Investigation on Nuclear Transport of Trypanosoma brucei: An in silico Approach

1.1. Trypanosomiasis

A group of animal and human diseases caused by parasitic protozoan trypanosomes is called trypanosomiases. The final decade of the 20th century witnessed a frightening revival in sleeping sickness (human African trypanosomiasis) in sub-Saharan Africa. Meanwhile, Chagas' disease (American trypanosomiasis) remains one of the most widespread infectious diseases in South and Central America. Arthropod vectors are responsible for the spread of African and American trypanosomiases, and disease restraint through insect control programs is an attainable target. However, the existing drugs for both illnesses are far from ideal. The trypanosomes are some of the earliest diverging members of the Eukaryotae and share several biochemical oddities that have inspired research into discovery of new drug targets. Nevertheless, discrepancies in mode of interactions between trypanosome species and their hosts have spoiled efforts to design drugs effective against both species. Heightened awareness of these neglected diseases might result in progress towards control through increased financial support for drug development and vector eradication [1].

Trypanosome is a group of unicellular parasitic flagellate protozoa which mostly infects the vertebrate genera. A number of trypanosome species cause important veterinary diseases, but only two cause significant human diseases. In sub-Saharan Africa, Trypanosoma brucei causes sleeping sickness or human African trypanosomiasis whilst in America, Trypanosoma cruzi causes Chagas' disease (Figure 1) [2]. Meanwhile, the life cycle of these parasitic protozoa engage insect vectors and mammalian hosts (Figure 2) [1]. All trypanosomes require more than one obligatory host to complete their life cycle and are transmitted via vectors. Most of the species are transmitted by blood-feeding invertebrates, however there are distinct mechanisms among the varying species. In the invertebrate hosts they are generally found in the intestines as opposed to the bloodstream or any other intracellular environment in the mammalian host. As trypanosomes develop through their life cycle, they undergo a series of morphological changes [3] as is typical of trypanosomatids.

The life cycle often consists of the trypomastigote form in the vertebrate host and the trypomastigote or promastigote form in the gut of the invertebrate host. Intracellular lifecycle stages are normally found in the amastigote form. The trypomastigote morphology is unique to species in the genus Trypanosoma.

The genome organization of T. brucei is splitted into nuclear and mitochondrial genomes. The nuclear genome of T. brucei is made up of three classes of chromosomes according to their size on pulsed-field gel electrophoresis, large chromosomes (1 to 6 megabase pairs), intermediate chromosomes (200 to 500 kilobase pairs) and mini chromosomes (50 to 100 kilobase pairs) [4]. The large chromosomes contain most genes, while the small chromosomes tend to carry genes involved in antigenic variation, including the variant surface glycoprotein (VSG) genes. Meanwhile, the mitochondrial genome of the Trypanosoma, as well as of other kinetoplastids, known as the kinetoplast, is characterized by a highly complex series of catenated circles and minicircles and requires a cohort of proteins for organisation during cell division. The genome of T. brucei has been completely sequenced and is now available online [5].

1.3. Nuclear transport

Nuclear transport of proteins and ribonucleic acids (RNAs) between the nucleus and cytoplasm is a key mechanism in eukaryotic cells [6]. The transport between the nucleus and cytoplasm involves primarily three classes of macromolecules: substrates, adaptors, and receptors. The transport complex is formed when the substrates bind to an import or an export receptor. Some transport substrates require one or more adaptors to mediate formation of a transport complex. Once assembled, these transport complexes are transferred in one direction across the nuclear envelope via aqueous channels that are part of the nuclear pore complexes (NPCs). Following dissociation of the transport complex, both adaptors and receptors are recycled through the NPC to allow another round of transport to occur. Directionality of either import or export therefore depends on the formation of receptor-substrate complex on one side of the nuclear envelope and the dissociation of the complex on the other. The Ran GTPase is vital in producing this asymmetry. Modulation of nuclear transport generally involves specific inhibition of the formation of a transport complex, however, more global forms of regulation also occur [7]. The general concept of import and export process is shown in Figure 3 [8].

1.4. In silico approach

In silico study is defined as an analysis which is performed using computer or via computer simulation. It involves the strategy of managing, mining, integrating, and interpreting information from biological data at the genomic, metabalomic, proteomic, phylogenetic, cellular, or whole organism levels. The bioinformatics instruments and skills become crucial for in silico research as genome sequencing projects have resulted in an exponential growth in protein and nucleic acid sequence databases. Interaction among genes that gives rise to multiprotein functionality generates more data and complexity. In silico approach in medicine is not only reducing the need for expensive lab work and clinical trials but also is possible to speed the rate of drug discovery. In 2010, for example, researchers found potential inhibitors to an enzyme associated with cancer activity in silico using the protein docking algorithm EADock [9]. About 50 % of the molecules were later shown to be active inhibitors in vitro [9]. A unique advantage of the in silico approach is its worldwide accessibility. In some cases, having internet access or even just a computer is sufficient enough. Laboratory experiments either in vivo or in vitro both require more materials. In protein sequence analysis, in silico approach gives highly reproducible results in many cases or even exactly the same results because it only relies on comparison of the query sequence to a database of previously annotated sequences. However, in sophisticated analysis such as development of the 3-D structure of proteins from their primary sequences, discrepancies in results are to be expected due to the manual optimization which must consider several crucial steps such as template selection, target-template alignment, model construction and model evaluation.

1.5. Problem statements

Considering the importance of nuclear shuttling in many cellular processes, proteins responsible for the nuclear transport are vital for parasite survival. The presence of nuclear transport machinery was highlighted in the eukaryotic parasites such as Plasmodium falciparum, Toxoplasma gondii and Cryptosporidium parvum. However, the nuclear transport in T. brucei has not been established. Nuclear shuttling is one of the overlooked aspects of drug design and delivery. Exploitation of macromolecules movement across the nuclear envelope promises to be an exciting area of drug development. Furthermore, the divergence between host and parasite systems is always exploited as a strategy in drug development. Therefore, the exploitation of peculiarities of T. brucei nuclear transport machinery as compared to its host might be a promising strategy for the control of trypanosomiasis, which remains to be further investigated.

1.6. Objectives

This study is carried out to investigate the nuclear transport constituents of T. brucei by determining the functional characteristics of the parasite proteins. This includes functional protein domain, post translational modification sites and protein-protein interaction. The parasite proteins identified to exhibit the relevant functional protein domains, post translational modification sites and protein-protein interaction, are predicted as the true components for nuclear transport mechanism. This study also aims to evaluate the unique characteristics of proteins responsible for nuclear transport machinery between the parasites and human by determining the degree of protein sequence similarity. The information on the sequence level divergence between T. brucei proteins and their human counterparts may provide an insight into drug target discovery.

3.1. Parasite proteins involved in the nuclear transport machinery

Table 2 shows a summary of protein sequences used in this in silico analysis. A total of 904 and 642 protein sequences were retrieved in FASTA format from NCBI server and UniProt/SwissProt database respectively. A total of 18 protein sequences with less than 100 amino acid residues were excluded from the study as they were considered not completely functional [17]. Hence, 1528 protein sequences were used for protein sequence clustering. The 30% identity and above at the amino acid level is considered sufficient to imply functional relatedness [17]. Therefore, protein clustering with more than 30% similarity on the retrieved protein sequences produced a non-redundant data set of 248 protein sequences.

The BLASTp analyses against TriTrypDB using cut off point with E-value of less than 1e-06 and score of more than 100 for the whole 248 query protein sequences resulted in 34 hits of parasite proteins. However our approach failed to identify a Ran GTPase-activating protein (RanGAP) protein in this parasite. In reference [18] also reported that sequence similarity searches have been unable to identify a RanGAP protein in any protozoan. Keyword searches among annotated proteins in the T. gondii genome database identified one candidate which was shown to have strong similarity to Ran-binding protein 1 (RanBP1) based on sequence analysis. Perhaps the RanGAP function in apicomplexans is performed by a single protein with multiple cellular responsibilities (i.e., a fusion of Ran binding protein 1 and RanGAP). It is also possible that a completely unique parasite protein possesses the RanGAP function.

Table 3 shows the identified and characterized parasite proteins involved in the nuclear transport machinery. The functional annotation based on protein domains, showed that, out of 34, only 22 parasite protein sequences were predicted with high confidence level to be involved in the nuclear transport mechanism with the presence of relevant protein domains. This includes guanine triphosphate (GTP)-binding domain, Nucleoporin (NUP) C terminal domain, Armadillo repeat, Importin B N-terminal domain, regulator of chromosome condensation 1 (RCC1) repeat and Exportin domain (Table 4). All these protein domains were experimentally verified to regulate the nuclear transport mechanism in eukaryotes. There were seven T. brucei proteins that exhibited functional features of the Importin receptor. This finding is consensus with the number of Importin receptors in another eukaryotic pathogen, Toxoplasma gondii [8]. In addition, our results of other nuclear transport constituents in T. brucei such as RCC1, Ran, nuclear transport factor 2 (NTF2), cell apoptosis susceptibility (CAS), Exportin and Ran binding proteins were also in agreement with reference [18].

The nuclear and cytoplasmic compartments are divided by the nuclear envelope in eukaryotes. By using this compartmentalization and controlling the movement of molecules between the nucleus and the cytosol, cells are able to regulate numerous cellular mechanisms such as transcription and translation. Proteins with molecular size lower than 40 kDa are able to passively diffuse through the nuclear pore complex (NPC), whereas larger proteins require active transport through the assistance of Karyopherins, specific transport receptors that shuttle between the nucleus and cytosol. Karyopherins which are able to distinguish between the diverse proteome to target specific cargo molecules for transport, can be subdivided into those that transport molecules into the nucleus (Importins) and those that transport molecules out of the nucleus (Exportins). It has been reported that more than 2000 proteins are shuttled between the nucleus and the cytoplasm in yeast [19]. From our result, with the identification of Karyopherin and Nucleoporin proteins in T. brucei, we expect that the parasite employs the typical components for the nuclear transport machinery.

3.2. Regulatory aspect of the parasite nuclear transport

Table 5 shows the presence of phosphorylation and glycosylation sites in the parasite proteins. The phosphorylation sites were found to be present in all parasite proteins. It was predicted that the parasite proteins could be phosphorylated at Serine, Threonine and Tyrosine amino residues. However, the O-glycosylation sites were not present in three parasite proteins, namely Tb11.02.0870, Tb927.8.3370 and Tb927.7.5760.

Most of the parasite proteins were predicted to be involved in O-linked glycosylation. In eukaryotes, the O-linked glycosylation takes place in the Golgi apparatus. It also occurs in archaea and bacteria. Phosphorylation was reported to be crucial in the regulation of protein-protein interactions of the NADPH oxidase in the phagocytic cells [20]. The phosporylation-based signaling in T. brucei has been reported by reference [21]. Thus we believe that the phosphorylation could also regulate the nuclear transport components of T. brucei to participate in various functional interactions. Meanwhile, it was suggested that O-linked glycosylation may be analogous to protein phosphorylation. According to [22], phosphorylation by proline-directed kinases share the same sites with those potentially O-glycosylated by O-linked N-acetylglucosamine transferase (OGT). From this it is possible that O-glycosylation and phosphorylation may compete for sites of modification. Therefore, it is a strong likelihood that the nuclear transport of T. brucei could be regulated by both phosphorylation and O-glycosylation.

Apart from acting simply as an architectural structure which facilitates nuclear transport, the NPC may also play a more dynamic role in regulating transport. The specificity of import and export may be influenced by recognition of different substrates and alteration of the Nucleoporin expression. This would allow different interaction between the NPC and Karyopherins and modulate the nuclear import and export. However, the most common impact on nucleocytoplasmic movement stems comes from the post translational modifications of the cargo proteins themselves [23]. The post translational modification of NPC was reported by [24]. Post-translational modification of NUPs by ubiquitylation and phosphorylation can affect NUP turnover and pore disassembly, respectively. Our study identified four parasite proteins containing the Nucleoporin-related domain. We anticipate that the assembly and disassembly of the parasite Nucleoporin proteins might also be modulated by phosphorylation.

The NPC becomes an ideal target for inhibition of nuclear import or export. One of the most common features of Nucleoporins is the presence of conserved FG or FXFG repeats that bind to the Importin family members [25]. The monoclonoal antibodies such as mAb414 and RL2 can interrupt translocation through the NPC by blocking the FG and FXFG epitopes of the Nucleoporins. Consequently, several Nucleoporin proteins were identified by their reactivity against the anti-FG antibodies. Most of these FG repeat proteins exist as the cytoplasmic fibrils or projections on the nuclear side of the NPC. The monoclonal antibodies prevent cargo from associating with the edge of an NPC so it cannot cross the membrane [26]. Thus, there is a possibility that the pathogenesis of T. brucei could be controlled by inhibiting its Nucleoporin proteins.

3.3. Participation of parasite proteins in functional interaction network

Figure 5 illustrates the protein interaction data obtained from STRING 8.2 database. The mining of protein interaction data which is useful in contextual annotation of protein function showed that, out of 22 parasite homologues, only nine parasite proteins were interacting with each other. Out of the seven identified T. brucei Importins, only two namely Tb927.6.2640 and Tb10.70.4720 were found to be involved in that protein interaction network. This database mining approach indicated that T. brucei nuclear transport is typical of eukaryotic organisms. Importins initially recruit cargo at low RanGTP concentrations in the cytoplasm and release cargo at high RanGTP levels in the nucleus. Importin–RanGTP complexes return afterwards to the cytoplasm, where the Ran-bound GTP is finally hydrolysed and Ran dissociates from the receptor. The Importin can then bind and import another cargo molecule, while nuclear transport factor 2 (NTF2) recycles RanGDP back to nucleus. The cargo binding to exportins is controlled in a reverse manner compared to Importins; they recruit cargo at high RanGTP levels in the nucleus and release cargo at low RanGTP concentrations in the cytoplasm.

Table 5 shows evaluation of the obtained protein interaction data of the parasite nuclear transport. There were 13 functional interactions between parasite proteins identified from the mining of STRING 8.2 database. The score values of functional interactions range from 0.45 to 0.976. The Importin alpha (Tb927.6.2640) was found to be the most interactive parasite proteins by participating in six functional interactions. Based on the relevant protein domains and previous reports, four out of 13 functional interactions were considered with high confidence level. It should be emphasized that our approach only considered the protein interaction data derived from experiments, gene fusion and text mining. To our knowledge, this is the first report of functional protein interactions in the nuclear transport of the eukaryotic parasites. Whether other eukaryotic parasites share the common protein interaction network for the nuclear transport machinery remains to be elucidated.

To gain an insight into nuclear transport, understanding on interactions between transport receptors and proteins of the nuclear pore complex (NPC) is essential. According to [27], the fluorescence resonance energy transfer (FRET) can be employed between enhanced cyan and yellow fluorescent proteins (ECFP, EYFP) in living cells in order to explain the transport of receptor through the NPC. A FRET assay has been used to analyze a panel of yeast strains expressing functional receptor--ECFP and nucleoporin-EYFP fusions. Based on this approach, points of contact in the NPC for the related Importin Pse1/Kap121 and Exportin Msn5 were successfully characterized. That study proved the advantage of FRET in mapping dynamic protein interactions in a genetic system. In addition, both Importin and Exportin have overlapping pathways through the NPC. However, our database mining approach did not reveal any functional interaction between Nucleoporin and Karyopherin proteins of T. brucei.

3.4. Sequence similarity between parasite proteins and their human counterparts

Table 6 shows the degree of protein sequence similarity between parasite and human proteins. The similarity search for the sequence was carried out with the help of BLASTp tool. All the parasite proteins of nuclear transport machinery were found to have their counterparts in human. The degree of sequence similarity between parasite proteins and human counterparts range from 19% to 72%. The resulting score values range from 49.3 to 558. Meanwhile, all the identified human proteins contain the same protein domains involved in the nuclear transport.

A study reported by [28] showed that despite the high degree of similarity in the primary structure of human and T. cruzi ubiquitins, the three amino acid difference is sufficient to distinguish parasite versus host proteins. In this study, a simplified one step purification procedure to partially purify T. cruzi ubiquitin was performed. Following this preparation, ELISA and Western blots were carried out to show that chagasic sera recognise T. cruzi but not human or Leishmania ubiquitin indicating a species-specific response. Thus, it is probable that the T. brucei proteins could also be distinguished from human counterparts at primary sequence level by using the immunodetection method.

4.1. Transport of cargoes

In RanGTPase system, Ran-binding protein 1 (RanBP1) which is cytoplasmic localized binds RanGTP and eases the RanGAP-dependent conversion of RanGTP to RanGDP [29]. This indicates that RanBP1 catalyses the cytoplasmic disassembly of RanGTP–transport receptor complexes. These complexes are kinetically so stable that RanGAP alone fails to trigger GTP hydrolysis [30-32]. RanBP2 [33] is a major constituent of the cytoplasmic filaments of NPCs and exhibits similar functions as RanBP1. It has four RanBP1 homology domains and forms a stable complex with sumoylated RanGAP [34,35], in order to dismantle the RanGTP–transport receptor complexes that exit the nucleus. Importin- and exportin-mediated transport cycles can accumulate cargoes against gradients of chemical activity, which is an energy-dependent process. The RanGTPase system hydrolyses one GTP molecule per transport cycle, and a number of evidences suggest that this contains the sole input of metabolic energy [36-39]. We have successfully identified all the required key components in the T. brucei nuclear transport. Whether their functionalities in vivo are consensus with the known ones still remains to be further investigated.

4.2. Relationship between signaling pathways and nuclear transport

Many aspects of cell physiology are greatly dependent on the signaling pathways. This includes members of the mitogen activated protein (MAP) kinase family as well as phosphatidyl inositol 3 (PI3) and adenosine monophosphate (AMP) kinases which are crucial in controlling the cell growth, proliferation, apoptosis and the response to stress. By activating the signaling pathway through multiple kinase cascades, various stressors are able to regulate the nuclear transport. For example, oxidative and heat stress activate both MAP kinase kinase (MEK)-extracellular signal regulated kinase 1/2 (ERK1/2) and PI3 kinase-Akt pathways [40]. Based on these observations and the fact that many of the transport components are modified post-translationally, it was sensible to investigate whether these modifications are regulated by stress. A study reported by [41] showed that oxidant treatment induced phosphorylation and/or GlcNAc modification of soluble transport factors and nucleoporins. Interestingly, changes in transport factor modifications are not limited to stress conditions, as modifying ERK or PI3 kinase activities in unstressed cells also affect the transport factors. This is exemplified by the regulation of RanBP3 through ERK1/2-ribosomal S6 kinase (RSK) signaling, a regulatory link which ultimately controls the Ran concentration gradient. Furthermore, phosphorylation of Nup50 which is dependent on ERK, reduces its association with importin-β1 and transportin in vitro, and ERK2 is responsible to the oxidant-induced collapse of the Ran gradient [42]. It remains unknown how much modulating individual transport factors contributes to the overall regulation of nuclear trafficking. However, it is noteworthy that the kinase inhibitor PD98059, which targets ERK1/2 and ERK5, significantly increases classical nuclear import, both under normal and stress conditions. Taken together, these results highlight a critical role of ERK activity in nuclear transport, with ERK kinases targeting both soluble factors and nucleoporins [41]. Thus, there is an urgent need to investigate the possible connection between upstream signaling apparatus with nuclear transport components in T. brucei.

4.3. In silico approach for drug target discovery

We have provided interpretation of heterologous data sets for nuclear transport system of T. brucei from various resources. With the availability of protein databases and computer-aided softwares, we are able to explain various functional interactions between identified parasite proteins and how these functional interactions give rise to functionality and behavior of the parasite nuclear transport. This would partially facilitate the exhausted effort to obtain system-level understanding of T. brucei pathogenesis. Our in silico approach has the potential to speed up the rate of drug target discovery while reducing the need for expensive lab work and clinical trials. The conventional approaches in vivo and in vitro have high tendencies to produce inefficient results when investigating complex large scale data such as proteins associated with nuclear shuttling of macromolecules across the nuclear envelope. Therefore, the systematic in silico approach from this study provides a tremendous opportunity of cost effective drug target discovery for the pharmaceutical industry.

4.4. Experimental validation of in silico data

Experimental techniques such as yeast two-hybrid assay and affinity purification combined with mass spectrometry are useful to investigate the possible protein-protein interaction. However, they have their limitations in detecting certain types of interactions. They also have technical problems to scale-up for high-throughput analysis. In conjunction with this, in silico approach may solve those problems in inferring the protein function. The scope of experimental data can be expanded to increase the confidence of certain interacting protein pairs with the availability of databases containing in silico data such as protein domain and 3D structure. The databases integrate information from various resources such as computational prediction methods and public text collections. Since in silico and experimental approaches are complementary to each other, the combination of these different approaches is very useful to obtain a more accurate picture of T. brucei nuclear transport.

4.5. Our further direction

In silico approach offers various advantages over in vivo and in vitro approaches such as non-use of animals, low costs, and reduced execution time. This approach allows identification of proteins of interest from a particular biological study. From a protein function standpoint, transfer of annotation from known proteins to a novel target is currently the only practical way to convert vast quantities of raw sequence data into meaningful information. Many bioinformatics tools now provide more sophisticated methods to transfer functional annotation, integrating sequence, family profile and structural search methodology. Thus, in addition to data mining for protein-protein interaction, further in silico approach should also consider structural alignment, molecular docking and pathway modeling in order to obtain a comprehensive and more reliable insight into protein-protein interaction of T. brucei nuclear transport.