Báo cáo khoa học: In vitro embryonic developmental phosphorylation of the cellular nucleic acid binding protein by cAMP-dependent protein kinase, and its relevance for biochemical activities pdf

Abbreviations CCHC, Cys-Cys-His-Cys; CK2, casein kinase 2; c-myc, cellular myelocytomatosis oncogene; CNBP, cellular nucleic acid binding protein; Comp-CT, CT element complementary seque

cellular nucleic acid binding protein by cAMP-dependent protein kinase, and its relevance for biochemical activities Vero´nica A Lombardo, Pablo Armas, Andrea M J Weiner and Nora B Calcaterra

Divisio´n Biologı´a del Desarrollo, IBR – CONICET, Area Biologı´a General, Facultad de Ciencias Bioquı´micas y Farmace´uticas,

Universidad Nacional de Rosario, Argentina

The zinc-finger cellular nucleic acid binding protein

(CNBP) shows striking sequence conservation among

vertebrates [1] Recent works show that CNBP is

required for forebrain formation during vertebrate

organogenesis Cnbp-null mutant mice are embryonic lethal and show severe forebrain truncation and facial abnormalities due to a lack of proper morphogenetic movements of the anterior visceral endoderm during

Keywords

CNBP; Danio rerio; embryogenesis;

phosphorylation; PKA

Correspondence

N B Calcaterra, IBR – CONICET, A ´ rea

Biologı´a General, Dpto de Ciencias

Biolo´gicas, Facultad de Ciencias Bioquı´micas

y Farmace´uticas, Universidad Nacional de

Rosario, Suipacha 531 (S2002LRK) Rosario,

Argentina

Tel ⁄ Fax: +54 341 4804601

E-mail: calcaterra@ibr.gov.ar or

ncalcate@fbioyf.unr.edu.ar

Database

CNBPS158Anucleotide sequence data is

available in the GenBank database under the

accession number DQ519386

(Received 9 October 2006, revised 6

November 2006, accepted 15 November

2006)

doi:10.1111/j.1742-4658.2006.05596.x

The zinc-finger cellular nucleic acid binding protein (CNBP) is a strikingly conserved single-stranded nucleic acid binding protein essential for normal forebrain formation during mouse and chick embryogenesis CNBP cDNAs from a number of vertebrates have been cloned and analysed CNBP is mainly conformed by seven retroviral Cys-Cys-His-Cys zinc-knuckles and a glycine⁄ arginine rich region box CNBP amino acid sequences show a puta-tive Pro-Glu-Ser-Thr site of proteolysis and several putaputa-tive phosphoryla-tion sites In this study, we analysed CNBP phosphorylaphosphoryla-tion by embryonic kinases and its consequences on CNBP biochemical activities We report that CNBP is differentially phosphorylated by Danio rerio embryonic extracts In vitro CNBP phosphorylation is basal and constant at early embryonic developmental stages, it begins to increase after mid-blastula transition stage reaching the highest level at 48 hours postfertilization stage, and decreases thereafter to basal levels at 5 days postfertilization The cAMP-dependent protein kinase (PKA) was identified as responsible for phosphorylation on the unique CNBP conserved putative phosphoryla-tion site Site-directed mutagenesis replacing the PKA phospho-acceptor amino acid residue impairs CNBP phosphorylation, suggesting that phos-phorylation may not only exist in D rerio but also in other vertebrates CNBP phosphorylation does not change single-stranded nucleic acid bind-ing capability Instead, it promotes in vitro the annealbind-ing of complementary oligonucleotides representing the CT element (CCCTCCCC) from the human cellular myelocytomatosis oncogene (c-myc) promoter, an element responsible for c-myc enhancer transcription Our results suggest that phos-phorylation might be a conserved post-translational modification that allows CNBP to perform a fine tune expression regulation of a group of target genes, including c-myc, during vertebrate embryogenesis

Abbreviations

CCHC, Cys-Cys-His-Cys; CK2, casein kinase 2; c-myc, cellular myelocytomatosis oncogene; CNBP, cellular nucleic acid binding protein; Comp-CT, CT element complementary sequence; CT, CCCTCCCC containing sequence element; dpf, days postfertilization; GST, glutathione S-transferase; hpf, hours postfertilization; hnRNP K, heterogenous ribonucleic protein K; PEST, Pro-Glu-Ser-Thr; PKA, cAMP-dependent protein kinase; PKA-Ca, PKA subunit catalytic a; PKC, protein kinase C; PKI, protein kinase inhibitor; rp-mRNA, ribosomal protein mRNAs; RGG, glycine ⁄ arginine rich region; 5¢ TOP, 5¢ terminal oligopyrimidine tract.

pregastrulation stage [2] In chick embryos, Cnbp is

expressed in the equivalent tissues of the mouse

embryo and, furthermore, CNBP siRNA knockdown

produces forebrain truncation [3] It was proposed

that the CNBP role during vertebrate organogenesis

is to regulate the forebrain formation by controlling

the expression of a number of rostral head

trans-cription factors, such as c-myc [2], BF-1, Six3, and

Hesx1 [3]

A vast range of cellular functions has been assigned

to CNBP It was reported acting both as a negative

[4–6] and positive [7,8] transcriptional regulator Apart

from this, CNBP showed interaction with several RNA

targets [9–12] From these reports emerge two main

possible models where CNBP might act as

single-stran-ded nucleic acid binding protein: the first one involves

CNBP and heterogenous nuclear ribonucleic protein K

(hnRNP K) as transcriptional activators of the c-myc

gene via the CT promoter element (CCCTCCCC)

[7,13]; the second model proposes CNBP as a

transla-tional inhibitor of ribosomal proteins mRNAs

(rp-mRNAs) through binding to the 5¢ UTR containing a

5¢ terminal oligopyrimidine tract (5¢ TOP) motif [14]

CNBP cDNAs have been cloned from mammals

[7,12,15–17], chicken [18,19], amphibians [1,20,21] and

fish [10,22] The protein shows a highly conserved

struc-tural organization and amino acid sequence sharing

the classical arrangement of seven tandem canonical

Cys-Cys-His-Cys (CCHC)-type zinc knuckle domains

(C-/-X-C-G-X3-H-X4-C, where / is an aromatic amino acid and X is a variable amino acid) typical of retroviral nucleocapsid proteins and the presence of an glycine⁄ arginine rich region (RGG) box between the first and second Zn knuckles In silico analysis of CNBP amino acid sequence reveals the presence of a putative nuclear localization signal PKKEREQ, a putative Pro-Glu-Ser-Thr (PEST) site of proteolysis, and several putative phosphorylation sites [1,10,21,22] (Fig 1) The relationship between CNBP developmental behaviour and phosphorylation has not been analysed yet To gain an understanding of the biochemical con-sequences of CNBP phosphorylation, it was important

to analyse CNBP phosphorylation during embryonic development and to identify kinases that catalyse CNBP phosphorylation We report here that CNBP

is in vitro phosphorylated by zebrafish embryonic extracts from different developmental stages The cAMP-dependent protein kinase (PKA) was identified

as the predominant kinase that catalyses CNBP phos-phorylation and, furthermore, it was found that CNBP major phosphorylation occurred on the unique con-served putative phosphorylation site Finally, we observed that CNBP phosphorylation promoted the annealing of complementary oligonucleotides in vitro but did not modify single-stranded nucleic acid binding capability Based on these results, we discuss the role

of phosphorylation-dephosphorylation events in the biological function of CNBP

Fig 1 (A) In silico analysis of zebrafish CNBP amino acid sequence The amino acid sequence of zebrafish CNBP was analysed using the Prosite server The amino acid residues involved in CCHC motifs are indicated in light grey boxes Amino acids that participate in the coordi-nation of the Zn atom are boxed in dark grey The RGG box is boxed and the putative nuclear localization signal is underlined The arrow shows a putative site of proteolysis (B) In silico analysis of putative phosphorylation sites Zebrafish CNBP putative phosphorylation sites were analysed by Prosite, NetPhos 2.0, ScanSite 2.0, GPS, and PredPhospho servers Boxed amino acid residues were predicted by at least three phosphorylation site servers.

Results and Discussion

CNBP shows several putative phosphorylation

sites

The amino acid sequence of zebrafish CNBP (Fig 1A)

was analysed in silico searching for putative

phos-phorylation sites by using Prosite [23], NetPhos2.0

[24], ScanSite [25], GPS (group-based phosphorylation

scoring method) [26], and PredPhospho [27] servers

The analysis yielded potential phosphorylation sites on

several amino acid residues (Fig 1B) The amino acid

residues Ser4, Thr56, Ser70, and Ser158 were predicted

as putative phospho-acceptors by at least three

phos-phosite prediction servers (Fig 1B), reinforcing the

hypothesis of CNBP phosphorylation

Sequence alignment did not show major putative

phosphorylation site conservation among CNBP

sequences except for the site located immediately after

the seventh CCHC Zn knuckle motif (Table 1) The

conservation among CNBP vertebrates suggests that

phosphorylation in this site, which in zebrafish was

predicted by four of the five employed servers

(Fig 1B), may play a relevant role in the regulation of

CNBP biochemical activities

CNBP is phosphorylated in vitro by embryonic

kinases

To facilitate the biochemical characterization of

CNBP phosphorylation, the protein was expressed in

Escherichia coli as fusion proteins with N-terminal

hexa-histidine (His6-CNBP) and N-terminal

glutathi-one S-transferase (GST) (GST-CNBP) tags and used

for in vitro kinase activity assays Both GST-CNBP

(Fig 2A,B) as well as His6-CNBP (Fig S1), were

phosphorylated by extracts prepared from 24 hours postfertilization (hpf) embryos This extract failed to phosphorylate GST (Fig 2A,B) demonstrating that the kinase phospho-acceptor site is in CNBP itself From the analysis of gels, it was noticeable the exist-ence of phospho-peptides with lower molecular masses than the recombinant fusion protein GST-CNBP This phenomenon may be the consequence of CNBP pro-teolysis by either bacterial or embryonic proteases, as

it was previously reported [11] Indeed, the radioactive polypeptides were recognized by a polyclonal antibody raised against CNBP in western blot assays (Fig S2) Table 1 C-terminal amino acid region alignment of CNBPs from different organisms The sequences shown correspond to the last 29 amino acid residues from each CNBP protein Numbers in Danio rerio sequence represent the amino acid position in the primary structure shown in Fig 1A CCHC zinc knuckles are shaded gray Consensus sequence for PKA is boxed Amino acid residues predicted as phospho-acceptors are shaded black.

Fig 2 In vitro CNBP phosphorylation In vitro phosphorylation assays were performed using the recombinant CNBP (GST-CNBP) and zebrafish embryonic extract prepared from embryos at 24 hpf Phosphorylated GST-CNBP was run in 12% (w ⁄ v) SDS ⁄ PAGE (A) Gel autoradiography (B) Coomassie Blue stained gel Arrow on the right indicates GST-CNBP; asterisk on the right indicates GST protein.

Taken together, these results confirm that CNBP is

phosphorylated by embryonic zebrafish kinases and

allow us to further characterize CNBP

phosphoryla-tion and to identify the candidate kinases

CNBP is differentially phosphorylated in vitro

by protein extracts prepared from embryos at

different developmental stages

To further analyse CNBP phosphorylation, we carried

out in vitro recombinant CNBP phosphorylation using

extracts prepared from embryos at different

develop-mental stages A basal and constant phosphorylation

level was consistently observed at early embryonic

developmental stages until the stage of 8-somite

Beyond the 8-somite stage, CNBP phosphorylation

began to increase, reaching a maximum at 48 hpf

stage Afterwards, phosphorylation level decreased

progressively (Fig 3A,B) returning to a basal level at

5 days postfertilization (dpf) (Fig 3C) Similar results

were obtained when His6-CNBP was analysed

(Fig S3) The radioactivity⁄ protein ratio raised a

maximum of approximately 10-fold at 48 hpf with

respect to the basal phosphorylation level (Fig 3C)

These results confirm the existence of CNBP

phos-phorylation by embryonic kinases and, furthermore,

suggest a transient change in protein kinase activity

during zebrafish development

It is worth mentioning that CNBP phosphorylation

began to increase at the 8-somite stage (Fig 3C), the

developmental stage in which a CNBP subcellular

localization change became detectable in zebrafish as

well as in Bufo arenarum [1,10] Therefore, it is

tempting to speculate that CNBP phosphorylation

would be the post-translational modification that allows CNBP to move from the cytoplasm to the nucleus during early development The role of CNBP phosphorylation in vivo, as well as the putative corre-lation between the nucleo-cytoplasmic translocation and the CNBP phosphorylation status remains to be established These issues are subject for further research and are currently under investigation in our laboratory

CNBP is phosphorylated on Ser⁄ Thr residues

In silico analysis showed that CNBP amino acid sequences possess putative phosphorylation sites mainly on Ser⁄ Thr residues (Fig 1) To confirm this, GST-CNBP was in vitro phosphorylated and then sub-jected to an alkaline treatment This treatment removes most of the phosphates bound to serine or threonine but not to tyrosine residues from a phos-pho-protein [28] Radiolabelled GST-CNBP was run

on duplicated SDS⁄ PAGE gels that were subsequently autoradiographed (Fig 4A) After that, one of the gels was incubated in potassium hydroxide solution while the other was used as control Both gels were again subjected to autoradiography (Fig 4B) and finally stained with Coomasie-blue (Fig 4C) No radioactivity was observed in the alkaline-treated gel while the pro-tein was clearly detected by Coomasie-blue staining, suggesting that CNBP phosphorylation occurs on Ser and⁄ or Thr residues In agreement, specific phospho-tyrosine monoclonal antibody failed to recognize radioactively phosphorylated CNBP, but it detected phospho-proteins from control extracts (Fig 5) These results allow us to rule out CNBP phosphorylation on

Fig 3 Differential in vitro phosphorylation

of CNBP during zebrafish embryogenesis Embryonic extracts from different develop-mental stages (8 cell to 120 hpf) were used

to perform in vitro CNBP phosphorylation Phosphorylated GST-CNBP was run in 12% (w ⁄ v) SDS ⁄ PAGE (A) Gel autoradiography (B) Coomassie Blue stained gel (C) CNBP phosphorylation level ⁄ fusion protein amount ratio (Means ± SEM, n ¼ 3).

Tyr residues and, furthermore, to confirm the data

obtained from phospho-site predictor servers

CNBP is mainly phosphorylated by

cAMP-dependent protein kinase

To identify kinases capable of phosphorylating CNBP,

we carried out in vitro phosphorylation assays using

embryonic extracts previously incubated with specific

kinase inhibitors Ser4 and Thr56 were predicted as

putative amino acid targets of casein kinase 2 (CK2)

and, furthermore, developmental CNBP

phosphoryla-tion profile encompassed profiles previously reported

for CK2 during zebrafish development [29] Thus, we analysed GST-CNBP phosphorylation in the presence

of heparin, a highly selective and potent inhibitor of CK2 [30] In several independent experiments, no signi-ficant heparin phosphorylation inhibition was observed (Fig 6), suggesting that CNBP is not a CK2 substrate

To confirm this, we used purified CK2 subunit alpha (CK2a) enzyme for in vitro phosphorylation assays In all the conditions assayed, CK2a failed to phosphory-late CNBP (Fig S4) Taking into account that embryo-nic development proceeds at high cell-division rates, we analysed inhibition of cdc2, cdk2, erk1, and cdk5

kinas-es by preincubation of S100 embryo extracts with the cell cycle inhibitor olomoucine Figure 6 shows that no statistically significant inhibition was observed when ol-omoucine was assayed, suggesting that CNBP is not a substrate of cell cycle kinases Finally, we studied the effect of H-89 inhibitor on CNBP phosphorylation because Ser158 was predicted as a cAMP-dependent protein kinase (PKA) phosphorylation target We observed that CNBP phosphorylation was significantly inhibited by H-89 (Fig 6) and, furthermore, the inhibi-tion was proporinhibi-tional to the amount of H-89 used (Fig S5) Because most protein kinase C (PKC) iso-types may be inhibited by H-89, we analysed the effect

of the specific PKC inhibitor chelerytrine on CNBP phosphorylation We did not observe a significant decrease in CNBP phosphorylation level, thus, indica-ting that H-89 inhibition was predominantly on PKA activity In agreement, PKI, a specific PKA inhibitor, significantly impaired CNBP phosphorylation (Fig 6) Finally, we carried out in vitro CNBP phosphorylation using purified rat recombinant PKA catalytic subunit a (PKA-Ca) CNBP was effectively phosphorylated

by PKA-Ca and, as expected, phosphorylation was

Fig 5 Analysis of phospho-Tyr residues presence in phosphorylated CNBP In vitro phosphorylation assays were performed using GST-CNBP and zebrafish embryonic extracts prepared from different developmental stages as indicated Proteins were run in 12% (w ⁄ v) SDS ⁄ PAGE and blotted onto nitrocellulose membrane (A) Proteins stained with Ponceau Red (B) Autoradiography of the membrane (C) Tyrosine-phosphorylated proteins visualized by western blotting using monoclonal antibody raised against phosphotyrosine Epidermal growth factor-stimulated A431 cell lysate (Upstate Cell Signalling Solutions) was used as positive control GST was used as negative control Arrow on the right indicates GST-CNBP; asterisk on the right indicates GST protein.

Fig 4 Alkaline treatment of phosphorylated CNBP In vitro

phos-phorylation assays were performed using GST-CNBP and zebrafish

embryonic extracts prepared from different developmental stages

as indicated Proteins were run in 12% (w ⁄ v) SDS ⁄ PAGE (A)

Auto-radiography of untreated gels (B) AutoAuto-radiography of untreated

(left) and treated (right) gels (C) Untreated (left) and treated (right)

Coomassie Blue stained gels.

inhibited by H-89 (see below) Taken together, our data

indicate that PKA is one of the kinases responsible for

CNBP phosphorylation

PKA normally exists in the cytoplasm as an inactive

tretameric holoenzyme that is dissociated into two

catalytic and two regulatory subunits when cAMP is

generated as a consequence of adenylate cyclase

activa-tion Alternatively, PKA activity may be induced by

c-myc, which provides an endogenous activation of the

cAMP signal transduction pathway independent of

extracellular signals [31] A number of reports have

demonstrated that CNBP enhances c-myc transcription

[7,16] If c-myc increases PKA activity and,

conse-quently, CNBP becomes phosphorylated, thus, CNBP

phosphorylation would function as a fine tuning for

c-myc transcriptional regulation

Identification of CNBP phospho-acceptor residue

Zebrafish CNBP amino acid sequence shows a unique

putative PKA site spanning amino acids ARDCSIEAS

(154–162) In this site, the Ser158 is the hypothetical phospho-acceptor residue (Table 1) To evaluate whe-ther CNBP phosphorylation was on Ser158, we per-formed site-directed mutagenesis on CNBP cDNA in order to replace the Ser158 for an Ala residue The CNBPS158Amutant failed to be phosphorylated in vitro

by any of the S100 analysed embryonic extracts (Fig 7), even when using higher amounts of extracts and longer reaction times than those used for wild-type phosphorylation assays To further analyse CNBPS158A phosphorylation, we subjected it to phosphorylation

by PKA-Ca Actually, CNBPS158Awas phosphorylated

by PKA-Ca, and 32P incorporation was inhibited by H-89 (Fig 8A) However, radioactivity incorporation

in CNBPS158A was less than 20% with respect to the radioactivity incorporated in the wild-type protein (Fig 8B) CNBPS158A phosphorylation by PKA might

be due to the high amount of PKA-Ca used in the phosphorylation assay and, consequently, may not be relevant Alternatively, CNBPS158A phosphorylation might occur by conformational changes in the mole-cule due to the mutation that affect the accessibility of other PKA phosphorylation site(s) Thus, these results indicate that CNBP is mainly phosphorylated in Ser158 residue and, moreover, that this phosphoryla-tion is performed by PKA

A comparison of CNBP primary structures from

a variety of species revealed that the amino acid sequence for PKA phosphorylation is evolutionarily conserved (Table 1) While Ser158 residue is found in zebrafish, threonine residues occur in all known verteb-rate sequences at equivalent positions The residues around this phosphorylation site also display a high degree of conservation, which matches with the protein kinase phosphorylation sequence for PKA site This

Fig 7 In vitro phosphorylation of wild-type CNBP and CNBP S158A

by embryonic zebrafish extracts Wild-type GST-CNBP (CNBPWT) and mutant GST-CNBPS158A(CNBPS158A) were in vitro phosphoryl-ated by embryonic extracts from different developmental stages

as indicated Proteins (5 lg) were run in 12% (w ⁄ v) SDS ⁄ PAGE (A) Gel autoradiography (B) Coomassie Blue stained gel.

Fig 6 In vitro phosphorylation assays in the presence of different

specific kinase inhibitors In vitro phosphorylation assays were

per-formed using GST-CNBP and zebrafish embryonic extract from

48 hpf stage in the presence of specific kinase inhibitors as

indica-ted Phosphorylated GST-CNBP (5 lg) was run in 12% (w ⁄ v)

SDS ⁄ PAGE (A) Gel autoradiography (B) Silver stained gel.

(C) CNBP phosphorylation level ⁄ fusion protein amount ratio for

each treatment (Means ± SEM, n ¼ 3) **Indicates P < 0.01 in

respect of controls (one way ANOVA, Tuckey test) Controls:

kin-ase buffer for heparin and PKI-(6-22)-amide; 4% (v ⁄ v)

dimethylsulf-oxide for olomoucine, chelerytrine and H-89 Inhibitors were

grouped according their respective controls.

fact suggests that CNBP phosphorylation may not

only exist in zebrafish, but also in other vertebrates

playing the same biological function

Effect of CNBP phosphorylation on its

biochemical activities

CNBP was reported as single-stranded nucleic acid

binding protein able to recognize RNA as well as

sin-gle-stranded DNA molecules in electrophoretic

mobil-ity shift assays (EMSA) [10] Here we studied the

biochemical consequences of CNBP phosphorylation

on nucleic acid binding by means of EMSAs following

reported protocols [7,10] We assayed two ssDNA

probes representative of the informed CNBP

single-stranded nucleic acid targets, the L4-rp-mRNA

5¢ UTR [9] and the CT element complementary sequence (Comp-CT) from the human c-myc promoter [7] EMSAs were performed using wild-type CNBP and CNBPS158A treated with alkaline-phosphatase (npGST-CNBPWT and npGST-CNBPS158A) or incuba-ted with ATP and embryonic extracts as the in vitro phosphorylation assay conditions (pGST-CNBPWTand pGST-CNBPS158A) Regardless of the evaluated CNBP form or the nature of the analysed probe, band-shift intensity increased along with the protein increasing amount Nevertheless, no significant differences in binding affinities were observed as a consequence of phosphorylation (Fig S6) This finding suggests that

in vitro phosphorylation does not modify the CNBP binding capability for either L4 5¢ UTR or Comp-CT nucleic acid probes Moreover, the conservation of the nucleic acid binding activity suggests that the replace-ment of the Ser residue by an Ala does not signifi-cantly modify the CNBP conformation

The CCHC Zn finger domains found in CNBP are structurally similar and functionally equivalent to those present in retroviral proteins [32] CCHC retro-viral domains are responsible for nucleic acid structure remodelling, a biochemical activity typically tested by analysing the promotion of the annealing of comple-mentary oligonucleotide sequences [33,34] We wonder

if CNBP performs annealing promotion activity and, furthermore, if phosphorylation modifies this bio-chemical activity To this purpose, we analysed CT (sense) and [32P]-5¢ end-labelled Comp-CT (antisense) oligonucleotides annealing profile in the presence of phosphorylated and nonphosphorylated CNBP forms (Fig 9) Oligonucleotides were preincubated with the different CNBP forms and then mixed to let the reac-tions start Aliquots of annealing reaction mixture were taken at specific time points, and annealing reac-tion was stopped as described in Experimental proce-dures Samples were run in native PAGEs, which were followed by autoradiography Radioactive bands were quantified and annealing percentage was determined as indicated in Experimental procedures npGST-CNBPWT and npGST-CNBPS158A as well as pGST-CNBPS158A failed to promote complementary oligonucleotide annealing Conversely, the annealing activity signifi-cantly increased in presence of pGST-CNBPWT GST per se failed to promote complementary oligonucleotide annealing proving that this activity was dependent on phosphorylated CNBP itself

Our results show that CNBP phosphorylation does not change single-stranded nucleic acid binding activity Instead, it promotes the annealing of complementary DNA strands Similar biochemical behaviour was observed for hnRNP Al, a nucleo-cytoplasmic shuttling

Fig 8 In vitro phosphorylation of CNBP WT and CNBP S158A by

PKA-Ca Wild-type GST-CNBP (CNBPWT) and mutant GST-CNBPS158A

(CNBPS158A) were in vitro phosphorylated by PKA-Ca Proteins

(5 lg) were run in 12% (w⁄ v) SDS ⁄ PAGE (A) Autoradiography

(upper) and Coomassie Blue stained (lower) gels for wild-type (left)

and mutant (right) recombinant CNBPs (B) Phosphorylation

level ⁄ protein amount ratio for wild-type and mutant CNBP

subjec-ted to the different treatments Controls: buffer kinase,

dimethyl-sulfoxide (DMSO), and H-89 inhibition for PKA activity specificity.

protein that binds single-stranded nucleic acids and is

able to promote interstrand reannealing of DNA as

well as RNA [35,36] hnRNP Al is phosphorylated

by CK2 and PKA PKA phosphorylation suppresses

the ability of hnRNP Al to promote strand annealing

in vitro, without any detectable effect on its nucleic acid

binding capacity [37] Taken together, it seems that

phosphorylation by PKA would be a post-translational

modification that affects the properties of a variety of

proteins with common biochemical activities, such as

the binding to single-stranded nucleic acids and the

remodelling of their secondary structures

A model involving CNBP has been proposed for

5¢ TOP-mRNA translational control in which the

5¢ UTR is in equilibrium between open and closed

con-formations A closed CNBP-bound form would result

in translation inhibition, while an open La autoantigen

(La)-bound form would, instead, allow translation Cell

signals might influence the affinities of La or CNBP for

the 5¢ UTR, and the alternative binding of these

pro-teins may lead, either alone or together with additional

factors, to differential effects on the translation of

rp-mRNAs [14] It was recently reported that

hypo-phosphorylation of human La protein increases binding

to and shifts the 5¢ TOP-mRNA encoding rpL37 off the

polysome distribution suggesting that increased levels

of nonphosphorylated La protein may have a negative effect on rp-mRNA translation [38] According with our data, CNBP phosphorylation status would not be a determinant for 5¢ TOP-mRNAs binding, but might affect the stability of mRNA conformation through the nucleic acid structure remodelling activity and, conse-quently, might affect the translational control

In the other reported model for CNBP function, CNBP was found binding to the purine-rich strand of the CT element from the human c-myc promoter, opposite to hnRNP K that binds to the pyrimidine-rich strand These interactions determine the formation

of an open complex at the CT element composed of unpaired strands, CNBP, hnRNP K, and additional factors [7] Modulation of the concentrations and⁄ or the biochemical activities of either CNBP or hnRNP

K could influence the ability of the other factors to bind to the CT element and, thus, modulate CT ele-ment activity Here we report that in vitro phosphory-lation does not modify Comp-CT binding capability Instead, it promotes the annealing of the CT and Comp-CT sequences, an activity that may indicate the ability of CNBP to remodel nucleic acid conformation Alteration of the CT element conformation might either facilitate or block the action of mutually exclu-sive transcriptional factors, thereby allowing a single

Fig 9 Effect of CNBP phosphorylation on nucleic acid annealing promotion activity (A) Annealing assay performed without added protein or with pGST-CNBPWT, pGST-CNBPS158A, np-GST-CNBPWTand np-GST-CNBP S158A GST was used as a control Protein concentrations used (0.3 l M ) represent a protein ⁄ probe ratio of 30 : 1 The last two lanes on the right correspond

to mobility controls: labelled Comp-CT alone and a completely annealed reaction Sam-ples were taken at 0, 1, 3, 10, and 30 min Single-stranded (ss) and double-stranded (ds) species are indicated (B) Bar chart representing annealing percentage (%) obtained at 30 min (means ± SEM, n ¼ 3).

**Indicates P < 0.01 in respect of controls (one way ANOVA, Tuckey test).

regulatory sequence to confer different properties upon

nearby promoters Remodelling of nucleic acid

confor-mation has been documented as a possible mechanism

for transcriptional regulation [13] The presence or

absence of a particular single-strand DNA binding

protein could promote or restrict the interaction of a

conventional transcription factor with the promoter

Thus, single-strand loop formation directed by

consti-tutive and regulated sequence-specific DNA binding

proteins may provide an effective tool for customizing

the action of other factors and their upstream

signal-ling systems In the case of c-myc gene, this mechanism

is proposed to proceed through the binding of trans

factors, such as CNBP and hnRNP K, to a

single-stranded DNA hinge on the cis CT element [7,39]

Thus, it is tempting to speculate that CNBP annealing

activity, as a consequence of phosphorylation, would

generate a chromatin remodelling that influences trans

factors binding, directly linking alterations in DNA

conformation and topology with potentially changes in

gene transcription efficiency

The results presented here allow us to hypothesize

that CNBP phosphorylation might be a conserved

post-translational modification that enables CNBP to

perform a fine tune of the expression of a number of

target genes, including c-myc, during vertebrate

embry-ogenesis The role of c-myc in embryogenesis has been

clearly demonstrated in mice [40] and Xenopus [41]

Myc family proteins have been extensively studied for

almost 25 years, and have been implicated in a

pleth-ora of essential cellular processes, including cell

growth, cell proliferation, apoptosis and cellular

differ-entiation [42,43] Therefore, a central focus of the

c-myc field of research is to understand the key control

mechanisms responsible for the synthesis of this

important regulatory protein Although at present we

cannot completely explain the participation of CNBP

phosphorylation in c-myc expression regulation, future

investigations may uncover the involved mechanisms

Experimental procedures

Biological material

Adult zebrafish (Danio rerio) were raised and maintained

on a 14⁄ 10 h light ⁄ dark cycle at 28.5 C and bred in

mar-bled tanks as described [44] Embryos were staged

accord-ing to Kimmel [45]

S100 embryonic extract preparation

Zebrafish embryos from different developmental stages

were homogenized in four volumes of ice-cold buffer kinase

[50 mm Tris⁄ HCl (pH 8), 10 mm MgCl2, 1 mm EGTA,

1 mm phenylmethanesulfonyl fluoride, 1 mm dithiothreitol]

in a Potter-Elvehjem (Thomas, Philadelphia, PA, USA) at

0C Homogenates were centrifuged twice for 15 min at

22 000 g at 4C and then for 60 min at 100 000 g at 4 C Supernatant (S100) was used as cytosolic embryonic extract Protein concentration was estimated as described previously [46]

Generation and expression of recombinant CNBP protein forms

Site-directed CNBPS158Amutant (GenBank accession num-ber DQ519386) was generated from the wild-type cDNA (GenBank accession number AY228240) by PCR amplifica-tion Oligonucleotides used for mutant generation were:

AG-3¢, and reverse 5¢-GGAATTCCTACGCAGACGCTT

CNBP wild-type and CNBPS158A mutant were sequenced and subsequently cloned in pGEX-3X plasmid vector (Amersham Biosciences, Piscataway, NJ, USA) Fusion proteins were expressed in Escherichia coli DH5a and purified by gluthathione sepharose (Amersham Biosciences) affinity chromatography according to the manufacturer’s instructions For in vitro phosphorylation assays, proteins were reloaded into the matrix and stored at 4C in NaCl⁄ Pi containing 1% (v⁄ v) Triton X-100 until used Fusion protein molar concentrations were accurately esti-mated by densitometic analyses of SDS⁄ PAGE Coomassie Blue stained gels

In vitro phosphorylation assays

For in vitro kinase assays, 4–6 lg of GST-fusion proteins bound to gluthathione sepharose beads were incubated with

100 lg of S100 embryonic extract in kinase buffer containing

100 lm ATP and 5 lCi [32P]ATP[cP] at 30C for 5 min Beads were subsequently washed three-fold with 10 volumes

of NaCl⁄ Pi, proteins were eluted, and run in 12% (w⁄ v) SDS⁄ PAGE Phosphorylation of CNBP was visualized by using a GP-Storage Phosphor Screen and storm scanner and software (Amersham Biosciences) and proteins were Coo-massie Blue stained Phosphorylation levels were normalized with respect to the amount of protein loaded For PKA sub-unit catalytic a (PKA-Ca) in vitro phosphorylation, 4–6 lg

of GST-fusion proteins bound to gluthathione sepharose beads were incubated with 0.3 mgÆmL)1PKA-Ca in kinase buffer containing 50 lm ATP and 5 lCi [32P]ATP[cP]

Alkaline treatment

Alkaline treatment was performed essentially as described [28] Briefly, phosphorylated proteins were run in two

similar 12% (w⁄ v) SDS ⁄ PAGE, exposed to a GP-Storage

Phosphor Screen and visualized using a storm scanner and

software (Amersham Biosciences) One of the gels was

incu-bated with 1 m KOH at 55C for 2 h Then, both gels were

re-exposed and subsequently Coomasie Blue stained

Western blotting

For western blot analysis, 4 lg of radioactively

phosphor-ylated GST-CNBP by different zebrafish embryonic

extracts as indicated, 4 lg of GST, and 4 lg of antigen

control (epidermal growth factor-stimulated A431 cell

lysate; Upstate Cell Signaling Solutions, Lake Placid,

NY, USA) were separated by 12% (w⁄ v) SDS ⁄ PAGE

electrophoresis and blotted onto nitrocellulose membrane

The monoclonal antiphosphotyrosine, clone 4G10 (Upstate

Cell Signalling Solutions) was used as primary antibody

Immunoblots were performed according to the

manu-facturer’s instructions Peroxidase-conjugated antimouse

secondary antibody (Amersham Bioscience) was used

and developed using chemiluminescence (ECL, Amersham

Bioscience) according to the manufacturer’s instructions

Kinase inhibition assays

One hundred micrograms of S100 extract from 48 hpf

embryos were incubated during 15 min with different kinase

inhibitors at 30C and then used to analyse in vitro

recom-binant CNBP phosphorylation in kinase buffer containing

100 lm ATP and 5 lCi [32P]ATP[cP] at 30C for 5 min

Kinase inhibitors analysed (Sigma Aldrich, St Louis, MO,

USA) were olomoucine (1 mm final concentration); H-89

(N-[2-(p-bromocinnamylamino)ethyl]-5-isoquinoline

sulfona-mide) (200 nm final concentration); chelerytrine (200 lm

final concentration), heparin (500 lgÆmL)1 final

concentra-tion), and PKI (10 nm final concentration) Except for

heparin and PKI kinase phosphorylation inhibition assays,

4% (v⁄ v) dimethylsulfoxide was added to kinase reaction

mixture as control

Electrophoretic mobility shift assay (EMSA)

Binding reactions were performed in 20 mm Hepes, pH 8,

10 mm MgCl2, 1 mm EDTA, 1 mm dithiothreitol,

1 lgÆlL)1 BSA, 0.01 lgÆlL)1 of nonspecific competitor

DNA [poly(dI:dC)Æ(dC:dI) from Pierce Nucleic Acid

Tech-nologies, Milwaukee, NY, USA], 0.5 lgÆlL)1 heparin and

10% (v⁄ v) glycerol Boiled [32P]-5¢ end-labelled probes were

added to a final concentration of  2 nm, and accurate

molar amounts of fusion proteins were added as indicated

Final reaction volume was 20 lL Binding reactions were

incubated for 30 min at 37C and then loaded onto 8%

(w⁄ v) polyacrylamide gels containing 5% (v ⁄ v) glycerol in

Tris-Borate-EDTA ·0.5 After electrophoresis, gels were

dried and autoradiography visualized by using GP-Storage Phosphor Screen and storm scanner and software (Amer-sham Biosciences) Single-stranded probes used were L4-UTR representing the 5¢ UTR from Xenopus laevis L4 rp-mRNA (5¢-CCTTTTCTCTTCGTGGCCGCTGTGGAG AAGCAGCGAGGAGATG-3¢), and Comp-CT represent-ing the complementary (antisense) strand of the CT element from the human c-myc promoter (see below)

Strand annealing assays

Oligonucleotide annealing assays were performed essentially

as previously described [47] Probes for this assay were oligonucleotides representing the CT element from the human c-myc promoter: CT (sense): 5¢-CACCCTCCCCAC CCTCCCCATAAGCGCCCCTCCCGGGTTCCCAAA-3¢; and Comp-CT (antisense): 5¢-TTTGGGAACCCGGG AGGGGCGCTTATGGGGAGGGTGGGGAGGGTG-3¢

CT (10 nm) and [32P]-5¢ end-labelled Comp-CT (5 nm) oligonucleotides were separately denatured by heating and then independently preincubated at 37C in 50 mm Tris⁄ HCl, pH 8.0, 0.1 mm dithiothreitol, 6 mm MgCl2,

80 mm KCl, and 0.1 mm ZnCl2 in the absence or presence

of different GST-CNBP forms for 2 min, at the indicated accurate molar amounts To start the reactions, the prein-cubated probes were mixed and aliquots were taken at spe-cific time points as indicated Annealing reaction was stopped by adding stop solution [0.25% (w⁄ v) bromphenol blue, 0.25% (w⁄ v) xilene cyanol, 20% (v ⁄ v) glycerol,

20 mm EDTA, 0.2% (w⁄ v) SDS, and 0.4 mgÆmL)1 yeast tRNA] Reactions were incubated in stop solution at 37C for 1 min before being transferred to ice, and then subjec-ted to 15% native PAGE in Tris-Borate-EDTA 1· Gels were dried and autoradiography visualized by using GP-Storage Phosphor Screen and storm scanner and software (Amersham Biosciences) Radioactive bands were quantified and annealing percentage was determined by dividing the amount of annealed oligonucleotide (ds) by the total amount [annealed plus single-stranded (ss)] in each lane, and multiplying by 100 [% annealing¼ 100 · ds ⁄ (ds + ss)] Mobility controls for the oligonucleotides were the not annealed labelled Comp-CT probe alone, and the completely annealed labelled Comp-CT probe heated and slowly cooled in the presence of unlabelled CT probe

Acknowledgements

We thank Dr J Allende for the generous gift of zebra-fish CK2 clone and Dr S Moreno for the generous gift

of rat recombinant PKACa and PKI inhibitor This work was supported by ANPCyT PICT 01-8754, CONICET PIP N 03073, and Fundacio´n Josefina Prats grants AW, VL, and PA are Fellows of CONI-CET; NBC is CONICET staff member