Gliding motility is a dynamic event involving temporally and spatially controlled actin polymerization. Typically, formins assisted by profilin, bind to an FH1 domain to facilitate a rapid processive assembly of actin filaments 
. Despite the importance of TgPRF in parasite motility and invasion 
, none of the parasite formins carry a canonical FH1 domain. Since the region right upstream of FH2 domains of both formins is rich in short stretches of proline residues their potential to act as binding site for profilin was investigated but in vitro
pull-down experiments and in vivo
coIPs failed to show interaction. New subclasses of formins apparently lack FH1, suggesting that an FH1-independent pathway may mediate actin assembly 
. The crystal structure of the Plasmodium PRF in complex with an octa-proline peptide was solved and implicated the N-terminal tyrosine residue (Tyr5) in tethering to the poly-proline 
. In our hands, all apicomplexan PRFs showed either very low or no affinity for poly-proline (Plattner F., unpublished). It is plausible that the apicomplexan PRFs bind to a divergent unrecognizable domain on formins and further work is required to unravel how TgPRF contributes to actin filament formation.
TgFRM1 and TgFRM2 bind to TgACT1 and share biochemical characteristics with their counterparts in Plasmodium 
. Both formins are underneath the PM and aerolysin treatment revealed their differential affinity to membranes within the narrow space separating the IMC from the PM. Like TgMyoA, both formins homogenously distribute at the periphery of invading parasites, compatible with the notion that they may nucleate actin at any time and at any point of contact between the parasite and its substrate.
Conditional knockout of TgFRM1 established its role in motility and invasion although the effects were modest and impact on egress was minor. Despite multiple attempts, generation of a conditional knockout for TgFRM2 failed. In the course of this study, parasite lines were generated with triple Ty-1 tags inserted at the C-terminus of each formin by single crossing over in the ku-80-ko strain. RT-PCR analysis confirmed in frame integration of the tags but no signal was detectable by IFA or Western blot in these transgenic parasites (data not shown). These results indicate that the endogenous levels of both formins are extremely low and hence explain the weak phenotype observed upon mycFRM1i depletion. In the same context, the lack of success in replacing the endogenous TgFRM2 promoter with an inducible promoter might be due to a deleterious effect of mycFRM2i expression if the level is too high.
The function of TgFRM2 and possible redundancy with TgFRM1 was assessed by the expression of FH2 mutants to poison individually or simultaneously the two endogenous formins. This strategy also showed some limitations since the co-IP experiments revealed that 30% of FRM1 and 17% of FRM2 were not sequestered in defective heterodimers. This suggests that the affinity and or the stability of the homodimers (FRM-FRM and FH2-FH2) are higher than the heterodimer (FRM-FH2).
The FH2 WT, F1 and F2 are potent actin nucleators and their overexpression had a severe impact on parasite replication that was not dependent on TgFRM1 and TgFRM2. Points mutations were introduced in the FH2 domains to disrupt actin nucleation and hence eliminate this non-specific effect. As with Bni1p 
, a single point mutation in TgFRM2 (F2-R/A) was sufficient to abrogate its activity whereas a double mutation (F1-IR/AA) was needed to abolish actin nucleation of TgFRM1. The IR/AA double mutation conferred to F2 an unexpected barbed end capping activity. This mutation may impair the flexibility of the FH2 domain and prevent the switch from the closed to the open configuration during elongation. To understand this phenomenon, the resolution of the FH2 domain structure of TgFRM2 in presence of actin would be necessary.
The different effects of the R/A, and IR/AA mutations on the activities of TgFRM1 and TgFRM2 further testify that these two formins have different modes of interaction with actin. The fact that mutations affect differently barbed end growth and depolymerization processes, in which ATP/ADP-Pi-actin and ADP-actin are respectively exposed at barbed ends, suggests that these mutations may affect their interactions with ATP-actin and ADP-actin differently. Similar differences have already been observed with twinfilin, a capping protein that binds preferentially to ADP-bound barbed ends 
. The R/A mutation does not affect any of the activities of TgFRM1. In contrast, the R/A mutation of TgFRM2 may weaken its interaction with ATP-actin but not with ADP-actin. The IR/AA double mutation abolishes all activities of TgFRM1. The same double mutation transforms TgFRM2 into a strong barbed end capper in nucleation and barbed end growth, while leaving the barbed end depolymerization unaffected, which suggests that the double mutation reinforces binding of FRM2 to the ATP-terminal subunits in its “closed” configuration and abolishes its binding to ADP-terminal subunits.
Stabilization of DD-F1-IR/AA and DD-F2-R/A did not affect intracellular growth and revealed that both TgFRM1 and TgFRM2 play a role in gliding, invasion and egress. All phenotypes were aggravated when both dominant mutants were expressed in the same parasite (). These results give a strong indication that the two formins act in concert. However, since the stabilization of each FH2 mutants failed to sequester all the formins, invasion only dropped to 50% and in consequence it is not possible to completely rule out some level of functional redundancy between the two formins. Nevertheless the results demonstrate that both TgFRM1 and TgFRM2 contribute additively to the three vital aspects of the glideosome function namely gliding motility, host cell invasion and egress from the infected cells.
The refined analysis of the gliding motility phenotypes by video microscopy revealed that interfering with TgFRM1 and TgFRM2 preferentially affected helical and circular gliding, respectively, illustrating distinct contributions of the two formins in gliding. This study revealed that TgFRM1 is preferentially positioned at the PM, where fast nucleation occurs in close proximity to the complex formed between actin filaments and the aldolase-MIC2 tail complex. The filaments likely elongate over only a short distance with TgFRM2 potentially serving to stabilize and control the size of the filament close to the IMC.
Given the importance of these formins for parasite infection, it will be imperative to elucidate their mode of regulation and interaction with profilin as these unique features might become relevant therapeutic targets.