The Giardia duodenalis 14-3-3 Protein Is Post-translationally Modified by Phosphorylation and Polyglycylation of the C-terminal Tail*

The flagellated protozoan Giardia duodenalis (syn. lamblia or intestinalis) has been chosen as a model parasite to further investigate the multifunctional 14-3-3s, a family of highly conserved eukaryotic proteins involved in many cellular processes, such as cell cycle, differentiation, apoptosis, and signal transduction pathways. We confirmed the presence of a single 14-3-3 homolog gene (g14-3-3) by an in silico screening of the complete genome of Giardia, and we demonstrated its constitutive transcription throughout the life stages of the parasite. We cloned and expressed the g14-3-3 in bacteria, and by protein-protein interaction assays we demonstrated that it is a functional 14-3-3. Using an anti-peptide antibody raised against a unique 18-amino acid sequence at the N terminus, we observed variations both in the intracellular localization and in the molecular size of the native g14-3-3 during the conversion of Giardia from trophozoites to the cyst stage. An affinity chromatography, based on the 14-3-3 binding to the polypeptide difopein, was set to purify the native g14-3-3. By matrix-assisted laser desorption ionization mass spectroscopy analysis, we showed that polyglycylation, an unusual post-translational modification described only for tubulin, occurred at the extreme C terminus of the native g14-3-3 on Glu246, Glu247, or both and that the Thr214, located in the loop between helices 8 and 9, is phosphorylated. We propose that the addition of the polyglycine chain can promote the binding of g14-3-3 to alternative ligands and that the differential rate of polyglycylation/deglycylation during the encystation process can act as a novel mechanism to regulate the intracellular localization of g14-3-3.

14-3-3s are a family of highly conserved dimeric proteins with an approximate molecular mass of 30 kDa. With the exception of prokaryotes, they have been found in all eukaryotes studied to date, including protozoa, yeasts, plants, and animals (1). Generally, unicellular organisms have either one or two 14-3-3s (e.g. the slime mold Dictyostelium discoideum has one, and the yeast Saccharomyces cerevisiae has two), whereas multicellular organisms contain several. Mammals, including Homo sapiens, contain seven 14-3-3 genes (i.e. ␤, ␥, ⑀, , , , ), and 15 genes coding for 14-3-3s have been identified in the plant Arabidopsis thaliana. 14-3-3 homologs have also been found in several protozoan and metazoan parasites (1,2). The crystal structure of both animal and plant 14-3-3s has been shown to be conserved. The monomers, consisting of nine ␣-helices organized in a cup-like shape, are able to interact by their N-terminal portions to form U-shaped dimers (3)(4)(5).
The main property of 14-3-3s is their ability to bind other proteins containing a consensus Ser/Thr-phosphorylated binding sequence by the interaction with conserved residues located in the amphipathic groove of each monomer (6, 7). 14-3-3s were first described in mammals as activators of the brain enzymes tyrosine and tryptophan hydroxylases (8). To date, 14-3-3 interactors, identified in both animal and plant cells, have been shown to include enzymes, transcriptional factors, and structural proteins (9 -11). The phosphorylation (or dephosphorylation) of the 14-3-3 ligands determines their intracellular localization, complex formation, enzyme activation, and conformational changes through the association or dissociation with 14-3-3s.
With regard to protozoan parasites, only limited information is available on 14-3-3s. Preliminary studies, which have mainly focused on expression and localization, have been conducted on Plasmodium falciparum, Plasmodium knowlesi, Toxoplasma gondii, and Eimeria tenella, and the results have suggested that 14-3-3s play a role in the transmission of the regulatory signals involved in a wide array of biological processes, including the proliferation and migration of the parasite during infection (2,12,13).
To date, no analyses of 14-3-3s have been performed for Giardia duodenalis. The flagellate protozoan G. duodenalis (syn. lamblia or intestinalis), which parasitizes the upper part of the small intestine of mammals, including humans, is one of the major causes of non-bacterial diarrhea worldwide (14). This parasite belongs to the order Diplomonadidae, the earliest known branching eukaryotic lineage. Giardia is an "ancient" eukaryote, as inferred from the phylogenetic analysis of different genes and proteins (15,16) and from the finding that it retains many prokaryotic characteristics, such as the size of the small subunit rRNA and the presence of bacterial-like metabolic enzymes. Moreover, it lacks typical eukaryotic structures (i.e. peroxisomes) and has a reduced Golgi apparatus and mitochondrial remnant organelles (known as "mitosomes"), probably as an adaptation to parasitism (17). The Giardia life cycle consists of two main stages, (i) the trophozoite, a teardrop-shaped binucleated cell that colonizes the host intestine and reproduces through binary fission, and (ii) the cyst, an infective stage that is able to survive in the external environment. Given that the life cycle can be reproduced in vitro and that the entire genome is available (18), Giardia is a suitable model for analyzing the functions of 14-3-3s in eukaryotes and the peculiar role of this class of protein in parasites.
In the present study we describe the cloning and expression of the * This work was partially funded by the project "Vaccini anti-infettivi per l'uomo nel terzo millennio: integrazione di approcci molecolari, microbiologici ed immunologici," code 3AIF of the Italian Ministry of Health and by the European project MED-VET-NET, code WP12. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The nucleotide sequence(s) reported in this paper has been submitted to GenBank TM  14-3-3 protein homolog of G. duodenalis. The recombinant 14-3-3 was biochemically characterized, and the expression and intracellular localization of the 14-3-3 protein were studied during the parasite life stages using specific antibodies. Moreover, mass spectroscopy analysis of the purified protein demonstrated that the Giardia 14-3-3 is phosphorylated and, for the first time, that it is a target for polyglycylation.
Cell Culture and Differentiation-Trophozoites of the G. duodenalis WB-C6 strain were axenically grown for 72 h at 37°C in TYI-S-33 medium supplemented with 10% bovine serum and bovine bile at pH 7.0. Encystation was induced essentially as described (19). Briefly, the cells were grown until confluence, and the medium was replaced with TYI-S-33 supplemented with 10% bovine serum and 10 mg/ml bovine bile at pH 7.8 and then incubated at 37°C for the indicated time as shown in the figures.
Nucleic Acid Isolation-Genomic DNA was isolated from 10 9 cells of G. duodenalis using the phenol/chloroform extraction method. Total RNA was extracted from 10 7 trophozoites, or encysting parasites, using the RNAeasy mini kit (Qiagen, Hilden, Germany) following the manufacturer's instructions. Plasmid DNA was isolated from bacteria using the QIAPrep kit (Qiagen).
Northern Blot Analysis-Total RNA (10 g) was separated on 1.2% agarose gel containing 37% formaldehyde. RNA was capillary-transferred with 20ϫ standard saline phosphate-EDTA onto Hybond-N nylon membranes (Amersham Biosciences), and Northern hybridization was performed at 50°C, as described by Sambrook et al. (20), using specific probes previously labeled with [␣-32 P]dCTP using the RadPrime DNA labeling system kit (Invitrogen) following the manufacturer's instructions.
Vector Construction-The g14-3-3 coding sequence was amplified by PCR as described above. For overlay experiments, a short sequence, coding for a protein kinase A phosphorylation site, was introduced by PCR at the 5Ј end of the g14-3-3 coding sequence using the designed primers Gd14PKforw (5Ј-ggatcccgtcgtgcatctgttatggccgaggcatttacgcgt-3Ј) (the BamHI site is underlined, and the phosphorylation site coding sequence is in italics) and Gd14rev to produce the modified gPK14-3-3 protein. The PCR conditions were the same as those described for g14-3-3 amplification. Both fragments were cloned in the BamHI and EcoRI sites of the pGEX6-P1 vector (Amersham Biosciences) in-frame with the glutathione S-transferase (GST) 3 and introduced in Escherichia coli JM109 competent cells. The cloned fragments were verified by DNA sequencing. The plasmids were named p14-X and pPK14-X, and the expressed proteins were named g14-3-3 and gPK14-3-3, respectively.
The mutants PK-K53E and T214A were obtained by site-directed mutagenesis of the g14-3-3 coding sequence using the QuikChange site-directed mutagenesis methods (Stratagene, La Jolla, CA). The plasmid pPK14-X was used as template to produce the PK-K53E mutant, and the plasmid p14-X was used as a template for the T214A mutant. The primers used for mutagenesis were: K53Eforw (5Ј-ctgctttctgtagcctacgagaacgtcatcggc-3Ј) and K53Erev (5Ј-gcgggggccgatgacgttctcgtaggctacaga-3Ј) and T214Aforw (5Ј-acagatctggacaagctggccgaggagtcttac-3Ј) and T214Arev (5Ј-cgagtccttgtaagactcctcggccagcttgtc-3Ј) (the mutated triplets are underlined). After 18 cycles of PCR (95°C for 30 s, 53°C for 1 min, and 68°C for 8 min), 10 units of DpnI were added to the mixture to digest the templates, and the reactions were carried out at 37°C for 2 h. Twenty l of each reaction mixture was used to transform E. coli JM109 competent cells. The presence of the mutations was determined by DNA sequencing.
The plasmid pSCM138A, containing the difopein coding sequence fused to the enhanced green fluorescent protein (21), was kindly provided by Dr. Haian Fu. A 169-bp EcoRI difopein fragment was excised from the plasmid and cloned into the EcoRI-digested and dephosphorylated pGEX-6P1 vector in-frame with the GST coding sequence to produce the plasmid pDIF-X.
Expression and Purification of the Recombinant Proteins-E. colitransformed cells were grown in SOB medium (2% bacto tryptone, 0.5% yeast extract, 0.05% NaCl, 2.5 mM KCl, 10 mM MgCl 2 , pH 7.2) at A 600 ϭ 0.6, and the expression of recombinant proteins was induced in the presence of 1 mM isopropyl thio-␤-D-galactoside at 37°C for 4 h. GST-fused proteins were purified by affinity chromatography on glutathione-Sepharose 4B (Amersham Biosciences) and eluted with 10 mM reduced glutathione, pH 8.0, or released from GST by digestion with the appropriate amount of PreScission protease (Amersham Biosciences) in digestion buffer (50 mM Tris-HCl, 15 mM NaCl, 1 mM dithiothreitol, and 1 mM EDTA, pH 7.5) at 4°C for 16 h following the manufacturer's instructions.
Protein Preparation-Total proteins from soluble and membrane fractions were prepared according to the method of Moss et al. (22), with minor modifications. Briefly, 2 ϫ 10 9 trophozoites or encysting trophozoites were collected by chilling on ice and washed 3 times with cold PBS (140 mM NaCl, 2.7 mM KCl, 10 mM Na 2 HPO 4 , 1.8 mM KH 2 PO 4 , pH 7.2), and the cell pellet was frozen at Ϫ70°C overnight. Cells were resuspended in 2 volumes of extraction buffer (30 mM Tris-HCl, 1 mM dithiothreitol, and 1 mM EDTA, pH 7.4), supplemented with a protease-inhibitor mixture (P8340, Sigma) and a phosphatase-inhibitor mixture (P2850, Sigma), and then destroyed by sonication (5 times for 30 s at 60% power and 10% duty cycle) with a Sonoplus ultrasonic homogenizer (Bandelin electronic, Berlin, Germany). The lysate was centrifuged at 24,000 ϫ g for 30 min at 4°C, and the supernatant was collected and designated as soluble cytosolic fraction. The sediment containing the membranous material was washed twice with cold PBS and centrifuged both times for 30 min at 24,000 ϫ g at 4°C. The pellet was then resuspended in 3 ml of 8 M urea, 0.05 M Tris-HCl, 0.3 M KCl, and 0.002 M EDTA, pH 8.0, and constantly stirred at 4°C overnight. Further solubilization was achieved by sonication (5 times for 30 s at 50% power and 20% duty cycle), and centrifugation was performed at 24,000 ϫ g for 30 min at 4°C. The supernatant was collected, and the urea was removed by diafiltration against TBE (40 mM Tris, 54 mM boric acid, 1 mM EDTA, pH 8.3) using a PM-5 membrane and then concentrated using Centricon 10 (Millipore Corp., Bedford, MA). The protein concentration was measured with the method of Bradford (52) (Bio-Rad), and the material was stored at Ϫ70°C.
Overlay Assay-The overlay assay was carried out according to Lalle et al. (24). GST-gPK14-3-3 and the GST-PK-K53E mutant were labeled with [␥ 32 P]ATP on a cAMP-dependent protein kinase phosphorylation site using the protein kinase A catalytic subunit and then released from GST by cleavage with the PreScission protease. Proteins were separated by SDS-PAGE and transferred to a nitrocellulose membrane with 25 mM Tris, 192 mM glycine, and 0.01% SDS, pH 8.3, and blocked for 1 h with 5% nonfat dried milk in HT buffer (20 mM Hepes-KOH pH 7.6, 75 mM KCl, 5 mM MgCl 2 , 1 mM dithiothreitol, 0.1 mM EDTA, 0.04% Tween 20). The membrane was incubated at 4°C overnight in HT buffer with 3% nonfat dried milk and with 1 g/ml 32 P-labeled gPK14-3-3 or PK-K53E (9 kBq/ml). Where indicated, 100 M concentrations of the soluble phosphopeptide Raf259p (LSQRQRST(pS)TPNVHMV), reproducing one of the 14-3-3 binding motifs of the human Raf-1 protein, was added during the incubation with 32 P-gPK14-3-3. After thorough washing with HT buffer, the interaction was revealed by autoradiography. For Pep-spot binding assay, kindly provided by Drs. G. Cesareni and S. Panni, 100 nmol/cm 2 of phosphorylated and unphosphorylated forms of the Raf259 and Raf621 peptides were directly synthesized on cellulose-(3-amino-2-hydroxy-propyl) ether membranes, and after prewashing the filter with 100% ethanol and HT buffer, the overlay assay was performed as described above.
Immunofluorescence Microscopy-Trophozoites or encysting cells were collected and fixed at 37°C with 4% formaldehyde in PBS for 30 min. Fixed cells were stored in PBS at 4°C until permeabilization, or they were immediately permeabilized with 0.1% Triton X-100 in PBS at room temperature for 5 min and then spotted on slides. Slides were blocked with 1% gelatin in PBS for 30 min and then incubated at room temperature with polyclonal anti-g14-3-3 antiserum (1:100) for 1 h. After extensive washes with PBS, the parasites were incubated with a fluorescein isothiocyanate-conjugated anti-rabbit secondary Ab (Kirkegard and Perry Laboratories, Gaithersburg, MD) and the Texas Redconjugated anti-CWP monoclonal Ab A300-TR at 1:30 dilution (Waterborne Inc., New Orleans, LA), and the nuclei were stained with 300 nM 4Ј,6-diamidino-2-phenylindole. After washes with PBS, the slides were mounted with an anti-fading agent (Vectashield, Vector Laboratories, Burlingame, CA), and microscopy was performed using a Leica (Bensheim, Germany) confocal laser-scanning microscope with appropriate filter sets. A series of 20 optical sections with an approximate z distance of 0.3 m was performed for each cell. The image processing was performed using the Huygen program (Scientific Volume Imaging BV, Hilversum, The Netherlands) and the Imaris program (Bitplane, Zurich, Switzerland).
Pull-down Assay-For pull-down assays, 15 g of glutathione-Sepharose immobilized GST or GST-difopein were incubated with 3 mg of Giardia cytosolic fractions from trophozoites or from trophozoites grown for 12 h in encystation medium (12 h encysting cells) in HT buffer at 4°C for 2 h. After extensive washes with HT buffer, the 14-3-3 proteins bound to the beads were eluted with 100 l of 5 mM soluble Raf259p phosphopeptide in HT buffer. An aliquot was run in SDS-PAGE and subjected to immunoblot with the N14 serum as described above.
Mass Spectrometry Analysis-Aliquots of the PreScission-cleaved recombinant g14-3-3, the T214A mutant or the purified native g14-3-3 were separated on a one-dimensional gel NuPAGE 4 -12% (Novex, Invitrogen) run in MOPS buffer and stained with the Colloidal Blue staining kit (Invitrogen). Slices were excised, treated essentially as previously described (25), and digested with modified trypsin, sequencing grade (Promega Corp.). Peptide mixtures were desalted on a POROS-R2 (Applied Biosystems) handmade microcolumn or directly analyzed by MALDI-time of flight on a Voyager-DE STR mass spectrometer (Applera Corp.) in linear and/or reflector positive mode. As a matrix, we used re-crystallized ␣-cyano-4-hydroxycinnamic acid (␣-cyano, Sigma) dissolved in 50% CH 3 CN, 0.1% trifluoroacetic acid (10 mg/ml) or 2,5dihydroxybenzoic acid (DHB) (Sigma) in 50% CH 3 CN, and 1% orthophosphoric acid (30 mg/ml), which has been reported to enhance the efficiency of phosphopeptide ionization in MALDI by diminishing the loss of phosphoric acid (26). The presence of phosphopeptides in MALDI-MS spectra was confirmed by direct comparison between spectra acquired using 1% orthophosphoric DHB and ␣-cyano; the metastable decomposition product of the phosphopeptide is prominent with ␣-cyano-4-hydroxycinnamic acid, whereas it is present in lower quantities or not detectable in a 1% orthophosphoric DHB spectrum. Samples were loaded onto the instrument target using the dried droplet technique. Spectra were externally calibrated using a standard peptide mixture.
For the liquid chromatography-mass spectrometry analysis, LC/ESI-MS/MS and the LC/ESI-MS 3 experiments were performed on an LCQ-DECA XP instrument equipped with a Surveyor MS pump (Thermo Electron Corp.). Peptide mixtures were analyzed on a capillary column, BioBasic C18, 100 ϫ 0.180 mm, 5-m particle size (Thermo Electron Corp.); the operating flow (2 l/min) was obtained by the Surveyor MS pump through a homemade splitting device. The LCQ-ESI probe was equipped with a 34-gauge inner diameter metal needle (Thermo Electron Corp.). Peptides were eluted from the column in 50 min using a linear 5-60% acetonitrile gradient in 0.1% formic acid. The acquisition method for LC/ESI-MS/MS experiments was set to perform MS/MS data-dependent scanning on the three most abundant ions, enabling the dynamic exclusion function (repeat count 2). For LC-MS 3 experiments, parent ions at m/z 1054.6 (for MS/MS) and 1006.4 (for MS 3 ) were isolated (3.0 m/z window) and fragmented using 35% collision energy.
In Vitro Kinase Assays-Approximately 20 g of the glutathione-Sepharose immobilized recombinant GST-g14-3-3 or GST-T214A mutant protein was mixed with 50 l of the reaction buffer containing 100 mM Tris-HCl, pH 7.5, 20 mM MgCl 2 , 200 M ATP, 2.5 Ci of [␥-32 P]-ATP, protease and phosphatase inhibitors, and 0.4 mg of the trophozoite cytosolic fraction. After incubation at 30°C for 1 h, the reactions were stopped by adding 5 mM EDTA, and the beads were extensively washed with T-TBS. The recombinant wild type g14-3-3 and the T214A mutant protein were recovered by digestion with the PreScission protease, as described above, and 2 g of each protein was run in 12% SDS-PAGE and stained with Coomassie, and the phosphorylated polypeptides were detected by autoradiography of dried gels.
Densitometric Analysis-The image was acquired using the GS-690 Imaging Densitometer (Bio-Rad) and analyzed with the Multi-Analyst 1.1 software (Bio-Rad).
Sequence Analysis-General homology searches with DNA and protein sequences were conducted on non-redundant GenBank TM databases using the BLAST algorithm, available at www.ncbi.nlm.nih.gov/BLAST. Giardia genome databases were analyzed on-line (www.mbl.edu/Giardia) by submitting the protein sequence on translated databases (TBLASTN). Sequence data were provided by the National Centre for Biotechnology Information. Multiple alignments were performed using the ClustalW program at www.ebi.ac.uk/clustalw.

RESULTS
Identification and Cloning of the 14-3-3 Coding Sequence from G. duodenalis-A single putative 14-3-3 homolog gene has been annotated in the genome of the G. duodenalis WB-C6 clone (www.mbl.edu/Giar-dia). To verify the presence of other 14-3-3 homologs, we carried out an in silico screening of the G. duodenalis genome using the BLAST algorithm. As probes, we selected two amino acid sequences, RNLLS-VAYKN(V/I) and SYKDSTLIMQLL(R/H)DNLTLWTD(S/A), previously defined as 14-3-3 signatory motifs (27). The search confirmed the presence of the single open reading frame of 744 bp (GenBank TM /EBI accession number AACB01000009), previously identified by the Giardia genome project (18). The gene lacks introns and codes for a polypeptide of 248 amino acids with a predicted molecular mass of 28.5 kDa and a theoretical isoelectric point of 5.1 (data not shown). The sequence was confirmed by PCR amplification and sequencing of a 2.5kilobase region spanning the ATG and the stop codon of the open reading frame (data not shown). This gene was named g14-3-3 (Gen-Bank TM /EBI accession number DQ146480).
Using the ClustalW program, we performed a multiple alignment between the g14-3-3 protein and 117 sequences of different 14-3-3 isoforms from protozoa, fungi, plants, and animals, obtained from the GenBank TM data base. The g14-3-3 shows a degree of identity, ranging from 22 to 60%, with the other members of the 14-3-3 family (data not shown). In particular, we found 60% identity with the ⑀ subgroup of the animal 14-3-3s (Fig. 1, panel A). To evaluate the expression of g14-3-3 mRNA during the trophozoite stage of the parasite, a 750-bp fragment of the g14-3-3 gene was used as a probe in Northern blot analysis, showing that the mRNA is highly expressed (Fig. 1, panel B).
Expression of the Recombinant g14-3-3 and in Vitro Analysis of Its Biochemical Properties-To confirm that g14-3-3 is a functional member of the 14-3-3 family, the protein was expressed in bacteria as fusion with the GST and used in the overlay assay. Because 14-3-3s are characterized by the ability to form homo-and heterodimers and to bind target proteins containing 14-3-3 binding motifs, a dimerization experiment was performed using the 32 P-labeled gPK14-3-3 protein as bait and the g14-3-3 as prey. The g14-3-3 was able to dimerize with itself but not with GST alone, which was used as control (Fig. 2, panel A).
The binding properties of the recombinant g14-3-3 were tested by overlay assays using both phosphorylation-dependent and -independent binding motifs. To evaluate the specificity and the mechanism of the binding, we also produced the g14-3-3 point mutant K53E, in which the positively charged residue Lys 53 was replaced by glutamic acid, a negatively charged residue. The gPK14-3-3 was able to bind to two synthetic phosphopeptides, reproducing Raf-1 14-3-3 binding motifs, when they were used as prey but not their unphosphorylated version (Fig. 2, panel  B). On the contrary, the interaction of the K53E mutant with Raf257 phosphopeptides was abolished, and that with the Raf261 phosphopeptides was strongly reduced (Fig. 2, panel B).
At the same time we evaluated the interaction with a phosphorylation-independent motif using as target a GST-fused difopein (21), a polypeptide containing a tandem repeat of the WLDLE sequence and originally identified in the R18 peptide by phage display screening (7). The gPK14-3-3 interacted with GST-difopein but not with GST alone, whereas the K53E mutant did not interact at all (Fig. 2, panel C).
We examined the interaction of recombinant g14-3-3 with cytosolic and membrane protein fractions extracted from Giardia trophozoites. The recombinant gPK14-3-3 bound a large number of proteins on both fractions (Fig. 2, panel D). The specificity of the binding was confirmed by competition with soluble Raf259p synthetic phosphopeptide; in the presence of the phosphopeptide, the interaction with the g14-3-3 was completely abolished (Fig. 2, panel D).
Subcellular Distribution and Immunolocalization of g14-3-3 in the Giardia Trophozoite-The g14-3-3 was localized in cultured trophozoites. In permeabilized cells, the anti-g14-3-3 serum (N14) produced a spotted and intense labeling within the cytosol, whereas the nuclei were poorly stained. No signal was detected in the central area of the cell corresponding to the median body, and no labeling of the flagella was evident (Fig. 3, panel A).
Furthermore, the presence of g14-3-3 was assessed in different protein fractions obtained from trophozoites. The 14-3-3 protein was mainly detected in the soluble cytosolic fraction and, in a lower amount, in the membrane fraction (Fig. 3, panel B). A main band, corresponding to a molecular mass of ϳ32 kDa, was present in both fractions, whereas a second product of about 28 kDa was only detected in the membrane fraction. The difference between the theoretical molecular mass, 28.5 kDa, of the recombinant g14-3-3 protein and the 32 kDa of the observed main band could suggest that the native protein was subjected to one or more post-translational modifications.
The Subcellular Localization of the g14-3-3 Changes during Encystation-Two processes are essential in the life cycle of G. duodenalis; these are the development of the trophozoite into a cyst, or encystation, and the emerging of the infective trophozoites from the cyst, or excystation. We focused on encystation and investigated the expression levels of both mRNA and the g14-3-3 protein at different points in time during the process. For the Northern blot analysis, total RNA was probed with a 750-bp fragment encompassing the g14-3-3 coding sequence, and the expression of the g14-3-3 mRNA was compared with the mRNA levels of gap1, a constitutively expressed gene, and of cwp1, an encystations-induced gene (29). The g14-3-3 mRNA appeared to be constitutively expressed (Fig. 4 panel A).
The immunoblot analysis was performed using proteins extracted from trophozoites at different times during encystation. When the cytosolic soluble fractions were probed with the N14 serum, the 32-kDa band was immuno-decorated in all of the samples. A second band (of ϳ30 kDa) became progressively detectable in the samples obtained after 6 h of encystation and decreased after 24 h of induction, when almost 80% of the cells were developed into the cyst stage (Fig. 4, panel B). We did not obtain similar results for the membrane fraction, where the antiserum recognized a pattern of bands similar to that observed during the trophozoite stage (data not shown). Overlay assay. A, homodimerization of the g14-3-3. One g of GST or PreScission-cleaved g14-3-3 was separated on 12% SDS-PAGE, transferred onto nitrocellulose membrane, and probed with the 32 Plabeled gPK14-3-3. Ponceau staining of the membrane is shown on left, and autoradiography is shown on the right. Arrows indicate the protein molecular masses (kDa). B, interaction of the g14-3-3 with the phosphorylation-dependent binding sequence. The phosphopeptides Raf259 (SQRQRSTpSTPNVHM, pS is phosphoserine at position 259) and Raf621 (PKINRSApSEPSLHRA, pS at 621) and their unphosphorylated versions were synthesized on a cellulose-(3amino-2-hydroxy-propyl) ether membrane and probed with 32 P-labeled gPK14-3-3 or 32 P-labeled PK-K53E, and binding was revealed by autoradiography. C, interaction of the g14-3-3 with a phosphorylationindependent binding motif. One g of GST-difopein or GST was loaded in duplicate, separated on a 12% polyacrylamide gel, transferred onto nitrocellulose, and probed with 32 P-labeled gPK14-3-3 or 32 P-labeled PK-K53E. Half of the Ponceau stained filter is shown on the left;theautoradiographyisontheright.Thearrowsindicate molecular masses. D, binding of radiolabeled gPK14-3-3toGiardiaproteinextracts.Leftpanel,overlay of 32 P-labeled gPK14-3-3 with the trophozoite protein extracts after separation on 10% SDS-PAGE. S, soluble fraction (50 g); M, membrane fraction (50 g). Right panel,theoverlayexperimentwasrepeatedinthepresence of 100 M soluble synthetic phosphopeptide Raf259p. Molecular masses are on the right.
We also evaluated whether or not the intracellular localization of the protein during encystation was modified. The process was monitored using an anti-CWP2 monoclonal Ab (A300-TR), which immuno-decorates the encystation-specific vesicles in the encysting cells or the wall of the cysts. As in the case of the trophozoites, the encysting cells as well as the cysts showed the previously observed staining of the cytoplasm (Fig.  4, panel C). By contrast, an increasing nuclear localization of g14-3-3 was evident during the transition of cells from the early to the late encysting stage, yet the nuclei of the cysts were poorly stained (Fig. 4,  panel C). The g14-3-3 was not present in the encystation-specific vesicles of the encysting cells, as demonstrated by the absence of co-localization signal with the CWP2 protein (Fig. 4, panel C).
The g14-3-3 Is Post-translationally Modified by Phosphorylation and Polyglycylation-To determine the possible presence and the nature of post-translational modifications, the native g14-3-3 was purified by affinity chromatography. For this purpose, we used the soluble proteins  extracted from trophozoites and 12 h encysting cells, in which, according to the immunoblot analysis, the signal corresponding to the 30-kDa band was most evident. The complete set of immunolabeled bands, present in the soluble fraction from both trophozoites and 12-h encysting cells, was recovered in the purified material, as demonstrated by Western blot analysis with the N14 serum (Fig. 5, panel A). Furthermore, the calculated relative intensity of the 30-and 32-kDa bands at 12 h was comparable in the purified proteins and in the input soluble protein fraction (Fig. 5, panel B), suggesting that the modification(s) of g14-3-3 did not affect the affinity for either the difopein or the soluble Raf259p phosphopeptide.
The trophozoite-purified and the recombinant g14-3-3s were subjected to mass spectrometry analysis. The direct comparison of the peptide mass fingerprint of both proteins clearly showed largely comparable MALDI-MS spectra, confirming that the purified protein really was the g14-3-3. The spectra of the purified native sample lacked the internal tryptic peptide 202 AFDAAITDLDKLTEESYK 219 (202-219), expected at m/z 2030.00, and the peptide 230 DNLNLWVTDSAG-DDNAEEK 248 (230 -248), expected at m/z 2105.93 and corresponding to the extreme C-terminal peptide of the protein (Fig. 5, panel C).
In the MALDI-MS spectrum of the purified native sample (Fig. 5, panel C), acquired with 1% orthophosphoric acid DHB, the presence of the signal at m/z 2109.98, instead of the signal at m/z 2030.02, found in the recombinant protein, could be ascribed to a phosphorylation of the peptide (202-219). This hypothesis is supported by the presence in the MALDI-MS spectrum, acquired with ␣-cyano in reflector mode, of a peak at m/z 2012, not resolved, corresponding to the metastable decomposition product indicative of the loss of phosphoric acid (⌬m ϭ Ϫ98 Da).
To confirm this hypothesis, a tryptic digest of the trophozoite-puri-  (Fig. 6, panel A) revealed that the phosphorylated residue was Thr 214 . Stoichiometry of Thr 214 phosphorylation appears to be very high in trophozoite and 12-h encysting g14-3-3, as also confirmed by the presence in the MALDI/MS spectrum of detectable species corresponding to phosphorylated peptides (213-229) and (202-229), whereas their unphosphorylated counterparts are completely absent (data not shown).
To confirm the site of phosphorylation, a point mutant of g14-3-3 was generated, in which the threonine in position 214 was replaced with an alanine. The T214A mutant was expressed, and the presence of the mutation was verified by mass spectrometry (data not shown). Both the wild type and the T214A recombinant proteins were subjected to an in vitro phosphorylation assay in the presence of [␥ 32 P]ATP and of the soluble protein fraction from trophozoites. A protein kinase activity was able to phosphorylate the recombinant wild type protein but not the T214A mutant (Fig. 6, panel B). However, no electrophoretic shift of the wild type recombinant protein occurred after the in vitro phosphorylation.
The absence in the trophozoite-purified protein of the C-terminal peptide (229 -248), expected at m/z 2105.93, suggested that there was another post-translational modification(s) and/or a cleavage of this portion of the protein. Intriguingly, further inspection of MALDI-MS spec-tra at higher m/z values revealed the presence of peaks differing by multiples of 57 Da, which correspond to the expected mass increases due to successive additions of glycine residues (Fig. 7, panel A). This signal pattern can be assigned to polyglycylation of the C-terminal peptide. This post-translational modification, previously reported only for ␣and ␤-tubulin (30), results from the addition of one or more glycine residues to the ␥-carboxyl group of a glutamate. Two different glutamic acid residues, Glu 246 and Glu 247 , are contained in this portion of the g14-3-3 protein, and both could be sites for the polyglycylation. Polyglycylation could completely explain the broadness and the slower migration of the band corresponding to the native form in comparison with the recombinant protein.
When the 14-3-3 doublet from the 12-h encysting cells was analyzed, we observed that both forms were phosphorylated on Thr 214 (data not shown) and polyglycylated (Fig. 7, panel B). The extent of the polymodification differed between trophozoite-and 12-h encysting cell-purified proteins. As can be inferred by molecular mass calculations in the trophozoite-purified g14-3-3, glycine residues ranged from 10 to 31, with the greatest peak corresponding to 21 residues, and from 6 to 22 residues in the 12-h encysting lower band, with the main peak corresponding to the addition of 10 residues (Fig. 7, panel A and B).
The nature of this post-translational modification was assessed using the monoclonal antibody AXO49, previously developed against the polyglycylated tubulin and able to detect polyglycine chains of at least four residues (31). The antibody recognized g14-3-3 from both the trophozoites and the 12-h encysting cells both in the soluble fraction and in the purified g14-3-3, yet it failed to detect the unmodified recombinant protein expressed in bacteria (Fig. 7, panel C). However, when AXO49 was used to probe the membrane fraction from both the trophozoites and the 12-h encysting cells, it strongly recognized only a band with a molecular mass compatible with tubulin even though the identity of this band could not be confirmed (data not shown).
Because nothing is known about polyglycylating enzymes to explain how g14-3-3 can be a target for this enzymatic activity, we attempted to define a putative polyglycylation target sequence common to tubulins and g14-3-3. The last 13 C-terminal amino acids of the g14-3-3 protein, 236 VTDSAGDDNAEEK 248 , are unique to this isoform when compared with all of the other 14-3-3s and contain the polyglycylated glutamic acid(s). We aligned this portion of the g14-3-3 with the last 16 C-terminal residues surrounding the polyglycylation sites of the ␣and ␤-tubulin from G. duodenalis, Tetrahymena thermophila, and Paramecium tetraurelia, in which polyglycylation has been described and the polyglycylated residues have been identified (32)(33)(34). As result of the alignment, we observed that conserved residues were present in the tubulins as well as in the g14-3-3 (Fig. 7, panel D). These conserved amino acids defined a motif (T/G)X 0 -1 (D/E)X 1-3 G(D/E)X 1-2 (gE) [2][3][4] , where X is a polar or a negative charged amino acid, and gE is a polyglycylated glutamic acid.

DISCUSSION
We report on the cloning and characterization of a new member of the 14-3-3 protein family isolated from the parasite protozoa G. duodenalis. Our investigation of the G. duodenalis genome confirmed the presence of a single 14-3-3-homolog gene, previously annotated by the Giardia genome project (18). The presence of a single 14-3-3 coding gene is also consistent with data on other unicellular organisms, including parasites such as T. gondii, Neospora caninum, and Plasmodium spp. (12,35,36), although the ciliate Entodinium caudatum and the parasite protozoa Entamoeba histolytica harbor, respectively, four and three different 14-3-3 genes, probably as the result of gene duplication (1). The comparison of the amino acid sequences revealed a high degree of identity of the g14-3-3 protein with the ⑀ subgroup of Metazoan 14-3-3s. This is particularly important when considering that the ⑀ isoform is the most closely related to the hypothetical ancestral animal 14-3-3 protein (27). However, a phylogenetic tree based on the alignment of 14-3-3 proteins cannot be used to argue the evolutionary position of G. duodenalis given that, as previously proposed, the protozoan 14-3-3 isoforms are highly divergent from the other 14-3-3s, and it is difficult to select one of them to root a phylogenetic tree (1).
With regard to the properties of the g14-3-3 protein, in vitro interaction experiments using both its recombinant and native form have clearly demonstrated that it is a functional member of the 14-3-3 family with the ability to form homodimers and bind both phosphorylationdependent and -independent binding motifs. The specificity of these interactions is also demonstrated by the use of the K53E mutant in which, as for the well characterized K49E mutant of the mammalian 14-3-3 isoform (7, 28), the charge inversion at that particular position hampers the binding ability of g14-3-3. This experimental evidence strongly suggests that the general mechanism regulating the 14-3-3target interaction is well conserved in the g14-3-3.
The involvement of the g14-3-3 in a complex network of interactions in vivo is supported in vitro by the specific binding to a large number of proteins and by the presence of g14-3-3 not only in the cytosol but also in association with membrane proteins. The localization of 14-3-3 proteins in different subcellular compartments and in association with membrane proteins has been widely demonstrated (37,38), and a lipid raft-associated form of the T. gondii 14-3-3 homolog has been recently described (13).
In other parasites, such as P. falciparum, the expression of the 14-3-3s during the erythrocyte stage of the parasite has been suggested to depend on both the formation of a specific membranous structure and the inhibition of the translocation of the protein kinase C to the membrane of the infected cell (12). It has been suggested that 14-3-3s could be involved in the stage-specific development of T. gondii (35).
Our data demonstrate that the g14-3-3 mRNA and the g14-3-3 protein are constitutively expressed both in the Giardia trophozoite and during the encystation stage, and transcription and translation levels are stably maintained. However, encystation signals affect both the rate of post-translational modifications and the intracellular localization of the encoded protein.
To the best of our knowledge this is the first study to show that polyglycylation, previously detected only in tubulins, occurs in the g14-3-3 protein and that the amount of glycines bound to the protein is high in the trophozoite yet drastically reduced during encystation. Polyglycylation, differently from other post-translational modifications such as phosphorylation and tyrosinylation, dramatically influences the protein structure through the addition of a polyglycine side chain covalently attached by an isopeptide bond to the ␥-carboxylate group of a glutamate residue. Polyglycylation has been described as a post-translational modification of the C-terminal portion of the tubulin isolated from the  . Polyglycylation of g14-3-3. MALDI-MS spectra, acquired in reflector mode using ␣-cyano as the matrix, of the trophozoite-purified g14-3-3 (A) and 12h-encysting g14-3-3 (B) recovered from SDS-PAGE and digested with trypsin is shown. Three panels were selected from the spectra to better compare the distribution of glycylated species in the two proteins. An asterisk indicates polyglycylated peptides; the numbers of glycine residues are reported. C, Western blot analysis with the anti-polyglycylated tubulin-specific antibody (AXO49) (upper panel) and the N14 serum (g14-3-3) (lower panel). An aliquot ( 1 ⁄100) of the input soluble protein fraction from trophozoites (T) and encysting cells (12 h), an aliquot (1/20) of pull-down-purified g14-3-3 from trophozoites (T) and 12-h encysting cells (12 h), and 0.5 g of recombinant g14-3-3 (R) were loaded on 4 -20% SDS-PAGE. The immunodetection was visualized using horseradish peroxidase-conjugate secondary antibody and a chemiluminescent assay. Arrows indicate the different g14-3-3 forms. D, alignment of g14-3-3 with ␣and ␤-tubulin of G. duodenalis (GenBank TM /EBI accession numbers AAN46106 and P05304), ␣and ␤-tubulin of P. tetraurelia (GenBank TM /EBI accession numbers CAA67848 and CAE75646), and ␣and ␤-tubulin of T. thermophila (GenBank TM /EBI accession numbers P41351 and P41352). The alignment was performed with the ClustalW program and manually refined. The amino acids in gray define the hypothetical polyglycylation sequence (T/G)X 0 -1 (D/E)X 1-3G(D/E)X 1-2 (E) [2][3][4] . The asterisks indicate the two glutamic acid residues of g14-3-3 possibly glycylated. Putative or experimentally defined (in bold) polyglycylated glutamic acid residues are underlined. cilia of Paramecium (30) and for both ␣and ␤-tubulin of the ciliate and flagellate cells (39). Polyglycylation has already been demonstrated in Giardia, where it is the main post-translational modification of tubulins (32). We show that the polyglycylation occurs at the extreme C-terminal portion of the g14-3-3 protein, as in the case of tubulins. However, we are not able to define whether one or both glutamic acid residues, Glu 246 and Glu 247 , are subjected to polymodification. Polyglycylase and deglycylase activities have not been previously described even if the presence of the complete set of glycylases in Giardia has been inferred (40), yet in our study the sequence homology between the C-terminal tails of g14-3-3 and tubulins from Giardia, Tetrahymena, and Paramecium, where polyglycylation is predominant, suggests that the motif (T/G)X 0 -1 (D/E)X 1-3 G(D/E)X 1-2 (gE) [2][3][4] , which we identified in all of these proteins, could be a possible target sequence for polyglycylation.
Despite the generally high level of sequence homology among 14-3-3s, the C termini of different isoforms are hypervariable and differ in length; they are shorter in mammalian 14-3-3s (excluding the ⑀ isoform) and longer in yeast and plant isoforms (27). The g14-3-3 protein belongs to the group of short isoforms, as evident in the multiple alignment ( Fig. 1).
In view of the general high flexibility of the 14-3-3 C-terminal tails (3)(4)(5), we propose that the addition of polyglycine chain(s) to the extreme C terminus of the g14-3-3 could act as a removable tail affecting the intracellular localization of the protein and/or allowing the g14-3-3 dimer to interact at the same time with canonical ligands and with other molecules by means of a further mechanism mediated by the glycine chain(s).
It has recently been demonstrated that, in barley (Hordeum distichum), proteolytic cleavage of the C-terminal tail occurs for the 14-3-3A during seed germination, yet this has not been reported for other barley isoforms (1). The truncation does not seem to alter the binding property, yet it excludes the protein from the nucleus (41).
In a similar way our data suggest that the addition of polyglycines does not affect the binding properties of g14-3-3. The affinity of both the high and low glycylated g14-3-3 for the GST-difopein is comparable, since the relative amount of the two forms was constant before and after the affinity purification. This result is also in agreement with the observation that the binding affinity of human 14-3-3 C-terminal-tagged with green fluorescent protein is not altered (41), although a quantitative assay that includes a non-polyglycylated g14-3-3 would be necessary to fully support our result.
By contrast, we observed that the appearance of a less polyglycylated g14-3-3 during encystation coincided with the translocation of the protein inside the nuclei. In plants and animals the nuclear localization of 14-3-3s and their association with transcription factors have been shown, although their precise role in these transcription complexes is still unclear (43,44). In human cells it has been also demonstrated that 14-3-3s are actively transported in and out of the nucleus with a distinct subcellular distribution of 14-3-3 and -isoforms caused by differences in the nuclear export (42). In this scenario the observed change in the polyglycylation level of the g14-3-3 could affect a differential rate of nuclear import/export of the protein. Nuclear accumulation could also depend on the binding of g14-3-3 to nuclear proteins (e.g. transcription factors) as a consequence either of their stage specific phosphorylation or of an increased availability of the less polyglycylated 14-3-3 protein for other interactions. In Giardia, the activation of specific genes and changes in the metabolism occur during encystation with the activation/inactivation of protein kinases and phosphatases and the consequent change in the phosphorylation status of Giardia proteins (45,46).
In vivo studies, conducted in Tetrahymena with various polyglycyla-tion mutants of ␣and ␤-tubulin, have focused on the essential role of polyglycylation for correct axonemal organization, ciliary motility, and complete cytokinesis (47,48). In these cases tubulin polyglycylation has been suggested to play a role in the regulation of intraflagellar transport, a process necessary for the assembly and maintenance of cilia and flagella, by the movement of large protein complexes containing accessory proteins such as microtubule-associated proteins, dynein, and kynesin along the doublet microtubules (34). Canonical interactions of 14-3-3 proteins with these accessory proteins have also been proven (11), and the addition of polyglycine chain(s) could allow g14-3-3 to also participate in the microtubules turnover. We also identified the phosphorylation of Thr 214 as an extensive g14-3-3 post-translational modification. In our experimental conditions this residue appears to be the only one phosphorylated, as shown by the complete lack of 32 P incorporation into the T214A point mutant (Fig.  6B). The protein was phosphorylated both in the trophozoite and in the encysting cells. A three-dimensional model of g14-3-3, based on the crystal structure of other 14-3-3s (data not shown), localizes the Thr 214 in the loop between ␣-helices 8 and 9. Recently, a comparison of the resolved crystal structure of human 14-3-3, -, and -isoforms, free or in a complex with different ligands, focused on the high flexibility of this region, with differences in folding and conformation between the isoforms, suggesting that this portion of the protein could be mainly responsible for the ligand specificity observed for different 14-3-3 isoforms (49). Moreover, the binding of polyamines and divalent cations to an EF hand-like region present in this portion of the A. thaliana GF14 isoform results in the modulation of the interaction with phosphorylated nitrate reductase (50). There is increasing evidence that 14-3-3 post-translational modifications act as mechanisms for regulating interactions; in particular, the direct phosphorylation of 14-3-3s, mediated by a variety of protein kinases, results in the increase or decrease of the binding affinity and in the alteration of the dimerization property of different 14-3-3 isoforms (51). The phosphorylation of g14-3-3 could play a similar role, yet as for the polyglycylation, further analyses will be necessary to answer this question.
The present work demonstrates that G. duodenalis could be a promising and interesting system to explore new mechanisms of the regulation of 14-3-3 proteins, suggesting that the potential of this protein class, which is already numerous, has yet to be completely explored.