Identification of T. gondii myosin light chain-1 as a direct target of TachypleginA-2, a small-molecule inhibitor of parasite motility and invasion

Abstract

Motility of the protozoan parasite Toxoplasma gondii plays an important role in the parasite's life cycle and virulence within animal and human hosts. Motility is driven by a myosin motor complex that is highly conserved across the Phylum Apicomplexa. Two key components of this complex are the class XIV unconventional myosin, TgMyoA, and its associated light chain, TgMLC1. We previously showed that treatment of parasites with a small-molecule inhibitor of T. gondii invasion and motility, tachypleginA, induces an electrophoretic mobility shift of TgMLC1 that is associated with decreased myosin motor activity. However, the direct target(s) of tachypleginA and the molecular basis of the compound-induced TgMLC1 modification were unknown. We show here by "click" chemistry labelling that TgMLC1 is a direct and covalent target of an alkyne-derivatized analogue of tachypleginA. We also show that this analogue can covalently bind to model thiol substrates. The electrophoretic mobility shift induced by another structural analogue, tachypleginA-2, was associated with the formation of a 225.118 Da adduct on S57 and/or C58, and treatment with deuterated tachypleginA-2 confirmed that the adduct was derived from the compound itself. Recombinant TgMLC1 containing a C58S mutation (but not S57A) was refractory to click labelling and no longer exhibited a mobility shift in response to compound treatment, identifying C58 as the site of compound binding on TgMLC1. Finally, a knock-in parasite line expressing the C58S mutation showed decreased sensitivity to compound treatment in a quantitative 3D motility assay. These data strongly support a model in which tachypleginA and its analogues inhibit the motility of T. gondii by binding directly and covalently to C58 of TgMLC1, thereby causing a decrease in the activity of the parasite's myosin motor.

Figure 1. The tachyplegin analogue tachypleginA-4 is covalently bound to the compound-induced, faster-migrating form of rTgMLC1.

(A) Structures and monoisotopic molecular weights (MW (mono)) of the tachyplegin analogues tachypleginA-2 (tA-2) and the alkyne-containing tachypleginA-4 (tA-4). (B) Infected Sf9 cells expressing recombinant FLAG-tagged wild-type TgMLC1 (rTgMLC1) were treated for 20 min with 100 µM tA-2, tA-4 or an equivalent amount of DMSO. Cell lysates were then prepared, labelled with biotin-azide and resolved and visualized by SDS-PAGE/western blotting. Both tA-2 and tA-4 induce a shift in the electrophoretic mobility of rTgMLC1 (anti-FLAG western blot, top panel: red arrowhead, modified rTgMLC1; blue arrowhead, unmodified rTgMLC1). However, only the lower form of tA-4-treated rTgMLC1 was labelled by streptavidin (streptavidin western blot, middle panel), indicating that the compound is exclusively bound directly and covalently to the faster-migrating form of rTgMLC1.

Figure 2. TachypleginA-2 treatment generates an adduct of 225.118 Da on S57 and/or C58 of rTgMLC1.

Low energy collision-induced dissociation MS/MS spectrum for the doubly-charged ion corresponding to a modified form of the tryptic V46-R59 peptide. This spectrum was averaged from three independent scans, and is representative of three independent experiments. S∼ and C∼ indicate serine and cysteine residues with a combined adduct mass of 225.118 Da. Coverage of the b- and y-ions in this modified peptide is indicated in green.

Figure 3. The modified V46-R59 peptide is quantitatively enriched in the faster-migrating, compound-induced form of rTgMLC1.

(A) The averaged light and heavy isotopic envelopes from the unmodified and modified (faster-migrating) forms of rTgMLC1, respectively, for the unmodified (left panel) and modified (right panel) V46-R59 peptide. Ratios for dimethyl labelled samples were generated by comparing the average relative abundances of the light vs. heavy monoisotopic peaks (filled and open stars, respectively). (B) The dimethyl labelling ratios (light:heavy) for the four most readily identifiable peptides, in addition to the two forms of the V46-R59 peptide in the unmodified and modified rTgMLC1 forms. The dimethyl labelling ratios for the unmodified and modified V46-R59 peptide (14.04 and < 0.13, respectively, underlined) were strikingly different from the ratio calculated using the other four peptides (2.24±0.27, average±standard deviation). Note that the ratio for the modified peptide was calculated using the abundance of a peptide (m/z = 864.742) closest to the expected m/z for the light form, since this unmodified form of rTgMLC1 peptide (calculated m/z = 863.371) could not be detected. Results are representative of two independent experiments.

Figure 4. Treatment with the heavy tachyplegin analogue D10-tachypleginA-2 increases the mass of the adduct by 5.03 Da.

(A) Structure and monoisotopic molecular weight (MW (mono)) of the heavy analogue D10-tachypleginA-2 (D10-tA-2). (B) Infected Sf9 cells expressing rTgMLC1 were treated for 20 min with 100 µM tA-2, D10-tA-2 or an equivalent amount of DMSO and samples were resolved by SDS-PAGE/western blotting with an anti-FLAG antibody. The unmodified and modified forms of rTgMLC1 are indicated with blue and red arrowheads, respectively. D10-tA-2 induces an electrophoretic mobility shift of rTgMLC1 similar to that observed upon treatment with tA-2. (C) Low energy collision-induced dissociation MS/MS spectrum for the doubly-charged ion corresponding to a modified form of the tryptic peptide V46-R59. This spectrum is averaged from twenty independent scans, and is representative of three independent experiments. S§ and C§ indicate serine and cysteine residues with a combined adduct mass of 230.149 Da, which corresponds to an increase in mass of five deuterium atoms. Coverage of the b- and y-ions in this modified peptide is indicated in green.

Figure 5. Recombinant TgMLC1 containing a C58S, but not S57A, mutation prevents covalent modification by tachypleginA-4.

Infected Sf9 cells expressing wild-type rTgMLC1 (rWT), or rTgMLC1 containing either a S57A (rS57A) or C58S (rC58S) mutation were treated for 20 min with 100 µM tA-2, tA-4 or an equivalent amount of DMSO. Cell lysates were then prepared, labelled with biotin-azide (see text for details) and resolved and visualized by SDS-PAGE/western blotting. The unmodified and modified forms are indicated with blue and red arrowheads, respectively. Treatment of rS57A with tA-2 or tA-4 resulted in a TgMLC1 electrophoretic mobility shift and labelling of the faster-migrating of TgMLC1 similar to that observed for rWT, as shown in the anti-FLAG and streptavidin blots, respectively. However, not only did rC58S not shift in response to tA-4 treatment, but no streptavidin signal was detected, indicating that tA-4 was not covalently bound to rC58S. The lanes shown were from the same blot, and were exposed and adjusted for brightness and contrast identically. Results shown are representative of three independent experiments.

Figure 6. Parasites expressing TgMLC1 with the C58S mutation are significantly less sensitive to the motility-inhibiting effect of tachypleginA-2.

(A) Maximum intensity projections (MIPs) for FLAG-TgMLC1-WT (WT) and FLAG-TgMLC1-C58S (C58S) knock-in parasites in a 3D motility assay, treated with the indicated concentrations of tA-2. Scale bar = 50 µm. The signal intensities in the MIPs were inverted for better visualization of parasite trajectories. (B-E) Graphs comparing the (B) percent moving, (C) mean trajectory length, (D) mean velocity and (E) maximum velocity of WT (white bars) and C58S (grey bars) knock-in parasites in the 3D motility assay. All values from compound-treated samples were normalized to those for DMSO; see Figure S6 for the non-normalized data. The total number of WT parasites analyzed was 7,123 for DMSO; 4,662 for 25 µM tA-2; 5,255 for 50 µM tA-2 and 4,328 for 100 µM tA-2. The total number of C58S parasites analyzed was 5,484 for DMSO; 3,325 for 25 µM tA-2; 4,587 for 50 µM tA-2 and 4,417 for 100 µM tA-2. Data shown are the results of three independent experiments, with each experiment performed in triplicate. Datasets were compared by two-way ANOVA (** p<0.001); error bars = standard deviation.