Báo cáo khoa học: Two functionally redundant isoforms of Drosophila melanogaster eukaryotic initiation factor 4B are involved in cap-dependent translation, cell survival, and proliferation doc

Eur. J. Biochem. 271, 2923–2936 (2004) Ó FEBS 2004

doi:10.1111/j.1432-1033.2004.04217.x

Two functionally redundant isoforms of Drosophila melanogaster
eukaryotic initiation factor 4B are involved in cap-dependent
translation, cell survival, and proliferation
´
´
´
Greco Hernandez1, Paula Vazquez-Pianzola1, Andreas Zurbriggen2, Michael Altmann2, Jose M. Sierra3
and Rolando Rivera-Pomar1
1

Max-Planck-Institute fuăr biophysikalische Chemie, Goăttingen, Germany; 2Institut fuăr Biochemie und Molekularbiologie,
Universitaăt Bern, Switzerland; 3Centro de Biologı´a Molecular ‘Severo Ochoa’, Universidad Auto´noma de Madrid, Spain

Eukaryotic initiation factor (eIF) 4B is part of the protein
complex involved in the recognition and binding of mRNA
to the ribosome. Drosophila eIF4B is a single-copy gene
that encodes two isoforms, termed eIF4B-L (52.2 kDa) and
eIF4B-S (44.2 kDa), generated as a result of the alternative
recognition of two polyadeynlation signals during transcription termination and subsequent alternative splicing of
the two pre-mRNAs. Both eIF4B mRNAs and proteins are
expressed during the entire embryogenesis and life cycle.
The proteins are cytoplasmic with polarized distribution.
The two isoforms bind RNA with the same affinity. eIF4B-L

and eIF4B-S preferentially enhance cap-dependent over
IRES-dependent translation initiation in a Drosophila cellfree translation system. RNA interference experiments
suggest that eIF4B is required for cell survival, although

only a modest reduction in rate of protein synthesis is
observed. Overexpression of eIF4B in Drosophila cells in
culture and in developing eye imaginal discs promotes cell
proliferation.

The control of mRNA translation is a central process in
the regulation of gene expression. Regulation of mRNA
translation preferentially takes place at the initiation level,
and mRNA binding to the ribosome is a rate-limiting step in
translation initiation [1]. Translation initiation begins with
the recognition of the 5¢-UTR of an mRNA by proteins that
catalyze the landing of 40S ribosomes such as the eukaryotic
initiation factor (eIF)4F complex, which is composed of
factors eIF4E, eIF4A and eIF4G [2]. Other initiation factors
such as eIF1, eIF1A, eIF4B, and eIF5 are also required.
eIF4E recognizes the cap structure (m7GpppN) at the
5¢-UTR end of cellular mRNAs. eIF4A is an ATPase/RNA
helicase which is thought to unwind secondary structure at
the 5¢-UTR of mRNAs to help 40S ribosomes to scan to the
initiation codon. eIF4G is a scaffold/adaptor protein which
binds to the cap-binding protein eIF4E, as well as to eIF4A
and to further factors such as poly(A)-binding protein
(PABP) and ribosome-associated eIF3 [2]. For picornaviral
mRNAs and some cellular mRNAs, 5¢-UTR recognition
occurs independently of the cap structure and is mediated

by an internal ribosome entry site (IRES) located in
proximity to the initiation codon [3].
eIF4B was identified, and its genes were cloned from
mammals [4,5], yeast [6,7], and wheat [8,9]. The role of

eIF4B in the initiation of translation is not well understood.
Its main function is assumed to be in the scanning process
because eIF4B stimulates ATP-dependent unwinding of
the mRNA 5¢-UTR by eIF4F/eIF4A. eIF4B transiently
associates with eIF4F [4,10,11] and stimulates the ATPdependent RNA-helicase activity of eIF4A and eIF4F
in vitro [7,10,12–15]. eIF4B is able to anneal complementary
RNA strands in the absence of eIF4A [16,17]. It binds
nonspecifically to RNA via an RNA recognition motif
(RRM) located at the N-terminus and also via a basic
amino-acid sequence located at its C-terminus [5–7,11,
16–18]. There is also evidence that eIF4B might facilitate
binding of 40S ribosomes to the mRNA. The RRM of
mammalian eIF4B specifically binds to 18S rRNA [19]. The
central part of the mammalian protein contains a region
rich in Asp, Arg, Tyr and Gly (DRYG), which, in addition
to mediating homodimerization, is required for its interaction with eIF3 [20]. The N-terminus of human eIF4B
mediates the interaction with PABP [21]. Genetic evidence
[6,22] and in vitro experiments [23,24] support the model
that eIF4B bridges eIF3 to the 40S subunit. Additional
experiments suggest that mammalian eIF4B binds to
picornaviral IRESs and is involved in IRES-dependent
translation of these mRNAs, although this interaction is
apparently not essential for IRES-dependent translation
[24–29]. Yeast eIF4B is not essential, because knockout
strains are viable but exhibit a temperature-sensitive and
cold-sensitive phenotype [6,7]. In addition, a new initiation
factor termed eIF4H has been identified in mammals. This

Correspondence to R. Rivera-Pomar, Centro Regional de Estudios
´

Genomicos, Av. Calchaqui km 35, 500, 1888-Florencio Varela,
Argentina. Fax: + 54 (11) 4275 8379, Tel.: + 54 (11) 4275 8100,
E-mail:
Abbreviations: eIF, eukaryotic initiation factor; PABP, poly(A)binding protein; IRES, internal ribosome entry site; RRM,
RNA recognition motif; GST, glutathione S-transferase; RNAi,
RNA interference.
(Received 11 October 2003, revised 1 May 2004,
accepted 14 May 2004)

Keywords: cell survival; Drosophila; eukaryotic initiation
factor 4B (eIF4B); proliferation; translation.


´
2924 G. Hernandez et al. (Eur. J. Biochem. 271)

factor shows sequence similarity to and functional conservation with human eIF4B [30]. Schizosaccharomyces pombe
SCE3 encodes an RNA-binding protein involved in cell
division which has been found to share high sequence
similarity with human eIF4B [31].
In Drosophila, the mechanism of translational control is
not well understood because of the lack of detailed
information about the functional contribution of individual
components of the translational machinery. Only some of
the factors involved in the initiation of translation have been
characterized, including eIF4E [32,33], eIF4G [34,35],
eIF4A [36] and eIF2 [37]. The completion of the Drosophila
genome project [38] has opened up the possibility of
identifying most if not all translation factors. Here we
report the characterization of Drosophila melanogaster

eIF4B gene. We show that, in contrast with other analysed
organisms, Drosophila possesses two eIF4B isoforms encoded by a single gene. Both eIF4B isoforms are expressed
throughout development, bind RNA, and stimulate capdependent synthesis of proteins in a redundant manner. We
also show that Dm-eIF4B is required for cell survival and
that it stimulates cell proliferation.

Experimental procedures
Construction of plasmids
PCR amplification was performed on the EST LD09953 to
introduce BamHI sites flanking the ORF of Dm-eIF4B-L.
The PCR product was cloned into pGEM-T (Promega) to
create pGEMT-4BL, and into the BamHI site of the vectors
pGEX-2T and pGEX-6P2 (Amersham Pharmacia Biotech)
to create pGEX2T-4BL and pGEX6P2-4BL, respectively.
Derivatives corresponding to Dm-eIF4B-S were generated
by mutagenesis on Dm-eIF4B-L plasmids using the QuikChange Site-Directed Mutagenesis Kit (Stratagene) and the
primers 5¢-GTGTCCAGGTGAATAACAGCTGGACG
AGGAA-3¢ and 5¢-TTCCTCGTCCAGCTGTTATTCA
CCTGGACAC-3¢ to introduce a stop codon in nucleotides 1171–1173 of the Dm-eIF4B-L ORF. The yeast
plasmids pTRP1-4BL, pTRP1-4BS and pHIS3-4A were
generated by inserting Dm-eIF4B-L, Dm-eIF4B-S, and
eIF4A [36] ORFs into the unique BamHI site of p301TRP1/GAL and p301-HIS3/GAL [7]. The 5¢-UTR cDNA
sequence of Ultrabithorax (Ubx) [39] and caudal [40] were
cloned into the EcoRI site of pBluescript (Invitrogen) to
create the constructs pBS-Ubx and pBS-Cad, respectively.
The 5¢-UTR cDNA sequence of Ubx was cloned into the
SacI site of the pCap-FLuc (pLUC-cassette) vector which
contains the Firefly luciferase (FLuc) ORF [41] to create
the plasmid pUbx-FLuc. The dicistronic reporter vector
pFLuc/RLuc was generated by cloning the Renilla luciferase ORF (RLuc) into the HpaI site of the pLUCcassette. The Ubx 5¢-UTR was also cloned into the BglII

site of pFLuc/RLuc to create the construct pFLuc/Ubx/
RLuc. Nucleotides 1215–1416 and 1311–1547 of the DmeIF4B-L and Dm-eIF4B-S cDNAs, respectively, were
cloned into the EcoRV site of pBluescript vector to create
the plasmids pBS-4BLduplex and pBS-4BSduplex. The
construct pUAS-4BL was generated by cloning the fulllength Dm-eIF4B-L cDNA from the EST LD09953 into
the NotI–XhoI sites of pUAST [42].

Ó FEBS 2004

Protein expression and purification, antibody production,
and Western blot analysis
In vitro transcription/translation of the ESTs was carried
out using the TNT-coupled Reticulocyte Lysate System
(Promega) in the presence of a [35S]Met and [35S]Cys
mixture (14.3 mCiỈmL)1; Amersham) as described by the
manufacturer. Labeled proteins were resolved by SDS/
PAGE (12.5% gel) and detected by autoradiography. Glutathione S-transferase (GST)-eIF4B-L and GST-eIF4B-S
fusion proteins were expressed from pGEX6P2-4BL and
pGEX6P2-4BS in Escherichia coli BL21 CodonPlus
(Novagen) and purified with glutathione–Sepharose 4 Fast
Flow (Amersham Pharmacia Biotech) according to the
manufacturers instructions. The GST tag was removed by
proteolytic digestion with PreScission Protease (Amersham
Biosciences) and further purified on glutathione–Sepharose
to remove both GST and protease. GST-eIF4B-L was
produced from pGEX2T-4B-L in E. coli BL21(DE3)pLysS
cells (Novagen). Polyclonal antibodies to Dm-eIF4B were
generated in rabbit by immunization with 200 lg GSTeIF4B-L protein and Titer Max Adjuvant (Sigma). Lysates
of wild-type Drosophila staged animals were freshly prepared by disrupting the samples on dry ice in a buffer
containing 40 mM Hepes/KOH, pH 7.5, 200 mM KCl,

4 mM EDTA, 1% (v/v) Triton X-100, 0.6 mL)1 aprotinin, 20 lgỈmL)1 leupeptin, 200 lgỈmL)1 soybean trypsin
inhibitor, and EDTA-free protease inhibitor cocktail Complete (Roche Diagnostics GmbH), and centrifuged at
15 800 g for 10 min at 4 °C. The supernatant was recovered
on ice and immediately used without storing. For Western
blot analysis, 5 lg total protein extracts were mixed with
an equal volume of 2· sample buffer, boiled for 2 min,
and loaded on to a gel. The protein concentration
of the samples was quantified with the Protein-Assay Kit
(Bio-Rad). Western blots were performed using rabbit anti(Dm-eIF4B) (1 : 20 000–1 : 80 000) and developed using
the ECL detection kit (Amersham Pharmacia Biotech).
Northern blot and quantitative real-time RT-PCR
Total RNA of staged wild-type D. melanogaster (Oregon R)
was isolated using the RNeasy Mini Kit (Qiagen) and
digested with RNase-free DNase I. Northern blot was
performed as described [33]. The constructs pBS-4BLduplex
and pBS-4BSduplex bearing isoform-specific sequences
were linearized with PstI and transcribed with T3 RNA
polymerase to synthezize 32P-labeled antisense RNAs.
Quantitative real-time RT-PCR was performed with
100 ng RNA and the QuantiTect SYBR green RT-PCR
kit (Qiagen) in an Engine Opticon System (M.J. Research
Inc., Reno, NV, USA). Primers (25-mer) were designed to
amplify 100 bp-long fragments.
Embryo whole-mount in situ hybridization and
immunostaining
Whole-mount in situ hybridization of embryos was
performed as described [43] using antisense or sense
(control) RNA probes against Dm-eIF4B-L (spanning
nucleotides 1426–1215) or Dm-eIF4B-S (nucleotides 1547–
1311) mRNAs. Images were acquired with an Axioplan

Ó FEBS 2004

Functional analysis of Drosophila eIF4B (Eur. J. Biochem. 271) 2925

Microscope coupled to a Kontron CCD camera. Embryo
immunohistochemistry was performed using rabbit anti(Dm-eIF4B) (1 : 500) and Cy3-labeled goat anti-rabbit Ig
(Jackson, West Grove, PA, USA). Images were acquired
with a CLS 310 confocal scanning microscope (Zeiss).

In vivo complementation in Saccharomyces cerevisiae
Constructs p301-TRP1/GAL1-TIF3 [16], pTRP1-4BL,
pTRP1-4BS and pHIS3-4A were used to transform the
yeast strain RCB-1C (MATa tif3::ADE2 ade2 his3 leu2 trp1
ura3 canR [6]). Cells were transformed using the lithium
acetate method [44], plated and tested at 22 °C, 30 °C and
37 °C for growth complementation by Dm-eIF4B as
described [45].
RNA-binding assays
Filter RNA-binding assays were performed using recombinant Dm-eIF4B isoforms as previously described [46].
For cross-linking experiments, 32P-labeled RNA probes
(Ubx and caudal 5¢-UTR) were generated by transcription
of linearized pBS-Ubx and pBS-Cad with the T7 Mega
transcription kit (Ambion, Austin, TX, USA) and radioactive ATP, GTP and UTP (Amersham). RNA probes were
treated with RNase-free DNAse I (Qiagen), further purified
using the RNeasy kit (Qiagen), and integrity was assessed by
agarose gel electrophoresis. RNA probes were diluted in
10 mM Hepes/K+, pH 7.6, containing 15 mM KCl and
2.5 mM MgCl2. The cross-linking was performed in 10 lL

cross-linking buffer (10 mM Hepes/K+, pH 7.6, 1 mM
dithiothreitol, 5% glycerol, 1 mM ATP, 100 ngỈlL)1 total
yeast tRNA, and 10 lgỈlL)1 heparin), 1.6 lg Dm-eIF4B
or GST protein and 100 000 c.p.m.ỈlL)1 RNA (previously
treated for 15 min at 70 °C). The reaction mixtures were
incubated for 15 min at room temperature and then
irradiated for 35 min on ice at 254 nM and digested for
45 min with 1 lL RNAse A (1 lgỈlL)1)/T1 (5 lL)1) at
room temperature. RNA–protein complexes were resolved
in 10% SDS/PAGE and imaged in a Phosphoimager.

In vitro translation assays
Translation extracts were prepared from 0–2 h-old Drosophila embryos as described [47]. In vitro translation was
performed as described [32,41]. Translation assays in the
presence of m7GpppG analog or from heat shockedembryos were as described [48]. Transcripts were synthesized using the Ampliscribe mRNA transcription kit.
Reporter gene expression was determined using the Dualluciferase reporter assay system (Promega) and detected in a
Monolight 2010 luminometer (Analytical Luminescence
Laboratory, San Diego, CA, USA).
RNA interference (RNAi)
Sense and antisense RNAs were prepared from linearized
pBS-4BLduplex and pBS-4BSduplex using the Ampliscribe
mRNA transcription kit (Biozym Diagnostics, Hessich
Olendorf, Germany) in the presence of m7GpppG (New
England BioLabs, Frankfurt am Main, Germany), digested
with DNAse I and purified using the RNeasy kit. Isoform-

specific dsRNAs were produced by hybridization of an
equimolar amount of sense and antisense RNAs in 50 mM
NaCl and 20 mM Tris/HCl pH 8.0 (3 min at 85 °C, 60 min
at 65 °C, chilled on ice and stored at )20 °C). The quality of

the dsRNA molecules was monitored by agarose gel
electrophoresis. Drosophila Schneider S2 cells (1 · 106 in a
35-mm dish) were transfected with 7.5 lg dsRNA using the
Effectene reagent (Qiagen). After 17 h, the medium was
removed, the cells were resuspended in 4.5 mL medium,
counted, and split into four dishes. Then 24 h after transfection, the medium was removed from two wells and the cells
subjected to starvation [medium containing 0.1% (v/v) fetal
bovine serum]. Two wells were kept in medium containing
10% (v/v) serum as fed controls. At 48 h and 72 h after
transfection (24 h and 48 h after starvation), the cells were
harvested for counting and used for Western blotting using
antibodies to Dm-eIF4B as described above.
Overexpression of Dm-eIF4B in S2 cells
For the overexpression of Dm-eIF4B-L, 1 · 106 Drosophila
S2 cells in 35 mm dishes were transfected with either 100 ng
pAct-Gal4 and 300 ng pUAS-4BL, or 100 ng pAct-Gal4
and 300 ng pUAS using the Effectene reagent (Qiagen). At
17 h after transfection, the medium was removed, and the
cells were resuspended in 4.5 mL medium, counted, and
split into four dishes. At 24 h after transfection, the medium
was removed from two wells, and the cells were subjected
to starvation [medium containing 0.1% (v/v) fetal bovine
serum]. Two wells were kept in medium containing 10%
(v/v) serum as fed controls. At 48 h and 72 h after
transfection (24 h and 48 h after starvation) the cells were
counted.
Transgenic flies and overexpression analysis
Flies were raised as described [49]. The construct pUAS-4BL
was used to generate transgenic flies as described [50] by
microinjection in yw embryos. Ectopic overexpession of

eIF4B-L in imaginal eye discs was achieved using the Gal4
system [42]. The transgenic strain yw; P{w UAS-eIF4B-L}
(this study) was crossed to the strain w[*]; P{w[ + mC] ¼
GAL4-ninaE.GMR}12 (Bloomington Drosophila Stock
Center), which drives expression of Gal4 in and behind
the morphogenetic furrow of the developing eye imaginal
disc [51]. The F1 progeny was raised at 25 °C until thirdinstar larvae for imaginal disc analysis or until adulthood
for phenotypic analysis. Eye imaginal discs were dissected
in 1 · NaCl/Pi, fixed for 20 min in 6% (v/v) formaldehyde/
1 · NaCl/Pi, blocked for 1 h in 5% (v/v) normal horse
serum in 1 · NaCl/Pi, and immunostained at 4 °C overnight with rabbit anti-(eIF4B-L) (1 : 7000) or rabbit anti(phospho-histone 3) (1 : 500; Upstate Biotechnology, Lake
Placid, NY, USA) in 5% (v/v) normal horse serum/
1 · NaCl/Pi. Discs were washed three times for 20 min
with 1 · NaCl/Pi at room temperature and incubated for
2 h with either Cy5-conjugated goat anti-rabbit Ig
(1 : 1000) or Cy3-coupled donkey anti-rabbit Ig (1 : 250).
Discs were further washed 3 times for 20 min in 1 · NaCl/
Pi and, when indicated, incubated for 8 min with Alexa
488-conjugated phalloidin (1 : 100) in 1 · NaCl/Pi. Fluorescent signals were acquired and images analyzed using a


´
2926 G. Hernandez et al. (Eur. J. Biochem. 271)

confocal scanning microscope LSM 510 Meta (Zeiss) and
the apropriate filters set. Adult flies were dehydrated in
ethanol, dried, gold coated, and images acquired using a
scanning electron microscope (Zeiss) as described [52].

Results

A single eIF4B gene encodes two eIF4B polypeptides
in Drosophila
A screening of Drosophila ESTs data base using the human
eIF4B cDNA [5] revealed the existence of two groups of
cDNAs, each with a distinct restriction pattern (data not
shown). The comparison of these cDNAs with the genome
of Drosophila [38] indicated the presence of a gene
(CG10837) within the genomic contig AE003089 encoding
the putative fly homolog of eIF4B. D. melanogaster eIF4B
is a single-copy gene spanning 15.5 kb in the chromosome
3R within the Antp locus. A detailed comparison of the
sequences of each group of cDNAs with the genomic
sequence supported the expression pattern of the eIF4B
gene as proposed in Fig. 1A. The existence of one canonical
polyadenylation signal for Dm-eIF4B-S pre-mRNA at
nucleotide 9459, and three for Dm-eIF4B-L pre-mRNA
at nucleotides 23144, 23148 and 23154 of the genomic
fragment AE003089 gives rise to the synthesis of two
pre-mRNAs of different length (Fig. 1A). A subsequent
alternative splicing (from nucleotide 9066 in the long premRNA) allows the specific removal of a premature
termination codon (nucleotide 9071) from the long mRNA
(Fig. 1A). As a result of this expression pattern, we expected
two mRNAs encoding two protein isoforms of Dm-eIF4B
(termed Dm-eIF4B-L and Dm-eIF4B-S). Both mRNAs
( 1.5 and 1.6 kb long, respectively; not shown) have an
identical short 5¢-UTR (39 nucleotides) but differ in their
respective 3¢-UTRs (62 nucleotides for the long and 409
nucleotides for the short form). The ORFs predict a
Dm-eIF4B-L isoform of 459 amino acids (52.2 kDa) and a
Dm-eIF4B-S isoform of 390 amino acids (44.2 kDa).

Interestingly, the two polypeptides share the first 389 amino
acids, as the last amino acid of Dm-eIF4B-S is not present in
Dm-eIF4B-L.
The alignment of the predicted amino-acid sequences
of Dm-eIF4Bs with those of human [5] and yeast [6,7]
counterparts shows a significant overall similarity, in
particular in the N-terminal region (Fig. 1B). A 50%
similarity and 38% identity between the 346 first amino
acids of Drosophila and human eIF4B is found. Amino
acids 75–277 of both Dm-eIF4Bs are 37% similar to and
27% identical with yeast eIF4B. The similarity and identity
between amino acids 90–346 of Dm-eIF4B and Sch. pombe
Sce3p are 44% and 36%, respectively, and 34% and 24%
between amino acids 157–419 of Dm-eIF4B-L and both
eIF4B1 from A. thaliana and eIF4B from wheat, respectively (not shown). Amino acids 74–160 showed 29%
identity with and 49% similarity to human eIF4H. The
RRM described for eIF4B homologs is conserved in
Drosophila (Fig. 1B, boxed), suggesting that both isoforms
should be able to bind RNA. On the other hand, the
PABP-interacting motif [21] is not conserved, suggesting
that the Dm-eIF4B and Dm-PABP do not interact, as
confirmed by our negative interaction results in vitro (not

Ó FEBS 2004

shown). The region corresponding to the DRYG motif [5] is
partially conserved. A similar conservation pattern is
observed in the putative Anopheles gambiae eIF4B gene
´
(G. Hernandez and R. Rivera-Pomar, unpublished observations).

We analysed the size of the two Dm-eIF4B proteins by
in vitro transcription/translation in rabbit reticulocyte lysate
of cDNAs representative of each group of ESTs, which
yielded two polypeptides with approximately the expected
molecular mass (Fig. 1C). To confirm the existence of two
eIF4B protein isoforms in Drosophila cells, we raised antibodies against recombinant GST-eIF4B-L and performed
Western blot experiments. As shown in Fig. 1D, the anti(Dm-eIF4B-L) serum was able to detect the recombinant
proteins eIF4B-L (lane 4) and eIF4B-S (lane 5) as well as
both endogenous Dm-eIF4B isoforms in S2 cell extracts
(lane 6). The preimmune serum did not recognize the
recombinant proteins or any polypeptide present in the cell
extract (lanes 1–3).
Developmental expression of the Dm-eIF4B gene
The electrophoretic resolution of both mature eIF4B
mRNA isoforms was not easily obtained when whole
cDNA probes were used. We observed a smear of 1.5–
1.6 kb, probably as a result of different poly(A) tails
(Northern blot analysis; not shown). Thus, the occurrence
and size of the transcripts was confirmed by Northen
blotting using isoform-specific probes. Developmental
analysis revealed that both isoforms are expressed and
detected as a single transcript from embryos to adulthood
and that the isoform corresponding to eIF4B-S mRNA is
the more abundant (Fig. 2A). A significant maternal
contribution was deduced from the high level of expression of eIF4B-S observed in very early embryos (0–3 h of
development). The expression of Dm-eIF4B-L and DmeIF4B-S mRNAs was analyzed by quantitative real-time
RT-PCR using total RNA derived from different life stages and specific oligonucleotide primers for each messenger (Fig. 2B). Using this method, we observed a major
contribution of eIF4B-S, which was on average four times
more abundant than eIF4B-L. We also performed Western blots using cell extracts from the same life cycle stages.
In agreement with our mRNA quantification experiments,

we observed that both Dm-eIF4B-L and Dm-eIF4B-S
proteins are expressed during the entire life cycle, but a higher
level of Dm-eIF4B-S expression was detected (Fig. 2C). In
summary, a higher level of both Dm-eIF4B-S mRNA and
protein with respect to Dm-eIF4B-L was detected in 0–3 h
embryos.
Whole mount in situ hybridization in embryos using
specific probes showed a ubiquitous but specific signal for
Dm-eIF4B-L and Dm-eIF4B-S transcripts from early stages
on (Fig. 2D). A strong maternal component of both
transcripts was observed (Fig. 2D, syncytial blastoderm
stage). At stages 14–16, they slightly accumulate in the
nervous system (Fig. 2D, stage 16). No signal was obtained
when we used sense RNA probes as controls (Fig. 2D,
bottom). Immunohistochemistry and confocal imaging of
embryos and cells in culture showed that the proteins are
also ubiquitous (Fig. 2E–J). However, as our antibody does
not distinguish between isoforms, we cannot exclude the


Ó FEBS 2004

Functional analysis of Drosophila eIF4B (Eur. J. Biochem. 271) 2927

Fig. 1. A single gene encodes two eIF4B polypeptides in Drosophila. (A) Gene structure of eIF4B. The composition of the two pre-mRNAs was
deduced from sequence comparison of the annotated gene (CG10837) (38) and ESTs LD09953 and LD14038, and confirmed by RT-PCR and
sequencing of the exon–exon junctions. Numbers refer to the genomic fragment AE003089. The first nucleotide of the donor splicing site of the
second intron ( 9 kb) in the long pre-mRNA is numbered (9066). Black boxes refer to the ORFs in each mRNA. Stop codons are indicated by
asterisks and numbered (9071 and 23122). The genes bcd and sd are located within the second intron and are transcribed in the opposite direction.
(B) Alignment of the deduced amino-acid sequences of Drosophila eIF4B-S and eIF4B-L with those of the human (h4B) [5] and yeast (y4B) [6,7]

counterparts. Conserved residues in all proteins or in two species are indicated as black and gray boxes, respectively. Both elements of the RRM are
squared. (C) Autoradiography of [35S]Met incorporation in in vitro transcription/translation products of ESTs LD09953 (lane 1) and LD14038
(lane 2). Molecular mass markers are indicated on the left. (D) Detection of both eIF4B isoforms in Drosophila cells. Recombinant eIF4B-L (2 ng;
lanes 1 and 4), eIF4B-S (2 ng; lanes 2 and 5) or S10 extracts (5 lg) from Schneider S2 cells (lanes 3 and 6) were subjected to Western blot analysis
using preimmune serum (lanes 1–3) or serum containing polyclonal antibodies against recombinant GST-eIF4B-L (lanes 4–6) (dilution 1 : 20 000).
Molecular mass protein markers are shown on the left.


´
2928 G. Hernandez et al. (Eur. J. Biochem. 271)

Ó FEBS 2004

Fig. 2. Expression of eIF4B during Drosophila life cycle and embryogenesis. (A) Northern blot detection of Dm-eIF4B-S and Dm-eIF4B-L mRNAs
during life cycle by isoform-specific probes (see Experimental procedures; also scheme in Fig. 6A). Only a single mRNA of  1.6 kb was detected
with each probe using total RNA derived from early embryos (0–3 h), embryos (0–18 h), first (1), second (2) and third (3) instar larvae, pupae (P)
and adults (A). (B) Relative levels of Dm-eIF4B-S (dashed bars) and Dm-eIF4B-L (black bars) mRNAs measured by quantitative real-time RTPCR of total RNA from the stages shown in (A). The data represent three independent experiments. (C) Detection of eIF4B-S and eIF4B-L by
Western blot analysis of protein extracts prepared from the same developmental stages using anti-(GST-eIF4B-L) IgG (dilution 1 : 20 000).
Molecular mass markers are indicated on the right. (D) Localization of Dm-eIF4B-L mRNA (left column) and Dm-eIF4B-S mRNA (middle
column) during embryogenesis as revealed by in situ hybridization. Developmental stages are indicated at the left side according to Campos-Ortega
and Hartenstein (1997) [63]. As a negative control, parallel in situ hybridization experiments were performed with sense probes and the same
development time (lower panels). Embryos are oriented anterior to the left and dorsal to the top. (E–J) Detection of Dm-eIF4B isoforms by
immunohistochemistry of syncytial blastoderm (E), stage 11 embryo (F) and stage 14 embryo (G). Confocal sections corresponding to the embryos
displayed in (E) and (F) show apical accumulation in the external and pole cells (H) and apical accumulation in tracheal pits (I) of the proteins.
(J) Immunohistochemistry of Schneider cells using anti-eIF4B IgG (dilution 1 : 500).


Ó FEBS 2004

Functional analysis of Drosophila eIF4B (Eur. J. Biochem. 271) 2929

occurrence of isoform-specific distribution. As expected,
Dm-eIF4B is localized to the cytoplasm (Fig. 2E–J). It is
preferentially found in the apical region of polarized cells
(Fig. 2H–I). In cultured S2 cells, Dm-eIF4B-L and
Dm-eIF4B-S are also strictly localized to the cytoplasm
(Fig. 2J). Transfection of S2 cells with C-terminal yellow
fluorescent protein (YFP) fusions confirmed the cytoplasmic localization of both YFP-Dm-eIF4B-L and
YFP-Dm-eIF4B-S (data not shown).
Dm-eIF4B isoforms do not substitute for yeast eIF4B
in vivo
We investigated whether both Dm-eIF4B isoforms alone
or in combination with Drosophila eIF4A are able to
complement the lack of eIF4B in a knockout yeast strain.
For this purpose, we cloned Dm-eIF4B-L, Dm-eIF4B-S
and Dm-eIF4A ORFs into the yeast vectors p301-TRP1Gal1/10 and p301-HIS3-Gal1/10, which allow expression
of cDNAs on galactose-containing media [16] as indicated
in the Experimental procedures. The constructs were
introduced into the yeast strain RCB-1C, a temperaturesensitive and cold-sensitive null mutant of yeast eIF4B [6].
In contrast with yeast eIF4B, none of the Dm-eIF4B
isoforms alone or together with Dm-eIF4A were able to
support a significant growth of the null mutant at 37 °C
or 22 °C (data not shown). We also performed in vitro
translation of a luciferase reporter mRNA using a cell-free
translation yeast system derived from the eIF4B-deficient
yeast strain [7] and supplemented it with our recombinant
proteins. In agreement with the in vivo results, the
addition of Dm-eIF4B-L, Dm-eIF4B-S or Dm-eIF4A
alone or in combinations showed no stimulation of
luciferase synthesis (data not shown). Together, these

results suggest that Dm-eIF4B is unable to replace the
function of its yeast counterpart.
Binding of Dm-eIF4B-L and Dm-eIF4B-S to RNA
Dm-eIF4B isoforms contain one putative RRM that is
conserved in other eIF4Bs (Fig. 1B, boxed). To test
whether Dm-eIF4B binds to RNA, filter-binding assays
[46] using purified Dm-eIF4B-L and Dm-eIF4B-S recombinant proteins and 32P-labeled RNA were performed.
Both isoforms bound to single-stranded RNA with similar
affinity, with an estimated Kd,approx of 2.6 · 10)6 M and
2.8 · 10)6 M for Dm-eIF4B-L and Dm-eIF4B-S, respectively (Fig. 3A). This value is significantly lower than the
Kd,approx measured for Tif3p, which is in the range of
3.4 · 10)7 M in the same assays (not shown). Nevertheless, the lower values for Dm-eIF4B determined in our
assays are consistent with the weak binding observed for
human the eIF4B RRM domain [53]. The RNA-binding
activity of Dm-eIF4B was confirmed by cross-linking to
the radiolabeled 5¢-UTR of Drosophila Ultrabithorax and
caudal mRNA. As shown in Figs 3B and 4C, both RNAs
are cross-linked to Dm-eIF4B recombinant proteins, but
not to GST. Preliminary evidence for cross-linking of
endogenous Dm-eIF4B to different RNAs was also
´
obtained using extracts from Drosophila S2 cells (S. Lopez
de Quinto, E. Martı´ nez-Salas and J. M. Sierra, unpublished results).

Fig. 3. Dm-eIF4B-L and Dm-eIF4B-S are RNA-binding proteins. (A)
Filter-binding assay showing the titration curves of Drosophila eIF4B
proteins against 32P-labeled 38-nucleotide single-stranded RNA
(0.25 pmol) [46]. At the top, Kd,approx values are shown. Seventy percent of the 38-nucleotide RNA is functional, as estimated by testing
1000 · molar excess of eIF4B over RNA. For eIF4B-L and eIF4B-S,
an approximate Kd ¼ 2.6 · 10)6 M and Kd ¼ 2.8 · 10)6 M, respectively, was estimated. Owing to the extensive washing, the Kd values

were rather underestimated. (B, C) Cross-linking experiments were
carried out in the absence of protein or the presence of GST, recombinant Dm-eIF4B-L or Dm-eIF4B-S proteins, with Ultrabithorax (B)
or caudal (C) radiolabeled 5¢-UTRs. (D) Coomassie staining of the gel
shown in (C). Molecular mass markers are shown on the left.

Redundant function of Dm-eIF4B-L and Dm-eIF4B-S
in translation
To study the effect of Dm-eIF4B isoforms in cap-dependent
and IRES-dependent translation, we analyzed the effect of
both proteins on the translation of different mRNA
reporters in a cell-free Drosophila embryo translation system
[32,41]. We used different in vitro transcribed, polyadenylated mRNA reporters, containing either the firefly (FLuc)
or Renilla (RLuc) luciferase cistrons. The cap-FLuc reporter [41] was used to monitor the cap-dependent translation
as shown in Fig. 4A. Whereas the addition of BSA to the


´
2930 G. Hernandez et al. (Eur. J. Biochem. 271)

Ó FEBS 2004

Fig. 4. Recombinant Dm-eIF4B-L and
Dm-eIF4B-S enhance cap-dependent, but not
IRES-dependent, translation in Drosophila
cell-free extracts. (A) Cap-dependent translation using cap-FLuc reporter mRNA in the
absence or presence of increasing amounts of
the indicated proteins. (B) Functionality of the
Ubx-IRES in a dicistronic reporter mRNA
in vitro. The effect of the addition of the cap
analog m7GpppG on cap-dependent (black

bars) and IRES-dependent (gray bars) translation using the dicistronic mRNAs (drawn on
top) is shown. (C) Effect of the addition of the
indicated proteins on cap-dependent (black
bars) and IRES-dependent (gray bars) translation using the dicistronic mRNA (drawn on
top) shown. FLuc, firefly luciferase; Ubx,
Ultrabithorax IRES; RLuc, Renilla luciferase.
The data from four independent experiments
are presented as percentage of control samples
(no protein added).

translation extracts did not have any effect on the translation of the reporter mRNA, translation of firefly luciferase
(FLuc) mRNA increased up to twofold upon the addition
of recombinant Dm-eIF4B-L, Dm-eIF4B-S upon an equimolecular mix of both to the translation system (Fig. 4A),
indicating a positive, specific and redundant effect on
cap-dependent translation.
To assess the IRES-dependent translation, we performed
similar experiments but using a capped dicistronic reporter

mRNA bearing firefly luciferase (FLuc) as first cistron, and
Renilla luciferase (RLuc) as the second cistron. As an
intercistronic element, we used the IRES of Ultrabithorax
(Ubx), a well-defined IRES in Drosophila [39,54]. To prove
the functionality of the Ubx IRES in our system, we first
performed competition experiments by addition of the cap
analog m7GpppG to the translation reaction to mimic the
lack of eIF4E in the lysate [48]. Simultaneously we
compared the translation of cistrons from both capped


Ó FEBS 2004

Functional analysis of Drosophila eIF4B (Eur. J. Biochem. 271) 2931

reporter FLuc/RLuc mRNA, lacking intercistronic
sequences (negative control), and capped FLuc/Ubx/RLuc
(Fig. 4B). As expected, competition experiments with
increasing cap analog concentrations and FLuc/RLuc
(Fig. 4B, left panel) showed that the efficiency of translation
of the first cistron (black bars) decreased with increasing
concentrations of cap analog, whereas the second cistron
remained untranslated (gray bars). In the case of FLuc/
Ubx/RLuc mRNA (Fig. 4B, right panel) translation of
capped FLuc also decreased in the presence of the cap
analog (black bars). Conversely, the presence of the cap
analog enhanced the translation of the second cistron
(RLuc, gray bars) to levels comparable to those obtained for
the first cistron in the absence of cap inhibitor. This
indicated that the Ubx IRES is indeed driving internal
initiation of the second cistron in our in vitro translational
system. We then analyzed the effect of the addition of
recombinant eIF4B proteins or BSA as a control on FLuc/

Fig. 5. Inhibition of cap-dependent, but not
IRES-dependent, translation by anti-eIF4B IgG
and rescue of the translational activity by
recombinant Dm-eIF4B proteins. (A) Functionality of the Ubx IRES in the uncapped
monocistronic mRNA reporter Ubx-FLuc.
Top, cap-FLuc and Ubx-FLuc mRNA
reporters used. Left, translation of cap-FLuc
(circles) and Ubx-FLuc (squares) in the

presence of cap analog m7GpppG. Right,
translation of cap-FLuc (black bars) and
Ubx-FLuc (gray bars) in extracts prepared
from heat-shock-treated embryos. (B–D) Differential inhibition of cap-dependent vs.
IRES-dependent translation after incubating
extracts with antibodies against Drosophila
eIF4B. Translation of the monocistronic
reporter mRNAs cap-FLuc (B), Ubx-FLuc
(C) and the dicistronic capped FLuc/Ubx/
RLuc (D) in the absence or presence of purified IgG fraction from preimmune serum or
from serum containing anti-eIF4B IgG. In (D)
cap-dependent and IRES-dependent translation values are indicated as black and gray
bars, respectively. (E) Rescue of the translational activity of an eIF4B-immunodepleted
extract by addition of recombinant proteins
Dm-eIF4B-L and Dm-eIF4B-S. The data
from four independent experiments are represented as percentage of the control samples
without the addition of protein.

Ubx/RLuc mRNA in our translational assay. Simultaneous
determination of both cap-dependent and IRES-dependent
initiation indicated preferential translation of the first
cistron when eIF4B-L or eIF4B-S, but not BSA was added
(Fig. 4C, black bars). However, no significant effect on
IRES-dependent initiation was detected upon addition of
the Dm-eIF4B proteins or BSA (Fig. 4C, gray bars). Similar
results were obtained using the IRES of Antennapedia [55]
(data not shown). We conclude that cap-dependent initiation is enhanced by Dm-eIF4B-L and Dm-eIF4B-S and
that their effect on cap-dependent translation is equivalent.
Our results also indicate that IRES-dependent translation
is not affected by either form of Dm-eIFB.

We also wanted to prove the functionality of Ubx IRES
when using a monocistronic reporter mRNA Ubx-FLuc
(see below). For this purpose, we performed cap analog
competition experiments and compared the translational
efficiency of noncapped Ubx-FLuc reporter mRNA with


´
2932 G. Hernandez et al. (Eur. J. Biochem. 271)

that of cap-FLuc (Fig. 5A, left panel). Addition of 0.1 mM
cap analog resulted in a 60% inhibition of translation of
cap-FLuc (circles), whereas that of uncapped Ubx-FLuc
remained unchanged (squares). Further addition of cap
analog resulted in 90% inhibition of cap-FLuc mRNA
translation. Conversely, translation of Ubx-FLuc increased
in the presence of cap analog, reaching a plateau at 140% of
translation. In parallel, we tested these reporters in translation extracts derived from untreated and heat-shocked
embryos [48] (Fig. 5A, right), e.g. under conditions when
cap-dependent initiation is severely impaired [48,56]. In
extracts derived from untreated embryos, both cap-FLuc
and Ubx-FLuc mRNAs were efficiently translated (control).
In translation extracts derived from heat-shocked embryos
(heat-shocked) translation of cap-FLuc decreased to 30% of
the control (untreated extracts; black bars), whereas that of
Ubx-FLuc mRNA increased five times (gray bars). These
results confirm the ability of Ubx-FLuc mRNA to be
translated in an IRES-dependent manner and were used
in further experiments. We translated cap-FLuc and

Ó FEBS 2004

Ubx-FLuc mRNA in the presence of purified IgG against
Dm-eIF4B-L (Fig. 5B–D). Addition of anti-(Dm-eIF4B-L)
IgG (but not IgG purified from preimmune serum) resulted
in a twofold decrease in cap-FLuc translation (Fig. 5B).
Conversely, anti-(Dm-eIF4B-L) IgG had only a minor
effect on translation of Ubx-FLuc mRNA (Fig. 5C).
Translation of the first cistron of the dicistronic transcript
FLuc/Ubx/RLuc (Fig. 5D) was also affected by anti-(DmeIF4B-L), but had no significant effect on IRES-dependent
translation of the second cistron (Fig. 5D). Inhibition of
cap-dependent translation caused by the anti-(Dm-eIF4BL) IgG could be reversed to some extent by the addition of
recombinant Dm-eIF4B-L or Dm-eIF4B-S to the cell-free
extract (Fig. 5E). The relative stimulation of luciferase
synthesis obtained in the Dm-eIF4B-inhibited lysate was
similar to that obtained in the noninhibited lysate. Taken
together, these results suggest that both Dm-eIF4B isoforms
have a redundant positive effect on cap-dependent
mRNA translation but do not intervene in IRES-dependent
translation.

Fig. 6. Drosophila eIF4B is required for cell
survival. (A) Localization of the dsRNAs
(arrows) used to knock down each eIF4B
transcript. (B) Levels of eIF4B isoforms in
Schneider S2 mock cells (lane C) or in cells
transfected with dsRNA specific for
Dm-eIF4B-L (lanes L and L + S) or
Dm-eIF4B-S (lanes S and L + S). Total
protein (1 lg per lane) from cells incubated

for 48 h after transfection was probed with
anti-(Dm-eIF4B-L) IgG. (C) Effect of RNAi
on growth on fed (10% fetal bovine serum) S2
cells. The cells were mock transfected (circles)
or transfected with a mixture of the dsRNAs
for both Dm-eIF4B isoforms (squares) and,
after 24 h, re-plated in the same medium
containing 10% fetal bovine serum and incubated for 24 h or an additional 48 h. The
number of viable cells at the time of re-plating
(0 h) is taken as 100%. (D) Protein synthesis
in fed (10% fetal bovine serum) eIF4B knockdown S2 cells. Aliquots of the cells incubated
for 48 h (C) were pulsed with [35S]methionine
(50 lCiỈmL)1) for 2 h, and the labeled proteins analysed by SDS/PAGE (lower panel).
Numbers correspond to the incorporation of
[35S]methionine. The levels of eIF4B isoforms
were estimated by Western blot (upper panel).
(E) Effect of RNAi on growth of starved
(0.1% fetal bovine serum) S2 cells. The
experiment was carried out as in (C) except
that the cells were re-plated in medium containing 0.1% fetal bovine serum. All experiments were performed in triplicate.


Ó FEBS 2004

Functional analysis of Drosophila eIF4B (Eur. J. Biochem. 271) 2933

Dm-eIF4B is involved in cell survival and stimulates
cell proliferation
To determine the in vivo requirement for Dm-eIF4B, we used
RNAi to knock-down Dm-eIF4B in Drosophila S2 cells [57].

Specific dsRNAs for each Dm-eIF4B mRNA isoform were
used in our transfection experiments (for a scheme, see
Fig. 6A). An efficient reduction of Dm-eIF4B expression
was observed by Western blot analysis. Dm-eIF4B-L
(Fig. 6B, lane L) and most of Dm-eIF4B-S, the more
abundant isoform (Fig. 6B, lane S), were knocked down
48 h after transfection with specific dsRNAs. Surprisingly,
the Dm-eIF4B-S dsRNA also resulted in a reduction in the
levels of the Dm-eIF4B-L (Fig. 6B, lane S). Whether this was
due to a direct effect of the interfering RNA or a secondary
effect on the synthesis of eIF4B-L by the removal of eIF4B-S
could not be established. However, the simultaneous transfection with both dsRNAs was effective in completely
removing Dm-eIF4B-L and 90% of Dm-eIF4B-S at 48 h
(Fig. 6B, lane L + S) and 72 h after transfection. Doubletransfected cells were used for a parallel analysis of cell
growth and protein synthesis. Removal of both eIF4B
isoforms resulted in a slight inhibitory effect on cell
proliferation in the presence of 10% serum (Fig. 6C). This
is analogous to the knockout phenotype observed in yeast,
where lack of Tif3 is not essential for cell viability and has
only a moderate effect on cell proliferation at 30 °C.
Accordingly, a small but reproducible effect on protein
synthesis (measured as incorporation of methionine) was
observed (Fig. 6D). Interestingly, when knocked-down cells
were grown at low serum concentrations (0.1%), a significant
reduction in cell survival was observed (Fig. 6E). From our
results we conclude that, under optimal growth conditions
(10% serum), eIF4B is not essential, but it significantly
stimulates cell proliferation under limiting growth conditions
(0.1% serum).
A more drastic effect was observed by overexpressing

Dm-eIF4B-L. S2 cells were cotransfected with plasmid
pUAS-4B-L, a vector bearing the cDNA of Dm-eIF4B-L
under the control of a UAS promoter, and plasmid pActinGAL4 to activate expression of Dm-eIF4B-L (Fig. 7A).
Transfected cells stimulated proliferation independently of
the serum concentration used (Fig. 7B). We also investigated if the overexpression of Dm-eIF4B-L was able to alter
the pattern of cell proliferation in eye imaginal discs. The
Drosophila eye develops from the imaginal discs as a very
precisely patterned structure, which is a powerful tool for
studying cell proliferation and differentiation [58]. In the
developing eye imaginal disc, there are two waves of cells
undergoing mitosis with a dynamic structure between them,
the morphogenetic furrow. The morphogenetic furrow
separates a region of differentiated cells from a region of
proliferating cells [58] (Fig. 7C, arrow). We produced
transgenic flies bearing the same construct as the one
previously used in overexpression experiments in S2 cells
and genetically manipulated transgenic descendants to
express Dm-eIF4B-L in the differentiating part of the eye
imaginal disc (for details see Experimental procedures).
Immunofluorescence and Western blot experiments confirmed the overexpression of Dm-eIF4B-L in the differentiated, nonproliferative tissue (Fig. 7C, left and center
panels). We estimated that the level of Dm-eIF4B-L

Fig. 7. Effect of overexpression of Dm-eIF4B-L in S2 cells and imaginal
discs. (A) Plasmids used in this study. Act, constitutive Actin5C
promoter; GMR, glass promoter, which drives gene expression in the
posterior region of the morphogenetic furrow of the developing eye
imaginal disc; UAS, upstream activator sequence; 4BL, Dm-eIF4B-L
cDNA. (B) Effect of overexpression of Dm-eIF4B-L on cell proliferation of fed (10% serum) or starved (0.1% serum) S2 cells. The
experiment was carried out as in Fig. 6 except that cells were cotransfected with pAct-Gal4 and either pUAS-4BL or pUAS (control).
(C) Overexpression of eIF4B-L in eye imaginal discs. Immunodetection

of eIF4B in wild-type (left) or transgenic UAS-4BL/GMR (middle) eye
antennal imaginal discs. Arrows indicate the morphogenetic furrow of
the eye imaginal disc. The level of eIF4B-L was estimated by Western
blot in wild-type (1; 20 discs) or UAS-4BL/GMR transgenic (2; 15 discs)
eye imaginal discs. (D) Scanning electron microscopy (200 ·) of an adult
eye from wild-type (left) or UAS-4BL/GMR transgenic (right) flies.
Note the disarray of the otherwise patterned structure of the eye, the
increased size of the ommatidia, and the presence of extra chaeta. (E) As
in (D) but at 2000 · magnification. (F) Staining of mitotic cells in wildtype (left) or UAS-4BL/GMR transgenic (middle) eye antennal imaginal
discs with anti-(phosphorylated histone 3) Ig (red). Actin was labeled
with Alexa-488-conjugated phalloidin (green). Extramitotic cells in the
differentiated tissue, posterior to the morphogenetic furrow (arrows),
are evident in the disc overexpressing eIF4B-L. Asterisks indicate the
two waves of mitotic cells. Discs are oriented anterior to the left.


´
2934 G. Hernandez et al. (Eur. J. Biochem. 271)

expression increased 10 times relative to the expression of
eIF4B-S (Fig. 7C, right panel). The analysis of the adult eye
showed abnormal growth and disruption of the periodic
pattern of ommatidia (Fig. 7D) with a penetrance > 70%.
This is consistent with overproliferation phenotypes, a fact
reinforced by the presence of extra chaete (Fig. 7E). In the
wild-type eye, there is a periodic pattern of individual chaete
surrounding one ommatidia (Fig. 7E, left panel). In the
transgenic eye, we consistently observed the presence of
groups of two to three chaeta (Fig. 7E, right panel). As
these structures derive from a single cell in the differentiation

region, we conjectured that the observed phenotype may be
the result of extra mitosis rounds in an otherwise nonproliferative tissue. To confirm this hypothesis, we analysed the
pattern of cell proliferation in the developing eye by in situ
detection of phosphorylated histone 3, which specifically
marks mitotic cells and filamentous actin allowing the
regular pattern of the differentiating tissue to be seen
(a scheme of the eye imaginal disc is shown in the right
panel; the arrow indicates the morphogenetic furrow and
the two mitosis waves). In the wild-type eye imaginal disc,
cell division is restricted to the antennal disk and to the two
mitotic waves flanking the morphogenetic furrow (Fig. 7F,
left panel). The discs overexpressing Dm-eIF4B-L display
additional rounds of mitosis, in particular in the otherwise
nondividing cells posterior to the morphogenetic furrow
(Fig. 7F, middle panel). The rate of proliferation observed
in cells expressing higher levels of Dm-eIF4B-L precisely
correlated with the eye phenotype.

Discussion
We have characterized two isoforms of eIF4B from
D. melanogaster, Dm-eIF4B-L and Dm-eIF4B-S, which
are encoded by a single gene and generated by the use of two
alternative polyadenylation sites and subsequent alternative
splicing of the second intron. This likely reflects similarities
to human eIF4B, which is encoded by two different
mRNAs also produced by alternative polyadenylation sites,
although both mRNAs encode the same polypeptide [5]. In
S. cerevisiae [6,7] and wheat [9], a single eIF4B polypeptide
is produced, whereas in A. thaliana two eIF4B isoforms are
encoded by two different genes [9]. In Anopheles gambiae,

only a single gene could be detected from the genome
´
sequence (G. Hernandez and R. Rivera-Pomar, unpublished observation).
The presence of most initiation factors in all eukaryotic
organisms and the high sequence similarity across the phyla
are evidence of the evolutionary conservation of the
translational machinery. Human and Drosophila eIF4E
are able to substitute for their yeast homolog in vivo ([45];
´
G. Hernandez, M. Altmann and R. Rivera-Pomar, unpublished data). However, not all factors exhibit this property;
mammalian and Drosophila eIF4A and Drosophila eIF4G
are not able to substitute for their yeast counterparts ([59];
´
G. Hernandez, M. Altmann and R. Rivera-Pomar, unpublished data). This is also true for Dm-eIF4B. It may reflect
the existence of subtle differences in the basic translational
machinery between species, both concerning the required
components and their interactions. Another explanation
could be that the lower affinity for RNA of Dm-eIF4B,
which also accounts for the mammal counterpart, is not

Ó FEBS 2004

sufficient to fulfil its function in yeast. Furthermore, factors
involved in more complex interactions with other proteins
may require the presence of species-specific components, a
subject that deserves future investigation.
In addition to its role in cap-dependent translation,
human eIF4B is involved in IRES-dependent translation of
picornaviral mRNAs [24,26–29]. In this study we observed
that both Dm-eIF4B isoforms preferentially promote capdependent translation, which is sensitive to immunodepletion and the addition of recombinant protein Dm-eIF4B.

As the excess of eIF4B or the presence of anti-(Dm-eIF4B)
IgG did not show a significant effect on the in vitro activity
of Ultrabithorax and Antennapedia IRES, we conclude that,
at least for these mRNAs, eIF4B might not be involved.
In vitro binding of Dm-eIF4B to these two IRESs is likely
due to unspecific RNA-binding activity. Whether this
implies a more important role of eIF4B for the translation
of viral rather than for cellular IRESs or, alternatively, other
Drosophila IRESs are eIF4B-dependent has not been
established and requires further investigation. However,
we can clearly conclude that eIF4B has a role in capdependent translation in Drosophila embryonic extracts.
It has been proposed that eIF4B, eIF4A and eIF4F are
involved in the scanning process complementing the role of
eIF1 and eIF1A, and that in the presence of eIF4F, eIF4B
and eIF4A, scanning is less dependent on eIF1 and eIF1A
[60,61]. We observed that reduction of eIF4B decreases the
level of cap-dependent translation while addition of recombinant eIF4B increases translation activity in vitro, arguing
for a central although not exclusive role of eIF4B in the
initiation process. Our depletion experiments support this
conclusion.
Deletion of yeast Tif3 has been shown to cause slow
growth and temperature sensitivity in yeast cells [6,7]. Our
RNAi-based knock-down experiments in cultured cells
show reduced effects on the rate of protein synthesis and,
accordingly on cell proliferation. This suggests that in higher
eukaryotes also (as shown for yeast) eIF4B is not essential,
although we cannot exclude the possibility that the residual
level of Dm-eIF4B after RNAi is enough to keep Drosophila
S2 cells alive. Under starvation conditions, i.e. a situation
in which the cells do not divide, levels of endogenous

Dm-eIF4B were reduced four times (data not shown). As
RNAi treatment performed in S2 cells during starvation
showed that eIF4B is required for cell survival, additional
eIF4B reduction by RNAi treatment may explain why the
cells died. In yeast, eIF4A is required for cell survival during
starvation [62], suggesting that its helicase activity is
required to survive longer periods of treatment, as observed
here for Dm-eIF4B. Overexpression of Dm-eIF4B-L in
Drosophila cell cultures and in developing tissues provided
interesting insights into the function of eIF4B. In mammalian cell cultures, a 50-fold overexpression of eIF4B resulted
in the inhibition of the protein synthesis of a reporter
mRNA [5] and no obvious phenotype. Here we show that a
10-fold increase in Dm-eIF4B-L promotes cell proliferation
in differentiated imaginal tissues. Although we could not
prove that this is a direct effect due to enhanced translation
rates, it correlates with our in vitro experiments demonstrating enhanced translation by the addition of recombinant
Dm-eIF4B-L. We provide here the first in vivo evidence for
increased cell proliferation in response to changes in eIF4B


Ó FEBS 2004

Functional analysis of Drosophila eIF4B (Eur. J. Biochem. 271) 2935

levels. This opens up the way for further biochemical and
genetics experiments to elucidate the mode of action of
eIF4B.

13.

Acknowledgements
´
We thank Annelies Zechel for excellent technical assistance, Jose
Alcalde for antibody production, Marian Bienz for the Ubx cDNA,
´
Fatima Gebauer for the pLUC-cassette vector, Iris Plischke for the
injection of embryos, Peter Schwartz for the scanning electronic
microscopy of the samples, Martin Zeidler and Tina Mukherjee for
sharing their knowledge on eye development, and Stephan Hoppner for
ă
proofreading the manuscript. This work was supported by the Max
Planck Gesellschaft and the Bundesministerium fuer Bildung und
´
Forschung (R.R.-P.), and by grants from the Ministerio de Educacion y
´
´
Cultura and Fundacion Ramon Areces to the CBM, Spain (J.M.S.).

14.

15.

16.

References
1. Hershey, J.W.B. & Merrick, W.C. (2000) The pathway and
mechanism of inititiation of protein synthesis. In Translational
Control of Gene Expression (Sonenberg, N., Hershey, J.W.B. &
Mathews, M.B., eds), pp. 33–88. Cold Spring Harbor Laboratory
Press, Cold Spring Harbor, NY.

2. Gingras, A.C., Raught, B. & Sonenberg, N. (1999) eIF4 initiation
factors: effectors of mRNA recruitment to ribosomes and regulators of translation. Annu. Rev. Biochem. 68, 913–963.
3. Jackson, R.J. (2000) Comparative view of initiation site selection
mechanisms. In Translational Control of Gene Expression
(Sonenberg, N., Hershey, J.W.B. & Mathews, M.B., eds), pp. 127–
184. Cold Spring Harbor Laboratory Press, Cold Spring Harbor,
NY.
4. Grifo, J.A., Tahara, S.M., Morgan, M.A., Shatkin, A.J. &
Merrick, W.C. (1983) New initiation factor activity required for
globin mRNA translation. J. Biol. Chem. 258, 5804–5810.
5. Milburn, S.C., Hershey, J.W.B., Davies, M.V., Kelleher, K. &
Kaufman, R.J. (1990) Cloning and expression of eukaryotic
initiation factor 4B cDNA: sequence determination identifies a
common RNA recognition motif. EMBO J. 9, 2783–2790.
6. Coppolecchia, R., Buser, P., Stotz, A. & Linder, P. (1993) A new
yeast translation initiation factor suppresses a mutation in the
eIF-4A RNA helicase. EMBO J. 12, 4005–4011.
7. Altmann, M., Muller, P.P., Wittmer, B., Ruchti, F., Lanker, S. &
Trachsel, H. (1993) A Saccharomyces cerevisiae homologue of
mammalian translation initiation factor 4B contributes to RNA
helicase activity. EMBO J. 12, 3997–4003.
8. Browning, K.S., Fletcher, L., Lax, S.R. & Ravel, J.M. (1989)
Evidence that the 59-kDa protein synthesis initiation factor from
wheat germ is functionally similar to the 80-kDa initiation factor
4B from mammalian cells. J. Biol. Chem. 264, 8491–8494.
9. Metz, A.M., Wong, K.C.H., Malmstrom, S.A. & Browning, K.S.
ă
(1999) Eukaryotic initiation factor 4B from wheat and Arabidopsis
thaliana is a member of a multigene family. Biochem. Biophys. Res.
Commun. 266, 314–321.

10. Jaramillo, M., Dever, T.E., Merrick, W.C. & Sonenberg, N.
(1991) RNA unwinding in translation: assembly of helicase complex intermediates comprising eukaryotic initiation factors eIF-4F
and eIF-4B. Mol. Cell. Biol. 11, 5992–5997.
11. Methot, N., Pause, A., Hershey, J.W.B. & Sonenberg, N.
(1994) The translation initiation factor eIF-4B contains an
RNA-binding region that is distinct and independent from its
ribonucleoprotein consensus sequence. Mol. Cell. Biol. 14, 2307–
2326.
12. Abramson, R.D., Dever, T.E., Lawson, T.G., Ray, B.K., Thach,
R.E. & Merrick, W.C. (1987) The ATP-dependent interaction of

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

eukaryotic initiation factors with mRNA. J. Biol. Chem. 262,
3826–3832.
´
Pause, A., Methot, N., Svitkin, Y., Merrick, W.C. & Sonenberg, N.
(1994) Dominant negative mutants of mammalian translation
initiation factor eIF-4A define a critical role for eIF-4F in
cap-dependent and cap-independent initiation of translation.
EMBO J. 13, 1205–1215.
Lawson, T.G., Lee, K.A., Maimone, M.M., Abramson, R.D.,
Dever, T.E. & Merrick, W.C. (1989) Dissociation of doublestranded polynucleotide helical structures by eukaryotic initiation
factors, as revealed by a novel assay. Biochemistry 28, 4729–4734.
Rozen, F., Edery, I., Meerovitch, K., Dever, T.E., Merrick, W.C.
& Sonenberg, N. (1990) Bidirectional RNA helicase activity of
eukaryotic translation initiation factors 4A and 4F. Mol. Cell.
Biol. 10, 1134–1144.
Altmann, M., Wittmer, B., Methot, N., Sonenberg, N. & Trachsel,
H. (1995) The Saccharomyces cerevisiae translation initiation
factor Tif3 and its mammalian homologue, eIF-4B, have RNA
annealing activity. EMBO J. 14, 3820–3827.
Niederberger, N., Trachsel, H. & Altmann, M. (1998) The RNA
recognition motif of yeast translation initiation factor Tif3/eIF4B
is required but not sufficient for RNA strand-exchange and
translational activity. RNA 4, 1259–1267.

Naranda, T., Strong, W.B., Menaya, J., Fabbri, B.J. & Hershey,
J.W.B. (1994) Two structural domains of initiation factor eIF-4B
are involved in binding to RNA. J. Biol. Chem. 269, 14465–14472.
Methot, N., Pickett, G., Keene, J.D. & Sonenberg, N. (1996)
In vitro RNA selection identifies RNA ligands that specifically
bind to eukaryotic translation initiation factor 4B: the role of the
RNA recognition motif. RNA 2, 38–50.
Methot, N., Song, M.S. & Sonenberg, N. (1996) A region rich in
aspartic acid, arginine, tyrosine and glycyne (DRYG) mediates
eukaryotic initiation factor 4B (eIF4B) self association and interaction with eIF3. Mol. Cell. Biol. 16, 5328–5334.
Bushell, M., Wood, W., Carpenter, G., Pain, V.M., Morley, S.J. &
Clemens, M.J. (2001) Disruption of the interaction of mammalian
protein synthesis eukaryotic initiation factor 4B with the poly(A)binding protein by caspase- and viral protease-mediated cleavages.
J. Biol. Chem. 276, 23922–23928.
De la Cruz, J., Lost, I., Kressler, D. & Linder, P. (1997) The p20
and Ded1 proteins have antagonistic roles in eIF4E-dependent
translation in Saccharomyces cerevisiae. Proc. Natl Acad. Sci. USA
94, 5201–5206.
Benne, R. & Hershey, J.W.B. (1976) Purification and characterization of initiation factor IF-E3 from rabbit reticulocytes. Proc.
Natl Acad. Sci. USA 73, 3005–3009.
Pestova, T.V., Hellen, C.U.T. & Shatsky, I.N. (1996) Canonical
eukaryotic initiation factors determine initiation of translation by
internal ribosomal entry. Mol. Cell. Biol. 16, 6859–6869.
Meyer, K., Petersen, A., Niepmann, M. & Beck, E. (1995) Interaction of eukaryotic initiation factor eIF-4B with a picornavirus
internal translation initiation site. J. Virol. 69, 2819–2824.
Ochs, K., Rust, R.C. & Niepmann, M. (1999) Translation initiation factor eIF4B interacts with a picornavirus internal ribosome
entry site in both 48S and 80S initiation complexes independently
of the initiator AUG location. J. Virol. 73, 7505–7514.
´
Lopez de Quinto, S. & Martı´ nez-Salas, E. (2000) Interaction of the

eIF4G initiation factor with the aphothovirus IRES is essential for
internal translation initiation in vivo. RNA 6, 1380–1392.
´
Lopez de Quinto, S., Lafuente, E. & Martı´ nez-Salas, E. (2001)
IRES interaction with translation initiation factors: functional
characterization of novel RNA contacts with eIF3, eIF4B, and
eIF4GII. RNA 7, 1213–1226.
Ochs, K., Saleh, L., Bassili, G., Sonntag, V.H., Zeller, A. &
Niepmann, M. (2002) Interaction of translation initiation factor


´
2936 G. Hernandez et al. (Eur. J. Biochem. 271)

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.
40.

41.

42.

43.
44.

eIF4B with the poliovirus internal ribosome entry site. J. Virol. 76,
2113–2122.
Richter-Cook, N.J., Dever, T.E., Hendsold, J.O. & Merrick, W.C.
(1998) Purification and characterization of a new eukaryotic
protein translation factor: eukaryotic initiation factor 4H. J. Biol.
Chem. 273, 7579–7587.
Schmidt, S., Hofmann, K. & Simanis, V. (1997) Sce3, a suppressor
of the Schizosaccharomyces pombe septation mutant cdc11,
encodes a putative RNA-binding protein. Nucleic Acids Res. 25,
3433–3439.
Maroto, F.G. & Sierra, J.M. (1989) Purification and characterization of mRNA cap-binding protein from Drosophila melanogaster embryos. Mol. Cell. Biol. 9, 2181–2190.
´
Hernandez, G. & Sierra, J.M. (1995) Translation initiation factor
eIF-4E from Drosophila: cDNA sequence and expression of the
gene. Biochim. Biophys. Acta 1261, 427–431.
Zapata, J.M., Martı´ nez, M.A. & Sierra, J.M. (1994) Purification
and characterization of eukaryotic polypeptide chain initiation
factor 4F from Drosophila melanogaster embryos. J. Biol. Chem.

269, 18047–18052.
´
Hernandez, G., Castellano, M.M., Agudo, M. & Sierra, J.M.
(1998) Isolation and characterization of the cDNA and the gene
for eukaryotic translation initiation factor 4G from Drosophila
melanogaster. Eur. J. Biochem. 253, 27–35.
Dorn, R., Morawietz, H., Reuter, G. & Saumweber, H. (1993)
Identification of an essential Drosophila gene that is homologous
to the translation initiation factor eIF-4A of yeast and mouse.
Mol. Gen. Genet. 237, 233–240.
Mateu, M.G., Vicente, O. & Sierra, J.M. (1987) Protein synthesis
in Drosophila melanogaster embryos. Purification and characterization of polypeptide chain initiation factor 2. Eur. J. Biochem.
162, 221–229.
Adams, M.D., Celniker, S.E., Holt, R.A., Evans, C.A., Gocayne,
J.D., Amanatides, P.G., Scherer, S.E., Li, P.W., Hoskins, R.A.,
Galle, R.F., George, R.A., Lewis, S.E., Richards, S., Ashburner,
M., Henderson, S.N., Sutton, G.G., Wortman, J.R., Yandell,
M.D., Zhang, Q., Chen, L.X., Brandon, R.C., Rogers, Y.H.,
Blazej, R.G., Champe, M., Pfeiffer, B.D., Wan, K.H., Doyle, C.,
Baxter, E.G., Helt, G., Nelson, C.R., Gabor, G.L., Abril, J.F.,
Agbayani, A., An, H.J., Andrews-Pfannkoch, C., Baldwin, D.,
Ballew, R.M., Basu, A., Baxendale, J., Bayraktaroglu, L., Beasley,
E.M., Beeson, K.Y., Benos, P.V., Berman, B.P., Bhandari, D.,
Bolshakov, S., Borkova, D., Botchan, M.R., Bouck, J., Brokstein,
P., Brottier, P., Burtis, K.C., Busam, D.A., Butler, H., Cadieu, E.,
Smith, H.O., Gibbs, R.A., Myers, E.W., Rubin, G.M. & Venter,
J.C. (2000) The genome sequence of Drosophila melanogaster.
Science 287, 2185–2195.
Hart, K. & Bienz, M. (1996) A test for cell autonomy based on a
dicistronic messenger translation. Development 122, 747–751.

Mlodzik, M. & Gehring, W.J. (1987) Expression of caudal gene
in the germ line of Drosophila: formation of RNA and
protein gradient during early embryogenesis. Cell Death Diff. 48,
464–478.
Gebauer, F., Corona, D.F.V., Preiss, T., Becker, P.B. & Hentze,
M.W. (1999) Translational control of dosage compensation in
Drosophila by Sex-lethal: cooperative silencing via the 5¢ and 3¢
UTRs of msl-2 mRNA is independent of the poly (A) tail. EMBO
J. 18, 6146–6154.
Brand, A. & Perrimon, N. (1993) Targeted gene expression as a
means of altering cells fates and generating dominant phenotypes.
Development 118, 401–415.
Klinger, M. & Gergen, P. (1993) Regulation of runt transcription
by Drosophila segmentation genes. Mech. Dev. 43, 3–19.
Ito, H., Fukuda, Y., Murata, K. & Kimura, A. (1983) Transformation of intact yeast cells treated with alkali cations. J. Bacteriol.
153, 163–168.

Ó FEBS 2004
45. Altmann, M., Muller, P.P., Pelletier, J., Sonenberg, N. &
ă
Trachsel, H. (1989) A mammalian translation initiation factor
can substitute for its yeast homologue in vivo. J. Biol. Chem. 264,
12145–12147.
46. Berset, C., Zurbriggen, A., Djafarzadeh, S., Altmann, M. &
Trachsel, H. (2003) RNA-binding activity of translation initiation factor eIF4G1 from Saccharomyces cerevisiae. RNA 9,
871–880.
47. Tuschl, T., Zamore, P.D., Lehmann, R., Bartel, D.P. & Sharp,
P.A. (1999) Targeted mRNA degradation by double-stranded
RNA in vitro. Genes Dev. 13, 3191–3197.
48. Maroto, F. & Sierra, J.M. (1988) Translational control in heatshocked Drosophila embryos. J. Biol. Chem. 263, 15720–15725.

49. Ashburner, M. (1989) Drosophila. A Laboratory Manual. Cold
Spring Harbor Press, Cold Spring Harbor, NY.
50. Rubin, G.M. & Spradling, A.C. (1982) Genetic transformation
of Drosophila with transposable element vectors. Science 218,
348–353.
51. Freeman, M. (1996) Reiterative use of the EGF receptor triggers
differenciation of all cell types in the Drosophila eye. Cell 87,
651–660.
52. Kimmel, B.E., Heberlein, U. & Rubin, G.M. (1990) The homeodomain protein rough is expressed in a subset of cells in the
developing Drosophila eye where it can specify photoreceptor cell
subtype. Genes Dev. 4, 712–727.
53. Fleming, K., Ghuman, J., Yuan, X., Simpson, P., Szendroi, A.,
ă
Matthews, S. & Curry, S. (2003) Solution structure and RNA
interactions of the RNA recognition motif from eukaryotic
translation initiation factor 4B. Biochemistry 42, 8966–8975.
54. YeX., Fong, P., Izuka, N., Choate, D. & Cavener, D.R. (1997)
Ultrabithorax and Antennapedia 5¢ untranslated regions promote
developmentally regulated internal translation initiation. Mol.
Cell. Biol. 17, 1714–1721.
55. Oh, S.K., Scott, M.P. & Sarnow, P. (1992) Homeotic gene
Antennapedia mRNA contains 5¢-noncoding sequences that confer
translational initiation by internal ribosome binding. Genes Dev. 6,
1643–1653.
56. Schneider, R.J. (2000) Translational control during heat shock. In
Translational Control of Gene Expression (Sonenberg, N.,
Hershey, J.W.B. & Mathews, M.B., eds), pp. 615–635. Cold
Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
57. Clemens, J.C., Worby, C.A., Simonson-Leff, N., Muda, M.,
Maehama, T., Hemmings, B.A. & Dixon, J.E. (2000) Use of

double-stranded RNA interference in Drosophila cell lines to dissect signal transduction pathways. Proc. Natl Acad. Sci. USA 97,
6499–6503.
58. Wolff, T. & Ready, D.R. (1993) Pattern formation in the
Drosophila. In The Development of Drosophila Melanogaster
(Martinez-Arias, M. & Bate, M., eds), pp. 1277–1325. Cold Spring
Harbor Laboratory Press, NY.
59. Prat, A., Schmid, S.R., Buser, P., Blum, S., Trachsel, H. & Linder,
P. (1990) Expression of translation initiation factor 4A from yeast
and mouse in Saccharomyces cerevisiae. Biochim. Biophys. Acta
1050, 140–145.
60. Pestova, T.V., Borukhov, S.I. & Hellen, C.U. (1998) Eukaryotic
ribosomes require initiation factors 1 and 1A to locate initiation
codons. Nature (London) 394, 854–859.
61. Pestova, T.V. & Kolupaeva, V.G. (2002) The roles of individual
eukaryotic translation initiation factors in ribosomal scanning
and initiation codon selection. Genes Dev. 16, 2906–2922.
62. Paz, I. & Choder, M. (2001) Eukaryotic translation initiation
factor 4E-dependent translation is not essential for survival of
starved yeast cells. J. Bacteriol. 183, 4477–4483.
63. Campos-Ortega, J.A. & Hartenstein, V. (1997) The Embryonic
Development of Drosophila Melanogaster, 2nd edn. SpringerVerlag, Berlin, Heidelberg.