Nucleomorphin. A novel, acidic, nuclear calmodulin-binding protein from dictyostelium that regulates nuclear number.

Probing of Dictyostelium discoideum cell extracts after SDS-PAGE using (35)S-recombinant calmodulin (CaM) as a probe has revealed approximately three-dozen Ca(2+)-dependent calmodulin binding proteins. Here, we report the molecular cloning, expression, and subcellular localization of a gene encoding a novel calmodulin-binding protein (CaMBP); we have called nucleomorphin, from D. discoideum. A lambdaZAP cDNA expression library of cells from multicellular development was screened using a recombinant calmodulin probe ((35)S-VU1-CaM). The open reading frame of 1119 nucleotides encodes a polypeptide of 340 amino acids with a calculated molecular mass of 38.7 kDa and is constitutively expressed throughout the Dictyostelium life cycle. Nucleomorphin contains a highly acidic glutamic/aspartic acid inverted repeat (DEED) with significant similarity to the conserved nucleoplasmin domain and a putative transmembrane domain in the carboxyl-terminal region. Southern blotting reveals that nucleomorphin exists as a single copy gene. Using gel overlay assays and CaM-agarose we show that bacterially expressed nucleomorphin binds to bovine CaM in a Ca(2+)-dependent manner. Amino-terminal fusion to the green fluorescence protein (GFP) showed that GFP-NumA localized to the nucleus as distinct arc-like patterns similar to heterochromatin regions. GFP-NumA lacking the acidic DEED repeat still showed arc-like accumulations at the nuclear periphery, but the number of nuclei in these cells was increased markedly compared with control cells. Cells expressing GFP-NumA lacking the transmembrane domain localized to the nuclear periphery but did not affect nuclear number or gross morphology. Nucleomorphin is the first nuclear CaMBP to be identified in Dictyostelium. Furthermore, these data present the first identification of a member of the nucleoplasmin family as a calmodulin-binding protein and suggest nucleomorphin has a role in nuclear structure in Dictyostelium.

Probing of Dictyostelium discoideum cell extracts after SDS-PAGE using 35 S-recombinant calmodulin (CaM) as a probe has revealed approximately three-dozen Ca 2؉dependent calmodulin binding proteins. Here, we report the molecular cloning, expression, and subcellular localization of a gene encoding a novel calmodulin-binding protein (CaMBP); we have called nucleomorphin, from D. discoideum. A ZAP cDNA expression library of cells from multicellular development was screened using a recombinant calmodulin probe ( 35 S-VU1-CaM). The open reading frame of 1119 nucleotides encodes a polypeptide of 340 amino acids with a calculated molecular mass of 38.7 kDa and is constitutively expressed throughout the Dictyostelium life cycle. Nucleomorphin contains a highly acidic glutamic/aspartic acid inverted repeat (DEED) with significant similarity to the conserved nucleoplasmin domain and a putative transmembrane domain in the carboxyl-terminal region. Southern blotting reveals that nucleomorphin exists as a single copy gene. Using gel overlay assays and CaM-agarose we show that bacterially expressed nucleomorphin binds to bovine CaM in a Ca 2؉ -dependent manner. Amino-terminal fusion to the green fluorescence protein (GFP) showed that GFP-NumA localized to the nucleus as distinct arc-like patterns similar to heterochromatin regions. GFP-NumA lacking the acidic DEED repeat still showed arc-like accumulations at the nuclear periphery, but the number of nuclei in these cells was increased markedly compared with control cells. Cells expressing GFP-NumA lacking the transmembrane domain localized to the nuclear periphery but did not affect nuclear number or gross morphology. Nucleomorphin is the first nuclear CaMBP to be identified in Dictyostelium. Furthermore, these data present the first identification of a member of the nucleoplasmin family as a calmodulin-binding protein and suggest nucleomorphin has a role in nuclear structure in Dictyostelium.
Calmodulin (CaM 1 ), the major, essential Ca 2ϩ -binding protein of all eukaryotes, is highly conserved such that the CaM of mammals and of eukaryotic microbes, such as Dictyostelium discoideum, differs in only a few amino acids leaving them functionally identical (1)(2)(3)(4). CaM is a small acidic protein consisting of a flexible ␣-helical tether joining two globular domains each of which contain two Ca 2ϩ -binding sites (5,6). Upon Ca 2ϩ binding, CaM undergoes a large conformational change exposing two hydrophobic patches that allow for targetprotein interaction (7). CaM binding to its targets does not operate through a conserved motif, because CaM binding regions on target proteins show little sequence homology. Many Ca 2ϩ -dependent CaMBPs have one or more CaM-binding domains characterized by a basic amphipathic helix commonly with positively charged residues interspersed among hydrophobic and aromatic residues (7). ␣-Helical wheel analysis typically shows a segregation of hydrophobic residues on one side and basic charged residues on the other (8). Ca 2ϩ -dependent CaMBPs can be grouped into two related motifs (1-8-14 and 1-5-10) based on conserved hydrophobic residues (9 -11). Proteins containing the 1-8-14 motif include calcineurin (CN), nitric-oxide synthase, adenylyl cyclase, and myosin light chain kinase, whereas those using the 1-5-10 motif include CaM kinase I (CaMKI), CaMKII, and synapsin (11)(12)(13)(14). CaMBPs that do not recognize these motifs include adenylyl cyclase and dystrophin (15,16). Many Ca 2ϩ -independent CaMBPs exist, adding to the complexity of CaM regulation and function. CaM also interacts in a Ca 2ϩ -independent manner through an IQ motif, as has been shown for conventional type II myosin light chains and the unconventional myosins, neuromodulin and neurogranin (17)(18)(19). Because CaM's target proteins (CaMBPs) do the work, a full understanding of their structure, function, and regulation is essential, but, as yet, the full complement of CaMBPs has not been determined for any cell type.
Dictyostelium has long been used as a model organism for the study of the molecular biology of cell function and differentiation in eukaryotic cells (20). Dictyostelium is haploid and contains one CaM gene, and null mutants are lethal (1). Screening of D. discoideum cell extracts after SDS-PAGE using a recombinant CaM probe ( 35 S-VU1-CaM) reveals Ͻ12 Ca 2ϩ -independent, ϳ3 dozen Ca 2ϩ -dependent CaMBPs plus a sole Ca 2ϩ -inhibited CaMBP (21)(22)(23)(24). CaM and certain CaMBPs have been linked to specific events during starvation, asexual development, chemotaxis, gametogenesis, fertilization, and spore germination in D. discoideum (21)(22)(23)(24)(25)(26). Despite their number and essential roles in a number of events, only a few CaMBPs have been characterized fully in Dictyostelium. Myosin heavy chain kinase, a Ca 2ϩ /CaM-dependent enzyme involved in myosin assembly, has been shown to be integral in cell motility (27,28). Myosin I isoforms are Ca 2ϩ -independent CaM targets (17). Knockout mutants for ␣-actinin, a Ca 2ϩ /CaM-dependent CaMBP that binds actin, show defects in motility and orientation (29). Knockout mutants for Dictyostelium IQGAP, a Ca 2ϩindependent CaMBP, are unable to complete cytokinesis (30). Dictyostelium CN has been extensively characterized (31)(32)(33). FK506 and cyclosporin A do not inhibit growth or aggregation but do affect cell differentiation, yet the regulation and roles of Dictyostelium CN are still under analysis (34,35). In addition to acting as CaM's effectors, CaMBPs bind different targets based upon cytosolic Ca 2ϩ levels; they localize CaMs to subcellular locales and act as capacitators in CaM function (36). Until all of the CaMBPs have been revealed and their functions elucidated, any model of cellular processes involving Ca 2ϩ and CaM will remain incomplete.
Our approach has been to screen a ZAP cDNA expression library of cAMP-chemoresponsive cells from multicellular development using a recombinant CaM probe ( 35 S-VU1-CaM) to isolate cDNAs encoding putative CaMBPs. This approach was validated when the first cDNA that we isolated and sequenced encoded calcineurin A (GenBank TM accession number U22397), the sole Ca 2ϩ /CaM-dependent protein phosphatase of eukaryotes. Similar approaches have been used to identify novel CaMBPs in plants, including maize pollen calmodulin-binding protein, and a tobacco plasma membrane calmodulin-binding channel protein (37,38). Here we report on the isolation of a 1119-bp cDNA encoding a novel, acidic, nuclear protein of 340 amino acids with a predicted molecular mass of 38.7 kDa that contains several nuclear localization sequences (NLS) plus an extensive, continuous 52-amino acid inverted repeat of glutamic and aspartic acid residues (DEED repeat). This region is characteristic of the conserved nucleoplasmin domain suggesting it is a member of the nucleoplasmin superfamily of nuclear proteins. We have named this gene numA for nucleomorphin. Southern blotting and PCR indicate that a single gene exists in the genome of D. discoideum. Binding assays using truncated fusion proteins show nucleomorphin to be a Ca 2ϩ -dependent CaM target protein that contains at least two CaM-binding domains. Nucleomorphin mRNA and protein are each expressed continuously throughout development. GFP-NumA localized specifically to several arc-like, heterochromatic regions along the periphery of the nucleus. GFP-NumA lacking the putative transmembrane domain and the DEED repeat still accumulated at the periphery of the nucleus; however, removal of the acidic region leads to an increase in multinuclearity. Together these data indicate that nucleomorphin is a nuclear calmodulin-binding protein and is related to nucleoplasmin, which may be involved in nuclear structure in Dictyostelium.

EXPERIMENTAL PROCEDURES
Materials-All restriction enzymes used were purchased from Amersham Biosciences, Inc. T4 DNA ligase and T4 polynucleotide kinase were purchased from New England BioLabs. All chemicals, including media reagents, were purchased from BioShop Canada, VWR Scientific Products, or Sigma-Aldrich. TRIzol RNA isolation reagent and low DNA mass ladders were purchased from Invitrogen. Expand High Fidelity PCR kit, High Pure PCR Template purification kit, DIG DNA Labeling and Detection kit, DIG Easy-Hyb, nylon membranes, PCR molecular weight markers, and DIG-labeled DNA and RNA molecular weight markers were from Roche Molecular Biochemicals. PCR primers were purchased from Sigma-Genosys. Qiaex II Gel Extraction system was purchased from Qiagen. The pET21-b(ϩ) vector was obtained from Novagen.
Screening of the ZAP cDNA Expression Library Using 35 S-VU1-CaM-Titration of bacteriophage and screening of the cDNA library were essentially carried out as described previously (39). Variations in protocol were implemented with respect to the use of 35 S-radiolabeled recombinant CaM as a probe. An overnight culture of an XL1-Blue strain of Escherichia coli in LB media containing 0.2% maltose was pelleted and resuspended in 10 mM MgSO 4 . Serial dilutions of bacteriophage ranging from 10 Ϫ4 to 10 Ϫ6 were prepared in SM buffer (100 mM NaCl, 20 mM MgSO 4 ⅐7H 2 O, 50 mM Tris-HCl, pH 7.5, and 0.01% gelatin). Diluted phage was mixed 1:1 with XL1-Blue E. coli and incubated at 37°C. Three milliliters of liquid LB top agarose was added and spread evenly over a 90-mm LB agar plate then incubated at 42°C for 2 h. Nitrocellulose membranes were soaked in 10 mM isopropyl-␤-D-thiogalactopyranoside (IPTG), dried, and placed over the surface of the agar plates, and incubated at 37°C overnight. Membranes were removed from the plates, dried, and then washed in CaM-probing buffer (50 mM Tris-HCl, pH 7.0, 200 mM KCl, 1 mM CaCl 2 , and 0.05% Tween 20). Membranes were blocked for 2-3 h in CaM-probing buffer with 3% BSA and 0.05% sodium azide added, then probed at room temperature for 3 h with 3 ml of recombinant 35 S-VU1-CaM (3.6 ϫ 10 5 to 9.0 ϫ 10 5 cpm/ml) dissolved in CaM-probing buffer supplemented with 3% BSA and 0.05% sodium azide. The production of 35 S-VU1-CaM is detailed elsewhere (40). The membranes were then washed with CaM-probing buffer four consecutive times for a period of 10 min each. Membranes were dried and exposed to autoradiography film (Kodak X-Omat AR Scientific Imaging film) for 1-2 days. Single plaques corresponding to positive signals on the autoradiography film were located and subjected to three successive screens before being referred to as cDNA inserts coding for putative calcium-dependent calmodulin-binding proteins. Excision of the pBluescript phagemid and infection of host bacterial cells was conducted according to the protocol provided with the ExAssist Interference-Resistant Helper Phage kit purchased from Stratagene. Plasmids were isolated from positive clones by way of alkali lysis (39) and subjected to restriction analysis using StyI, HindIII, BstEII, and PstI. Clones with dissimilar restriction enzyme fragments were sequenced at the Core Molecular Biology Facility, York University. Sequences were compared with those in the GenBank TM Data base using BLAST (41). Protein motifs were identified, and the predicted molecular weight and isoelectric point were determined from the deduced amino acid sequences using the programs PSORT, version 6.4 (42), Prosite, and the ExPASy Molecular Biology Server (Swiss Institute of Bioinformatics).
Cell Lines and Cultures-D. discoideum AX3 was used as the wild type strain. Cells were cultured axenically in HL-5 at 22°C. For development, spores were mixed with a suspension of E. coli B/r and spread onto SM plates incubated at room temperature for 40 h. Plates were flooded with lower pad solution (LPS), and the cells were transferred to 50-ml centrifuge tubes. Dictyostelium amoebae were washed free of bacteria with fresh LPS by repetitive centrifugation. Cells were suspended at a density of 1 ϫ 10 8 cells/ml in LPS, deposited evenly onto the surface of Millipore filters over LPS-saturated pads in 60-mm plastic Petri dishes and incubated in a humidity chamber.
Northern and Southern Blot Analyses-For the genomic Southern analysis, Dictyostelium nuclear DNA was isolated. Approximately 5 ϫ 10 7 cells were harvested by centrifugation. The cells were resuspended in 1 ml of Nuclei buffer, pH 7.6 (20 mM Tris-HCl, pH 7.4, 5 mM MgOAc, 0.5 mM EDTA, and 5% sucrose). A volume of 0.2 ml of 20% Triton X-100 was added, mixed, and incubated on ice for 5 min. Nuclei were pelleted by centrifugation at 5000 rpm for 5 min at 4°C and resuspended in 0.3 ml of Protease K buffer, pH 7.5 (100 mM Tris-HCl, pH 7.4, 5 mM EDTA, pH 8.0, 0.1 mg/ml proteinase K, and 1% SDS) and incubated for 2 h at 65°C. Samples are extracted once with an equal volume of phenolchloroform and again with chloroform-isoamyl alcohol. Sodium acetate was added to 0.3 M, and the DNA was precipitated with 2.5 volumes of ice-cold 100% ethanol. The DNA was resuspended in 1ϫ TE supplemented with RNase A. DNA was digested using AseI, BglII, HindIII, and EcoRI, and 1 g was electrophoresed on a 0.7% agarose gel and transferred to a nylon membrane (39). DIG-labeled nucleomorphin cDNA was used as a probe. For Northern analysis, total RNA was isolated from both vegetative amoebas, and cells were developed for the indicated times. 30 g of total RNA was size-fractionated on a 1% agarose gel containing formaldehyde and transferred to nylon membrane by capillary blotting (39). Both Southern and Northern hybridization and detections were carried out using the DIG DNA labeling and detection kit according to the manufacturer's recommendation.
Construction of pMAL and pTX-GFP Expression Vectors-PCR primers were designed to amplify the entire nucleomorphin open reading frame coding for amino acids 1 through 340. EcoRI and HindIII sites were added to the 5Ј-and 3Ј-ends, respectively, using the primers 38F-ER1 (5Ј-TGAATTCATGGATGTTCATTTAACATCAT-3Ј) and 38R-H3 (5Ј-TAATTTAAGCTTTTAATTTGAGGGTA-3Ј). Using the Expand High Fidelity PCR kit from Roche Molecular Biochemicals, PCR was performed in a GeneAmp PCR System 9600 (PerkinElmer Life Sciences) for 25 cycles (each consisting of a denaturation at 94°C for 2 min, annealing at 50°C for 2 min, and extension at 72°C for 1 min). A single 1035-bp fragment was amplified. Prior to sub-cloning into pUC19, which had previously been digested with SmaI, the 5Ј-ends were phosphorylated using T4 polynucleotide kinase and then ligated to pUC19 using T4 DNA ligase and transformed into E. coli strain DH5␣. The recombinant plasmid was purified and was designated pUC38. Digestion of pUC38 with EcoRI and HindIII liberated the 1035-bp fragment, which was agarose gel-purified using the Qiagen QIAEX II gel purification kit and ligated into pMALc-2 digested with EcoRI and HindIII and transformed into E. coli strain TB1. The recombinant plasmid was purified, sequenced, and designated pMAL-NumA.
NumA⌬C36 -PCR primers were designed to amplify the region encoding amino acids 1 through 304. EcoRI and HindIII sites were added to the 5Ј-and 3Ј-ends, respectively, using the primers 38F-ER1 and 38TM-H3 (5Ј-AAGCTTACTTTAACTAGATTTGAGTGA-3Ј). PCR was performed as outlined above generating a single 926-bp fragment that was sub-cloned into pUC19 and pMAL as described, sequenced and named pMAL-NumA⌬C36.
NumA⌬C160 -PCR primers were designed to amplify the region encoding amino acids 1 through 180. EcoRI and HindIII sites were added to the 5Ј-and 3Ј-ends, respectively, using the primers 38F-ER1 and 38-CtBR (5Ј-AAGCTTTTACTTTCCTTTAATAAATCTTG-3Ј). PCR was performed as outlined above generating a single 551-bp fragment and was subcloned into pUC19 and pMAL as described, sequenced, and named pMAL-NumA⌬C160.
NumA⌬C218 -PCR primers were designed to amplify the region encoding amino acids 1 through 122. EcoRI and HindIII sites were added to the 5Ј-and 3Ј-ends, respectively, using primers 38F-ER1 and 38-Nt (5Ј-AAGCTTCTATTAATCATCATATTTATAC-3Ј). PCR was performed as outlined above producing a single fragment of 380 bp that was subcloned into pUC19 and pMAL as described, sequenced, and named pMAL-NumA⌬C218.
NumA⌬N165-PCR primers were designed to amplify the region encoding amino acids 166 through 304. EcoRI and HindIII sites were added to the 5Ј-and 3Ј-ends, respectively, using the primers 38-Ct (5Ј-GAATTCGATTTTGATAGTGATG-3Ј) and 38TM-H3. PCR was performed and generated a single 437-bp fragment that was sub-cloned into pUC19 and pMAL as described, sequenced, and named pMAL-NumA⌬N165.
GFP-NumA-PCR primers were designed to amplify the entire nucleomorphin open reading frame coding for amino acids 1 through 340. To facilitate cloning into pTX-GFP, a SacI site was added to the forward primer (5Ј-AGAGCTCATGGATGTTCATTTAACATCATCGAC-ATC-3Ј), and an XhoI site (5Ј-ACTCGAGTTAATTTGAGGGTAAAGAC-ATAAAAAATGCACT-3Ј) to the reverse primer. PCR was performed and generated a single 1043-bp fragment that was sub-cloned into pUC19 and pTX-GFP as described, sequenced, and named pGFP-NumA.
GFP-NumA⌬C36 -Removal of residues 305-340 is described above. PCR was performed as outlined for pGFP-NumA with the exception of the reverse primer. This primer incorporates an XhoI site to (5Ј-ACTC-GAGACTTTAACTAGATTTGAGTGA-3Ј) facilitate cloning into pTX-GFP. The 926-bp fragment was cloned into pTX-GFP, sequenced, and named pGFP-NumA⌬C36. ⌬N and ⌬C indicate amino-and carboxylterminal mutants, and the numbers specify the number of amino acids deleted from the respective end.
PCR-mediated Site-directed Mutagenesis of Nucleomorphin-To express nucleomorphin maximally we used the PCR to remove 46 amino acids between residues 118 and 167. Using the same forward primer for the GFP construct and a mutagenic reverse primer designated mut38N (5Ј-ATCATCCCATGGATACTCGTTTAATTTGATTGATCT-3Ј) with a restriction site for NcoI, we amplified a 377-bp fragment, treated the ends with calf intestinal alkaline phosphatase (CIAP) (New England BioLabs) and digested it with the enzyme NcoI. A second primer pair consisted of the reverse GFP primer and a mutagenic forward primer designated mut38C (5Ј-GATGAACCATGGTTTGATAGTGATGAA-GATGTTTCA-3Ј) that contained a restriction site for NcoI. The PCR amplified a 548-bp fragment that was treated with CIAP and digested with the enzyme NcoI. The digested PCR products were purified using the High Pure PCR purification kit, and the fragments were ligated using T4 DNA ligase. The PCR was then used to amplify the desired ligation product using the forward and reverse primers described above in the construction of pGFP-NumA. A single 905-bp fragment was amplified, subcloned into pUC19 and pET21-b(ϩ), and named pETNumA⌬118 -167.
Protein Purification and Production of Antibodies-E. coli strain TB1 expressed only pMAL-NumA⌬N165 and pMAL-NumA⌬C218 to any detectable levels, and both were purified by maltose affinity chromatography as described by the manufacturer. Each fusion protein was of the expected size viewed on SDS-PAGE (39). Four rabbits were each initially immunized with 750 g of purified recombinant MBP-NumA⌬C218. Boosts of 400 g of purified protein were performed at 3-week intervals, and two rabbits were exsanguinated 11 days after the third boost. Antibodies were affinity purified before use as described previously (43).
Immunodetection of Nucleomorphin through Development-Developmental cells were harvested from filters at the various developmental time points indicated, washed in 1ϫ phosphate-buffered saline, solubilized with lysis buffer (20 mM Tris-HCl, pH 6.8, 1% SDS, 1 mM EDTA, and 1 mM phenylmethylsulfonyl fluoride), and heated to 100°C for 5 min. Total protein content was measured spectrophotometrically using the Bio-Rad protein dye assay. 20 g of protein per lane was resolved using 10% SDS-PAGE, transferred onto a PVDF membrane, and blocked overnight in 5% nonfat dry milk in TTBS (20 mM Tris-HCl, 150 mM NaCl, 0.1% Tween 20, and 0.05% sodium azide). Membranes were incubated with purified primary antibodies directed against NumA⌬C218 at a dilution of 1:100 at 37°C for 1 h. Blots were washed with agitation for 90 min, changing the wash buffer every 10 min. Bound antibody was detected with the ECL Western blotting kit (Amersham Biosciences, Inc.) using HRP-conjugated anti-rabbit-IgG exposed to autoradiography film (1:20,000) (Kodak, USA).
Calmodulin Binding to Fusion Proteins-Purified fusion proteins MBP-NumA⌬C218 and MBP-NumA⌬N165 were resolved using 10% SDS-PAGE. After electrophoresis, the gels were fixed for 30 min in methanol/acetic acid (40%/7%) followed by rinsing with distilled water. Gels were washed overnight in 10% ethanol. Gels were rinsed in 10% ethanol, followed by 100 mM imidazole (pH 7.0), for 30 min; 20 mM imidazole (pH 7.0), for 15 min; and a final incubation with probe buffer (200 mM KCl, 0.1% BSA, and 1 mM CaCl 2 ) for 15 min. For Ca 2ϩindependent assays, 2 mM EGTA is substituted for the 1 mM CaCl 2 in the probe buffer. Gels were incubated overnight at 4°C with 1 Ci/ml 35 S-VU1-CaM dissolved in the appropriate probe buffer (40). Gels were then washed three to four times for 2 h each in probe buffer and then fixed overnight in methanol/acetic acid (40%/7%). Gels were then stained with Coomassie Blue, destained, dried onto filter paper, and exposed to Beta-Max autoradiography film (Amersham Biosciences, Inc.) for about 2 weeks before developing.
Pull-down Assay with CaM-Agarose Beads-Interaction studies between calmodulin and pETNumA⌬118 -167 were performed essentially as described with minor modifications (44). Ten micrograms of partially purified recombinant pETNumA⌬118 -167 in 100 l of CaM-binding buffer (20 mM Tris-HCl, pH 7.6, 100 mM KCl, 0.1 mM dithiothreitol, 2 mM Ca 2ϩ ) was incubated for 1 h at 4°C on a rotator with 50 l of CaM-agarose (Sigma) pre-equilibrated in the same buffer. To assess Ca 2ϩ -independent binding, Ca 2ϩ was replaced with 5 mM EGTA throughout the experiment. The supernatant was removed after brief centrifugation, and the resin was washed five times with 500 l of binding buffer. Bound protein was eluted by the addition of 20 l of sample buffer containing 5 mM EGTA and boiling. CaM-protein interaction was detected by SDS-PAGE or immunoblotting.
Transformation of Dictyostelium with pGFP-NumA, pGFP-NumA⌬C36, and pGFP-NumA⌬118 -167-Cells were transformed by electroporation essentially as described elsewhere with minor modifications (45). Cells growing in HL-5 were maintained in log phase for a period of 3 days and were routinely monitored to ensure they were of uniform size and healthy. Approximately 5 ϫ 10 6 cells were collected by centrifugation. Cells were washed three times with ice-cold H-50 buffer, pH 7.0 (20 mM HEPES, 50 mM KCl, 1 mM MgSO 4 , 5 mM NaHCO 3 , 10 mM NaCl, 1 mM NaH 2 PO 4 ) and resuspended in 100 l of H-50, and 5 g of plasmid DNA was added. Cells were incubated on ice for 5 min and then transferred to cold 0.4-cm gap cuvettes. Cells were electroporated with two consecutive pulses of 0.85 kV and a capacitance of 3 microfarads with 5 s between pulses. Cells were transferred to 10 ml of fresh HL-5 and allowed to recover for 16 -20 h. The medium was gently aspirated off and replaced with HL-5 containing 10 g/ml G418. The medium was replaced every 3 days for a period of 10 days. Colonies could be observed by the third day of selection. Single colonies were aspirated using sterile Pasteur pipettes into 2 ml of HL-5 containing 10 g/ml G418 and grown axenically to saturation as described above.
Imaging of GFP-NumA Clones-To record the distribution of GFP-NumA in living cells, cells were grown to a density of 3 ϫ 10 6 cells/ml, washed in low ionic strength DB buffer, resuspended at a density of 1 ϫ 10 7 cells/ml, and applied to cover slips. Cells were allowed to adhere to the cover slips for a period of 1 h, which also served as a period of starvation to allow cells to digest endocytosed nutrient medium that is autofluorescent. To observe nuclei, cells were stained with Hoechst 33258. Phase contrast and fluorescent images were taken using a Nikon Optiphot-2 epifluorescence microscope with a 100ϫ Neofluar oil immersion objective and a differential interference contrast filter. For exciting GFP, the microscope was equipped with a UV excitation immunofluorescence fluorescein isothiocyanate filter set with a dichroic mirror (DM-400, Nikon) and barrier filter (BA-435, Nikon). Images were captured using a Nikon FX-35DX camera attached to a Microflex electronic shutter (UFX-DX, Nikon). At least 400 cells from each experiment were examined, and the number of nuclei per cell was counted.

Isolation of a Novel Dictyostelium CaM-binding Protein,
Nucleomorphin-We have used a protein-protein interaction screening system to isolate cDNAs encoding putative Ca 2ϩ -dependent CaM-binding proteins from cAMP-chemoresponsive cells of D. discoideum. A ZAP cDNA expression library constructed from cells 10 -16 h through development was screened using recombinant 35 S-VU1-CaM. Single plaques corresponding to positive signals on autoradiographic film were subjected to three successive screens in which the amount of 35 S-VU1-CaM was reduced by a factor of 10, respectively (data not shown). After the high stringency screening, positive signals were deemed putative Ca 2ϩ -dependent CaMBPs. Of those recombinants detected, two clones were isolated that coded for nucleomorphin, all of which bound CaM in a Ca 2ϩ -dependent manner. Restriction enzyme analysis of each cDNA revealed identical banding patterns for each, suggesting they were derived from the same clone. Sequencing of the two cDNAs revealed that they were identical.
Sequence and Structural Features of Dictyostelium Nucleomorphin-Nucleomorphin is encoded by a cDNA of 1119 bp. The nucleotide sequence and the deduced amino acid sequence of this gene are shown in Fig. 1. The cDNA contains a 49-bp extension at the 5Ј-end, and the sequence upstream of the ATG start codon (GTAATGG) differs slightly from a consensus translation initiation sequence (AAAATGG) of D. discoideum (46). The termination codon (UAA) occurs at 1070 bp. The A/T-rich Dictyostelium genome has resulted in an extreme codon bias and thus Dictyostelium genes almost exclusively use UAA as their termination codon (47,48). Translation of the sequence predicts a single open reading frame encoding a protein of 340 amino acids with a calculated molecular mass of 38.7 kDa and an isoelectric point of 4.83. Fig. 1 shows the amino acid sequence of nucleomorphin and highlights interesting domains contained within. One feature is the somewhat long stretches containing serine (amino acids 6 -16 and 295-306). A stretch of serine, threonine, and asparagine is commonly seen in many Dictyostelium proteins but is of unknown function (49). Blast analysis of nucleomorphin did not reveal any significant matches to known protein homologs. Within nucleomorphin resides an extensive, highly acidic domain (amino acids 121-172) that is made up of 35% aspartic acid and 54% glutamic acid residues, which we call the DEED repeat. This region has an interesting inverted repeat of which the significance, if any, remains unclear. A PSI-and PHI-BLAST search of this domain produced significant similarities to Pfam nucleoplasmin proteins from a variety of species including mitotic apparatus protein p62 from sea urchin (50), nucleoplasmin-like protein (CRP-1) from Drosophila (51), and nucleophosmin from chicken (52). An alignment of the conserved domain using ClustalW shows the presence of high percentage similarities (Fig. 2). Based on further analysis with PSORT, version 6.4 (42), nucleomorphin contains a putative transmembrane region (residues 319 -335) typical of an Nt (N tail) membrane protein, because it does not contain a detectable cleavage signal sequence and the region exists in the carboxyl terminus (53). Discrimination of nuclear localization signals (NLS) reveals the presence of four possible regions. At position 61 in the amino acid sequence exists a classical NLS (residues RPRK) with two more found at positions 31 and 246 (residues PKSKKKF and PTKKRSL), respectively. The fourth NLS is consistent with that of the bipartite classification, found at position 48 in the amino acid sequence (residues KKSYQDPEIIAHSRPRK). There are also four conserved dileucine residues within nucleomorphin positioned at residues 23, 44, 181, and 234. If nucleomorphin undergoes post-translational modifications at any of the potential sites listed in Fig. 1, it may explain why this protein whose predicted molecular mass of 38.7 kDa migrates on SDS-PAGE gels with an apparent molecular mass of 43 kDa. Although, one cannot rule out the possibility that the highly acidic nature of nucleomorphin results in anomalous migration rates through SDS-PAGE due to its poor ability to bind SDS. This has been demonstrated directly by fusion protein studies using cloned centromere protein B cDNAs (54). The sequences believed responsible for high affinity CaM binding, nucleomorphin (81-94 and 172-194), are presented in Fig. 3. Within the amino terminus of nucleomorphin is a predicted CaM-binding domain (residues 81-94) of the 1-14 motif; the carboxyl terminus contains two more potential domains, a 1-14 motif is present (residues 182-194) and a putative domain that contains characteristics of the 1-10 motif (residues 172-185). Nucleomorphin Is a Ca 2ϩ -dependent Calmodulin-binding Protein-The full-length cDNA was cloned in-frame into both prokaryotic expression vectors pMALc2 and pET21-b(ϩ). Attempts at overexpression of nucleomorphin as a fusion protein for purification purposes using either system was unsuccessful. The fusion protein from IPTG-induced cell cultures could not be detected by Coomassie Blue staining of SDS-PAGE gels and was only detected using serum antibodies against MBP or T7 tag (data not shown). We hypothesized that the 52-amino acid stretch (residues 121-172) of repeating Asp/Glu was affecting the translation of nucleomorphin in E. coli. To test this, we constructed a number of truncated MBP-nucleomorphin fusion proteins: those containing the DEED repeat (pMAL-NumA⌬C36 and pMAL-NumA⌬C160) and those that lacked the DEED repeat (pMAL-NumA⌬C218, pMAL-NumA⌬N165, and pETNumA⌬118-167) (Fig. 4A). Again, expression of those constructs containing the DEED repeat could only be detected using serum antibodies against either MBP or the T7 tag (data not shown). Expression of constructs lacking the DEED repeat was successful and allowed for the purification of each fusion protein as described (Fig. 4B). Interestingly, the pETNumA⌬118-167 product was detected only in the insoluble inclusion bodies. Each construct was used to begin mapping the location of the CaM-binding domain. Overlay analysis with 35 S-CaM showed that at least two regions of CaM binding exist. The maltose-binding protein does not bind CaM, thus binding must reside within the nucleomorphin sequence. Fig. 5 presents the predicted CaM binding domains as helical wheels. Each domain clearly shows a segregation of basic charged residues to one side of the helix and hydrophobic residues on the other. Binding of CaM to either pMAL-NumA⌬C218 or pMAL-NumA⌬N165 was found to be Ca 2ϩdependent, because no binding to CaM was observed in the presence of EGTA (Fig. 5B). A second pull-down assay using CaMagarose also verified the Ca 2ϩ -dependent nature of CaM binding to pETNumA⌬118-167 (Fig. 5C). We have begun further deletion mapping of nucleomorphin to identify the exact sequences responsible for CaM interaction.
Expression and Immunodetection of Nucleomorphin during Development-Nucleomorphin expression was assessed by Northern blots of total RNA isolated from cells at 4-h intervals during filter development. A single transcript of about 1.4 kb was detected and found to be expressed constitutively throughout development (Fig. 6A). To determine the native size of the protein, total protein was also extracted at 4-h intervals during filter development, separated by SDS-PAGE, and blotted onto PVDF membranes, and nucleomorphin was detected using affinity-purified nucleomorphin antibody. As shown in Fig. 6B, nucleomorphin migrates with an apparent molecular mass of 43 kDa in SDS-PAGE and is present throughout development. The protein levels slightly increase during the first 4 h and then begin to decrease through development (Fig. 6B). These results may indicate an essential role for nucleomorphin in the life cycle of Dictyostelium. Genomic Southern blot analysis was performed using the DIG-labeled cDNA probe after digestion of genomic DNA with the restriction enzymes AseI, BglII, Hin-dIII, and EcoRI (Fig. 7). The BglII digest resulted in the detec-tion of two hybridizing bands as expected, because a single site for this restriction enzyme occurs in the middle of the cDNA and only one band was detected in the remaining lanes. This indicates that a single gene encoding nucleomorphin resides in the genome of D. discoideum. The PCR was used with primers designed from the cDNA to amplify the entire open reading frame using genomic DNA as the template. A single product was amplified identical in size to the cloned cDNA suggesting the gene for nucleomorphin is uninterrupted by introns (data not shown). DNA sequencing confirmed this and also verified the integrity of the cDNA encoding nucleomorphin.
Subcellular Localization of GFP-NumA in Vegetative Amoebae-We have used a GFP tag to study the localization of nucleomorphin in vivo. Because the carboxyl terminus contains a predicted single-pass transmembrane domain that may provide structural elements necessary for membrane associations, we decided to fuse GFP to the amino-terminal end of nucleomorphin. The intensity of GFP-NumA fluorescence in all of our transformants was moderate compared with cells transfected with GFP alone (Fig. 8A). To ensure that the observed fluorescence is due to the intact fusion protein, Western blots were performed on total protein extracts using a monoclonal anti-GFP antibody (Sigma). A band was detected with a molecular mass of ϳ70 kDa corresponding to the predicted mass of the fusion protein (data not shown). However, a band of ϳ31 kDa was also observed representing the GFP, which may explain the presence of slight background intracellular fluorescence. Additionally, we also observed that fluorescence levels varied broadly from cell to cell, a common problem related to the actin15-promoter used to drive expression of the GFP fusion protein (55,56). In vegetative cells, GFP-NumA was almost exclusively within the nucleus appearing as distinct arc-like bands that corresponded to heterochromatin-like domains adjacent to the nuclear membrane (57) (Fig. 8B). Cells were treated with Hoechst 33258, a specific stain for AT-rich regions of double-stranded DNA to verify the nuclear localization of the GFP constructs. Fig. 8C shows GFP-NumA is localized within nuclei stained with Hoechst. . For Ca 2ϩ -independent assays, 2 mM EGTA was substituted for the 1 mM CaCl 2 in the probe buffer. After washing and fixing, bound CaM was detected through exposure to autoradiography film. C, CaM-agarose pull-down assay of pETNumA⌬118 -167. The assay was performed as described under "Experimental Procedures." Insoluble extracts from cells expressing pETNumA⌬118 -167 were incubated with CaM-agarose, washed free of non-bound protein, eluted by boiling in sample buffer, and resolved using 10% SDS-PAGE. Gels were stained with Coomassie Blue, destained, and dried. Lane 1, insoluble protein fraction; lane 2, eluted protein from CaM-agarose. Molecular mass of the fusion proteins in kilodaltons is shown. The Effects of Expression of GFP-NumA Constructs Lacking the Tm or DEED Repeat-Cells expressing GFP-NumA⌬C36 lacking the putative transmembrane domain or the acidic DEED repeat also retain their localization to heterochromatinlike domains at the periphery of Hoechst-stained nuclei (Fig.  9). However, those cells expressing the GFP-NumA⌬118 -167 construct consistently displayed a dramatic increase in multinuclearity with as many as 16 nuclei in one cell (Fig. 9A). In contrast, wild type AX3 cells and the other GFP-NumA cell lines typically had one or two nuclei (Figs. 8 and 9C). It must be noted that pTX-GFP is an extrachromosomal vector, and integration into the genome does not occur. Therefore, the phenotypes observed are a result of overexpression of GFP-NumA⌬118 -167 in the presence of the wild type nucleomorphin. The number of nuclei per cell was counted for each GFP transformant (Fig. 10). Cells expressing GFP-NumA⌬118 -167 were rarely mononucleate compared with GFP-NumA and GFP-NumA⌬C36 transformants with 28.9% of the cells having four or more nuclei. None of the constructs seemed to alter the size of the nuclei except in the case of cells with extremely high numbers of nuclei. To this end, we are beginning detailed analyses of the nuclear and cytoplasmic volumes during growth and development. DISCUSSION To understand the roles of calmodulin, the major eukaryotic Ca 2ϩ sensor protein in Dictyostelium, we screened a ZAP cDNA expression library with 35 S-radiolabeled CaM to isolate cDNAs encoding some of its target proteins. Single plaques corresponding to positive signals identified on autoradiography film were selected resulting in the isolation of a cDNA encoding a novel CaM-binding protein. Using different approaches we have shown that the protein product, which we have called nucleomorphin, encoded by the cDNA binds to CaM in a Ca 2ϩdependent manner. First, the cDNA was identified through plaque screening where binding to CaM was detected in the presence of Ca 2ϩ , but not EGTA (data not shown). Second, in experiments using truncated forms of nucleomorphin fused to MBP, nucleomorphin bound to 35 S-CaM in a Ca 2ϩ -dependent manner using an SDS-gel blot overlay assay (Fig. 5). The CaM binding intensities observed on the gel overlay are very similar suggesting CaM binds each fusion with comparable affinity. Proteins, including dystrophin have been shown to contain multiple CaM binding sequences found in both the amino and carboxyl termini (16). Specifically, we have looked for amino acid sequences, ϳ20 -24 residues in length, that match the described structural features of known CaM-binding proteins, including helical wheel analysis, conserved hydrophobic residues, propensity to form an ␣-helix, and net charge (8,11). We have identified three potential CaM-binding domains in nucleomorphin based upon these criteria (Fig. 3). These regions of sequence show characteristics typical of other calmodulinbinding sequences in that they are predominantly positively charged and contain an abundance of hydrophobic residues with aromatic residues interspersed. The calmodulin-binding sequences of CaMKI, CaMKII, utrophin, caldesmon, smooth muscle myosin light chain kinase, myristoylated alanine-rich c kinase substrate protein, synapsin, and myosin all show these cationic-hydrophobic chemical characteristics (3,7). Within the amino terminus of nucleomorphin is a predicted CaM-binding domain (residues 81-94) of the 1-14 motif that was initially described (11). The main feature of this domain is the spacing of 12 amino acids between two bulky hydrophobic residues. Some CaMBPs of this type also contain conserved hydrophobic residues at positions 8 and/or 5 within the spacer region, although it is still unclear at this time whether the presence of these residues affects the binding affinities of CaM. Within the carboxyl terminus of nucleomorphin, a second potential 1-14 motif is present (residues 182-194). A third putative CaMbinding domain can be found in nucleomorphin (residues 172-185) that contains characteristics of the 1-10 motif, in which two bulky hydrophobic residues are spaced by eight amino acids. Many, but not all, known CaMBPs with this domain also have a conserved bulky hydrophobic residue positioned at amino acid 5 (11). The Ca 2ϩ -dependent binding of nucleomorphin was further confirmed using bacterial expressed protein on CaM-agarose (Fig. 5). This assay used the pETNumA⌬118 -167 construct lacking the DEED repeat (residues 121-172) but retained the predicted CaM-binding domains. We hypothesized that this long stretch of acidic residues was affecting the translation of nucleomorphin in E. coli by putting high demands on the need for each charged tRNA, respectively. This would likely result in pausing of the ribosome complex, halting translation, and leading to instability in the transcript. Expression of those constructs containing the Asp/Glu repeat could only be detected using serum antibodies against MBP or the T7 tag (data not shown). The concentration of CaM used in the screening of the library and subsequent CaM binding series of experiments falls within the range of physiological CaM concentrations in the cell supporting evidence that nucleomorphin plays a physiological role in Dictyostelium. To this end, we are beginning sitedirected mutagenesis experiments to identify the exact sequences responsible for CaM binding.
Data base searches failed to reveal a homolog of nucleomor-phin. However, the sequence contains a putative conserved domain of the superfamily of proteins known as nucleoplasmins. Other evidence also suggests nucleomorphin is a nuclear protein.
There are four predicted NLS within the amino acid sequence of nucleomorphin. Of these, one NLS is consistent with that of the bipartite classification, found at position 48 in the amino acid sequence (residues KKSYQDPEIIAHSRPRK), which is consistent with nucleoplasmin (58). The significance of the NLS in nucleomorphin has yet to be shown, but experiments to delete these regions by site-directed mutagenesis will likely resolve this question. In addition to this is the presence of multiple potential phosphorylation sites for such kinases as casein kinase II, protein kinase A, and protein kinase C. Multiple phosphorylation sites and nuclear localization signals are also characteristics of nucleoplasmin and nucleophosmin (59,60). Nucleomorphin is a highly acidic protein largely in part to the 52-amino acid region of Asp/Glu residues. Amino acid comparison of nucleomorphin with members of the nucleoplasmin family reveals sequence identities and similarities confined to the acidic domain. All members of this family have one or more acidic domains consisting of a total of 17 to more than 100 glutamic acid and aspartic acid residues per molecule. Sequence analysis shows that these two residues can comprise more than 25% of the total amino acids in some proteins of this superfamily; in nucleomorphin, they make up ϳ21% of the total residues. Other chromatin-associated proteins, such as p62 (50), also contain acidic domains. Although nucleomorphin does not contain an identifiable DNA-binding domain, members of the nucleoplasmin family, including nucleoplasmin 3 and nucleoplasmin also lack such domains (61). The DEED repeat may also explain the anomalous molecular mass in SDS-PAGE. Like p62, CRP-1 from Drosophila and Xenopus nucleoplasmin, the predicted molecular mass of nucleomorphin (38 kDa) is less than the apparent mass calculated from SDS-PAGE mobility (43 kDa). The larger apparent mass may be due to the acidic properties of the protein as was shown for p62 (50). Although phosphorylation can contribute to the increase in mass, the phosphorylation state of nucleomorphin is still unknown. In keeping with the acidic properties of nucleomorphin, the classes of proteins called A-proteins have also been shown to contain extensive stretches of acidic residues (50). These proteins are a complex group that vary in both structure and function, yet demonstrate a significant binding to the core histones in vivo (62,63). This suggests that the acidic domain within nucleomorphin could serve to bind to core histones or other positively charged chromatin-associated proteins. It has been shown that nucleoplasmin in binding to the core histones mediates chromatin decondensation, nucleosome formation, and DNA transcription and is believed to be essential in assembling nucleosomal arrays, although precise functions and the mode of action remain to be elucidated (58,(63)(64)(65).
In Dictyostelium, nucleomorphin mRNA and protein are maintained throughout development at relatively steady levels. Expression of GFP-NumA was analyzed by Western blotting using monoclonal anti-GFP antibodies, and it was found that the construct, although present in appreciable amounts, was being degraded. Furthermore, it accounts for the appearance of weak GFP fluorescence throughout the cell. This suggests nucleomorphin expression is tightly regulated so as to maintain a specific physiological level. GFP-NumA was almost exclusively localized to heterochromatin-like patches at the periphery of the nucleus. Truncated forms of GFP-NumA lacking the putative transmembrane domain also localized to the nuclear periphery. The fluorescence observed was similar to the wild type, but the arc-like pattern was less distinct. Thus we cannot rule out the possibility that nucleomorphin attaches to the nuclear envelope. If nucleomorphin associates with chromatin-binding proteins, the mode of interaction may be strong enough to keep it localized to the nuclear periphery even in the absence of the TM domain. GFP-NumA⌬118 -167 lacking the DEED repeat provided clues to its potential role in Dictyostelium. Cells expressing this construct display a dramatic increase in the number of nuclei per cell yet appear healthy otherwise with a normal growth rate. Although the intranuclear localization of nucleomorphin was not affected, it suggests that the DEED repeat may serve to negatively regulate processes involved in nuclear structure or organization. The wild type numA gene is present in cells transfected with each of the GFP constructs but likely not at a level to compensate for the effects of overexpressing GFP-NumA⌬118 -167. This provides evidence of functionality associated with the DEED-repeat regarding the regulation of nuclear number. The size of the nuclei did not seem to be affected by any of the constructs used except in the case of cells with extremely large numbers of nuclei as was seen in one cell with 16 nuclei. These cells also appeared to be somewhat larger in size. We have begun detailing these characteristics through the analyses of nuclear and cytosolic volumes during growth and development. It will also be important to determine if the DNA content of the nuclei in the multinucleated cells is altered under any conditions. Several attempts to knock out nucleomorphin by way of homologous recombination have been unsuccessful providing more evidence of the need to regulate nucleomorphin levels in vivo. Similar to nucleomorphin, both Drosophila CRP-1 and CRP-2 are detectable at steady-state levels throughout development (63).
The role of CaM in the nucleus is well known, and several nuclear CaMBPs have been identified in other organisms (11). Gauthier et al. (21) had previously shown that about one dozen Ca 2ϩ -dependent nuclear CaMBPs exist in Dictyostelium, but nucleomorphin is the first nuclear CaMBP to be characterized in this organism. Furthermore, this work presents the first evidence demonstrating the CaM-binding ability of a nucleoplasmin-like protein. The apparent heterochromatic association of nucleomorphin at the nuclear periphery and the in-crease in nuclear number in constructs lacking the DEED repeat strongly suggest that this protein is involved in some fundamental organization of the Dictyostelium nucleus.