Inhibitors of ApiAP2 protein DNA binding exhibit multistage activity against Plasmodium parasites

Abstract

Plasmodium parasites are reliant on the Apicomplexan AP2 (ApiAP2) transcription factor family to regulate gene expression programs. AP2 DNA binding domains have no homologs in the human or mosquito host genomes, making them potential antimalarial drug targets. Using an in-silico screen to dock thousands of small molecules into the crystal structure of the AP2-EXP (Pf3D7_1466400) AP2 domain (PDB:3IGM), we identified putative AP2-EXP interacting compounds. Four compounds were found to block DNA binding by AP2-EXP and at least one additional ApiAP2 protein. Our top ApiAP2 competitor compound perturbs the transcriptome of P. falciparum trophozoites and results in a decrease in abundance of log2 fold change > 2 for 50% (46/93) of AP2-EXP target genes. Additionally, two ApiAP2 competitor compounds have multi-stage anti-Plasmodium activity against blood and mosquito stage parasites. In summary, we describe a novel set of antimalarial compounds that interact with AP2 DNA binding domains. These compounds may be used for future chemical genetic interrogation of ApiAP2 proteins or serve as starting points for a new class of antimalarial therapeutics.

Fig 1. Nine putative competitors of DNA binding by AP2-EXP were selected by an in-silico screen and tested in vivo and in vitro.

A) Each compound that was prioritized based on the in-silico screen was assigned an identifier (A-I) and tested for anti-Plasmodium activity in a 48-hour growth inhibition assay against asexual blood stage P. falciparum. B) Each putative AP2-EXP competitor was added to an EMSA containing the AP2-EXP AP2 domain. Activity against AP2-EXP will result in a loss of the shifted DNA probe. DMSO vehicle was used as a control for normal DNA binding by AP2-EXP. 150 fmoles of DNA probe, 125ng of AP2-EXP, and 300μM of each compound were used for each lane. C) Chemical structures of Compounds A, B, C, and I. Three of the four compounds that prevent AP2-EXP DNA binding in vitro (Compounds A, B, and C) have a benzoxazole core moiety, denoted by a red box.

Fig 2. Analogues of Compound B prevent AP2-EXP DNA binding with differing efficacy.

A) Compounds B, B-1, B-2, B-3, and B-4 were added to an EMSA with AP2-EXP. Activity against AP2-EXP will result in the loss of the shifted probe. 40 fmoles of DNA probe, 125ng of AP2-EXP, and 300μM of each compound were used for each lane. B) Compounds B, B-1 and B-4 were titrated at 25, 50, and 150μM into an EMSA with AP2-EXP to determine whether there are differences in efficacy against AP2-EXP. DMSO vehicle was used as a control for normal DNA binding. 40 fmoles of DNA probe and 125ng of AP2-EXP were used for each lane. The concentration increase for each compound is indicated by the triangle from left to right. C) Chemical structures of Compounds B, B-1, B-2, B-3, and B-4. Compounds that prevent DNA binding by AP2-EXP are highlighted in black. Each compound has a different substitution pattern on the right-side benzene ring, denoted by a red box. The IC₅₀ against asexual P. falciparum of compounds B-1, B-2, B-3, and B-4 measured by Gamo et al [42] is indicated at the bottom of each identifier. The IC₅₀ of Compound B was determined in this study.

Fig 3. Compounds B and C affect P. falciparum in the mid trophozoite stage, coinciding with the maximum expression of AP2-EXP.

A-D) Highly synchronous wild type Pf3D7 parasites at 10% starting parasitemia were spiked with DMSO vehicle control (A), 40μM Compound B (B), 40μM Compound C (C), or 40μM Compound H (D) and parasite staging was recorded at 6-hour intervals throughout the IDC. Each growth assay was performed in technical and biological triplicate. Error bars represent standard deviation of the mean. The color key indicates the asexual blood stage morphologies recorded at each time point. E) Representative Giemsa-stained parasite images corresponding to the growth assay from panels (A-D). F) Highly synchronous asexual blood stage wild type Pf3D7 parasites at 1% starting parasitemia were dosed with 40μM Compound C for 8-hour intervals starting at 6 hpi. Compound C was removed following each dose interval. At 54 hpi, the ratio of reinvaded rings to stalled trophozoites was recorded. DMSO was used as a vehicle control for normal growth. Each assay was performed in biological and technical triplicate. Error bars represent standard deviation of the mean. Significance of the difference between the DMSO control and Compound C dosed values was tested using the Mann Whitney U-test. G) AP2-EXP protein expression in the asexual blood stages ring (R), trophozoite (T), and schizont (S), and Stage III gametocytes (Gam), was probed by western blot against AP2-EXP endogenously tagged with GFP (AP2-EXP::GFP). The expected molecular weight of AP2-EXP::GFP (147kDa) is indicated by an arrow. Wild type Pf3D7 protein was used as a negative control. The full-length western blot is provided in S10B Fig.

Fig 4. Compound C disrupts the P. falciparum transcriptome specifically at 30 hpi, with bias towards AP2-EXP target genes predicted by ChIP-seq.

A) Highly synchronous wild type Pf3D7 parasites were spiked with 12μM (0.66xIC₅₀) Compound C or DMSO vehicle control at 10 hpi. Total RNA was harvested at 6, 12, 18, 24, 30, 42, and 48 hpi for quantification by DNA microarray. The full transcriptome Spearman Correlation between Compound C and DMSO control spiked parasites was plotted as a correlogram. An asterisk indicates p-value of correlation < 0.05. B) Three replicates of AP2-EXP Chip-seq were collected at 30 hpi using two genetically tagged parasite lines (AP2-EXP::GFP and AP2-EXP::HA). A no epitope control was collected using wild type Pf3D7 parasites and the anti-GFP antibody. Log₂ enrichment of the immunoprecipitate (ChIP) over Input DNA for one replicate of AP2-EXP::GFP, AP2-EXP::HA, and No Epitope Control is plotted relative to the coding sequence start site of AP2-EXP target genes identified in at least 2/3 replicates. CATGCA is the most overrepresented DNA motif within the peaks of AP2-EXP occupancy conserved in at least 2/3 ChIP-seq experiments. C) A volcano plot of the changes in abundance for AP2-EXP target genes at 24–30 hpi in the presence of Compound C. 46/93 detected AP2-EXP target transcripts decrease in abundance by log₂ fold change > 2 (Adjusted p-value cutoff < 0.1). D) GO-term analysis of transcripts that are both AP2-EXP targets and decreased in abundance with respect to Compound C (Bonferroni p-value cutoff < 0.05).

Fig 5. Compounds B and C are active against mosquito stage P. berghei parasites.

A) Compounds B, C and F were added to an EMSA with PbAP2-Sp. Activity against PbAP2-Sp will result in the loss of the shifted probe. 150 fmoles of DNA probe, 125ng of PbAP2-Sp, and 300μM of each compound were used for each lane. DMSO vehicle was used as a control for normal DNA binding by PbAP2-Sp. B) Midgut oocyst counts per mosquito for P. berghei infected mosquitoes on day 14 post infection following compound injection on day 10 post infection. Compound identity is indicated in the legend. Rep. 1 and Rep. 2 correspond to biological replicates one and two, respectively. Significance of difference between DMSO vehicle injected samples and compound injected samples was assessed using the Mann Whitney U-test. Bars indicate which compound injected dataset has been compared to the DMSO control dataset. Error bars represent the standard deviation of the mean. C) Midgut sporozoite counts per five infected mosquitoes on day 14 post infection following compound injection on day 10 post infection. Compound identity is indicated in the legend. Rep. 1 and Rep. 2 correspond to biological replicates one and two, respectively. Significance of difference between DMSO vehicle injected samples and compound injected samples was assessed using the Mann Whitney U-test. Bars indicate which compound injected dataset has been compared to the DMSO control dataset. Error bars represent the standard deviation of the mean.