Figure 1.
Comparisons of primary structures between mammalian BAT5 orthologs.
Predicted acyltransferase motif (HxxxxD) [3] and predicted active site nucleophile (#) [3] are idicated. In addition, two lipase-like motifs (GxSxxG) are highlighted. Gray shading indicates amino acid residues deviating from the human sequence. Comparison to the human sequence indicated the following identity: mouse (96%), rat (95%), naked mole rat (96%), bat (95%), alpaca (97%) and camel (97%). The following gene bank accession numbers were used for the sequences: Human (Homo sapiens) BAB63383.1, mouse (Mus muculus) NP_848707.1, rat (Rattus norvegicus) NP_997696.1, naked mole-rat (Heterocephalus glaber) XP_004847034, bat (Myotis lucifugus) XP_006104671, alpaca (Vicugna pacos) XP_006215394, camel (Camelus ferus) XP_006178831.
Figure 2.
Activity-based protein profiling (ABPP) to visualize catalytically active BAT5 protein in lysates of HEK293 cells after transient expression of human (h) or mouse (m) BAT5 orthologs.
Serine hydrolases were labeled using the active site serine targeting fluorescent probe TAMRA-FP. After separation in SDS-electrophoresis gel (10%), serine hydrolase activity was visualized by in-gel fluorescent gel scanning as detailed in the Methods section. Molecular weight markers (MW) are indicated at left. Transient transfections with the cDNAs encoding hBAT5 (A) or mBAT5 (B) results in robust labeling of a ∼63 kDa protein band that is absent from parental HEK293 cells. Data are from one typical transfection, transfections were repeated twice with similar outcome.
Figure 3.
In vitro glycerolipid substrate profile of hBAT5.
HEK293 cells were transiently transfected with the cDNA encoding hBAT5, as detailed in the Methods section. After 48 h, cells were harvested and lysates prepared for hydrolase activity measurements using a sensitive fluorescent glycerol assay. The substrate panel included monoacylglycerols (MAGs) with the indicated acyl chain length, isomer and degree of saturation, the diacylglycerol (DAG) 1,2-dioleoyl(C18∶1)-rac-glycerol, the triacylglycerols (TAG-1 = 1,2,3-trioleoyl(C18∶1)glycerol; TAG-2 = 1-palmitoyl(C16∶0)-2-oleoyl(C18∶1)-3-linoleoyl(C18∶2)-rac-glycerol, as well as the prostaglandin glycerol esters PGD2-G, PGE2-G, PGF2α-G and 15d-PGJ2-G. Cellular lysates (0.3 µg/well) were incubated together with the indicated substrates [25 µM final concentration, added from 10 mM stock solutions in ethanol into the glycerol assay mix containing 0.5% (w/v) fatty acid free BSA and 1% (v/v ethanol). Glycerol production was determined at time-point 60 min. A. Background activity for the tested substrates. Cellular background was similar between HEK and Mock-transfected cells and the values shown are combined from the two. B. Substrate profile of hBAT5. C. The MAG substrate profile of hBAT5 in assays conducted without BSA. D. Linear, time-dependent generation of glycerol as a result of 1-LG (25 µM) hydrolysis in hBAT5-HEK lysates (0.3 µg/well) in incubations with or without BSA. Note ∼2.5-fold higher hydrolysis rate in the absence of BSA. Data are mean + SEM from 3–7 (A and B) or three (C and D) independent experiments using hBAT5 lysates from one transfection. Statistical comparisons between the MAG 1(3)- and 2-isomers were done by using unpaired (B) or paired (C) t-test and the significance is indicated with an asterix (NS, non-significant; *, p<0.05; **, p<0.01).
Figure 4.
Fatty acid liberation from various substrates when incubated together with lysates of HEK293 cells expressing hBAT5 or hABHD12.
HEK293 cells were transiently transfected with the cDNA encoding hBAT5 or hABHD12, as detailed in [10] and the Methods section. After 48 h, cells were harvested and lysates prepared for lipase activity measurements based on the Cayman’s FFA fluorescence assay kit. For validation purposes, lysates of parental HEK293 cells were tested in parallel, and the substrate panel included the MAGs 1-AG (A) and 1-LG (B), the lysophospholipids C18∶1-LPA (C) and C18∶1-LPS (D), the DAG 1,2-dioleoyl(C18∶1)-rac-glycerol (E), and the TAG 1,2,3-trioleoyl(C18∶1)glycerol (F). Cellular lysates (0.3 µg/well) were incubated together with the indicated substrates [25 µM final concentration, added from 10 mM stock solutions in ethanol into the FFA assay mix containing 0.1% (v/v ethanol). Assay blank and fatty acid standard (C18∶1) were included for each condition. FFA content was determined at time-points 10 and 20 min. Data are mean ± SEM from quadruplicate wells in one experiment. Statistical comparisons to values of the HEK lysates were done using paired t-test and the significance is indicated with an asterix (NS, non-significant; **, p<0.01; ***, p<0.001).
Figure 5.
pH optima of mouse (m) and human (h) BAT5 hydrolase activities towards 1-LG and 15d-PGJ2-G.
Lysates (0.3 µg/well) of HEK293 cells transiently expressing mBAT5 (B, E) or hBAT5 (C, F) were incubated at the indicated pH for 60 min using 1-LG (A, B, C) or 15d-PGJ2-G (D, E, F) as the substrate (25 µM final concentration). For comparative purposes, substrate hydrolysis in lysates of parental HEK293 cells is also illustrated (A, D). The incubations additionally contained 0.1% (v/v) ethanol, 0.1% (v/v) DMSO and 0.5% (w/v) BSA. After 60 min incubation, the pH of the assay system was brought to the neutral range by the addition of Glycerol Assay Mix containing additionally 100 mM Tris-HCl, pH 7.4 and THL (10−5 M) to quench hydrolase activity. Assay blanks and glycerol standards were included for each pH condition. Glycerol content was determined at time-point 60 min. The buffer systems were as follows: 10 mM K-phosphate buffer (KPB) covering the pH range 5.3–8.0 and 10 mM Tris-HCl covering the pH range 7.2–9.1. For comparative purposes, hydrolase activity in the routinely used hydrolase assay buffer (TEMN, pH 7.4) is shown. Data are mean + SEM of duplicate wells from two independent experiments.
Figure 6.
Km and Vmax values for hBAT5-dependent hydrolysis of 1-LG and 15d-PGJ2-G.
HEK293 cells were transiently transfected with the cDNA encoding hBAT5, as detailed in the Methods section. After 48 h, cells were harvested and lysates prepared for hydrolase activity measurements. Cellular lysates (0.3 µg/well) were incubated together with the indicated concentrations of 1-LG in assay mixes containing (+ BSA) or not (– BSA) 0.5% (w/v) BSA (A) or with 15d-PGJ2-G in assay mix containing BSA (B). The Km and Vmax values (C) were determined at time-point 90 min, where substrate utilization was <10%, and were calculated as non-linear regressions using GraphPad Prism 5.0 for Windows. As crude cellular lysates were used, total protein concentration was used instead of true enzyme concentration. The Lineweaver-Burk plots are shown inside rectangles. Statistical comparison of the Km and Vmax values for 1-LG between the + and - BSA conditions was done using unpaired t-test and the significance is indicated with an asterix (*, p<0.05). Values are mean ± SEM from three independent experiments.
Figure 7.
β-lactones as BAT5 inhibitors.
A. Potency of β-lactones to inhibit 1-LG hydrolysis in lysates of hBAT5-HEK293 cells. Lysates were treated for 30 min with the inhibitors after which glycerol generated in the hydrolysis of 1-LG (25 µM final concentration) was monitored for 90 min. Inhibitor dose–response curves and IC50 values derived thereof were calculated from non-linear regressions using GraphPad Prism 5.0 for Windows. The logIC50 values are means ± SEM from three independent experiments. IC50 values (mean) are shown in parenthesis. B–C. Competitive ABPP of hBAT5-HEK293 cell lysates showing that palmostatin B inhibits probe binding to hBAT5 with the same potency (IC50 ∼ 100 nM) as it blocks hBAT5-dependent 1-LG hydrolysis (A). C. Quantitative data (mean + SEM) on the effects of palmostatin B and THL on probe labeling of the hBAT5 band (black arrow in B). The data are derived from three separate ABPP experiments. D. Competitive ABPP demonstrating that the potency order of the β-lactones to inhibit probe binding to mBAT5 (palmostatin B > THL > ebelactone A << hymeglusin (inactive) is identical to that observed for hBAT5 (A). In B and D, lysates (25 µg) were pretreated for 1 h with DMSO or the indicated concentrations of the inhibitors, after which TAMRA-FP labelling was conducted for 1 hour at RT. The reaction was stopped, 5 µg protein was loaded per lane and the proteins separated in SDS-electrophoresis minigel (10%). TAMRA-FP labeling was visualized after in-gel fluorescence imaging as described in the Methods section. The image in D is representative from two ABPP experiments with similar outcome.
Figure 8.
Competitive ABPP unveiling palmostatin B targets among the serine hydrolases in mouse brain membrane proteome.
Membranes (100 µg) were pretreated for 1 h with DMSO or the indicated concentrations of the serine hydrolase inhibitors, after which TAMRA-FP labelling was conducted for 1 hour at RT. The reaction was stopped, 10 µg protein (10 µl) was loaded per lane and the proteins were separated in SDS-electrophoresis gel (10%). TAMRA-FP labeling was visualized after in-gel fluorescence imaging as described in the Methods section. Molecular weight markers are indicated at left. Reference inhibitors JJKK-048, WWL70, THL and JZL195 were used at the indicated concentrations to identify the following serine hydrolases from the gel: fatty acid amide hydrolase (FAAH), inhibited by JZL195 [21]; MAGL doublet, inhibited by JJKK-048 [22] and JZL195 [21]; ABHD6, inhibited by WWL70 [23]–[24], THL [18], and also by JZL195 at the used concentration; BAT5, inhibited by THL [4], [18]. Note that palmostatin B dose-dependently inhibits probe labeling of BAT5, ABHD12 and LYPLA1/2, two closely-related proteins migrating at ∼25 kDa [25]. At the highest concentration, probe labeling of MAGL and ABHD6 is also inhibited. Note the absence of additional visible targets for palmostatin B at the tested concentrations. Data are representative from three ABPP experiments with similar outcome.
Table 1.
Sensitivity of hABHD6, hMAGL, hABHD12 and hFAAH towards palmostatin B and related β-lactones.
Figure 9.
Inhibition of hBAT5 by phenyl and benzyl substituted 1,3,4-oxadiazol-2(3H)-ones 38–50.
Lysates of hBAT5-HEK293 cells were treated for 30 min with the inhibitors after which glycerol liberated from 1-LG hydrolysis (25 µM final concentration) was kinetically monitored for 90 min. Inhibitor dose–response curves and IC50 values derived thereof were calculated from non-linear regressions using GraphPad Prism 5.0 for Windows. Values are mean ± SEM from three independent experiments using lysates from one bacth of transfection.
Figure 10.
Inhibition of hBAT5 by 5-alkoxy and phenoxy substituted 3-(3-nitrophenyl)-1,3,4-oxadiazol-2(3H)-ones 40, 51–55.
Lysates of hBAT5-HEK293 cells were treated for 30 min with the inhibitors after which glycerol liberated from 1-LG hydrolysis (25 µM final concentration) was kinetically monitored for 90 min. Inhibitor dose–response curves and IC50 values derived thereof were calculated from non-linear regressions using GraphPad Prism 5.0 for Windows. Values are mean ± SEM from three independent experiments using lysates from one bacth of transfection.
Figure 11.
Competetive ABPP profiling of the selectivity of the HSL inhibitor C7600 (38) and its analogues 40 and 44 among the serine hydrolases in mouse wholebrain membrane proteome.
Membranes (100 µg) were treated either with DMSO or the indicated concentrations of the inhibitors for 60 min at RT. After this, serine hydrolases were labeled for 60 min using the active site serine targeting fluorescent probe TAMRA-FP. After separation in SDS-electrophoresis gel (10%), serine hydrolases were visualized by in-gel fluorescent gel scanning as detailed in the Methods section. Molecular weight markers (MW) are indicated at left. Consistent with previous reports [28]–[29], C7600 (38) targets the endocannabinoid hydrolases FAAH and MAGL, as evidenced by dose-dependent inhibition of probe binding to these serine hydrolases. In addition, 38 inhibits probe labeling of KIAA1363, a protein doublet migrating at ∼50 kDa [4] and of LYPLA1/2, two closely-related proteins migrating at ∼25 kDa [25]. Note sensitivity of BAT5 to all tested inhibitors. Note also improved BAT5 selectivity of 40 and 44 as compared to 38. The gel is representative from two (four in the case of 40) independent ABPP experiments with similar outcome.
Figure 12.
Schemes outlining the synthesis procedures for the C7600 analogues of this study.