Subscriber access provided by University of Chicago Library
Characterization and use of TurboLuc luciferase as a reporter for high-throughput assays Douglas Auld, Janaki Narahari, Pei-i Ho, Dominick Casalena, Vy Nguyen, Evelina Cirbaite, Douglas Hughes, John Daly, and Brian Webb Biochemistry, Just Accepted Manuscript • DOI: 10.1021/acs.biochem.8b00290 • Publication Date (Web): 11 Apr 2018 Downloaded from http://pubs.acs.org on April 11, 2018
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
Characterization and use of TurboLuc luciferase as a reporter for high-throughput assays
Douglas S. Auld1†*, Janaki Narahari2†, Pei-i Ho1, Dominick Casalena1, Vy Nguyen1, Evelina Cirbaite3, Doug Hughes2, John Daly4, and Brian Webb2* 1
Novartis Institutes for Biomedical Research, Chemical Biology and Therapeutics, 250 Mass. Ave.,Cambridge, MA; 2Thermo Fisher Scientific, Rockford, IL ; 3Thermo Fisher Scientific, Vilnius ; 4Gene Stream Pty Ltd, Perth, Australia †
These authors contributed equally to the work.
*Corresponding
authors.
Email:
[email protected];
[email protected] 1 ACS Paragon Plus Environment
Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 2 of 24
Abstract: Luciferase-based reporter assays are powerful tools for monitoring gene expression in cells because of their ultrasensitive detection capacity and wide dynamic range. Here we describe the characterization and use of a luciferase reporter enzyme derived from the marine copepod Metridia luciferase family, referred to as TurboLuc™ luciferase (TurboLuc). To develop TurboLuc, the wild-type luciferase was modified to decrease its size, increase brightness, slow luminescent signal decay, and provide for efficient intracellular expression. To determine the enzyme susceptibility to compound inhibition and judge the suitability of using of TurboLuc as a reporter in screening assays, purified TurboLuc enzyme was screened for inhibitors using two different compound libraries. No inhibitors of this enzyme were identified in a library representative of typical diverse low molecular weight (LMW) compounds using a purified TurboLuc enzyme assay supporting that such libraries will show very low interference with this enzyme. We were able to identify a few inhibitors from a purified natural product library which can serve as useful tools to validate assays using TurboLuc. In addition to the inhibitor profile for TurboLuc we describe the use of this reporter in cells employing miniaturized assay volumes within 1536-well plates. TurboLuc luciferase is the smallest luciferase reporter enzyme described to date (16kDa), shows bright luminescence and low interference by LMW compounds and therefore should provide an ideal reporter in assays applied to high-throughput screening.
2 ACS Paragon Plus Environment
Page 3 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
Enzymes producing fluorescent or luminescence products have been widely employed as reporters in both biochemical and cell-based assays. One of the most widely used reporter enzymes is firefly luciferase derived from Photinus pyralis (FLuc) applied to genetic and low molecular weight (LMW) compound screening due to the high sensitivity and dynamic range of luminescence, available constructs, and detection reagents.1 The motivation to find improved luciferases has been partly driven by the realization that FLuc can show high interferences when screening typical LMW compound libraries.2-11 One of the primary modes of compound interference for reporter enzymes which take luciferins as substrates are enzyme inhibitors which often act either competitively or noncompetitively yielding different effects on the assay signal depending on the format (biochemical or cell-based) and composition of the detection reagent6. Counterintuitively, certain luciferase inhibitors can lead to an increase signal in cell-based reporter gene assays due to ligand-based stabilization of the enzyme structure in cells leading to a prolonged half-life relative to untreated cells.
3, 5, 6, 12, 13
This response mimics gene activation and is readily detected for competitive inhibitors which are removed from the enzyme following addition of lytic detection reagents which contain high concentrations of luciferin substrates. These compounds often act specifically and show overlapping structure-activity-relationships (SAR) between related luciferases as many bind in the luciferin pocket and therefore show selectivity between luciferases of unrelated sequences and mechanisms. In this regards, while clearly a nuisance when interpreting results of luciferase-based assays, luciferase inhibitors are not among the so called “pan assay interference compounds” (PAINS) 14 as the majority are acting as specific enzyme inhibitors and do not interfere with non-luciferase-based
3 ACS Paragon Plus Environment
Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
assays.3,
6, 11
Page 4 of 24
To mitigate against interference with luciferase assays, improved
luciferases and assays are needed which could substitute for FLuc or serve as orthogonal reporters to confirm activity in luciferase-based assays. New luminescence reporters have focused on non-ATP dependent luciferases and aim to provide stable lower molecular weight enzymes with improved brightness over FLuc.15-17 Naturally secreted luciferases have yielded a variety of extremely bright reporters. Gaussia luciferase (GLuc) is a 20kDa protein from the marine copepod, Gaussia princeps. This bioluminescent enzyme is significantly brighter than FLuc and Renilla reniformis (RLuc) and is secreted into the cell culture media, allowing for convenient monitoring of reporter activity.17 The recently introduced NanoLuc luciferase which is an engineered enzyme based on the natively secreted luciferase found in the deep sea shrimp Oplophorus gracilirostris shows ~100-fold improved brightness over FLuc and RLuc and high stability.15 Importantly, for high-throughput screening (HTS) applications and use as a reporter in cells, the low-molecular weight of NanoLuc (19kDa) provides for efficient expression as well as potentially lower susceptibility to interferences in HTS assays. Indeed, NanoLuc and GLuc showed to be less vulnerable to inhibition by LMW compounds compared to FLuc, perhaps due to the smaller size and non-ATP dependent enzyme mechanism of these enzymes.18 We have recently developed a reporter gene assay system using a novel 16kDa luciferase derived from a marine copepod Metridinidae luciferase family. To develop TurboLuc, the wild-type Metridia pacifica luciferase was modified to express intracellularly by removal of the secretory signal sequence. The intracellular luciferase was further modified to decrease its size, increase brightness, and provide stable glow 4 ACS Paragon Plus Environment
Page 5 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
kinetics by deleting N-terminal sequences, modification of the leader sequence and amino acid changes. Within the conserved region (e.g. without the N-terminal region which includes a signal sequence), TurboLuc shows 92.9% identity to Metridia pacifica luciferase (MpLuc1), 88.6% Metridia longa luciferase (MlLuc), and 80.7% identity to GLuc (Fig. 1a) but is not related to NanoLuc (amino acid sequence conservation of ~30%) which is part of the Oplophoridae luciferase family. The TurboLuc enzyme does not require ATP, takes coelenterazine as substrate, and is 100-fold brighter than FLuc (Table 1).
Figure 1
Consistent with other coelenterazine-utilizing luciferases, TurboLuc shows an emission maximum of 480 nm (Fig. 1b). The spectral properties of TurboLuc will enable this reporter to be multiplexed with red-shifted luciferases and could be utilized in BRET applications. To optimize the sensitivity of TurboLuc as a reporter assay, the TurboLuc 5 ACS Paragon Plus Environment
Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 6 of 24
gene was further modified to include destabilization elements to reduce accumulation of both the TurboLuc mRNA and protein in cells, effectively reducing the basal expression of the TurboLuc reporter in uninduced conditions (Figure 1c i).19 This dual destabilized TurboLuc (TurboLucDD) showed a significantly greater response in a cell-based CRE reporter assay than the non-destabilized TurboLuc (Figure 1c, ii). A biochemical enzyme assay for purified TurboLuc enzyme was constructed (Table 2). To develop the enzyme assay, we first varied the coelenterazine substrate concentration in the supplied buffer to determine an apparent KM value of 38 µM. Metridia longa luciferase has been reported to have a cooperative mechanism of substrate binding.20 A sigmoidal fit to the data yielded a Hill coefficient =1.55 (using the equation Y=Vmax*Xh/(Khalfh+ Xh; h = Hill coefficient) and a Khalf = 20 µM (Fig 2a, i). The assay signal was found to be linear using a 10 µM substrate concentration and low (nM) enzyme concentrations within a 20 minute incubation time (Fig. 2a, ii). The final enzyme assay used a 1536-well assay protocol compatible with robotic automation using miniaturized assay volumes (Table 2). A low substrate concentration was chosen to help sensitize the enzyme assay to competitive inhibitors which has been shown to be a prominent mode of inhibition for luciferase inhibitors.6, 18 Solvent sensitivity was measured for TurboLuc diluted in DMEM Hi-Glucose media containing 10% FBS at a final concentration of 50 pM (Fig 2b, i). Solvents were added to various concentrations and incubated for 10 minutes. This showed no effect with up to 2% DMSO which is well above the DMSO concentration where typical HTS assays are performed (typically assays contain 30% inhibition at the chosen
Figure 2 7 ACS Paragon Plus Environment
Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 8 of 24
screening concentration. This level of inhibition corresponded to 3 s.d. of the control wells treated with DMSO. The TurboLuc enzyme assay against the diversity set showed excellent performance with a Z΄ = 0.58±02 (Fig. 2c, i). The Z΄ value was calculated as = 1 −
( ) ( )
where σn and σa are the standard deviations of the neutral and positive controls, respectively, and µn and µa are the means of the neutral and active controls, respectively.22 In the diversity library of compounds, chosen based on medicinal chemistry tractability and properties desirable for lead discovery, we found a very low hit rate using a 10 µM screening concentration (10 hits showed >30% inhibition; 0.05% hit rate; Fig. 2d). In a previous screen of NanoLuc against the same 44K diversity set we found a hit rate of 2% using KM levels of substrate in the assay (Fig. 2d) and 1.2% in the presence of detection reagent buffer containing excess substrate (fumarizine, a coelenterazine analog).18 None of these weak hits against TurboLuc validated upon retesting in triplicate from a resupplied sample - we found that the average inhibition determined at 20 µM was 50% inhibition as the activity threshold the hit rates for FLuc, NanoLuc, and TurboLuc were 4.5% (116 hits), 1.2% (32 hits),
and
0.5
%
(14
hits),
respectively. Approximately half of the TurboLuc hits showed similar potency in the NanoLuc assay, with the majority of these also inhibiting FLuc suggesting nonspecific interference with the luminescent enzyme assays (Fig. 3a,b). Compounds showing activity at either NanoLuc or TurboLuc were validated using 8 concentrations to determine concentration-response curves (triplicate determinations). Example inhibitors from the natural product collection are shown in Figure 3. Few potent compounds were found which were selective for TurboLuc over NanoLuc and FLuc. Compound CID: 132904 (Fig. 4a) showed a potency of 2.6 µM and 0.5 µM 9 ACS Paragon Plus Environment
Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 10 of 24
against TurboLuc and NanoLuc, respectively and also showed high inhibition in the FLuc screen (98% at 25 µM, Fig. 3b). Compound CID: 467299 (Fig. 4b) was found to be selective against TurboLuc, with an IC50 3.3 µM against TurboLuc and >25 µM against NanoLuc although with some potency against FLuc (80% at 25 µM, Fig. 3). Compound CID: 13887800 (Fig. 4c) validated with weak activity (50% inhibition at 25 µM) and did not show activity at NanoLuc or FLuc (Fig3a,b). Compound CID: 12443420 (Fig. 4d) showed selectivity for NanoLuc (IC50 = 1.2 µM) over TurboLuc (IC50 = 18 µM) and also inhibited FLuc in the primary screen (94% at 25 µM, Fig. 3b). To directly measure the susceptibility of TurboLuc to so-called PAINS compounds, we screened a library of 1,408 compounds annotated to have promiscuous assay interference mechanisms in the 1536-well enzyme assays for TurboLuc (Table 2) and FLuc. The assays were performed in triplicate using a single compound concentration of 10 µM. Examining hits with > 30% inhibition showed a hit rate of 0.5% for TurboLuc and 6.7% for FLuc. Although this library contain over 100 compounds confirmed to show aggregation by differential light scattering measurements at concentrations below 20 µM, none of these compounds scored as hits in the TurboLuc assay. The primary mode of interference in the TurboLuc assay, comprising four of the seven of the hits, was associated with diazo compounds such as CID:6805369 which is a close analog to dyes such as Solvent Red, therefore light absorption is the likely mode of interference. The most potent hit for TurboLuc (97% average inhibition) was the polyphenolic compound tannic acid (CID:16129778). The average inhibition for all other TurboLuc hits in this library was 40±10%, therefore the expected IC50s should be >10 µM. In contrast, the FLuc assay showed 20 hits with >80% inhibition and the majority of these were 10 ACS Paragon Plus Environment
Page 11 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
annotated as poorly soluble or aggregating compounds as well as the diazo dye CID:6805369 (tannic acid was not found as an active in the FLuc assay). This data supports improved resilience of TurboLuc to PAINS mechanisms.
Figure 4
A TurboLuc cell-based assay where TurboLucDD was stably expressed under the control of a 4X CRE (cAMP response element) promoter in HEK293 cells was used to evaluate the performance of this reporter in cells. The CRE-TurboLucDD reporter gene assay was performed in 1536-well plates using 2,500 cells/well in a 5 µL assay (Table 3). Compound incubation was for 24 hrs at 37° C and 5% CO2. Five µL of 2X detection was then added and after 3 minutes of incubation the plates were read on a Perkin Elmer ViewLux. The optimized assay was screened against a library containing compounds of known mechanism of action (MoA).23 The MoA library was screened as a qHTS using eight different compound concentrations titrated with an inter-plate dilution protocol.24,
25
The cell-based assay showed a Zʹ of 0.51±0.04 (Fig. 5a) and
11 ACS Paragon Plus Environment
Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 12 of 24
retrospective analysis of the qHTS data as single concentration screening datasets showed that screening at 25 or 12.5 uM yielded acceptable hit rates of either 3 and 2%, respectively, with very low false positive rates (~0.5%) and high confirmation rates (8089%).25 Examination of compound-target annotations showed that the enriched mechanisms found among actives included targets expected to score as agonists such as an adenylate cyclase activator (Fig. 5b) and epigenetic modulators such as BRD4 inhibitors (Fig. 5c) and several HDAC inhibitors (examples shown in Fig. 5d). Figure 5
Summary TurboLuc is the smallest reporter enzyme described to date and has been optimized for efficient intracellular expression through removal of the secretory signal. We observed excellent performance in a cell-based assay using lytic detection reagents containing excess coelenterazine substrate. The low compound interference observed here for TurboLuc should make this enzyme a good choice to design assays applied to HTS of chemical libraries. Hit rates in optimized HTS assays are typically 80% identity to TurboLuc, show cooperative substrate binding.20 Splitting GLuc via two conserved domains (Fig. 1) showed that each domain maintains catalysis activity with substrate specificity identical to the full length enzyme but with much reduced efficiency (30% inhibition in the PubChem library show a very low hit rate for TurboLuc (10 compounds). Screening concentration was 10 µM at n = 1 for the PubChem library and n = 2 for the natural product library. An activity of -100% equals 100% inhibition and 0% represents full enzyme activity.
Figure 3. Comparison of hits in the natural product library. Data shown represents the average activity (n = 2) for screens against the various luciferase enzyme assays. Compounds are shown annotated by the PubChem CID numbers which showed either non-selective activity or selective activity against either TurboLuc, FLuc, or NanoLuc. 22 ACS Paragon Plus Environment
Page 23 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Biochemistry
Screening concentration was 25 µM. a) Activity in the NanoLuc screen vs. the TurboLuc screen. b) Activity of the FLuc screen vs the TurboLuc screen. An activity of -100% equals 100% inhibition and 0% represents full enzyme activity. Figure 4. Example actives from the natural product library. Concentrationsresponse curves for the example compounds identified as hits in the natural product library (highlighted in Figure 3). Concentration response curves against TurboLuc and NanoLuc are shown for compounds CID:132904 (a), CID:467299 (b), CID: 13887800 (c), and CID:12443420 (d). Figure 5. Cell-based TurboLuc reporter gene assay. The TurboLuc reporter gene assay showed excellent performance a) Z΄-factor vs assay plates. b) An adenylate cyclase activator, colforsin daropate, which is a water-soluble derivative of forskolin that directly activates adenylate cyclase. c) Two BRD4 inhibitors found as agonists in the assay. d) Example HDAC inhibitors found as agonists in the assay.
23 ACS Paragon Plus Environment
Biochemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
54x45mm (300 x 300 DPI)
ACS Paragon Plus Environment
Page 24 of 24
