Structure-guided development of selective caseinolytic protease P agonists as antistaphylococcal agents

Abstract

Methicillin-resistant Staphylococcus aureus is a ubiquitous pathogen, posing a serious threat to human health worldwide. Thus, there is a high demand for antibiotics with distinct targets. Caseinolytic protease P (ClpP) is a promising target for combating staphylococcal infections; however, selectively activating S. aureus ClpP (SaClpP) rather than Homo sapiens ClpP (HsClpP) remains challenging. Herein, we rationally design and identify ZG297 by structure-based strategy. It binds and activates SaClpP instead of HsClpP. This is due to differentiated ligand binding attributed to crossed "tyrosine/histidine" amino acid pairs. ZG297 substantially inhibits the growth of a broad panel of S. aureus strains in vitro, outperforming the selective (R)-ZG197 agonist. ZG297 also functions as a potent antibiotic against multidrug-resistant S. aureus infections in Galleria mellonella larvae, zebrafish, murine skin, and thigh infection models. Collectively, we demonstrate that ZG297 is a safer and more potent antistaphylococcal agent than acyldepsipeptide 4 and (R)-ZG197.

Keywords: ClpP; MRSA; ZG297; antibiotic; antistaphylococcal agents; caseinolytic protease P; methicillin-resistant Staphylococcus aureus; murine skin infection model; murine thigh infection model; selective agonist; structure guided design; tyrosine/histidine pairs.

Graphical abstract

Figure 1

Rational design of selective SaClpP agonists (A) Structure-based design of selective SaClpP agonists rather than HsClpP in the previous study. (B) Close view of the naphthyl group involved in the binding of ZG180 to the hydrophobic pockets of SaClpP and HsClpP. The structures of the ZG180/SaClpP (PDB: 7WID) and ZG180/HsClpP (PDB: 7WH5) complexes were used, with the color yellow representing ZG180. The surfaces of SaClpP and HsClpP are colored gray and pale green, respectively. (C) Volumes and surface areas for the naphthyl group binding in the hydrophobic pockets of SaClpP and HsClpP, abstracted from the structural complexes of ClpP bound to ZG180. (D) Structures of ZG180 derivatives with different groups at the R position. (E) Effect of ZG180 derivatives on SaFtsZ degradation by SaClpP or HsClpP. The EC₅₀ values (μM) and selectivity windows are provided. (F and G) Quantification of the accelerated effect of ZG180 derivatives ONC212 and ADEP 4 on FITC-casein hydrolysis by SaClpP (F) and HsClpP (G) (n = 3 biological replicates). Mean ± SD (error bars) are shown. See also Figure S1.

Figure 2

Interactions of ZG283 and ZG297 toward SaClpP in vitro (A) Effect of the agonists on the shifts in melting temperatures of SaClpP and HsClpP, determined using the DSF assay (n = 3 biological replicates). Mean ± SD (error bars) are shown. (B) Representative western blot images showing the effect of agonists on the thermal stability of SaClpP. CETSA was performed on intact bacterial cells of the S. aureus 8325-4 strain in the presence of 10 μM compound for 2 h. (C) ΔT_m values revealing the effect of agonists on the thermal stability of cellular SaClpP (n = 3 biological replicates). Mean ± SD (error bars) are shown. (D) Representative western blot images showing the effect of agonists on the thermal stability of HsClpP. CETSA was conducted on intact HK-2 cells with or without treatment with 10 μM compound for 2 h. (E) ΔT_m values revealing the effect of agonists on the thermal stability of cellular HsClpP (n = 3 biological replicates). Mean ± SD (error bars) are shown. See also Figure S2.

Figure 3

X-ray crystal structure of SaClpP bound with ZG297 (A) Overall structure of SaClpP bound to ZG297 (PDB: 9IRP). Top and side views of the agonist-bound SaClpP complex are presented. The surface of SaClpP is colored gray, and ZG297 is shown as cyan spheres. (B) 2fo-fc map contoured at 3.5σ showing continuous electron densities for atoms of ZG297 in the hydrophobic pocket of SaClpP. (C) Stereo view of the interactions between ZG297 and SaClpP in the hydrophobic pocket. Two images from slightly different angles are displayed. Cyan stick represents ZG297. Dark dashed lines represent hydrogen bonds. Chains A and G are labeled gray and yellow, respectively. Key amino acids involved in interactions with ZG297 are shown as sticks. (D) Structural alignment of ZG297 in the hydrophobic pocket of SaClpP (left) and HsClpP (right). The protein surface is colored gray. ZG297 is colored cyan. The diffused electron densities of ZG297 are shown in dots. See also Figure S3 and Table S1.

Figure 4

Identification of key residues responsible for selective activation of SaClpP rather than HsClpP by ZG297 (A) Sequence alignment and structural superimposition of ZG297-bound SaClpP and HsClpP. Residues in SaClpP and HsClpP are shown as gray and light pink sticks, respectively. (B) Quantification of the effect of ZG297 on the hydrolysis of SaFtsZ by SaClpP and Y61A and H83Y SaClpP mutants, detected using SDS-PAGE (n = 3 biological replicates). Mean ± SD (error bars) are shown. (C) Quantification of the effect of ZG297 on the hydrolysis of SaFtsZ by HsClpP and H116Y and Y138H HsClpP mutants, detected using SDS-PAGE (n = 3 biological replicates). Mean ± SD (error bars) are shown. (D) Representative western blot images showing the effect of ClpP agonists on the degradation of cellular SaFtsZ in cell lysates of the ΔclpP S. aureus strain supplemented with or without ClpP variants as indicated. See also Figure S4.

Figure 5

ZG297 exerts SaClpP-dependent antistaphylococcal activity (A) Representative western blot images showing the effect of ZG297 on the abundance of cellular SaFtsZ in the intact cells of 8325-4, ΔclpP, and clpP-complemented S. aureus strains. (B) Representative images showing the effect of ZG297 on cell diameters of 8325-4, ΔclpP, and clpP-complemented S. aureus strains, detected using SEM. S. aureus cells were treated with 5 μg/mL ZG297 for 3 h. The scale bar represents 1 μm. (C) MIC (μg/mL) values of SaClpP agonists and antibiotics for the inhibition of the growth of a panel of S. aureus strains. The clpP-complemented strain is insensitive to erythromycin because the complemented pYJ335 plasmid contains an erythromycin-resistant gene. See also Figure S5.

Figure 6

Antistaphylococcal infection therapy with ZG297 in vivo (A) Therapeutic windows of ClpP agonists. Therapeutic windows were calculated by dividing the MTT value by the MIC value. Six mammalian cell lines were used in the MTT assay, while the S. aureus USA300 strain was used in the MIC assay. (B) Effect of ONC212 and ZG297 on cellular ATP levels in HK-2 (left) and HEK 293T/17 (right) cells (n = 3 biological replicates). Mean ± SD (error bars) are shown. (C) Effect of agonists on the abundance of the protein of the mitochondrial respiratory chain complex in HK-2 cells, detected using western blotting. (D) Therapeutic effect of compounds on USA300 and XJ049 infection in the G. mellonella larvae model. Larvae (n = 16 animals) were infected with 1 × 10⁶ CFU of USA300 or XJ049, followed by treatment with a single dose of ZG297, erythromycin, or oxacillin. (E) Therapeutic effect of compounds on USA300 and XJ049 infection in the zebrafish model. Zebrafish (n = 11 animals) were infected with 2 × 10⁷ CFU of USA300 or 1 × 10⁸ CFU of XJ049, followed by treatment with a single dose of 5 mg/kg of compounds. (F) Necrotic skin lesion size at day 0 (left) and four days post infection (right). BALB/c mice (n = 6 animals) were infected with 2 × 10⁷ CFU of XJ049 and treated with 5 mg/kg of compounds twice a day for three days. Mean ± SEM (error bars) are shown. (G) Bacterial load in skin samples at the end of the experiment (n = 6 animals). Skin tissue was excised from mice, and the homogenate was plated onto TSA for measurement. Mean ± SEM (error bars) are shown. (H) Representative H&E-stained images showing infected skin tissues at the end of the experiment. The scale bar represents 500 μm. (I) Therapeutic effects of ClpP agonists and the conventional antibiotics on deep-seated chronic infections in a murine thigh infection model. ICR (CD1) mice (n = 7 animals) were infected with 2 × 10⁶ CFU of stationary-phase S. aureus USA300. Deep-seated, infected muscle tissue was excised from mice, and homogenates were plated onto TSA for measurement. Mean ± SEM (error bars) are shown. The p value was determined using the two-tailed unpaired Student’s t test (B, F, G, and I) and the Mantel-Cox test (D and E). ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001, ns: not significant. See also Figure S6 and Tables S2 and S3.