CNBP regulates wing development in Drosophila melanogaster by promoting IRES-dependent translation of dMyc

Abstract

CCHC-type zinc finger nucleic acid binding protein (CNBP) is a small conserved protein, which plays a key role in development and disease. Studies in animal models have shown that the absence of CNBP results in severe developmental defects that have been mostly attributed to its ability to regulate c-myc mRNA expression. Functionally, CNBP binds single-stranded nucleic acids and acts as a molecular chaperone, thus regulating both transcription and translation.

 

In this work we report that in Drosophila melanogaster, CNBP is an essential gene, whose absence causes early embryonic lethality. In contrast to what observed in other species, ablation of CNBP does not affect dMyc mRNA expression, whereas the protein levels are markedly reduced. We demonstrate for the first time that dCNBP regulates dMyc translation through an IRES-dependent mechanism, and that knockdown of dCNBP in the wing territory causes a general reduction of wing size, in keeping with the reported role of dMyc in this region. Consistently, reintroduction of dMyc in CNBP-deficient wing imaginal discs rescues the wing size, further supporting a key role of the CNBP-Myc axis in this context.

Collectively, these data show a previously uncharacterized mechanism, whereby, by regulating dMyc IRES-dependent translation, CNBP controls Drosophila wing development. These results may have relevant implications in other species and in pathophysiological conditions.

Introduction

CCHC-type zinc finger nucleic acid binding protein (CNBP, also named ZNF9) is a conserved single-stranded DNA binding protein of 19 kDa.¹ Mammalian CNBP contains 7 highly conserved CCHC zinc finger motifs, an arginine/glycine-rich domain (RGG), a putative PEST, nuclear localization sequence, and PKA phosphorylation site, and its mRNA is expressed in most cells and tissues, with higher levels in highly proliferating cells.²

Functionally, CNBP binds single-stranded DNA and RNA and promotes rearrangements of their secondary structure, thus acting as a nucleic acid chaperone.³ CNBP may regulate both transcription and translation, by affecting chromatin or RNA structural remodeling. It was originally identified as a transcriptional factor, bound to the sterol regulatory element (SRE) of the HMGCoA gene, involved in cholesterol metabolism.⁴ As a transcriptional regulator, CNBP has been also described as both inhibitor and enhancer of transcription of some genes, such as β-myosin heavy chain, and the macrophage colony-stimulating factor.^5,6 Most notably, CNBP was shown to regulate transcription of c-myc, through its ability to bind and promote the formation of G quadruplexes (G4) in the c-myc promoter.^1,7

CNBP has been also described as a translational regulator, acting with 2 distinct mechanisms. The first one is the enhancement of translation upon binding to the G-rich sequences downstream of the TOP regions of several mRNAs encoding ribosomal proteins and elongation factors, thus increasing the rate of global translation.^8,9 Alternatively, CNBP has been shown to promote IRES-dependent translation of the ornithine decarboxylase (ODC) mRNA, thus acting as an ITAF member.^10,11 The ability to regulate IRES-dependent translation has been observed in human and yeast CNBP,¹² suggesting a high conservation of this mechanism throughout the species.

A mutation in the gene encoding CNBP has been found in patients with myotonic distrophy type 2 (DM2) and consists of an expansion of CCTG repeats in the first intron.¹³ Studies in these patients have reported an impairment of global protein synthesis and altered IRES-dependent translation.^9,11

Genetic studies in different species have highlighted the essential role of CNBP during embryonic development. Mice lacking both alleles of the CNBP gene die around E10.5, and embryos show severe forebrain truncation and facial abnormalities.¹⁴ A similar phenotype is also observed in chicken¹⁵ and zebrafish¹⁶ upon CNBP depletion, indicating the conserved role in controlling embryonic development. Associated to CNBP depletion is a decrease of cell proliferation, which has been mainly attributed to the ability of CNBP to regulate c-myc expression.

In the present work, we have analyzed the structure and function of CNBP in Drosophila melanogaster, a model where the function of CNBP has not been investigated yet. We have observed that CNBP is highly conserved in fly and, in analogy with other species, it is essential for development. Importantly, we have found that dCNBP promotes IRES-dependent translation of diminutive (dm), the c-myc Drosophila ortholog (herein called dMyc), and that targeted knockdown of dCNBP in the wing results in a reduction in size similar to dMyc phenotype.

Results

The structure and function of CNBP is conserved in Drosophila melanogaster

Amino acid alignment of Drosophila and human CNBP showed a high degree of sequence similarity between the 2 proteins. According to ClustalW algorithm, among 74% conserved amino acids, 43% are identical, 15.7% strongly conserved, and 15.1% weakly conserved (Fig. 1A).

Figure 1. CNBP is conserved in Drosophila melanogaster. (A) ClustalW alignment of human and D. melanogaster CNBP gene products. Identical, strongly conservative and weakly conservative aminoacids are indicated by asterisk, colon, and dots, respectively, according to ClustalW convention. (B) Schematic representation of CNBP in Homo sapiens and Drosophila melanogaster. CNBP contains 7 conserved CCHC Zn-finger repeats in Homo sapiens and 6 in Drosophila; the arginine/glycine-rich (RGG) box is conserved in both species.

Primary sequence analysis of Drosophila CNBP (dCNBP) revealed the presence of 6 CCHC zinc finger sequences and a putative RGG box, while the mammalian orthologs contain 7 CCHC domain and 1 RGG box (Fig. 1B; ref. 1).

The high degree of conservation of dCNBP suggests that its physiological function should also be conserved. To determine whether dCNBP plays a role in development, we used 2 VDRC UAS-RNAi transgenic lines, carrying a GAL4-inducible dCNBP-RNAi construct against the CNBP coding gene (GC3800), and crossed them with driver lines expressing the GAL4 under the control of 3 different ubiquitous promoters: tubulin, actin, and 69B. As shown in Figure 2A, when crossed with tubGAL4, expression of CNBP RNAi led to a partial depletion of the protein, whereas the combination of 2 copies UAS-dCNBP-RNAis resulted in a complete loss of dCNBP protein. Notably, the combination of single or double CNBP RNAi lines with any of the 3 GAL4-expressing transformants always resulted in lethality before the pupal stage (Fig. 2B). Thus, in analogy with other species, CNBP deficiency is developmentally lethal, even with incomplete ablation of the protein.

Figure 2. dCNBP is essential for development. (A) RNAi efficiency in larvae expressing a single (dCNBP^RNAi) or double copy (2XdCNBP^RNAi) of dCNBP RNAi, under the constitutively active tubulin expression driver. (B) Phenotypic consequences of RNAi-mediated inactivation of dCNBP using the reported VDRC UAS-RNAi lines in combination with either tubGAL4, actGAL4, or 69BGAL4. (C) dCNBP and dMyc mRNA levels (QPCR), normalized with the housekeeping RpL32 mRNA in wild-type and tubGAL4 double copy (2XdCNBP^RNAi) dCNBP RNAi-expressing larvae. **P < 0.05 vs. WT. (D) Effect of dCNBP depletion (2XdCNBP^RNAi) on dMyc expression in WT or tubGAL4-driven UAS- dCNBP^RNAi-expressing larvae. Actin loading control.

dCNBP regulates IRES-dependent translation of dMyc

Gene deletion studies in mice, chicken, Xenopous, and zebrafish have revealed that the absence of CNBP is associated with a lower proliferation rate that has been mainly attributed to the downregulation of c-myc expression.¹ Since it has been described that CNBP induces c-myc transcription in vertebrates,⁷ we wondered whether dMyc mRNA levels were downregulated in the absence of dCNBP. Surprisingly, we found that in larvae ubiquitously expressing dCNBP RNAi, the levels of dMyc mRNA were not changed (Fig. 2C), whereas dMyc protein levels were strongly reduced (Fig. 2D), suggesting a post-transcriptional regulation. As a control, we tested dCNBP mRNA and protein levels, which were both reduced upon knockdown (Fig. 2C and D).

CNBP has been characterized as an IRES trans-acting factor (ITAF) with the ability to promote IRES-dependent translation of the ODC mRNA.¹¹ Thus, since c-myc can be translated from internal initiation sequences,¹⁷ we hypothesized that dCNBP could promote IRES-dependent translation of dMyc.

Indeed, in analogy with other IRES-containing mRNAs, dMyc 5′UTR is longer than the average UTR length, it is rich in GC nucleotides, and appears to have a stable secondary structure.¹⁸

We first tested if the translational function of dCNBP is conserved in Drosophila by studying its co-sedimentation with polyribosome fractions. To this end, we performed a polysomes pelleting of wild-type embryo lysates. As shown in Figure 3A, dCNBP was associated to the polysome pellet, together with the ribosomal RpS6 protein, and this association was prevented by EDTA treatment that separates ribosome into their subunits (Fig. 3B).

Figure 3. Translational control of dMyc by CNBP (A) dCNBP levels in supernatant (super) and polysome fractions (pellet) of WT embryos extracts after ribosome pelletting (B) Polysomal recruitment of dCNBP in non-treated (Ctr) and EDTA-treated embryos lysates. To assess the specificity of protein co-fractionation, 100 mM EDTA, pH 7.4 was added to cell extracts and loaded onto a 100 mM EDTA (+EDTA) sucrose cushion. Distribution of ribosomal proteins along the gradient and purity of fractions controlled with RpS6 and Vinculin staining, respectively. (C) Analysis of dCNBP association with polysome in S2 cells. Cell lysate was separated on 15–50% sucrose gradients. The presence of dCNBP protein in each fraction was analyzed by western blotting. Distribution of ribosomal proteins along the gradient was controlled with RpS6 staining. (D) dCNBP controls IRES-dependent translation of dMyc. IRES activity in HEK293T cells transfected with dMyc-Luc or Empty-Luc reporters and dCNBP or Empty expression vectors as indicated. *P < 0.01, dMyc-Luc vs. Empty-Luc, **P < 0.01 CNBP vs. Empty. Results are shown as the average ± SD of triplicate experiments (n = 4).

To further study the association of dCNBP to polysomes, we performed sucrose fractionation of lysates from Drosophila S2 cells. dCNBP was detected in the polysomes and subpolysome fractions and co-purified with the ribosomal protein RpS6 (Fig. 3C). Therefore, these data demonstrated that dCNBP associates to active polyribosomes.

We next tested whether dCNBP can regulate IRES-dependent translation of dMyc. To this end, we used a reporter plasmid expressing a single bicistronic transcript containing the Renilla reporter coding sequence, translated via a cap-dependent mechanism, and a downstream Luciferase reporter open reading frame, which can be translated only in the presence of an upstream IRES sequence.¹¹ We cloned the 5′UTR fragment of dMyc, between the 2 cistrones and generated the dMyc-Luc reporter (Fig. 3D, bottom). Compared with the empty bicistronic vector, dMyc-Luc vector showed a significant basal IRES activity (Fig. 3D, top), thus demonstrating that the 5′UTR region of dMyc can be translated through an IRES-mediated mechanism. Importantly, this activity was further enhanced by 3-fold upon overexpression of dCNBP (Fig. 3D, top), demonstrating that dCNBP increases IRES-dependent translation of dMyc mRNA.

dCNBP regulates wing development

We next studied if this newly identified ability of dCNBP to upregulate dMyc may affect proper development of a district where this transcription factor plays an established role.

To this end, we knocked down dCNBP in wing imaginal discs, a territory where an appropriate level of dMyc expression is necessary to achieve a proportionately correct size.¹⁹

We crossed the nubGAL4 line, expressing GAL4 in the wing pouch, with dCNBP RNAi lines carrying either 1 or 2 RNAi constructs (2×dCNBP^RNAi). In agreement with the role played by dMyc, we observed a significant reduction of wing size with each single RNAi lines (Fig. 4A and data not shown). Expression of the 2 RNAis in combination resulted in a more pronounced reduction of wing size, associated to loss of patterning elements such as veins (Fig. 4A). Thus, in analogy with dMyc deletion, knockdown of dCNBP reduces wing size. To further validate the involvement of dMyc in this context, we reintroduced dMyc in dCNBP-deficient wings. As shown in Figure 4B, reintroduction of dMyc rescued the effect of dCNBP knockdown by partially restoring normal size of the wings. Thus, these data supported the hypothesis that dCNBP is involved in the control of wing size by regulating dMyc levels.

Figure 4. Expression of dMyc rescues undergrowth caused by dCNBP depletion by RNAi-mediated knockdown. Adult wings of flies expressing the indicated UAS-constructs under the control of the nubGAL4 driver. (A) nubGAL4-mediated expression of UAS-dCNBP^RNAi causes a reduction in size; expression of the 2 RNAis in combination (2xdCNBP^RNAi) causes stronger reduction of wing size and loss of patterning elements, such as veins. (B) Reconstitution of dMyc expression in 2xdCNBP^RNAi wings results in a partial rescue of tissue growth. Wings areas on the indicated media were measured using ImageJ (National Institutes of Health); error bars represent SEM (n ≥ 15). *P < 0.05.

Discussion

In the present work we have characterized the function of dCNBP in regulating embryonic and wing development in Drosophila melanogaster.

The amino acid sequence of dCNBP is highly conserved and shares most of the relevant domains with members of other species, thus suggesting that the protein retains the same molecular functions of the other orthologs. One of such function is the ability to promote IRES-dependent translation, a feature that appears to be conserved even in Gis2p protein, the CNBP ortholog in S. cerevisiae, which lacks the RGG domain.¹² The conserved translational function of dCNBP is supported by the evidence that the protein co-fractionates with actively translating ribosomes and increases the rate of IRES-dependent translation.

With the exception of yeast Gis2p, ablation of CNBP in different species results in embryonic lethality, with severe defects of forebrain formation. Structure–function relationship studies have highlighted the importance of the N-terminal domain (i.e., the first Zn knuckle and the RGG domain) in promoting the biological function of CNBP, mainly attributed to changes in cell proliferation in particular territories.² In agreement with this observation, yeast Gis2p, which is the only CNBP ortholog lacking the RGG domain, does not seem to be essential for cell replication. Conversely, in keeping with gene ablation studies in other species, we have observed that knockdown of dCNBP, which has a highly conserved N-terminal region, causes lethality during development.

The role of CNBP during development has been mainly linked to its ability to promote c-myc mRNA expression. However, in the present work, we have observed that the mRNA levels of dMyc were not changed in CNBP-deficient larvae, while the protein levels were strongly downregulated. We have demonstrated that this is linked to the ability of dCNBP to enhance IRES-dependent translation of dMyc. This observation is supported by studies using a bicistronic expression plasmid containing the 5′UTR IRES of dMyc, which is robustly induced by overexpression of dCNBP.

Supporting the identified role in regulating dMyc protein levels, ablation of dCNBP in the wing discs results in reduction of the wing size, which is partially restored upon reintroduction of dMyc. While we cannot rule out the involvement of other targets regulated by dCNBP, these data document a relevant biological role played by the dCNBP-dMyc axis during Drosophila development. It would be interesting to study whether CNBP regulates IRES-dependent translation of c-myc in other species as well, thus adding a further level of complexity to the current model of CNBP function.

Another relevant evidence stemming from these data is that the IRES-dependent regulation of myc seems to be an evolutionary conserved mechanism. In this regard, it has been reported a key role of IRES-dependent translation of c-myc during mouse embryogenesis.²⁰ Therefore, our data further support the important developmental role of the IRES-dependent control of myc and link this regulation to CNBP, another critical conserved regulator of developmental processes.

In summary, in this work we have demonstrated for the first time the essential role of dCNBP during Drosophila development. By identifying the ability of dCNBP to promote IRES-dependent translation of dMyc, we have unveiled a novel mechanism of regulation of a critical growth regulator, which may have potential relevant pathophysiological implications.

Materials and Methods

Cell culture, transfection, and bicistronic assay

HEK293T cells were cultured as previously described.²¹ S2 cells were grown in complete Schneider medium at 25 °C.

HEK293T cells were transfected using Lipofectamine 2000 Reagent (Invitrogen) according to the manufacturer’s protocol.

Bicistronic assays were performed measuring luciferase activity with the Promega dual-luciferase assay system, as described previously.^21,22

The plasmid pcDNA3.1CMV-renilla-TAA-luciferase (Empty-Luc) was provided by AJ Link (Venderbilt University).

pcDNA3.1CMV-renilla-TAA-IRES^dMyc-luciferase (IRES-containing regions of dMyc, dMyc-Luc) plasmid was cloned by inserting the 5′UTR-region of dMyc mRNA (Flybase CG10798) in the Empty-Luc plasmid; dMyc 5′UTR was amplified from Drosophila melanogaster cDNA, with the following primers:

Fw: 5′ –ATACCATGGG CTGTTGATTT GACTGGGTG-3′

Rev: 5′ –ATACCATGGC ATTGCGATTA TGTTGTCTGG G-3′

Ligation was performed after cutting dMyc 5′UTR, and the Empty-Luc vector with PvuII and NcoI enzyme (New England Biolabs).

The mammalian expression vector pcDNA3HA-dCNBP was obtained by PCR-amplification of the CNBP homolog from Drosophila melanogaster cDNA.

RNA extraction and quantitative PCR

Total RNA was extracted from wild-type or RNA-interference-expressing third instar larvae with TRI Reagent (AM9738, Ambion), reverse transcribed, and analyzed by quantitative real-time PCR with the custom designed Syber Green oligonucleotide listed below:

dMyc QPCR_Fw: 5′-CCGAAAGCGA CTGGAAAGC;

dMyc QPCR_Rev: 5′-TGAGTTATAT GCCTCGCTGA ACA;

RpL32 QPCR_Fw: 5′-CAAGAAGTTC CTGGTGCACA AC

RpL32 QPCR_Rev: 5′-AAACGCGGTT CTGCATGAG

dCNBP QPCR_Fw: 5′-GAGGCCGTCA ACGAACGT

dCNBP QPCR_Rev: 5′-CGCAGCCGTA GCAAGTCTTC

Immunoblot and antibodies

For immunoblot analysis, third instar larvae were collected, lysed with 2× Laemli Reducing Buffer, incubated for 10 min at 98 °C, and centrifuged for 5 min at 10 000 g. The extracts were analyzed by SDS-PAGE as previously described.²¹

The following antibodies were used: polyclonal anti-CNBP (ab48027 AbCam, 1:500), anti-S6 Ribosomal Protein (Cell Signaling #2317 D. melanogaster 1:500), anti-Myc Antibody (d1–717) (sc-28207 SantaCruz 1:500), anti-Vinculin (MH24, developed by Robert H Waterston, and obtained from DSHB).

Polysome analysis

S2 cells were lysed with ice-cold TKM Lysis Buffer (100mM TRIS-HCl pH 7.4, 100 mM KCl, 10 mM MgCl₂, 0.1% TritonX-100, 1 mM DTT, 40 mg/ml pepstatin, leupeptin, and aprotinin, 0.5 mM PMSF, RNAguard [Promega] 2.5 U/μl), incubated 5 min on ice and centrifuged at 13 500 g for 5 min. Supernatants were loaded onto 15–50% sucrose gradients in TNaM buffer (30 mM TRIS-HCl, pH 7.4, 100 mM NaCl, and 5 mM MgCl₂) and centrifuged for 110 min in a Beckman SW41 rotor at 37 000 rpm at 4 °C. Gradients were automatically collected, monitoring the optical density at 260 nm, and the fractions were precipitated for protein analysis as described.²³ Distribution of ribosomal proteins along the gradient and purity of fractions was controlled with RpS6 western blot.

Ribosome pelleting experiments were performed as described previously.¹¹ Wild-type control flies were incubated at 25 °C, and embryos were collected on apple plates (cleared apple juice 25%, sucrose 2,5 %, agar 2,5%) for 16–18 h. Embryos were washed twice with washing buffer (NaCl 0,7% TritonX-100 0,04%) and dechorionated with 13% sodium hypoclorite for 3 min. After 3 washes with washing buffer, embryos were dried and lysed with ice-cold TKM lysis buffer, incubated 8 min on ice, and centrifuged at 11 300 rpm for 5 min. The supernatants were loaded onto a 1 M sucrose cushion (−/+100 mM EDTA) and centrifuged at 38 000 g for 1 h in a rotor type 80Ti.

Distribution of ribosomal proteins along the fractions was controlled with RpS6 and Vinculin western blot.

Fly stocks

The RNAi transgenic fly lines were obtained from the Vienna Drosophila RNAi Center (VDRC) (Dietzl et al., 2007). Transgenic UAS-dCNBP^RNAi fly lines obtained from the VDRC (ID 16283 and 16284) were crossed to nubGAL4, tubGAL4, actGAL4, or 69BGAL4 driver flies and then incubated at 29 °C or 25 °C. The named 2xdCNBP^RNAi stock consisted of a line carrying 2 UAS-dCNBP^RNAi constructs (ID 16283 and 16284) inserted into different chromosomes. All the driver lines used have been previously described. The nubGAL4 was a kind gift from M Mìlan; the other driver lines and the UAS-dMyc line (#9675) used were obtained from Bloomington stock center (http://flystocks.bio.indiana.edu/).

Statistical analysis

Statistical analysis was performed using StatView 4.1 software (Abacus Concepts). Results are expressed as mean ± s.d. of at least 3 separate experiments, each performed in triplicate. Statistical differences were analyzed with the Mann–Whitney U test for non-parametric values and a P < 0.05 was considered significant.

Glossary

Abbreviations:

CNBP

CCHC-type zinc finger nucleic acid binding protein

dm

diminutive

DM2

myotonic distrophy type 2

EDTA

Ethylenediaminetetraacetic acid

IRES

internal ribosome entry site

ITAF

IRES trans-acting factor

ODC

ornithine decarboxylase

PEST

proline, glutamate, serine, and threonine domain

RGG

arginine glycine rich domain

RpL32

60S ribosomal protein L32

RpS6

ribosomal protein S6

TOP

terminal oligopyrimidine

VDRC

Vienna Drosophila RNAi Center

10.4161/cc.27268