Phosphonate and Bisphosphonate Inhibitors of Farnesyl Pyrophosphate Synthases: A Structure-Guided Perspective

Abstract

Phosphonates and bisphosphonates have proven their pharmacological utility as inhibitors of enzymes that metabolize phosphate and pyrophosphate substrates. The blockbuster class of drugs nitrogen-containing bisphosphonates represent one of the best-known examples. Widely used to treat bone-resorption disorders, these drugs work by inhibiting the enzyme farnesyl pyrophosphate synthase. Playing a key role in the isoprenoid biosynthetic pathway, this enzyme is also a potential anticancer target. Here, we provide a comprehensive overview of the research efforts to identify new inhibitors of farnesyl pyrophosphate synthase for various therapeutic applications. While the majority of these efforts have been directed against the human enzyme, some have been targeted on its homologs from other organisms, such as protozoan parasites and insects. Our particular focus is on the structures of the target enzymes and how the structural information has guided the drug discovery efforts.

Keywords: bisphosphonate; farnesyl pyrophosphate synthase; isoprenoid biosynthesis; phosphonate; structure-based drug design.

Figure 1

Mevalonate pathway and downstream metabolites. (A) An overview of the MVA pathway and downstream metabolites. Enzyme names are in italics. Dotted arrows represent multiple enzymatic steps. Sites of pharmacological intervention by current clinical drugs are indicated in red. (B) Catalytic steps of FPPS and GGPPS reactions.

Figure 2

Structures of clinical N-BP drugs.

Figure 3

Overall structure and substrate/N-BP binding of human FPPS. (A) The overall structure of hFPPS. Left side: the homodimeric biological assembly in the unliganded open conformational state [Protein Data Bank (PDB) ID: 2F7M]. One subunit is represented in a rainbow color scheme (blue to red from the N- to C-terminus). The locations of the conserved aspartic acid-rich motifs are indicated by arrows. Right side: superimposed structures of hFPPS in the open (semi-transparent white; PDB ID: 2F7M) and fully closed (cyan; PDB ID: 4H5E) states. For clarity, only one subunit is displayed. (B) DMAPP (magenta) bound to the active site of hFPPS. The residues of the DDXXD motifs are in stick representation. Mg²⁺ ions are shown as yellow spheres. DMAPP was modeled into the PDB structure 4H5E based on an analog-bound E. coli FPPS structure (PDB ID: 1RQI). (C) The binding of risedronic acid (PDB ID: 1YV5). Yellow dashes indicate a bifurcated H-bond. (D) The hydrophobic pocket of the allylic substrate site in semi-transparent surface representation. Displayed in magenta is a modeled GPP molecule. (E) The IPP sub-pocket in the enzyme-substrates ternary complex (DMAPP in magenta and IPP in light purple). (F) The conformational cascade required for the C-terminal tail closure.

Figure 4

Catalytic mechanism of FPPS reaction. Only the first catalytic cycle (i.e., the condensation of DMAPP and IPP to produce GPP) is represented for simplicity. In the subsequent cycle, GPP is condensed with another unit of IPP to produce the final product FPP.

Figure 5

Exploratory bisphosphonate inhibitors of human FPPS. (A) Molecular structures of exploratory bisphosphonate inhibitors of hFPPS. (B) The binding of inhibitor 4c (magenta; PDB ID: 4PVY). The residues of interest displayed as sticks are the same as in Figures 3B–D. (C) The binding of 5b (magenta; PDB ID: 4L2X).

Figure 6

Discovery of non-bisphosphonate hFPPS inhibitors. (A) The binding of inhibitor 6a (green; see panel B for its molecular structure) to the new druggable pocket of hFPPS (PDB ID: 3N6K). To provide a reference point, the binding sites for DMAPP (magenta), IPP (purple), and Mg²⁺ ions (yellow) are indicated in the overall structure via superposition. (B) Structures of non-bisphosphonate inhibitors of hFPPS. R¹ and R² in compounds 7 and 8 represent very broadly defined substituents, including hydrogen, halogen, and optionally substituted heterocyclic groups.

Figure 7

Bisphosphonate inhibitors with a dual binding mode. (A) Examples of thienopyrimidine and benzimidazole bisphosphonates. (B) The binding of inhibitor 5d to the active site (left panel; PDB ID: 4JVJ) and the new druggable pocket (right; PDB ID: 4LPG) in the presence and absence of Mg²⁺ ions.

Figure 8

Allosteric binding of FPP to hFPPS. (A) A schematic representation of FPPS catalytic cycle and allosteric product inhibition. FPP binding locks the enzyme in its open, inactive state. (B) A crystal structure showing FPP bound to the allosteric pocket (PDB ID: 5JA0). (C) Residues making direct interactions with the pyrophosphate moiety. (D) Conformational changes induced by FPP binding. Superimposed in gray is the structure of hFPPS in the unliganded state (PDB ID: 2F7M).

Figure 9

Examples and binding modes of allosteric inhibitors of hFPPS. (A) Thienopyrimidine-based monophosphonate inhibitors of hFPPS. (B) The binding of inhibitor 10b at the allosteric pocket (PDB ID: 6N83). (C) The binding of 10c (PDB ID: 6N82). (D) Structures of benzoindole-, salicylic acid-, and quinoline-based inhibitors of hFPPS. (E) Superimposed thienopyrimidine-, salicylic acid-, and quinoline-based monophosphonate inhibitors in their binding poses. Protein structures are omitted. The phosphorus and oxygen atoms of the phosphonate groups are colored in orange and red, respectively. All other atoms are colored in white.

Figure 10

Prodrugs to target hFPPS. (A) The structure of the pivaloyloxymethyl (POM) group. (B) Examples of POM-protected N-BP prodrugs.

Figure 11

Structures of trypanosomatid FPPS. (A) Superimposed structures of T. cruzi (green; PDB ID: 4DWG), T. brucei (light purple; PDB ID: 4RXD), L. major (yellow; PDB ID: 4JZX) and human FPPS (white; PDB ID: 4H5E). Conserved residues of the active site are displayed in the inset. (B) Aromatic residues of the allylic substrate site hydrophobic pocket.

Figure 12

Bisphosphonate inhibitors of trypanosomatid FPPS and their target binding. (A) Examples of exploratory bisphosphonate inhibitors of trypanosomatid FPPS. (B) The binding of 12a (top panel; PDB ID: 4DXJ) and 12d (bottom; PDB ID: 4DWG) to TcFPPS. The conformational changes in Tyr94 and Gln167 induced by 12d results in an increase in the volume of the binding site hydrophobic pocket. (C) The binding of 13b (top; PDB ID: 4JZX) and 13c (bottom; PDB ID: 4JZB) to LmFPPS. (D) Superimposed binding poses of 13d (yellow; PDB ID: 5AEL), 13e (magenta; PDB ID: 3EFQ), 14a (white; PDB ID: 5AFX), and 14b (orange; PDB ID: 5AHU).

Figure 13

Biosynthesis of juvenile hormones and bisphosphonate inhibitors of CfFPPS2. (A) Molecular structures of juvenile hormones (JHs). (B) The MVA pathway and its promiscuous ethyl-branched metabolites (in brackets). The intermediate steps catalyzed sequentially by mevalonate kinase, phosphomevalonate kinase, and mevalonate pyrophosphate decarboxylase are omitted for simplicity (dotted arrow). Abbreviations: HEG CoA, hydroxylethylglutaryl coenzyme A; HIPP, homoisopentenyl pyrophosphate; HDMAPP, homodimethylallyl pyrophosphate. (C) The production of FPP/ethyl-branched analogs of FPP and their conversion to JH. Various combinations of the starting substrates lead to different forms of JH. For example, the condensation of HDMAPP with HIPP and IPP produces 7,11-bishomofarnesyl pyrophosphate, which is then converted to JH I. Notably, methyl farnesoate is the immediate precursor of JH III and lacks the epoxide moiety characteristic of JH. While there is a long-standing debate, the potential role of methyl farnesoate as a JH has been recognized. (D) N-alkylated ortho-substituted bisphosphonate inhibitors of CfFPPS2.

Figure 14

Structures of bisphosphonate-bound CfFPPS2. (A) The overall structure of the CfFPPS2-15b-IPP ternary complex (cyan; PDB ID: 6B06). Superposed in white is the structure of a human FPPS ternary complex (PDB ID: 4H5E). (B) The binding of inhibitor 15b. The structure is of the CfFPPS2-15b binary complex (PDB ID: 6B04). (C) The binding of inhibitor 15d (PDB ID: 6B07). Displayed residues are identical as in (B).