Insights into the role of the junctional region of Plasmodium falciparum dihydrofolate reductase-thymidylate synthase

Background Plasmodium falciparum dihydrofolate reductase-thymidylate synthase (pfDHFR-TS) is a well-defined target of anti-malarial drug, such as pyrimethamine and cycloguanil. Emergence of malaria parasites resistant to these drugs has been shown to be associated with point mutations of the gene coding for the target enzymes. Although the 3D-structure of P. falciparum bifunctional pfDHFR-TS has been reported previously, relatively little is known about the interactions between the pfDHFR and pfTS domains and the roles of the junctional region that links the two domains together. Therefore, a thorough understanding of the interaction of the two domains and the role of the junctional region of this target is important as the knowledge could assist the development of new effective anti-malarial drugs aimed at overcoming drug-resistant malaria. Methods A system was developed to investigate the interaction between pfDHFR and pfTS domains and the role of the junctional region on the activity of the recombinant pfTS. Based on the ability of co-transformed plasmids coding for pfDHFR and pfTS with truncated junctional region to complement the growth of TS-deficient Escherichia coli strain χ2913recA(DE3) on minimum media without thymidine supplementation, active pfTS mutants with minimal length of junctional region were identified. Interactions between active pfDHFR and the pfTS domains were demonstrated by using a bacterial two-hybrid system. Results Using TS-deficient E. coli strain χ2913recA(DE3), the authors have shown for the first time that in P. falciparum a junctional region of at least 44 amino acids or longer was necessary for the pfTS domain to be active for the synthesis of thymidylate for the cells. Truncation of the junctional region of the bifunctional pfDHFR-TS further confirmed the above results, and suggested that a critical length of the junctional peptide of pfDHFR-TS would be essential for the activity of TS to catalyze the synthesis of thymidylate. Conclusion The present study demonstrated the interactions between the pfDHFR and pfTS domains of the bifunctional pfDHFR-TS, and revealed that the junctional region linking the two protein domains is essential for the expression of catalytically active pfTS domain. The findings could be useful since inhibition of the pfDHFR-TS domain-domain interaction could form a basis for the development of new anti-malarial drugs based on targeting the non-active site region of this important enzyme.


Background
Malaria remains an important disease in many tropical and sub-tropical countries [1]. The disease has become a global health threat, with over one million deaths annually, mostly children in sub-Saharan Africa [2]. The emergence of malaria resistance to almost all the currently available antimalarial drugs has highlighted an urgent need to identify new malarial targets, and develop new effective drugs to combat the drug-resistant parasites [3][4][5][6][7][8]. Plasmodium falciparum dihydrofolate reductase-thymidylate synthase (pfDHFR-TS) is a well-defined target of antifolate drugs such as pyrimethamine and cycloquanil. The enzyme is responsible for the production of folates as well as thymidylate (dTMP) required for DNA synthesis [9]. Unfortunately, the emergence of anti-folate resistance has compromised the utility of the drugs and presented an urgent need to discover new drug targets and to develop novel effective drugs to combat drug-resistant parasites.
Structural studies of P. falciparum DHFR-TS revealed that the native enzyme is a homodimeric protein comprising 231 residues of DHFR domain (~27 kDa) at the N-terminus, followed by a short junctional region of 89 residues (~11 kDa) and 288 residues (~34 kDa) of the TS domain at the C-terminus of the protein [10,11]. It has been postulated that such a bifunctional arrangement could have evolved as a mechanism for the tight coupling generation of reduced folates required for the synthesis of amino acids, purine, pyrimidine, and dTMP, a phenomenon called "substrate channelling". Support for this hypothesis comes from the evidence of metabolic channelling of the H 2 folate produced in the TS-catalyzed reaction, which was found to proceed at a faster rate than the diffusion rate [12]. Data from the bifunctional DHFR-TS of Leishmania major and Toxoplasma gondii also supported the substrate channelling hypothesis [12,13]. Nevertheless, the mechanism of substrate channelling for pfDHFR-TS remains unclear since, based on the structure of pfDHFR-TS, such a mechanism could not explain the delivery of H 2 folate from the pfTS domain to the active site of pfDHFR, and an electrostatic channelling mechanism was proposed as a possible alternative method [14][15][16].
The junctional region (JR) linked between the DHFR and TS domains in parasitic protozoa reported thus far varies significantly in length depending upon the source [9,10,16,17]. The long JR found in P. falciparum (89 amino acids) provides a number of interactions which facilitate contacts with the DHFR domain of the opposite half of the DHFR-TS dimer, and brings the two DHFR domains closer together. Structural alignments of DHFR-TS enzymes from Cryptosporidium hominis and P. falciparum revealed that the JR played an important role in the orientation of the DHFR domain relative to TS [17]. Therefore, inhibition of the interaction between the JR, the DHFR and TS domains could be a possible approach for the development of novel effective anti-malarial drugs [18]. The present study therefore describes an approach towards understanding the interactions between DHFR and TS domains of the bifunctional pfDHFR-TS, and the important role of JR on the activity of the TS domain. This is the first successful expression of the catalytically active pfTS domain that has JR attached at its N-terminus, as all attempts in the past to express pfTS domain failed. Through deletion of the JR, it is demonstrated that a critical length of JR is required for proper folding of the pfTS domain leading to an active molecule of pfTS. The findings highlight the importance of the JR which links DHFR and TS domains and suggest a possible alternative in exploiting the non-active site region of this important enzyme in developing new anti-malarial drugs to overcome drug-resistant parasites.

Construction and transformation of recombinant plasmids
The gene coding for P. falciparum DHFR (amino acids 1-228) was amplified using a template DNA from pETpfDHFR-TS [21]. The pfJRTS mutants were constructed by the PCR mutagenesis method. The PCR reaction (100 μl) is composed of 50 ng pETpfDHFR-TS as a DNA template, 25 ρmole of each primer (Table 1), 200 μM of dNTPs, 1.5 mM MgCl 2 , and 2.5 units of Taq DNA polymerase in 1 x reaction buffer. The PCR conditions were as follows: 1 cycle of 94°C for 3 min, then 25 cycles of 94°C for 45 sec, annealing at 45°C for 30 sec, and extension at 72°C for 1 min. This was followed by a final extension at 72°C for 5 min. Deletion mutants of bifunctional pfDHFR-TS were constructed using the whole plasmid amplification PCR approach [22] employing recombinant plasmid pET15b carrying the P. falciparum DHFR-TS(3D7) gene as a template and sets of primers as shown in Table 2. The PCR products were analysed by agarose electrophoresis, and were further purified using a Qiaquick Gel Extraction kit.

Genetic complementation studies
Overnight culture of the co-transformants was streaked on a minimal agar plate [24] containing 100 μg/ml ampicillin, 30 μg/ml kanamycin, and 0.025 mM IPTG without thymidine supplementation. The plates were incubated at 37°C for 48 hours. Colonies appearing on this plate were considered to have positive genetic complementation, and were selected for subsequent characterization and verification of the expressed enzymes.

Expression and purification of the expressed enzymes
An overnight culture of E. coli χ2913 harbouring two plasmids, i.e., pAC-pfDHFR and truncated pET-pfJRTS with different lengths of truncated JR, was inoculated at 1% inoculum in LB containing 100 μg/ml of ampicillin, 30 μg/ml kanamycin. The bacteria were grown at 37°C until the OD 600 of the cell suspension reached~0.5-0.6, and isopropyl β-D thiogalactopyranoside (IPTG) was added at a final concentration of 0.025 mM. The culture was allowed to grow with shaking at 20°C for 20 hours. The cells were harvested by centrifugation at 6,500 g for 10 min at 4°C, washed once with 250 ml cold phosphate buffered saline, pH 7.4, resuspended in buffer A (20 mM potassium phosphate buffer, pH 7.0, 0.1 mM EDTA, 10 mM DTT, 20% glycerol) containing 0.2 M KCl, and passed through a French Pressure Cell (American Instruments Co Inc, USA) at 12,000 psi three times. After centrifugation at 20,000 g for 1 hour at 4°C, the clear supernatant of the crude sample was circulated at a flow rate of~0.5 ml/min in a methotrexate-sepharose CL-6B column (1.5 × 5.0 cm) pre-equilibrated with buffer A containing 0.2 M KCl. After overnight circulation, the column was washed with 30 ml of buffer A containing 0.75 M KCl, followed by 20 ml of buffer A containing 0.2 M KCl. The column was then washed with 30 ml of elution buffer (50 mM TES pH 7.8, 0.1 mM EDTA, 10 mM DTT, 20% glycerol, 50 mM KCl) containing 4 mM H 2 folate to elute DHFR. Fractions of 1 ml were collected. Active fractions with DHFR activity were pooled, concentrated, and H 2 folate in the pooled fraction was removed by passing the pooled fraction through a pre-packed NAP-25 column (Pharmacia) pre-equilibrated with buffer A.

Enzyme assays and protein analysis
The activity of DHFR was determined spectrophotometrically by monitoring the rate of decrease in absorbance at 340 nm [12,25]. The standard DHFR assay (1 ml) in a 1-cm path-length cuvette was composed of 100 μM H 2 folate, 100 μM NADPH, 50 mM TES, pH 7.0, 75 mM β-mercaptoethanol, 1 mg/ml bovine serum albumin, and 0.01 units of enzyme. The reaction was initiated with H 2 folate. One unit of DHFR activity is defined as the amount of enzyme that produces 1 μmole of product per minute at 25°C. The activity of TS was determined by monitoring the increase of absorbance at 340 nm due to the formation of H 2 folate at 25°C [26]. The reaction (1 ml) in 1-cm pathlength cuvette was composed of 50 mM TES, pH 7.4, 25 mM MgCl 2 , 1 mM EDTA, 6.5 mM HCHO, 75 mM βmercaptoethanol, 100 μM (6R) CH 2 H 4 folate, 125 μM dUMP and the enzyme. The reaction components, except for dUMP, were incubated at 25°C for at least 5 min to obtain the baseline prior to initiation the reaction with dUMP. One unit of TS activity is defined as the amount of enzyme that produces 1 nmole of product per minute at 25°C. The activities of the expressed enzymes were reported as mean ± standard deviation.

[ 3 H]-FdUMP binding assay
The ability of the pfJRTS mutants with truncated JR to form covalent complex with [ 3 H]-FdUMP and CH 2 H 4 folate was investigated by incubating the mutant proteins with 0.5 μM [ 3 H]-FdUMP (19.3 Ci/mmol), 0.1 mM CH 2 H 4 folate, and 6.5 mM formaldehyde in 50 mM TES pH 7.4, 25 mM MgCl 2 , 1 mM EDTA, and 75 mM β-mercaptoethanol for 15 min at room temperature. The reaction was then electrophoresed on 12.5% SDS-PAGE. After Coomassie-Blue staining and destaining, the gel was soaked with autoradiography enhancer (EN 3 HANCE™, New England Nuclear) with gentle shaking for 30 min at room temperature. This was followed by washing the gel with cold water and drying the gel on a piece of filter paper under vacuum at 80°C for 30 min. The dried gel was then exposed on AGFA X-ray film at −80°C with intensifying screens for three to five days before development of the X-ray film.

Bacterial two-hybrid system
Interactions between pfDHFR domain and pfTS domain including those pfTS mutants with truncated JR were investigated using an E. coli two-hybrid system (Stratagene Inc) [27,28]. The sequence coding for pfDHFR domain was cloned into pBT (bait) vector to yield pBT-pfDHFR, whereas the pfJRTS was cloned into pTRG (target) vector to yield the pTRG-pfJRTS plasmid. The two plasmids were co-transformed into BacterioMatch two-hybrid system reporter strain E. coli XL1-blue MRF'. Positive interaction is indicated by the ability of the bacteria to grow on LB agar plate supplemented with 250 μg/ml carbenicillin, 15 μg/ml tetracycline, 34 μg/ml chloramphenical, and 50 μg/ ml kanamycin (LB-CTCK).

Construction of pfJRTS and pfDHFR-TS mutants and growth complementation studies
Unlike the DHFR domain of the bifunctional pfDHFR-TS, which can be heterologously expressed to yield the catalytically active form of enzyme [29][30][31][32], all attempts to express the catalytically active domain of pfTS have failed. Evidence from deletion of pfDHFR-TS suggested that the amino terminus of the pfDHFR domain is important for the function of the pfTS domain, and interactions between the pfDHFR and pfTS domains are important [18]. To address the important function of JR, mutant constructs of the pfJRTS domain with various lengths of JR sequences were constructed and co-transformed with plasmid expressing the catalytically active pfDHFR domain (pAC-pfDHFR) in TS-deficient E. coli strain χ2913recA(DE3). Successful expression of catalytically active pfTS was monitored by the ability of the co-transformed bacteria to complement growth upon plating on minimum agar plate and the detectable TS activity in the crude extract of the cells. Figure 1 illustrates the constructs of pfJRTS mutants containing varying lengths of JR attached to the Cterminus of the pfDHFR domain (Asn 231 ). The mutant with the longest JR (pET-pfJRTS△232-235) had only four amino acids of the JR (Lys 232 -Asn 235 ) deleted, while mutants with the shortest JR (pET-pfJRTS△232-299) had 68 residues (Lys 232 -Lys 299 ) removed. A similar approach was undertaken to construct bifunctional pfDHFR-TS deletion mutants of which the JR sequence was shortened to compare the effects of JR deletion with the truncated mutants of pfJRTS constructs. The bifunctional pfDHFR-TS deletion mutants being constructed include pET-pfDHFR-TS△229-265, pET-pfDHFR-TS△ 229-271, pET-pfDHFR-TS△229-274, pET-pfDHFR-TS△229-275, pET-pfDHFR-TS△229-276, and pET-pfDHFR-TS△229-277 ( Figure 2). Figure 3 shows the 89 amino acids of P. falciparum JR sequence and indicates the positions of amino acids that were chosen to prepare truncated and deletion mutants. Results from growth complementation monitored after incubation of the plates overnight at 37°C revealed that pET-pfJRTS△232-235, pET-pfJRTS△232-251, pET-pfJRTS△232-265, pET-pfJRTS△ 232-271, pET-pfJRTS△232-274, and pET-pfJRTS△232-276 with corresponding JR length of 85, 69, 55, 49, 46, 44 amino acid residues, respectively, could grow on minimum media ( Figure 4A, lanes 2-7), while pET-pfJRTS△232-277 and pET-pfJRTS△232-299 which had the JR length of 43 and 21 residues, respectively, did not show growth complementation ( Figure 4A, lanes 8-9). The results suggest that the length of JR was important for the proper folding of the TS domain and hence affected its function. The data revealed that at least 44 residues (the mutant pET-pfJRTS△232-276) were required in the case of JR of P. falciparum in order for the pfTS domain to function properly.

Purification and characterization of the expressed enzymes
Crude extracts of the transformed TS-deficient E. coli that were both positive and negative from the growth complementation studies were assayed for DHFR and TS activities. Extracts that were positive were passed through a MTX-affinity column for purification of the pfDHFR domain [32]. The results from enzyme assays  revealed that the specific activity of pfDHFR from all the co-transformants were not significantly different among the crude extracts (average 36.2 ± 3.1 nmole/min/mg), but the activity was only about half of that observed from the crude extract of the bifunctional pfDHFR-TS (Table 3).
The specific activities of the partially purified pfDHFR upon affinity purification were 1,371 ± 208 nmole/min/mg (Table 3), about 3 times lower than that obtained from the purified pfDHFR-TS. Interestingly, the TS activities of the purified enzymes from the truncated pfJRTS mutants are  in most cases 1-2 times higher than that from the purified pfDHFR-TS. The activity of pfTS was found to co-elute with that of pfDHFR, suggesting that the expressed pfTS domain was somehow associated with the pfDHFR domain. As indicated in Table 3, the TS activity of the pfJRTS mutants was found to decrease upon shortening the length of JR in the construct.
The importance of the length of the JR sequence on the activity of pfTS was also investigated using the bifunctional pfDHFR-TS. Site-directed mutagenesis of the bifunctional pfDHFR-TS was performed to yield mutants containing the same lengths of JR as for the pfJRTS mutant constructs described above. These mutants include pET-pfDHFR-TS△229-265, pET-pfDHFR-TS△229-271, pET-pfDHFR-TS△ 229-274, pET-pfDHFR-TS△229-275, pET-pfDHFR-TS△229-276, and pET-pfDHFR-TS△229-277. Table 4 summarizes the pfDHFR and pfTS activities of the bifunctional DHFR-TS and the bifunctional mutants from the crude extracts and upon partial purification. In agreement with the growth complementation experiments, no pfTS activity was detected from the crude extracts and partial purification of pET-pfDHFR-TS△229-275, pET-pfDHFR-TS△229-276, and pET-pfDHFR-TS△229-277 mutants. However, the activities of the pfDHFR domain of these mutants remained active, though they were found to gradually decrease upon shortening the length of the JR sequence.

Demonstration of pfDHFR-TS domain-domain interaction by Escherichia coli two-hybrid system
The interaction between the pfDHFR domain and pfJRTS was demonstrated using E. coli BacterioMatch™ two-hybrid system (Strategene). The pfDHFR domain was cloned into NotI-BamHI sites of a pBT bait plasmid, resulting in a pBT-pfDHFR plasmid, which contains the pfdhfr gene fused at the end of the bacteriophage λcI gene. The pfJRTS domain was cloned into BamHI-XhoI sites of the pTRG target plasmid, resulting in constructs that express truncated pfJRTS domains fused with the αsubunit of RNA polymerase. If the pfJRTS domain could interact with the pfDHFR domain, then this would stabilize the binding of RNA polymerase located close to the promoter and activate the transcription of the Amp R reporter gene. As a consequence, the transformed reporter strain (E. coli XL1-blue MRF') showed growth Table 3 DHFR and TS activities in crude extracts and purified enzymes from TS-deficient Escherichia coli χ2913 harbouring pAC-pfDHFR and pET-pfJRTS with different lengths of JR on Luria-Bertani agar containing carbenicillin, tetracycline, chloramphenical and kanamycin (LB-CTCK).
The interaction between the DHFR monomer and the donated helix within the JR sequence has earlier been noted from the structures of the bifunctional DHFR-TS enzymes of C. hominis and P. falciparum [17,33]. An interaction of the pfDHFR and pfJRTS domains was demonstrated using the E. coli two-hybrid system [27,28,34]. The E. coli reporter strain co-transformed with plasmids pBT-pfDHFR and pTRG-pfJRTS (with full-length JR sequence) can grow on a LB-CTCK agar plate ( Figure 6). The data is in agreement with the complementation results and results from [ 3 H]-FdUMP binding studies supporting the importance of JR on the folding of the pfTS domain.

Discussion
The JR represents a junctional region linked between the DHFR and the TS domain of parasitic protozoa. This region has been proposed as potential target for drug development in many parasitic protozoa. Indeed, this region in C. hominis from one monomer was reported to make extension contacts with the DHFR active site of the other monomer [17]. In P. falciparum, the amino acid residues Asp 283 -Asn 296 of JR are strongly predicted to be involved in domain-domain interaction. Unfortunately, the major portions of JR (Lys 232 -Asn 280 ) were not seen in the crystal structure previously reported [33]. Therefore, it remains unclear whether each DHFR domain of P. falciparum is linked to the TS domain as seen in C. hominis or there is a domain swapping assembly. The role of JR was characterized with respect to interdomain interaction.
A series of truncated mutants of pfJRTS were constructed containing varying lengths of JR and their interactions with the active pfDHFR domain were examined. By employing TS-deficient E. coli χ2913 and monitoring its growth complementation in minimum media without thymidine supplementation, the study showed that the pfJRTS construct with the shortest length of JR that could still show growth complementation in TS-deficient E. coli χ2913 was pfJRTS△232-276. The specific activity of DHFR determined for the crude extract of the monofunctional pfDHFR upon co-expression with pfJRTS△232-276 was 34.2 ± 9.1 nmole/ min/mg, a value which was about half that obtained from the bifunctional pfDHFR-TS (70.3 ± 3.6 nmole/min/mg), but was about the same level for all truncated pfJRTS mutants. However, the specificity of TS of this truncated construct was only 1.1 ± 0.1 nmole/min/mg (Table 3), which is about 29% of the wild-type enzyme (3.7 ± 0.4 nmole/ min/mg). It is noteworthy that the DHFR specific activities among the co-transformants investigated were comparable whereas the TS specific activities were dramatically reduced upon shortening of the JR sequence. The pfDHFR and pfJRTS domains could be co-purified by using a MTX-affinity column, suggesting that the pfDHFR and pfTS domains interacted with each other because methotrexate binds only to the pfDHFR domain. For the co-transformants that lose the pfTS activity including pfJRTS△232-277 and pfJRTS△232-299, the constructs still expressed active pfDHFR ( Table 3). The expression of catalytically active pfTS upon cotransformation with pfDHFR These results agree well with the previous studies, suggesting that the pfDHFR domain is essential for pfTS to be active [18,35]. However, one interesting piece of data from this laboratory showed that co-expression of  Figure 6 Escherichia coli two-hybrid showing interaction between pfDHFR and pfTS domains. Escherichia coli BacterioMatch™ two-hybrid system (Strategene) was employed for the study of the domain-domain interaction between P. falciparum DHFR and TS domains. The gene coding for pfDHFR was cloned into the bait plasmid pBT, whereas the gene for pfJRTS was cloned into the target plasmid pTRG. The two recombinant plasmids were transformed into the E. coli XL1-blue MRF' reporter strain. Positive control is E. coli XL1-blue MRF' co-transformed with pBT-LDF2 and pTRG-Gal11 p provided by the manufacturer. Negative control is E. coli XL1-blue MRF'. Agar plate on the left is LB-kanamycin (LB-K) and that on the right is LB-carbenicillin, tetracycline, chloramphenical and kanamycin (LB-CTCK).
the death mutant pfDHFR (Ser 108 to Trp 108 mutation) [36] and pfJRTS△232-235 or P. vivax DHFR with pfJRTS△232-235 could not confer expression of catalytically active pfTS (data not shown). This finding suggests that not only minimal length of JR is required but active pfDHFR domain is also important for the pfTS domain to be active. Although endogenous DHFR from an E. coli host exists, the enzyme could not promote the growth of TS-deficient E. coli. It is possible that amino acids within the pfDHFR domain are critically necessary to stabilize the DHFR tertiary structure which could indirectly influence the correct folding and hence the activity of the pfTS domain. The domain-domain interaction of the bifunctional pfDHFR-TS was further demonstrated using an E. coli two hybrid system and the results obtained were consistent with other experiments. The pBT-pfDHFR and pTRG-pfJRTS co-transformant showed interaction and turned on the Amp R gene, allowing the growth of E. coli XL1-blue MRF'. The results suggest that the minimal length of JR which could allow expression of the catalytically active pfTS was 44 amino acids, approximately half the full length of JR of the P. falciparum DHFR-TS.
The importance of the length of JR was investigated by the ability to complement the growth of TS-deficient E. coli in the absence of thymine of co-transformed plasmids harbouring the genes for pfDHFR and truncated pfJRTS compared to plasmid harbouring the gene for the bifunctional pfDHFR-TS. Complementation experiments support the conclusion that the length of JR sequence is crucial. The finding that positive complementation was still observed from the construct of truncated bifunctional pfDHFR-TS△229-274 which possesses two more amino acid residues than the case of monofunctional pfJRTS△232-276 could be due to the more complex structure of the bifunctional protein as compared to the monofunctional enzymes resulting from the interaction, or associated with the sensitivity of the complementation assay system. To further address this discrepancy, a more sensitive [ 3 H]-FdUMP binding assay method was used to confirm that pfDHFR-TS△229-276 is the construct with the shortest length of JR that could still express active pfTS. The result was in good agreement with the results from complementation studies using monofunctional pfJRTS△229-276. The fact that the bifunctional mutant failed to show complemention in TS-deficient E. coli but could not detect the activity of pfTS by in vitro enzymatic assay could be due to the insufficient expression of protein in combination with the sensitivity of the complemention method, since a minimum TS activity of~1-5 nmole/min/ mg is necessary for complementation to be observed in E. coli χ2913 [37]. It should also be noted that the low expression observed in the truncated bifunctional pfDHFR-TS was consistent with a previous mutation study which proposed that the significant intracellular proteolysis activity associated with the α-helix structure in this region could be attributed to the low expression of the mutant proteins [38].
The present results clearly demonstrated that shortening the JR sequence in the pfDHFR-TS could affect the activity of both pfTS and pfDHFR. Contrary to that observed in T. gondii and C. hominis which have longer JR sequences, mutations of the JR sequence were reported to affect only the activity of DHFR but not the activity of TS [39,40]. This and other data support the conclusion that the length of JR sequence is necessary for the activity of both pfDHFR and pfTS domains, and hence highlights the importance of the JR sequence in P. falciparum bifunctional DHFR-TS.
It was not so surprising that the constructs pfJRTS△232-277 and pfDHFR-TS△229-277 showed no expression of catalytically active pfTS, despite the fact that the constructs contain a sequence that encompasses Asp 283 -Asn 296 , the region previously predicted to form α-helix and responsible for electrostatic interaction between the negatively charge residues and positive charge of the pfDHFR and pfTS from different domains [33]. A number of possible explanations could be as follows: (i) the length of JR in the constructs is too short, and this could affect the formation of a stable helix and consequently disturb the interaction with pfDHFR and interfere with the proper formation of pfTS conformation; (ii) too short a JR sequence could affect the structure which could result in a steric constraint or functional infringement of one domain upon the other; (iii) the existence of translational autoregulation could block translation of mRNA coding for bifunctional enzyme from binding to the DHFR domain of the protein [41]; and, (iv) the mutation could cause exposure of some parts of the enzyme which are sensitive to proteolysis [38]. Therefore, too short a JR sequence could neither support interaction between pfDHFR and pfTS domains nor contribute to proper folding of the pfDHFR-TS bifunctional protein.

Conclusions
The data presented in this study show that: a) the presence of an active pfDHFR domain and the appropriate length of JR are critical for pfTS to fold correctly to be catalytically active; and, b) deletion of JR affects both the activities of the pfTS and the pfDHFR domain of the bifunctional pfDHFR-TS. From this study, the JR of P. falciparum has a unique role which is different from other long linker parasitic protozoa. The data described here could be useful for the development of compounds that bind to the JR and interfere with the dimerization of the pfTS subdomains. These studies could potentially lead to novel means of development of inhibitors targeting the JR sequence of the parasite.