Tài liệu Báo cáo khoa học: Molecular characterization of Arabidopsis thaliana PUF proteins – binding specificity and target candidates doc

Molecular characterization of Arabidopsis thaliana PUF
proteins – binding specificity and target candidates
Carlos W. Francischini and Ronaldo B. Quaggio
Departamento de Bioquı
´
mica, Instituto de Quı
´
mica, Universidade de Sa˜o Paulo, Brazil
Introduction
The translational control of RNA is an important reg-
ulatory process in animal development. This regulation
is accomplished by sequence-specific RNA-binding
proteins that recognize cis-acting elements usually
located in the 3¢ UTR. In recent years, and as a result
of great efforts aiming to understand the mechanism
of RNA control in animals, the function of a diverse
number of RNA-binding proteins has been elucidated
[1–4]. Despite this, translational control through the
binding of RNA-binding proteins to 3¢ UTR tran-
scripts has been poorly described in plants.
PUF proteins are a large family of RNA-binding
proteins found in all eukaryotes. These proteins reduce
the expression of mRNA targets by binding in 3¢ UTR
regulatory elements, thus controlling translation or
mRNA stability [5]. Members of the PUF family have
been implicated in diverse processes in development. In
Drosophila, Pumilio binds to the Nanos response ele-
ment (NRE) sequence within the 3¢ UTR of maternal
hunchback mRNA and reduces its expression in the
posterior pole of the embryo. This control is essential
for abdomen formation [6]. In Caenorhabditis elegans

hermaphrodites, the Pumilio homolog FBF binds to
the 3¢ UTR of fem-3 mRNA, repressing its translation
and controlling the sperm–oocyte switch [7]. Dictyoste-
lium PufA represses pkaC mRNA and inhibits the
Keywords
Arabidopsis; PUF proteins; RNA-binding
protein; three-hybrid system; translational
control
Correspondence
R. B. Quaggio, Instituto de Quı
´
mica,
Departamento de Bioquı
´
mica, Universidade
de Sa˜o Paulo, Avenida Professor Lineu
Prestes, 748, Sa˜o Paulo 05508-000, Brazil
Fax/Tel: +55 11 3091 2171
E-mail:
(Received 24 March 2009, revised 15 July
2009, accepted 22 July 2009)
doi:10.1111/j.1742-4658.2009.07230.x
PUF proteins regulate both stability and translation through sequence-spe-
cific binding to the 3¢ UTR of target mRNA transcripts. Binding is medi-
ated by a conserved PUF domain, which contains eight repeats of
approximately 36 amino acids each. Found in all eukaryotes, they have
been related to several developmental processes. Analysis of the 25 Arabid-
opsis Pumilio (APUM) proteins presenting PUF repeats reveals that 12
(APUM-1 to APUM-12) have a PUF domain with 50–75% similarity to
the Drosophila PUF domain. Through three-hybrid assays, we show that

APUM-1 to APUM-6 can bind specifically to the Nanos response element
sequence recognized by Drosophila Pumilio. Using an Arabidopsis RNA
library in a three-hybrid screening, we were able to identify an APUM-
binding consensus sequence. Computational analysis allowed us to identify
the APUM-binding element within the 3¢ UTR in many Arabidopsis tran-
scripts, even in important mRNAs related to shoot stem cell maintenance.
We demonstrate that APUM-1 to APUM-6 are able to bind specifically to
APUM-binding elements in the 3¢ UTR of WUSCHEL , CLAVATA-1,
PINHEAD ⁄ ZWILLE and FASCIATA-2 transcripts. The results obtained
in the present study indicate that the APUM proteins may act as regulators
in Arabidopsis through an evolutionarily conserved mechanism, which may
open up a new approach for investigating mRNA regulation in plants.
Abbreviations
APBE, APUM-binding element; APUM, Arabidopsis Pumilio; IRP, iron regulatory protein; NRE, Nanos response element.
5456 FEBS Journal 276 (2009) 5456–5470 ª 2009 The Authors Journal compilation ª 2009 FEBS
development of fruiting bodies [8], whereas, in yeast,
both Puf3 and Puf5 (Mpt5) proteins promote the
decay of COX17 and HO mRNA, respectively,
through binding to their 3¢ UTR sequences [9,10].
Although members of this family of proteins have
been shown to play distinct roles in different organ-
isms, the maintenance and self-renewal of stem cells
appears to be an ancestral function [5,11]. Drosophila
Pumilio binds to a NRE-like sequence within the
3¢ UTR of cyclin B1, repressing its translation and
promoting germline stem cell development [12–14].
C. elegans FBF also controls germline stem cell main-
tenance by regulating gld-1 mRNA expression and sus-
taining mitosis [15]. The Planaria PUF homolog
DJPum is expressed in neoblasts, which are capable of

self-renewal and differentiation during Planaria regen-
eration. DJPum inactivation by dsRNA was found to
cause a dramatic reduction in the number of neoblasts
and impaired tissue regeneration [16]. In mammals,
PUM2 is expressed in human germline stem cells [17],
whereas the mouse homologs Pum1 and Pum2 are
expressed in fetal and adult hematopoietic stem cells,
as well as in fetal neural stem cells [11].
The canonical PUF domain comprises eight PUF
repeats of approximately 36 amino acids each,
arranged in tandem to form a single concave structure,
usually located in the C-terminal region of the protein.
Each repeat is formed by three a-helices that align
with the equivalent helices in the adjacent repeat,
forming three ladders of helices running through the
domain [18,19]. The crystal structure of the human
Pumilio ⁄ NRE complex demonstrated that each repeat
of the PUF domain recognizes a single nucleotide in
the RNA. Sequence-specific recognition is mediated by
three conserved amino acids residues present at posi-
tions 12, 13 and 16, located in the second helix of each
repeat [20]. Recently, it was shown that these residues
are also important for C. elegans FBF specificity,
suggesting that PUF proteins of different organisms
recognize RNA with the same modularity [21].
Although PUF proteins have been shown to regu-
late distinct mRNA targets across species, the nucleo-
tides recognized appear to be conserved because all
known mRNAs regulated by these proteins contain a
UGURN

1-3
AU(A ⁄ U) sequence [7–10,15,22–29]. In
addition to its ability to bind RNA, the PUF
domain was demonstrated to take part in the pro-
tein–protein contacts necessary for RNA regulation
[6,30,31].
In the present study, we report the first analysis of
plant proteins possessing PUF repeats. Using compu-
tational analyses and yeast three-hybrid assays, we
found that at least six Arabidopsis thaliana proteins
possess eight PUF repeats and can specifically recog-
nize the NRE sequence of Drosophila hunchback
mRNA. Through a yeast three-hybrid screening using
an Arabidopsis RNA hybrid library, we identified
mRNAs that may be target candidates of Arabidopsis
Pumilio (APUM) regulation. The screen also allowed
us to determine a consensus sequence recognized by
the six APUM proteins that can bind to the NRE
sequence (APUM-1 to APUM-6). Using this consen-
sus, we show that APUM proteins are able to bind to
the 3¢ UTR of transcripts related to self-renewal and
stem cell maintenance in the shoot apical meristem.
Moreover, the consensus sequence suggests that a great
number of Arabidopsis transcripts are potential targets
for regulation by the PUF family of proteins. The
results obtained reveal a molecular conservation of
PUF proteins in Arabidopsis thaliana and suggest that
translational regulation via binding to 3¢ UTR in
plants may have a role as important as that previously
described in animals.

Results
PUF proteins in A. thaliana
blast-p analysis of the Arabidopsis genome database
(The Arabidopsis Information Resource – TAIR;
) with Drosophila Pumilio
was carried out to localize PUF proteins. Further
pfam analyses using a cut-off E-value of 1.0 over the
blast-p output identified 25 proteins containing PUF
repeats, which is the largest number of putative PUF
proteins found in a single organism to date. We
named the putative A. thaliana Pumilio homologs
APUM-1 to APUM-25 (Fig. 1). clustal w alignment
of these protein sequences was used to generate a
phylogenetic tree [32], indicating that they may be
separated into four distinct groups of similar pro-
teins, which we named groups I, II, III and IV
(Fig. 1A). Only 3 out of 25 proteins were found to
fall outside these groups. Within each group, some
proteins show a degree of primary sequence identity,
from 40% to 90% along their entire lengths, and
from 63% to 96% among their PUF domains
(Table 1). The analysis also showed that the PUF
domain of the six proteins from group I are highly
similar (approximately 50% identical and 75% simi-
lar) to the Drosophila PUF domain (Fig. 2A and
Table 2), whereas the six proteins from group II have
lower levels of similarity (30% identical and 50%
similar) with the Drosophila PUF domain (Fig. 2B
and Table 2). Moreover, these 12 putative APUM
proteins from groups I and II have eight PUF

C. W. Francischini and R. B. Quaggio PUF proteins in Arabidopsis
FEBS Journal 276 (2009) 5456–5470 ª 2009 The Authors Journal compilation ª 2009 FEBS 5457
repeats in the C-terminal region (Fig. 1B), equivalent
to the number found in the well-characterized PUF
proteins [5,11]. Proteins from group III, group IV
and the three outsiders show more similarity among
themselves than they do with the Drosophila PUF
domain (data not shown).
A
B
Fig. 1. Analysis of the 25 putative APUM
proteins. (A) Phylogenetic tree constructed
based on
CLUSTAL W alignment of all putative
APUM proteins and Drosophila Pumilio
(accession number A46221). Numbers rep-
resent the bootstrap analysis from 1000
trials. (B) Number of PUF repeats identified
for each APUM in the
PFAM analysis. Gray
circles represent the localization of repeats
in the protein and the numbers indicate the
position of each repeat in the PUF domain.
Black circles represent repeats identified in
the
PFAM that fall outside the C-terminal
region. APUM proteins were named
APUM-1 to APUM-25.
PUF proteins in Arabidopsis C. W. Francischini and R. B. Quaggio
5458 FEBS Journal 276 (2009) 5456–5470 ª 2009 The Authors Journal compilation ª 2009 FEBS

Prediction of APUM-binding specificity
In the human Pumilio–NRE complex, residues 12 and
16 of each repeat make hydrogen bonds or Van der
Waals contacts with a specific RNA base, whereas res-
idue 13 makes stacking interactions [20]. An analysis
of these residues for each PUF repeat of all putative
APUM proteins showed that APUM-1 to APUM-6
have the amino acids in positions 12, 13 and 16, simi-
lar to human and Drosophila Pumilio. On the other
hand, the amino acids in these positions in APUM-7
to APUM-12 are more similar to those found in yeast
Puf4 and Puf5 proteins (Table 3; data not shown).
The remaining APUM proteins do not show conserva-
tion of residues 12, 13 and 16 with any well-character-
ized PUF homolog (data not shown) and possess
less than eight PUF repeats in their PUF domains
(Fig. 1B).
The analysis allows us to suggest that the group I
proteins (APUM-1 to APUM-6) share the same
RNA-binding specificity of Drosophila and human
Pumilio-1. Thus, we expected that these APUM
proteins should bind to the NRE sequence within
the 3¢ UTR of Drosophila Pumilio mRNA target
hunchback. The group II proteins APUM-7 to
APUM-11, which have a nonconservative Asn fi His
substitution in residue 13 of repeat 7 (Table 3),
would be expected to have binding to the second
nucleotide in the UGU triplet impaired. Binding to
this nucleotide is essential for RNA recognition
[24,25,33–36].

Table 1. Amino acid identity between some putative Arabidopsis
PUF proteins in the full-length and PUF domain.
Gene ID Similar to:
Full protein
identity (%)
PUF domain
identity (%)
At2g29190 At2g29140 ⁄ At2g29200 90 95
At3g20250 A4g25880 38 63
At1g78160 At1g22240 65 84
At1g35730 At1g35750 78 80
At5g43090 At5g43110 64 68
A
B
Fig. 2. CLUSTAL W alignment of the APUM proteins with the PUF domains most similar to Drosophila PUF domain. (A) APUM proteins of
group I. (B) APUM proteins of group II.
C. W. Francischini and R. B. Quaggio PUF proteins in Arabidopsis
FEBS Journal 276 (2009) 5456–5470 ª 2009 The Authors Journal compilation ª 2009 FEBS 5459
Binding of APUM to NRE
To test the predictions regarding the RNA-binding
specificities of the putative APUM proteins, we investi-
gated the capacity of APUM to bind to the NRE
sequence of hunchback mRNA (Fig. 3B) [36]. We used
the APUM-2 protein as a representative member of
group I proteins and APUM-7 as a representative of
group II APUM proteins (Table 3).
Protein–RNA interactions was evaluated using yeast
three-hybrid system, which was shown to be a reliable
approach for identifying true interactions [25,33,37–39].
This system uses LexA ⁄ MS2 coat protein fusion to

tether the RNA hybrid to the promoter of reporter
genes. The RNA-binding protein is produced as a tran-
scription activation fusion domain through which the
reporter genes (HIS3 or LacZ) are transcribed when the
RNA–protein interaction is established (Fig. 3A) [37].
The pYESTrp3 ⁄ APUM-2 and pYESTrp3 ⁄ APUM-7
vectors were transformed in the yeast YBZ-1 strain
[40] together with the pRH5¢⁄NRE vector. After
growth in selective medium, individual colonies were
tested for LacZ reporter activation. The results showed
that LacZ reporter was activated in yeast colonies
transformed with APUM-2 and NRE, but not with
APUM-7 and NRE (Fig. 3C).
In the NRE fragment, two UGU sequences, named
Box A and Box B (Fig. 3B), were shown to be essential
for Drosophila Pumilio recognition because UGU resi-
dues substitution in both boxes abolished the interaction
with the protein [36]. To verify the specificity of APUM-
2 for the NRE, three mutant NREs with nucleotide sub-
stitutions in the UGU sequence of Box A, NRE(A
)
B
+
);
Box B, NRE(A
+
B
)
); and in both Box A and B,
NRE(A

)
B
)
), were used as baits in the yeast three-hybrid
assay (Fig. 3B) [33]. The results of reporter activation
indicated that APUM-2 interacts with NRE(A
)
B
+
),
Table 2. The 12 Arabidopsis PUF domains most similar to the Dro-
sophila PUF domain.
Gene ID
Similarity to
Drosophila PUF
domain (%)
Identity to
Drosophila PUF
domain (%)
Group I At2g29200 (APUM-1) 74 54
At2g29190 (APUM-2) 75 54
At2g29140 (APUM-3) 74 54
At3g10360 (APUM-4) 73 54
At3g20250 (APUM-5) 73 55
At4g25880 (APUM-6) 69 52
Group II At1g78160 (APUM-7) 56 29
At1g2240 (APUM-8) 55 30
At1g35730 (APUM-9) 51 29
At1g35750 (APUM-10) 53 29
At4g08840 (APUM-11) 57 29

At5g56510 (APUM-12) 53 31
Table 3. Alignment of the nucleotide binding residues of human Pumilio-1 and the corresponding residues in the APUM proteins. The well-
characterized Drosophila Pumilio and yeast Puf4 are also shown. Amino acids at position 12, 13 and 16, respectively, of each repeat are
boxed in gray.
a
Well-characterized Drosophila Pumilio and yeast Puf4 proteins included as comparison.
b
Preferential sequences recognized by Drosophila
Pumilio [27].
c
Preferential sequences recognized by yeast Puf4 [23].
PUF proteins in Arabidopsis C. W. Francischini and R. B. Quaggio
5460 FEBS Journal 276 (2009) 5456–5470 ª 2009 The Authors Journal compilation ª 2009 FEBS
whereas no interaction was observed with NRE(A
+
B
)
)
and NRE(A
)
B
)
) (Fig. 3C). Quantitative analysis of
LacZ expression showed that the binding affinity of
APUM-2 for NRE(A
)
B
+
) was not significantly altered
with respect to the wild-type NRE sequence, whereas

the interaction with NRE(A
+
B
)
) and NRE(A
)
B
)
) was
fully abolished (Fig. 3D). Furthermore, assays using
APUM-7 as prey did not interact with the wild-type or
any of the mutant NREs (Fig. 3C).
To confirm that the result of binding specificity
observed between APUM-2 and NRE can be extended
to the remaining group I proteins, we tested the interac-
tion of APUM-1, APUM-3, APUM-4, APUM-5 and
APUM-6 with wild-type and mutant NREs. Qualitative
(data not shown) and quantitative analysis of LacZ
activity (Fig. 3D) revealed that all five APUMs tested
recognized the NRE and NRE(A
)
B
+
) sequences, but
did not bind to NRE(A
+
B
)
) or NRE(A
)

B
)
).
Together, these results confirmed our predictions
regarding the binding specificity of the subset of group
I APUM proteins, showing that A. thaliana has at least
six PUF proteins with conserved RNA-binding and
similar specificity. The group I APUM proteins recog-
nize Box B within the NRE sequence because UGU
substitutions in Box A did not abolished the interac-
tion. The Box B sequence presents a trinucleotide AUA
downstream of the UGU motif (Fig. 3B), indicating
that APUM proteins could recognize the sequence
UGUANAUA, as do PUF proteins of other organisms
(Table 3) [11]. These observations allow us to speculate
that, although NRE is not the natural RNA target in
Arabidopsis, the Box B sequence should mimic the
authentic Arabidopsis targets. On the other hand,
APUM-7 was unable to bind to NRE, possibly as a
result of the nonconservative substitution at repeat 7.
Influence of the Asp fi His substitution on the
APUM-7 RNA-binding capacity
The APUM-7 protein has the same binding residues as
yeast Puf4 and Puf5 (Table 3; data not shown), except
for an Asp fi His substitution at repeat 7. The yeast
proteins have been shown to recognize sequences similar
to the NRE Box B sequence (Fig. 3B and Table 3) [23].
We considered that, if APUM-7 did not bind to
NRE because of the nonconservative substitution at
residue 13 of the repeat 7, then changing this back to

Asp may restore APUM-7 binding to NRE. In a simi-
lar manner, if this Asp is critical for interaction, its
substitution for a His would be expected to abolish
binding of APUM-2 to NRE.
To evaluate these hypotheses, we tested the interaction
of APUM-2 ⁄ N fi H (APUM-2 with the Asp fi His
A
B
C
D
E
Fig. 3. Interaction analysis between APUM and the NRE transcript.
(A) Schematic representation of the yeast three-hybrid system. (B)
Sequence of the wild-type NRE transcript (WT) and NRE mutants
with nucleotides substitutions in Box A, NRE(A
)
B
+
); Box B,
NRE(A
+
B
)
); and in both Box A and B, NRE(A
–
B
–
). (C) Qualitative
analysis of LacZ reporter activation in the interaction of APUM-2
and APUM-7 with NRE WT and NRE mutants. The iron responsive

element RNA and the IRP protein were used as positive controls
for the interaction. (D) Quantitative analysis of LacZ reporter activa-
tion in the interaction between APUM-1, APUM-2, APUM-3,
APUM-4, APUM-5 and APUM-6 with NRE WT and mutants. (E)
Interaction assay of APUM-2 with Asn to His substitution in the
residue 13 of the repeat 7 (APUM-2 ⁄ N fi H) and the protein
APUM-7 with His fi Asn substitution in the same position (APUM-
7 ⁄ H fi N) with the NRE transcript.
C. W. Francischini and R. B. Quaggio PUF proteins in Arabidopsis
FEBS Journal 276 (2009) 5456–5470 ª 2009 The Authors Journal compilation ª 2009 FEBS 5461
substitution at residue 13 of repeat 7) and of APUM-
7 ⁄ H fi N (APUM-7 with the His fi Asp substitution at
residue 13 of repeat 7) with the NRE transcript in the
three-hybrid system. The results obtained showed that
APUM-2 ⁄ N fi H continued to recognize the NRE,
whereas APUM-7 ⁄ H fi N did not (Fig. 3E), indicating
that the failure to bind to NRE is not a result of
Asp fi His substitution.
Because APUM-8 to APUM-11 proteins share the
same substitution in repeat 7, it is expected that they
will behave as APUM-7 does (i.e. they will not bind to
NRE). Similarly, the APUM-12 protein has exactly the
same amino acids binding residues as APUM-
7 ⁄ H fi N, which suggests that they may exhibit simi-
lar binding behaviors.
Yeast three-hybrid screen to identify
APUM-binding RNA
To identify putative mRNA targets of APUM pro-
teins, we used a yeast three-hybrid screen, which was
shown to be a useful and reliable approach for profil-

ing mRNAs that bind directly to a specific RNA-bind-
ing protein [41–43]. Accordingly, we generated an
Arabidopsis RNA hybrid library of small fragments
(50–150 bp) and used this as prey in a three-hybrid
screen with APUM-2 as bait (Fig. 4A).
From approximately eight million independent Ara-
bidopsis RNA sequences screened, 189 positive interac-
tions derived from 63 distinct sequences were isolated
(Fig. 4B). Of these 63 clones, 27 (43%) were insert
cloned in antisense position. The other 36 clones
(57%) were sense sequences, with five (14%) 3¢ UTR
transcripts (Fig. 4C and Table 4).
Computational analysis of RNA sequences
identified in the yeast three-hybrid screen
Although only five of 63 transcripts identified by the
three-hybrid screens were derived from 3¢ UTR
regions, all of them (sense and antisense) bound to bait
specifically, suggesting the existence of a consensus
motif within these 63 distinct transcripts recognized by
APUM-2. We therefore analyzed these sequences using
multiple expectation maximization for motif elicitation
(meme) as a motif discovery tool [44] (http://meme.
nbcr.net/meme/intro.html). The analysis identified an
eight nucleotide motif present in all 63 transcripts
(Fig. 5A). The consensus possesses a UGUR tetranu-
cleotide sequence, which has been reported to be pres-
ent in all targets of the PUF family [5,11]. In addition,
a(A⁄ U)(U ⁄ G)(A ⁄ U ⁄ C) sequence located one nucleo-
tide downstream of the UGUR motif is highly similar
to the trinucleotide AUA and AUU present in the

target consensus of many other PUF members
[21,23,27,39,45]. In some transcripts, these last three
nucleotides were AGA and AGC, which have not been
described for any other PUF protein to date.
On the basis of these results, we were able to iden-
tify two NRE Box B-like consensus sequences, which
we named the APUM-binding elements (APBE)
(Fig. 5B). The APBE of the 3¢ UTR sequences identi-
fied in the screening is shown in Table 4.
Evaluation of the APBE identified by means of
yeast three-hybrid screen
Because the deduced binding consensus is very small,
it must be present in a large number of Arabidopsis
transcripts. Indeed, a search for the APBE motif in all
5¢ UTR, 3¢ UTR and ORFs annotated at the TAIR
database showed that approximately 56% of all ORF
A
B
C
Fig. 4. Screen of an Arabidopsis RNA hybrid library to identify RNA
bound by APUM-2. (A) Scheme of the three-hybrid strategy used in
the screen. (B) Number of colonies identified in each step of the
screen. (C) Distribution of the 63 distinct sequences in relation of
their position in the Arabidopsis transcriptome.
PUF proteins in Arabidopsis C. W. Francischini and R. B. Quaggio
5462 FEBS Journal 276 (2009) 5456–5470 ª 2009 The Authors Journal compilation ª 2009 FEBS
sequences and 43% of all 3¢ UTR sequences have at
least one binding consensus for APUM proteins,
whereas, in 5¢ UTR, its occurrence is significantly
lower (Table 5).

As a result of the high occurrence of the APUM
binding sites in the plant genome, we decided to focus
in the binding of APUM consensus to 3¢ UTR tran-
scripts expressed in the tissue related to plant meris-
tems because the regulation of transcripts related to
stem cell maintenance is considered to be an ancestral
function of PUF proteins in animals. Thus, a 32 nucle-
otide region of the 3¢ UTR of CLAVATA-1(CLV-1)
(At1g75820), ZWILLE ⁄ PINHEAD (ZLL) (At5g43810),
WUSCHEL (WUS) (At2g17950) and FASCIATA-2
(FAS-2) (At5g64630) transcripts was cloned in the
pRH5¢ vector and tested with APUM-2 protein in the
three-hybrid system (Fig. 5C). These four transcripts
have been described to code for proteins involved in
diverse developmental processes, including shoot meri-
stem organization, stem cell maintenance and mainte-
nance of cellular organization of apical meristems [46–
50]. The LacZ reporter was activated in all assays
tested (Fig. 5D), indicating that the APBE motif is
sufficient for APUM-2 recognition. The APUM-1,
APUM-3, APUM-4, APUM-5 and APUM-6 proteins
also interacted with these transcripts, whereas APUM-
7 did not (data not shown). These results confirm that
the APBEs can be recognized by proteins of group I
and also indicate that these consensus can be useful to
identify putative mRNAs targeted by APUM-1 to
APUM-6.
Group I APUM proteins requires nucleotides in
both 5¢ and 3¢ of the APBE motif
In the computational analysis used to identify a con-

sensus binding motif, no biases towards nucleotides
outside the APBE were identified (Fig. 5A,B). How-
ever, the interactions between APUM-2 with the NRE
transcript and with the four 3¢ APBE UTR sequences
chosen by bioinformatics analysis showed distinct val-
ues of LacZ reporter activation (Figs 3D and 5E).
These data suggest that binding affinity may be influ-
enced either by nucleotides outside of the consensus
motif or by small variations within the consensus.
The interaction of APUM-2 with the FAS-2 tran-
script was the strongest among the interactions tested
in the three-hybrid system (Figs 3D and 5E). The
FAS-2 transcript used in the binding assay differs from
that of WUS, CLV-1 and ZLL sequences in both
APBE and flanking nucleotides (Fig. 5C), whereas its
binding core element is exactly the same as that of
Box B present in the NRE transcript (Fig. 3B).
Because APUM-2 binds to FAS-2 approximately five-
fold more strongly than to NRE (Fig. 5E), we can sug-
gest that specific nucleotides flanking the core element
of FAS-2, which are not present in the NRE sequence,
may contribute to APUM-2 binding.
To examine the contributions of flanking nucleotides
in the affinity between APUM-2 and FAS-2, we pro-
duced double mutations in nucleotides upstream and
downstream of the APBE (Fig. 6A). Quantification
analysis of b-galactosidase activity showed that several
substitutions reduced the binding affinity to different
degrees (Fig. 6B,C). Most significantly, mutations at
positions )1 ⁄ )2 abolished the interaction with

APUM-2 (Fig. 6B). The interaction of APUM-1,
APUM-3, APUM-4, APUM-5 and APUM-6 proteins
with the FAS-2 transcript was also abolished when
the nucleotides at positions )1 ⁄ )2 were substituted
(Fig. 6D).
These results demonstrate that nucleotides upstream
and downstream of the binding consensus are critical
for interaction with APUMs from group I. We can
therefore consider the APBE as the core binding
element, whereas other flanking nucleotides contribute
to the accomplishment of strong or weak inter-
actions.
Table 4. 3¢ UTR transcripts identified in the yeast three-hybrid screening. Upper case letters and boxed sequences indicate the presumptive
APUM binding sites. Information about each gene product was obtained from the TAIR database.
Gene ID (number of
times isolated) Coding for: Sequence identified (5¢-to3¢)
At3g63500 (7) Protein containing PHD domain;
unknown function
ugcgucugaca
UGUACAGCcccugccaaauuuuaauaggcaat
AGUAAAUAaauaacgacaagaagcaaaugg
At5g24490 (1) Ribosomal protein; unknown function cucaucucuccuuacaguuuaccuguguaggaguuaggguucuuga
auaaacaaugcaacaaagauuguagaagucag
UGUACAUA
At4g36040 (1) Protein containing DNAJ domain;
unknown function
cuacgucggacggaacugggaaaccgaucaguguugguagugaguuaa
cucggugaccgaguuaguagaacgaguuaauuag
UGUAAAUAcgaagcca
At4g39090 (1) ‘Embryo defective’ (RD19); response to

physiological stress
uuuaucucugcuucuugcu
UGUAAAUAaa
At3g47470 (1) Chlorophyll a ⁄ b-binding protein cuccaugaacaaauuuggaaucuucaa
UGUACAGA
C. W. Francischini and R. B. Quaggio PUF proteins in Arabidopsis
FEBS Journal 276 (2009) 5456–5470 ª 2009 The Authors Journal compilation ª 2009 FEBS 5463
Discussion
Multiple PUF members in A. thaliana
Currently, the largest number of PUF proteins found
in a single organism was in C. elegans, which has
eleven homologs, whereas yeast has six; human and
mouse possess two; and Drosophila and Dyctiostelium
have only one member [5]. Recently, new studies have
revealed the presence of ten, two and one homologs
in Trypanosome, Plasmodium and Planaria, respec-
tively [16,33,51]. In the present study, we showed that
the A. thaliana genome may contain the largest
number of putative PUF proteins described to date
(Fig. 1).
Functional characterizations of different homologs
have shown that a single PUF protein may be associ-
ated with several distinct developmental processes.
Moreover, PUF proteins in the same organism may
have overlapping and independent functions. In C. ele-
gans, FBF-1 and FBF-2, which share 90% sequence
identity, act redundantly in sperm–oocyte switch and
germ stem cell maintenance [7,15]. However, these two
proteins show distinct patterning functions in the distal
germ line, independently affecting the number of cells

in the mitotic region [29]. Also in C. elegans, the lack
of PUF-8, which is more similar to Drosophila Pumilio
than to FBF, causes germ line dedifferentiation and
the formation of fast growing tumors [52]. In Drosoph-
ila, the single Pumilio has been related to many inde-
pendent processes [14,53–56], and five of the six yeast
PUF homologs, which are significantly divergent in
sequence, appear to have predominately distinct func-
tions [23].
In A. thaliana, we have identified three highly con-
served gene families that account for 22 of 25 putative
PUF proteins. The three remaining proteins can be
divided into a closely-related pair and a single outsider
(Fig. 1A). The large number of copies of highly similar
proteins (Table 1) could be an indicative of redundant
functions in the plant. However, these functions might
be specific to each group of duplicated genes. We
A
B
C
D
E
Fig. 5. Identification and evaluation of a common sequence motif
in the mRNA obtained from yeast three-hybrid screen. (A) Eight
nucleotide motif found by
MEME analysis in all 63 distinct clones.
(B) Deduced APBE. (C) Computational identification of an APBE
(boxed sequences) in the 3¢ UTR region of transcripts FASCIATA-2
(FAS-2), WUSCHEL (WUS), CLAVATA-1 (CLV-1) and ZWILLE ⁄ PIN-
HEAD (ZLL). The sequences shown are the 3¢ UTR regions used in

the yeast three-hybrid assays. (D) Qualitative analysis of LacZ repor-
ter activation in the interaction between APUM-2 and the tran-
scripts FAS-2, WUS, CLV-1 and ZLL. The NRE sequence (Fig. 2B)
was used as a positive control. (E) Quantitative analysis of LacZ
activity in the interactions shown in (D).
Table 5. Occurrence of the APBE in the A. thaliana transcriptome.
Consensus 5¢ UTR
a
3¢ UTR
a
ORF
a
UGURNAKH 1831 7630 16803
UGURNUUA 377 1881 3844
a
Known and putative sequences in the A. thaliana database (TAIR).
A total of 21835 3¢ UTR sequences, 20 564 5¢ UTR sequences and
36 690 ORFs were analyzed separately and independently of
length.
PUF proteins in Arabidopsis C. W. Francischini and R. B. Quaggio
5464 FEBS Journal 276 (2009) 5456–5470 ª 2009 The Authors Journal compilation ª 2009 FEBS
therefore predict that various PUF of A. thalina may
be involved in many different processes in the plant.
RNA-binding capacity of the APUM proteins
pfam analysis of all putative APUM proteins showed
that the six APUM group I proteins, all highly similar
to Drosophila Pumilio (Figs 1A and 2A and Table 2),
have the eight conserved repeats characteristic of the
PUF family of proteins (Fig. 1B). These six homologs
have the same residues necessary to confer RNA speci-

ficity in human Pumilio-1 (Table 3) and can bind to
the NRE sequence specifically (Fig. 3). Six group II
APUM proteins (APUM-7 to APUM-12) (Fig. 1B)
also possess eight PUF repeats, some of which do not
show conservation in residues directly involved in
nucleotide recognition (Table 3). Through site-directed
mutagenesis and interactions assays, we showed that
this substitution is not responsible for the APUM-7
binding impairment (Fig. 3E).
Although PUF proteins have been shown to recog-
nize RNA through a UGUR tetranucleotide followed
by an AU(A ⁄ U) sequence, the number of nucleotides
between these two sequences is variable among different
homologs. For example, C. elegans FBF recognize
RNA that have the UGUR and AUA sequence sepa-
rated by two nucleotides, whereas C. elegans PUF-8,
Drosophila and human Pumilio and yeast Puf3 recog-
AB
CD
Fig. 6. Analysis of binding affinity between APUM-2 and the FAS-2 3¢ UTR transcripts with nucleotide substitutions upstream and down-
stream of the APBE. (A) Double substitutions in flanking nucleotides of APBE (lower case). Bold letters in the wild-type sequence indicate
the APBE. The first nucleotide of the motif is numbered base one. The individual adenine to guanine substitution at nucleotide four was per-
formed to confirm the deduced APBE, which admits a guanine in this position (Fig. 5). (B) Quantitative analysis of LacZ reporter activation in
the interactions between APUM-2 and the FAS-2 transcripts with substitutions in nucleotides upstream of the APBE. (C) Quantitative analy-
sis of LacZ reporter activation in the interactions between APUM-2 and the FAS-2 transcripts with substitutions in the nucleotides down-
stream of the APBE. (D) Quantitative analysis of LacZ activation in the interactions of APUM-1, APUM-3, APUM-4, APUM-5 and APUM-6
with the FAS-2 transcript wild-type (WT) and FAS-2 transcript with substitutions at positions )2 ⁄ )1.
C. W. Francischini and R. B. Quaggio PUF proteins in Arabidopsis
FEBS Journal 276 (2009) 5456–5470 ª 2009 The Authors Journal compilation ª 2009 FEBS 5465
nize the same two sequences separated by only one

nucleotide [21,23,25,27]. In the RNA targets of yeast
Puf5, three nucleotides separate the UGUR and AUA
trinucleotides, whereas the separation is only two nucle-
otides in the RNA targets of Puf4 [23]. Thus, we specu-
late that, if APUM-7 to APUM-12 proteins can bind to
RNA, the UGUR and AUA sequence must be sepa-
rated by more than one nucleotide. This hypothesis,
however, remains to be tested. The other more dis-
tantly-related APUM proteins, in which residues 12, 13
and 16 of each repeat are not conserved (data not
shown), either bind to RNA targets that deviate from
targets of typical PUF proteins or do not bind to RNA.
The results obtained provide strong evidence that at
least six APUM proteins must function as translational
regulators in a manner similar to that of other well-
characterized members of PUF family. These proteins
must bind to transcripts in the plant and regulate their
translation and ⁄ or turnover.
A large number of APUM target candidates: is
this plausible?
RNA-binding proteins have been associated with
diverse aspects of post-transcriptional gene regulation,
including RNA processing, export, localization, degra-
dation and translational regulation [57,58]. Recent
studies have shown that specific RNA-binding proteins
associate with large and distinct mRNA populations,
supporting a role of post-transcriptional operon, in
which mRNA encoding functionally-related proteins
should be coordinately regulated by specific mRNP
components [59].

Systematic identification of the mRNA targets for
five of six yeast PUF proteins showed that each homo-
log interacts with specific subpopulations of mRNA.
Moreover, Puf3, Puf4 and Puf5 were shown to bind,
respectively, to 56%, 26% and 49% of all known and
putative 3¢ UTR sequences possessing the binding con-
sensus identified [23]. The same strategy was used to
identify mRNA bound to Drosophila Pumilio in
embryos and adult ovaries. The results obtained were
similar to that for yeast because Pumilio was shown to
interact with approximately 24% of all 3¢ UTR
possessing the motif UGUAHAUA identified in that
study [27].
In the present study, we identified two APUM con-
sensus sequences, named APBEs (Fig. 5A,B), which
are recognized by six APUM proteins. Together, these
two consensus sequences occur in approximately 43%
of all Arabidopsis 3¢ UTR annotated in the Arabidopsis
database (Table 5). The results obtained demonstrate
that transcripts possessing the consensus motif are
strong candidates for in vivo regulation because all
four 3¢ UTR transcripts chosen by bioinformatics anal-
ysis interacted with APUM-1, APUM-2, APUM-3,
APUM-4, APUM-5 and APUM- 6 in the three-hybrid
system assays (Fig. 5C,D; data not shown).
The large numbers of PUF homologs and 3¢ UTR
transcripts with APBE in A. thaliana indicate that
APUM proteins should regulate a large number of
mRNAs in the plant. However, the number of tran-
scripts regulated in vivo could be limited by nucleotides

flanking the APBE because some substitutions in the
FAS-2 transcript abolished the interaction with the
protein (Fig. 6).
Therefore, as has been suggested in yeast and Dro-
sophila, APUM proteins may be involved in the regu-
lation of many aspects of growth and development.
Interaction between APUM proteins and
transcripts possessing the APBEs
In the present study, we have shown that APUM-1 to
APUM-6 proteins are evolutionarily conserved PUF
proteins. Moreover, we have presented two binding
consensus motifs that allow the efficient identification
of target candidates to translational regulation by
APUMs in A. thaliana. The results obtained show that
the presence of the APBE is evidence of a possible
interaction, although this requires confirmation for
each particular case. The results of nucleotide substitu-
tions in the FAS-2 transcript reveal that binding is
highly affected by nucleotides flanking the binding site.
Recently, Stumpf et al. [60] described a similar strategy
to identify mRNA sequence motifs bound by PUF-5
and PUF-6 proteins from C. elegans. Using a three-
hybrid assay, they screened a random small fragment
RNA library to define the binding motif sequence.
When the motif was used to search the C. elegans
mRNA database, they found a high number of
mRNAs that have the binding motif on their 3¢ UTR
sequence. Some of the mRNAs were shown to be
potential targets of the PUFs homologs in a three-
hybrid interaction assay.

In the present study, we described a physical interac-
tion between group I APUM proteins and the meriste-
matic transcripts CLV-1, WUS, ZLL and FAS-2
(Fig. 5C,D). Because we have obtained experimental
evidence indicating that some APUM proteins are
localized in to shoot and root meristems, these four
transcripts comprise potentials targets candidates for
APUM regulation. Interestingly, preliminary molecular
assays with APUM antisense plants have confirmed a
potential biological relationship between APUM
proteins and at least two of the four transcripts.
PUF proteins in Arabidopsis C. W. Francischini and R. B. Quaggio
5466 FEBS Journal 276 (2009) 5456–5470 ª 2009 The Authors Journal compilation ª 2009 FEBS
The results obtained in the present study therefore
open up a new area of investigation, suggesting that
many different aspects of the plant can be coordi-
nately regulated by Arabidopsis PUF homologs,
in the same way as that shown for PUF proteins in
animals.
Experimental procedures
Materials
All restrictions enzymes and polymerases were obtained
from Invitrogen (Sa
˜
o Paulo, SP, Brazil). The RNA–protein
interaction Hybrid Hunter kit was obtained from Invitro-
gen (Sa
˜
o Paulo, SP, Brazil). All oligonucleotides were
obtained from Invitrogen (Sa

˜
o Paulo, SP, Brazil). The
Mutagenesis Kit was obtained from Strategene (Sa
˜
o Paulo,
SP, Brazil). Nitrocellulose membrane was obtained from
Amersham Biosciences (Sa
˜
o Paulo, SP, Brazil). All amino
acids, Gal-ONp, sodium carbonate (Na
2
CO
3
), yeast nitro-
gen base, ammonium sulfate, succinic acid, glucose, agar
and agarose were obtained from Sigma (St Louis, MO,
USA). Trizol reagent was obtained from Invitrogen (Sa
˜
o
Paulo, SP, Brazil). The YBZ-1 strain was obtained from
Marvin Wickens (Madison, WI, USA).
Three-hybrid system assays
Three-hybrid assays were performed in accordance with the
RNA–protein interaction Hybrid Hunter kit protocol (Invi-
trogen) and as previously described [38].
Cloning of APUM cDNAs and 3¢ UTR sequences
in the yeast three-hybrid system vectors
APUM proteins-VP16 activation domain fusions were con-
structed by insertion of the RT-PCR amplified PUF
domains of APUM-1, APUM-2, APUM-3, APUM-4,

APUM-5, APUM-6 and APUM-7 into the pYESTrp3
vector (Invitrogen) through recombination in the Gateway
system (Invitrogen). To clone the 3¢ UTR transcripts, two
complementary oligonucleotides corresponding to the
NRE wild-type [36], NRE mutants [33], FAS-2 wild-type
and CLV-1 were synthesized and annealed. Each annealed
pair of oligonucleotides possess a 32 bp sequence of a
3¢ UTR region of transcripts containing the APBE, a 5¢
AvrII or XbaI site and a 3¢ SmaI. After digestion, the
inserts were ligated to an AvrII-SmaI-digested pRH5¢ vec-
tor (Invitrogen). To clone the FAS-2 mutants, WUS,
CLV-1 and ZLL sequences, only sense primers were
synthesized. Each primer was annealed with an antisense
oligonucleotide 5¢-TCCCCCGGGGG-3¢ and extended with
Klenow fragment of Escherichia coli DNA polymerase.
After extension, the double-stranded DNA were digested
with XbaI and SmaI and ligated to an AvrII-SmaI-digested
pRH5¢ vector. The sequences of primers used are given in
Tables S1 and S2.
Single amino acid change in APUM-2 and
APUM-7 proteins
The amino acid substitution to generate pYESTrp3APUM-
2 ⁄ N fi H and pYESTrp3APUM-7 ⁄ H fi N proteins was
obtained according to the protocol described in the Quik-
Change Site-Directed Mutagenesis Kit (Strategene). The
primers used were: 5¢-GAGCCAACAGAAGTTTGCT
TCACACGTTGTTGAGAAATGTTT-3¢ (forward) and
5¢-GTCAAACATTTCTCAACAACGTGTGAAGCAAAC
TTCTGTTGG-3¢ (reverse) to APUM-2 and 5¢-CGATG
CAGAAATTCAGTAGCAACATGGTGGAACGATGTC

TCA-3¢ (forward) and 5¢-GCATGAGACATCGTTCCAC
CATGTTGCTACTGAATTTCTGCA-3¢ (reverse) to APUM-7.
Qualitative and quantitative b-galactosidade
activity assays
For qualitative assays, colonies grown on solid selective
medium lacking uracil and Trp were transferred to a nitro-
cellulose membrane, lysed in liquid nitrogen and exposed to
X-gal for 1–2 h [38]. For quantitative assays, cells were first
grown overnight on liquid selective medium lacking uracil
and Trp, then diluted 1 : 5 in YPDA medium (10 gÆL
)1
yeast extract, 20 gÆL
)1
peptone, 20 gÆL
)1
glucose and 0.2%
adenine hemisulfate salt) and grown again until an A
600
of
1.0–1.5 was reached. After growth, 1.5 mL of cells were
centrifuged, washed, lysed in liquid nitrogen and incubated
with 160 lL of Gal-ONp 4 mgÆmL
)1
for 1 h at 30 °C. The
reactions were stopped with 400 lLofNa
2
CO
3
, centrifuged
and A

420
was measured. For each assay, at least three
independent yeast colonies were used.
Direct interaction between APUM and RNA in the
yeast three-hybrid system
To investigate direct interaction of the APUM proteins with
RNA, the yeast YBZ-1 strain [40], in which the Lex-MS2
fusion protein was stably integrated, was co-transformed
with both pYESTrp3 ⁄ APUM and one of the bait pRH5¢⁄
RNA plasmids. As a negative control, the pYESTrp3 ⁄
APUM prey was co-transformed with the transcript control
pRH5¢⁄iron responsive element or empty pRH5¢ vector.
As a positive control, yeast was co-transformed with
pYESTrp3-iron regulatory protein (IRP) and pRH5¢⁄IRE.
Transformed yeast was plated on yeast complete synthetic
medium (0.12% yeast nitrogen base, 0.5% ammonium sul-
fate, 1% succinic acid, 2% glucose, amino acid mix, 2 %
agar) lacking uracil and Trp and the b -galactosidase activ-
ity of individual colonies was determined.
C. W. Francischini and R. B. Quaggio PUF proteins in Arabidopsis
FEBS Journal 276 (2009) 5456–5470 ª 2009 The Authors Journal compilation ª 2009 FEBS 5467
Construction of the Arabidopsis RNA hybrid
library for the yeast three-hybrid screen
The Arabidopsis hybrid RNA library was constructed from
total RNA isolated from wild-type plants grown for 2 weeks
in MS medium (Murashige and Skoog medium from Sigma),
using Trizol reagent (Invitrogen), in accordance with the
manufacturer’s instructions. Random hexamer-primed
cDNAs were synthesized using the CloneMiner cDNA
Library Construction Kit (Invitrogen). Double-stranded

cDNA was digested with AluI, HaeIII, DraI and SspI and
then fractionated on a 2% agarose gel. Fragments ranging
from 50 to 150 bp were purified from the gel and ligated to
the pRH5¢ vector (Invitrogen) in a single step reaction, using
PmeI and T4 DNA ligase at 25 °C overnight. The ligation
mixtures were used to transform competent DH10B cells.
Screening of the Arabidopsis RNA library
Yeast strain YBZ-1 [40] was co-transformed with APUM-2
protein hybrid and the hybrid RNA plasmids. Cells were
grown on yeast complete synthetic medium lacking His,
uracil and Trp and containing 1 mm 3-aminotriazole,
a competitive inhibitor of the His3p enzyme. Within
4–5 days, colonies were picked into fresh synthetic medium
and tested to LacZ reporter activation. To eliminate false
positives, the APUM-2 protein plasmid was purged from
LacZ positive colonies by growing in YPDA liquid medium
at 30 °C overnight. Cells that lost the APUM-2-activation
domain were again tested for b-galactosidase activity. The
true positive plasmids were isolated from yeast cells and
re-transformed into YBZ-1 cells with APUM-2 plasmid or
the pYESTrp3 ⁄ IRP plasmid control. Transformants that
activated the LacZ reporter with APUM-2 plasmid, but not
with the pYESTpr3 ⁄ IRP plasmid control, were then
sequenced.
Computational analyses of protein and RNA
sequences
Sequences of RNA fragments obtained from screening were
identified using blastn on the TAiR database. The sequences
were searched for motifs in the sense strand with meme soft-
ware [44]. Nucleotides that had an occurrence at least 10%

among the MEME selected sequences were used to compile
the APUM consensus element for searching. Phylogenetic
and molecular evolutionary analyses were performed using
the mega, version 3.1 [32]. Transcripts possessing the APBE
were searched for within the TAIR database.
Acknowledgements
We are grateful to Dr Shaker Chuck Farah and
Dr Carla Columbano de Oliveira for their helpful
comments and Dr Marvin Wickens for providing the
YBZ-1 strain. We thank Dr Jerry Ostrowski for help
in constructing the RNA library. This work was
supported by FAPESP and CNPq.
References
1 Wells DG (2006) RNA-binding proteins: a lesson in
repression. J Neurosci 26, 7135–7138.
2 Wallander ML, Leibold EA & Eisenstein RS (2006)
Molecular control of vertebrate iron homeostasis by
iron regulatory proteins. Biochim Biophys Acta 1763,
668–689.
3 Cho PF, Gamberi C, Cho-Park YA, Cho-Park IB,
Lasko P & Sonenberg N (2006) Cap-dependent transla-
tional inhibition establishes two opposing morphogen
gradients in Drosophila embryos. Curr Biol 16, 2035–
2041.
4 Colegrove-Otero LJ, Minshall N & Standart N (2005)
RNA-binding proteins in early development. Crit Rev
Biochem Mol Biol 40, 21–73.
5 Wickens M, Bernstein DS, Kimble J & Parker R (2002)
A PUF family portrait: 3¢ UTR regulation as a way of
life. Trends Genet 18, 150–157.

6 Wharton RP, Sonoda J, Lee T, Patterson M & Murata
Y (1998) The Pumilio RNA-binding domain is also a
translational regulator. Mol Cell 1, 863–872.
7 Zhang B, Gallegos M, Puoti A, Durkin E, Fields S,
Kimble J & Wickens MP (1997) A conserved RNA-
binding protein that regulates sexual fates in the C.
elegans hermaphrodite germ line. Nature 390, 477–484.
8 Souza GM, da Silva AM & Kuspa A (1999) Starvation
promotes Dictyostelium development by relieving PufA
inhibition of PKA translation through the YakA kinase
pathway. Development 126, 3263–3274.
9 Olivas W & Parker R (2000) The Puf3 protein is a tran-
script-specific regulator of mRNA degradation in yeast.
EMBO J 19, 6602–6611.
10 Tadauchi T, Matsumoto K, Herskowitz I & Irie K
(2001) Post-transcriptional regulation through the HO
3¢-UTR by Mpt5, a yeast homolog of Pumilio and
FBF. EMBO J 20, 552–561.
11 Spassov DS & Jurecic R (2003) The PUF family of
RNA-binding proteins: does evolutionarily conserved
structure equal conserved function? IUBMB Life 55,
359–366.
12 Forbes A & Lehmann R (1998) Nanos and Pumilio
have critical roles in the development and function of
Drosophila germline stem cells. Development 125, 679–
690.
13 Asaoka-Taguchi M, Yamada M, Nakamura A,
Hanyu K & Kobayashi S (1999) Maternal Pumilio acts
together with Nanos in germline development in
Drosophila embryos. Nat Cell Biol 1, 431–437.

PUF proteins in Arabidopsis C. W. Francischini and R. B. Quaggio
5468 FEBS Journal 276 (2009) 5456–5470 ª 2009 The Authors Journal compilation ª 2009 FEBS
14 Parisi M & Lin H (1999) The Drosophila pumilio
gene encodes two functional protein isoforms that
play multiple roles in germline development, gonado-
genesis, oogenesis and embryogenesis. Genetics 153,
235–250.
15 Crittenden SL, Bernstein DS, Bachorik JL, Thompson
BE, Gallegos M, Petcherski AG, Moulder G, Barstead
R, Wickens M & Kimble J (2002) A conserved RNA-
binding protein controls germline stem cells in Caenor-
habditis elegans. Nature 417, 660–663.
16 Salvetti A, Rossi L, Lena A, Batistoni R, Deri P,
Rainaldi G, Locci MT, Evangelista M & Gremigni V
(2005) DjPum, a homologue of Drosophila Pumilio, is
essential to planarian stem cell maintenance. Develop-
ment 132, 1863–1874.
17 Moore FL, Jaruzelska J, Fox MS, Urano J, Firpo MT,
Turek PJ, Dorfman DM & Pera RA (2003) Human
Pumilio-2 is expressed in embryonic stem cells and germ
cells and interacts with DAZ (Deleted in AZoospermia)
and DAZ-like proteins. Proc Natl Acad Sci USA 100,
538–543.
18 Edwards TA, Pyle SE, Wharton RP & Aggarwal AK
(2001) Structure of Pumilio reveals similarity between
RNA and peptide binding motifs. Cell 105, 281–289.
19 Wang X, Zamore PD & Hall TM (2001) Crystal struc-
ture of a Pumilio homology domain. Mol Cell 7, 855–
865.
20 Wang X, McLachlan J, Zamore PD & Hall TM (2002)

Modular recognition of RNA by a human pumilio-
homology domain. Cell 110, 501–512.
21 Opperman L, Hook B, DeFino M, Bernstein DS &
Wickens M (2005) A single spacer nucleotide determines
the specificities of two mRNA regulatory proteins. Nat
Struct Mol Biol 12, 945–951.
22 Nakahata S, Katsu Y, Mita K, Inoue K, Nagahama Y
& Yamashita M (2001) Biochemical identification of
Xenopus Pumilio as a sequence-specific cyclin B1
mRNA-binding protein that physically interacts with a
Nanos homolog, Xcat-2, and a cytoplasmic polyadeny-
lation element-binding protein. J Biol Chem 276 , 20945–
20953.
23 Gerber AP, Herschlag D & Brown PO (2004) Extensive
association of functionally and cytotopically related
mRNAs with Puf family RNA-binding proteins in
yeast. PLoS Biol 2, E79.
24 Jackson JS Jr, Houshmandi SS, Lopez Leban F &
Olivas WM (2004) Recruitment of the Puf3 protein to
its mRNA target for regulation of mRNA decay in
yeast. RNA 10, 1625–1636.
25 Fox M, Urano J & Reijo Pera RA (2005) Identification
and characterization of RNA sequences to which
human PUMILIO-2 (PUM2) and deleted in Azoosper-
mia-like (DAZL) bind. Genomics 85, 92–105.
26 Thompson BE, Bernstein DS, Bachorik JL, Petcherski
AG, Wickens M & Kimble J (2005) Dose-dependent
control of proliferation and sperm specification by
FOG-1 ⁄ CPEB. Development 132, 3471–3481.
27 Gerber AP, Luschnig S, Krasnow MA, Brown PO &

Herschlag D (2006) Genome-wide identification of
mRNAs associated with the translational regulator
PUMILIO in Drosophila melanogaster. Proc Natl Acad
Sci USA 103, 4487–4492.
28 Eckmann CR, Crittenden SL, Suh N & Kimble J (2004)
GLD-3 and control of the mitosis ⁄
meiosis decision in
the germline of Caenorhabditis elegans. Genetics 168,
147–160.
29 Thompson BE, Lamont LB & Kimble J (2006) Germ-
line induction of the Caenorhabditis elegans vulva. Proc
Natl Acad Sci USA 103, 620–625.
30 Sonoda J & Wharton RP (1999) Recruitment of Nanos
to hunchback mRNA by Pumilio. Genes Dev 13, 2704–
2712.
31 Sonoda J & Wharton RP (2001) Drosophila brain
tumor is a translational repressor. Genes Dev 15, 762–773.
32 Kumar S, Tamura K & Nei M (2004) MEGA3: inte-
grated software for molecular evolutionary genetics
analysis and sequence alignment. Brief Bioinform 5,
150–163.
33 Cui L, Fan Q & Li J (2002) The malaria parasite Plas-
modium falciparum encodes members of the Puf RNA-
binding protein family with conserved RNA binding
activity. Nucleic Acids Res 30, 4607–4617.
34 Zamore PD, Williamson JR & Lehmann R (1997) The
Pumilio protein binds RNA through a conserved
domain that defines a new class of RNA-binding
proteins. RNA 3, 1421–1433.
35 Cheong CG & Hall TM (2006) Engineering RNA

sequence specificity of Pumilio repeats. Proc Natl Acad
Sci USA 103, 13635–13639.
36 Murata Y & Wharton RP (1995) Binding of pumilio to
maternal hunchback mRNA is required for posterior
patterning in Drosophila embryos. Cell 80, 747–756.
37 SenGupta DJ, Zhang B, Kraemer B, Pochart P, Fields
S & Wickens M (1996) A three-hybrid system to detect
RNA–protein interactions in vivo. Proc Natl Acad Sci
USA 93, 8496–8501.
38 Bernstein DS, Buter N, Stumpf C & Wickens M (2002)
Analyzing mRNA–protein complexes using a yeast
three-hybrid system. Methods 26, 123–141.
39 Bernstein D, Hook B, Hajarnavis A, Opperman L &
Wickens M (2005) Binding specificity and mRNA
targets of a C. elegans PUF protein, FBF-1. RNA 11,
447–458.
40 Hook B, Bernstein D, Zhang B & Wickens M (2005)
RNA–protein interactions in the yeast three-hybrid sys-
tem: affinity, sensitivity, and enhanced library screening.
RNA 11, 227–233.
41 Sengupta DJ, Wickens M & Fields S (1999) Identifica-
tion of RNAs that bind to a specific protein using the
yeast three-hybrid system. RNA 5, 596–601.
C. W. Francischini and R. B. Quaggio PUF proteins in Arabidopsis
FEBS Journal 276 (2009) 5456–5470 ª 2009 The Authors Journal compilation ª 2009 FEBS 5469
42 Seay D, Hook B, Evans K & Wickens M (2006) A
three-hybrid screen identifies mRNAs controlled by a
regulatory protein. RNA 12, 1594–1600.
43 Ostrowski J, Wyrwicz L, Rychlewski L & Bomsztyk K
(2002) Heterogeneous nuclear ribonucleoprotein K pro-

tein associates with multiple mitochondrial transcripts
within the organelle. J Biol Chem 277, 6303–6310.
44 Bailey TL & Elkan C (1994) Fitting a mixture model by
expectation maximization to discover motifs in biopoly-
mers. Proc Int Conf Intell Syst Mol Biol 2, 28–36.
45 White EK, Moore-Jarrett T & Ruley HE (2001) PUM2,
a novel murine puf protein, and its consensus RNA-
binding site. RNA 7, 1855–1866.
46 Laux T, Mayer KF, Berger J & Jurgens G (1996) The
WUSCHEL gene is required for shoot and floral meri-
stem integrity in Arabidopsis . Development 122, 87–96.
47 Mayer KF, Schoof H, Haecker A, Lenhard M, Jurgens
G & Laux T (1998) Role of WUSCHEL in regulating
stem cell fate in the Arabidopsis shoot meristem. Cell
95, 805–815.
48 Schoof H, Lenhard M, Haecker A, Mayer KF, Jurgens
G & Laux T (2000) The stem cell population of Arabid-
opsis shoot meristems in maintained by a regulatory
loop between the CLAVATA and WUSCHEL genes.
Cell 100, 635–644.
49 Kaya H, Shibahara KI, Taoka KI, Iwabuchi M,
Stillman B & Araki T (2001) FASCIATA genes for
chromatin assembly factor-1 in Arabidopsis maintain
the cellular organization of apical meristems. Cell 104,
131–142.
50 Moussian B, Haecker A & Laux T (2003) ZWILLE
buffers meristem stability in Arabidopsis thaliana. Dev
Genes Evol 213, 534–540.
51 Caro F, Bercovich N, Atorrasagasti C, Levin MJ &
Vazquez MP (2006) Trypanosoma cruzi: analysis of the

complete PUF RNA-binding protein family. Exp Paras-
itol 113, 112–124.
52 Subramaniam K & Seydoux G (2003) Dedifferentiation
of primary spermatocytes into germ cell tumors in
C. elegans lacking the pumilio-like protein PUF-8. Curr
Biol 13, 134–139.
53 Baines RA (2005) Neuronal homeostasis through trans-
lational control. Mol Neurobiol 32, 113–121.
54 Menon KP, Sanyal S, Habara Y, Sanchez R, Wharton
RP, Ramaswami M & Zinn K (2004) The translational
repressor Pumilio regulates presynaptic morphology
and controls postsynaptic accumulation of translation
factor eIF-4E. Neuron 44, 663–676.
55 Schweers BA, Walters KJ & Stern M (2002) The
Drosophila melanogaster translational repressor pumilio
regulates neuronal excitability. Genetics 161 , 1177–1185.
56 Gamberi C, Peterson DS, He L & Gottlieb E (2002) An
anterior function for the Drosophila posterior determi-
nant Pumilio. Development
129, 2699–2710.
57 Mazumder B, Seshadri V & Fox PL (2003) Transla-
tional control by the 3¢-UTR: the ends specify the
means. Trends Biochem Sci 28, 91–98.
58 Dreyfuss G, Kim VN & Kataoka N (2002) Messenger-
RNA-binding proteins and the messages they carry. Nat
Rev Mol Cell Biol 3, 195–205.
59 Keene JD & Tenenbaum SA (2002) Eukaryotic mRNPs
may represent posttranscriptional operons. Mol Cell 9,
1161–1167.
60 Stumpf CR, Kimble J & Wickens M (2008) A Caenor-

habditis elegans PUF protein family with distinct RNA
binding specificity. RNA 14, 1550–1557.
Supporting information
The following supplementary material is available:
Table S1. List of primers used in the PUF domains
amplification.
Table S2. List of primers synthesized to cloning of
3¢ UTR transcripts in the pRH5¢ vector.
This supplementary material can be found in the
online version of this article.
Please note: As a service to our authors and readers,
this journal provides supporting information supplied
by the authors. Such materials are peer-reviewed and
may be re-organized for online delivery, but are not
copy-edited or typeset. Technical support issues arising
from supporting information (other than missing files)
should be addressed to the authors.
PUF proteins in Arabidopsis C. W. Francischini and R. B. Quaggio
5470 FEBS Journal 276 (2009) 5456–5470 ª 2009 The Authors Journal compilation ª 2009 FEBS