Here, we present a protocol to measure KDM1A target engagement in a human or animal cell, tissue or blood samples treated with KDM1A inhibitors. The protocol employs chemoprobe tagging of the free KDM1A enzyme and direct quantification of the target occupation using chemoprobe-based immunoassays and can be used in preclinical and clinical studies.
The assessment of the target engagement, defined as the interaction of a drug with the protein it was designed for, is a basic requirement for the interpretation of the biological activity of any compound in drug development or in basic research projects. In epigenetics, target engagement is most often assessed by the analysis of proxy markers instead of measuring the union of the compound to the target. Downstream biological readouts that have been analyzed include the histone mark modulation or gene expression changes. KDM1A is a lysine demethylase that removes methyl groups from mono- and dimethylated H3K4, a modification associated with the silencing of gene expression. Modulation of the proxy markers is dependent on the cell type and function of the genetic make-up of the cells investigated, which can make interpretation and cross-case comparison quite difficult. To circumvent these problems, a versatile protocol is presented to assess the dose effects and dynamics of the direct KDM1A target engagement. The assay described makes use of a KDM1A chemoprobe to capture and quantify uninhibited enzyme, can be broadly applied to cells or tissue samples without the need for genetic modification, has an excellent window of detection, and can be used both for basic research and analysis of clinical samples.
Lysine specific demethylase 1 (KDM1A)1 is a demethylase involved in the control of gene transcription. This protein has emerged as a candidate pharmacological target2 in oncology; including Acute Myeloid Leukemia3 (AML), Myelodysplasia Syndrome (MDS)4, Myelofibrosis (MF)5,6, Small Cell Lung Cancer (SCLC)7; in Sickle Cell Disease (SCD)8,9, and in central nervous system diseases including Alzheimer's Disease (AD), Multiple Sclerosis (MS); and in aggression10.
Most of the KDM1A inhibiting compounds in clinical development are cyclopropylamine derivatives and inhibit the protein by covalent binding to its flavin adenine dinucleotide (FAD) cofactor11. Inhibition of KDM1A induces gene expression changes, but these changes vary enormously across tissues, cell types, or disease cases. Inhibition of KDM1A also changes histone marks12, yet these changes are generally produced locally at a specific site in the genome, and are again, highly tissue and cell specific.
The protocol was developed to directly measure KDM1A target engagement in biological samples and has been optimized for the use with cyclopropylamine derived inhibitors. The assay is based on ELISA technology and analyzes, in parallel, Total and Free (i.e. unbound by inhibitor) KDM1A in a native protein extract from a biological sample in a solid phase assay. As a first step, the biological sample is lysed in the presence of the biotinylated KDM1A selective chemoprobe OG-88113,14, derived from the selective KDM1A inhibitor ORY-1001 (iadademstat), a potent inhibitor of KDM1A in clinical development for the treatment of oncological disease. The chemoprobe has an IC50 for KDM1A of 120 nM and includes a FAD binding moiety linked to a biotinylated polyethylene glycol (PEG)-tail. The chemoprobe exclusively binds to the free KDM1A, but not to the inhibitor-bound KDM1A in the sample. After the chemoprobe binding, the KDM1A containing complexes in the sample are captured on microtiter plates with streptavidin coated surface to determine free KDM1A, or on plates coated with a monoclonal anti-KDM1A capture antibody to determine total KDM1A. After washing, both plates are incubated with an anti-KDM1A detection antibody, washed again, and incubated with a secondary HRP-conjugated donkey anti-rabbit IgG antibody for detection using a luminescent substrate and quantification by measuring relative light units (RLU) in a luminometer (Figure 1).
Figure 1. Schema of ELISA Enzyme linked chemoprobe immunoabsorbent assay for KDM1A target engagement: A) Determination of total KDM1A using sandwich ELISA and B) Determination of free KDM1A using chemoprobe ELISA. Please click here to view a larger version of this figure.
A standard curve is included in both ELISA plates to verify the linearity of each assay. The determination of KDM1A target engagement in each sample is then calculated as a relative value to the pre-dose or vehicle treated sample.
Blood samples were obtained from the Instituto de Investigación Biomédica Sant Pau Biobank according to Spanish legislation (Real Decreto de Biobancos 1716/2011) and approval of the local ethics committees. Studies with animal tissues were performed in accordance with the institutional guidelines for the care and use of laboratory animals (European Communities Council Directive 86/609/EEC) established by the Ethical Committee for Animal Experimentation at the PRAAL-PCB.
1. Preparation of biological samples for the assay.
CAUTION: This protocol involves manipulation of biological samples which may be subjected to the Occupational Safety and Health Administration (OSHA) Blood Borne Pathogens standard (29 CFR 1910.1030), Directive 2000/54/EC of the European Parliament and of the Council of 18 September 2000 or equivalent regulations. In addition, the biological samples may contain traces of biologically active investigational chemical compounds and the protocol may involve further manipulation of such compounds. Review the safety data sheet (SDS) of the compounds used prior to the initiation of the experiment and strictly observe all applicable safety measures established in the research center, including the use of adequate personal protective equipment (PPE). Wear proper protective clothing and use proper shielding during the course of the experiment. Discard residues in the appropriate waste containers (biological/cytotoxic waste).
NOTE: This protocol starts with cells or samples of subjects treated with a KDM1A inhibitor and their untreated or vehicle/placebo treated controls3.
2. Solution preparation
3. Native protein extraction
4. Quantification of native protein using Bradford assay
5. Luminescent ELISAs for Total and Free KDM1A determination
NOTE: Keep the lab temperature constant at 23-24 °C (RT).
STANDARD SERIES | KDM1A Standard working solution (µL) | |
(pg KDM1A/well) | PBS (µL) | |
2500 for C–* | 800 | – |
2500 | 800 | – |
1750 | 560 | 240 |
1250 | 400 | 400 |
750 | 240 | 560 |
250 | 80 | 720 |
25 | 8 | 792 |
0 | 0 | 800 |
NOTE: | ||
(1) The volume prepared of each dilution is enough to run in triplicate 2 plates of assay. | ||
(2) The recommended range is between 2.5 and 5,000 pg / well | ||
* For the negative control C–, without KDM1A detection antibody |
Table 1: Standard Preparation. To prepare the Standard Series of KDM1A protein, pipette the indicated volumes of KDM1A Standard working solution and PBS into eight properly labeled 1.5 mL microcentrifuge tubes.
Table 2: Deep Well Plate Design.Standards (blue) and samples (yellow) from step 5.4.2. were pipetted into the reflected positions of the Deep Well Plate to facilitate loading into the ELISA plates following the direction of the blue (standard) and yellow (samples) arrows. Please click here to download this file.
Table 3: ELISA Plate Design. The assay plate includes the standard curve with decreasing amounts of recombinant KDM1A target (in blue); the biological samples (S) in yellow; and corresponding negative controls (contain the samples but not the primary detection antibody) in white, to be loaded on the ELISA plates from the Deep Well Plate. The blank (0 in standard curve) contains all capture and detection reagents but no sample. Please click here to download this file.
6. Calculation of the target engagement
The Linearity of Total and Free KDM1A determination.
A Standard Series was prepared as described in step 5.3.2., using 0 to 2500 pg of full-length human recombinant KDM1A enzyme. The RLU values of Total and Free rKDM1A were assessed to verify the linearity (Figure 2A and 2B). Data are represented as mean from 3 experiments with 3 technical replicates (n) ± SD. The RLU values of Total and Free KDM1A detected in human PBMCs from the blood of 3 independent volunteers are superposed on the Standard curve in Figure 2C and 2D. Blood samples were obtained from the Instituto de Investigación Biomédica Sant Pau Biobank according to Spanish legislation (Real Decreto de Biobancos 1716/2011) and approval of the local ethics committees.
Figure 2. Determination of Total and Free rKDM1A in PBMCs of healthy volunteers. RLU values of Total rKDM1A assessed by ELISA (A) and of Free rKDM1A assessed by chemoprobe capture ELISA (B). Data were obtained from 3 replicate experiments, each analyzed in triplicate (N = 3; n = 3). RLU values of Total KDM1A assessed by ELISA (C) and of Free KDM1A (D) for the PBMCs of 3 independent untreated volunteers (red, blue and green squares) superposed on Standard Curve. Data were obtained from one experiment analyzed in triplicate (N = 1; n = 3). Values represented are the means ± SD. Please click here to view a larger version of this figure.
Analysis of KDM1A target engagement in cells
AML cells were cultured following provider recommendations. Cells were treated with vehicle or ORY-1001 at different concentrations (0.25; 0.5; 1; 5 and 25 nM) (Figure 3). The native protein extracts were obtained in presence of 25 nM OG-881 chemoprobe. 0.5 µg of total protein was used to perform the target engagement analysis as described previously. Total and free KDM1A were determined, and the percentage of target engagement of ORY-1001 to KDM1A was calculated relative to the vehicle as described.
Figure 3. Dose-response of KDM1A Target Engagement in a human AML cell line. Cells were treated with vehicle or ORY-1001 at different concentrations (0.25; 0.5; 1; 5 and 25 nM) and used for determination of target engagement as described. Data were obtained from 3 replicate experiments, each analyzed in triplicate (N = 3, n = 3). Values represented are the means ± SD. Please click here to view a larger version of this figure.
Analysis of in vivo KDM1A target engagement
The objective of this experiment was to characterize the target engagement of ORY-1001 in different rat tissues, in function of the dose level. To achieve this goal, 15 Sprague-Dawley rats (200-250 g) were housed in a cytostatic security room to avoid potential contamination by the tested compound. A maximum of 3 rats/cage were randomly assigned to 5 study groups. The 5 different study groups received, respectively, vehicle; 1; 3; 10 or 30 µg / kg of ORY-1001 by oral administration for 4 consecutive days. Compound stock solutions were prepared daily. Animals were weighed before each administration to adjust the required volume. All animals were housed at constant room temperature (20 – 24 ºC) and relative humidity (45 – 65 %) under a 12 h light-dark cycle (lights on at 6:00 AM). Food and water were available ad libitum. Blood samples were collected 2 h after last administration in K2EDTA tubes and PBMCs were isolated according to the procedure described previously in step 1.2.2. and preserved at -80 °C until native protein extraction. Lung samples were also collected 2 h after the last drug administration, frozen immediately in liquid nitrogen, and stored at -80 °C. Studies were performed in accordance with the institutional guidelines for the care and use of laboratory animals (European Communities Council Directive 86/609/EEC) established by the Ethical Committee for Animal Experimentation at the PRAAL-PCB.
After pulverization, the native protein extracts from lung were obtained as described and quantified. 5 µg of total protein from pooled PBMCs or 7.5 µg of total protein from lung from 3 animals were used per dose group to run the KDM1A target engagement assay.
The dose-response of KDM1A target engagement in PBMCs and in lung treatment of rats with ORY-1001 by oral gavage, calculated relative to the vehicle group is shown in Figure 4A and 4B. As can see in Figure 4C, the ex vivo incubation with 25 nM ORY-1001 of lung protein extracts from the vehicle treated animals yields full TE yet but does not further increase TE in samples from rats treated for 4 days with 30 µg/kg ORY-1001, confirming KDM1A was already fully inhibited in vivo.
Figure 4. In vivo and ex vivo native KDM1A target engagement. Dose-response of KDM1A target engagement in PBMCs (A) and lung samples (B) from rats treated with ORY-1001 for 4 consecutive days (p.o). Data were obtained from pooled PBMCs extracts from 3 animals per cohort, analyzed in duplicate (N = 1, n = 2) or from the lungs from 3 individual animals per cohort, analyzed in triplicate (N = 3, n = 3). C. Comparison of TE in pooled lung protein extracts of rats treated with vehicle (left) or 30 µg/kg ORY-1001; after 1 h ex vivo incubation of the extracts without (grey bars) or with 25 nM ORY-1001 (blue bars) (N = 3, n = 3). All data are represented as means ± SD. Please click here to view a larger version of this figure.
The protocol presented here was developed to directly measure KDM1A target engagement using a novel KDM1A chemoprobe capture based ELISA. The method has been validated on cultured human cell lines and ex vivo samples from human, rat and mouse and baboon (including PBMCs, lung, brain, skin, tumors), but can be readily applied to other species in which the KDM1A antibody target epitopes and catalytic center are conserved. As OG-881 is an activity based chemoprobe, the sample quality is important and proper manipulation and conservation of samples should be pursued especially during the initial steps of the protocol, to ensure the KDM1A activity is conserved.
The current experimental protocol was optimized to analyze KDM1A target engagement by covalent FAD targeting inhibitors. It can also be used with reversible inhibitors that block the access to the FAD cofactor of KDM1A. Potent reversible inhibitors with long residence times may employ the unmodified protocol.
The OG-881 chemoprobe may not be suitable for low potency reversible inhibitors with high off-rates. The particular chemoprobe used in this manuscript is not cell penetrant and therefore analyses are performed ex vivo on lysed samples.
The method can be run on instruments that are broadly available in research and analytical laboratories; it does not require genetic modifications to be introduced into cells, and it can easily be applied to different sample types. Another advantage is that it can be used on samples derived from different species that are frequently used in the preclinical proof of concept studies and in toxicology models and that it has successfully been translated to analyze clinical samples.
Other methods have been used for analysis of KDM1A target engagement. Many of these methods use proxy markers like changes of the H3K4me2 histone mark, using AlphaLisa15; or induction of expression markers using qRT-PCR or FACS analysis16. However, in cells or tissues, the histone marks are controlled by multiple factors, and assays that measure changes in the histone mark do not always provide a good dynamic range for analysis. KDM1A inhibition can induce potent changes in gene and protein expression, but the response tends to very heterogeneous and highly cell context dependent, which can complicate analyses of dose response3,7.
Direct assessment of the occupation of the target is, therefore, the best option to measure target engagement. One assay that has been proposed for this is the cellular thermal shift assay (CETSA), based on the increase of thermal stability of target proteins upon binding of inhibitors. This method may, in principle, be applied to unmodified cells and different tissue types and has recently been used to assess the cellular activity of KDM1A inhibitors in cultivated cells17. However, this technology has rarely been used for in vivo pharmacodynamics studies18 and to the best of our knowledge, its use has not been reported in clinical trials.
The protocol provided here describes a fully validated chemoprobe based method which has been used to determine KDM1A target engagement in cells and tissue samples. The method has been successfully translated to analyze samples of human subjects treated with a KDM1A inhibitor19 and will be of great use to model PK/PD responses in clinical trials.
The authors have nothing to disclose.
This study was financed by Oryzon Genomics. S.A., Hoffman-La Roche, and partially supported by the CIIP-20152001 and RETOS collaboration program RTC-2015-3332-1.
0,05% Trypsin-EDTA (1X) | Thermo Scientific | #25300-062 | |
10 X Protease Inhibitor Tablets | Roche | #11836153001 | |
96 deep well storage block | VWR | #734-1679 | |
96 well ELISA plates | Nunc | #436110 | |
Adhesive black Film | Perkin Elmer | #6050173 | |
Adhesive transparent Film | VWR | #60941-062 | |
Biotinylated KDM1A probe OG-881 | Oryzon Genomics S.A. | NA | |
Bovine Serum Albumin | Sigma | # 3117057001 | |
Bovine Serum Albumin Standard | Thermo Scientific | #23208) | |
Bradford Protein Assay | BioRad | #500-0001 | |
Cell lysis buffer 10X | Cell Signaling | #9803 | |
Centrifuge for 96- well plates | Hettich | Rotina 420R | |
Flask | Thermo Scientific | #156499 | |
Full length, enzymatically active human Recombinant LSD1 / KDM1A | Active Motif | #31426 | |
Graphpad Prism 5 Project | GraphPad Software | NA | |
Luminol-Enhacer and Peroxide Solution (Chemiluminescent Substrate) | Thermo Scientific | #37074 | |
Micro Centrifuge | Eppendorf | 5415 R | |
Microplate reader Infinite 200-Tecan | Tecan | Infinite 200 | |
Mouse monoclonal capture antibody Anti-KDM1A (N-terminal epitope) | Abcam | #ab53269 | |
Needle G18 gauge blunt | BD | #303129 | |
ORY-1001 (iadademstat) | Oryzon Genomics S.A. | NA | |
PBMC separation tubes 10 ml | Greiner bio-one | #163288 | |
PBMC separation tubes 50 ml | Greiner bio-one | #227288 | |
PBS 1x | Sigma | #D8537 | |
Plate shaker | Heidolph Instruments | Rotamax 120 | |
Polysorbate 20 | Sigma | #P7949 | |
Rabbit monoclonal detection antibody Anti-KDM1A (C-terminal epitope) | Cell Signaling | #672184BF-100 | |
Secondary antibody Peroxidase-conjugated Donkey Anti-rabbit IgG | Thermo Scientific | #31458 | |
Spectrophotometer cuvette 1.5 | Deltalab | #302100 | |
Spectrophotometer for cuvette | GE Healthcare | GeneQuant 1300 | |
Streptavidin | Promega | #Z704A | |
Syringe | BD | #303172 | |
Type 1 ultrapure water | Millipore | Milli-Q Advantage A10 | |
Ultrasonic cleaner | VWR | USC200T |