results in thicker, smaller ring canals. To investigate the role of Akap200 in regulating ring canal growth, genetic interactions with other genes that are known to regulate ring canal morphology were examined. Akap200 mutations suppress the small ring canal phenotype produced by Src64B mutants, linking Akap200 with the non-receptor tyrosine kinase pathway. Together, these results provide the first evidence that PKA localization is required for morphogenesis of actin structures in an intact organism (Jackson, 2002).

During development, cell function and morphology are influenced by exposure to myriad extracellular signals. Protein kinase A, a ubiquitous, highly conserved serine-threonine kinase, is a key intracellular transducer of many hormonal and other extracellular signals. In the absence of cAMP, inactive Protein kinase A is a holoenzyme (PKA), a heterotetramer of two identical catalytic subunits (Pka-C) and two identical regulatory subunits (Pka-R). When cAMP is present, it binds to the regulatory subunits and releases catalytic subunits from the holoenzyme, allowing phosphorylation of target substrates. In metazoans (except C. elegans), the PKA regulatory subunits are of two types: type I (Pka-RI) or type II (Pka-RII). Mammals have two isoforms of each type of regulatory subunit (e.g. Pka-RIa or Pka-RIß) but Drosophila has only one isoform of each type of regulatory subunit. Once activated, the catalytic subunits are capable of phosphorylating a large number of protein substrates, both in vitro and in vivo (Jackson, 2002 and references therein).

The promiscuous enzymatic activity of the Pka-C subunit raises questions as to how cAMP-mediated signaling can achieve specific cellular responses. One proposed mechanism is that PKA holoenzyme is regulated by sequestration near specific targets, ensuring a rapid local response to an extracellular signal. This sequestration is achieved through A-kinase anchoring proteins (AKAPs), a heterogeneous family of proteins that bind to PKA regulatory subunits and anchor PKA holoenzyme (Scott, 1994). Many different subcellular regions are targets for PKA anchoring (Colledge, 1999). Pka-RI can interact with dual-specificity AKAPs that bind to either Pka-RI or Pka-RII (Huang, 1997). In mammalian cells, the function of AKAPs can be disrupted by treatment with Ht31, a dominant negative peptide that interferes with AKAP-Pka-RII interactions (Carr, 1991). Disrupting PKA localization with Ht31 interfers with several cellular functions, including the PKA-dependent regulation of L-type calcium channels in HEK293 cells (Gao, 1997) and skeletal muscle cells (Johnson, 1994), AMPA or kainate receptor potentiation in hippocampal neurons (Rosenmund, 1994), calcium-activated potassium channels in isolated smooth muscle patches (Wang, 1996), aquaporin channel function in pancreatic cells (Klussmann, 1999) and apoptotic pathways (Harada, 1999). These analyses reveal the importance of AKAP-Pka-RII interactions and support the hypothesis that sequestration of PKA may facilitate localized and/or specific response to cAMP signaling (Jackson, 2002).

In Drosophila, two AKAPs have been cloned using in vitro gel overlay assays. Akap550 encodes a protein that has no known functional domains other than a coiled-coil Pka-RII binding site (Han, 1997). Immunocytochemical analysis of Akap550 shows that it is present in the cytoplasm of developing gut, trachea and salivary gland cells and is particularly elevated in the nervous system (Han, 1997) and in other adult tissues. The second AKAP, Akap200, contains both a coiled-coil Pka-RII binding domain and a myristoylated alanine-rich C-kinase substrate (MARCKS) domain (Li, 1999). Alternate splicing of this gene produces two transcripts; one contains the Pka-RII binding domain and one lacks it, but the MARCKS domain is retained in both. Akap200 protein associates with membranes and subcortical regions of cells when transfected into cultured Schneider cells, and purified Akap200 binds to purified actin filaments in vitro (Rossi, 1999). These observations suggest that Akap200 is involved in linking PKA signaling to the actin cytoskeleton (Jackson, 2002).

Although Akap200 is expressed throughout development in the fly, its function is not known. Drosophila oogenesis provides an ideal model system for analyzing AKAP-mediated regulation of PKA signaling events in a developmental context. The egg chambers present in the Drosophila ovary contain large cells with well-defined and easily observed actin-derived structures. Furthermore, large numbers of egg chambers are easy to obtain and egg chambers at different stages of development are present within a single ovary. Oogenesis begins in the germarium, a structure at the anterior tip of each ovariole that contains the germline and somatic precursor cells. Approximately halfway down the germarium, interleaving follicle cells migrate to surround a single germline cyst and form an egg chamber. Thus, each egg chamber consists of an epithelial monolayer of somatically-derived follicle cells surrounding a syncytial cyst of 16 germline-derived cells. The germline cyst arises by four incomplete cytokineses of a germline cystoblast cell and contains 15 polyploid nurse cells and a single diploid oocyte. The nurse cells and oocyte are connected to each other via specialized channels known as ring canals. These ring canals are rich in filamentous actin as well as several other proteins that regulate the morphology and growth of the ring canals during oogenesis (Jackson, 2002 and references therein).

The role of PKA anchoring has been examined in this developmental context. Akap200 and Pka-RII colocalize to membranes and ring canals during oogenesis and loss of Akap200 function diminishes Pka-RII membrane localization. Moreover, loss of Akap200 function affects the structural integrity of the germline-derived cells. Expressing a mutant PKA catalytic subunit, incapable of associating with regulatory subunits and therefore incapable of being properly sequestered, produces phenotypes similar to loss of Akap200 function. Akap200 mutations result in large, thin ring canals; conversely, Akap200 overexpression in the germline yields smaller, thicker ring canals. Finally, loss of Akap200 function suppresses the small ring canal phenotype produced by Src64B mutants (Jackson, 2002).

These results provide in vivo support of a model for regulating Pka-C activity by subcellular sequestration. It is proposed that Akap200 in the vicinity of ring canals and nurse cell membranes anchors Pka-RII to these same subcellular regions. In turn, Pka-RII binds and inactivates Pka-C; the Akap200-Pka-RII complex may be thought of as a 'sponge' for active Pka-C subunit. Some, if not a majority, of the Pka-C subunits in the germline are sequestered in the vicinity of membranes as well and may result from associations with the Akap200-PkaRII complex. Several different cytoskeleton-associated proteins that regulate the morphology of filamentous actin might be regulated by Pka-C phosphorylation, and the combination of activities of these other proteins and Pka-C ensures that the actin filaments are properly assembled and their morphology maintained (Jackson, 2002).

In this model, Akap200 could have two roles. By binding to other proteins besides Pka-RII, Akap200 may act as a scaffold that brings substrates and other regulatory proteins to the membrane, ensuring rapid, local responses to PKA activation. Alternatively, by interacting with the inhibitory regulatory subunits, Akap200 may act as a 'sponge' to ensure that concentrations of active Pka-C subunit are precisely controlled at the subcellular level. These results, together with studies of Pka-C loss-of-function in the ovary, suggest that the precise concentration of Pka-C activity at the membrane is critical to regulate the morphology of actin structures in egg chambers, and that one of the functions of Akap200 is to control this activity. In Pka-C loss of function mutants, some of the cytoskeleton-associated proteins that require Pka-C phosphorylation may fail to become active. As a result, local instabilities in subcortical actin filaments may provide weak points for rapid disassembly of the remaining filaments, ultimately producing a binucleate nurse cell. In Akap200 mutants, Pka-RII cannot be sequestered; therefore, the regulatory 'sponge' discussed above does not form properly. Similarly, when a mutant Pka-C incapable of being properly sequestered is expressed in the nurse cells, it does not matter that AKAP200/Pka-RII 'sponge' is localized appropriately. These two changes produce higher local concentrations of active Pka-C in the vicinity of subcortical actin and ring canals, which in turn phosphorylate other actin-associated proteins and modify their activity inappropriately. Although a different set of cytoskeleton-associated proteins may be misregulated by ectopic Pka-C activity than loss of Pka-C activity, the end result of local destabilization of actin filaments and nurse cell fusion could be the same. This hypothesis predicts that loss of Pka-RII function would also result in ectopic, active Pka-C and would produce a similar phenotype. Indeed, nurse cell fusions with very similar morphology have been reported in Pka-RII mutant ovaries (Park, 2000). Finally, one of the predictions of a scaffold model is that expressing Pka-C* throughout the germline cells might be expected to overcome the lack of a scaffold in Akap200 mutants. It was found, however, that expressing Pka-C* in Akap200 mutants exacerbates the frequency of binucleate cell production. These results support the hypothesis that the amount of active Pka-C is regulated locally in the ovary and argue against the need for Akap200 to bring Pka-C to the vicinity of substrate molecules. Nevertheless, these two mechanisms are not mutually exclusive, and Pka-C regulation by Akap200 in these cells could be complex. The identities of the hypothesized actin-associated proteins remain to be determined. One intriguing candidate is the p21-PAK -- a regulator of actin filament dynamics via LIM kinase and cofilin, shown recently to be down-regulated by Pka-C phosphorylation (Howe, 2000). Other likely candidates include those uncovered in a screen (Jackson, 1999) for mutations that produce binucleate cells (Jackson, 2002).

Loss of Akap200 does not completely eliminate all Pka-RII membrane localization, possibly because even the strongest alleles, which eliminate all detectable Akap200 protein, may not be null. Other proteins besides Akap200 are also likely to contribute to Pka-RII sequestration in the vicinity of the membranes. In mammalian cells, the widely expressed AKAP75/79/150 sequesters Pka-RII subunit to membranes in transfected HEK293 cells (Li, 1996), and ezrin, an actin-binding protein, is an AKAP in gastric mucosal cells (Dransfield, 1997). However, neither of these proteins can be responsible for the residual Pka-RII membrane staining in Akap200 mutants because no obvious homologs to either AKAP75/79/150 or ezrin exist in the fly genome. In fact, the overall lack of strong sequence homology among AKAPs hinders the ability to identify homologs by database searching. Other Pka-RII binding proteins that may be present in the fly ovary must therefore be identified using other methods (Jackson, 2002).

The subcellular localization of endogenous Akap200 in ovaries is consistent with previous studies showing that the myristoylation site within the MARCKS domain is responsible for Akap200 membrane association in transfected Schneider cells (Rossi, 1999). The current studies do not distinguish whether the localization in the ovary results from direct membrane association via the myristoylation modification, or from associations with subcortical actin via the actin-binding motif found within MARCKS domain proteins. The localization of Akap200 on the outer rim of ring canals could be explained by direct interactions with the filamentous actin found on the inner rim of the ring canal, or by embedding Akap200 into the plasma membrane near the outer rim via its myristoylation motif. Interestingly, no Akap200 is detected associated with other filamentous actin-containing structures in the ovary, such as the spectrosome, the fusome or the actin fibers that anchor nurse cell nuclei during dumping. This observation raises the possibility that Akap200 does not bind to ovarian actin filaments directly; rather, the myristoylation domain may embed Akap200 in the membrane near subcortical actin and ring canals. Alternatively, the filamentous actin-binding activity of Akap200 may be regulated, such that Akap200 is capable of binding only to subsets of actin filaments in the ovary. This regulation may be mediated by Protein kinase C or CAM kinase modifications in the MARCKS (PKC substrate) domain, or phosphorylation elsewhere in the protein by non-receptor tyrosine kinases. Further experimentation is needed to distinguish among these possibilities (Jackson, 2002).

One surprising phenotype observed in Akap200 mutants is that ring canals are larger. The mechanism by which ring canal size is regulated during oogenesis is not yet clear. Mutations in two non-receptor tyrosine kinase genes (Src64B and Tec29) make smaller ring canals and are important for regulating ring canal size. Interestingly, these mutations also produce binucleate cells; however, abnormally sized ring canals have not been described in other mutants that make multinucleate cells (such as chickadee or cappuccino). Unexpectedly, it was found that changing Akap200 levels in the germline influences the size of the ring canals. Two general mechanisms are envisioned that could produce changes in ring canal size: defective architecture or defective temporal control. First, defective assembly of ring canal components could produce ring canals that cannot respond to structural stresses appropriately. For example, in Akap200 mutants, a component of the ring canal may be missing or misregulated, resulting in a complex that is not tightly held together. The diameter of the ring canal could therefore expand abnormally, producing larger, thinner ring canals. Conversely, the smaller, thicker ring canals observed when Akap200 was overexpressed and in DSr64B mutants could be produced if the entire ring canal complex contracted abnormally as a result of a misregulated or missing component. The second possibility is that stage-specific assembly of the entire ring canal complex may be accelerated (or arrested), resulting in larger (or smaller) ring canals that are inappropriately sized for the stage of the egg chamber. In other words, the large ring canals seen in Akap200 mutants could be explained by hypothesizing that stage nine nurse cells 'think' they should be making stage ten-sized ring canals. Because Src64B, Tec29 and Akap200 are all linked to the actin cytoskeleton, the first hypothesis is favored, but the second possibility could not be discounted because the order of assembly and the regulation of ring canal growth are incompletely understood processes (Jackson, 2002).

Nevertheless, the Akap200 gene product likely contributes to altered ring canal size in two non-exclusive ways. (1) Proteins regulating ring canal growth and morphology could affect the functional activity of Akap200. (2) Akap200 could influence the expression levels, subcellular localization or functional activity of proteins associated with ring canals (Jackson, 2002).

The first possibility was tested by examining Akap200 localization in mutants that affect ring canal size and structure. However, no changes in Akap200 expression levels or localization were detected in Src64B^PI or in hts¹ mutants. Nevertheless, future analyses may demonstrate that the localization pattern of Akap200 in other mutants is directly regulated by a gene that regulates ring canal size. Alternatively, Akap200 may regulate the localization or function of a protein that is assembled into ring canals. This second possibility was tested by examining ring canal structure in Akap200 mutants. No significant differences could be detected in the expression levels or subcellular localization of known ring canal-associated proteins or markers (such as phalloidin, Hts-RC, anti-phosphotyrosine, Kelch, Filamin and Src64B). Critical gene products that do undergo such a change may exist, but these remain hitherto unidentified. Furthermore, covalent modifications that change the function of the protein but not the expression levels or localization would not be detected by these assays and would require biochemical experiments to detect such modifications. The analysis of other ring canal-associated mutants and their interaction with Akap200 may eventually provide insights into the mechanism of ring canal growth (Jackson, 2002).

How do the ring canals fit into the model explaining the nurse cell fusions? The larger, thinner ring canals could be a manifestation of weakened supporting elements in the vicinity of the ring canal that result from ectopic active Pka-C-mediated phosphorylation of the other cytoskeleton-associated proteins. This structural weakening could provide an opportunity for the ring canal to expand and ultimately break, resulting in a retraction of membrane from the ring canal and the freeing of a ring canal remnant (Jackson, 2002).

Alternatively, the increased ring canal size may be unrelated to the formation of multinucleate cells; rather, a second mechanism prevents fusion of the nurse cells. This hypothesis raises the intriguing possibility that the two domains of Akap200 provide different functions: the MARCKS domain may regulate ring canal size, whereas the Pka-RII binding region may be involved in preserving the cytoarchitecture of the cells. In support of this second hypothesis, multinucleate cells with approximately normal-sized ring canals are produced in the hypomorphic Akap200^ix4 allele, in other PKA signaling- and cytoskeleton-associated mutants, and by overexpressing Pka-C in the germline cells. Thus, the large ring canal phenotype may be separable from the nurse-cell fusion phenotype. Although other binding partners for Akap200 are not currently known, this protein may bring together both PKA and PKC signaling pathways at or near the actin cytoskeleton (Li, 1999). One of these other pathways may be linked to regulating ring canal size, but not the formation of multinucleate cells (Jackson, 2002).

The importance of PKA localization has been the subject of investigation and speculation for several years. For example, PKA localization is important in regulating the activity of ion channels in the context of cultured cells. In contrast, other studies found that localization of Pka-RII by AKAPs may not be required for calcium channel modulation in skeletal muscle or for mouse sperm motility. These analyses of Akap200 mutants provide evidence that PKA localization is important for regulating the function of specialized actin structures during oogenesis in Drosophila. The analysis of AKAP mutant phenotypes in model organisms may provide the tissue- and cell-specific context necessary to detect subtle phenotypes that have profound consequences on cellular functions (Jackson, 2002).

GENE STRUCTURE

BDGP P1 clone DS02110 (GenBank accession number AC004423) contains the Akap200 gene as well as complete 5'- and 3'-flanking regions. Alignment of Akap200 cDNA with portions of the DS02110 sequence revealed the organization of the cognate gene. The Akap200 structural gene spans 15,220 bp and is composed of six exons and five introns. The first exon serves as a template for transcription of a short segment of 5'-untranslated mRNA. Exon 1 precedes a large intron (~11 kbp) that accounts for >70% of the nucleotides at this locus. Exons 2-6, which encode the Akap200 open reading frame and 3'-untranslated nucleotides in mRNA, contain a total of 2773 bp. Small and medium size introns (introns 2-5) contribute only 1200 bp of intervening DNA sequence in this region. Exon 5 is atypically large and includes a block of 381 codons that direct the incorporation of ~50% of the constituent amino acid residues into the Akap200 polypeptide. A Pro-rich region and the RII tethering domain are encoded by exon 5. The Akap200 gene lies between genes named fuzzy and gurken at positions 2-32 on the physical (genome) map (Li, 1999).

A novel, 1.9-kbp insert was discovered by sequencing candidate Akap200 cDNAs obtained from phage and bacterial libraries. The sequence of 5'-untranslated nucleotides and a large contiguous DNA segment comprising 344 codons was identical with nucleotides 94-1186 in Akap200 cDNA (GenBank accession number AF132884). However, nucleotide 1186 was directly linked to a 3' sequence that matches perfectly with nucleotides at positions 2330-2964 in Akap200 cDNA. Alignment of the 1.9-kbp cDNA (GenBank accession number AF132885) and Akap200 gene sequences has revealed that the smaller cDNA was produced by the precise excision of exon 5 from the Akap200 transcript. The shorter cDNA and its cognate polypeptide were named "deleted Akap200," or DAkap200, to indicate their derivation from Akap200 via exon elimination. Excision of exon 5 results in the loss of a block of 381 residues that includes the RII binding region and a Pro-rich sequence that is a candidate Src homology 3 binding site. Other structural features of Akap200 are retained in the 372-residue DAkap200 polypeptide. DAkap200 is a highly acidic protein (pI ~4.2) in which Glu, Asp, Ala, Pro, Ser, and Thr contribute 62% of total amino acids. The protein lacks Tyr, whereas Met, Cys, and Trp appear only once in the conceptual translation of the protein. DAkap200 has a calculated M_r of 38,000, but its atypical migration in denaturing electrophoresis yields an apparent M_r of 95,000. Moreover, the N-terminal myristoylation site (residues 1-7), the highly basic PSD (phosphorylation site domain)-like region (residues 118-148), and a Pro-rich (and Thr-rich) segment (residues 316-335) are present in both DAkap200 and the larger Akap200 isoform (Li, 1999).

PROTEIN STRUCTURE

Amino Acids - 753 (AKAP200-P1) and 273 (AKAP200-P2)

Structural Domains

A unique Drosophila gene encodes two novel signaling proteins. Drosophila A kinase anchor protein 200 binds regulatory subunits of protein kinase AII (PKAII) isoforms in vitro and in intact cells. The acidic Akap200 polypeptide (pI ~3.8) contains an optimal N-terminal myristoylation site and a positively charged domain that resembles the multifunctional phosphorylation site domain of vertebrate myristoylated alanine-rich C kinase substrate proteins. The 15-kilobase pair Akap200 gene contains six exons and encodes a second protein, DAkap200. DAkap200 is derived from Akap200 transcripts by excision of exon 5 (381 codons), which encodes the PKAII binding region and a Pro-rich sequence. DAkap200 appears to be a myristoylated alanine-rich C kinase substrate analog. Akap200 and DAkap200 are evident in vivo at all stages of Drosophila development. Thus, both proteins may play important physiological roles throughout the life span of the organism. Nevertheless, Akap200 gene expression is regulated. Maximal levels of Akap200 are detected in the pupal phase of development; DAkap200 content is elevated 7-fold in adult head (brain) relative to other body parts. Enhancement or suppression of exon 5 excision during Akap200 pre-mRNA processing provides potential mechanisms for regulating anchoring of PKAII and targeting of cAMP signals to effector sites in cytoskeleton and/or organelles (Li, 1999).

The sequence of the largest Akap200 cDNA (3053 bp) is deposited in GenBank^TM (accession number AF132884). A predicted initiator Met codon (nucleotides 155-157) is incorporated within an optimal translation start motif (ANNATGG, nucleotides 152-158). The putative 154-bp 5'-untranslated portion of the cDNA lacks alternative ATG sequences but includes translation stop codons in all reading frames. In addition, amino acid residues 1-7 constitute an acceptor site for myristoylation, a modification that occurs at the N terminus of proteins. An open reading frame of 752 codons follows the initiator ATG and precedes a translation stop signal at nucleotides 2414-2416. The 3'-untranslated region comprises 614 bp and is followed by a polyadenylate tail. Processing of the 3'-end of Akap200 mRNA appears to be governed by either of two overlapping, atypical poly(A) addition signals (GATAAA or AATATA, nucleotides 3004-3014) that precede the polyadenylate tail by 22 or 17 nucleotides (Li, 1999).

Akap200 is composed of 753 amino acids and has a calculated M_r of 79,075. The anchor protein is exceptionally acidic (pI ~3.8) and has an atypical amino acid composition. Glu, Asp, Ser, Thr, Ala, and Pro account for 61% of the residues in Akap200 . In contrast, Cys, Met, Trp, and Tyr are included at only five positions (0.7% of total amino acids) in the polypeptide chain. Akap200 has an apparent M_r of 200,000 upon electrophoresis under denaturing conditions. High net negative charge and aberrantly reduced electrophoretic mobility are properties shared between Akap200 and mammalian AKAPs. Most AKAPs have been assigned names that correspond to their apparent M_r values. Thus, the name A kinase anchor protein of 200 kDa, or Akap200, is formulated in accord with standard nomenclature for anchor proteins (Li, 1999).

The Akap200 sequence is not highly homologous with sequences of previously studied polypeptides. However, several segments provide clues about potential targeting, tethering (RII binding), and regulatory domains. The N terminus of Akap200 includes a Gly residue adjacent to the initiator Met and a Ser-Lys dipeptide at positions 6 and 7. This sequence (MGXXXSK) constitutes an optimal target site for the ubiquitous enzyme N-myristoyl-CoA transferase. A segment of Akap200 that encompasses residues 119-148 includes a large cluster of Lys residues that create a highly basic (pI ~11.5) and positively charged (+12) domain in the midst of a protein with a predicted overall charge of -116 at pH 7. The compositions of the basic region in Akap200 and a central portion of MARCKS and MARCKS-related proteins (MacMARCKS, F52, MRP) are similar, although the degree of amino acid identity is modest (<30%) when the indicated sequences are aligned. The basic region of MARCKS proteins [named the phosphorylation site domain (PSD)] is a major target for protein kinase C-catalyzed phosphorylation in many mammalian cells. The Drosophila anchor protein also possesses Pro-rich regions (amino acids 328-332 and 468-479) that may serve as docking sites for proteins with Src homology 3 domains (Li, 1999).

DAkap200 has the hallmark features of MARCKS proteins. It is myristoylated; highly acidic; heat-stable; enriched in Ala, Pro, Ser, and Thr; deficient in aromatic and sulfur-containing amino acids; contains a basic PSD region; and behaves like an extended rod upon sedimentation; and its M_r value is overestimated by 150% upon denaturing electrophoresis. It binds Ca²⁺-calmodulin and F-actin and is phosphorylated and regulated by PKC. Moreover, the M_r of DAkap200 (38,000) closely approaches calculated molecular weight values of mammalian MARCKS proteins (30,000-33,000). Unlike the 79-kDa anchor protein (Akap200), MARCKS and DAkap200 lack a discrete domain that binds RII subunits. Together, these properties make DAkap200 the first candidate analog of MARCKS in Drosophila (Rossi, 1999).

date revised: 12 October 2002

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.