Distinct substrate specificities of the three catalytic subunits of the Trichomonas vaginalis proteasome

Abstract

The protozoan parasite Trichomonas vaginalis (Tv) causes trichomoniasis, the most common non-viral sexually transmitted infection in the world. Although Tv has been linked to significant health complications, only two closely related 5-nitroimidazole drugs are approved for its treatment. The emergence of resistance to these drugs and lack of alternative treatment options poses an increasing threat to public health, making development of novel anti-Trichomonas compounds an urgent need. The proteasome, a critical enzyme complex found in all eukaryotes has three catalytic subunits, β1, β2, and β5 and has been validated as a drug target to treat trichomoniasis. With the goal of developing tools to study the Tv proteasome, we isolated the enzyme complex and identified inhibitors that preferentially inactivate either one or two of the three catalytic subunits. Using a mass spectrometry-based peptide digestion assay, these inhibitors were used to define the substrate preferences of the β1, β2 and β5 subunits. Subsequently, three model fluorogenic substrates were designed, each specific for one of the catalytic subunits. This novel substrate profiling methodology will allow for individual subunit characterization of other proteasomes of interest. Using the new substrates, we screened a library of 284 peptide epoxyketone inhibitors against Tv and determined the subunits targeted by the most active compounds. The data show that inhibition of the Tv β5 subunit alone is toxic to the parasite. Taken together, the optimized proteasome subunit substrates will be instrumental for understanding the molecular determinants of proteasome specificity and for accelerating drug development against trichomoniasis.

Keywords: drug discovery; drug screening; parasite; protease inhibitor; proteasome; substrate specificity; trichomonas.

FIGURE 1

Gel electrophoresis‐based characterization of Tv20S. (a) Tv lysate, isolated Tv20S, and human c20S were resolved on a non‐denaturing NuPAGE 3%–8% tris‐acetate gel, silver stained and imaged. (b) The same protein samples were incubated for 1 h with 2 μM of the fluorescent Me4BodipyFL‐Ahx3Leu3VS probe and imaged with 470 nm excitation and 530 nm emission wavelengths. (c) Tv lysate was incubated with 2 μM of the fluorescent probe for the indicated times and the subunits were resolved by electrophoresis using a NuPAGE 12% bis‐tris denaturing gel.

FIGURE 2

Subunit‐specific inhibition of Tv20S by carmaphycin‐17 (CP‐17) and carfilzomib (CFZ). Tv20S was preincubated with 1–5 μM of CP‐17 (a) or CFZ (c) for 1 h prior to labeling with the Me4BodipyFL‐Ahx3Leu3VS probe and size separation by SDS‐PAGE gel electrophoresis. The bands from (a) and (c) were respectively quantified with ImageJ software (b, d), and the intensities were plotted relative to the band intensity in vehicle (DMSO) treated lanes. Three independent assays were performed using the same preparation of Tv20S. Results represent means and SEM of the triplicate assays and * represents band intensity that is significantly changed relative to the non‐inhibitor treated control.

FIGURE 3

Characterization of peptides cleaved by Tv20S. (a) Distribution of cleavage sites within the 14‐mer peptides. (b) IceLogo frequency plot of the peptides cleaved by Tv20S showing the amino acids that are increased (above mid‐line) and decreased (below mid‐line) in the P4 to P4′ positions. The red arrow indicates the cleavage site. Lowercase n in the plot designates norleucine. (c) Abundance of individual cleaved peptides is expressed as a percentage relative to their abundance in DMSO treated controls. Cyan colors designate peptides whose cleavage was decreased by the indicated inhibitors, while purple colors depict peptides that were more abundant in the presence of the inhibitors compared to DMSO controls. Cleavage products clustered into three main groups, designated as 1 to 3.

FIGURE 4

Design of a Tv20S β5‐specific substrate. (a) IceLogo frequency plot showing the amino acids that are increased (above mid‐line) and decreased (below mid‐line) in the P4 to P4′ positions of the Tv20S β5 subunit. Lowercase n corresponds to norleucine. (b) Example peptide from the MSP‐MS substrate library that is cleaved by Tv20S and whose cleavage is inhibited by 1 μM CP‐17 (mean ± SE, n = 4 replicates). The N‐terminal cleavage product in bold was identified by LC–MS/MS. (c) Michaelis Menten plots with the indicated fluorogenic substrates (mean ± SE, n = 3 replicates).

FIGURE 5

Design of a β2‐specific substrate. (a) IceLogo frequency plot showing the amino acids that are increased (above mid‐line) and decreased (below mid‐line) in the P4 to P4′ positions of the Tv20S β2 subunit. Lowercase n corresponds to norleucine. (b) Example peptide that is cleaved by Tv20S, partially inhibited by 1 μM CP‐17, and fully inhibited by 10 μM each of CP‐17 and CFZ. The cleavage product in bold was identified via LC–MS/MS (mean ± SE, n = 4 replicates). (c) Michaelis Menten plot with the substrates Ac‐FRSR‐amc and Boc‐LRR‐amc (mean ± SE, n = 3 replicates).

FIGURE 6

Design of a β1‐specific substrate. (a) IceLogo frequency plot showing the amino acids that are increased (above mid‐line) and decreased (below mid‐line) in the P4 to P4′ positions of cleaved peptides in Group 3. Lowercase n corresponds to norleucine. (b) Example peptide product that is cleaved independent of CP‐17 and/or CFZ (n = 4). (c) Enzyme activity using Ac‐PSRN‐amc and the legumain substrate Z‐AAN‐amc (n = 3). (d) Example peptide product that is cleaved by Tv20S β1 where activity increases in the presence of CP‐17 and CFZ (mean ± SE, n = 4 replicates). (e) Michaelis Menten plot of Ac‐RYFD‐amc and Z‐LLE‐amc (mean ± SE, n = 3 replicates). (f) Enzyme activity using Ac‐RYFD‐amc and Z‐LLE‐amc following a 30 min preincubation of Tv20S with the indicated inhibitors (mean ± SE, n = 3 replicates).

FIGURE 7

Validation of subunit‐selective substrates. Concentration‐response curves for Tv20S subunit inhibition by (a) CP‐17, (b) CFZ, (c) leupeptin, and (d) ixazomib, with activity normalized to a 0.025% DMSO control. Assays were performed using 30 μM Ac‐GWYL‐amc (β5), 30 μM Ac‐FRSR‐amc (β2) or 50 μM Ac‐RYFD‐amc (β1). Fluorescence was measured over 2 h and the velocity of each reaction was plotted to generate concentration‐response curves (data points show mean ± SE, n = 3 replicates). IC₅₀ values were calculated by fitting to a concentration–response curve using GraphPad Prism 9.

FIGURE 8

Inhibition of Tv growth by two β5 subunit specific proteasome inhibitors. (a) Whole‐cell activity (EC₅₀) screen of 284 proteasome inhibitors comprising the Kezar Life Sciences library against Tv. Each data point represents the mean pEC₅₀ for one compound (n = 3 biological replicates). The top 2 most active compounds are highlighted in red. (b) Chemical structures of KZR‐58100 and KZR‐58165. Subunit staining by a 24 h incubation with Me4BodipyFL‐Ahx3Leu3VS following pre‐incubation of Tv lysate with KZR‐58100 and KZR‐58165. The sample was then size‐separated by SDS‐PAGE. (c, d) Concentration‐response curves for Tv20S subunit inhibition by KZR‐58100 and KZR‐58165. Activity was normalized to a 0.025% DMSO control. IC₅₀ values were calculated by fitting to a dose–response curve using GraphPad Prism 9 (mean ± SE, n = 3 replicates).