Characterization of potential drug targeting folate transporter proteins from Eukaryotic Pathogens

Introduction

A heterogeneous diversity of eukaryotic pathogens is responsible for the most economically important diseases of humans and animals^1,2. As a result of underdevelopment, a lack of social infrastructure and insufficient funding of public health facilities, most of these pathogens are endemic to resource-poor countries in sub-Saharan Africa, South-East Asia and South America, where they are responsible for high morbidity and mortality^1–3. Of these, parasitic protozoa form a major group, with the apicomplexans and kinetoplastid parasites represented by important members, which cause diseases such as malaria, cryptosporidiosis, toxoplasmosis, babesiosis, leishmaniasis, Human African trypanosomiasis and south American trypanosomiasis or Chagas’ disease causing most of the morbidity and mortality^4,5. Other important diseases caused by protozoans include giardiasis, amoebic dysentery^6,7 and trichomoniasis⁸. A vicious cycle of poverty and disease exists for most of these parasites with a high infection and death rate in affected populations^9–11. The appreciable burden of disease caused by these parasites has been aggravated by the lack of a licensed vaccine for most of them¹². Furthermore, current drugs of choice for treatment for many of the parasites have significant side effects, with the added emergence of drug resistant strains^13–15. Despite the urgent demand for new therapies for control, few drugs have been developed to combat these parasites¹⁶. A major limitation to the development of new drugs is the paucity of new drug targets. There is therefore a need for discovery of novel and alternate potential chemotherapeutic targets that can help in drug development efforts for disease control^16–18. A possible approach to selective antimicrobial chemotherapy has been to exploit the inhibition of unique targets, vital to the pathogen and absent in mammals^17,18.

A metabolic pathway that has been exploited considerably for the development of drugs is the folate biosynthetic pathway¹⁹. Antifolate drugs target this pathway and are the most important and successful antimicrobial chemotherapies targeting a range of bacterial and eukaryotic pathogens. While most parasitic protozoa can synthesize folates from simple precursors, such as GTP, p-aminobenzoic acid (pABA) and glutamate, higher animals and humans cannot²⁰. Additionally, a few of these parasites can also salvage folate as nutrient from their host²¹. These folate compounds are important for synthesis of DNA, RNA and membrane lipids and are transported via receptor-mediated or/and carrier-mediated transmembrane proteins; folate transporters^20–22. Importantly, antifolate chemotherapies that target the biosynthesis and processing of folate cofactors have been effective in the chemotherapy of bacterial and protozoan parasites²¹. More importantly, the folate pathway has also been confirmed as being essential in some eukaryotic pathogens such as Plasmodium, trypanosomes and Leishmania¹⁹.

In addition to the folate biosynthesis pathway, proteins that mediate transport of useful nutrients such as folic acid have been identified as important chemotherapeutic drug targets^18,19,23. Hence, the folate pathway, metabolites and transporters continue to be extensively studied for identification of new enzymes including transporters, which may serve as new drug targets²². Recent estimates have ascribed eight different membrane transporters to eukaryotes²⁴.

Proteins that mediate transportation of folates have been well studied in a few parasites such as Plasmodium falciparum, Trypanosoma brucei, Leishmania donovani and Leishmania major^25,26. These studies have provided information on mode of action of drugs^25,27,28 in addition to studies describing mechanisms of parasite drug resistance^25–32. However, folate transport proteins remain unidentified and uncharacterized in many other eukaryotic pathogens. This is despite the sequencing of the genomes of most eukaryotic microbes, which has produced a vast wealth of data that could aid in identification of druggable pathogen-specific proteins^33–39. It is therefore imperative to search and identify from these parasite genomes additional proteins such as folate transporters that may serve as novel drug targets^40,41.

Therefore, in an attempt to identify and characterize targets for novel therapeutics, we report herein an extensive search of folate transporters from pathogen genomes. In addition, we investigated the evolutionary relationship of these transporters in a bid to determine similarities and differences that make them attractive drug targets. The knowledge provided may assist in the design of new antifolates for protozoan parasites.

Methods

We extracted protein sequences of approximately 200-pathogens that mediate transportation or salvage of folates from Eukaryotic Pathogen Genome Database Resources (http://eupathdb.org/eupathdb/), and from the literature using a key-word search. We also searched the 200-pathogen genome sequences archived at the Eukaryotic Pathogen Genome Database Resources (http://eupathdb.org/eupathdb/). The search was for all proteins that mediate transportation or folate salvage alone or folate salvage and related compounds (such as pteridine, biopterin and methotrexate) together. This database gives public access to most sequenced emerging/re-emerging infectious pathogen genomes⁴². We utilized the word “folate” for search on the gene text and “folic acid” was used to confirm the hits. Hit results containing proteins annotated as folate-binding protein YgfZ, folate/pteridine transporter, folate/biopterin transporter, reduced folate carrier family protein, folate/methotrexate transporter FT1, Folate transporters alone and other folate related proteins were retrieved. The complete list of proteins extracted from EuPathdb is presented in Dataset 1⁴³. The folate transporters were classified based on type of transporter, number of transmembrane helix (TMH) and localization (either cell or mitochondrial membrane) of transporter. Gene sequences were obtained in FASTA format for transporter proteins using the sequence download tool on EuPathDB (http://eupathdb.org/eupathdb/).

Figure 1.

Categorization of proteins identified based on [A] Type of transporters and [B] Number of TMHs.

To ensure that most of the retrieved proteins had not been previously studied, we performed a literature search on PubMed (http://www.ncbi.nlm.nih.gov/pubmed/?term=) and Google Scholar (https://scholar.google.com) using the query “folate transporter + Parasite name”. The protein sequence information (Dataset 1⁴³) obtained from literature search was used for a BLAST search on EupathDB (http://eupathdb.org/eupathdb/), UniprotDB (http://www.uniprot.org) and GeneDB (http://www.genedb.org/Homepage). The protein information are included in Dataset 1⁴³ and summarised in Dataset 2⁴⁴, which are marked as either identified in other literature or in this research work. Sequence data were edited on textEdit mac version and uploaded to Molecular Evolutionary Genetics Analysis (MEGA) platform version 7.0 obtained from http://www.megasoftware.net⁴⁵. The 234 sequences were aligned using muscle tools with large alignment (Max iterations = 2) selected while other settings were left at defaults. Evolutionary history was inferred using the Neighbor-Joining method⁴⁶. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (500 replicates) was also analysed⁴⁷. The tree was drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances were computed using the number of differences method⁴⁸. While uniform rate and complete deletion was selected for substitution rates and data subset, respectively. Other parameters were at default settings. All positions containing gaps and missing data were eliminated. The newick format of the tree was exported and opened on FigTree 1.4.2 platform downloaded from http://tree.bio.ed.ac.uk/software/figtree/⁴⁹. The final tree was constructed using radial tree layout. Additional analysis consisted of sub-phylogenies based on the transporter type. Since folate-binding protein YgfZ, folate/pteridine transporter, folate/biopterin transporter, putative, reduced folate carrier family protein, folate/methotrexate transporter FT1, putative folate transporters alone and others have 10, 25, 132, 2, 7, 49 and 9. So we decided to reconstruct the phylogeny based folate transporter, folate-biopterin transporter after considering the identification number, the species diversity in each category.

Results

A methodological search for folate transporters in all eukaryotic microbe genomes examined under EuPathDB with validation via GenBank, GeneDB and UniProt contained a total of 234 proteins (detail features of proteins are presented in Dataset 1⁴³). We identified these transporters in 28 pathogen species (containing 63 strains) cutting across 12 phyla (Dataset 2⁴⁴). The parasites with the highest number of folate transporters are Phytophthora parasitica INRA-310, P. infestans T30–4 and Leptomonas pyrrhocoris H10 with 20, 16 and 16 proteins, respectively. While Aspergillus clavatus NRRL 1, A. flavus NRRL3357, A. macrogynus ATCC 38327, Crithidia fasciculata strain Cf-Cl and others have one folate transporter protein each (Dataset 2⁴⁴). The different proteins identified to be involved in folate salvage or related molecules were folate-binding protein YgfZ, folate/pteridine transporter, folate/biopterin transporter, reduced folate carrier family protein, folate/methotrexate transporter FT1 and folate transporters having a 4%, 11%, 56%, 1%, 3% and 21% identity, respectively. Proteins that did not belong to these groups were classified as others (4%) (Figure 1A). A good number of the proteins identified had predicted transmembrane helixes with a few having none (Figure 1B). Furthermore, a number of the transporters possess signal peptides (Dataset 1⁴³), which may be required for targeting to cellular locations. Deciphering the sequence of the targeting signal may indicate its product destination.

Figure 2. Phylogenetic tree showing the relatedness of all the proteins identified to transport folate in different pathogen species.

Tree was inferred using the Neighbor-Joining method with bootstrap test at 1000 replicates.

Our literature search for parasite folate transporters on PubMed and Google Scholar indicated 60% (38 out 63) of the proteins were identified for the first time as presented in Dataset 2⁴⁴, while 40% have been previously investigated. Besides, the Leishmania folate transporters we came across were not found on the EupathDB resource. We thus performed a BLAST search of Kinetoplastida on EupathDB, the returned hits were folate/biopterin transporter for L. infantum. The only Plasmodium species with results for proteins that salvage folate was P. falciparum. Our study, however, describes for the first time the presence of these transporters in other Plasmodium species. There were no transporter proteins deposited in EupathDB for P. malariae and P. ovale. However, folate transporters I and II were retrieved from our search of GeneDB for P. malariae and P. ovale curtisi, respectively.

Our analysis of folate transporters indicate the presence in Plasmodium species of two proteoforms; folate transporter I and II (Dataset 1⁴³). All Leishmania species identified possess folate/biopterin transporters and not folate transporters. Trypanosome species have both folate/pteridine and folate transporters; T. cruzi Dm28c, T. cruzi Sylvio X10/1 and T. cruzi CL Brener Esmeraldo-like, T. cruzi CL Brener Non-Esmeraldo-like and T. cruzi marinkellei strain B7 all have folate/pteridine transporter while T. brucei TREU927, T. brucei Lister strain 427, T. brucei gambiense DAL972, T. congolense IL3000 possess folate transporters. Eukaryotic parasites like Eimeria acervulina Houghton, E. brunetti Houghton, E. maxima Weybridge, E. necatrix Houghton, E. praecox Houghton, E. tenella strain Houghton and Neospora caninum Liverpool all boast folate/methotrexate transporter FT1. The folate-binding protein YgfZ was found in the fungus, Allomyces macrogynus ATCC 38327, the protist C. fasciculata strain Cf-Cl, C. immitis RS, the feline protozoon, Hammondia hammondi strain H.H.34, Sarcocystis neurona SN3, S. punctatus DAOM BR117, T. gondii GT1, T. gondii ME49, T. gondii VEG and T. brucei TREU927. Parasites such as Microsporidium daphniae UGP3 and the amoeba Naegleria fowleri ATCC 30863 possess the reduced folate carrier family protein. We observed that 7% of the identified proteins are localized on the mitochondrial membrane of some pathogens such as the fungi Aspergillus clavatus NRRL 1, A. flavus NRRL3357, C. immitis RS, the yeast Cryptococcus neoformans var. grubii H99, Fusarium graminearum PH-1, A. capsulatus G186AR, Leptomonas pyrrhocoris H10, the food fungus Neosartorya fischeri NRRL 181, Phytophthora parasitica INRA-310 and P. ultimum DAOM BR144. The remaining proteins are localized on the plasma membrane (Dataset 1⁴³).

Approximately 15% (34/234) of the folate transporters identified possess signal peptides with the trypanosomes with the most signal peptides. Deductions can be made of the probable destination within the cell of any transporter by its signal peptide sequence; thus, further work may seek to decipher the sequence of the targeting signal to determine its localization. The proteins identified all have transmembrane helixes with the exception of the alveolate Chromera velia CCMP2878, apicomplexan P. berghei ANKA, S. neurona SN3, the kinetoplastid T. brucei TREU927, T. grayi ANR4 and protist Vitrella brassicaformis CCMP3155 with Gene ID’s Cvel_17766, PBANKA_0713700, SN3_01500005, Tb927.8.6480, Tgr.2739.1000 and Vbra_15327, respectively (Dataset 1⁴³).

The phylogenetic tree (Figure 2) shows the evolutionary position, history and relationship of all the folate transporters identified in this work. The type of transporter or species/strain was used for constructing phylogenic trees, with the 234 proteins identified forming two clades, a major and minor. The major clade lacked a sub-clade, while the minor clade possessed a sub-clade. All proteins identified were distributed between the two major clades; except for folate/methotrexate transporter and mitochondrial folate transporter, with the latter present on the major clade and the former on the minor clade exclusively. All the species are represented on both clades, however, V. brassicaformis CCMP3155, Plasmodium species, A. clavatus NRRL, A. flavus NRRL3357, A. macrogynus ATCC 38327, C. fasciculata strain Cf-Cl, C. immitis RS, C. immitis RS, C. muris RN66, C. neoformans var. grubii H99, C. neoformans var. grubii H99, Leishmania species, N. bombycis CQ1, N. caninum Liverpool, F. graminearum PH-1 and H. hammondi strain H.H.34 are exclusively on the major clade. There are some parasites that were identified once, as shown in Dataset 1⁴³; these are mostly in the large clade. Some of these pathogens include P. ultimum DAOM BR144, which has mitochondrial folate transporter/carrier proteins similar to Homo sapiens, E. cuniculi GB-M1, which has proteins similar to folate transporter, and S. punctatus DAOM BR117, which has folate-binding protein YgfZ. These were the only proteins of the aforementioned species identified in this work. However, M. daphniae UGP3, which had reduced folate carrier domain containing protein, was the only parasite that was found in the small clade. Improving on our phylogenetic analysis, we performed a sub-phylogenetic reconstruction (Figure 3–Figure 5) based on the substrate type of the transport proteins. After phylogenetic analysis each sub-phylogeny show a clear characterization except for folate-biopterin transporters (Figure 4), which fell in a different clade save for Leptomonas species and C. velia. The newick formats of the phylogenetic trees in Figure 2–Figure 5 are presented as Supplementary Dataset 1–Supplementary Dataset 4.

Figure 3. Phylogenetic tree showing the relatedness of folate transporters alone in different pathogen species.

Tree was inferred using the Neighbor-Joining method with bootstrap test at 1000 replicates.

Figure 4. Phylogenetic tree showing the relatedness of folate/biopterin transporter alone in different pathogen species.

Tree was inferred using the Neighbor-Joining method with bootstrap test at 1000 replicates.

Figure 5. Phylogenetic tree showing the relatedness of folate-binding protein YgfZ alone in different pathogen species.

Tree was inferred using the Neighbor-Joining method with bootstrap test at 1000 replicates.

Discussion

Folate transporters are important proteins involved in the salvage of folate, cofactors and related molecules in eukaryotic pathogens important for metabolism and survival in their respective hosts²¹. We identified proteins that could mediate the salvage of folates into cells and/or mitochondria from eukaryotic microbe genomes in EuPathDB. Many of these proteins are involved in folate biosynthesis or transport and are present in most of the eukaryotic microbe genomes we queried. In this study, 234 genes encoding homologues of folate salvaging proteins were identified in the genome of 64 strains, representing 28 species of eukaryotic microbes. Some of the pathogens among the microbes queried include P. falciparum 3D7 and IT, P. knowlesi H, P. berghei ANKA, P. chabaudi chabaudi, T. brucei Lister 427, T. brucei TREU927, T. brucei gambiense DAL972, Encephalitozoon cuniculi GB-M1. The pathogens range from bacteria through to fungi, intracellular parasites such as Plasmodium and leishmania species, to extracellular parasites such as trypanosome species. This suggests a widespread presence of the proteins cutting across a range of pathogens that infect humans and animals.

About 40% of the proteins we identified have been previously identified and characterized in parasites such as Plasmodium falciparum^22,30, Trypanosome species²⁶, Leishmania species and Toxoplasma gondii⁵⁰. It has been estimated that over half of the drugs currently on the market target integral membrane proteins ion channel blockers like verapamil, and serotonin transporter inhibitors feature prominently in the WHO model list of essential medicines^51,52. Unfortunately, a great number of these transporters have not been adequately explored as drug targets⁵³. Folate transporters therefore represent attractive drug targets for treatment of infectious diseases. Thus their identification from other eukaryotic pathogens could open a window for novel chemotherapeutics for disease control^54,55.

In Plasmodium two folate transporters have been identified, namely PfFT1 and PfT2. These transporters have been shown to mediate the salvage of folate derivatives and precursors in P. falciparum and proposed blocking of their salvage activities may improve the antimalarial efficacy of several classes of antimalarial drugs. In our work we identified folate transporters for other plasmodial species, which, like P. falciparum, may also be chemotherapeutic targets. Transport of folate in higher eukaryotes is made possible by a high affinity folate-biopterin transporters FBT or BT1 family^22,30. In the trypanosomes and related kinetoplastids, a member of these transporters, the folate biopterin transporter (FBT) family of proteins was identified in Leishmania²⁸. It is thought that MFS proteins are related to the FBT. These proteins have been characterized in a few protozoa and cyanobacteria⁵⁶. Results from our study describing the presence of these transporters across several phyla corroborate results from other works establishing the conservation of folate transport function among FBT family proteins from plants and protists^22,56.

Malaria parasites encode transporters belonging to the organo anion transporter (OAT) folate-biopterin transporter (FBT), glycoside-pentoside-hexuronide: cation symporter (GPH), families, which are closely related to the major facilitator super-family of membrane proteins⁵⁷. The inhibition of these transporters by blockers of organic anion transporters such as probenecid has been implicated in sensitization of Plasmodium resistant parasites to antifolates^58,59. Thus, in Plasmodium chemotherapy, identification of folate transporters could lead to screening for compounds that interfere with folate transport and salvage for antimalarial chemotherapy^22,30. We identified several types of folate transporters that have been described and functionally characterized in Leishmania with some implicated in the import of the antifolate methotrexate^60,61. Thus far, only protozoan transporters in Plasmodium, Leishmania, and Trypanosoma brucei have been characterized and these are known to mediate the uptake of the vitamins folate and/or biopterin^22,62,63. Thus in parasites species of medical importance folate transporter proteins may provide new targets for therapy.

We also identified folate salvaging proteins from fungi such as Coccidioides immitis and A. clavatus, fungi found in soil^64–66, vegetable⁶⁴ and waters in tropical and subtropical areas⁶⁷. These fungi are known to occasionally become pathogenic and act as opportunistic pathogens for animals and man⁶⁶. Coccidioidomycosis caused by C. immitis in association with AIDS has been known to be a fatal disease⁶⁸. Treatment of acute and chronic infections with antifungals such as amphotericin B have not been adequate, hence folate transporters may present new targets in these group of pathogens. Identification was also made on pathogens such as C. fasciculata that parasitize several species of insects including mosquitoes and has been widely used to test new therapeutic strategies against parasitic infections⁶⁹. As a model organism, folate transporters identified in C. fasciculata may be useful in research for developing new drugs in medically important Kinetoplasts as has been shown for other targets in this protozoan parasite⁷⁰.

We noticed that P. parasitica INRA-310 and L. pyrrhocoris H10 had the highest number of folate transporters identified. Their utility as model fungal (P. parasitica) and monoxenous kinetoplast may provide models instrumental for developing new antifolates for fungal and protozoan diseases. The relatedness of these proteins across the different pathogens show that there are two major phylogenetically distinct clades in the eukaryotic pathogens examined. The clustering of these proteins suggests that these transport proteins have highly conserved regions often required for basic cellular function or stability^60,71–87. Thus, antifolate chemotherapic drugs that are effective against one pathogen might have some effect on others. However, the converse may be the case for the free-living non-parasitic photosynthetic algae, Chromera velia and Vitrella brassicaformis, protists related to apicomplexans^88,89. These groups of algae live freely in their environment, which unlike apicomplexans that depend on a host animal to survive⁸⁸. This adaptation may explain the difference in the clustering of their transporters after phylogenetic analysis, which separated on the minor clade from other apicomplexans that separated on the major clade. This suggests a high level of evolutionary divergence between folate transporters in both the apicomplexans and these algae based on life-style adaptations.

Conclusion

In summary, we have retrieved information on 234 folate transporter proteins from Eukaryotic Pathogen Database (EuPathDB) resources. The folate transporter proteins were categorized into potential drug targeting features including mitochondrial localization, number of transmembrane helix, and protein sequence relatedness. The identification of folate salvage proteins in diverse eukaryotes extend the evolutionary diversity of these proteins and suggests they might offer new possibilities for potential drug development targeting folate-salvaging routes in eukaryotic pathogens.

Data availability

Dataset 1: Complete list of proteins extracted from Eupthadb and literature search, including their properties. These data are available in a .xlsx file. Doi, 10.5256/f1000research.10561.d167325⁴³

Dataset 2: Eukaryotic microbes from which folate transporters were identified. These data are available in a .xlsx file. Doi, 10.5256/f1000research.10561.d167326⁴⁴