Precise regulation of the NFAT (nuclear factor of activated T cells) family of transcription factors (NFAT1-4) is essential for vertebrate development and function. In resting cells, NFAT proteins are heavily phosphorylated and reside in the cytoplasm; in cells exposed to stimuli that raise intracellular free Ca2+ levels, they are dephosphorylated by the calmodulin-dependent phosphatase calcineurin and translocate to the nucleus. NFAT dephosphorylation by calcineurin is countered by distinct NFAT kinases, among them casein kinase 1 (CK1) and glycogen synthase kinase 3 (GSK3). This study used a genome-wide RNA interference (RNAi) screen in Drosophila to identify additional regulators of the signalling pathway leading from Ca2+-calcineurin to NFAT. This screen was successful because the pathways regulating NFAT subcellular localization (Ca2+ influx, Ca2+-calmodulin-calcineurin signalling and NFAT kinases) are conserved across species, even though Ca2+-regulated NFAT proteins are not themselves represented in invertebrates. Using the screen, DYRKs (dual-specificity tyrosine-phosphorylation regulated kinases) has been identified as novel regulators of NFAT. DYRK1A and DYRK2 counter calcineurin-mediated dephosphorylation of NFAT1 by directly phosphorylating the conserved serine-proline repeat 3 (SP-3) motif of the NFAT regulatory domain, thus priming further phosphorylation of the SP-2 and serine-rich region 1 (SRR-1) motifs by GSK3 and CK1, respectively. Thus, genetic screening in Drosophila can be successfully applied to cross evolutionary boundaries and identify new regulators of a transcription factor that is expressed only in vertebrates (Gwack, 2006).
To validate the use of genome-wide RNAi screening in Drosophila to identify regulators of the Ca2+-calcineurin-NFAT signalling pathway, an NFAT-GFP (green fluorescent protein) fusion protein containing the entire regulatory domain of NFAT1 was used. This domain bears >14 phosphorylated serines, 13 of which are dephosphorylated by calcineurin. Five of the thirteen serines are located in the SRR-1 motif, which controls exposure of the nuclear localization sequence (NLS) and is a target for phosphorylation by CK1; three are located in the SP-2 motif, which can be phosphorylated by GSK3 after a priming phosphorylation by protein kinase A (PKA); and four are located in the SP-3 motif, for which a relevant kinase had yet to be identified at the time this study was initiated. The SP-2 and SP-3 motifs do not directly regulate the subcellular localization of NFAT1, but their dephosphorylation increases both the probability of NLS exposure and the affinity of NFAT for DNA. NFAT-GFP was correctly regulated in Drosophila S2R+ cells: it was phosphorylated and localized to the cytoplasm under resting conditions; it became dephosphorylated and translocated to the nucleus with appropriate kinetics in response to Ca2+ store depletion with the sarcoplasmic/endoplasmic reticulum ATPase (SERCA) inhibitor thapsigargin; and its dephosphorylation and nuclear translocation were both sensitive to the calcineurin inhibitor cyclosporin A (CsA). S2R+ cells treated with limiting amounts of thapsigargin displayed intermediate phosphorylated forms of NFAT-GFP, most likely reflecting progressive dephosphorylation of serines within individual conserved motifs of the regulatory domain. Depletion of the primary NFAT regulator, calcineurin, by RNAi in S2R+ cells inhibited thapsigargin-dependent dephosphorylation and nuclear import of NFAT-GFP. Thus, the major pathways regulating NFAT phosphorylation and subcellular localization (store-operated Ca2+ influx, calcineurin activation and NFAT phosphorylation/dephosphorylation) are conserved in Drosophila and appropriately regulate vertebrate NFAT (Gwack, 2006).
A genome-wide RNAi screen on unstimulated S2R+ cells was performed, and aberrant nuclear localization of NFAT–GFP was scored. Positive candidates obtained in the screen include (1) Na+/Ca2+ exchangers and SERCA Ca2+ ATPases, the knockdown of which would be expected to increase basal levels of intracellular free Ca2+ ([Ca2+]i); (2) the scaffold protein Homer, which has been linked to Ca2+ influx and Ca2+ homeostasis; (3) stromal interaction molecule (STIM), a recently identified regulator of store-operated Ca2+ influx, and (4) several protein kinases that control NFAT function either directly via phosphorylation or indirectly via basal [Ca2+]i levels, calcineurin activity, or other kinases. To identify kinases that directly phosphorylate the NFAT regulatory domain, Flag-tagged human homologues of selected Drosophila kinases were expressed in HEK293 cells, and anti-Flag immunoprecipitates were tested in an in vitro kinase assay for their ability to phosphorylate a GST-NFAT1(1-415) fusion protein. Three kinases -- protein kinase cGMP-dependent (PRKG1), DYRK2 and interleukin (IL)-1 receptor-associated kinase 4 (IRAK4) -- showed strong activity in this assay. In cells, only DYRK2 countered the dephosphorylation of NFAT-GFP by calcineurin, even though both PRKG1 and DYRK2 were expressed at high levels. CD4+ TH1 cells isolated from Irak4-/- mice showed normal NFAT1 dephosphorylation, re-phosphorylation and nuclear transport compared to control TH1 cells. Therefore focus was placed on DYRK-family kinases as potential direct regulators of NFAT (Gwack, 2006).
DYRKs constitute an evolutionarily conserved family of proline- or arginine-directed protein kinases belonging to the CMGC family of cyclin-dependent kinases (CDKs), mitogen-activated protein kinases (MAPKs), GSK and CDK-like kinases (CLKs). The DYRK family has multiple members that can be predominantly nuclear (DYRK1A and DYRK1B) or cytoplasmic (DYRK2-4 and homeodomain interacting protein kinase 3 (HIPK3)/DYRK6). RT-PCR and western blotting suggested that DYRK1A and DYRK2 were major representatives of nuclear and cytoplasmic DYRKs, respectively, in Jurkat T cells, a conclusion supported by several additional observations. (1) Overexpression of DYRK2 prevented dephosphorylation of NFAT-GFP after ionomycin treatment; overexpression of wild-type DYRK2, but not a kinase-dead mutant of DYRK2, prevented NFAT nuclear localization in thapsigargin-treated cells. The slower-migrating form of NFAT might indicate that DYRK2, a proline-directed kinase, acts in part by phosphorylating the SPRIEITP docking sequence on NFAT1 and thereby blocking the calcineurin-NFAT interaction. However, DYRK was still effective when the potential DYRK phosphorylation sites in the docking sequence were eliminated by substituting HPVIVITGP for SPRIEITPS. (2) Both wild-type and kinase-dead DYRK co-immunoprecipitated with NFAT1. (3) Depletion of the DYRK-family candidate CG40478 in S2R+ cells did not affect (and DYRK2 overexpression in Jurkat T cells only slightly diminished) Ca2+ mobilization in response to thapsigargin. (4) Most importantly, depletion of endogenous DYRK1A with DYRK1A-specific short interfering RNAs (siRNA) in HeLa cells stably expressing NFAT-GFP increased the rate and extent of NFAT1 dephosphorylation and nuclear import while slowing re-phosphorylation and nuclear export. These results show that DYRK1A and DYRK2 are physiological negative regulators of NFAT activation in cells. The absence of basal NFAT dephosphorylation in DYRK1A-depleted cells may reflect both the expression of other DYRK family members in human cells and the predominantly nuclear localization of DYRK1A (Gwack, 2006).
DYRKs are direct NFAT1 kinases that selectively phosphorylate the SP-3 motif, but nevertheless control the overall phosphorylation of NFAT1. Flag-tagged DYRK2 expressed in HEK cells, as well as bacterially expressed recombinant DYRK1A and DYRK2, phosphorylated peptides corresponding to the SP-3 motif of the NFAT1 regulatory domain in vitro, but did not phosphorylate SRR-1 or SP-2 peptides or an SP-3 peptide with serine to alanine substitutions in the known phosphoserine residues. Two serine residues (underlined) in the SP-3 motif (SPQRSRSPSPQPSPHVAPQDD) fit the known sequence preference of DYRK kinases [R(x)xx(S/T)(P/V)], and both are known to be phosphorylated in cells. DYRK is reported to prime for GSK3-mediated phosphorylation of eukaryotic initiation factor 2B-epsilon (eIF2B-epsilon) and the microtubule-associated protein tau, as well as for GSK3- and CK1-mediated phosphorylation of OMA-1. Therefore whether DYRK kinases could also prime for GSK3- and CK1-mediated phosphorylation of NFAT1 was investigated. Pre-phosphorylation of the NFAT1 regulatory domain by DYRK2 led to robust phosphorylation by GSK3 and induced the mobility shift characteristic of phosphorylation of the SP-2 and SP-3 motifs; this shift was not observed after pre-phosphorylation by PKA. Furthermore, pre-phosphorylation by DYRK2 accelerated CK1-mediated phosphorylation of GST-NFAT1(1-415) by at least twofold. In contrast, DYRK2 did not prime phosphorylation of the SP-2 peptide by GSK3 nor the SRR-1 peptide by CK1, consistent with the fact that neither motif is a substrate for DYRK. This 'discontiguous' priming mechanism is distinct from conventional priming, which requires phosphorylation at +4 and -3 for GSK3 and CK1, respectively. A less likely interpretation is that the conventional priming sites for CK1 and GSK3 are efficiently phosphorylated by DYRK in the context of the GST-NFAT1(1-415) protein, although they are not phosphorylated in the peptide context (Gwack, 2006).
The kinase-dead mutant of DYRK2 shows that DYRK regulates the transcriptional activity of NFAT. Wild-type DYRK2 strongly diminished NFAT-dependent activity, whereas the kinase-dead mutant increased NFAT-dependent luciferase activity of the IL2 promoter, of an NFAT-activating protein 1 (AP-1) reporter, or of an AP-1-independent promoter (the kappa3 site of the tumour-necrosis factor-alpha (TNF-alpha) promoter). Similarly, wild-type DYRK2 diminished production of endogenous IL-2 by stimulated Jurkat T cells, whereas kinase-dead DYRK2 had the opposite effect (Gwack, 2006).
These data indicate that DYRK is a key kinase that regulates NFAT1 phosphorylation. DYRK, GSK3 and CK1 target completely distinct motifs of the NFAT1 regulatory domain, but DYRK-mediated phosphorylation of the SP-3 motif primes for further phosphorylation of the distinct SRR-1 and SP-2 motifs by CK1 and GSK3, respectively, thus facilitating complete phosphorylation and deactivation of NFAT1. This mechanism, which has been termed 'discontiguous priming', is reminiscent of that recently proposed for C. elegans oocyte maturation protein 1 (OMA-1), in which phosphorylation of Thr 239 by the DYRK-family kinase minibrain kinase 2 (MBK-2) potentiates GSK3-mediated phosphorylation of Thr 339. It is likely that DYRK2, DYRK3 and DYRK4, which are localized to the cytoplasm, function primarily as 'maintenance' kinases that sustain the phosphorylation status of cytoplasmic NFAT in resting cells, whereas DYRK1A and DYRK1B, which are localized to the nucleus, re-phosphorylate nuclear NFAT and promote its nuclear export. Notably, NFAT dephosphorylation may also proceed through a sequential mechanism, with dephosphorylation of the SRR-1 motif promoting dephosphorylation of the SP-2 and SP-3 motifs by increasing their accessibility to calcineurin. DYRK1A and the endogenous calcineurin regulator DSCR1/RCN/calcipressin-1 are both localized to the Down's syndrome critical region on human chromosome 21; thus, overexpression of these negative regulators of NFAT could contribute (by inhibiting NFAT activation) to the neurological and immunological developmental anomalies observed in individuals with chromosome 21 trisomy (Gwack, 2006).
Genome-wide RNAi screening in Drosophila is a valid and powerful strategy for exploring novel aspects of signal transduction in mammalian cells, provided that key members of the signalling pathway are evolutionarily conserved and represented in the Drosophila genome. This study used the method to identify conserved regulators of the purely vertebrate transcription factor NFAT; this is the first example of a genome-wide RNAi screen that crosses evolutionary boundaries in this manner. It is likely that conserved aspects of the regulation of other mammalian processes will also be successfully defined by developing assays in Drosophila cells (Gwack, 2006).
Calcineurin is a calcium-activated protein phosphatase involved in multiple aspects of cardiac and skeletal muscle development and disease. Genes encoding calcineurin subunit proteins are highly conserved among animal species. Toward the goal of identifying new calcineurin-interacting loci that function in myogenic processes, an activated form of mouse calcineurin A was expressed in Drosophila and suppressors of the phosphatase-induced lethality were screened for. A mutation in the canB2 gene, which encodes a regulatory subunit of Drosophila calcineurin, can suppress a pupal developmental arrest phenotype to adult viability. Since canB2 is an essential gene and rare homozygous escapers are flightless, canB2 expression and function was further analyzed in pupae and adults. The gene is expressed in the forming indirect flight muscles and central nervous system during pupal development. A canA gene is comparably expressed in these tissues. Consistent with the observed muscle expression, canB2 mutants exhibit severe defects in the organization of their indirect flight muscles, a phenotype that is likely caused by muscle hypercontractility. Together, these findings demonstrate a vital role for the phosphatase in this specific facet of Drosophila myogenesis and show conserved fly and vertebrate calcineurin genes contribute prominently to fundamental processes of muscle formation and function (Gajewski, 2003).
The Drosophila modulatory calcineurin-interacting protein (MCIP) sarah (sra) is essential for meiotic progression in oocytes. Activation of mature oocytes initiates development by releasing the prior arrest of female meiosis, degrading certain maternal mRNAs while initiating the translation of others, and modifying egg coverings. In vertebrates and marine invertebrates, the fertilizing sperm triggers activation events through a rise in free calcium within the egg. In insects, egg activation occurs independently of sperm and is instead triggered by passage of the egg through the female reproductive tract; it is unknown whether calcium signaling is involved. MCIPs [also termed regulators of calcineurin (RCNs), calcipressins, or DSCR1 (Down's syndrome critical region 1)] are highly conserved regulators of calcineurin, a Ca2+/calmodulin-dependent protein phosphatase 1 and 2. Although overexpression experiments in several organisms have revealed that MCIPs inhibit calcineurin activity, their in vivo functions remain unclear. Eggs from sra null mothers are arrested at anaphase of meiosis I. This phenotype was due to loss of function of sra specifically in the female germline. Sra is physically associated with the catalytic subunit of calcineurin, and its overexpression suppresses the phenotypes caused by constitutively activated calcineurin, such as rough eye or loss of wing veins. Hyperactivation of calcineurin signaling in the germline cells resulted in a meiotic-arrest phenotype, which can also be suppressed by overexpression of Sra. All these results support the hypothesis that Sra regulates female meiosis by controlling calcineurin activity in the germline. This is the first unambiguous demonstration that the regulation of calcineurin signaling by MCIPs plays a critical role in a defined biological process (Takeo, 2006; Horner, 2006).
sarah mutation disrupts several aspects of egg activation in Drosophila. Eggs laid by sarah mutant females arrest in anaphase of meiosis I and fail to fully polyadenylate and translate bicoid mRNA. Furthermore, sarah mutant eggs show elevated cyclin B levels, indicating a failure to inactivate M-phase promoting factor (MPF). Taken together, these results demonstrate that calcium signaling is involved in Drosophila egg activation and suggest a molecular mechanism for the sarah phenotype. The conversion of the sperm nucleus into a functional male pronucleus is compromised in sarah mutant eggs, indicating that the Drosophila egg's competence to support male pronuclear maturation is acquired during activation. Despite its independence from a sperm trigger, egg activation in Drosophila involves calcium-mediated pathways that are likely to be analogous to those in other animals. It is intriguing that among these downstream events is the acquisition of the egg's competence to remodel the sperm nucleus into the male pronucleus (Horner, 2006).
To explore the in vivo function of the Drosophila MCIP Sra, a null mutation in this locus was created by gene targeting. Homozygotes of the null allele sraKO are semilethal during larval or pupal stages. In addition, sraKO females were sterile, and their ovulation is abnormal (Ejima, 2004). These phenotypes were rescued by either sarah transgenes. Taken together, these results unambiguously demonstrate that sra is responsible for the phenotypes associated with sraKO, which include developmental defects in both sexes and ovulation and sterility in females (Takeo, 2006).
There is no apparent morphological abnormality in ovarian development in sraKO females, but eggs from sraKO mothers failed to hatch. Wild-type eggs at 2 hr after deposition already have completed meiosis and undergo synchronous mitotic nuclear division. In contrast, eggs—hereafter referred to as sra eggs—from sraKO mothers had a localized DAPI-stained signal in the cortical region near the anterior pole, indicating that sra eggs are arrested during meiosis. To analyze this phenotype in detail, the pattern of chromosome segregation and the spindle shape of sra eggs were examined. Spindle microtubules were visualized with tubulin antibody staining. In wild-type females, mature oocytes are arrested at metaphase of meiosis I, during which the chromosomes are seen as a large mass of chromatin. After the release of meiotic arrest during ovulation, individual chromosome arms become visible and migrate toward the poles; the chromosomes subsequently undergo meiosis II, after which nuclear fusion and the mitotic divisions of the zygote take place. In sra eggs, the meiotic chromosomes were seen in between the metaphase plate and the poles. It was confirmed that the oocytes taken from sra mutants including sraGS3080 and sraGS3168 are arrested at metaphase I as in the wild-type. Therefore, sra eggs are arrested at anaphase I shortly after the meiotic resumption from the metaphase I arrest (Takeo, 2006).
To determine whether the meiotic-arrest phenotype of sra eggs is caused by loss of function in the germline or somatic cells of sra mothers, mutant germline clones were generated by the flippase-dominant female sterile (FLP-DFS) technique. All eggs laid by wild-type females completed meiotic divisions, whereas most eggs (98%) from sra mothers arrested at anaphase of meiosis I. sra germline clones were also arrested at anaphase I, reproducing the phenotype of sraKO. A few sra eggs, including germline clones, were arrested at meiosis II. In these eggs, the two spindles were perpendicular to each other, rather than in tandem as in the wild-type. These results demonstrate that the meiotic defects in sra eggs are specifically attributable to the loss of function of sra in the female germline. Consistent with this conclusion, sra is highly expressed in the female germline during oogenesis (Ejima, 2004) and in early embryos. Furthermore, the meiotic-arrest phenotype caused by sra mutations was almost fully rescued by nos-GAL4/UASp-sra transgenes. These results establish that sra is required in the germline for meiotic progression in Drosophila females (Takeo, 2006).
Sra is a Drosophila member of the modulatory calcineurin-interacting protein (MCIP) family of proteins, which are known to function as endogenous regulators of calcineurin. Calcineurin consists of two subunits, CnA and CnB. The Drosophila genome contains three genes encoding CnA subunits (CanA1, Pp2B-14D, and CanA-14F) and two genes encoding CnB subunits (CanB and CanB2). The functions of calcineurin have been poorly analyzed in Drosophila. It is hypothesized that sra functions as an endogenous regulator of calcineurin in Drosophila. Analyses of the expression pattern of Drosophila calcineurin genes by RT-PCR revealed that all three CnA and both CnB genes were expressed in larvae and adult females, but among these, only Pp2B-14D, CanA-14F, and CanB2 are expressed in early embryos and ovaries. Therefore, these three calcineurin subunits are candidates for interacting with Sra in the female germline (Takeo, 2006).
A constitutively active form of calcineurin can be created by truncating the C-terminal part of CnA. Misexpression of the active form of Pp2B-14D (Pp2B-14Dact) causes morphological abnormalities in eyes and wings. To determine the effects of sra on calcineurin signaling, whether activated calcineurin-dependent phenotypes can be modified by coexpression of sra was tested. Overexpression of sra alone in developing eyes by using GMR-GAL4 did not induce any phenotypic change. Flies misexpressing Pp2B-14Dact showed a mild rough-eye phenotype, which was completely suppressed by coexpression of sra. Similarly, misexpression of Pp2B-14Dact in the posterior compartment of developing imaginal discs by using en-GAL4 resulted in loss of wing veins and reduction of wing size. These wing phenotypes were also completely suppressed by coexpression of sra, whereas overexpression of sra alone had no effect on wing morphology. Furthermore, overexpression of sra rescued the lethality induced by the muscle-specific expression of Pp2B-14Dact by using 24B-GAL4. All these results clearly show that sra has an inhibitory effect on calcineurin signaling (Takeo, 2006).
If Sra acts as an inhibitor of calcineurin in vivo, it was speculated that calcineurin signaling is hyperactivated in sra mutants; that is, hyperactivation of calcineurin signaling might also affect the meiotic phenotype as in sra mutants. It was found that females carrying nos-GAL4 and UASp-Pp2B-14Dact had fully developed ovaries, but were sterile or semisterile, depending on the transgenic lines. The sterility or semisterility caused by nos>Pp2B-14Dact was effectively rescued by co-overexpression of sra, demonstrating that sra counteracts activated calcineurin (Takeo, 2006).
To characterize the meiotic phenotype caused by Pp2B-14Dact, eggs were stained from semisterile females expressing nos>Pp2B-14Dact (Ejima, 2004) with DAPI and tubulin antibody to visualize chromosomes and spindles, respectively. A total of 63 eggs were examined. Of these, 10% (6/63) developed normally, whereas 14% (9/63) had neither DAPI signaling nor tubulin antibody staining. The remaining 76% showed complex abnormalities that could be classified into three types: (1) dispersed chromatins with no obvious spindle (33%); (2) apparently normal chromosomes with an abnormal spindle (38%), and (3) a mass of chromatin with an apparently normal spindle (5%). Also the nuclei of mature oocytes taken from nos>Pp2B-14Dact (Ejima, 2004) females were observed to see whether meiotic arrest at metaphase I is normal. It was found that the majority had abnormal nuclei containing dispersed chromatins, and that the remaining were arrested at metaphase I or anaphase I. These results suggest that calcineurin signaling was activated to a greater extent in the germline of females constructed in this way than in sra mutants. Taken together, these results demonstrate that the regulation of calcineurin signaling is critical for female meiosis, and its regulator Sra/MCIP is essential for meiotic progression at the time of egg activation in the Drosophila female (Takeo, 2006).
In vertebrates whose meiotic arrest occurs at metaphase II, arrest is released at the time of fertilization. The mechanisms of meiotic arrest and resumption have been extensively studied in mice and frogs, and several key components have been identified, including Cdc2/Cyclin B (MPF) and MAP kinase. In addition, the so-called “Ca2+ transient” mediated by IP3 signaling has been linked to egg activation; this transient promotes completion of meiosis, ion-channel opening, and cortical granule exocytosis. Recent studies have revealed that Ca2+/calmoduin-dependent protein kinase II (CaMKII) is physiologically activated in mouse oocytes in response to fertilizing sperm. CaMKII is implicated in the regulation of the timing of re-entry into mitosis through the phosphorylation of Cdc25C, a phosphatase mediating G1/M transition by dephosphorylating MPF in Xenopus. More recently, CaMKII was shown to phosphorylate an anaphase-promoting complex/cyclosome (APC/C) inhibitor, Emi1-related protein (Erp1), resulting in its degradation and thereby releasing the brakes on the cell cycle from metaphase II in Xenopus eggs (Takeo, 2006).
Less is known about the mechanism of egg activation in Drosophila. Genetic screens for female-sterile mutations have identified several genes involved in female meiosis. For example, twine, a homolog of Drosophila cdc25, is required for arrest at metaphase I in mature oocytes. cortex (cort) and grauzone (grau) mutant eggs exhibit meiotic arrest at meiosis II with defects in cytoplasmic polyadenylation and translation of maternal bicoid mRNAs. cort encodes an APC/C activator protein Cdc20, suggesting that APC/C-Cdc20Cort-mediated inactivation of MPF is required for the translational control of poly(A)-dependent maternal mRNA. grau encodes a member of the C2H2-type zinc-finger protein family and activates transcription of cort to induce the completion of female meiosis. In addition, a recent study reported that a small cell-cycle regulator, Cks30A, plays an essential role in meiotic progression by associating with Cdk1 (Cdc2)/cyclin complexes and mediating Cyclin A degradation in the female germline. Therefore, cell-cycle regulators involved in meiotic progression are likely to be conserved between vertebrates and Drosophila (Takeo, 2006).
The involvement of calcineurin signaling in female meiosis has not previously been described in any organism. These studies on sra are the first to demonstrate that regulation of a Ca2+-dependent phosphatase is critical for the progression of female meiosis. Analyses of mutations in calcineurin genes and identification of the substrates in germline cells should facilitate further understanding of the role of calcineurin signaling in female meiosis (Takeo, 2006).
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.