Identification of a Novel Nucleocytoplasmic Shuttling RNA Helicase of Trypanosomes

Gene expression in trypanosomes is controlled mostly by post-transcriptional pathways. Little is known about the components of mRNA nucleocytoplasmic export routes in these parasites. Comparative genomics has shown that the mRNA transport pathway is the least conserved pathway among eukaryotes. Nonetheless, we identified a RNA helicase (Hel45) that is conserved across eukaryotes and similar to shuttling proteins involved in mRNA export. We used in silico analysis to predict the structure of Trypanosoma cruzi Hel45, including the N-terminal domain and the C-terminal domain, and our findings suggest that this RNA helicase can form complexes with mRNA. Hel45 was present in both nucleus and cytoplasm. Electron microscopy showed that Hel45 is clustered close to the cytoplasmic side of nuclear pore complexes, and is also present in the nucleus where it is associated with peripheral compact chromatin. Deletion of a predicted Nuclear Export Signal motif led to the accumulation of Hel45ΔNES in the nucleus, indicating that Hel45 shuttles between the nucleus and the cytoplasm. This transport was dependent on active transcription but did not depend on the exportin Crm1. Knockdown of Mex67 in T. brucei caused the nuclear accumulation of the T. brucei ortholog of Hel45. Indeed, Hel45 is present in mRNA ribonucleoprotein complexes that are not associated with polysomes. It is still necessary to confirm the precise function of Hel45. However, this RNA helicase is associated with mRNA metabolism and its nucleocytoplasmic shuttling is dependent on an mRNA export route involving Mex67 receptor.


Introduction
Chagas disease is a neglected disease endemic to Latin America, where about eight million people are affected [1]. This disease is caused by infection with Trypanosoma cruzi (T. cruzi). In addition to their medical importance, trypanosomatids are interesting experimental models because the regulation of gene expression in these organisms have some unusual features. Several genes are grouped together under the control of a single promoter region [2,3] and give rise to long polycistronic transcripts. These transcripts are processed by trans-splicing and polyadenylation to form monocistronic messenger RNA (mRNA) [4,5,6]. The resulting mature mRNAs are then transported from the nucleus to the cytoplasm, where protein synthesis occurs. Gene expression is controlled mostly by post-transcriptional events and the mechanisms controlling mRNA processing [4,5,6] and stability [7,8,9] are becoming increasingly understood. However, little is known about the mechanisms of mRNA nucleocytoplasmic transport in these parasites, and the identity of factors that determine the fate of mRNA in the cytoplasm remains to be unveiled.
In yeast and mammalian cells, the bidirectional translocation of macromolecules between the nucleus and cytoplasm (e.g. RNAs to the cytoplasm and transcription factors to the nucleus) involves the nuclear pore complex (NPC), which is composed largely of nucleoporins [10]. The NPC mediates the transport of molecules by interacting transiently with proteins from the b-karyopherin family [11], which are conserved nuclear receptors known as importins and exportins [12]. Crm1 is the major exportin in many organisms [13]. It recognizes a nuclear export sequence (NES) motif in shuttling proteins and its activity is dependent on RanGTP [14,15]. Only some mRNAs are transported by Crm1 [16,17,18,19]. Instead, most mRNAs are exported by the Mex67 receptor [19], which does not belong to the karyopherin family and functions in a RanGTP-independent manner. During transcription, RNA-binding proteins of the THO complex associate with the nascent mRNA and initiate the formation of the ribonucleoprotein complex (mRNP). This involves the recruitment of processing and export factors, such as Sub2 and Yra1, resulting in the formation of the transcription/export (TREX) complex. The TREX complex interacts with the spliceosome and processed mRNAs are exported through the NPC by the Mex67 receptor and are disassembled in the cytoplasm by the ATP-dependent helicase Dbp5 (DDX19 in humans) [20,21]. Comparative genomic analyses have demonstrated that the mRNA export pathway is poorly conserved in many parasites, including T. cruzi [22], suggesting distinct mechanisms in mammals and parasites.
In this study, we investigated the function of an ATP-dependent DEAD/H RNA helicase (Hel45) in T. cruzi. We show that Hel45 shuttles between the nucleus and cytoplasm and is located near to NPCs. We demonstrate that the export of Hel45 is dependent on transcription and we show that it forms ribonucleoprotein complexes not associated with polysomes in the cytoplasm. We also show that Hel45 has a nuclear export signal and its shuttling is dependent on an mRNA export pathway, involving a homolog of the Mex67 nuclear receptor. Our findings also suggest that this protein is involved in mRNA metabolism and its nucleocytoplasmic shuttling is dependent on an mRNA export route involving Mex67 receptor.
Structural homology-based molecular modeling of Hel45 (GeneID 3541696) was carried out by protein searches with the BLASTP of the protein data bank (PDB) database [27]. Alignments of proteins, based on primary and secondary structures, with low levels of sequence identity were generated with the GenTHREADER program [28]. A model was constructed with MODELLER 9v11 [29]. Figures of the structural model were generated with PyMOL software (available at http://www. pymol.org).

Polyclonal antibody production
The Hel45 open reading frame (ORF) was amplified by PCR with the oligonucleotide primers Hel45F and Hel45R (Table 1). T. cruzi Dm28c was used as the DNA template. The PCR product was cloned into the pDONR TM 221 vector from Gateway technology (Invitrogen) and was then recombined into the pDEST TM 17 vector (Invitrogen) to produce a his-tagged Hel45 recombinant, according to the manufacturer's protocol. Production of recombinant protein was induced in Escherichia coli BL21 (DE3) by addition of 1 mM IPTG and incubation for 3 h at 37uC. His-tagged Hel45 protein was purified by affinity chromatography on Ni-NTA resin (Qiagen) under denaturing conditions, and was used to inoculate mice to produce polyclonal antibodies (according to De Souza et al. (2010) [33]).
Antibodies were incubated with the membrane for 1 hour. The membrane was then washed three times in 0.1% Tween 20 in PBS. Bound antibodies were detected by the alkaline phosphatase [34] or peroxidase [33] method.
The nuclear and cytoplasmic extracts for cellular fractionation analysis were obtained by hypotonic lysis of epimastigote forms, as described by Roberts et al. (1998) [35].

Light microscopy
The modified pTcGW vector [36] was used to tag the NT with PTP [37]. The oligonucleotides used to clone the Hel45 ORF are shown in Table 1. The nuclear export signal was deleted by fusion PCR. For this, two fragments of Hel45 were amplified by PCR with the Hel45F/NESR and NESF/Hel45R oligonucleotides (Table 1) and these two amplicons were mixed prior to another round of PCR. The fragment obtained was sequenced and inserted into the pTcGW vector to create a Hel45DNES mutant tagged at its N-terminal end with PTP. epimastigotes were transfected with these plasmids, et al. (1991) [38]. Stable lines were selected by adding 500 mg/ml G418 to the culture medium. The endogenous and PTP-tagged Hel45 proteins were localized by indirect immunofluorescence assays, as described by   [39]. Mouse anti-Hel45 polyclonal antibodies (1:100 dilution) or rabbit anti-protein A (ProtA) antibodies (1:40,000 dilution) were incubated with the parasites for 1 hour at 37uC. The parasites were then washed with PBS and incubated with Alexa Fluor 488-conjugated goat anti-mouse IgG, Alexa Fluor 633conjugated rabbit anti-mouse IgG or Alexa Fluor 594-conjugated goat anti-rabbit IgG antibodies (Invitrogen, 1:600 dilution), as

Ultrastructural microscopy
Ultrastructural immunocytochemistry of Hel45 in T. cruzi epimastigote forms were performed as described by Motta et al. (2003) [40]. Samples were blocked for 30 minutes with 3% BSA, 0.5% teleostean gelatin, and 0.02% Tween 20 in PBS pH 8.0, and were then incubated with anti-Hel45 antiserum (1:50 dilution) for 1 hour. The parasites, on grids, were treated for 30 minutes with blocking solution and were incubated with 10 nm gold-labeled anti-mouse IgG (Sigma-Aldrich) diluted 1:250 in blocking solution. The grids were washed in blocking solution, stained with uranyl acetate and lead citrate, and were observed with a Zeiss EM-900 transmission electron microscope.

Mex67 RNAi interference
The ortholog of Mex67 in T. brucei was named TbMex67 (GeneID 3664369). For gene knockdown by RNA interference, the region corresponding to 362-845 of the nucleotide sequence was chosen with the RNAit program [41]. A DNA fragment was amplified with the oligonucleotides Mex67RNAiF (forward) and Mex67RNAiR (reverse) (Table1) for cloning into the p2T7-117 vector [42]. A total of 10 mg insert-containing vector was linearized with NotI enzyme and was transfected into procyclic forms of T. brucei 29-13 cell line [32].
Transfected parasites were selected by the addition of 5 mg/ml phleomycin to the medium. RNAi was induced by adding 2 mg/ ml tetracycline to log phase parasites, and the knockdown confirmed by western blotting with polyclonal antisera anti-Mex67. Anti-GAPDH was used as a loading control.

Fluorescence in situ hybridization (FISH) for the detection of mRNA
For the detection of poly(A) + RNA in T. brucei and T. cruzi, the parasites were harvested, washed in PBS (pH 7.4), fixed by incubation in 4% paraformaldehyde and were allowed to adhere to poly-L-lysine-coated slides for 10 minutes. The slides were washed in PBS and the parasites were permeabilized by incubation with 0.2 M HCl (diluted in PBS) for 10 minutes. For T. cruzi, the cells were incubated with prehybridization buffer (35% formamide, 0.02% BSA in 2X SSC buffer) supplemented with 25 mg/ml tRNA, 1 mg/ml salmon sperm DNA (Sigma-Aldrich) and 40 U/ml RNaseOUT (Invitrogen) for 30 minutes at 37uC. For T. brucei, the cells were incubated with prehybridization buffer containing 10X Denhardt's solution, 1 mM EDTA, 35% formamide in 4X SSC and supplemented with 0.5 mg/ml tRNA and 2 mU/ml RNaseOUT for 30 min at room temperature. Digoxigenin-conjugated oligo(dT) probes (6 ng/ml) were diluted in prehybridization buffer and denatured by heating at 65uC for 3 minutes. Hybridization was performed for 16 hours at 37uC. Probe binding was detected by indirect immunofluorescence analysis with mouse monoclonal anti-digoxigenin antibody (Sigma-Aldrich, 1:300 dilution) and Alexa Fluor 488-conjugated goat anti-mouse IgG secondary antibody (Invitrogen, 1:600 dilution), as described previously. As a control, 100 mg/ml RNase A was added to the pretreatment buffer before probe hybridization.

Polysome sedimentation profiles
Polysome sedimentation profiles were obtained by the ultracentrifugation of cytoplasmic extracts (1610 9 cells) of epimastigotes on sucrose density gradients. The cells were treated with 100 mg/ml cycloheximide for 10 minutes at 28uC and were harvested by centrifugation at 5,0006g. Parasites were washed in cold TKMC buffer (10 mM Tris-HCl pH 7.4; 10 mM MgCl 2 ; 300 mM KCl) supplemented with 100 mg/ml cycloheximide. Cell pellet was resuspended in 900 ml TKMC buffer supplemented with 100 mg/ml cycloheximide, 10 mg/ml heparin, 10 mM E-64, 10 mM PMSF and was transferred to a new tube containing 100 ml TKMC buffer to which 10% (v/v) NP-40 and 2 M sucrose were added. The lysate was centrifuged at 16,0006g at 4uC for 5 minutes and 500 ml of cleared supernatant was centrifuged on a linear sucrose density gradient from 15 to 55% [43]. For micrococcal nuclease treatment, the supernatant was incubated with 500 U/ml micrococcal nuclease in the presence of 2 mM CaCl 2 for 30 min, and the reaction was blocked by adding 2.5 mM EGTA.
Parasites were treated with 2 mM puromycin for 1 hour at 28uC and were then washed in cold TKMP buffer (10 mM Tris-HCl pH 7.4, 2 mM MgCl 2 , 500 mM KCl). Cells were centrifuged at 5,0006g and were resuspended in 900 ml TKMP buffer supplemented with 10 mg/ml heparin, 10 mM E-64, 10 mM PMSF and 1 mM puromycin. Parasites were lysed with 100 ml TKMP buffer supplemented with 10% (v/v) NP-40 and 2 M sucrose. The cell lysate was centrifuged and 500 ml of the clear supernatant was layered onto a sucrose density gradient (15-55%) prepared with TKMP buffer. All gradients were centrifuged at 192,0006g for 2 hours at 4uC and fractions were separated with an ISCO gradient fractionation system. Aliquots of 30 ml of each fraction were collected for western blotting with mouse anti-Hel45 and mouse anti-S7 antibodies.

mRNP isolation
The cytoplasmic fraction of epimastigotes was obtained by hypotonic lysis, as described by Roberts et al. (1998) [35]. Poly(A) + molecules were isolated from the cytoplasmic extract with 1 mg magnetic oligo(dT)-conjugated beads from the PolyATract mRNA Isolation System IV (Promega) according to the manufacturer's protocol. Cytoplasmic extracts were incubated with the beads for 16 hours at 4uC. The beads were washed three times with hypotonic buffer containing 5 mM 2-mercaptoethanol and 1% NP-40, and bound particles were eluted with 0.2% SDS with heating at 95uC for 5 minutes. As a control, 10 mg/ml RNase A was added to the protein extract before its incubation with beads.

Results
Comparative analysis and prediction of the structure of the ATP-dependent DEAD/H RNA helicase (Hel45) Previous analysis demonstrated the existence of a T. cruzi protein sequence that we named Hel45, conserved across different eukaryotic supergroups examined, including several species of Excavata and Chromalveolata [22]. It is a protein with a predicted molecular weight of 44.9 kDa that belongs to the ATP-dependent DEAD/H RNA helicase family. Comparative analyses by multiple sequence alignment demonstrated that the nine characteristic motifs (Q, I, Ia, Ib, II, III, IV, V, VI) of the DEAD-box helicase protein family [44] were conserved in Hel45 ( Figures 1A and 1B). Besides, we observed that Hel45 is similar to shuttling proteins involved in mRNA export routes, as DBP5/DDX19 [45] and the eukaryotic initiation factor 4AIII (eiF4AIII) [46,47]. This comparative analysis showed that Hel45 is more similar to eiF4A group than to DBP5/DDX19 group of RNA helicases. It had 80.7%, 76.8% and 76.0% of similarity to the eiF4AIII in humans, yeast, and P. falciparum, respectively (Table S1). Whereas the similarity with DBP5/DDX19 in humans, yeast and P. falciparum was 60.8%, 60.3%, and 52.7%, respectively (Table S1). Hel45 is highly conserved in other trypanosomatids, showing 92.7% and 82.1% of identity, and 97.4% and 91.8% of similarity with T. brucei and L. major, respectively (Table S1).
We used molecular modeling to predict the structure of Hel45 ( Figure 1C) based on structural similarity with related proteins (yeast and human eIF4A; accession numbers in the protein database: 2VSO and 2ZU6, respectively, and human eIF4AIII; accession numbers in the protein database: 2HYI, 2J0S, and 2HXY). We sought to assess the potential role of the protein as an RNA helicase based on this structure. The predicted structure of the presumptive RNA helicase comprised two functional domains ( Figure 1C): the N-terminal (NT) domain and the C-terminal (CT) domain. The model of this protein suggested a dynamic spatial conformation of the NT and CT domains, due to a deep cleft between these domains. The two domains were linked by a flexible loop, which is characteristic of proteins with RNA helicase activity.
Hel45 is present in ribonucleoprotein complexes not associated with polysomes in the cytoplasm Comparative analysis suggests that Hel45 is very similar to the members of human eukaryotic initiation factor 4A group (eIF4A). We therefore carried out polysome fractionation to assess the potential role of Hel45 in translation. We found that low-density polysome-independent fractions were enriched in Hel45 (Figure 2A). Remarkably, the sedimentation profile of Hel45 on the sucrose density gradient was not modified by treatment with puromycin ( Figure 2B), in contrast with that of the ribosomal protein S7 (Figures 2A and 2B). The sedimentation profile of Hel45 was altered only by treatment of the cytoplasmic extract with micrococcal nuclease ( Figure 2C). These results suggest that the sedimentation profile of Hel45 is dependent on RNA integrity. Moreover, mRNP precipitation with oligo(dT)-conjugated beads indicated that Hel45 was a component of ribonucleoprotein complexes in the cytoplasm ( Figure 2D). These data demonstrate that Hel45 forms mRNPs, but is not associated with polysomes.

Hel45 is found in both the nucleus and the cytoplasm and clusters around NPCs
Western-blot analysis of cellular fractions of T. cruzi (Figure 3A.1) showed that about 70% of Hel45 is present in the cytoplasmic fraction and 30% is present in the nuclear fraction ( Figure 3A.2). Indirect immunofluorescence assays showed that Hel45 was dispersed throughout the cytoplasm, but showed enrichment around the nucleus ( Figure 3B). Ultrastructural immunocytochemical analysis of T. cruzi epimastigotes confirmed that Hel45 was present in both the nuclear and cytoplasmic compartments ( Figures 3C.1 and 3C.2). In the nucleus, gold particles were present in the periphery of electron-dense chromatin regions and in the periphery of the nucleolus (Figure 3C.1). The protein was dispersed throughout the cytoplasm (not shown) and also accumulated close to NPCs (Figures 3C.1 and 3C.2).

A predicted nuclear export signal (NES) is essential for Hel45 shuttling between the nucleus and cytoplasm
The presence of Hel45 in the nucleus and cytoplasm suggests that it acts as a shuttling protein (Figure 3). We therefore searched for nuclear export signals (NES) with the NESsential program, which predicts NES sites on the basis of protein sequence, regional disorder and solvent accessibility criteria [30]. This program recognized a classic nuclear export signal at position 255-261 of the CT region, which consisted of the sequence LYDTLTI. The probability that this putative NES was functional was 63% based on sequence alone. The predicted NES motif is highlighted in Figures 1B and 1C (yellow region). The NES region is located at the end of a helix in the predicted structure ( Figure 1C). Structural analysis showed that the side chains of amino-acids D257 and T258 located within the NES motif makes hydrogen-bonds with D393 in the CT region, which keeps the NES motif close to the ahelix ( Figure 1C, detailed inset). We addressed the role of this signal by performing fusion PCR to delete the sequence corresponding to the predicted NES ( Figure S1). This NES deletion significantly modified the distribution of Hel45, and caused its accumulation in the nucleus ( Figure 4A). Western blot analyses confirmed the ectopic expression of tagged Hel45 (Figures 4B and 4C). We also investigated the presence of a conserved nuclear localization signal (NLS) with the PredictNLS program, but no such motif was identified in Hel45 (data not shown).

The inhibition of transcription blocks Hel45 export to the cytoplasm
We found that Hel45 is present in mRNP complexes ( Figure 2); therefore, we examined whether blocking the transcription affects Hel45 nucleocytoplasmic export. The treatment of T. cruzi with actinomycin D (ACTD) resulted in the accumulation of Hel45 in  Figure 5A). For control of actinomycin D activity, it was observed the accumulation of mRNA in the nucleolus ( Figure 5B), as previously described [48]. This suggests that Hel45 transport to the cytoplasm is dependent on active transcription.
Hel45 shuttling is dependent on Mex67 but not on the Crm1 protein Crm1 is the major exportin in many organisms that recognizes nuclear export sequence (NES) motifs in shuttling proteins [14,15]. We first investigated if the exportin Crm1 is involved in Hel45 shuttling by treating epimastigotes with leptomycin B (LMB), which specifically inhibits Crm1 activity [49,50]. LMB did not alter the distribution of Hel45 ( Figure 6A) even after treatment with lethal concentrations of the drug, or for long incubation times that affected the growth rate of parasite ( Figure 6B). This indicates that Hel45 is not exported in a Crm1-dependent manner.
Mex67 is a nuclear mRNA export receptor in Trypanosoma brucei [51,52]. We hypothesized that shuttling of Hel45 occurs by an mRNA export pathway involving the receptor Mex67, because the export of Hel45 depends on transcription ( Figure 5). We tested this hypothesis by knocking down the expression of Mex67 with an inducible system of RNAi in T. brucei, because T. cruzi does not have a functional RNA interference machinery [53] and lacks an inducible system for gene silencing. The protein orthologous to Hel45 in T. brucei showed a predicted molecular weight of 45.51 kDa and we named this protein TbHel46. The amino acid sequence of Hel45 is 92.7% identical and 97.4% similar to that of TbHel46 (Table S1). We used an inducible RNAi system to knockdown the expression of TbMex67, as described previously [52]. The induction successfully prevented the expression of TbMex67 protein ( Figure 6D) and impaired the growth of T. brucei ( Figure 6C). TbMex67 RNAi induction also caused the accumulation of both polyadenylated mRNA ( Figures 6E.1 and S2) and TbHel46 (Figures 6E.1 and S3) in the nucleus after 48 hours, which we quantified from the fluorescence intensity of labelling ( Figures 6E.2 and 6E.3). This demonstrates that shuttling of TbHel46 depends on the Mex67-mRNA export pathway in trypanosomes.

Discussion
The mechanisms of molecular exchange between the nucleus and cytoplasm are well characterized in mammals and yeast. However, the proteins and mechanisms involved in the mRNA nucleocytoplasmic transport in several species of parasites are poorly understood. Trypanosomes branched off early from the metazoan lineage, which may account for the conservation of only a few proteins of the mRNA export network in these highly divergent organisms [22]. Regarding to comparative analyses of Ib for RNA-binding, and III for ATP hydrolysis [44]. The C-terminal domain (CTD) contains the motifs IV and V for RNA-binding, and VI for ATPase and unwinding activities [44]. The predicted nuclear export signal (NES) in the LYDTLTI sequence (255-261 position) is shown in yellow. (C) Molecular modeling of Hel45. The nine motifs are highlighted in orange, the predicted NES (yellow) is close to the CT extremity (green). A zoom of this region (box) shows the side chains of amino-acids D257, T258 and D393, and the interactions that maintain the structure at its C-terminal extremity. The organization of the NES in the CT is shown in the inset (upper right corner). doi:10.1371/journal.pone.0109521.g001   were collected after the sedimentation of cytoplasmic extract from T. cruzi treated with 100 mg/ml cycloheximide (A), 2 mM puromycin (B), or 500 U/ml micrococcal nuclease in the presence of 2 mM CaCl 2 (C). The 40S and 60S ribosomal subunits, the 80S ribosome monomer and polysomes are indicated. A western blot was performed with an anti-Hel45 antibody for each fraction. S7, a small ribosomal subunit protein, was used as a control. (D) mRNP isolation assay. Western-blot analysis with anti-Hel45 and anti-S7 antibodies and mRNPs obtained from the T. cruzi cytoplasmic fraction after elution from oligo(dT)-conjugated magnetic beads (El). As a control, cytoplasmic extract was treated with 10 mg/ml RNaseA before mRNP capture. FT = flow-through from cytoplasmic extract not bound to the oligo(dT). El = eluted fraction. doi:10.1371/journal.pone.0109521.g002 proteins involved in RNA export, our previous work [22] identified a RNA DEAD-box helicase protein named Hel45. We have considered that Hel45 is conserved across eukaryotes and has similarity to shuttling RNA helicases from mammalian and yeast, like DBP5/DDX19 [45] and eiF4AIII [46,47]. Then, we decided to investigate the role of Hel45 in mRNA nucleocytoplasmic transport in Trypanosoma cruzi.  RNA helicases of the DEAD/H-box family are characterized by the presence of nine conserved motifs that are incorporated into two RecA-like domains. These helicases are involved in several biological steps in RNA metabolism, from transcription to translation [44]. We demonstrate that these nine motifs are conserved in Hel45 and are clustered into the two domains typical of DEAD-box helicases (Figures 1B and 1C). These findings suggest that Hel45 belongs to the RNA helicase family. Comparative analyses of protein sequences from members of eIF4A family and other RNA helicases involved in mRNA metabolism have shown that Hel45 is more similar to eiF4AIII ( Figure 1A, Table S1). In metazoan, eiF4AIII is a nuclear protein [54] that associates to mRNA during splicing at region containing the Exon Junction Complex (EJC) and shuttles to the cytoplasm probably to function during Non-sense Mediated Decay (NMD) pathway [46]. However, in trypanosomes no orthologue of eIF4AIII was identified yet as component of EJC core and the role of EJC in trans-splicing remain unclear [55]. Furthermore, it is not clear that a classical NMD exist in trypanosomes [56]. Further investigation is necessary to identify proteins associated to Hel45 before speculating a functional correlation with metazoan eiF4AIII.
Hel45 is present in both cytoplasm and nucleus (Figure 3), and deletion of the predicted NES motif in the CT region resulted in the accumulation of the protein in the nucleus ( Figure 4A). These observations confirm that the NES-containing Hel45 is a shuttle protein. The transport of NES-containing cargo is usually mediated by exportins, which interact with domains of inner nucleoporins [57] to mediate transient docking at the NPC [21]. Comparative genomic analysis has shown that Crm1 is the most conserved exportin in diverse organisms [22]. Furthermore, the T. cruzi Crm1 contains the CRIME domain, which interacts with RanGTP, and the CCR domain, which is a target of leptomycin B [58]. Surprisingly, we demonstrated that Hel45 nuclear export was not blocked by sustained treatment with leptomycin B ( Figure 6A). This indicates that Crm1 is not the receptor involved in transporting Hel45 through the NPC, suggesting that NES is not recognized by Crm1. The NES is essential for Hel45 shuttling, but it may not be the only signal. Indeed, deletion did not result in the complete retention of Hel45 in the nucleus, and some was still present in the cytoplasm ( Figure 4A). RNA-binding motifs have also been reported to mediate the nuclear transport of proteins in T. cruzi [59]. Therefore we cannot rule out the possibility that the RNA-binding motifs of Hel45 (Figure 1) are also important for transport. Interestingly, only a small number of shuttle proteins have been identified in trypanosomatids and only two RNAbinding proteins identified thus far have NES motifs [7,60].
Our data indicate that Hel45 is localized at the periphery of dense chromatin domains ( Figure 3C.1) called interchromatin granule clusters (IGCs) [39], thought to correspond to regions of active transcription and splicing. Many studies have shown that nascent mRNAs move into these interchromatin spaces [61,62]. It is likely that an interaction with mRNA is also essential for Hel45 shuttling because Hel45 accumulated in the nucleus when transcription was blocked by actinomycin D ( Figure 5). Therefore, Hel45 appears to interact with mRNA during transcription and is transported through the NPC by a specific receptor.
Exportin activity is dependent on RanGTP and, apart from Crm1, other exportins are not conserved in T. cruzi [22]. Treatment with leptomicyn B did not block the shuttling of Hel45; therefore, Hel45 may be exported by a RanGTP-independent pathway. Mex67 functions as the receptor for a RanGTPindependent mRNA export pathway in several eukaryotic species. This receptor also mediates mRNA export in T. brucei, because knockdown of Mex67 expression leads to accumulation of mRNA in the nucleus [51,52]. We found that knockdown of Mex67 in T. brucei caused an accumulation of the T. brucei ortholog of Hel45 in the nucleus ( Figure 6E.1). This suggests that Hel45 shuttling is dependent on the Mex67 pathway in trypanosomes.
Even if the molecular modeling has shown structural similarity with members of eIF4A family, Hel45 does not seem to function as a translational factor, because the inhibition of translation did not  change the sedimentation profile of Hel45 on sucrose density gradients ( Figure 2B). Instead, the sedimentation profile of Hel45 was affected by the treatment of the protein extract with a nuclease ( Figure 2C). In addition, Hel45 was present in mRNPs not associated with polysomes ( Figure 2D). These results are consistent with previous findings that identified Hel45 as a component of polysome-independent mRNP complexes [43]. Based on these findings, we suggest that Hel45 also interacts with mRNA in the cytoplasm and it is not associated with polysomes. Besides, members of eIF4A family have been shown to possess rather diverse roles in the mRNA lifecycle, although they are highly similar. Their specific and diverse functions are often regulated and dictated by interacting partner proteins [63].
Taking together, our findings suggest that the nucleocytoplasmic shuttling of Hel45 is dependent on a Mex67 mRNA export pathway. However, additional studies are required to assess the precise function of Hel45 in mRNA metabolism. Previous work, indicate that components of the mRNA export pathway in parasites, such as Mex67 [51,52], present distinct features. This means that the function of specific components needs to be dissected within the context of these particular organisms. Lastly, most factors that play a role in post-transcriptional regulation in parasites are cytoplasmic proteins; therefore, we believe that the identification of nucleocytoplasmic shuttling proteins will improve the knowledge of the factors involved in post-transcriptional regulation of gene expression in parasites. Figure S1 Deletion of the predicted NES of the Hel45 gene. Alignment of the Hel45 gene and Hel45DNES sequences, obtained with Clustal W2 software. Hel45DNES was sequenced and the deletion of NES was confirmed. (N) Nucleotide sequence. (AA) Deduced amino-acid sequence translated from the nucleotide sequence. Asterisks (*) indicate consensus nucleotide sequence. NES = nuclear export sequence. (TIF) Figure S2 Localization of polyadenylated mRNA after induction of RNAi against Mex67 in T. brucei. Cellular localization of mRNA with a digoxigenin-conjugated oligo(dT) probe, by fluorescence in situ hybridization (FISH). The probe was detected by indirect immunofluorescence assays with a mouse anti-digoxigenin monoclonal antibody (Sigma-Aldrich, 1:300 dilution) followed by a secondary Alexa Fluor 488-conjugated antibody. As a control, 100 mg/ml RNase A was incubated with the parasites before probe hybridization (RNase A). DAPI = DNA stained with DAPI. MERGE = merged images for DAPI staining and FISH. N = nucleus. K = kinetoplast. Bar = 5 mm.