Bis(benzoyl) phosphate inactivators of beta-lactamase C from Mtb

Abstract

The class A β-lactamase BlaC is a cell surface expressed serine hydrolase of Mycobacterium tuberculosis (Mtb), one of the causative agents for Tuberculosis in humans. Mtb has demonstrated increased susceptibility to β-lactam antibiotics upon inactivation of BlaC; thus, making BlaC a rational enzyme target for therapeutic agents. Herein, we present the synthesis and structure-activity-relationship data for the 1st-generation library of bis(benzoyl) phosphates (1–10). Substituent effects ranged from σ_p = −0.27 to 0.78 for electronic and π = −0.41 to 1.98 for hydrophobic parameters. Compounds 1, 4 and 5 demonstrated the greatest inhibitory potency against BlaC in a time-dependent manner (k_obs = 0.212, 0.324, and 0.450 mn⁻¹ respectively). Combined crystal structure data and mass spectrometric analysis of a tryptic digest for BlaC inactivated with 4 provided evidence that the mechanism of inactivation by this bis(benzoyl) phosphate scaffold occurs via phosphorylation of the active-site Ser-70, ultimately leading to an aged form of the enzyme.

The bacterial pathogen Mycobacterium tuberculosis (Mtb) is the main causative agent for Tuberculosis (TB) in humans¹ and a serious global public health threat lacking effective treatment options for patients infected with multidrug-resistant (MDR) and extensively drug resistant (XDR) strains. While β-lactam antibiotics have been historically successful and safe against most bacterial pathogens, this class of antibiotics is ineffective in treating Mtb infections² primarily due to the β-lactamase BlaC expression and activity in Mtb. As a β-lactamase, BlaC can efficiently hydrolyze 3rd and 4th generation β-lactam antibiotics to near diffusion rates.^3,4 The large and open BlaC active site accommodates a variety of β-lactam scaffolds, providing Mtb with a broad spectrum β-lactamase. One common approach to combat the β-lactamase in drug-resistant pathogens is to couple a β-lactam antibiotic with a β-lactam inhibitor, enhancing Mtb susceptibility to carbapenems and cephalosporins when the antibiotics are coupled with the β-lactam inhibitor clavulanic acid.^5–8

Inspired by the pioneering work of Pratt⁹ who investigated a bis(benzoyl) phosphate (Fig. 1, R = H) as an inactivator of the OXA β-lactamase, our group prepared a small library of bis(benzoyl) phosphates as putative inactivators of BlaC from Mtb. The scope of the library was aimed at exploring substituent effects on the inhibitory potency of these compounds. As a proof-of-concept study to determine the feasibility of inactivating BlaC with a bis(benzoyl) phosphate scaffold, we limited the library to commercially accessible precursors. Originally, we hypothesized that the trend in inhibition would closely follow Hammett principles with electron withdrawing groups increasing the electrophilicity of the phosphorous center for nucleophilic attack of the phosphorous by the active site Ser-70. The Hammett premise that the main factor dictating inhibition potency the electrophilicity of the phosphorus center. In addition, to the electronics of the phosphorous center, we hypothesized that the large open active site would favor inhibitors with large hydrophobic characteristics for optimal active site interactions. In order to test these hypotheses, benzoyl compounds were chosen that had a variety of σ and π values. Based on the results of this study, we will seek to expand the library in a subsequent study.

Fig. 1.

Synthesis of bis(benzoyl) phosphate inactivators of BlaC.

The bis(benzoyl) phosphates (Table 1) were readily prepared as shown in Fig. 1. The facile method of preparation was conducted on gram scale which involved the benzoylation of dibasic phosphate in 28% to quantitative yields (Table 1). Yields, ³¹P NMR chemical shifts, as well as relevant σ and π values are presented in Table 1. The hydrolytic stability of the bis(benzoyl) phosphates 1–10 were determined in a ³¹P NMR study in which the target compounds were incubated in the enzyme assay buffer. Apart from compounds 9 and 10, no hydrolysis was observed for the bis(benzoyl) phosphates over 8 h.¹⁰

Table 1.

The inhibitory potency of bis(benzoyl) phosphate inactivators of BlaC.

Entry	R=	% yield	31P NMR	σpa	πb	% Inhibitionc, d [I] = 100 µM
1	4-OCH3	67	−17.15	−0.27	−0.09	65.5 (4.8)
2	4-CH3	Quant.	−17.24	−0.17	0.22	41.2 (2.6)
3	4-C6H5	37	−17.34	−0.01	1.98	0.0 (1.3)
4	H	Quant.	−17.33	0.00	0.00	52.3 (0.7)
5	4-F	78	−17.58	0.06	0.01	84.3 (3.7)
6	4-Cl	28	−17.66	0.23	0.58	24.8 (3.8)
7	3,4-DiCl	74	−18.09	n/a*	1.21	20.2 (2.9)
8	4-CF3	88	−17.87	0.43	1.14	42.5 (1.6)
9	4-CN	47	−17.91	0.66	−0.41	†
10	4-NO2	62	−18.05	0.78	−0.20	†

Values in bold represent > 50% inhibition at 100 µM.

^a,bLiterature values.^11–13

^cn ≥ 4 for each % inhibition value.

^dStandard deviation presented in parenthesis.

^*Values not available.

^†Unstable in enzyme assay conditions.

Bis(benzoyl) phosphates 1–8 were initially screened at a single concentration (100 µM) for inhibitory potency against purified BlaC (Table 1). Compounds exhibiting greater than 50% inhibition were selected as lead candidates for additional characterization. From our initial pilot library, we identified three compounds (1, 4, and 5) that exhibited promising potency and were subsequently assayed for timedependent inactivation of BlaC. The overall rate constants for BlaC inactivation (k_obs) for compounds 1, 4, and 5 were 0.212, 0.324, and 0.450 mn⁻¹ respectively.

Based on the % inhibition values (Table 1), a correlation was not observed between the substituent group on the benzoyl ring and inhibitory activity of BlaC. When comparing compounds that demonstrated 50+% inhibition of BlaC, compounds 1, 4, and 5, the electrostatic properties range widely from −0.27 to 0.06. In contrast, the hydrophobic properties of these compounds were in a narrow range within −0.09 to 0.01. These two observations suggest that any substituent group in the para position of the benzoyl scaffold can impart a variety of inhibition potency if the hydrophobic character of the group is kept relatively small.

Based on our above observation, we have proposed two mechanistic pathways in which the active-site Ser-70 of BlaC is inactivated in a time-dependent fashion by our bis(benzoyl) phosphates (Fig. 2).^14,15 It is expected that both the carboxyl and the phosphoryl centers in our molecule would be sufficiently electrophilic to covalently modify Ser-70 in BlaC. However, recent X-ray crystallographic data revealed a phosphorylated Ser-70 in BlaC following inactivation by compound 4 (Fig. 3) supports Path A (Fig. 2). These results suggest that after an initial inactivation step, in which an acyl phosphoryl adduct of BlaC is formed, it is subsequently hydrolyzed leading to an aged and enzyme.

Fig. 2.

Phosphorylation (Path A) vs acylation (Path B) of BlaC as putative mechanisms of BlaC inactivation by bis(benzoyl) phosphates.

Fig. 3.

X-ray crystal structure of BlaC (A) and BlaC inactivated by bis(benzoyl) phosphate 4 in which Ser-70 is phosphorylated (B). PDB 6N14.

To further confirm the results of the crystal structure analysis that suggested Ser-70 phosphorylation, we conducted mass spectrometric analysis on the peptide fragments generated by a tryptic digest of BlaC inactivated by compound 4.¹⁰ As shown in Fig. 4, a mass consistent with the sequence FAFCS(PO₃)TFK (AA 66–73) was identified (10.74 mM, 37 °C, 30 min). The peptide FAFCSTFK (AA 66–73) was identified from all three samples with common and distinct posttranslational modifications (PTMs). Beside the carbamidomethylation on Cys-69, the two control protein digests didn’t contain the peptide FAFCSTFK (AA 66–73) with any other modifications. For the reaction of BlaC with bis(benzoyl), phosphate inhibitor a doubly charged ion at 544.2202 m/z was detected, which included a phosphorylation site. Fig. 4 The kinetic and crystal structure data results suggest that the mechanism of inactivation of BlaC by bis(benzoyl) phosphates follows Path A as described in Fig. 2.

Fig. 4.

Mass spectrometry data for the tryptic digest of BlaC inactivated by bis (benzoyl) phosphate 4. Phosphorylation of Ser-70 on the peptide sequence FAFCS(PO₃)TFK (AA 66–73) was confirmed by the presence of the singly charged peptide fragments b7, y4 and y6.

In summary, we generated a first-generation library of bis(benzoyl) phosphates as putative inactivators of BlaC. With the exception of compounds 9 and 10, which were unstable in the assay conditions, all of the entries in the library exhibited inhibition activity against BlaC with compounds 1, 4, and 5 having the greatest inhibitory potency against BlaC. Even though there is no observable trend in the electronics of the substituent group and hydrophobicity there are still aspects of the scaffold design that can still be optimized to increase the inhibitor potency against BlaC. While these bis(benzoyl) phosphates could be reactive at either their carboxyl or phosphoryl center, X-ray crystallographic results confirmed that phosphorylation is the dominant mode of inactivation. In conclusion, the initial results from this study are promising and suggest that a scaffold based on analogs 1, 4, or 5 could be expanded to generate BlaC inhibitors with greater inhibitory potency and contribute to the generation of small molecule therapeutics for Mtb infection.