|Home | About | Journals | Submit | Contact Us | Français|
Lipoprotein-associated phospholipase A2 (Lp-PLA2 or PLA2G7) binds to low-density lipoprotein (LDL) particles, where it is thought to hydrolyze oxidatively truncated phospholipids. Lp-PLA2 has also been implicated as a pro-tumorigenic enzyme in human prostate cancer. Several inhibitors of Lp-PLA2 have been described, including darapladib, which is currently in phase 3 clinical development for the treatment of atherosclerosis. The selectivity that darapladib and other Lp-PLA2 inhibitors display across the larger serine hydrolase family has not, however, been reported. Here, we describe the use of both general and tailored activity-based probes for profiling Lp-PLA2 and inhibitors of this enzyme in native biological systems. We show that both darapladib and a novel class of structurally distinct carbamate inhibitors inactivate Lp-PLA2 in mouse tissues and human cell lines with high selectivity. Our findings thus identify both inhibitors and chemoproteomic probes that are suitable for investigating Lp-PLA2 function in biological systems.
Lipoprotein-associated phospholipase A2 (Lp-PLA2 or PLA2G7) is a secreted serine hydrolase that is found on low-density lipoprotein (LDL) particles and, in response to oxidation of LDL particles (ox-LDL), is able to hydrolyze oxidized phosphatidylcholine into lysophosphatidylcholine (LPC) and oxidized fatty acids.1 The released LPC and oxidized fatty acids are thought to promote inflammation,2, 3 thus designating Lp-PLA2 as a potential risk factor in the development of atherosclerosis. This premise is also supported by a positive correlation between Lp-PLA2 levels and the incidence of coronary heart disease.4 Genetic studies have further shown that a small percentage of the Asian population carries an inactivating mutation of a conserved valine (V279F) in Lp-PLA2,5 and both positive and negative correlations between this mutation and cardiovascular disease have been reported.6–8 Based on the connection between Lp-PLA2 and cardiovascular disease, as well as the genetic evidence that inactivation of Lp-PLA2 may protect against coronary heart disease, this enzyme is considered a potential biomarker and therapeutic target for the treatment of atherosclerosis. Several studies have also connected elevated levels of Lp-PLA2 expression to the progression of prostate cancer.9–11 Interestingly, siRNA-induced Lp-PLA2 inactivation in the VCaP human prostate cancer cell line has been shown to induce apoptosis and sensitize the cells to oxidative stress.9 Thus, Lp-PLA2 may also serve as a biomarker and therapeutic target for prostate cancer.
Due to the interest in Lp-PLA2 as a therapeutic target, several inhibitors of this enzyme have been reported. These include β-lactams,12 amides of xanthurenic acid,13 and the compound darapladib,14 which is in Phase III clinical trials for the treatment of atherosclerosis. Treatment with darapladib has been shown to significantly decrease plasma Lp-PLA2 activity and atherosclerotic plaque formation in ApoE-deficient mice,15 in diabetic and hypercholesterolemic swine,16 and in human patients with coronary heart disease.17, 18
Despite the generation of several classes of Lp-PLA2 inhibitors, including a compound in late-stage clinical development, we are not aware of any detailed reports on the selectivity of these agents. This question is especially pertinent for inhibitors of serine hydrolases, like Lp-PLA2, given the huge number of enzymes from this class in humans.19 Lp-PLA2 also lacks precise assays that report on its activity in complex biological systems. Measurements of Lp-PLA2 activity are typically performed using either the artificial fluorescent substrate 1-decanoyl-2-(4-nitrophenylglutaryl)phosphatidylcholine (DNPG)20 or radioactive substrate [3H]-PAF,21 neither of which can readily distinguish Lp-PLA2 from several other PAF hydrolases that exist in mammalian proteomes.22, 23
With these considerations in mind, we set out to develop and apply activity-based protein profiling (ABPP) methods24, 25 for characterizing Lp-PLA2 and its inhibitors, including: 1) performing a high-throughput screen for novel classes of inhibitors, 2) identifying tailored activity-based probes for detecting Lp-PLA2 activity in cell and tissue proteomes, and 3) evaluating established and newly discovered Lp-PLA2 inhibitors for target engagement and selectivity in human cells. The high-throughput screen was performed using purified human Lp-PLA2 protein26, 27 and a recently described fluorescence polarization-ABPP (FluoPol-ABPP) platform28 in collaboration with the Molecular Libraries Screening Center Network at The Scripps Research Institute. FluoPol-ABPP assays showed time-dependent increases in fluorescence polarization of Lp-PLA2 labeled with the serine hydrolase-directed activity-based probe fluorophosphonate-rhodamine (FP-Rh), while catalytically inactive enzyme (Lp-PLA2SA) and no-enzyme control samples showed no increase in fluorescence polarization (Fig. 1). From a library of 326,141 compounds screened at 3.39 µM (Pub Chem AID: 463082), we identified 4,934 compounds (1.5%) that reduced the fluorescence polarization (> 24%) of Lp-PLA2 labeling by the FP-Rh probe. A confirmation screen (AID 463230) was performed on 2,500 compounds at the same concentration in triplicate and 1,675 (71.6%) of the compounds were confirmed as active. From these results, we cherry-picked 153 compounds based on their high (> 60%) inhibition of Lp-PLA2, selectivity for Lp-PLA2 compared to other screened enzymes and bioassays listed in PubChem, lack of undesired functional groups, and medicinal chemistry potential. We rescreened these 153 compounds versus recombinant mouse Lp-PLA2 by gel-based competitive ABPP.29 Briefly, lysates from HEK-293T cells transfected with a mouse Lp-PLA2 cDNA were incubated with each test compound (5 µM) for 30 min at 37 °C and then reacted with FP-Rh (1 µM, 30 min, 25 °C). Proteomic reactions were run on SDS-PAGE and visualized by in-gel fluorescence scanning. Twelve compounds were confirmed to inhibit Lp-PLA2 by > 75% at 5 µM, and eight of these compounds were carbamates, two of which contained a biphenyl moiety (P3 and P9, Fig. 2). A previous study30 from our lab identified a structurally distinct carbamate WWL153 (Fig. 2) as a lead inhibitor for Lp-PLA2 that displayed good potency (IC50 = 290 nM), but sub-optimal selectivity (see below). We therefore combined the 2-methyl-4-piperazinoquinoline portion of WWL153 with the biphenyl moiety of P3 and P9 to create the compound JMN4 (Fig. 2).
Additional competitive gel-based ABPP assays revealed that JMN4 exhibited superior potency for inhibiting recombinant mouse (IC50 = 90 nM) and human (IC50 = 5.9 nM) Lp-PLA2 compared to either WWL153 (IC50 = 290 nM for mouse Lp-PLA2; IC50 = 250 nM for human Lp-PLA2) or P9 (IC50 = 470 nM for mouse Lp-PLA2; IC50 = 100 nM for human Lp-PLA2) (Fig. 3A, B). Competitive ABPP experiments in the mouse brain proteome also revealed that JMN4 cross-reacted with fewer serine hydrolases than WWL153 and P9, both which inhibited FAAH at high concentrations (Fig. 3C). WWL153 also inhibited MAGL and ABHD6 (Fig. 3C). These data, taken together, designated JMN4 as a potent and selective inhibitor of mouse and human Lp-PLA2.
It is important to note that, although Lp-PLA2 is expressed in the mouse brain,30 it is not visualized by gel-based ABPP using FP-Rh, presumably due to the presence of more abundant, comigrating serine hydrolases. We reasoned that a more selective activity-based probe could facilitate the detection of endogenous Lp-PLA2 in complex proteomes such as mouse brain. Recently, we reported a 1,2,3-triazole urea activity-based probe termed HT-01 and its use to visualize the low-abundance serine hydrolase DAGLβ in cell and tissue proteomes.31 Examination of the HT-01 labeling profile of mouse brain membrane proteome revealed the presence of an additional tetrahydrolipstatin (THL)-sensitive serine hydrolase activity that exhibited a molecular weight (~55 kDa doublet) consistent with glycosylated Lp-PLA2. We had previously shown that mouse Lp-PLA2 was inhibited by THL32 and various 1,2,3-triazole ureas31 and found here that this enzyme was also inhibited by HT-01 with an IC50 of 39 nM (Fig. 4A). Based on these results, we tested both darapladib and JMN4 in gel-based competitive ABPP experiments with HT-01 and found that these compounds blocked HT-01 labeling of the 55 kDa mouse membrane enzyme, supporting its designation as endogenous Lp-PLA2 (Fig. 4B). JMN4 showed lower potency against mouse brain Lp-PLA2 (IC50 ~ 1 µM) compared to recombinant mouse Lp-PLA2, which we speculate may be due to the greater amount of total proteome needed in the assay to detect endogenous Lp-PLA2 in mouse brain (1.0 mg brain protein/mL versus 0.25 mg transfected cell protein/mL). HT-01 also identified a ~70 kDa protein that was inhibited by JMN4, but not darapladib (Fig. 4B). This enzyme is likely the blood-derived serine hydrolase carboxylesterase ES1, which has been a common off-target for many carbamate inhibitors.30 Finally, we performed a competitive gel-based ABPP analysis with FP-Rh, which showed that neither darapladib nor JMN4 cross-reacted with other serine hydrolases detected in mouse brain at concentrations up to 10 µM (Fig. 4C).
The aforementioned studies indicated that JMN4 and darapladib were selective inhibitors of mouse Lp-PLA2 and that HT-01 served as a tailored activity-based probe for detection of this enzyme in native proteomes. We next investigated the performance of these inhibitors/probes in human proteomes. As previously mentioned, human prostate cancer cells have been found to express Lp-PLA2.9–11 Consistent with these past studies, we detected an ~50 kDa protein in the PC3 human prostate cancer cell proteome that exhibited darapladib- and JMN4-sensitive labeling by the HT-01 probe (Fig. 4D). These features indicated that the 50 kDa protein was human Lp-PLA2. Interestingly, as we previously observed in the mouse brain proteome (Fig. 4C), human Lp-PLA2 could not be detected by the general serine hydrolase probe FP-Rh in the PC3 cell proteome (Fig. 4E), presumably due to overlapping signals from more abundant, co-migrating serine hydrolases. These FP-Rh profiles did, however, provide evidence that darapladib and JMN4 exhibited good selectivity for Lp-PLA2 in that other detected serine hydrolase activities in the PC3 cell proteome were not inhibited by these compounds (Fig. 4E).
We next asked whether darapladib and JMN4 could inhibit human Lp-PLA2 in situ by treating PC3 cells with these compounds and then analyzing the serine hydrolase activity profiles by the liquid chromatography-mass spectrometry (LCMS) platform ABPP-SILAC.33–36 PC3 cells were cultured in isotopically heavy or light media and treated with 100 nM darapladib or JMN4 (heavy cells) or DMSO (light cells) for 4 hours. Cells were harvested, lysed, and labeled with the activity-based probe FP-biotin.29 Light and heavy fractions were mixed in a 1:1 ratio, enriched with avidin-conjugated beads, digested on-bead with trypsin and analyzed by liquid chromatography–tandem MS (LC-MS/MS) using an LTQ-Velos Orbitrap instrument. SILAC ratios for each peptide were generated using the CIMAGE software37 to quantify light and heavy signals from the parent ion (MS1) peaks. Protein identities were determined from product ion peptide profiles (MS2) using the ProLuCID search algorithm.38 We also performed a control experiment where both heavy and light cells were treated with DMSO to control for the effects of isotopic growth media on protein expression and activity. This analysis showed both darapladib (Fig. 5A) and JMN4 (Fig. 5B) completely inhibited Lp-PLA2 in human prostate cancer cells, while not affecting any of the other ~40 serine hydrolases detected in these cells.
In conclusion, we have shown that both darapladib and the newly developed compound JMN4 are potent and selective inhibitors of Lp-PLA2 in both mouse tissues and human cell lines, indicating that these compounds can be used as a structurally and mechanistically39 distinct pair of probes to study Lp-PLA2 function in biological systems. We also believe that the tailored probe HT-01 should prove useful for profiling Lp-PLA2 activity in native proteomes, which is otherwise difficult to detect by convenient gel-based ABPP assays using more general serine hydrolase-directed probes. We speculate, for instance, that HT-01 could form the basis for an ex vivo target engagement assay to confirm Lp-PLA2 inhibition in human clinical studies, as has recently been performed for proteasome inhibitors using FPbiotin.40
We thank Mr. Pierre Baillargeon and Mrs. Lina DeLuca (Lead Identification Division, TSRI Florida) for their assistance with compound management. This work was supported by the National Institutes of Health (DA033760, MH084512, HL084366), the Prostate Cancer Foundation, the Achievement Rewards For College Scientists (ARCS) Foundation (fellowship for J.M.G.N.) and the Skaggs Institute for Chemical Biology.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.