ZHAO Ying*,ZHANG Jia-huai*
(1.Institute of Materia Medica,Chinese Academy of Medical Sciences&Peking Union Medical College,Beijing 100050,China;2.Chinese Pharmacological Society,Beijing 100050,China;3.Center for Clinical Laboratory,Capital Medical University,Beijing 100069,China)
Abstract:OBJECTIVE To develop a method to evaluate the compatibility of compounds in the fluores⁃cence resonance energy transfer(FRET)model for β-secretase(BACE1)inhibitor screening.METHODS Two commercially available BACE1 inhibitor screening systems based on FRET were selected to eval⁃uate the BACE1 inhibitory activities of(-)-epigallocatechin-3-gallate(EGCG)and Compound 1 according to the supplier′s protocol.The inhibitory rates and slopes of the catalytic curves of the inhibitors were calculated.The effect of inhibitors on the fluorescence intensity of the systems were quantitatively calculated and the comparatively evaluated.RESULTS EGCG,a reported non-competitive inhibitor of BACE1,directly induced the reduction of fluorescence intensity of one of the systems.The slope of the line with the addition of EGCG(10.8±2.6)conformed to that of the line of EGCG inhibition(10.2±3.4),which indicated that EGCG was a pseudo-positive inhibitor of BACE1.Compound 1 had little effect on the fluorescence intensity of the systems,so the inhibitory activity of Compound 1 was confirmed.The compounds which showed inhibi⁃tory activity in preliminary screening should be checked in the blank control without BACE1 to calibrate the effect of compound on the system fluorescence intensity.The applicability of the tested compounds in the screening system could thus be evaluated to prevent pseudo-positive results.CONCLUSION This fluorescence calibration method with compound control can be universally used for assays based on FRET theory to evaluate the applicability of tested BACE1 inhibitors.
Key words:fluorescence resonance energy transfer;β-secretase;fluorescence intensity;(-)-epigallo⁃catechin-3-gallate
Alzheimer disease(AD)is a neurodegenerative disorder that mainly affects the elders.Although the exact causes of AD remain unclear,amyloid-β(Aβ)formation and deposition are considered to be the key process during the pathogenesis of AD[1-2].β-secretase(BACE1),a transmembrane aspartyl protease,is regarded as the key enzyme during the process of A β deposition.Several studies showed that BACE1 might be a good therapeutic target for AD[3-4].Thus to explore BACE1 inhibitors from either natural resources[5]or synthe⁃sized compounds[3]has become a hot spot for anti-AD drug research and development.Several BACE1 assay kits or substrates are commercially available for high-throughput screening(HTS)of BACE1 inhibitors,such as PanVera®BACE1 FRET assay kit,red(part No.P2985),R&D Systems®Fluoro⁃genic peptide substrate Ⅳ(Catalog No.ES004),and a series of Calbiochem®BACE1 substrate.So far,most of the BACE1 assays are based on fluorescence resonance energy transfer(FRET)technology.In this method,two fluorophores(a fluorescent donor and a quenching acceptor)are synthesized in the substrate.When the substrate is cleaved by BACE1,the quenching acceptor is separated from the donor,resulting in the increase in fluorescence.The increase in fluorescence is linearly related to the rate of proteolysis[6-7].It is a quick and homogenous method that is easily for HTS.Using these assays,some small molecular BACE1 inhibitors from natural resources were identified and reported,such as flavanones[8]and catechins[9].Because of their high stability and permeability,the small molecular inhibitors are regarded as good drug candidates.
However,the fluorescence in FRET assay may be interfered by the test compounds,which limits the application of the assay in BACE1 inhibitor screening.The fluorescence intensity could be either amplified or attenuated,which may lead to pseudonegative or pseudo-positive results[10].In this study,we aim to develop a universal method for fluores⁃cence based assays to measure the capacity of test compounds,especially natural products.
Recombinant human BACE1 and(-)-epigal⁃locatechin-3-gallate(EGCG)were purchased from Sigma-Aldrich Co.LLC.(St.Louis,MO,USA).Compound 1,namely 2-[(Z)-heptadec-11-enyl]-6-hydroxybenzoic acid,is a phenolic acid isolated fromHomalomena occultaby our group[11].Fluorogenic peptide substrate Mca-Ser-Glu-Val-Asn-Leu-Asp-Ala-Glu-Phe-Arg-Lys(Dnp)-Arg-Arg-NH[where Mca is(7-methoxycoumarin-4-yl)-acetyl,and Dnp is 2,4-dinitrophenyl]were pur⁃chased from R&D Systems,Inc.(Minneapolis,MN,USA).BACE1 FRET assay kit was purchased from PanVera Co.(Madison,WI,USA).DMSO was purchased from AMRESCO LLC.(Solon,OH,USA).Other agents were domestic ones.All inhibitory activity assays were performed with TECAN Infinite F200 plate reader(Männedorf,Switzerland).
BACE1 inhibitory activity assays were performed in 384-well black plates according to the supplier′s protocol.For PanVera BACE1 FRET assay,substrate(Rh-EVNLDAEFK-Quencher)250 nmol·L-1,BACE1 10 mU,and different concentrations of inhibi⁃tors(EGCG 10 μmol·L-1or Compound 12.7 μmol·L-1,dissolved in small volumes of DMSO prior to addi⁃tion to the buffer)were mixed in assay buffer(sodium acetate 50 mmol·L-1,pH 4.5).The total volume was 30 μL.A well without BACE1 was set as blank control.The real-time fluorescence intensity in the initial 60 min was recorded using a TECAN Infinite F200 plate reader at Ex545nm/Em590nmat 25℃with a time interval of 1 min.In R&D BACE1 FRET assay,the final concentration of the substrate[Mca-SEVNLDAGFRK(Dnp)RR-NH]was 400 nmol·L-1,and the excitation/emission wavelengths were 320 nm/405 nm.The other parameters were the same as those in PanVera assay.
The inhibitory rate was calculated by the following equation:Inhibitory rate(%)=[1-(FSFS0)/(FC-FC0)]×100%,whereFS0andFSare the flu⁃orescence of samples at zero time and 60 min,andFC0andFCare the fluorescence of control at zero time and 60 min,respectively.
PanVera substrate 250 nmol·L-1or R&D substrate 400 nmol·L-1,BACE1 10 mU,and inhibi⁃tors(EGCG 10 μmol·L-1or Compound 12.7 μmol·L-1)were mixed in assay buffer(sodium acetate 50 mmol·L-1,pH 4.5).The total volume was 30 μL.A series of wells containing no inhibitor were set to be control and starting treatments(ST).The baseline control contained substrate in the assay buffer alone.The reaction was performed at 25℃ for 60 min on a TECAN Infinite F200 fluo⁃rescent plate reader with Ex545nm/Em590nmfor PanVera assay,and Ex320nm/Em405nmfor R&D assay.Fluores⁃cence in each well was tracked kinetically in realtime at a 10-min interval.The velocity of each treatment was represented by the slope of cata⁃lyzing curve(S60).Compound 1,which was proved to be a non-competitive inhibitor of BACE1[11],was used as a positive control.S60was calculated by the following equation:S60=(F60-F0)/60,whereF0andF60are the fluorescence at zero time and 60 min,respectively.
EGCG 10 μmol·L-1or Compound 12.7 μmol·L-1was added to the ST at the end of1.3(60 min),and fluorescent intensities were recorded imme⁃diately.A straight line was drawn by the initial and end points,and the slope.Sa0was calculated by the following equation:Sa0=(Fa0-F0)/60,whereF0andFa0are the fluorescence at zero time and immediately after the inhibitors were added,respectively.
Fluorescence intensities were recorded at the end of 20 min incubation after1.4.The velocity in the last 20 min was represented by the slope of catalyzing line asSa20.Sa20was calculated by the following equation:Sa20=(Fa20-Fa0)/20,whereFa0andFa20are the fluorescence immediately after the inhibitors were added and at the end of 20 min incubation,respectively.
In PanVera BACE1 FRET assay system,the inhibitory rate of EGCG 10 μmol·L-1was calculated to be(67.8±2.1)%,which was consistent with the previous report(IC=1.6×10-6mol·L-1)[6].However,
50the inhibitory rate of EGCG to BACE1 detected in R&D BACE1 FRET assay system was(7.6±5.5)%,which was much lower than the activity calculated in PanVera assay system(Tab.1).
The catalyzing curves were shown in Fig.1,and slope of each line was calculated as in Tab.2.It could be found in Tab.2 thatS60,Sa0andSa20for control were similar,indicating that the catalyzing velocity was constant during 85 min.The catalytic curves of EGCG 10 μmol· L-1orCompound 12.7 μmol·L-1were also linear.However,when EGCG or Compound 1 was added in ST,there was a drop in the slope.We selected 2.7 μmol·L-1as the concentration of Compound 1,for at this concentration,the slope drops in the same way as that of EGCG 10 μmol·L-1.At this concentra⁃tion of Compound 1,the activity of BACE1 was completely inhibited.It could be seen thatSa0andSa20of ST+EGCG 10 μmol·L-1were equal,and equal to the counterparts of EGCG 10 μmol·L-1,indicating that the difference ofS60of control and EGCG 10 μmol·L-1was totally induced by the effect of EGCG on fluorescence rather than the inhibi⁃tion to BACE1.However,althoughSa0of ST+Compound 12.7 μmol·L-1was lower thanSa0of control,and equal toSa0of ST+EGCG 10 μmol·L-1,Sa20of ST+Compound 12.7 μmol·L-1was appar⁃ently lower thanSa0of ST+Compound 12.7 μmol·L-1,and similar toS60of Compound 12.7 μmol·L-1,indicating that although the fluorescence intensity might be affected by Compound 1,the main cause of slope drop was the inhibition of Com⁃pound 1 to BACE1.
Tab.1 Inhibitory rates of(-)-epigallocatechin-3-gallate(EGCG)to β-secretase(BACE1)detected in PanVera and R&D BACE1 system
Fig.1 Effect of EGCG(A)and Compound 1(B)on fluo⁃rescence intensity in FRET model in Pan Vera system.The fluorescence intensity dropped after EGCG 10 μmol·L-1or Compound 12.7 μmol· L-1was added into starting treatment(ST).In order to show the slope changes of the lines,the data of fluorescence intensity of control was original,those of ST+EGCG or Compound 1 were(original data-500),and those of EGCG 10 μmol·L-1were(original data-1000).A:the line given out by the starting point and the point immediately after EGCG was added(dashed line)was parallel to the catalyzing line of EGCG 10 μmol·L-1,indicating the fluorescence reduction was totally induced by the quenching effect of EGCG.B:the slope of the line drawn by the starting point and the point immediately after Compound 1 was added(dashed line)was apparently higher than that of the catalyzing line of Compound 12.7 μmol·L-1,indi⁃cating that fluorescence reduction was induced by the inhibition of BACE1.±s,n=4.
Tab.2 Slope of catalyzing line affected by EGCG and Compound 1 in FRET model in PanVera system
FRET models are widely used in inhibitor screening.However,the test compound may affect fluorescence intensity and yield pseudo-positive results.Generally,there are two factors that may affect the fluorescence in a FRET assay.①Some natural products contain fluorophores whose fluo⁃rescent spectra overlay those of the substrate.As a result,the background fluorescent intensity becomes too high to be measured.②The fluores⁃cence of a substrate may be quenched by the test compound,which leads to pseudo-positive results.Factor 1 could be easily found due to the very high and instable fluorescence intensity.However,factor 2 might be neglected because the fluores⁃cence intensity could also be reduced by the inhi⁃bition of BACE1 activity.This phenomenon has also been reported in other FRET-based assays[12].A typical example is of EGCG,which was reported to be a non-competitive inhibitor of BACE1 but showed no inhibitory activity in R&D BACE1 assays.
EGCG had been reported as a non-competi⁃tive inhibitor of BACE1[9].Similar results were obtained by us using the Panvera BACE1 systems.However,EGCG did not show inhibitory activity in our R&D screening system.Besides,although EGCG exhibited neuroprotective activities in a series of models[13-14],there was little evidence that BACE1 is a direct target of EGCG.A possible explanation for the inconsistency is that in the reported system[PanVera®BACE1 FRET assay kit,red],the fluorescence intensity was directly reduced by EGCG,instead of BACE1 inhibition.We conducted an experiment to verify our assump⁃tion using PanVera kit.Since Compound 12.7 μmol·L-1fully inhibited the activity of BACE1,the effect of Compound 1 on fluorescence intensity could be neglected.Our studies proved that at lower concentrations at which IC50could be calculated,Compound 1 had little effect on fluorescence intensity.
It can be seen from the example above that the influence of the test compound on fluores⁃cence could affect the reliability of FRET measure⁃ment,which is why a control test should be carried out to evaluate the compatibility of the test com⁃pounds that show BACE1 inhibitory activity in this FRET model.Here,we propose a method.①A series of STs that are completely the same as control are prepared and tracked under the stan⁃dard condition for 60 min.The slopes for all treat⁃ments and control should be equal.②Add the test compound into the starting treatments immedi⁃ately after step 1 is finished.Read the fluores⁃cence intensities and check if there is a sudden drop.If there is,step 3 should be taken.③ Calcu⁃late the slope from the straight line drawn from the fluorescence intensities at the beginning and after the compounds are added,and compare the slope value to that of the inhibiting curve.If the two values are similar,the apparent inhibitory ac⁃tivity of the compound to BACE1 is probably a pseudo-result caused by the influence of the compound on fluorescence.④The fluorescence intensities of all treatments should be tracked for extra 20 min to confirm that all catalyzing curves are linear.If the curves bend,the tracking time in step 1 should be reduced to ensure that the cata⁃lyzing velocity is constant in the system during the whole procedure.
There is also a possibility that the slope drops to some extent after the compound is added,but still higher than that obtained from the inhibi⁃tory curve.That means although the compound could affect the fluorescence,it does have the ability to inhibit BACE1.In this situation,the propor⁃tion of the fluorescence intensity that is reduced by the compound could be calculated from the difference in the slope before/after the addition of the compound,and the real inhibitory activity might be calibrated by the proportion.However,a better way is to use another model that is measured under different wavelengths,which should be far enough to those in the former model.Usually the compound would not simultaneously affect systems with far different wavelengths,but it should also be checked as mentioned above to make it sure.
In summary,we found a pseudo-positive BACE1 inhibitor that could induce fluorescence drop in FRET system.It is a universal problem in FRET and other fluorescence-based screening system.Thus,we established the method to evaluate the compatibility of compounds in the BACE1 FRET model.With this method,the compounds that possess pseudo-inhibitory activity on BACE1 could be found,and the real ability of the BACE1 inhibitors that may reduce the fluorescence inten⁃sity of the system could be calibrated.Although we develop the method for BACE1 inhibitor screening model,the method might be universal for other models based on FRET theory.