Mutant p53 Depletion by Novel Inhibitors for HSP40/J-Domain Proteins Derived from the Natural Compound Plumbagin

Abstract

Accumulation of missense mutant p53 (mutp53) in cancers promotes malignant progression. DNAJA1, a member of HSP40 (also known as J-domain proteins: JDPs), is shown to prevent misfolded or conformational mutp53 from proteasomal degradation. Given frequent addiction of cancers to oncogenic mutp53, depleting mutp53 by DNAJA1 inhibition is a promising approach for cancer therapy. However, there is no clinically available inhibitor for DNAJA1. Our in silico molecular docking study with a natural compound-derived small molecule library identified a plumbagin derivative, PLIHZ (plumbagin-isoniazid analog), as a potential compound binding to the J domain of DNAJA1. PLIHZ efficiently reduced the levels of DNAJA1 and several conformational mutp53 with minimal impact on DNA contact mutp53 and wild-type p53 (wtp53). An analog, called PLTFBH, which showed a similar activity to PLIHZ in reducing DNAJA1 and mutp53 levels, inhibited migration of cancer cells specifically carrying conformational mutp53, but not DNA contact mutp53, p53 null, and wtp53, which was attenuated by depletion of DNAJA1 or mutp53. Moreover, PLTFBH reduced levels of multiple other HSP40/JDPs with tyrosine 7 (Y7) and/or tyrosine 8 (Y8) but failed to deplete DNAJA1 mutants with alanine substitution of these amino acids. Our study suggests PLTFBH as a potential inhibitor for multiple HSP40/JDPs.

Keywords: DNAJA1; HSP40; inhibitor; mutant p53; natural compound.

Figure 1

Knockdown of DNAJA1 specifically reduces protein levels of conformational mutp53, but not DNA contact mutp53 and wtp53. (A,B) Western blotting for p53, DNAJA1, and GAPDH (A) and immunofluorescence for p53, DNAJA1, and DAPI (B) using multiple cancer cells with different p53 status with or without knockdown of DNAJA1 (JA1) and p53 (p53). Scale bar: 50 µm.

Figure 2

Identification of a compound that binds DNAJA1 to specifically reduce conformational mutp53. (A) Chemical structure of PLIHZ compound, derived from plumbagin, which was identified through a molecular docking (left). Ribbon and crystal structures of DNAJA1 protein (PDB: 2LOI) showing binding of DNAJA1 to PLIHZ at tyrosine residue Y7 (right) with a 2.6 Å bond distance and binding free energy of −6.6 kcal/mol. (B) CETSA showing intracellular binding of PLIHZ to DNAJA1. A representative Western blotting for DNAJA1 and GAPDH using protein extracts from CAL33 (p53^R175H) cells with treatment with DMSO and PLIHZ at 80 µM for 4 h, followed by incubation at different temperatures for 3 min (left). A representative blot using protein extracts from unheated cells are also shown. A summarized graph showing normalized DNAJA1 band densities at different temperatures of 40, 43, 46, 49, 52, 55, and 58 °C (right). Mean ± SEM from three independent experiments (n = 3). ** p < 0.01 for two-way ANOVA. (C) Western blotting for p53, DNAJA1, and GAPDH using indicated cells with different p53 status, treated with DMSO or PLIHZ at ~1/2 of 24h-IC50 for 24 h. (D) Immunofluorescence for p53, DNAJA1, and DAPI using indicated cells treated with DMSO or PLIHZ at ~1/2 of 24h-IC50 for 24 h.

Figure 3

PLTFBH, an analog of PLIHZ, specifically reduces conformational mutp53 levels similar to PLIHZ. (A) Chemical structures of PLIHZ analogs including PLIHZ, PLFBH, PLTFBH, PLFUH, and PLOCT. (B) Western blotting for p53 and GAPDH using HN31 cells treated with different PLIHZ analogs (40 µM for 24 h). (C) Western blotting for p53, DNAJA1, and GAPDH using HN31, MDA-MB-231, and U2OS cells treated with PLTFBH at ~1/2 of 24h-IC₅₀. (D) Immunofluorescence for p53, DNAJA1, and DAPI using HN31, MDA-MB-231, U2OS, and H1299 cells treated with PLTFBH at ~1/2 of 24h-IC₅₀. Scale bar: 50 µm. (E) CETSA showing intracellular binding of PLTFBH to DNAJA1 in CAL33 cells: a representative Western blotting for DNAJA1 and GAPDH using protein extracts from HN31 cells (left); a summarized graph showing normalized DNAJA1 band densities at different temperatures (right). Mean ± SEM from three independent experiments (n = 3). * p < 0.05 for two-way ANOVA. (F) Western blotting for p53, DNAJA1, and GAPDH using HN31 cells with or without knockout for DNAJA1 (JA1) or p53 (p53) using the CRISPR-Cas9 strategy. (G) Summary of MTT assays using control, DNAJA1 knockout, or p53 knockout HN31 treated with different concentrations of PLTFBH for 72 h. Mean ± SEM from three independent experiments (n = 3). n.s.: not significant for one-way ANOVA. IC₅₀ values of PLTFBH for each sub-cell line are shown on the right.

Figure 4

PLTFBH inhibits migratory potential of cancer cells in a manner dependent on DNAJA1 and conformational mutp53. (A) Transwell migration assays using indicated cells with different p53 status treated with PLTFBH at ~1/2 IC₅₀ for 12 h. All cells were pre-treated with PLTFBH for 12 h, followed by trypan blue staining and transwell migration assays. Top: summarized graphs. Bottom: representative images. Mean ± SEM from three independent experiments (n = 3). *** p < 0.001 for two-tailed Student’s t-test. n.s.: not significant. Scale bar: 100 µm. (B) F-actin staining showing inhibition of filopodia formation in HN31 cells, but not MDA-MB-231, U2OS, and H1299 cells, by PLTFBH (TF). Top: summarized graph. Bottom: representative images. Mean ± SEM from three independent experiments (n = 3). **** p < 0.0001 for two-tailed Student’s t-test. n.s.: not significant. Scale bar: 10 µm. (C) Transwell migration assays using DNAJA1- or p53-knockdown HN31 cells treated with PLTFBH at ~1/2 IC₅₀ for 12 h. Cells were pre-treated with PLTFBH for 12 h. Mean ± SEM from three independent experiments (n = 3). ** p < 0.01 for two-tailed Student’s t-test. n.s.: not significant. Scale bar: 100 µm. (D) Rac1/Cdc42 activation assays following pulldown of active Rac1 and Cdc42 using protein extracts from HN31 cells treated with DMSO (D) or PLTFBH (TF) at ~1/2 of 24h-IC₅₀. Left: representative immunoblots. Right: summarized graph. Mean ± SEM from three independent experiments (n = 3). * p < 0.05, *** p < 0.001 for two-tailed Student’s t-test.

Figure 5

PLTFBH selectively decreases protein levels of certain members of HSP40/JDPs. (A) Western blotting for several members of HSP40/JDPs and GAPDH using HN31 cells treated with DMSO (D) or PLFBH (TF) at ~1/2 IC50 for 24 h. (B) Three groups (good, moderate, little or no) of HSP40/JDPs based on their response to PLTFBH. Note that some blots were from the same membrane (DNAJB1 and DNAJB2; DNAJB6 and DNAJC6; DNAJC2 and DNAJC3), hence using the same GAPDH control. (C) Amino acid sequence alignment of the J domain of multiple HSP40/JDPs by centering the HPD sequence. (D) CETSA showing intracellular binding of PLTFBH with DNAJA4, but not DNAJC6, in CAL33 cells treated with 80 µM of PLTFBH (TF) for 4 h. Vinculin was used as a loading control. Summarized graphs showing normalized DNAJA4 and DNAJC6 band densities at different temperatures (right). Mean ± SEM from three independent experiments (n = 3). * p < 0.05 for two-way ANOVA.

Figure 6

Mutations at Y7 and Y8 residues in DNAJA1 abrogate the ability of PLTFBH to deplete DNAJA1 and conformational mutp53. (A) Western blotting to detect exogenous DNAJA1 and GAPDH, as well as endogenous p53 (p53^C176F), using DNAJA1-knockout HN31 cells (JA1KO) expressing exogenous wild-type (wt), Y7A mutant (Y7A), and Y8A mutant (Y8A) DNAJA1, treated with DMSO (D) or PLTFBH (TF) at ~1/2 IC₅₀ for 24 h. (B) Immunofluorescence for p53, DNAJA1, and DAPI using the same experimental set of HN31 sub-cell lines as in Figure 6A. Scale bar: 50 µm. (C) F-actin staining using the HN31 sub-cell lines treated with DMSO or PLTFBH at ~1/2 IC₅₀ for 24 h. Mean ± SEM from three independent experiments (n = 3). ** p < 0.01 for two-tailed Student’s t-test. n.s.: not significant. Scale bar: 10 µm. Left: representative images. Right: summarized graph.