Unique features of apicoplast DNA gyrases from Toxoplasma gondii and Plasmodium falciparum
© Nagano et al.; licensee BioMed Central. 2014
Received: 24 July 2014
Accepted: 10 December 2014
Published: 19 December 2014
DNA gyrase, an enzyme once thought to be unique to bacteria, is also found in some eukaryotic plastids including the apicoplast of Apicomplexa such as Plasmodium falciparum and Toxoplasma gondii which are important disease-causing organisms. DNA gyrase is an excellent target for antibacterial drugs, yet such antibacterials seem ineffective against Apicomplexa. Characterisation of the apicoplast gyrases would be a useful step towards understanding why this should be so. While purification of active apicoplast gyrase has proved impossible to date, in silico analyses have allowed us to discover differences in the apicoplast proteins. The resulting predicted structural and functional differences will be a first step towards development of apicoplast-gyrase specific inhibitors.
We have carried out sequence analysis and structural predictions of the enzymes from the two species and find that P. falciparum gyrase lacks a GyrA box, but T. gondii may retain one. All proteins contained signal/transport peptides for localization to the apicoplast but T. gondii Gyrase B protein lacks the expected hydrophobic region. The most significant difference is in the GyrA C-terminal domain: While the cores of the proteins, including DNA binding and cleavage regions are essentially unchanged, both apicoplast gyrase A proteins have C-terminal domains that are significantly larger than bacterial counterparts and are predicted to have different structures.
The apicoplast gyrases differ significantly from bacterial gyrases while retaining similar core domains. T. gondii Gyrase B may have an unusual or inefficient mechanism of localisation to the apicoplast. P.falciparum gyrase, lacks a GyrA box and is therefore likely to be inefficient in DNA supercoiling. The C-terminal domains of both apicoplast Gyrase A proteins diverge significantly from the bacterial proteins. We predict that an additional structural element is present in the C-terminal domain of both apicoplast Gyrase A proteins, including the possibility of a β-pinwheel with a non-canonical number of blades. These differences undoubtedly will affect the DNA supercoiling mechanism and have perhaps evolved to compensate for the lack of Topoisomerase IV in the apicoplast. These data will be useful first step towards further characterisation and development of inhibitors for apicoplast gyrases.
Apicomplexa are a group of unicellular protist parasites, most of which contain a plastid known as the apicoplast. Apicoplasts are the result of secondary endosymbiosis in which a prokaryotic cyanobacterium was incorporated into a unicellular eukaryote which was subsequently engulfed by a second eukaryote ,. As a result of this process, the apicoplast contains the remnants of the bacterial plasmid and has four membranes, which from outside inward are the second host’s endomembrane, the first host’s plasma membrane and the two apicoplast membranes .
The Apicomplexa are of interest because they count amongst their number several important human pathogens. Foremost amongst these are Plasmodium species, responsible for malaria, a disease which in 2010 infected approximately 200 million people resulting in over 600,000 deaths  and Toxoplasma gondii which can cause dangerous complications in the immune-compromised, is classified by the CDC as a “neglected parasitic disease” and is the biggest cause of death from foodborne illness in the USA .
Treatments for both toxoplasmosis (caused by T. gondii) and malaria exist but are not without their drawbacks. Most notably in the case of malaria, resistance to existing treatments is widespread  and has recently come to include artemisinin, the major component of the most effective current treatment ,. Progress has been made in development of a malaria vaccine but results of large scale clinical trials have been disappointing  although most recent small scale trials do show promise . Drug resistance in T. gondii is also regarded as problematic .
Apicoplasts are known to be essential for the survival of apicomplexan cells due to their numerous roles (reviewed by van Dooren and Striepen ) these include synthesis of heme, iron-sulfur clusters, fatty acids and isoprenoids. The requirement for an apicoplast was initially demonstrated in T. gondii, which was unable to survive when apicoplast DNA replication was inhibited , or when the apicoplast was absent . In Plasmodium their essential function in the blood stages appears to be synthesis of isoprenoid precursors .
The indispensability of the apicoplast together with fact that it is a eubacteria-derived plastid raises the possibility of exploiting it for specific targeting of pathogenic Apicomplexa with antibacterial drugs without affecting the human host .
Gyrase poisons work by inhibiting the religation of cleaved DNA, leading to fragmentation and cell death. The highly successful fluoroquinolone class of antibacterials function in this way and also target topoisomerase IV (“topo IV”), a closely related bacterial type II topoisomerase, in the same fashion. Topo IV itself is structurally highly similar to DNA gyrase, also functioning as a heterotetramer (ParC2, ParE2). Like gyrase it requires ATP hydrolysis to carry out its function, which is mainly decatenation of daughter chromosomes rather than negative supercoiling. Differences in the C-terminal domain of GyrA/ParC subunits appear to account for this difference in function .
Until recently, gyrase was thought to occur only in bacterial cells. It is now known that, while apparently not occurring in humans and most other higher eukaryotes, it is found in plant plastids . A malarial gyrase, whose existence had long been suggested by evidence such as the activity of fluoroquinolones against Plasmodium and T. gondii ,-, was finally shown to exist through sequencing of the P. falciparum genome  and is localized to the apicoplast. Apicoplast gyrase has been mooted as a potentially useful therapeutic target , although fluoroquinolone activity in vitro culture is usually much higher than in an infected host where results are variable. In the case of in vitro experiments with T. gondii for example, ciprofloxacin (CFX) has an IC50 of 27.9 μg/ml  and in the case of P. falciparum in in vitro culture, initial results suggested an IC50 for CFX of 1.7 μg/ml for 24-hour experiments . Mahmoudi et al. carried out a comprehensive in vitro test of 25 quinolones and fluoroquinolones against blood stages of P. falciparum  and found an IC50 for CFX of approximately 9.2 μg/ml and 3.4 μg/ml against chloroquine sensitive and resistant strains respectively. These numbers compare to IC50s for CFX against E. coli in the approximate range of 0.011-0.015 μg/ml .
Interestingly, the way in which CFX affects T. gondii appears different to the effect on P. falciparum as T. gondii shows a clear “delayed death” where extending treatment time lowers the IC50, something which appears not to occur in P. falciparum for CFX  although it is seen for some other molecules.
There is much less information regarding the effectiveness of these drugs against the pathogens in an infected host (“in vivo”). CFX, while active against T. gondii in in vitro culture, apparently shows little effect against the parasite in vivo . In contrast, the fluoroquinolone trovafloxacin is unusual in that it does show in vivo activity . Poor in vivo performance in general could be related to complications caused by parasite life cycle or simply the additional physical barriers in place in the host cell. Differences could also be due to differences in structure or mechanism of action of the enzymes themselves. Mature T. gondii gyrase (Tg-gyrase) is 142% the size of E. coli gyrase (Ec-gyrase) in terms of number of amino acids and P. falciparum gyrase (Pf-gyrase) is 116%, giving much scope for potential deviation from the E. coli “norm”. To understand the poor effectiveness of gyrase-poisoning antibacterials against Apicomplexa it is very important to biochemically and biophysically characterize purified Pf- and Tg-gyrases in vitro.
Summary of experiments previously carried out with P. falciparum gyrase proteins
GyrB (various constructs)
Decreased ATP turnover rate compared to E. coliprotein
ATPase activity can be enhance
DNA binding activity through 45 amino acid insert region
Dar et al., 2009 
Conformational change in response to nucleotide
Clamp closure can be modulated by nucleotide binding
Dar et al., 2009 
Dimerization occurred in the absence of ATP
Dar et al., 2007 
Ec-GyrA (full length)
Enhanced ATP turnover rate
Dar et al., 2007 
Lower supercoiling activity compared to E. coli gyrase
Introduced DNA breaks
Introduced DNA breaks
Dar et al., 2007 
Failed in DNA supercoiling
Dar et al., 2007 
Ec-GyrB (full length)
Introduced DNA breaks
Dar et al., 2007 
Pf-GyrB also shows ATP-hydrolysis and DNA supercoiling activities when complemented with Ec-GyrA . Furthermore the enzyme appears resistant to drugs that function by disrupting the ATP-binding site with the Ki value of coumermycin reported to be 500 times that of the E. coli protein  and the Ki value of novobiocin being approximately 3.5 times higher for Pf-GyrB compared to Ec-GyrB .
While full-length Pf-GyrA (residues 156–1222) has not been successfully expressed and purified, it has proved possible to produce the N-terminal DNA cleavage-reunion domain (residues 163–540) . This domain shows CaCl2-induced and CFX-induced DNA cleavage when combined with Pf-GyrB . The expression and characterization of the full-length CTD of Pf-GyrA has not been reported to date, although an N-terminal portion of it (residues 723–887) has been produced .
The lack of a purified full length Pf-GyrA has undoubtedly hampered progress in understanding Pf-gyrase. Still less is known of the Tg-gyrase proteins, the purification of which has not been reported to date (as is the case for all other apicomplexan gyrases). If these enzymes could be purified and characterized biochemically and structurally in comparison with standard eubacterial gyrase the data generated could help in the design of new inhibitors and may also point to a better understanding of the mechanism of action of gyrase itself.
In terms of structure, alignments for P. falciparum have suggested a close match between the 43 kDa N-terminal GyrB domain of Ec-gyrase containing the ATP-binding and hydrolysis site and the equivalent region of Pf-GyrB and between the 59 kDa N-terminal cleavage-reunion domain of Ec-GyrA and equivalent region of Pf-GyrA . However, to the best of our knowledge such alignments have not been carried out for the C-terminal regions of Pf-GyrA and GyrB which is where the majority of sequence divergence is located. Tg-gyrase proteins appear never to have been subject to structural prediction studies.
Summary of the protein sequences used in this study
Molecular weight (Da)
T. gondii GT1
T. gondii ME49
E. coli (ParC)
Molecular weight (Da)
[GenBank: NP_373243 ]
T. gondii GT1
T. gondii ME49
[GenBank: XP_002371198 ]
E. coli (ParE)
We show that the apicomplexan gyrases, as well as the obvious differences in size, contain signal and translocation peptides, which differ according to protein and species. Pf-gyrases contain numerous asparagines including repeat sequences. Perhaps the most significant differences between the proteins are found at the GyrA CTD, which varies significantly in size compared to the bacterial counterparts from which they show considerable divergence. We speculate that this divergence may include a DNA wrapping β-pinwheel domain with a different size/number of blades than has been seen to date in all known gyrases and/or an additional domain of unknown function. These differences may be related to modulation of supercoiling control in apicoplasts, which are not known to contain topoisomerase IV.
Results and discussion
Sequence analyses of GyrA
Bioinformatics analyses were carried out in order to gain deeper insights into the unique properties of Pf- and Tg-gyrases. Sequence alignments for Tg- and Pf- GyrAs and GyrBs all aligned best with topo IV structures. This likely reflects the lack of structural data for the relevant regions of gyrase proteins (i.e. full length GyrB and GyrA including the C-terminal DNA wrapping domains) rather than an indication that the structures actually more closely resemble E. coli topo IV.
Amino acid sequences of diverse gyrases were analysed using the MEGA analysis software . Examples of several bacterial GyrAs and GyrBs were compared to examples from Plasmodium species along with sequences from T. gondii and E. coli topoisomerase IV. Two analyses were carried out. Firstly pairwise distances between sequences were calculated (Additional file 1: Tables S1 and S2). These showed that for gyrases, the differences in amino acid sequences were greatest between the bacterial gyrases and the Plasmodium gyrases for both GyrA and GyrB. Topoisomerase IV, as would be expected, was also the most different in the case of ParE compared to GyrBs, but in fact less different than the Plasmodium GyrA proteins in the case of ParC. Tg-GyrA fell between bacterial and Plasmodium sequences in terms of difference and Tg-GyrB was similar to Plasmodium proteins. The maximum parsimony bootstrap consensus trees for the proteins show a similar result with two clear clusters, one for bacterial proteins and another for Plasmodium proteins with topoisomerase IV and T. gondii gyrases being somewhat variable outliers (See Additional file 1: Figure S1). Sequences were also aligned with ClustalW 2.1 , which showed that 25.6% of residues were identical between Ec-GyrA and Pf-GyrA. The value changes when analyses are delimited to individual domains identified by the Pfam server. Percentages of identical residues for the N-terminal domain (Ec-GyrA 32–507 encompassing WHD, tower, and coiled-coil domains) and the C-terminal domain (Ec-GyrA 538–840 encompassing the β-pinwheel domain) were 34.1% and 18.0%, respectively. These observations are rationalized by lower homology of the C-terminal domain of Pf-GyrA with respect to Ec-GyrA. Analysis of the amino acid sequences of Ec-GyrA using the Pfam domain prediction server results in six β-sheet blades of the β-pinwheel motif being predicted in accordance with reported experimental results . In contrast, only three such motifs are predicted from the Pf-GyrA sequence with the E-value set to 10, the maximum value. This prediction is consistent with secondary structures predicted by the SOPMA structure prediction server , where β-sheets are predicted for those regions of Pf-GyrA (Figure 1A). At face value, this suggests the presence of a small β-pinwheel domain consisting of a cluster of no more than three blades (Figure 1, Additional file 1: Figure S2). All known gyrases have 6 blades . Given that individual blades in Pf-GyrA appear to be approximately the same size as those found in other gyrases, it is difficult to envisage three blades alone to wrap DNA with a sufficiently acute angle to deliver it to the DNA clamp. Two possible explanations are that i) Pf-GyrA may be an example of a topo II with fewer than six β-pinwheel motifs (as is seen for E. coli topo IV ). ii) Pf-GyrA does in fact contain a β-pinwheel with a larger number of blades but they are of a sequence highly divergent from known structures and are thus not recognised by Pfam. For Tg-GyrA, the percentage of identical residues to Ec-GyrA is 33.5%. In terms of the distribution of identical residues, similarity can be observed to that seen between Ec-GyrA and Pf-GyrA, i.e. a higher identity is found in the N-terminal domain (41.5%) compared to the C-terminal domain (26.7%).
The N-terminal 59 kDa fragment of Ec-GyrA bears a domain similar to the DNA-binding domain of the catabolite-activator protein (CAP) containing a helix-turn-helix (HTH) motif and the tower domain . A hallmark of topoisomerase, the active site tyrosine that forms a transient phosphodiester bond with the end of the G-segment DNA is found in the HTH motif, which is part of the positively charged groove. A cluster of conserved residues comprises the active site of Ec-GyrA based on the crystal structure of the 59 kDa fragment. Those residues (Ec-GyrA Arg32, Lys42, Arg46, Arg47, Arg121, Tyr122) are perfectly conserved among Pf- and Tg- GyrAs, consistent with their role and confirming that the cleavage-reunion domain likely functions in an identical manner to that of other gyrases.
Next, the “GyrA-box” and its vicinity were considered. The GyrA-box is a positively charged motif (Sequence Q++GG + G, where + is any positively charged residue ) found in blade 1 of the β-pinwheel of the GyrA CTD and is known to be necessary for DNA supercoiling and wrapping ,. GyrA from species such as E. coli have a conserved proline at position 636 (Ec-GyrA numbering) close to the hinge between blades 1 and 2 and this was thought to introduce a tilt in the packing between blades, causing the pinwheel to be out of plane, giving positive handedness to the DNA wrap . Deletion of this proline in Ec-GyrA leads to a 2-3-fold decrease in supercoiling activity . In some other species lacking the proline, such as B. burdorferi, the CTD is planar . It is also planar in the case of 6-bladed Topo IV CTDs . However, in M. tuberculosis gyrase, there is no conserved proline and yet the CTD is tilted . These results suggest that other mechanisms in addition to/instead of the “conserved” proline are responsible for the non-planar shape.
Sequence analyses of GyrB
The overall percentage of identical residues between Ec-GyrB and Pf-GyrB was 38.5%. Unlike the case of Ec-GyrA and Pf-GyrA, the identity is spread more homogenously throughout the length of the proteins. The percentage of identical residues at the N-terminal domain (E. coli 31–391) is 35.7%, whereas for the C-terminal domain (E. coli 419–793) it is 40.7%. The percentage of identical residues between Ec-GyrB and Tg-GyrB is 35.1% (N-terminal domain: 35.6%, C-terminal domain: 32.9%).
Interestingly, Pf-GyrB includes an insertion of 45 amino acids in its TOPRIM domain which is essential for activity . This same insert is present in all Plasmodium species but is absent from the other 9 GyrB genes we considered. Our alignments also confirmed a 49-amino acid insert in part of the ATPase domain corresponding to the ATP lid of Ec-GyrB (Figure 2). We find that this insert, which has been noted previously  is unique to Plasmodium GyrBs amongst the proteins we investigated.
Proportions of asparagine residues in the coding sequences of gyrases A and B in four species of Plasmodium
Proportion of serine resides in gyrases of T. gondii GT1 with respect to the full-length sequence and the signal peptide
Full length protein
1 – 1594, 9.91%
1 – 258, 23.26%
1 – 1386, 20.06%
1 – 345, 40.58%
We also compared the signal and transit peptides of the proteins. A typical apicoplast-targeting protein contains an N-terminal signal peptide followed by the apicoplast transit peptide for protein import into the apicoplast . The amino acid sequence varies in transit peptides from different apicomplexans, for example, lysine is common in P. falciparum while arginine is common in T. gondii ,. Those basic residues near the N-terminus of the transit peptide are reported to be important for faithful transport into the apicoplast and an algorithm has been established to identify the signal and transit peptides in multiple Apicomplexa species . We analyzed N-terminal regions of Plasmodium and Toxoplasma gyrases that are expected to play key roles in translocation. Submission of Plasmodium and Toxoplasma gyrase sequences to the SOPMA  and Protscale  servers results in identification of a short hydrophobic α-helix predicted for each Plasmodium gyrase gene (Figure 3). This is consistent with previous work showing that the signal peptides of apicoplast-targeted proteins in P. falciparum comprise a hydrophobic region . In contrast, analyses of the Tg-GyrB sequence indicate different characteristics at the N-terminal region (Figure 3): Surprisingly for a protein that is predicted to be translocated to the apicoplast, the N-terminal signal peptide of Tg-GyrB is markedly non-hydrophobic which is in contrast to Tg-GyrA, Pf-GyrA and Pf-GyrB. This observation could imply an inefficient direction of Tg-GyrB across the endoplasmic reticulum in the first step of the secretory pathway.
Structure prediction and alignments
Pf-GyrA CTD structure prediction by I-TASSER differs from Pfam results in that residues identified by Pfam as part of the β-pinwheels in the CTD (residues 978–1121, shown in orange in Figure 4A) do not align with the equivalent structure in ParC but residues 710–968 do align with the 5-bladed ParC pinwheel. Given that the two predicted regions do not overlap, it may be that the actual structure consists of a large β-pinwheel consisting of both regions combined resulting in eight blades. This would be unusual as all gyrase β-pinwheels known have six blades. The equivalent region of topo IV in various organisms is more variable and includes both 3-bladed and 8-bladed pinwheels .
This inconsistency is attributable to a constraint to the structural homology calculations that derives from the fact that the ParC template structure features the pinwheel domain immediately C-terminal to the N-terminal domain. In order to create a homology model of the C-terminal domain of Pf-GyrA without such constraint, the sequence of the C-terminal region (671–1222) was submitted to I-TASSER on its own. This results in the pinwheel domain being predicted in a region different from when the fuller sequence is submitted (Figure 4B). Homology models based on the five best templates consistently features the β-pinwheel domains beginning in residues 875–877 and ending in residues 1220–1222 which covers all blades predicted by Pfam, and some blades predicted by I-TASSER using the mature sequence, to give a total of 6 blades. Also, the five best models consistently feature α-helices between residues 698–848. This is qualitatively similar to the prediction made by Dar et al.  in predicting a coiled-coil between the NTD and the CTD of GyrA, and hence legitimises the prediction of an additional structural element in Plasmodium GyrA compared to bacterial counterparts.
When the full length Tg-GyrA sequence (omitting the signal and transit sequences) was submitted to I-TASSER, the results as shown in Figure 4C were obtained. As with Pf-GyrA, the protein aligned best with the same E. coli ParC structure (pdb 1ZVU ). The rmsd was 1.91 Å. As was the case for the Pf-GyrA protein, there is good alignment with the cleavage reunion and N-terminal portion of ParC. In addition, of the residues identified as β-pinwheel blades by Pfam; (1052–1086, 1323–1371, 1405–1446 and 1523–1564) only the most N-terminal (1052–1086) overlaps with the β-pinwheel blades of ParC, the others being dispersed throughout the C-terminus in regions not aligned with any part of the ParC structure. Interestingly, the region predicted to be the most N-terminal blade is not recognized as any domain/motif by Pfam. The I-TASSER alignment places unaligned, largely unstructured sequences both N and C-terminal to the β-pinwheel. The C-terminal 500 residues beginning at residue A1096 of Tg-GyrA were not assigned as corresponding to any existing structure in ParC by I-TASSER. Similarly to Pf-GyrA, the sequence of the C-terminal region of Tg-GyrA (775–1594) was submitted to I-TASSER on its own (Figure 4D). The models based on the top four templates consistently predict α-helices in residue range of 788–945, followed by β-pinwheel motifs between 974–1505. The latter range only covers three out of four blades predicted by Pfam but, as for the CTD of Pf-GyrA, a total of 6 β-pinwheel blades is predicted in total. Clearly, the details of precisely which parts of the C-terminus are predicted to be a pinwheel depends on the software employed and the submitted amino acid sequence. Prediction of α-helices N-terminal to the β-pinwheel domain is similar to the case in Pf-GyrA, and suggests the presence of α-helices between the NTD and the CTD as being a common element among apicomplexan GyrAs.
Homology models obtained from I-TASSER reveal little evidence of gross structural deviation by Pf-GyrA and Tg-GyrA at the catalytic core and its surroundings. This is unsurprising given that the sequence identity is higher for the N-terminal domains in both GyrA and GyrB as described above. The most effective gyrase targeting antibacterials, the fluoroquinolone gyrase poisons, bind to a site made primarily of residues of the cleavage-reunion domains in GyrA and bound DNA . The absolute conservation of amino acid sequence of this region in Pf and Tg-GyrA means that the binding site for gyrase poisons is likely unchanged. In Tg-GyrA, the exact start residue of the mature protein is not fully established, thus the possibility of a relatively large structural element at the N-terminus is not completely ruled out.
Tg-GyrB was also submitted to I-TASSER and like Pf-GyrB aligns well with the same S. pneumoniae topo IV structure 4I3H  with an rmsd of 1.14 Å. Due to its extra length, Tg-GyrB contains a greater number of longer, unassigned loops. These loop regions are clustered around the GHKL region (Figure 5B). Overall, in both Pf- and Tg-GyrBs core domains seem to be largely preserved and this includes the regions involved in DNA binding (Figure 5).
In summary, we have carried out a comprehensive sequence and structural comparison of two apicomplexan gyrases to standard bacterial (E. coli) gyrase and have found a number of differences which may have both relevance to enzyme function and consequences in terms of effectiveness of therapeutics as well as providing useful data for future structural studies.
While the results presented in this work are not definitive there are sufficient data to enable some tentative speculation, which should be considered in the light of recent findings from M. tuberculosis: This organism also has a gyrase but lacks topo IV. M. tuberculosis gyrase has a greater decatenation ability relative to negative supercoiling ability compared to Ec-gyrase ,, something which is thought to be due to decreased DNA-stimulated ATPase activity and a truncated “tail” in the CTD which combine to inhibit the enzyme from achieving as high a level of DNA duplex under-winding as the E. coli enzyme . Further research has shown that this gyrase has a calcium binding site and that that gyrase in M. tuberculosis can shift between gyrase-like and topo IV-like activities via the modulatory action of Ca2+ which binds in the linker region between the NTD and CTD of GyrA . This is proposed to alter the position of the β-pinwheels relative to the remainder of the enzyme such that supercoiling may be favored in the absence of Ca2+ while disfavored in its presence, allowing relaxation/decantation to occur. The fact that the greatest variation in Pf- and Tg-gyrases compared to Ec-gyrase is in the CTD strongly suggests that this is due to a role in affecting or modulating the balance between gyrase-like and topo IV-like activities. Tg-gyrase, in contrast to Pf-gyrase, has a likely GyrA box but lacks an insert in the ATP lid meaning that these mooted activity modifiers are not available to it. In this instance, compensation for the lack of topo IV is presumably achieved at least in part through its unusually large GyrA CTD, which may affect extent and/or position of the DNA wrap. It is also the case that additional structural elements found in Plasmodium and Toxoplasma gyrases may be responsible for modulating their activity by means not found in prokaryotes. These may pose an additional target for drugs in the future. For example, we may envisage that inhibiting the apicoplast targeting of gyrases will be lethal to Apicomplexa.
Further biochemical studies on individual gyrase proteins from these organisms as well as holoenzymes in conjunction with high-resolution studies will be required to answer outstanding questions and assess their suitability as targets for development of therapeutics.
Amino acid sequences of proteins were aligned with ClustalW 2.1 using the default parameters . The percentages of sequence identities were calculated by dividing the number of identical residues by the total aligned sequence length without insertions. Aligned sequences were viewed using CLC Sequence Viewer (CLC bio). Aligned amino acid sequences of gyrase and topoisomerase IV proteins from various species were inputted into MEGA (v. 6.06 beta) . Pairwise distances between sequences were computed (the number of amino acid differences divided by the total number of amino acids compared). Gaps and missing data were deleted and no variance estimation method was used. A maximum parsimony tree was calculated using the Jones-Taylor-Thornton model assuming uniform rates among sites. Phylogeny was tested using the bootstrap method with 500 bootstrap replications.
Structure predictions were carried out using I-TASSER ,. In all cases, amino acid sequences for the proteins, lacking the predicted signal and transit sequences were submitted to the I-TASSER server. Default settings were used with no additional restraints employed. From the top 10 identified structural analogues returned the best alignment was chosen based on TM-score ranking. The top ranking alignment templates were E. coli ParC  for Pf-GyrA, Pf-GyrA CTD (residues 671–1222) and Tg-GyrA, S. pneumoniae ParC-ParE55 fusion protein  for Pf-GyrB and Tg-GyrB and X. campestris GyrA CTD for Tg-GyrA CTD (residues 775–1594). PDB coordinates were visualised using PyMOL . All of the proteins considered were significantly longer than the homologous regions in the model templates. This has an important consequence for the homology models generated. When additional sequences are present then clearly not all sequences can be aligned to a template. In some cases we observe that where sequences flanking these loop regions themselves have poor similarity to the template, then “loop swapping” may occur where the flanking sequence is in fact designated a loop and the loop is assigned structure. This does occur in parts of our structure but has little effect on the overall model.
Dimeric models of homology models of the proteins as generated by I-TASSER, were created by superposing each homology model onto dimeric crystal structures of the template topoisomerases using Coot . The SSM Superpose function was used, and root mean square deviation (RMSD) of the distances between alpha carbons of the aligned residues were calculated. This allowed us to see regions of the Pf or Tg protein, which matched well with a known gyrase protein structure and, more importantly those regions where poor homology was predicted.
Availability of supporting data
All supporting data associated with this manuscript are included as additional files.
The authors are grateful to Kam Zhang for critical reading of the manuscript.
This work was supported by RIKEN Initiative Research Funding awarded to JGH.
- Lim L, McFadden GI: The evolution, metabolism and functions of the apicoplast. Phil Trans R Soc B. 2010, 365 (1541): 749-763. 10.1098/rstb.2009.0273.View ArticlePubMed CentralPubMedGoogle Scholar
- van Dooren GG, Striepen B: The algal past and parasite present of the apicoplast. Annu Rev Microbiol 2013, 67(1):null.Google Scholar
- WHO: World Malaria Report. Geneva: World Health Organization Press; 2012.Google Scholar
- Centers for Disease Control and Prevention [http://www.cdc.gov/parasites/npi.html]
- Petersen I, Eastman R, Lanzer M: Drug-resistant malaria: molecular mechanisms and implications for public health. FEBS Lett. 2011, 585 (11): 1551-1562. 10.1016/j.febslet.2011.04.042.View ArticlePubMedGoogle Scholar
- Dondorp AM, Nosten F, Yi P, Das D, Phyo AP, Tarning J, Lwin KM, Ariey F, Hanpithakpong W, Lee SJ, Ringwald P, Silamut K, Imwong M, Chotivanich K, Lim P, Herdman T, An SS, Yeung S, Singhasivanon P, Day NP, Lindegardh N, Socheat D, White NJ: Artemisinin resistance in Plasmodium falciparum malaria. New Engl J Med. 2009, 361 (5): 455-467. 10.1056/NEJMoa0808859.View ArticlePubMed CentralPubMedGoogle Scholar
- Anderson TJ, Nair S, Nkhoma S, Williams JT, Imwong M, Yi P, Socheat D, Das D, Chotivanich K, Day NP, White NJ, Dondorp AM: High heritability of malaria parasite clearance rate indicates a genetic basis for artemisinin resistance in western Cambodia. J Infect Dis. 2010, 201 (9): 1326-1330. 10.1086/651562.View ArticlePubMed CentralPubMedGoogle Scholar
- The RTS, S Clinical Trials Partnership.: A phase 3 trial of RTS,S/AS01 malaria vaccine in African infants. New Engl J Med 2012, 367(24):2284–2295.Google Scholar
- Seder RA, Chang L-J, Enama ME, Zephir KL, Sarwar UN, Gordon IJ, Holman LA, James ER, Billingsley PF, Gunasekera A, Richman A, Chakravarty S, Manoj A, Velmurugan S, Li ML, Ruben AJ, Li T, Eappen AG, Stafford RE, Plummer SH, Hendel CS, Novik L, Costner PJM, Mendoza FH, Saunders JG, Nason MC, Richardson JH, Murphy J, Davidson SA, Richie TL, et al: Protection against malaria by intravenous immunization with a nonreplicating sporozoite vaccine. Science 2013. 341:1359-1365,Google Scholar
- Sims PG: Drug Resistance in Toxoplasma gondii. In: Antimicrobial Drug Resistance. Edited by Mayers D: Humana Press; New York: 2009: 1121–1126.Google Scholar
- Fichera ME, Roos DS: A plastid organelle as a drug target in apicomplexan parasites. Nature. 1997, 390 (6658): 407-409. 10.1038/37132.View ArticlePubMedGoogle Scholar
- He CY, Shaw MK, Pletcher CH, Striepen B, Tilney LG, Roos DS: A plastid segregation defect in the protozoan parasite Toxoplasma gondii. Embo J. 2001, 20 (3): 330-339. 10.1093/emboj/20.3.330.View ArticlePubMed CentralPubMedGoogle Scholar
- Yeh E, DeRisi JL: Chemical rescue of malaria parasites lacking an apicoplast defines organelle function in blood-stage Plasmodium falciparum . PLoS Biol. 2011, 9 (8): e1001138-10.1371/journal.pbio.1001138.View ArticlePubMed CentralPubMedGoogle Scholar
- Pradel G, Schlitzer M: Antibiotics in malaria therapy and their effect on the parasite apicoplast. Curr Mol Med. 2010, 10 (3): 335-349. 10.2174/156652410791065273.View ArticlePubMedGoogle Scholar
- Collin F, Karkare S, Maxwell A: Exploiting bacterial DNA gyrase as a drug target: current state and perspectives. Appl Microbiol Biotechnol. 2011, 92 (3): 479-497. 10.1007/s00253-011-3557-z.View ArticlePubMed CentralPubMedGoogle Scholar
- Nöllmann M, Crisona NJ, Arimondo PB: Thirty years of escherichia coli DNA gyrase: from in vivo function to single-molecule mechanism. Biochimie. 2007, 89 (4): 490-499. 10.1016/j.biochi.2007.02.012.View ArticlePubMedGoogle Scholar
- Corbett KD, Shultzaberger RK, Berger JM: The C-terminal domain of DNA gyrase A adopts a DNA-bending ß-pinwheel fold. Proc Natl Acad Sci U S A. 2004, 101 (19): 7293-7298. 10.1073/pnas.0401595101.View ArticlePubMed CentralPubMedGoogle Scholar
- Chaudhuri I, Soding J, Lupas AN: Evolution of the beta-propeller fold. Proteins. 2008, 71 (2): 795-803. 10.1002/prot.21764.View ArticlePubMedGoogle Scholar
- Aravind L, Leipe DD, Koonin EV: Toprim - a conserved catalytic domain in type IA and II topoisomerases, DnaG-type primases, OLD family nucleases and RecR proteins. Nucl Acids Res. 1998, 26 (18): 4205-4213. 10.1093/nar/26.18.4205.View ArticlePubMed CentralPubMedGoogle Scholar
- Punta M, Coggill PC, Eberhardt RY, Mistry J, Tate J, Boursnell C, Pang N, Forslund K, Ceric G, Clements J, Heger A, Holm L, Sonnhammer EL, Eddy SR, Bateman A, Finn RD: The Pfam protein families database. Nucl Acids Res 2012, 40(Database issue):D290–D301.,Google Scholar
- Geourjon C, Deleage G: SOPMA: significant improvements in protein secondary structure prediction by consensus prediction from multiple alignments. Comput Appl Biosci. 1995, 11 (6): 681-684.PubMedGoogle Scholar
- Schoeffler AJ, Berger JM: Recent advances in understanding structure-function relationships in the type II topoisomerase mechanism. Biochem Soc Trans. 2005, 33 (Pt 6): 1465-1470.View ArticlePubMedGoogle Scholar
- Wall MK, Mitchenall LA, Maxwell A: Arabidopsis thaliana DNA gyrase is targeted to chloroplasts and mitochondria. Proc Natl Acad Sci USA. 2004, 101 (20): 7821-7826. 10.1073/pnas.0400836101.View ArticlePubMed CentralPubMedGoogle Scholar
- Divo AA, Sartorelli AC, Patton CL, Bia FJ: Activity of fluoroquinolone antibiotics against Plasmodium falciparum in vitro. Antimicrob Agents Chemother. 1988, 32 (8): 1182-1186. 10.1128/AAC.32.8.1182.View ArticlePubMed CentralPubMedGoogle Scholar
- Weissig V, Vetro-Widenhouse TS, Rowe TC: Topoisomerase II inhibitors induce cleavage of nuclear and 35-kb plastid DNAs in the malarial parasite Plasmodium falciparum. DNA Cell Biol. 1997, 16 (12): 1483-1492. 10.1089/dna.1997.16.1483.View ArticlePubMedGoogle Scholar
- Krishna S, Davis TM, Chan PC, Wells RA, Robson KJ: Ciprofloxacin and malaria. Lancet. 1988, 1 (8596): 1231-1232. 10.1016/S0140-6736(88)92056-9.View ArticlePubMedGoogle Scholar
- Khan AA, Slifer T, Araujo FG, Remington JS: Trovafloxacin is active against Toxoplasma gondii. Antimicrob Agents Chemother. 1996, 40 (8): 1855-1859.PubMed CentralPubMedGoogle Scholar
- Gardner MJ, Hall N, Fung E, White O, Berriman M, Hyman RW, Carlton JM, Pain A, Nelson KE, Bowman S, Paulsen IT, James K, Eisen JA, Rutherford K, Salzberg SL, Craig A, Kyes S, Chan MS, Nene V, Shallom SJ, Suh B, Peterson J, Angiuoli S, Pertea M, Allen J, Selengut J, Haft D, Mather MW, Vaidya AB, Martin DM, et al: Genome sequence of the human malaria parasite Plasmodium falciparum. Nature. 2002, 419 (6906): 498-511. 10.1038/nature01097.View ArticlePubMedGoogle Scholar
- García-Estrada C, Prada CF, Fernández-Rubio C: Rojo–Vázquez F, Balaña-Fouce R: DNA topoisomerases in apicomplexan parasites: promising targets for drug discovery. P Roy Soc B-Biol Sci. 2010, 277 (1689): 1777-1787. 10.1098/rspb.2009.2176.View ArticleGoogle Scholar
- Gozalbes R, Brun-Pascaud M, Garcia-Domenech R, Galvez J, Girard PM, Doucet JP, Derouin F: Anti-toxoplasma activities of 24 quinolones and fluoroquinolones in vitro: prediction of activity by molecular topology and virtual computational techniques. Antimicrob Agents Chemother. 2000, 44 (10): 2771-2776. 10.1128/AAC.44.10.2771-2776.2000.View ArticlePubMed CentralPubMedGoogle Scholar
- Mahmoudi N, Ciceron L, Franetich JF, Farhati K, Silvie O, Eling W, Sauerwein R, Danis M, Mazier D, Derouin F: In vitro activities of 25 quinolones and fluoroquinolones against liver and blood stage Plasmodium spp. Antimicrob Agents Chemother. 2003, 47 (8): 2636-2639. 10.1128/AAC.47.8.2636-2639.2003.View ArticlePubMed CentralPubMedGoogle Scholar
- Piddock LJ, Zhu M: Mechanism of action of sparfloxacin against and mechanism of resistance in gram-negative and gram-positive bacteria. Antimicrob Agents Chemother. 1991, 35 (11): 2423-2427. 10.1128/AAC.35.11.2423.View ArticlePubMed CentralPubMedGoogle Scholar
- Goodman CD, Su V, McFadden GI: The effects of anti-bacterials on the malaria parasite Plasmodium falciparum. Mol Biochem Parasitol. 2007, 152 (2): 181-191. 10.1016/j.molbiopara.2007.01.005.View ArticlePubMedGoogle Scholar
- Raghu Ram EV, Kumar A, Biswas S, Chaubey S, Siddiqi MI, Habib S: Nuclear gyrB encodes a functional subunit of the Plasmodium falciparum gyrase that is involved in apicoplast DNA replication. Mol Biochem Parasitol. 2007, 154 (1): 30-39. 10.1016/j.molbiopara.2007.04.001.View ArticlePubMedGoogle Scholar
- Dar MA, Sharma A, Mondal N, Dhar SK: Molecular cloning of apicoplast-targeted plasmodium falciparum DNA gyrase genes: unique intrinsic ATPase activity and ATP-independent dimerization of PfGyrB subunit. Eukaryot Cell. 2007, 6 (3): 398-412. 10.1128/EC.00357-06.View ArticlePubMed CentralPubMedGoogle Scholar
- Ali JA, Jackson AP, Howells AJ, Maxwell A: The 43-kilodalton N-terminal fragment of the DNA gyrase B protein hydrolyzes ATP and binds coumarin drugs. Biochemistry. 1993, 32 (10): 2717-2724. 10.1021/bi00061a033.View ArticlePubMedGoogle Scholar
- Dar A, Prusty D, Mondal N, Dhar SK: A unique 45-amino-acid region in the toprim domain of plasmodium falciparum gyrase B is essential for its activity. Eukaryot Cell. 2009, 8 (11): 1759-1769. 10.1128/EC.00149-09.View ArticlePubMed CentralPubMedGoogle Scholar
- Tamura K, Stecher G, Peterson D, Filipski A, Kumar S: MEGA6: Molecular Evolutionary Genetics Analysis version 6.0. Mol Biol Evol. 2013, 30 (12): 2725-2729. 10.1093/molbev/mst197.View ArticlePubMed CentralPubMedGoogle Scholar
- Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA, McWilliam H, Valentin F, Wallace IM, Wilm A, Lopez R, Thompson JD, Gibson TJ, Higgins DG: Clustal W and Clustal X version 2.0. Bioinformatics. 2007, 23 (21): 2947-2948. 10.1093/bioinformatics/btm404.View ArticlePubMedGoogle Scholar
- Thompson JD, Higgins DG, Gibson TJ: CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucl Acids Res. 1994, 22 (22): 4673-4680. 10.1093/nar/22.22.4673.View ArticlePubMed CentralPubMedGoogle Scholar
- Ruthenburg AJ, Graybosch DM, Huetsch JC, Verdine GL: A superhelical spiral in the Escherichia coli DNA gyrase A C-terminal domain imparts unidirectional supercoiling bias. J Biol Chem. 2005, 280 (28): 26177-26184. 10.1074/jbc.M502838200.View ArticlePubMedGoogle Scholar
- Corbett KD, Schoeffler AJ, Thomsen ND, Berger JM: The structural basis for substrate specificity in DNA topoisomerase IV. J Mol Biol. 2005, 351 (3): 545-561. 10.1016/j.jmb.2005.06.029.View ArticlePubMedGoogle Scholar
- Morais Cabral JH, Jackson AP, Smith CV, Shikotra N, Maxwell A, Liddington RC: Crystal structure of the breakage-reunion domain of DNA gyrase. Nature. 1997, 388 (6645): 903-906. 10.1038/42294.View ArticlePubMedGoogle Scholar
- Kramlinger VM, Hiasa H: The “GyrA-box” is required for the ability of DNA gyrase to wrap DNA and catalyze the supercoiling reaction. J Biol Chem. 2006, 281 (6): 3738-3742. 10.1074/jbc.M511160200.View ArticlePubMedGoogle Scholar
- Ward D, Newton A: Requirement of topoisomerase IV parC and parE genes for cell cycle progression and developmental regulation in Caulobacter crescentus. Mol Microbiol. 1997, 26 (5): 897-910. 10.1046/j.1365-2958.1997.6242005.x.View ArticlePubMedGoogle Scholar
- Hsieh TJ, Yen TJ, Lin TS, Chang HT, Huang SY, Hsu CH, Farh L, Chan NL: Twisting of the DNA-binding surface by a beta-strand-bearing proline modulates DNA gyrase activity. Nucleic Acids Res. 2010, 38 (12): 4173-4181. 10.1093/nar/gkq153.View ArticlePubMed CentralPubMedGoogle Scholar
- Hsieh TJ, Farh L, Huang WM, Chan NL: Structure of the topoisomerase IV C-terminal domain: a broken beta-propeller implies a role as geometry facilitator in catalysis. J Biol Chem. 2004, 279 (53): 55587-55593. 10.1074/jbc.M408934200.View ArticlePubMedGoogle Scholar
- Tretter EM, Berger JM: Mechanisms for defining supercoiling Set point of DNA gyrase orthologs: II. The shape of the GyrA subunit C-terminal domain (Ctd) is Not a sole determinant for controlling supercoiling efficiency. J Biol Chem. 2012, 287 (22): 18645-18654. 10.1074/jbc.M112.345736.View ArticlePubMed CentralPubMedGoogle Scholar
- Holdgate GA, Tunnicliffe A, Ward WH, Weston SA, Rosenbrock G, Barth PT, Taylor IW, Pauptit RA, Timms D: The entropic penalty of ordered water accounts for weaker binding of the antibiotic novobiocin to a resistant mutant of DNA gyrase: a thermodynamic and crystallographic study. Biochemistry. 1997, 36 (32): 9663-9673. 10.1021/bi970294+.View ArticlePubMedGoogle Scholar
- Aravind L, Iyer LM, Wellems TE, Miller LH: Plasmodium biology: genomic gleanings. Cell. 2003, 115 (7): 771-785. 10.1016/S0092-8674(03)01023-7.View ArticlePubMedGoogle Scholar
- Muralidharan V, Goldberg DE: Asparagine repeats in plasmodium falciparum proteins: good for nothing?. PLoS Pathog. 2013, 9 (8): e1003488-10.1371/journal.ppat.1003488.View ArticlePubMed CentralPubMedGoogle Scholar
- Verra F, Hughes AL: Biased amino acid composition in repeat regions of Plasmodium antigens. Mol Biol Evol. 1999, 16 (5): 627-633. 10.1093/oxfordjournals.molbev.a026145.View ArticlePubMedGoogle Scholar
- Hughes AL: The evolution of amino acid repeat arrays in Plasmodium and other organisms. J Mol Evol. 2004, 59 (4): 528-535. 10.1007/s00239-004-2645-4.View ArticlePubMedGoogle Scholar
- Dutta R, Inouye M: GHKL, an emergent ATPase/kinase superfamily. Trends Biochem Sci. 2000, 25 (1): 24-28. 10.1016/S0968-0004(99)01503-0.View ArticlePubMedGoogle Scholar
- Waller RF, Keeling PJ, Donald RG, Striepen B, Handman E, Lang-Unnasch N, Cowman AF, Besra GS, Roos DS, McFadden GI: Nuclear-encoded proteins target to the plastid in Toxoplasma gondii and Plasmodium falciparum. Proc Natl Acad Sci U S A. 1998, 95 (21): 12352-12357. 10.1073/pnas.95.21.12352.View ArticlePubMed CentralPubMedGoogle Scholar
- Parsons M, Karnataki A, Feagin JE, DeRocher A: Protein trafficking to the apicoplast: deciphering the apicomplexan solution to secondary endosymbiosis. Eukaryot Cell. 2007, 6 (7): 1081-1088. 10.1128/EC.00102-07.View ArticlePubMed CentralPubMedGoogle Scholar
- Ralph SA, Foth BJ, Hall N, McFadden GI: Evolutionary pressures on apicoplast transit peptides. Mol Biol Evol. 2004, 21 (12): 2183-2194. 10.1093/molbev/msh233.View ArticlePubMedGoogle Scholar
- Cilingir G, Broschat SL, Lau AO: ApicoAP: the first computational model for identifying apicoplast-targeted proteins in multiple species of Apicomplexa. PLOS ONE. 2012, 7 (5): e36598-10.1371/journal.pone.0036598.View ArticlePubMed CentralPubMedGoogle Scholar
- Gasteiger E, Hoogland C, Gattiker A, Duvaud Se, Wilkins M, Appel R, Bairoch A: Protein Identification and Analysis Tools on the ExPASy Server. In: The Proteomics Protocols Handbook. Edited by Walker J: Humana Press; New York: 2005: 571–607.Google Scholar
- Foth BJ, Ralph SA, Tonkin CJ, Struck NS, Fraunholz M, Roos DS, Cowman AF, McFadden GI: Dissecting apicoplast targeting in the malaria parasite Plasmodium falciparum. Science. 2003, 299 (5607): 705-708. 10.1126/science.1078599.View ArticlePubMedGoogle Scholar
- Zhang Y: I-TASSER server for protein 3D structure prediction. BMC Bioinformatics. 2008, 9 (1): 40-10.1186/1471-2105-9-40.View ArticlePubMed CentralPubMedGoogle Scholar
- Roy A, Kucukural A, Zhang Y: I-TASSER: a unified platform for automated protein structure and function prediction. Nat Protocols. 2010, 5 (4): 725-738. 10.1038/nprot.2010.5.View ArticleGoogle Scholar
- Laponogov I, Veselkov DA, Crevel IM-T, Pan X-S, Fisher LM, Sanderson MR: Structure of an ‘open’ clamp type II topoisomerase-DNA complex provides a mechanism for DNA capture and transport. Nucl Acids Res. 2013, 41 (21): 9911-9923. 10.1093/nar/gkt749.View ArticlePubMed CentralPubMedGoogle Scholar
- Khor V, Yowell C, Dame JB, Rowe TC: Expression and characterization of the ATP-binding domain of a malarial Plasmodium vivax gene homologous to the B-subunit of the bacterial topoisomerase DNA gyrase. Mol Biochem Parasitol. 2005, 140 (1): 107-117. 10.1016/j.molbiopara.2004.12.013.View ArticlePubMedGoogle Scholar
- Ostrov DA, Hernandez Prada JA, Corsino PE, Finton KA, Le N, Rowe TC: Discovery of novel DNA gyrase inhibitors by high-throughput virtual screening. Antimicrob Agents Chemother. 2007, 51 (10): 3688-3698. 10.1128/AAC.00392-07.View ArticlePubMed CentralPubMedGoogle Scholar
- Yoshida H, Bogaki M, Nakamura M, Nakamura S: Quinolone resistance-determining region in the DNA gyrase gyrA gene of Escherichia coli. Antimicrob Agents Chemother. 1990, 34 (6): 1271-1272. 10.1128/AAC.34.6.1271.View ArticlePubMed CentralPubMedGoogle Scholar
- Yoshida H, Bogaki M, Nakamura M, Yamanaka LM, Nakamura S: Quinolone resistance-determining region in the DNA gyrase gyrB gene of Escherichia coli. Antimicrob Agents Chemother. 1991, 35 (8): 1647-1650. 10.1128/AAC.35.8.1647.View ArticlePubMed CentralPubMedGoogle Scholar
- Aubry A, Fisher LM, Jarlier V, Cambau E: First functional characterization of a singly expressed bacterial type II topoisomerase: the enzyme from Mycobacterium tuberculosis. Biochem Biophys Res Comm. 2006, 348 (1): 158-165. 10.1016/j.bbrc.2006.07.017.View ArticlePubMedGoogle Scholar
- Manjunatha UH, Dalal M, Chatterji M, Radha DR, Visweswariah SS, Nagaraja V: Functional characterisation of mycobacterial DNA gyrase: an efficient decatenase. Nucleic Acids Res. 2002, 30 (10): 2144-2153. 10.1093/nar/30.10.2144.View ArticlePubMed CentralPubMedGoogle Scholar
- Karkare S, Yousafzai F, Mitchenall LA, Maxwell A: The role of Ca2+ in the activity of Mycobacterium tuberculosis DNA gyrase. Nucleic Acids Res. 2012, 40 (19): 9774-9787. 10.1093/nar/gks704.View ArticlePubMed CentralPubMedGoogle Scholar
- Schrodinger, LLC: The PyMOL Molecular Graphics System, Version 1.3r1. In.; 2010.Google Scholar
- Emsley P, Lohkamp B, Scott WG, Cowtan K: Features and development of Coot. Acta Crystallogr D. 2010, 66 (Pt 4): 486-501. 10.1107/S0907444910007493.View ArticlePubMed CentralPubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.