Secondary Mutations Correct Fitness Defects in

0 downloads 0 Views 3MB Size Report
mutations associated with oryzalin resistance in T. gondii. (Morrissette et al. ...... high-affinity binding of the dinitroaniline herbicide oryzalin to tubulin from Zea ...
Copyright Ó 2008 by the Genetics Society of America DOI: 10.1534/genetics.108.092494

Secondary Mutations Correct Fitness Defects in Toxoplasma gondii With Dinitroaniline Resistance Mutations Christopher Ma, Johnson Tran, Catherine Li, Lakshmi Ganesan, David Wood and Naomi Morrissette1 Department of Molecular Biology and Biochemistry, University of California, Irvine, California 92697 Manuscript received June 10, 2008 Accepted for publication July 21, 2008 ABSTRACT Dinitroanilines (oryzalin, trifluralin, ethafluralin) disrupt microtubules in protozoa but not in vertebrate cells, causing selective death of intracellular Toxoplasma gondii parasites without affecting host cells. Parasites containing a1-tubulin point mutations are dinitroaniline resistant but show increased rates of aberrant replication relative to wild-type parasites. T. gondii parasites bearing the F52Y mutation were previously demonstrated to spontaneously acquire two intragenic mutations that decrease both resistance levels and replication defects. Parasites bearing the G142S mutation are largely dependent on oryzalin for viable growth in culture. We isolated 46 T. gondii lines that have suppressed microtubule defects associated with the G142S or the F52Y mutations by acquiring secondary mutations. These compensatory mutations were a1-tubulin pseudorevertants or extragenic suppressors (the majority alter the b1-tubulin gene). Many secondary mutations were located in tubulin domains that suggest that they function by destabilizing microtubules. Most strikingly, we identified seven novel mutations that localize to an eight-amino-acid insert that stabilizes the a1-tubulin M loop, including one (P364R) that acts as a compensatory mutation in both F52Y and G142S lines. These lines have reduced dinitroaniline resistance but most perform better than parental lines in competition assays, indicating that there is a trade-off between resistance and replication fitness.

OXOPLASMA gondii is a human pathogen grouped within the Apicomplexa, a phylum of protozoa that comprises all obligate intracellular parasites (Levene 1988). Infection with this parasite can be life threatening in immunocompromised individuals and cause birth defects or miscarriage during fetal infection (Black and Boothroyd 2000). Other apicomplexans such as Plasmodium spp. and Cryptosporidium spp. have considerable medical importance or are animal pathogens that affect livestock. Apicomplexans have a specialized apex that contains unique organelles that coordinate invasion of host cells (Morrissette and Sibley 2002a). These parasites are delimited by a pellicle, a composite structure formed by the association of the plasma membrane with the inner membrane complex, an assemblage of flattened vesicles. The invasive stages of apicomplexan parasites have two microtubule populations: subpellicular microtubules and spindle microtubules. Subpellicular microtubules are nondynamic; they maintain both apical polarity and the characteristic crescent shape of the parasite by interacting with the cytoplasmic face of the pellicle. Spindle microtubules form an intranuclear spindle to coordinate chromosome segregation. Both populations are critically important to parasite survival and replication.

T

1 Corresponding author: Department of Molecular Biology and Biochemistry, University of California, Irvine, CA 92697-3900. E-mail: [email protected]

Genetics 180: 845–856 (October 2008)

Microtubules are built by the polymerization of a-b-tubulin heterodimers and typically contain 13 protofilaments. Protofilaments are formed by longitudinal head-to-tail association of heterodimers and laterally associate to assemble the microtubule lattice (Downing and Nogales 1998). Biochemical and structural studies have elucidated the roles of tubulin domains, which are important to the regulation of microtubule polymerization and depolymerization. Both a- and b-tubulins have H1-S2 (N) and M loops, which coordinate contacts between adjacent protofilaments (Nogales et al. 1997, 1999; Downing and Nogales 1998; Lowe et al. 2001; Li et al. 2002). a-Tubulin contains an eight-amino-acid insert missing from the analogous region of b-tubulin, which stabilizes the a-tubulin M loop to promote protofilament lateral association. Both a- and b-tubulins bind to GTP; however, only b-tubulin GTP can be exchanged or hydrolyzed while a-tubulin GTP plays a structural role in this subunit (David-Pfeuty et al. 1977; Sage et al. 1995a,b; Nogales et al. 1998; Anders and Botstein 2001). The b-tubulin subunits are stimulated to hydrolyze GTP in a polymerization-dependent fashion by association of a GTPase-activating domain from the a-tubulin subunit of the adjacent dimer within the protofilament. Structural evidence indicates that GTPbound dimers have a conformation that facilitates protofilament contacts by both a- and b-subunits. After hydrolysis, GDP-tubulin dimers have a kinked confor-

846

C. Ma et al.

mation that is believed to weaken lateral contacts to promote microtubule disassembly (Nogales et al. 2003; Wang et al. 2005; Wang and Nogales 2005; Nogales and Wang 2006a,b). Protozoan parasites are sensitive to dinitroaniline compounds, which disrupt microtubules to inhibit replication (Stokkermans et al. 1996; Bogitsh et al. 1999; Makioka et al. 2000; Traub-Cseko et al. 2001; Bhattacharya et al. 2002). Given that these compounds are inactive against vertebrate or fungal tubulins, the mechanism of dinitroaniline action is of particular interest for development of new antimicrobial therapies. Our assays, as well as data from others, indicate that wild-type T. gondii parasites have an IC50 value of 0.25 mm for the dinitroaniline oryzalin and display aberrant phenotypes in culture at 0.5 mm oryzalin (Stokkermans et al. 1996; Morrissette and Sibley 2002b). Although extracellular parasites are refractory to the effects of microtubule-disrupting drugs, during intracellular growth, parasite microtubules are dynamic and sensitive to disruption by submicromolar concentrations of oryzalin or other dinitroanilines. Dinitroaniline selectivity is associated with restricted binding to sensitive (plants and protozoa) but not to resistant (vertebrates and fungi) tubulin (Hess and Bayer 1977; Morejohn et al. 1987; Chan and Fong 1990; Hugdahl and Morejohn 1993). Studies using molecular dynamics analysis and compound docking indicate that dinitroanilines interact with a consistent site on multiple conformations of a-tubulin derived from independent molecular dynamics models of T. gondii, Leishmania spp., and Plasmodium spp. tubulins (Morrissette et al. 2004; Mitra and Sept 2006). Parallel analysis of vertebrate (bovine) a-tubulin reveals that dinitroanilines have nonspecific, low-affinity interactions and no consensus binding site, consistent with in vivo and in vitro observations that these compounds do not bind to vertebrate tubulin or disrupt vertebrate microtubules. This work predicts a binding site on parasite a-tubulin that is located beneath the H1-S2 (N) loop. Since this loop participates in protofilament interactions, the binding site location suggests that dinitroanilines disrupt lateral contacts in the microtubule lattice. Consistent with this notion, a comparison of a-tubulin molecular dynamics simulations in the presence and absence of a bound drug suggests that dinitroaniline binding profoundly limits flexibility of the a-tubulin H1-S2 loop, which is drawn in toward the core of the tubulin dimer (Mitra and Sept 2006). T. gondii is a haploid organism for most of its life cycle, although it is capable of sexual recombination during a transient diploid zygote stage. There are three a- and three b-tubulin genes in the genome: the a2 gene appears to be gamete stage specific and the divergent a3 gene may be used to construct the conoid, an unusual structure built of tubulin sheets rather than microtubules (Hu et al. 2002). The a1- and b1-tubulin genes

are the dominantly expressed forms in the proliferative (tachyzoite) stage of the parasite studied here. We have previously identified a number of a1-tubulin point mutations associated with oryzalin resistance in T. gondii (Morrissette et al. 2004; Ma et al. 2007). The mutations are distributed throughout the linear sequence of a1tubulin. However, when mapped onto a model of T. gondii a1-tubulin based on the structure of the vertebrate tubulin dimer, many of the mutations cluster in specific domains of the protein. For example, many mutations are located in or near domains that are associated with microtubule stability or in the core of a-tubulin, adjacent to or within the dinitroanilinebinding site identified by computational methods. Our working model suggests that most tubulin mutations confer dinitroaniline resistance by one of two possible mechanisms: (1) by increasing subunit affinity within the microtubule to compensate for the action of a bound drug and (2) by reducing the affinity of the tubulin for dinitroanilines by changes to the compound binding site. Our previous observations that diverse mutations to a1-tubulin are sufficient for conferring resistance to dinitroanilines such as oryzalin in the protozoan parasite T. gondii might suggest that dinitroaniline-binding site ligands are not appropriate for development of antiparasitic agents. However, in many cases, drug resistance mutations isolated under laboratory conditions are not observed in clinical settings since parasites are incapable of robust growth in natural reservoirs. For example, a subset of laboratory-selected dihydrofolate reductase (DHFR) mutations that confer pyrimethamine resistance in Plasmodium falciparum are either rare or not observed in the field (Hankins et al. 2001; Hastings et al. 2002; Bates et al. 2004). Moreover, studies that have exploited growth competition assays to assess the fitness of T. gondii strains bearing wild-type or pyrimethamine-resistant DHFR genes have concluded that even mutant strains that behave similarly to wild-type parasites in vitro can display growth defects in vivo (Fohl and Roos 2003). We show here that dinitroaniline-resistant parasites have reduced fitness reflected by increased rates of replication defects and poor performance in growth competition assays. These lines rapidly and spontaneously acquire compensatory mutations that correct the fitness defects and reduce resistance to dinitroanilines. We conclude that drugs that selectively target parasite microtubules remain a realistic option for the development of new antiparasitic therapies.

MATERIALS AND METHODS Culture of Toxoplasma lines: T. gondii tachyzoites (the standard RH strain and mutants derived from the RH strain) were propagated in human foreskin fibroblast (HFF) cells in DMEM with 10% FBS as previously described (Roos et al. 1994). Oryzalin (Riedel-deHaen, Germany) stock solutions

Secondary Mutations in Tubulin were made up in DMSO. Lines bearing oryzalin-resistance mutations were propagated in 0.5 mm oryzalin to suppress selection of secondary mutations. Selection of suppressors: T. gondii lines bearing allelic replacements of the F52Y or G142S mutations to the a1-tubulin gene were generated as previously described (Morrissette et al. 2004; Ma et al. 2007). These lines were serially passed for 3–4 weeks in the absence of oryzalin to select for spontaneous mutants with improved growth. After single-cell cloning, the clones displayed robust growth and each of the lines analyzed represents an independently derived (nonsibling) lineage. Clonal lines were analyzed for changes to the a1- or b1-tubulin genes. Analysis of tubulin point mutations: The a1- and b1-tubulin genes were amplified from genomic DNA isolated from individual parasite lines as previously described (Morrissette et al. 2004; Ma et al. 2007). The a1-tubulin gene was amplified using thermal cycling with primers GAGTCTCGTAGAGAA CAAGC (59 UTRA) and CGTTTATACCTTCACCTTTTC (39 UTRA), and the purified fragment was sequenced as previously described. The b1-tubulin gene was amplified with primers GTGGTGTTGCGCCTTC (59 UTRB) and CGAGT GTTTAGGACAGTGAC (39 UTRB) and was sequenced with the following primers: (BT3E1) CATTCTCCGCGATTCTC; (BT5E2) GTCCGGGTGTTCCTAC; (BTF2a) CATCATGGAG ACTTTCTCC; (BTR2a) GGAGAAAGTCTCCATGATG; (BTF2b) CAAAGAACATGATGTGCG; (BTR2b) CGCACATCATGTTC TTTG; (BT3E2) CTCGTCCATACC TTCACC; (BT5E3) GAGA TGGCACATTTAGTGTG; (BT3E3) CTCCCTCTTCCTCTGC; and (BT5E4) CCGAGTATCAGCAGTACC. DNA sequences were analyzed using Sequencher software (GeneCodes) to identify point mutations in the coding sequence of the a1-tubulin or b1-tubulin genes. We also checked lines for a2-, a3-, b2-, and b3-tubulin mutations but did not identify any changes to these genes. Structural models of T. gondii tubulin: Point mutations were mapped onto models of T. gondii tubulin using PyMol (Delano 2002). The models (containing a rebuilt H1-S2 loop) were previously generated for computational studies (Morrissette et al. 2004; Mitra and Sept 2006). Immunofluorescence staining: Extracellular parasites were fixed, permeablized, and stained in suspension as previously described (Morrissette and Sibley 2002b). The T. gondii plasma membrane was labeled with DG52 antibody (Burg et al. 1988) and detected with an Alexa 594 secondary antibody (Invitrogen). DNA was visualized with DAPI. Suspension samples were allowed to settle onto 35-mm dishes with glass coverslip insets (MatTek). Images were collected on a Zeiss Axioskop using the Axiovision camera and software. Images were exported and manipulated in Photoshop 8.0. Quantification of replication defects: T. gondii were passed into HFF cells without oryzalin selection and allowed to grow until complete host cell lysis. Extracellular parasites were viewed in suspension in MatTek dishes with a coverslip inset using a 363 phase-contrast lens on a Zeiss Axioskop microscope. Images were captured as tif files and scored by counting as previously described (Ma et al. 2007). To avoid underrepresenting aberrant forms, we counted the number of apical regions to establish total parasites lost from the parasite population through replication defects. Competition assays: T25 flasks with confluent HFF cells were inoculated with a 1:1 ratio of RH-strain-derived lines described here and wild-type (RH strain) parasites expressing cytosolic GFP (1 3 107 parasites from each line) (Kim et al. 2001). After host monolayer lysis, a new T25 flask with confluent HFF cells was inoculated with 1.0 ml of the lysed culture and the remaining material (5 ml) was used for flow cytometry analysis. Extracellular parasites were purified away from host cell debris

847

by passing lysate through a 3-mm filter. After parasites were collected by centrifugation (all spins performed at 6000 rpm for 6 min), they were fixed in 1.0% formalin (Sigma) for 1 min at RT. After a PBS wash, parasites were blocked in 10% (w/v) BSA/ PBS (Fisher Scientific) for 30 min at 4° and surface labeled with DG52 anti-P30/SAG1 mouse antibody [30 min at 4°, 1:1000 in 10% (w/v) BSA/PBS] and a PE-Cy5.5-conjugated goat antimouse secondary antibody (Invitrogen) for 30 min at 4°, 1:1000 in 10% (w/v) BSA/PBS. The total number of parasites (PECy5.5 labeling) and the number of parasites with GFP fluorescence were enumerated by flow cytometry using a FACSCalibur flow cytometer (Becton Dickinson) and analyzed with Flowjo software (Tree Star). The trend lines represent the percentage of total mutant parasites at each time point, and the analysis was carried out for 7 and 15 serial parasite passages.

RESULTS

Toxoplasma a1-tubulin mutations are associated with oryzalin resistance but have fitness defects: We previously isolated a number of T. gondii lines that have mutations to a1-tubulin that are sufficient for conferring resistance to dinitroanilines such as oryzalin (Morrissette et al. 2004; Ma et al. 2007). Among the lines isolated in our work are two that are oryzalin resistant due to point mutations F52Y or G142S in the a1-tubulin gene (Figure 1). Parasites bearing the G142S mutation are resistant to 1.0 mm oryzalin. G142 participates in binding the a-phosphate portion of GTP (Gigant et al. 2000; Lowe et al. 2001), and parasites with the G142S mutation are largely dependent on the presence of oryzalin for viable growth in culture. F52 is located in the H1-S2 loop, a domain that is critically important for protofilament– protofilament associations in the microtubule lattice (Li et al. 2002). Parasites bearing the F52Y mutation have high rates of defective replication and have been shown to spontaneously acquire compensatory mutations. We previously reported that parasites with the F52Y mutation were resistant to 7 mm oryzalin (Ma et al. 2007). In the course of carrying out the experiments reported here, we discovered that parasites bearing this mutation rapidly acquire secondary mutations (even when under 0.5 mm oryzalin selection) and that the true resistance of the parental line is 13 mm. Relative to wild-type T. gondii, both F52Y and G142S parasite lines have high rates of overt replication defects (Figure 1B). We previously quantified the total percentage of aberrant extracellular parasites for wild type (4%) and the F52Y (13%) line (Ma et al. 2007). Using a similar analysis, the G142S line has 19% replication defects when grown in the absence of oryzalin (not shown). Defects are corrected by compensatory mutations in lines derived from the two parental parasite strains: When T. gondii parasites bearing an allelic integration of the F52Y or G142S mutation are passed for several generations in the absence of oryzalin selection, they spontaneously acquire mutations that allow them to grow with increased robustness. After single-cell cloning independently selected lines with improved growth, we

848

C. Ma et al.

Figure 1.—(A) The F52Y and G142S a-tubulin point mutations confer 13 and 1 mm oryzalin resistance when integrated into wild-type (sensitive) T. gondii parasites. The F52Y mutation (red) is located in the H1-S2 (N) loop (red) of a1-tubulin. The computationally determined dinitroaniline-binding site is below this loop (oryzalin in orange). Interactions between the H1-S2 (N) loop and the M loop of adjacent dimers coordinate lateral associations between protofilaments in the microtubule lattice. The G142S mutation is located in the core of a1-tubulin and the G142 residue represents the initial glycine in the GGGTGSG motif (teal), which is characteristic of tubulins. This motif is part of the GTP-binding site, which is essential for correct folding of the tubulin dimer and interacts with the phosphate portion of GTP (purple). (B) (Top) Immunofluorescence images of extracellular parasites of the G142S line labeled with DG52 (red labels the parasite surface) and DAPI (blue labels the parasite DNA). (Bottom) DNA distribution only. (1) A normal-appearing parasite has a crescent shape and DNA staining of the nucleus (arrow) and apicoplast DNA (arrowhead). (2–6) Parasite ‘‘monsters’’ result from replication defects due to improper coordination of spindle microtubule (nuclear division) and subpellicular microtubule (cytokinesis) functions. Although the extracellular parasites in 3 and 4 are similar to intracellular replicating stages (see diagram in Ma et al. 2007), parasites cannot complete division outside of host cells and their abnormal shape prevents invasion of new host cells. In addition, 2, 5, and 6 illustrate parasites with diffuse (2), improperly segregated (three nuclei) (5), and nonsegregated (6) nuclear staining (compare to 1).

analyzed the sequences of T. gondii a1- and b1-tubulin genes for mutations associated with suppression of defects. We identified both a1-tubulin pseudorevertants and extragenic mutations in b1-tubulin and other (unidentified) genes. We isolated 27 independent lines from the F52Y line and identified mutations that localize to a1-tubulin or b1-tubulin in 22 of these lines (Figure 2 and Table 1). Two of the a1-tubulin mutations are located in the M or H1-S2 (N) loops and seven mutations are located in the eight-amino-acid a-tubulinspecific insert. Six of the lines have mutations to the b1-tubulin gene; two of these are in the M or H1-S2 (N) loops. The G142S resistance mutation alters the GGGTGSG tubulin motif, which is associated with binding the phosphate portion of GTP. Of the 24 G142S-derived lines characterized here, 13 acquired a secondary mutation in the a1-tubulin gene (Figure 3 and Table 2). One of these mutations is associated with GTP binding (S171A), two localized in the eight-aminoacid a-tubulin-specific insert, and one is in the M loop. We also isolated 13 lines with extragenic suppressors, with 11 of these occurring in the b1-tubulin gene and 2 being in other, unidentified loci. During the course of these studies we did not identify any true revertants of F52Y or G142S parasites. Growth competition assays indicate that the derived lines have increased fitness relative to the parental strains: We have exploited an established GFP-expressing T. gondii line to assess the relative fitness of the F52Y and G142S parasite lines and the derived progeny (Kim et al. 2001). Both the GFP-expressing line and the

tubulin mutant lines are derived from RH strain T. gondii. We co-infected HFF cells with an equivalent number of GFP-expressing parasites and parasites of each tubulin mutant line and exploited green fluorescence to follow the relative rates of replication and growth over serial passage (Figure 4). Parasites from lysed flasks were used to inoculate new flasks such that they would lyse the host cell monolayer in 2 days. The remainder of each sample was analyzed by flow cytometry. In initial experiments, the parasites were labeled with the DG52 (anti-SAG1) antibody to ensure that the entire parasite population was visualized for analysis (Figure 4 inset). Once we were confident that our gating had captured the parasite population, we omitted the DG52 labeling, with consistent results. At each time point, the total number of parasites and the number of GFP-labeled parasites were determined. Nonfluorescing parasites represent the fraction of the population expressing the specific tubulin mutation(s). These growth competition assays indicate that the F52Y and G142S lines compete poorly with wild-type GFP parasites (Figure 4). After 6 days, the G142S line is not detectable and at 8 days the F52Y line cannot be detected. In contrast, the b1-P61A and a1-P364R lines that are derived from the G142S and F52Y parasites are still present at this time, although their declining percentage in the population indicates that they are less fit than the GFP-RH line. As a control, we also carried out growth competition for wild-type RH parasites vs. the GFP-expressing parasites. Over time, we reproducibly observed that the GFPexpressing parasites were at a disadvantage relative to the

Secondary Mutations in Tubulin

849

Figure 2.—Location of compensatory mutations obtained from the F52Y line mapped on a model of the Toxoplasma tubulin dimer. The a1-subunit is white and the b1-subunit is blue. (A) Most mutations fall between the M and H1S2 loops (red) facing the inner lumen of the microtubule. The F52Y mutation is located in the H1-S2 loop of a1-tubulin (red) and secondary mutations in a1-tubulin (black text) occur at T51A, V181I, A273V, K280N, N293D, A295G, S300T, M313T, P360A, T361P, P364A, P364R, G366R, D367V, L368F, and R373C (orange/yellow). Mutations in yellow localize to the a-tubulin-specific insert. Extragenic suppressors in the b1-tubulin subunit (white text) occur at G34S, G82D, P261S, G269V, L273V, and H396Q. (B) The b1-tubulin mutation P261S is the only mutation that faces the outer surface of the dimer within the microtubule lattice, while the b-subunit mutation H396Q is located at the dimer–dimer interface within the protofilament. (C) The mutations T361P, P364A, P364R, G366R, D367V, and L368F occur in the a-tubulin-specific insert (yellow) and the mutations P360A and R373C are adjacent to this insert (orange).

wild-type RH T. gondii. Therefore, the growth competition using GFP-RH parasites underestimates the fitness defects associated with the dinitroaniline resistance mutations and the associated suppressed lines. We analyzed all of the F52Y- and G142S-derived lines after 2 weeks in serial passage with GFP-RH parasites to assess relative fitness relationships in competition with wild-type parasites (Figures 5 and 6). In most cases, lines with compensatory mutations do not fully recover robust growth. The most effective F52Y-derived lines with respect to fitness are point mutations at T51A (the a-tubulin H1-S2 loop), P364A (the a-tubulin insert), and G34S (the b-tubulin H1-S2 loop), which were still all ,50% of the culture after 2 weeks (Figure 5 and Table 3). The most effective G142S-derived lines are S171A (an a-tubulin GTP-binding site residue), L230V (located in helix 7 of a-tubulin), and V363A (in the a-tubulin insert). The L230V line is present in 50% of the culture after 2 weeks (Figure 6 and Table 3). However, since the GFP-RH parasites have a fitness defect relative to unlabeled RH strain wild-type parasites, it is still likely that the L230V line would compete unfavorably with wild-type T. gondii. The derived lines have diminished oryzalin resistance relative to the parental strains: The F52Y and G142S

parasite lines display 13 and 1.0 mm oryzalin resistance, respectively. Lines derived from these parental strains display reduced oryzalin resistance (Figures 5 and 6 and Table 3). With respect to the F52Y-derived lines, the P364R line retains the highest level of oryzalin resistance (7 mm) and competes relatively well with the GFP-RH parasites. Since the G142S line parasites have 1 mm oryzalin resistance, the range of resistance observed in the derived lines is substantially lower. A number of the lines (the Q256H, S287P, and V363A mutations in a-tubulin and the P358R and A393V mutations in b-tubulin) do not have greater resistance to oryzalin than that observed for wild-type parasites (,0.5 mm by morphological criteria). Lines with secondary mutations at S171A and L230V retain the highest levels of oryzalin resistance in the context of the greatest degree of overall fitness in the competition assay. All 46 lines bearing compensatory mutations are associated with diminished oryzalin resistance. DISCUSSION

Previous genetic studies have identified secondary mutations that correct tubulin-associated defects. In

850

C. Ma et al. TABLE 1 F52Y secondary mutations

Mutant

Codon change

Location

T51A V181I A273V K280N

ACC to GCC GTT to ATT GCG to GTG AAG to AAC

AT AT AT AT

N293D A295G S300T

AAC to GAC GCT to GGT AGC to ACC

AT a-helix 9 AT a-helix 9 AT a-9 to b-8 loop

M313T P360A T361P P364A P364R G366R D367V L368F R373C G34S G82D P261S G269V L273V H396Q

ATG to ACG CCC to GCC ACT to CCT CCT to GCT CCT to CGT GGT to CGT GAC to GTC TTG to TTC CGC to TGC GGT to AGT GGC to GAC CCT to TCT GGG to GTG CTC to GTC CAC to CAG

AT AT AT AT AT AT AT AT AT BT BT BT BT BT BT

H1-S2 loop b-5 to a-5 loop M loop M loop

b-sheet 8 near insert insert insert insert insert insert insert b-sheet 10 H1-S2 loop a-2 to b-3 loop a-8 to b-7 loop b-sheet 7 M loop a-helix 11

Other nearby mutations F52I, F52L, F52Y: Tg dinitroanilineR (Morrissette et al. 2004) S180P, A180T: At twisting (monomer interface) (Ishida et al. 2007a) I275T: Tg dinitroanilineR (Morrissette et al. 2004) K279A-K281A: Sc benomylss (Richards et al. 2000), A281T: At twisting (Ishida et al. 2007a) A295V: Tg dinitroanilineR (Morrissette et al. 2004) A295V: Tg dinitroanilineR (Morrissette et al. 2004) S300T: Ant. fish coldR (Detrich et al. 2000), M301T: Tg dinitroanilineR (Morrissette et al. 2004) D310A-K312A: Sc benomylss (Richards et al. 2000) G354E: Ce mec-12 (Fukushige et al. 1999)

Also suppresses G142S

D373A-R374A: Sc benomylss (Richards et al. 2000) G34S: Ce mec-7 (Savage et al. 1989) F85L: Tg suppressor of G142S F261V: Hs paclitaxelR (Monzo et al. 1999) G269D: Ce mec-7 (Savage et al. 1989) T274I: Hs epothilone/taxaneR (Giannakakou et al. 2000) A393T: Ce mec-7 (Savage et al. 1989); A394T: At twisting (Ishida et al. 2007a)

At, Arabidopsis thaliana; Ant. fish, Antarctic fish; Ce, Caenorhabditis elegans; Hs, Homo sapiens; Sc, Saccharomyces cerevisiae; Tg, Toxoplasma gondii; R, resistant; ss, supersensitive.

studies of colchicine resistance, intragenic mutations correct defects associated with the D45Y mutation to btubulin, which confers colchicine resistance in CHO cells (Wang et al. 2004). This substitution is located in the H1-S2 loop and, since it increases microtubule stability, cells that express it are hypersensitive to taxol. Lines that suppress the taxol hypersensitivity and associated temperature-sensitive defects have point mutations at V60A (the H1-S2 loop) or Q292H (near the M loop) of b-tubulin that alter microtubule stability. When wild-type CHO cells express a tagged b-tubulin construct bearing the D45Y mutation, the microtubule array is strikingly denser relative to the array observed with expression of wild-type tubulin alone. Inclusion of secondary mutations in the b-tubulin constructs reduces the microtubule array, making it similar to that observed with expression of wild-type tubulin. When b-tubulin constructs with the V60A and Q292H secondary mutations alone are expressed, the microtubule arrays are dramatically reduced relative to those observed with wild-type tubulin expression. These observations are strikingly similar to our results described here. The F52Y mutation is predicted to increase microtubule stability to confer dinitroaniline resistance, but also causes increased rates of replication defects. Secondary mutations are located in regions such as the H1-S2 loop, the M loop, and the a-

tubulin insert domain, which are predicted to decrease microtubule stability. We attempted to express GFP-tagged vertebrate tubulin with analogous mutations in COS-7 cells to visualize differential assembly of these tubulins. Unfortunately, expression appeared to be somewhat toxic and the patterns were too variable to be used to assess the relative stability of tubulins with these mutations ( J. Tran and N. Morrissette, data not shown). Many of the secondary mutations are identical or quite similar to previously identified mutations that decrease microtubule stability in Caenorhabditis elegans and Arabidopsis thaliana. C. elegans sensory neuron function requires unusual 15-protofilament microtubules (Chalfie and Thomson 1979, 1982; Chalfie 1982; Chalfie et al. 1986; Savage et al. 1989, 1994). These microtubules are built from dimers encoded by distinct a- (mec-121) and b- (mec-71) tubulin genes (Savage et al. 1989, 1994; Gu et al. 1996; Fukushige et al. 1999). A number of mec-7 and mec-12 alleles have been identified in screens for sensory neuron defects. In addition, mutations in Arabidopsis that convert linear growth of plant parts (such as roots and stems) to a twisted phenotype have altered cortical microtubule arrays. This twisting morphology can be phenocopied by treatment of wild-type plants with the microtubule inhibitor propyzamide. Many of the twisting mutants have muta-

Secondary Mutations in Tubulin

851

Figure 3.—Location of compensatory mutations obtained from the G142S line mapped on a model of the Toxoplasma tubulin dimer (the a1-subunit is white and the b1-subunit is blue). The G142S mutation (yellow) is located in the GTP-binding domain of atubulin. (A) Most mutations (orange/yellow) are on the surface of the dimer that faces the inner lumen of the microtubule. Secondary mutations in a1-tubulin (black text) occur at I93V, G131R, S171A, I209V, L230V, Q256H, S287P, P348L, I355E, V363A, P364R, V371G, and F418I. Suppressors in the b1-tubulin subunit (white text) occur at residues A18P, P61A, F85L, K122E, H264Q, Q291E, R318C, N337T, P358R, M363V, and A393V. G142S a1-tubulin mutations at V363A and P364R occur in the a-tubulin-specific insert (yellow). (B) The a1-tubulin mutation at Q256 and the b1-tubulin mutation at H264 are the only mutations that face the outer surface of the dimer within the microtubule lattice. The b-subunit mutation A393V is located at the dimer–dimer interface within the protofilament. (C) The primary resistance mutation at G142 is located within the tubulin motif (teal) of the a-tubulin GTP-binding site. The GTP moiety is purple and residues contributing to GTP binding outside of the GGGTGS tubulin motif are red. The primary mutation G142S is located in the binding site and interacts with the a-phosphate. The S171A mutation also locates to the binding site; S171 interacts with the ribose portion of GTP (Gigant et al. 2000; Lowe et al. 2001).

tions to a- and b-tubulin that decrease microtubule stability and increase plant sensitivity to microtubuledisrupting drugs (Thitamadee et al. 2002; Abe et al. 2004; Nakajima et al. 2004, 2006; Shoji et al. 2004; Abe and Hashimoto 2005; Ishida and Hashimoto 2007; Ishida et al. 2007a,b). The b-tubulin mutation G34S (a F52Y suppressor) was previously identified as a mild,

recessive mec-7 allele (Savage et al. 1989). Other mec-7 alleles (P61L/S, G269D, R318G, and A393T) alter equivalent residues to different substitutions in T. gondii b-tubulin (the F52Y suppressor G269V and the G142S suppressors P61A, R318C, and A393V). The G142S secondary mutation I355E in a-tubulin is adjacent to the location of the mec-12 allele G354E. Moreover, the A393

852

C. Ma et al. TABLE 2 G142S secondary mutations

Mutant

Codon change

Location

I93V G131R S171A I209V L230V Q256H S287P P348L I355E

ATC to GTC GGT to CGT TCG to GCG ATC to GTC CTG to GTG CAG to CAC TCT to CCT CCC to CTC GAA to ATC

AT AT AT AT AT AT AT AT AT

b-sheet 3 a-3 to b-4 loop b-5 to a-5 loop a-helix 6 a-helix 7 a-helix 8 M-loop a-10 to b-9 loop b-sheet 9

V363A P364R V371G F418I

GTT to GCT CCT to CGT GTC to GGC TTC to ATC

AT AT AT AT

insert insert b-9 to b-10 loop a-helix 12

A18P P61A

GCC to CCC CCG to GCG

BT a-helix 1 BT H1-S2 loop

F85L K122E H264Q

TTC to TTG AAG to GAG CAC to CAG

BT a-2 to b-2 loop BT a-helix 3 BT a-8 to b-7 loop

Q291E

CAG to GAG

BT a-helix 9

R318C

CGT to TGT

BT b-sheet 8

N337T P358R M363V A393V

AAC to ACC CCG to CGG ATG to GTG GCT to GTT

BT BT BT BT

a-10 to b-9 loop b-9 to b-10 loop b-9 to b-10 loop a-helix 11

Other nearby mutations H89A-E91A: Sc benomylss (Richards et al. 2000) D128A-D131A: Sc colds, benomylss (Richards et al. 2000) K167A-E169A: Sc benomylss (Richards et al. 2000) A208F: At lefty pseudorevertant (Thitamadee et al. 2002) I231T: Tg dinitroanilineR (Morrissette et al. 2004) E255A: Sc dominant lethal (Richards et al. 2000) S287P: Antarctic fish coldR (Detrich et al. 2000) T349I: At twisting (Ishida et al. 2007a) G354E: Ce mec-12 (Fukushige et al. 1999); C356Y: Sp lethalts (Radcliffe et al. 1998) Also suppresses F52Y D373A-R374A: Sc benomylss (Richards et al. 2000) E415K: Ce mec-12 (Fukushige et al. 1999); E416A-E418A: Sc slow growth, benomylss (Richards et al. 2000) A19K Sc required for taxol binding (Gupta et al. 2003) V60A: CHO D45Y supp. (Wang et al. 2004); P61L, P61S: Ce mec-7 (Savage et al. 1989) G82D: Tg compensatory for F52Y R121A-R122A: Sc benomylss (Reijo et al. 1994) L263Q: Dm E194K suppr. (Fackenthal et al. 1995); G269D: Ce mec-7 (Savage et al. 1989) Q292G: Hs epothiloneR (He et al. 2001); Q292P: Ce mec-7 (Savage et al. 1989); Q292H: D45Y supp. (Wang et al. 2004) F317I, R318G: Ce mec-7 (Savage et al. 1989); R318A-K320A: Sc benomylR (Reijo et al. 1994) K336A-D339A: Sc recessive lethal (Reijo et al. 1994) C354S, C354A: Sc benomylR, cold stable (Reijo et al. 1994) A364T: Hs paclitaxelR (Giannakakou et al. 2000) A393T: Ce mec-7 (Savage et al. 1989); A394T: At twisting (Ishida et al. 2007a)

At, A. thaliana; Ce, C. elegans; CHO, Chinese hamster ovary cells; Dm, Drosophila melanogaster; Hs, H. sapiens; Sc, S. cerevisiae; Sp, Schizzosacharomyces pombe; R, resistant; s, sensitive; ss, supersensitive; ts, temperature sensitive.

residue is also identified as a threonine substitution causing twisting in Arabidopsis b-tubulin (A394T). The F52Y compensatory mutations V181I and K280N in atubulin and H396Q in b-tubulin and the G142S mutation P348L (a-tubulin) are near the twisting mutations S180P, A180T, A281T, and T349I (a-tubulin) and A394T (b-tubulin). Finally, the G142S suppressor Q291E is adjacent to a second glutamine at 292 that is a mec-7 allele (Q292P). This residue also influences microtubule stability in the context of epothilone resistance (Q292G) and suppresses hyperstabilized microtubules (Q292H) in D45Y colchicine-resistant cells (He et al. 2001; Wang et al. 2004). One of the novel findings presented in this work is the identification of a set of seven mutations that localize to the a-tubulin-specific insert (TVVPGGDL). The compensatory mutations in the F52Y line that localize to the insert are T361P, P364A, P364R, G366R, D367V, and L368F and in the G142S line mutations are V363A and P364R. To our knowledge, the insert substitutions rep-

resent the first description of mutations located in this highly conserved domain. We hypothesize that these substitutions impair the ability of the insert to promote M-loop lateral contacts with the adjacent protofilament. Indeed, the D367V mutation eliminates a salt bridge that occurs with R229 (Ma et al. 2007). Strikingly, the P364R mutation functions as a compensatory mutation in both lines, albeit more successfully in the F52Y line (compare competition results in Figures 5 and 6 and Table 3). Although the G142S/P364R line does not compete effectively enough with wild-type parasites to remain in the culture at 14 days, it was isolated as a line with improved growth over the G142S line and at some subtle level is likely to improve fitness of the parental line. This is also likely to be the explanation for the F52Y suppressor A295G in competition at 14 days (Table 3). The studies presented here show a correlation between acquisition of increased fitness and the appearance of mutations in the a1- or b1-tubulin genes. While

Secondary Mutations in Tubulin

853

Figure 4.—To assess relative fitness of the F52Y and G142S lines as well as the strains derived from these lines, equivalent numbers of parasites from a specific mutant line and wild-type RH strain parasites expressing GFP were inoculated into culture and the relative percentage of each population was quantified over time during serial passage using flow cytometry. The relative parasite concentrations were quantified at the time of complete host cell lysis when the lysate was passed into a new flask of host cells containing parasites. The red line indicates the behavior of G142S parasites in competition and the orange line is the G142S suppressor line b-P61A (located in the b-tubulin H1-S2 loop). The green line follows a competition assay containing F52Y and the teal line represents behavior of the F52Y secondary mutation a-P364R (located in the a-tubulin-specific insert). The G142S parasites are outcompeted by GFP-RH parasites by day 6 whereas the b-P61A suppressor line persists until day 24. The F52Y parasites are eliminated at day 12 but, in the presence of the a-P364R secondary mutation, they were still present, albeit at reduced levels, when the competition was terminated at 30 days. The purple line traces the control competition between wild-type (untransfected RH strain parasites) and the GFP-transfected ‘‘wild-type’’ RH strain line. Unlabeled parasites have a growth advantage over GFP-transfected parasites. (Inset) A representative FACS dot plot of DG52-labeled (red) parasites showing relative numbers of GFP-expressing wildtype Toxoplasma and a mutant line after a single passage.

the improved fitness observed in parasite lines derived from the F52Y and G142S parental strains is likely due to secondary tubulin mutations, other (nontubulin) mutations may also contribute to the observed phenotypes of decreased resistance and enhanced growth. Indeed, five of the lines derived from the F52Y suppressor screen and two lines derived from the G142S suppressor screen have improved growth but do not have altered a1-, a3-, b1-, or b3-tubulin genes, suggesting that nontubulin compensatory mutations do arise. Moreover, although we have created allelic replacements for two a-tubulin pseudorevertants in wild-type parasites (Ma et al. 2007), we cannot create allelic replacements for the b-tubulin suppressors since we cannot select for these in the

absence of dinitroaniline selection. When we analyzed the allelic replacements for the a-tubulin pseudorevertants (F52Y/A273V and F52Y/D367V) in the competition assay, they performed differently from the equivalent lines with these spontaneous mutations. Surprisingly, in both cases the transgene versions of the lines had greater fitness levels than the spontaneous mutant lines (data not shown). The suppressed lines described here were isolated after 4–6 weeks of growth in media without oryzalin. This represents the shortest time in which we could observe parasites with new mutations. In contrast, it takes at least 6 weeks to isolate transgene-bearing lines. We predict that further growth of lines with impaired fitness in the absence of oryzalin selection

Figure 5.—Replication fitness and oryzalin resistance in F52Y parasites and the derived lines. Solid bars: the percentage of mutant parasites relative to GFP-RH parasites after seven passages (14 days). The assays were carried out in triplicate and error bars indicate standard error of the mean. Shaded bars: oryzalin resistance levels observed for F52Y parasites and the derived lines.

854

C. Ma et al.

Figure 6.—Replication fitness and oryzalin resistance in G142S parasites and the 24 derived lines. Solid bars: the percentage of mutant parasites relative to GFP-RH parasites after seven passages (14 days). The assays were carried out in triplicate and error bars indicate standard error of the mean. Shaded bars: oryzalin resistance levels observed in G142S parasites and the derived lines.

would select for additional mutations that collectively contribute to increased fitness. Previous studies have established that there is a tradeoff between carrying genes that confer drug resistance and the fitness cost of these altered alleles, particularly in the absence of drug selection. Recent field studies from Malawi indicate that P. falciparum parasites in this region lost resistance to chloroquine in ,10 years after sulfadoxine–pyrimethamine treatment

replaced chloroquine treatment for individuals suffering from malaria (Laufer et al. 2006; Vogel 2006). These field observations, as well as the controlled laboratory studies of DHFR mutations in P. falciparum and T. gondii described in the Introduction, indicate that the existence of mutant lines that express resistant forms of target proteins does not rule out the development and use of agents that inhibit these proteins. With respect to the studies presented here, we believe

TABLE 3 Resistance and fitness levels

Mutation F52Y-parental F52Y-aT51A F52Y-aV181I F52Y-aA273V F52Y-aK280N F52Y-aN293D F52Y-aA295G F52Y-aS300T F52Y-aM313T F52Y-aP360A F52Y-aT361P F52Y-aP364A F52Y-aP364R F52Y-aG366R F52Y-aD367V F52Y-aL368F F52Y-aR373C F52Y-bG34S F52Y-bG82D F52Y-bP261S F52Y-bG269V F52Y-bL273V F52Y-bH296Q

% decrease resistance

% decrease population (day 14)

0 31 54 73 88 62 46 46 96 46 62 46 23 77 73 62 85 73 62 46 85 81 62

100 43 60 86 87 84 100 80 86 92 80 30 69 78 92 92 94 14 75 91 88 95 91

Mutation

% decrease resistance

% decrease population (day 14)

G142S-parental G142S-aI93V G142S-aG131R G142S-aS171A G142S-aI209V G142S-aL230V G142S-aQ256H G142S-aS287P G142S-aP348L G142S-aI355E G142S-aV363A G142S-aP364R G142S-aV371G G142S-aF418I G142S-bA18P G142S-bP61A G142S-bF85L G142S-bK122E G142S-bH264Q G142S-bQ291E G142S-bR318C G142S-bN337T G142S-bP358R G142S-bM363V G142S-bA393V

0 75 75 50 75 50 100 100 75 75 100 50 75 50 75 50 50 75 75 75 75 75 100 50 100

100 89 97 24 89 0 87 96 84 85 25 100 93 84 90 74 76 94 88 95 96 96 81 86 85

Secondary Mutations in Tubulin

that although dinitroanilines or other microtubuledisrupting compounds do not yet exist in an appropriate form for treatment of parasite infections, selectively targeting protozoan tubulin remains a viable strategy for eliminating parasite infections. We thank Dan Sackett for commenting on this manuscript, David Sibley and Michael Grigg for helpful conversations on competition assays, and Sheryl Tsai for bringing clarity to PyMOL commands. We thank Candice Kwark and Brian Luk, high school student participants in the American Cancer Society summer research program at the University of California at Irvine (UCI), who helped select and analyze the lines described here. Research presented in this article was supported by National Institutes of Health grants AI055790 and AI067981 to N.S.M. L.G. and D.W. are undergraduate students in the campus-wide honors program at UCI and L.G. was supported by a UCI summer undergraduate research program fellowship.

LITERATURE CITED Abe, T., and T. Hashimoto, 2005 Altered microtubule dynamics by expression of modified alpha-tubulin protein causes righthanded helical growth in transgenic Arabidopsis plants. Plant J. 43: 191–204. Abe, T., S. Thitamadee and T. Hashimoto, 2004 Microtubule defects and cell morphogenesis in the lefty1lefty2 tubulin mutant of Arabidopsis thaliana. Plant Cell Physiol. 45: 211–220. Anders, K. R., and D. Botstein, 2001 Dominant-lethal alpha-tubulin mutants defective in microtubule depolymerization in yeast. Mol. Biol. Cell 12: 3973–3986. Bates, S. J., P. A. Winstanley, W. M. Watkins, A. Alloueche, J. Bwika et al., 2004 Rare, highly pyrimethamine-resistant alleles of the Plasmodium falciparum dihydrofolate reductase gene from 5 African sites. J. Infect. Dis. 190: 1783–1792. Bhattacharya, G., M. M. Salem and K. A. Werbovetz, 2002 Antileishmanial dinitroaniline sulfonamides with activity against parasite tubulin. Bioorg. Med. Chem. Lett. 12: 2395–2398. Black, M. W., and J. C. Boothroyd, 2000 Lytic cycle of Toxoplasma gondii. Microbiol. Mol. Biol. Rev. 64: 607–623. Bogitsh, B. J., O. L. Middleton and R. Ribeiro-Rodrigues, 1999 Effects of the antitubulin drug trifluralin on the proliferation and metacyclogenesis of Trypanosoma cruzi epimastigotes. Parasitol. Res. 85: 475–480. Burg, J. L., D. Perelman, L. H. Kasper, P. L. Ware and J. C. Boothroyd, 1988 Molecular analysis of the gene encoding the major surface antigen of Toxoplasma gondii. J. Immunol. 141: 3584– 3591. Chalfie, M., 1982 Microtubule structure in Caenorhabditis elegans neurons. Cold Spring Harb. Symp. Quant. Biol. 46(Pt 1): 255– 261. Chalfie, M., and J. N. Thomson, 1979 Organization of neuronal microtubules in the nematode Caenorhabditis elegans. J. Cell Biol. 82: 278–289. Chalfie, M., and J. N. Thomson, 1982 Structural and functional diversity in the neuronal microtubules of Caenorhabditis elegans. J. Cell Biol. 93: 15–23. Chalfie, M., E. Dean, E. Reilly, K. Buck and J. N. Thomson, 1986 Mutations affecting microtubule structure in Caenorhabditis elegans. J. Cell Sci. Suppl. 5: 257–271. Chan, M. M., and D. Fong, 1990 Inhibition of leishmanias but not host macrophages by the antitubulin herbicide trifluralin. Science 249: 924–926. David-Pfeuty, T., H. P. Erickson and D. Pantaloni, 1977 Guanosinetriphosphatase activity of tubulin associated with microtubule assembly. Proc. Natl. Acad. Sci. USA 74: 5372–5376. DeLano, W. L., 2002 The PyMOL Molecular Graphics System. http:// www.pymol.org. Detrich, H. W., III, S. K. Parker, R. C. Williams, Jr., E. Nogales and K. H. Downing, 2000 Cold adaptation of microtubule assembly and dynamics. Structural interpretation of primary sequence changes present in the alpha- and beta-tubulins of Antarctic fishes. J. Biol. Chem. 275: 37038–37047.

855

Downing, K. H., and E. Nogales, 1998 Tubulin and microtubule structure. Curr. Opin. Cell Biol. 10: 16–22. Fackenthal, J. D., J. A. Hutchens, F. R. Turner and E. C. Raff, 1995 Structural analysis of mutations in the Drosophila b2tubulin isoform reveals regions in the b-tubulin molecular required for general and for tissue-specific microtubule functions. Genetics 139: 267–286. Fohl, L. M., and D. S. Roos, 2003 Fitness effects of DHFR-TS mutations associated with pyrimethamine resistance in apicomplexan parasites. Mol. Microbiol. 50: 1319–1327. Fukushige, T., Z. K. Siddiqui, M. Chou, J. G. Culotti, C. B. Gogonea et al., 1999 MEC-12, an alpha-tubulin required for touch sensitivity in C. elegans. J. Cell Sci. 112(Pt 3): 395–403. Giannakakou, P., R. Gussio, E. Nogales, K. H. Downing, D. Zaharevitz et al., 2000 A common pharmacophore for epothilone and taxanes: molecular basis for drug resistance conferred by tubulin mutations in human cancer cells. Proc. Natl. Acad. Sci. USA 97: 2904–2909. Gigant, B., P. A. Curmi, C. Martin-Barbey, E. Charbaut, S. Lachkar et al., 2000 The 4 A X-ray structure of a tubulin:stathmin-like domain complex. Cell 102: 809–816. Gu, G., G. A. Caldwell and M. Chalfie, 1996 Genetic interactions affecting touch sensitivity in Caenorhabditis elegans. Proc. Natl. Acad. Sci. USA 93: 6577–6582. Gupta, M. L., Jr., C. J. Bode, G. I. Georg and R. H. Himes, 2003 Understanding tubulin-Taxol interactions: mutations that impart Taxol binding to yeast tubulin. Proc. Natl. Acad. Sci. USA 100: 6394–6397. Hankins, E. G., D. C. Warhurst and C. H. Sibley, 2001 Novel alleles of the Plasmodium falciparum dhfr highly resistant to pyrimethamine and chlorcycloguanil, but not WR99210. Mol. Biochem. Parasitol. 117: 91–102. Hastings, M. D., S. J. Bates, E. A. Blackstone, S. M. Monks, T. K. Mutabingwa et al., 2002 Highly pyrimethamine-resistant alleles of dihydrofolate reductase in isolates of Plasmodium falciparum from Tanzania. Trans. R. Soc. Trop. Med. Hyg. 96: 674–676. He, L., C.-P. Huang Yang and S. B. Horwitz, 2001 Mutations in {beta}tubulin map to domains involved in regulation of microtubule stability in epothilone-resistant cell lines. Mol. Cancer Ther. 1: 3–10. Hess, F. D., and D. E. Bayer, 1977 Binding of the herbicide trifluralin to Chlamydomonas flagellar tubulin. J. Cell Sci. 24: 351–360. Hu, K., D. S. Roos and J. M. Murray, 2002 A novel polymer of tubulin forms the conoid of Toxoplasma gondii. J. Cell Biol. 156: 1039–1050. Hugdahl, J. D., and L. C. Morejohn, 1993 Rapid and reversible high-affinity binding of the dinitroaniline herbicide oryzalin to tubulin from Zea mays L. Plant Physiol. 102: 725–740. Ishida, T., and T. Hashimoto, 2007 An Arabidopsis thaliana tubulin mutant with conditional root-skewing phenotype. J. Plant Res. 120: 635–640. Ishida, T., Y. Kaneko, M. Iwano and T. Hashimoto, 2007a Helical microtubule arrays in a collection of twisting tubulin mutants of Arabidopsis thaliana. Proc. Natl. Acad. Sci. USA 104: 8544–8549. Ishida, T., S. Thitamadee and T. Hashimoto, 2007b Twisted growth and organization of cortical microtubules. J. Plant Res. 120: 61–70. Kim, K., M. S. Eaton, W. Schubert, S. Wu and J. Tang, 2001 Optimized expression of green fluorescent protein in Toxoplasma gondii using thermostable green fluorescent protein mutants. Mol. Biochem. Parasitol. 113: 309–313. Laufer, M. K., P. C. Thesing, N. D. Eddington, R. Masonga, F. K. Dzinjalamala et al., 2006 Return of chloroquine antimalarial efficacy in Malawi. N. Engl. J. Med. 355: 1959–1966. Levene, N. D., 1988 The Protozoan Phylum Apicomplexa. CRC Press, Boca Raton, FL. Li, H., D. DeRosier, W. Nicholson, E. Nogales and K. Downing, 2002 Microtubule structure at 8 a resolution. Structure 10: 1317. Lowe, J., H. Li, K. H. Downing and E. Nogales, 2001 Refined structure of alpha beta-tubulin at 3.5 A resolution. J. Mol. Biol. 313: 1045–1057. Ma, C., C. Li, L. Ganesan, J. Oak, S. Tsai et al., 2007 Mutations in {alpha}-tubulin confer dinitroaniline resistance at a cost to microtubule function. Mol. Biol. Cell 18: 4711–4720. Makioka, A., M. Kumagai, H. Ohtomo, S. Kobayashi and T. Takeuchi, 2000 Effect of dinitroaniline herbicides on the growth of Entamoeba histolytica. J. Parasitol. 86: 607–610.

856

C. Ma et al.

Mitra, A., and D. Sept, 2006 Binding and interaction of dinitroanilines with apicomplexan and kinetoplastid alpha-tubulin. J. Med. Chem. 49: 5226–5231. Monzo, M., R. Rosell, J. J. Sanchez, J. S. Lee, A. O’Brate et al., 1999 Paclitaxel resistance in non-small-cell lung cancer associated with beta-tubulin gene mutations. J. Clin. Oncol. 17: 1786– 1793. Morejohn, L. C., T. E. Bureau, J. Mole-Bajer, A. S. Bajer and D. E. Fosket, 1987 Oryzalin, a dinitroaniline herbicide, binds to plant tubulin and inhibits microtubule polymerization in vitro. Planta 172: 252–264. Morrissette, N. S., and L. D. Sibley, 2002a Cytoskeleton of apicomplexan parasites. Microbiol. Mol. Biol. Rev. 66: 21–38. Morrissette, N. S., and L. D. Sibley, 2002b Disruption of microtubules uncouples budding and nuclear division in Toxoplasma gondii. J. Cell Sci. 115: 1017–1025. Morrissette, N. S., A. Mitra, D. Sept and L. D. Sibley, 2004 Dinitroanilines bind alpha-tubulin to disrupt microtubules. Mol. Biol. Cell 15: 1960–1968. Nakajima, K., I. Furutani, H. Tachimoto, H. Matsubara and T. Hashimoto, 2004 SPIRAL1 encodes a plant-specific microtubule-localized protein required for directional control of rapidly expanding Arabidopsis cells. Plant Cell 16: 1178–1190. Nakajima, K., T. Kawamura and T. Hashimoto, 2006 Role of the SPIRAL1 gene family in anisotropic growth of Arabidopsis thaliana. Plant Cell Physiol. 47: 513–522. Nogales, E., and H. W. Wang, 2006a Structural intermediates in microtubule assembly and disassembly: How and why? Curr. Opin. Cell Biol. 18: 179–184. Nogales, E., and H. W. Wang, 2006b Structural mechanisms underlying nucleotide-dependent self-assembly of tubulin and its relatives. Curr. Opin. Struct. Biol. 16: 221–229. Nogales, E., S. G. Wolf and K. H. Downing, 1997 Visualizing the secondary structure of tubulin: three-dimensional map at 4 A. J. Struct. Biol. 118: 119–127. Nogales, E., K. H. Downing, L. A. Amos and J. Lowe, 1998 Tubulin and FtsZ form a distinct family of GTPases. Nat. Struct. Biol. 5: 451–458. Nogales, E., M. Whittaker, R. A. Milligan and K. H. Downing, 1999 High-resolution model of the microtubule. Cell 96: 79–88. Nogales, E., H. W. Wang and H. Niederstrasser, 2003 Tubulin rings: Which way do they curve? Curr. Opin. Struct. Biol. 13: 256–261. Radcliffe, P., D. Hirata, D. Childs, L. Vardy and T. Toda, 1998 Identification of novel temperature-sensitive lethal alleles in essential beta-tubulin and nonessential alpha 2-tubulin genes as fission yeast polarity mutants. Mol. Biol. Cell 9: 1757– 1771. Reijo, R. A., E. M. Cooper, G. J. Beagle and T. C. Huffaker, 1994 Systematic mutational analysis of the yeast beta-tubulin gene. Mol. Biol. Cell 5: 29–43.

Richards, K. L., K. R. Anders, E. Nogales, K. Schwartz, K. H. Downing et al., 2000 Structure-function relationships in yeast tubulins. Mol. Biol. Cell 11: 1887–1903. Roos, D. S., R. G. Donald, N. S. Morrissette and A. L. Moulton, 1994 Molecular tools for genetic dissection of the protozoan parasite Toxoplasma gondii. Methods Cell Biol. 45: 27–63. Sage, C. R., A. S. Davis, C. A. Dougherty, K. Sullivan and K. W. Farrell, 1995a beta-Tubulin mutation suppresses microtubule dynamics in vitro and slows mitosis in vivo. Cell Motil. Cytoskeleton 30: 285–300. Sage, C. R., C. A. Dougherty, A. S. Davis, R. G. Burns, L. Wilson et al., 1995b Site-directed mutagenesis of putative GTP-binding sites of yeast beta-tubulin: evidence that alpha-, beta-, and gammatubulins are atypical GTPases. Biochemistry 34: 7409–7419. Savage, C., M. Hamelin, J. G. Culotti, A. Coulson, D. G. Albertson et al., 1989 mec-7 is a beta-tubulin gene required for the production of 15-protofilament microtubules in Caenorhabditis elegans. Genes Dev. 3: 870–881. Savage, C., Y. Xue, S. Mitani, D. Hall, R. Zakhary et al., 1994 Mutations in the Caenorhabditis elegans beta-tubulin gene mec-7: effects on microtubule assembly and stability and on tubulin autoregulation. J. Cell Sci. 107(Pt 8): 2165–2175. Shoji, T., N. N. Narita, K. Hayashi, J. Asada, T. Hamada et al., 2004 Plant-specific microtubule-associated protein SPIRAL2 is required for anisotropic growth in Arabidopsis. Plant Physiol. 136: 3933–3944. Stokkermans, T. J., J. D. Schwartzman, K. Keenan, N. S. Morrissette, L. G. Tilney et al., 1996 Inhibition of Toxoplasma gondii replication by dinitroaniline herbicides. Exp. Parasitol. 84: 355–370. Thitamadee, S., K. Tuchihara and T. Hashimoto, 2002 Microtubule basis for left-handed helical growth in Arabidopsis. Nature 417: 193–196. Traub-Cseko, Y. M., J. M. Ramalho-Ortigao, A. P. Dantas, S. L. de Castro, H. S. Barbosa et al., 2001 Dinitroaniline herbicides against protozoan parasites: the case of Trypanosoma cruzi. Trends Parasitol. 17: 136–141. Vogel, G., 2006 Malaria. Chloroquine makes a comeback. Science 314: 904. Wang, H. W., and E. Nogales, 2005 Nucleotide-dependent bending flexibility of tubulin regulates microtubule assembly. Nature 435: 911–915. Wang, H. W., S. Long, K. R. Finley and E. Nogales, 2005 Assembly of GMPCPP-bound tubulin into helical ribbons and tubes and effect of colchicine. Cell Cycle 4: 1157–1160. Wang, Y., S. Veeraraghavan and F. Cabral, 2004 Intra-allelic suppression of a mutation that stabilizes microtubules and confers resistance to colcemid. Biochemistry 43: 8965–8973.

Communicating editor: S. Dutcher