Preferential translation of Hsp83 in Leishmania requires
a thermosensitive polypyrimidine-rich element
in the 39 UTR and involves scanning of the 59 UTR
MAYA DAVID,1IDAN GABDANK,2MIRIAM BEN-DAVID,1ALON ZILKA,1IRIT ORR,1DANNY BARASH,2
and MICHAL SHAPIRA1
1Department of Life Sciences, Ben Gurion University of the Negev, Beer Sheva 84105, Israel
2Department of Computer Sciences, Ben Gurion University of the Negev, Beer Sheva 84105, Israel
Heat shock proteins (HSPs) provide a useful system for studying developmental patterns in the digenetic Leishmania parasites,
since their expression is induced in the mammalian life form. Translation regulation plays a key role in control of protein coding
genes in trypanosomatids, and is directed exclusively by elements in the 39 untranslated region (UTR). Using sequential
deletions of the Leishmania Hsp83 39 UTR (888 nucleotides [nt]), we mapped a region of 150 nt that was required, but not
sufficient for preferential translation of a reporter gene at mammalian-like temperatures, suggesting that changes in RNA
structure could be involved. An advanced bioinformatics package for prediction of RNA folding (UNAfold) marked the
regulatory region on a highly probable structural arm that includes a polypyrimidine tract (PPT). Mutagenesis of this PPT
abrogated completely preferential translation of the fused reporter gene. Furthermore, temperature elevation caused the
regulatory region to melt more extensively than the same region that lacked the PPT. We propose that at elevated temperatures
the regulatory element in the 39 UTR is more accessible to mediators that promote its interaction with the basal translation
components at the 59 end during mRNA circularization. Translation initiation of Hsp83 at all temperatures appears to proceed
via scanning of the 59 UTR, since a hairpin structure abolishes expression of a fused reporter gene.
Keywords: Leishmania; translation regulation; Hsp83; 39 UTR; polypyrimidine tract; scanning of 59 UTR
Trypanosomatids are ancient eukaryotes that belong to the
kinetoplastid order and are known for their unusual
molecular features. Transcription of mRNAs is polycis-
tronic, and the primary transcripts are further processed by
trans-splicing and polyadenylation, to yield the monocis-
tronic mature transcripts. To date, no conventional RNA
pol II promoters for activating transcription of protein
coding genes were identified (Martinez-Calvillo et al. 2003),
although epigenetic effects appear to have a functional
impact (Siegel et al. 2009). As a result, steady-state levels of
differentially expressed genes are controlled mainly by post-
transcriptional events such as mRNA processing and sta-
bility (Clayton and Shapira 2007). Recent evidence ob-
tained from comparing the Leishmania transcriptome with
its proteome indicates that translation regulation plays a
key role in assigning the developmental pattern of gene
expression (Rosenzweig et al. 2008).
Heat shock genes are useful as a model system for
differential translation during the life cycle of Leishmania,
as their de novo synthesis increases dramatically at mam-
malian-like temperatures (Shapira et al. 1988; Garlapati
et al. 1999). This pattern of regulation can be efficiently
conferred onto a reporter gene when flanked by intergenic
regions (IRs) that are derived from the Hsp83 gene cluster,
as these provide signals for mRNA processing and trans-
lation. Using transgenic Leishmania cell lines we previously
showed that preferential translation of a reporter gene that
is fused to the Hsp83 IRs is controlled almost exclusively
by the 39 untranslated region (UTR). The 39 UTR alone
confers a pattern of regulation similar to that of the
endogenous gene, whereas the 59 UTR has only a synergistic
effect, and by itself cannot cause an increase in translation
during heat shock (Zilka et al. 2001). This pattern of
Reprint requests to: Michal Shapira, Department of Life Sciences, Ben
Gurion University of the Negev, POB 653, Beer Sheva 84105, Israel; e-mail:
email@example.com; fax: 972-8-6461710.
Article published online ahead of print. Article and publication date are
RNA (2010), 16:364–374. Published by Cold Spring Harbor Laboratory Press. Copyright ? 2010 RNA Society.
regulation in Leishmania varies from that of higher eu-
karyotes, since the 59 UTRs of the Drosophila and human
Hsp70 genes can efficiently induce preferential translation
of a reporter gene at elevated temperatures (Lindquist
1980). Former analysis of the 39 UTR of Leishmania
Hsp83 (888 nucleotides [nt]) using sequential deletions,
mapped the regulatory region to positions 201–472. How-
ever, this element was required, but not sufficient to induce
preferential translation onto the reporter gene. The mini-
mal region that could confer this pattern of expression
included the proximal half of the 39 UTR, positioned
between nucleotides 1 and 472. Interestingly, preferential
translation of the CAT transcript fused to this segment
occurred despite its reduced stability at elevated tempera-
tures (Zilka et al. 2001).
In prokaryotes, translation initiation is promoted by
base-pairing between the Shine–Dalgarno (SD) element
and a complementary sequence in the 16S rRNA. Hence,
accessibility of the SD element is fundamental for trans-
lation initiation to occur (Narberhaus et al. 2006). A
thermosensing mechanism caused by alterations in the
RNA structure was reported for several operons that en-
code mainly small heat shock proteins in various rhizobial
species (Chowdhury et al. 2003, 2006). Another example
for a thermosensing RNA is rpoH, which encodes the heat
shock factor sigma 32. Translation of its mRNA is up-
regulated in response to temperature elevation, as a result
of structural alterations in the 59 UTR. At normal temper-
atures the SD box is part of a double-stranded region and is
therefore not accessible to the 16S rRNA. During heat
shock this region melts and the SD element becomes single-
stranded, allowing it to pair with the 39 end of the 16S
rRNA and this promotes translation at elevated tempera-
tures (Morita et al. 1999).
Translation initiation in eukaryotes and prokaryotes
follow different pathways. The eukaryotic initiation com-
plex usually assembles on the cap structure at the 59 end of
the transcript and scans the 59 UTR until it reaches the first
AUG codon, where the ribosome assembles and translation
initiates. Alternatively, translation initiation can proceed in
a cap-independent manner that involves complex struc-
tures in the 59 UTR. In such cases, the small ribosomal
subunit binds directly to a region adjacent to the first AUG,
denoted as the internal ribosome entry site (IRES) (Holcik
et al. 2000). It has been suggested that, since cap-dependent
translation is inhibited during thermal stress, translation of
heat shock transcripts may be mediated by cap-indepen-
dent translation (Hernandez et al. 2004) or by ribosome
shunting (Yueh and Schneider 2000).
To examine whether the temperature switches which are
experienced by the parasites during their life cycle affect
the structure of the regulatory element in the Hsp83
39 UTR, we combined experimental and computational
approaches. We show that the regulatory region folds into
a highly probable structure, which contains a polypyrimi-
dine tract (PPT) and partially melts at mammalian-like
A region of 150 base pairs in the 39 UTR is required
for preferential translation of Hsp83
Temperature switches are a key parameter that drives stage-
specific expression during the life cycle of Leishmania
parasites. Heat shock transcripts are preferentially trans-
lated at temperatures typical of mammalian hosts, while
translation of cytoplasmic proteins is reduced. Different
species vary in their temperature sensitivities, thus stage
transformation occurs at a defined range of temperatures
(Shapira et al. 1988; Garlapati et al. 1999). The 39 UTR
plays a key role in translation regulation of stage-specific
genes such as Hsp83. Previous studies using sequential
deletions mapped the Hsp83 regulatory region to positions
201–472 in the 39 UTR. To narrow down the sequences that
are essential for conferring this pattern of regulation six
block deletions were introduced into the 39 UTR, removing
sequences 198–274, 249–286, 280–365, 346–399, 380–441,
and 428–476. The mutated IR regions were used to
construct a series of pHCH expression vectors, in which
the CAT gene (C) was flanked by a complete upstream IR
(H) and a mutated downstream IR (H) (Fig. 1A). The
downstream IR extended until the next coding region, to
maintain the signals for RNA processing (LeBowitz et al.
1993). Each of the six deletion constructs was intro-
duced into the pX expression vector and further used
to generate stable cell lines. Correct polyadenylation
was validated by 39 rapid amplification of cDNA ends
(39 RACE) analysis.
To follow the effect of heat shock on de novo synthesis of
the CAT reporter gene, cells were preincubated and then
metabolically labeled at 26°C, 33°C, and 37°C. Analysis of
the labeled proteins on SDS-polyacrylamide gels (Fig. 1B)
showed that deletions 198–274, 249–286, and 280–365
completely prevented preferential translation of the CAT
gene, whereas all the deletions within positions 346–476
had no such effect. It was therefore concluded that
nucleotides 198–346 were required for driving preferential
Incorporating the Hsp83 segment 201–472 within the
a-tubulin IR that was cloned downstream from the CAT
reporter gene failed to confer preferential translation at
elevated temperatures (data not shown), possibly due to its
spatial inaccessibility and spacing considerations. Further-
more, the minimal 39 UTR fragment that could confer
preferential translation consisted of its proximal half,
mapped to positions 1–472. Deletion of the distal half
(nucleotides 472–873) did not change this expression
pattern (Zilka et al. 2001). The relatively large size of the
39 UTR fragment that can independently confer preferential
Translation regulation of Hsp83 in Leishmania
translation raised the possibility that structural changes
RNA secondary structure prediction
for the regulatory region 1–472
RNA secondary structure can be predicted by software
packages that are based on energy minimization, such as
Mfold (Zuker 2003) and the Vienna RNA package
(Hofacker 2003). These methods give reliable predictions
for relatively short RNA segments (up to z150 nt), whereas
for larger fragments they generate multiple outputs. In this
study we used the recently developed algorithm 3.4-1
version of UNAfold with Mfold utilities version 4.0
(Markham and Zuker 2008) to generate a dot plot and
further give a visual impression of how ‘‘well defined’’ the
folding is (Zuker and Jacobson 1998). The predicted
structure is shown in Figure 2. Interestingly, the different
deletions that eliminate preferential translation (nucleo-
tides 201–346) are positioned on a highly probable sec-
ondary structure, in which the nucleotides are marked as
The PPT at positions 312–341 is
essential for preferential translation
at elevated temperatures
The Hsp83 ‘‘regulatory arm’’ (nucleo-
tides 201–364) (Fig. 2) contains a long
stretch of polypyrimidines, predicted to
be partly single- and partly double-
stranded. To test whether the PPT is
essential for preferential translation of
Hsp83 at elevated temperatures, small
deletions and targeted replacement mu-
tations were introduced into the 39 UTR
(Fig. 3A) and tested as described above.
The effects of these mutations on de
novo synthesis of the CAT–Hsp83 chi-
mera are shown in Figure 3B. Deletion
or exchange of the polypyrimidines that
are predicted to be single-stranded
(mutations D312–331 and the C-to-G
exchanges at positions 315–327), as well
as double-stranded (mutation D332–
341), eliminated preferential translation
of the reporter gene at elevated temper-
atures. Further downstream deletions
eliminating nucleotides that contain
non-PPT regions (D342–349) or PPT
elements (D350–364) did not interfere
with the induced translation at elevated
temperatures (Fig. 3B). It appears that
the regulatory element consists of the
PPT that is located between positions
312 and 341, and comprises both single-
and double-stranded regions (D312–331 and D332–341,
Western analysis showed that, similar to Hsp83 (Argaman
et al. 1994), a steady-state level of CAT expression was
observed at 26°C. Furthermore, it appeared that most of the
fine mutations caused a partial reduction in the steady-state
level of CAT expression, irrespective of their effect on de
novo synthesis at elevated temperatures (Fig. 3C).
Thermal melting of the regulatory RNA region
RNA is subject to thermal melting; however, different
RNAs may acquire variable melting patterns. To examine
how temperature elevation affects the regulatory element,
we tested its melting by monitoring changes in the UV
absorbance at 260 nm as a function of temperature. It has
been shown in the past that a collapse in the secondary
structure of nucleic acids in response to extreme conditions
leads to an increase in the optical density at 260 nm, by
z15%–20% (Thomas 1951, 1993). Thus, the thermal
stability of the wild-type and mutated 1–472 RNA frag-
ments was evaluated by UV spectroscopy. The RNAs were
FIGURE 1. Preferential translation of Hsp83 is directed by sequences 201–346 in the 39 UTR.
(A) The CAT reporter gene was flanked by complete IRs derived from the Hsp83 genomic
cluster, so that the signals for RNA processing were maintained on both sides. The resulting
constructs were cloned into the pX transfection vector of Leishmania and used to generate
stable cell lines. (B) De novo synthesis of the CAT reporter gene was examined by metabolic
labeling of the transgenic parasite cells grown at 26°C, and after their preincubation at 33°C, or
37°C, during 60 min. Protein extracts were separated over 12% polyacrylamide gels. Migration
of CAT, Hsp70, Hsp83, and the a- and b-Tubulins are marked with arrowheads.
David et al.
RNA, Vol. 16, No. 2
incubated at temperatures that increased by increments of
5°C, between 10°C and 45°C. To compare between the
melting patterns of the different RNA elements on a single
plot, we expressed the DOD260values for each curve as
relative numbers that ranged from 0% to 100% (the
DOD260at the highest temperature served as 100%). The
absorbance curves (Fig. 4) indicate a faster melting pattern
for the RNA fragments that direct preferential translation at
elevated temperatures (wild type and D350–364), as com-
pared to the noninducing mutated RNA fragments (D332–
341 and C to G exchange 315–327). The structure pre-
diction of UTR fragments containing mutations that in-
terfere with preferential translation at elevated tempera-
tures gives a somewhat lower dG values, as compared to
wild-type sequences (Supplemental Fig. 1).
The effect of temperature elevation on the secondary
structure of RNA within the regulatory region was also
monitored by the RNase H assay. RNase H cleaves RNA
which is hybridized to a complementary DNA strand. Thus
a short oligonucleotide can direct the cleavage of its
complementary RNA sequence, given that the latter is
single-stranded and accessible for hybridization with the
oligonucleotide. Double-stranded RNA regions are there-
fore resistant to the enzymatic cleavage.
The RNA fragment that corresponds to sequences 1–472
of the Hsp83 39 UTR was transcribed in vitro and labeled at
its 59 end. The radiolabeled RNA was incubated at 26°C
and at 37°C, annealed with specific antisense oligonucleo-
tides and subjected to cleavage by RNase H (Fig. 5A).
Analysis of the resulting fragments over denaturing gels
showed increased cleavage at 37°C in the presence of
oligonucleotides 317–329, 333–346, and 265–278, but not
with oligonucleotides 248–261 and 361–375 (Fig. 5B).
Although the PPT at positions 317–329 is predicted to be
single-stranded, its increased cleavage at elevated temper-
atures could indicate that this region stacks with short
complementary purines, which are found along the regu-
latory region 1–472, to create a complex tertiary structure.
Temperature elevation may melt these interactions, as
appears from the RNase H cleavage pattern. Cleavage of
the RNA fragment that corresponds to sequences 248–261
does not comply with the fact that part of this region is
predicted to be double-stranded. In view of this result, we
suggest that the short predicted stem that encompasses
two pairs of nucleotides (255 with 260 and 254 with
261) is mostly open, generating a large loop between
positions 252 and 267, which is constantly open. This
stem indeed is colored with yellow, indicating that it is
not ‘‘highly defined.’’ The results described above show
that the regulatory region is thermosensitive and that
specific regions are subject to melting that may increase
their accessibility to regulatory components during trans-
Translation of Hsp83 initiates by scanning
of the 59 UTR
The results obtained thus far indicate rather unequivocally
that preferential translation of Hsp83 in Leishmania is
mediated by a regulatory element in the 39 UTR unlike in
higher eukaryotes, where preferential translation of Hsps
requires the 59 UTR. However, at this stage, it is still
unclear how the 39 UTR promotes preferential translation
in these organisms. There are multiple reports on IRES-
mediated translation of Hsps in higher eukaryotes, which
does not involve scanning of the 59 UTR. We therefore
tested whether, despite the fact that the Hsp83 59 UTR is
FIGURE 2. Structure prediction of the Hsp83 39 UTR element 1–472
by UNAfold. Color annotation was used to indicate the propensity of
individual nucleotides to participate in base-pairing and whether or
not a predicted base pair is well determined. Forty colors that range
from red (unusually well determined) through orange, yellow, green,
blue, purple to black (poorly determined) are used (Zuker and
Jacobson 1998). The structure with the lowest DG that was obtained
using the Mfold program is shown, and the color of each nucleotide
indicates its estimated P-value. The deletions which are described in
Figure 1 are positioned on the predicted structure. Deletions that
abolished preferential translation are colored in green (D198–274),
purple (D249–286), and blue (D280–365). Deletions that did not
affect preferential translation are marked in gray (D346–399, D380–
441, and D428–476).
Translation regulation of Hsp83 in Leishmania
FIGURE 3. (Legend on next page)
David et al.
RNA, Vol. 16, No. 2
exchangeable, preferential translation of the Hsp83–CAT
hybrid mRNA proceeds via scanning of the 59 UTR.
To examine whether translation of Hsp83 initiates by
scanning of the 59 UTR, or whether the ribosome can
possibly target itself to the vicinity of the first AUG, a stable
hairpin structure was introduced in the middle of the
Hsp83 59 UTR. This structure consisted of five BamHI
linker repeats and was introduced at position 121 of the
z320-nt long 59 UTR (Fig. 6A). Addition of this strong
secondary structure prevented CAT translation at either
temperature, 26°C, 33°C, and 37°C, suggesting that the
scanning of the 59 UTR was interrupted (Fig. 6B). Fur-
thermore, the presence of the hairpin structure at the
59 UTR totally prevented any steady-state expression of
the CAT reporter, as shown by Western analysis (Fig. 6C).
Since preferential translation of the Hsp83 gene is
mediated by the 39 UTR, we also examined the effect of
a hairpin structure that was introduced downstream from
the regulatory region (at position 655). As expected, this
modification had no effect on de novo synthesis of the CAT
reporter mRNA (Fig. 6B).
In trypanosomatids, stability and translation of stage-
specific transcripts are regulated almost exclusively by
39 UTRs, yet little is known on their mode of function.
Since the developmental program of gene expression in
these digenetic parasites is directed by changes in temper-
ature and pH, heat shock genes provide an ideal system for
understanding how environmental switches are perceived
by the molecular machinery of these organisms.
In this study we performed a fine mapping of the 39 UTR
to characterize the element that is required for preferential
translation of the Leishmania Hsp83. The experiments are
based on former studies that identified the 39 UTR as the
region that confers this pattern of expression, originally
highlighting a rather large regulatory region between
nucleotides 201 and 472 (Zilka et al. 2001). However, the
minimal fragment that could independently induce de
novo synthesis at elevated temperatures consisted of the
proximal half of the 39 UTR (nucleotides 1–472). Implant-
ing a smaller fragment (nucleotides 201–472) in a nonheat
shock 39 UTR could not confer this effect, suggesting that it
possibly failed to assign a proper folding, or that it was not
accessible to the translational machinery.
The computer-based secondary structure prediction re-
veals that the regulatory region in the Hsp83 39 UTR is
found on a well-defined arm that contains a long stretch of
pyrimidines. The direct involvement of this PPT in pref-
erential translation during heat shock was established by
mutational analysis. A targeted deletion of the single- and
double-stranded pyrimidine-rich regions (D312–331 and
D332–341, respectively) prevented preferential translation
of the CAT reporter transcript at elevated temperatures.
Removal of sequences located further downstream (342–
349 and 350–364) had no such effect. Altogether, the
regulatory element lies between nucleotides 312 and 341.
The sequence of the Hsp83 39 UTR element is highly
conserved among different Leishmania species (Supple-
mental Fig. 2), but not with Trypanosoma brucei. However,
a partially conserved pyrimidine-rich region, (71%; 460-
CCACCUCACGUUCCUUUCCC-480), was found in the
39 UTR of T. brucei Hsp83. UNAfold analysis predicted that
FIGURE 3. Sequences 312–341 in the Hsp83 39 UTR are essential for preferential translation during heat shock. (A) A map of mutations within
the region 150–364 in the 39 UTR of Hsp83. The modified IRs were cloned downstream from the CAT gene to generate 39 UTRs. The upstream
Hsp83 IR was not modified. The chimeric CAT–Hsp83 genes were cloned in the pX vector for transfection of Leishmania, and used to generate
transgenic parasite cell lines. (B) Functional fine mapping of sequences that are required for preferential translation of the CAT–Hsp83 chimeric
gene. Cells expressing the CAT gene under control of the mutated Hsp83 39 UTR were grown at 26°C, or transferred to 33°C or 37°C for 1 h, and
metabolically labeled for 30 min at the corresponding temperatures with35[S]-labeled amino acids. Proteins were extracted, separated over 15%
SDS-polyacrylamide gels, and autoradiogrammed in a PhosphorImager. Migration of Hsp83. Hsp70, the a- and b-tubulins, and CAT are marked
with arrowheads. WT represents the nontransfected negative control cells which do not express CAT. pX-HCH represents transgenic cells
expressing CAT under control of nonmodified Hsp83 IRs. Deletion mutations D312–331, D337–364, D332–341, and the C/G exchange
mutation at positions 315–327 abrogated preferential translation of the CAT transcript. However, deletion mutations D350–364 and D342–349
did not interfere with the increased CAT translation at elevated temperatures. (C) Steady-state CAT expression at 26°C in cells transfected with
the different 39 UTR deletions. Cell extracts were separated over 15% SDS-polyacrylamide gels, blotted, and reacted with anti-CAT antibodies.
Protein loads were evaluated by control reactions with antibodies against Hsp70. The level of CAT expression was quantified with Multigauge
V3.0 and normalized against Hsp70. The values shown at the bottom of the figure represent a mean of two independent experiments.
FIGURE 4. RNA melting curves of the wild-type and mutated 1–472
RNA fragments. Representative melting profiles were obtained by
measuring the optical density of the RNA solutions (0.012 mg/mL) at
temperatures that increased by increments of 5°C. The RNA was
allowed to equilibrate for 10 min at each temperature, prior to
monitoring the absorbance at 260 nm. The change in absorbance for
each temperature (DOD260) was calculated relative to the starting
point. The DOD260values for each curve are expressed as relative
numbers that range from 0% to 100% (the DOD260at the highest
temperature served as 100%).
Translation regulation of Hsp83 in Leishmania
this element is located within a well-defined structural arm
(Supplemental Fig. 3), which is reminiscent of the parallel
structure in the Leishmania transcripts. The actual in-
volvement of this element in preferential translation is yet
to be examined experimentally.
A similar picture is revealed for the Leishmania Hsp70
genes, where the downstream untranslated region directs
their developmental expression pattern (Quijada et al.
2000). The Hsp70 gene cluster in Leishmania contains six
copies. The transcript of the first five copies increases at
elevated temperature and is translated with higher effi-
ciency (Folgueira et al. 2005), similar to Hsp83. The
transcript of the sixth copy contains a different 39 UTR
(Quijada et al. 1997) and follows a different regulatory
pathway. The UNAfold package predicts that only the
39 UTR of the heat-inducible Hsp70 I transcript contains
a well-defined structural arm with a short PPT (data not
shown). This finding is reminiscent of Hsp83 but should be
further tested for its involvement in directing preferential
translation of Hsp70.
Pyrimidine-rich regions serve as common regulatory
elements during RNA processing, and their function is
mediated by binding of specific splicing factors, such as the
PPT binding protein (PTB). In addition to its role in
alternative splicing (Singh et al. 1995; Perez et al. 1997),
PTB was also implicated as an IRES trans-acting auxiliary
factor (ITAF) that enhances cap-independent translation
via the 59 UTR (Mitchell et al. 2005; Song et al. 2005;
Spriggs et al. 2005; Bushell et al. 2006). More recent reports
indicate that PTB can promote translation through the
39 UTR, as shown for the hypoxia-inducible factor (HIF-1a)
(Galban et al. 2008) and of the ATP-Synthase b subunit
(Reyes and Izquierdo 2007). However, it can also act as
a translation repressor, based on studies with the 39 UTR of
Drosophila osk (Besse et al. 2009). It has been shown that
the human PTB1 can bind to PPTs of different lengths,
structured or single-stranded (Clerte and Hall 2009), thus
PTB can readily facilitate the formation of RNA–protein
regulatory complexes. The function of the PPT in the
Hsp83 39 UTR is not clear yet, and we do not have evidence
that PTB is involved. However, preliminary results show
that PTB1 from Leishmania (LmjF04.1170), a homolog of
the T. brucei protein (Stern et al. 2009), can interact with
the recently identified eIF4G of Leishmania (LeishIF4G-3;
Yoffe et al. 2009) in a yeast-two hybrid assay (data not
shown). We hypothesize that the regulatory PPT that we
identified in the 39 UTR of Hsp83 (nucleotides 312–334)
may function as a binding target for a regulatory protein
that in turn interacts with the basal translation initiation
The regulatory region in the 39 UTR of the Leishmania
Hsp83 transcript is thermosensitive and undergoes a limited
melting process, as observed by the RNase H assay, and by
monitoring changes in the optical density of the RNA. RNA
fragments that direct preferential translation during heat
shock (wild type and D350–364) seem to melt faster than
the noninducing fragments (D332–341 and C ! G ex-
change between positions 315 and 327). Melting of the
regulatory region may not be the only parameter that leads
to preferential translation. However, in view of the circular
model for translation initiation (Gingras et al. 1999), one
can speculate that changes in the melting status of the
39 UTR and spatial exposure of its regulatory elements can
affect their accessibility to molecules that also interact with
FIGURE 5. (Legend on next page)
David et al.
RNA, Vol. 16, No. 2
the translation initiation complex. The partial compatibility
between the RNA structure prediction and the RNase H
data may stem from the inability of current computer
algorithms to deal with tertiary RNA structures. Thus,
although the use of the older Mfold algorithm to predict
RNA folds at different temperatures did not reveal changes
in the ‘‘highly defined’’ region, tertiary and spatial alter-
ations as well as base stacking effects cannot be excluded.
Temperature elevation seems to assign a multitude of
regulatory traits on protein synthesis in Leishmania. Deci-
phering the mechanistic basis for this regulation is important
for our understanding of stage differentiation processes.
The 39 UTRs of several amastigote-specific genes contain
a conserved ‘‘amastin element’’ that is known to increase
their translation in the mammalian life form of the parasite
(Boucher et al. 2002). The amastin element has no sequence
conservation with the 39 UTR of Hsp83, and no structural
conservation as well. In fact, the UNAfold predictions of
this element do not reveal any well-defined structures as
in the Hsp83 transcript, suggesting that the two pathways
In Drosophila, the 59 UTR of Hsp70 mRNA is required
for efficient translation (Klemenz et al. 1985; McGarry and
Lindquist 1985). It was previously shown that cap-
dependent translation is inhibited during thermal stress
(Hellen and Sarnow 2001), and that translation initiation
of Hsp70 may be cap-independent and mediated by IRES
elements (Hernandez et al. 2004). Previous reports on
human Hsp70 showed that insertion of a stable hairpin
in the 59 UTR near the AUG initiation codon still allows
direct translation of a reporter gene, excluding a scanning
mechanism (Yueh and Schneider 2000). Other reports
suggested that, similar to Drosophila, translation of the
human Hsp70 is mediated by an IRES (Rubtsova et al.
For Hsp90, however, a dual regulation was established
(Hsp90 is parallel to Hsp83 in Leishmania). The 59 UTR of
Drosophila Hsp90 possesses significant secondary structure
elements that are typical to nonheat shock genes (Ahmed
and Duncan 2004). It was suggested that temperature
elevation reprograms translation of Hsp90 mRNA from
cap-dependent to a cap-independent mechanism. This was
shown by experiments in which translation of Hsp90 at a
normal growth temperature was strongly inhibited by treat-
ment with rapamycin, a drug that blocks cap-dependent
FIGURE 6. Preferential translation of Hsp83 occurs via scanning
of the 59 UTR. (A) Introduction of a hairpin structure at the Hsp83
59 and 39 UTRs. A scheme of plasmids carrying the CAT gene flanked
with Hsp83 IRs, with a foreign hairpin structure introduced either to
the 59 or to the 39 UTRs, is shown. The CAT constructs were cloned
into the pX transfection vector and used to generate transgenic
Leishmania lines. (B) A hairpin structure introduced in the 59 UTR
has an inhibitory effect on CAT translation at both temperatures. Cells
expressing the CAT gene under control of the Hsp83 IRs carrying
a hairpin structure either at the 59 or at the 39 UTR were incubated for
1 h at different temperatures, 26°C, 33°C, or 37°C, and metabolically
labeled with35[S]-labeled methionine and cystein during 30 min at
the corresponding temperatures. Proteins were extracted and sepa-
rated over 15% SDS-polyacrylamide gels. The migration distances of
Hsp83, Hsp70, tubulin, and the CAT reporter gene are marked by
arrows. Introduction of a hairpin structure into the 59 UTR inhibited
the de novo translation of the CAT–Hsp83 chimera at all tempera-
tures. (C) Steady-state CAT expression in cells transfected with
a CAT–Hsp83 chimera carrying a hairpin structure at the 59 (pX-
HCH-59hp) and 39 (pX-HCH-39hp) UTRs. Cell extracts were sepa-
rated over 15% SDS-polyacrylamide gels, blotted, and reacted with
anti-CAT antibodies. Protein loads were evaluated by control re-
actions with antibodies against Hsp70.
FIGURE 5. RNA melting measured by RNase H cleavage. (A)
Location of the antisense oligonucleotides on the predicted RNA
structure. The oligonucleotides used for the RNase H assay are
positioned on the predicted RNA structure (black lines), and the
ratio between the intensity of the two cleavage products is indicated
(25/37). Black lettering represents regions that melt at elevated
temperatures (25/37 < 1) and gray lettering marks regions that are
not affected by temperature elevation (25/37 = 1). (B) RNase H
cleavage directed by hybridization of the RNA fragment preincubated
at different temperatures. The end-labeled RNA (1–472) was exposed
to different temperatures (25°C or 37°C), then incubated with
different oligonucleotides and cleaved by RNase H at the correspond-
ing temperature. The RNase H cleavage products were separated over
6% denaturing polyacrylamide gels. Migration of the untreated full-
length 1–472 RNA product is shown at the left. Products of the RNase
H cleavage reaction following hybridization with oligonucleotides at
different temperatures are marked with a star. The oligonucleotide
positions in the 39 UTR are indicated above the lanes. Size markers of
100, 200, and 300 nt are depicted at the left side of each panel
(numbers at the far left and short lines between the panels). The ratio
between the intensity of the two bands (25/37) is shown at the bottom
of each panel.
Translation regulation of Hsp83 in Leishmania
protein synthesis. Such sensitivity was not observed during
heat shock (Duncan 2008).
Insertion of a strong secondary structure in the Hsp83
59 UTR prevents translation at normal temperatures as well
as during heat shock, indicating that initiation proceeds
through scanning of the 59 UTR at all temperatures. It
appears that the regulatory mechanism that drives prefer-
ential translation of Hsp83 in Leishmania varies dramati-
cally from the parallel mechanism in higher eukaryotes.
This is reflected not only by the key role assigned to the
39 UTR, but also by the strict requirement for scanning of
the 59 UTR at elevated temperatures. These findings most
probably rule out the involvement of an IRES element in
translation regulation of Hsp83 in Leishmania.
MATERIALS AND METHODS
Leishmania amazonensis isolate MHOM/BR/77/LTB0016 was
cultured in Schneider’s medium supplemented with 10% fetal
calf serum (FCS), 4 mM L-glutamine, and 25 mg/mL gentamycin.
Parasites were also grown in RPMI supplemented with 10% FCS,
4 mM L-glutamine, 25 mg/mL gentamycin, biotin 0.0001%, hemin
0.0005%, biopterin 0.002 mg/mL, HEPES 40 mM, and adenine
Site-directed mutagenesis was performed by overlap extension
PCR, or by one-step mutagenesis using primers that contained the
target mutation, on a pBluescript-based plasmid in which the
Hsp83 IR (M92925) was cloned between the PstI and SalI
restriction sites (pKS83PS). A list of primers that was used for
the mutagenesis is provided in Supplemental Table 1. The
mutated IR was excised from the pKS83PS plasmid by cleavage
with BclI and BamHI and the fragment was ligated into the
BamHI site downstream from the CAT coding gene, in a pX
plasmid that already contained the CAT coding gene (C) fused
with an upstream IR from the Hsp83 (H) cluster (pXHC). Thus,
the mutated IR was introduced downstream from the CAT coding
gene. Modification of the 59 UTR was done in a similar approach,
and the alteration was performed on a plasmid that contained the
IR of Hsp83 (pKS83). The modified IR was excised by BclI and
cloned into the BamHI site of pXCH.
Isolation of RNA and 39 RACE analysis
Total RNA was extracted from cells incubated at 26°C, and at
different time points following their transfer to 33°C with the
TRI Reagent (Scientific Research Laboratories). The proper poly-
adnylation sites were verified by 39 RACE, as previously described
(Zilka et al. 2001).
Plasmid DNA was electroporated into L. amazonensis parasites as
described (Laban and Wirth 1989), except that a double electrical
pulse of 5.5 kV/cm at 25 mF in a Bio-Rad Gene Pulser apparatus
was applied. Stably transfected lines were selected in the presence
of 60 mg/mL Geneticin (G418, Sigma). Neomycin-resistant
parasites appeared 10–14 d following transfection and were grown
in the presence of 200 mg/mL Geneticin.
Parasites were grown to a cell density of 3 3 107/mL. Cells (1 mL)
were preincubated at 26°C, 33°C, and 37°C for 1 h and then
labeled with 20 mCi of [35S] cystein-methionine protein labeling
mix (1175 Ci/mmol) for 30 min at the corresponding tempera-
tures. Following labeling, the cells were harvested at 4°C, washed
twice with cold phosphate buffered saline (PBS) and lysed in SDS-
PAGE sample buffer. Incorporation of the [35S] amino acids was
measured by precipitation with trichloroacetic acid (TCA). Pro-
tein samples containing the same amount of incorporated radio-
label corresponded to a similar number of cells and were separated
over 15% SDS-polyacrylamide gels. The gels were dried and
further processed for fluorography. Migration of the CAT reporter
gene was validated by the use of specific anti-CAT antibodies.
In vitro transcription and end labeling of RNA
RNA that corresponds to positions 1–472 in the 39 UTR of Hsp83
was transcribed in vitro from the T7 promoter, using a PCR-based
DNA template. Template DNA derived from the Hsp83 39 UTR
was amplified by PCR from pKS83 with the forward primer 4–21s
(59-GGTACGGCAGCGGCACAC-39) and the reverse primer 450–
lymerase promoter was added in a reamplification reaction with
the forward primer 4–19T7s (59-TAATACGACTCACTATAGGTA
CGGCAGCGGCAC-39; the promoter sequences are underlined)
and the reverse primer 450–472as (see above). Amplification was
performed in a total volume of 50 mL with 50 pmol of each
primer, 100 mM of each dNTP, and 4 U of BIO-X-ACT short
DNA polymerase (Bioline) in a buffer provided by the manufac-
turer. Cycling conditions were as follows: 5 min denaturation at
95°C, followed by 29 cycles of denaturation at 95°C for 30 sec,
annealing at 57°C for 40 sec, and elongation at 68°C for 40 sec.
Reamplification was performed under the same conditions, except
for the annealing temperature, which was 59°C.
Transcription was performed with 1 mg of PCR product as
template for 2 h at 37°C in a total volume of 100 mL. The reaction
contained 40 U of T3 RNA polymerase (Promega), 2.5 mM of
each rNTP, buffer (provided by the manufacturer: 40 mM Tris at
pH 7.9, 6 mM MgCl2, 2 mM spermidine, and 10 mM NaCl), and
100 U of RNasin (Promega). The reaction was stopped by freezing
in liquid nitrogen. Five units of DNase RNase free (Promega) were
added and the reaction was incubated for additional 30 min at
37°C. The DNase was inactivated by addition of 0.2 M EDTA at
pH 8 (2 mL) and by incubation at 65°C for 20 min. Loading buffer
was added (95% formamide, 20 mM EDTA at pH 8, 0.05%
bromophenol blue, 0.05% xylene blue) and samples were sepa-
rated on a 4% polyacrylamide/7 M urea gel. Following electro-
phoresis the product was excised and eluted in 600 mL of RNA
extraction buffer (0.3 M NaAc, 0.1% SDS, 1 mM EDTA at pH 8)
overnight at room temperature. The transcript was purified over
a column of the RNeasy MinElute cleanup kit (Qiagen) and eluted
in 30 mL of DEPC-treated ddH20.
For 59 end labeling the RNA (z0.1 mg) was dissolved in water,
incubated at 85°C for 10 min in the presence of DTT (4 mM), and
David et al.
RNA, Vol. 16, No. 2
cooled on ice for 2 min. Labeling was performed with [g32P] ATP
(20 mCi, Amersham) in a buffer containing 50 mM Tris-HCl at
pH 7.6, 10 mM MgCl2, 5 mM DTT, 0.1 mM spermidine, and 0.1
mM EDTA, by 10 U of T4 Polynucleotide kinase (Fermentas).
Unincorporated nucleotides were removed by applying the mix-
ture on a G-50 spin column, and the labeled fragment was gel-
RNase H protection analysis
RNA (z1 ng per reaction) labeled at the 59 end was incubated for
2 h at 26°C or 37°C in the reaction buffer (0.2 M Tris-HCl at pH
7.5, MgCl20.05 M, 0.02 M DTT), at a total volume of 8 mL per
reaction. The RNA was then annealed with 1 mL of a 1 mM
oligonucleotide solution for 5 min at the corresponding temper-
ature. One unit of RNase H (Ambion) was added to the reaction
mixture followed by a further incubation of 5 min at the
corresponding temperature. The reaction was stopped by the
addition of loading buffer and samples were separated over 6%
polyacrylamide/7 M urea gels. Dried gels were visualized by
Analysis by UV melting curves
RNA transcribed in vitro was adjusted to z0.3 OD at 260 nm.
The RNA was heated at increments of 5°C in ddH2O, or PBS at
pH 7 and maintained at each temperature for 10 min prior
to measuring absorbance in a JASCO V530 spectrophotometer
equipped with a cell holder that was used to maintain the re-
quested temperature by a circulating bath. Measurements were
taken at temperature increments that ranged between 10°C and
50°C. The OD260values for the melting process were normalized
and plotted against the temperature.
Computer prediction of the RNA secondary structure
The 3.4-1 version of UNAfold (Markham and Zuker 2005, 2008)
with Mfold utilities version 4.0 was used to determine the
probability for each nucleotide to be found in a single- or
double-stranded form. This software package, among many other
capabilities, generates a dot plot which gives a visual impression of
how well defined the folding is. The program evaluates each
nucleotide for its ability to pair with other nucleotides, and assigns
a P-value for each nucleotide. A low P-value indicates that there are
only a few base-pairing possibilities and nucleotides with P = 0 are
single-stranded. A color annotation is used to indicate the pro-
pensity of individual nucleotides to participate in base pairs and
whether or not a predicted base pair is well determined. Colors
range from red (unusually well determined) to black (poorly
determined) (Zuker and Jacobson 1998).
Supplemental material can be found at http://www.rnajournal.org.
This work was supported by the German–Israel Foundation, Grant
No. 728-23/2002, and the Israel Science Foundation, Grant No.
395/09, to M.S. We thank Michael Zuker from Rensselaer Poly-
technic Institute for providing us with the UNAfold algorithm.
Received June 8, 2009; accepted November 9, 2009.
Ahmed R, Duncan RF. 2004. Translational regulation of Hsp90 mRNA.
AUG-proximal 59-untranslated region elements essential for pref-
erential heat shock translation. J Biol Chem 279: 49919–49930.
Argaman M, Aly R, Shapira M. 1994. Expression of the heat shock
protein 83 in Leishmania is regulated post transcriptionally. Mol
Biochem Parasitol 64: 95–110.
Besse F, Lopez de Quinto S, Marchand V, Trucco A, Ephrussi A. 2009.
Drosophila PTB promotes formation of high-order RNP particles
and represses oskar translation. Genes & Dev 23: 195–207.
Boucher N, Wu Y, Dumas C, Dube M, Sereno D, Breton M,
Papadopoulou B. 2002. A common mechanism of stage-regulated
gene expression in Leishmania mediated by a conserved 39-un-
translated region element. J Biol Chem 277: 19511–19520.
Bushell M, Stoneley M, Kong YW, Hamilton TL, Spriggs KA,
Dobbyn HC, Qin X, Sarnow P, Willis AE. 2006. Polypyrimidine
tract binding protein regulates IRES-mediated gene expression
during apoptosis. Mol Cell 23: 401–412.
Chowdhury S, Ragaz C, Kreuger E, Narberhaus F. 2003. Temperature-
controlled structural alterations of an RNA thermometer. J Biol
Chem 278: 47915–47921.
Chowdhury S, Maris C, Allain FH, Narberhaus F. 2006. Molecular
basis for temperature sensing by an RNA thermometer. EMBO J
Clayton C, Shapira M. 2007. Post-transcriptional regulation of gene
expression in trypanosomes and leishmanias. Mol Biochem Para-
sitol 156: 93–101.
Clerte C, Hall KB. 2009. The domains of polypyrimidine tract binding
protein have distinct RNA structural preferences. Biochemistry 48:
Duncan RF. 2008. Rapamycin conditionally inhibits Hsp90 but not
Hsp70 mRNA translation in Drosophila: Implications for the mech-
Folgueira C, Quijada L, Soto M, Abanades DR, Alonso C, Requena JM.
2005. The translational efficiencies of the two Leishmania infantum
HSP70 mRNAs, differing in their 39-untranslated regions, are
affected by shifts in the temperature of growth through different
mechanisms. J Biol Chem 280: 35172–35183.
Galban S, Kuwano Y, Pullmann R Jr, Martindale JL, Kim HH, Lal A,
Abdelmohsen K, Yang X, Dang Y, Liu JO, et al. 2008. RNA-
binding proteins HuR and PTB promote the translation of
hypoxia-inducible factor 1a. Mol Cell Biol 28: 93–107.
Garlapati S, Dahan E, Shapira M. 1999. Effect of acidic pH on heat
shock gene expression in Leishmania. Mol Biochem Parasitol 100:
Gingras AC, Raught B, Sonenberg N. 1999. eIF4 initiation factors:
Effectors of mRNA recruitment to ribosomes and regulators of
translation. Annu Rev Biochem 68: 913–963.
Hellen CU, Sarnow P. 2001. Internal ribosome entry sites in
eukaryotic mRNA molecules. Genes & Dev 15: 1593–1612.
Hernandez G, Vazquez-Pianzola P, Sierra JM, Rivera-Pomar R. 2004.
Internal ribosome entry site drives cap-independent translation of
reaper and heat shock protein 70 mRNAs in Drosophila embryos.
RNA 10: 1783–1797.
Hofacker IL. 2003. Vienna RNA secondary structure server. Nucleic
Acids Res 31: 3429–3431.
Holcik M, Sonenberg N, Korneluk RG. 2000. Internal ribosome
initiation of translation and the control of cell death. Trends Genet
Klemenz R, Hultmark D, Gehring WJ. 1985. Selective translation of
heat shock mRNA in Drosophila melanogaster depends on se-
quence information in the leader. EMBO J 4: 2053–2060.
Laban A, Wirth DF. 1989. Transfection of Leishmania enriettii and
expression of chloramphenicol acetyltransferase gene. Proc Natl
Acad Sci 86: 9119–9123.
Translation regulation of Hsp83 in Leishmania
LeBowitz JH, Smith H, Beverley SM. 1993. Coupling of polyade- Download full-text
nylation site selection and trans-splicing in Leishmania. Genes &
Dev 7: 996–1007.
Lindquist S. 1980. Translational efficiency of heat induced messages in
Drosophila melanogaster cells. J Mol Biol 137: 151–158.
Markham NR, Zuker M. 2005. DINAMelt web server for nucleic acid
melting prediction. Nucleic Acids Res 33: W577–W581.
Markham NR, Zuker M. 2008. UNAFold: Software for nucleic acid
folding and hybridization. Methods Mol Biol 453: 3–31.
Martinez-Calvillo S, Yan S, Nguyen D, Fox M, Stuart KD, Myler PJ.
2003. Transcription of Leishmania major Friedlin chromosome 1
initiates in both directions within a single region. Mol Cell 11:
McGarry TJ, Lindquist S. 1985. The preferential translation of
Drosophila hsp70 mRNA requires sequences in the untranslated
leader. Cell 42: 903–911.
Mitchell SA, Spriggs KA, Bushell M, Evans JR, Stoneley M, Le
Quesne JP, Spriggs RV, Willis AE. 2005. Identification of a motif
that mediates polypyrimidine tract-binding protein-dependent
internal ribosome entry. Genes & Dev 19: 1556–1571.
Morita MT, Tanaka Y, Kodama TS, Kyogoku Y, Yanagi H, Yura T.
1999. Translational induction of heat shock transcription factor
s32: Evidence for a built-in RNA thermosensor. Genes & Dev 13:
Narberhaus F, Waldminghaus T, Chowdhury S. 2006. RNA ther-
mometers. FEMS Microbiol Rev 30: 3–16.
Perez I, Lin CH, McAfee JG, Patton JG. 1997. Mutation of PTB
binding sites causes misregulation of alternative 39 splice site
selection in vivo. RNA 3: 764–778.
Quijada L, Soto M, Alonso C, Requena JM. 1997. Analysis of post-
transcriptional regulation operating on transcription products of
the tandemly linked Leishmania infantum hsp70 genes. J Biol Chem
Quijada L, Soto M, Alonso C, Requena JM. 2000. Identification of
a putative regulatory element in the 39-untranslated region that
controls expression of HSP70 in Leishmania infantum. Mol
Biochem Parasitol 110: 79–91.
Reyes R, Izquierdo JM. 2007. The RNA-binding protein PTB exerts
translational control on 39-untranslated region of the mRNA for
the ATP synthase b-subunit. Biochem Biophys Res Commun 357:
Rosenzweig D, Smith D, Opperdoes F, Stern S, Olafson RW,
Zilberstein D. 2008. Retooling Leishmania metabolism: From sand
fly gut to human macrophage. FASEB J 22: 590–602.
Rubtsova MP, Sizova DV, Dmitriev SE, Ivanov DS, Prassolov VS,
Shatsky IN. 2003. Distinctive properties of the 59-untranslated
region of human hsp70 mRNA. J Biol Chem 278: 22350–22356.
Shapira M, McEwen JG, Jaffe CL. 1988. Temperature effects on molec-
ular processes which lead to stage differentiation in Leishmania.
EMBO J 7: 2895–2901.
Siegel TN, Hekstra DR, Kemp LE, Figueiredo LM, Lowell JE, Fenyo D,
Wang X, Dewell S, Cross GA. 2009. Four histone variants mark the
boundaries of polycistronic transcription units in Trypanosoma
brucei. Genes & Dev 23: 1063–1076.
Singh R, Valcarcel J, Green MR. 1995. Distinct binding specificities
and functions of higher eukaryotic polypyrimidine tract-binding
proteins. Science 268: 1173–1176.
Song Y, Tzima E, Ochs K, Bassili G, Trusheim H, Linder M,
Preissner KT, Niepmann M. 2005. Evidence for an RNA chaperone
function of polypyrimidine tract-binding protein in picornavirus
translation. RNA 11: 1809–1824.
Spriggs KA, Bushell M, Mitchell SA, Willis AE. 2005. Internal
ribosome entry segment-mediated translation during apoptosis:
The role of IRES-trans-acting factors. Cell Death Differ 12: 585–
Stern MZ, Gupta SK, Salmon-Divon M, Haham T, Barda O, Levi S,
Wachtel C, Nilsen TW, Michaeli S. 2009. Multiple roles for
polypyrimidine tract binding (PTB) proteins in trypanosome
RNA metabolism. RNA 15: 648–665.
Thomas R. 1951. On the existence of labile bonds in secondary
structures of the nucleic acids’ molecule. Experientia 7: 261–262.
Thomas R. 1993. The denaturation of DNA. Gene 135: 77–79.
Yoffe Y, Leger M, Zinoviev A, Zuberek J, Darzynkiewicz E, Wagner G,
Shapira M. 2009. Evolutionary changes in the Leishmania eIF4F
complex involve variations in the eIF4E–eIF4G interactions.
Nucleic Acids Res 37: 3243–3253.
Yueh A, Schneider RJ. 2000. Translation by ribosome shunting on
adenovirus and hsp70 mRNAs facilitated by complementarity to
18S rRNA. Genes & Dev 14: 414–421.
Zilka A, Garlapati S, Dahan E, Yaolsky V, Shapira M. 2001. De-
velopmental regulation of heat shock protein 83 in Leishmania.
39 processing and mRNA stability control transcript abundance,
and translation is directed by a determinant in the 39-untranslated
region. J Biol Chem 276: 47922–47929.
Zuker M. 2003. Mfold web server for nucleic acid folding and
hybridization prediction. Nucleic Acids Res 31: 3406–3415.
Zuker M, Jacobson AB. 1998. Using reliability information to
annotate RNA secondary structures. RNA 4: 669–679.
David et al.
RNA, Vol. 16, No. 2