The botulinum neurotoxin type A light chain (BoNT/A LC) is a metalloprotease that enters motor neurons, cleaves its substrate SNAP-25, and disrupts neurotransmission, thereby resulting in flaccid paralysis. Utilizing a high-throughput-compatible FRET-based assay, large libraries of small molecules can be screened for their impact on BoNT/A LC enzymatic activity.
Botulinum neurotoxin (BoNT) is a potent and potentially lethal bacterial toxin that binds to host motor neurons, is internalized into the cell, and cleaves intracellular proteins that are essential for neurotransmitter release. BoNT is comprised of a heavy chain (HC), which mediates host cell binding and internalization, and a light chain (LC), which cleaves intracellular host proteins essential for acetylcholine release. While therapies that inhibit toxin binding/internalization have a small time window of administration, compounds that target intracellular LC activity have a much larger time window of administrations, particularly relevant given the extremely long half-life of the toxin. In recent years, small molecules have been heavily analyzed as potential LC inhibitors based on their increased cellular permeability relative to larger therapeutics (peptides, aptamers, etc.). Lead identification often involves high-throughput screening (HTS), where large libraries of small molecules are screened based on their ability to modulate therapeutic target function. Here we describe a FRET-based assay with a commercial BoNT/A LC substrate and recombinant LC that can be automated for HTS of potential BoNT inhibitors. Moreover, we describe a manual technique that can be used for follow-up secondary screening, or for comparing the potency of several candidate compounds.
Botulinum neurotoxin A (BoNT/A), the most potent toxin currently known (LD50 ~1 ng/kg)1, is a potent neurotoxin that is produced by the bacterium Clostridium botulinum. Within an affected host, BoNT/A disrupts neurotransmission at the neuromuscular junction by binding motor neurons, internalizing into the cytosol, and ultimately cleaving neuronal proteins that are essential for acetylcholine exocytosis. Once inside a neuron, BoNT/A can persist for as long as several months2. Long-term inhibition of acetylcholine release hampers normal muscle contraction and results in flaccid paralysis, which, in severe cases, may result in cardiac and/or respiratory failure. Because of its extreme potency, ease of production, and long-term effects within the host, the CDC has labeled all BoNT serotypes as high-risk bioterrorism agents.
The mechanism of action of the toxin involves numerous steps, including binding to neuronal surface receptors, cellular uptake via receptor-mediated endocytosis, and translocation into the neuron cytosol. BoNT/A is comprised of two chains, a heavy and light chain, and both chains are required for toxicity. The heavy chain (HC) contains binding and translocation domains, while the light chain (LC) is a zinc-dependent metalloprotease that translocates from the endosome to the cytoplasm. Once inside the cytosol, the LC/A metalloprotease localizes to the inner cytoplasmic membrane and cleaves the membrane-bound host protein SNAP-25. SNAP-25 is a member of the SNARE (Soluble N-Ethylmaleimide-Sensitive Factor Attachment Protein Receptor) protein family, which plays a crucial role in acetylcholine exocytosis (reviewed in reference3). LC/A cleavage of SNAP-25 impairs SNARE complex function, which inhibits acetycholine neurotransmitter release and impairs muscle contraction.
Currently, treatment for botulism is limited and often includes administration of an equine neutralizing antibody4; however, because the toxin is rapidly internalized into neurons, the antibody has the antibody has a narrow time window of administration. Thus, many researchers believe that the BoNT/A LC may be a better therapeutic target. Because the LC is a zinc-dependent metalloprotease, one approach to inhibit LC/A activity has been to develop compounds that chelate the active-site zinc ion. For example, hydroxamate compounds chelate the active-site zinc and have excellent in vitro potency (Ki of the best BoNT/A LC small molecule inhibitor to date is 77 nM)5. However, many small molecules fail to advance as therapeutics due to various problems ex vivo or in vivo, including poor aqueous solubility, rapid metabolism, and/or high cytotoxicity. Therefore, new compounds with improved pharmacological and pharmacokinetic properties are needed. Small molecule compound identification often involves high-throughput screening (HTS) to identify novel scaffolds. Initial methods for BoNT/A LC activity screening were based on HPLC detection of short peptide substrate cleavage, which is time-consuming and not amenable to HTS applications6-8. Subsequently, Schmidt and colleagues9 developed a high-throughput BoNT/A LC activity assay that utilizes a fluorescein-labeled peptide substrate covalently attached to a microtiter plate. The BoNT/A LC cleaves the substrate and releases fluorescein, which can be quantified with a fluorometer. The plate format of this assay allows numerous compounds to be screened simultaneously; however, the assay requires labeling synthetic peptides with fluorescein and coating the assay plates with derivatized substrate molecules, which are cumbersome techniques. A much simpler method for detecting BoNT/A LC activity at low concentrations was later described by Schmidt et al., where a series of fluorogenic substrates were utilized to monitor BoNT LC activity in real time10. Additional techniques described in the literature include a depolarization after resonance energy transfer-based assay to detect and quantify BoNT activity in crude extracts; this method can be used for high-throughput applications10,11, although it requires sophisticated equipment to measure fluorescence resonance energy transfer (FRET) and polarization signals. Finally, several cell-based models for BoNT intoxication have been reported (reviewed in reference11) that will enable researchers to study the often limiting properties of compounds previously mentioned, including cytotoxicity, cell permeability, and stability. However, most of the existing cell-based assays are not amenable to HTS, and are labor and time intensive.
Herein, we describe a detailed protocol for a HTS method that utilizes the commercially available FRET-based BoNT/A LC substrate. The substrate is based on the SNAP-25 cleavage sequence and is a synthetic 13-mer peptide that contains a terminal fluorophore and quencher. BoNT/A cleavage separates the fluorophore and quencher, abolishing FRET and increasing measured fluorescence, which can be continually measured in a fluorometer plate reader. The assay is used routinely in our, as well as other laboratories, to identify new classes of BoNT/A LC inhibitors or to determine the relative potency of previously identified compounds5,12-15. This assay is suitable for HTS because of its simplicity, automation potential, low cost of materials, and ability to screen numerous compounds simultaneously (see reference16; Caglič et al., submitted; Bompiani et al., in preparation). In addition to HTS, this assay can be used to compare the relative potency of compounds by determining the IC50 value (concentration required to inhibit 50% of BoNT/A LC activity) of a compound. The assay can either be performed manually in a 96-well format (Manual Screening section of the Protocol Text) or can be automated in a 384-well format for HTS (Automated Operation section of the Protocol Text).
Manual Screening or IC50 Determination
This protocol can be used to determine the relative potency of a compound (IC50 value) by preparing a dilution series of the compound, or to manually screen for small-molecule inhibitors at a single concentration.
1. Preparation of Buffers, Reagents, and Required Instrumentation
2. Preparation of Compound Dilution Series
3. Plate Preparation and Spotting (Table 1)
4. Substrate Addition and Fluorescence Measurement
Table 1. Suggested assay plate layout.
Well # | Well description | Compound | Assay Buffer | Substrate | LC | DMSO |
---|---|---|---|---|---|---|
1 | Compound dose 1 | 1 μl | 79 μl | 10 μl | 10 μl | — |
2 | Compound dose 2 | 1 μl | 79 μl | 10 μl | 10 μl | — |
3 | Compound dose 3 | 1 μl | 79 μl | 10 μl | 10 μl | — |
4 | Compound dose 4 | 1 μl | 79 μl | 10 μl | 10 μl | — |
5 | Compound dose 5 | 1 μl | 79 μl | 10 μl | 10 μl | — |
6 | Compound dose 6 | 1 μl | 79 μl | 10 μl | 10 μl | — |
7 | Compound dose 7 | 1 μl | 79 μl | 10 μl | 10 μl | — |
8 | Compound dose 8 | 1 μl | 79 μl | 10 μl | 10 μl | — |
9 | Compound dose 9 | 1 μl | 79 μl | 10 μl | 10 μl | — |
10 | Negative control (100% LC activity) |
— | 79 μl | 10 μl | 10 μl | 1 μl |
11 | Compound intrinsic fluorescence control | 1 μl | 89 μl | 10 μl | — | — |
12 | Positive control (0% LC activity) |
1 μl | 79 μl | 10 μl | 10 μl | — |
Automated Operation for High-Throughput Screening
5. Preparation of Buffers, Reagents, Compound Plates, Barcodes, and Required Instrumentation
6. Preparation of Pin Tool
7. Preparation of Low-volume Liquid Dispenser
8. Dispensing BoNT/A Light Chain into Assay Plate
9. Stamping Compounds into Assay Plate
10. Dispensing Substrate into the Assay Plate
11. Fluorescence Measurements
12. Cleaning up
The FRET-based BoNT/A LC assay can be performed manually in a low-throughput manner to characterize known inhibitors or screen small libraries; alternatively, the protocol can be scaled-down and automated for high-throughput screening (HTS) with large libraries with the aim of identifying novel BoNT/A LC inhibitors. Regardless of the approach taken, an increase in fluorescence should be observed over time when BoNT/A LC is incubated with the substrate (Figure 1A shows a representative plot of fluorescence over time). If serial dilutions of a compound are tested, then a series of lines with varying slopes is often obtained (Figure 1B). Well 10 in the plate layout described here serves as a negative control as vehicle only (DMSO) is added; the rate of this reaction is defined as "100% enzyme activity" (or uninhibited activity), and rates in the presence of compound can be normalized to this rate to obtain the relative rate, or percent inhibition. Wells 11 (compound intrinsic fluorescence control) and 12 (positive control) should show horizontal curves (i.e. slopes close to 0).
When the assay is performed with serial dilutions of a compound, a graph of the rate of the reaction (assay slope) versus the inhibitor concentration can be used to determine the IC50 (Figure 1C). The dilution range should be chosen so that the plot appears sigmoid-shaped for optimal curve fitting. Most scientific graphing or statistical analysis software packages enable nonlinear fitting of the sigmoid curve to calculate the concentration of an inhibitor required to inhibit 50% of enzymatic activity (i.e. IC50 value). The IC50 value is a relative measure of potency best described as an apparent Ki value, as it depends upon the concentration of enzyme present in the assay. However, the potency of several compounds can be quickly compared and ranked with the protocol outlined here. It is important to note that in the assay described here one is capturing competitive inhibitors or the competitive component of noncompetitive inhibitors. This assay will not capture uncompetitive inhibitors or the uncompetitive component of noncompetitive inhibitors since these bind to the enzyme substrate complex, the concentration of which is vanishingly low when substrate concentration is much lower than the Km.
Proper mixing of the plate after adding each component is crucial to obtain a smooth linear increase in fluorescence over time, which is necessary to accurately determine the initial velocities. Figure 1D shows a representative experiment where there was insufficient mixing and the rate of product formation is not linear over time, confounding accurate data analysis.
Figure 1. A) Representative graph showing an increase in fluorescence (RFU, relative fluorescence units) upon cleavage of the FRET substrate by BoNT/A LC over time. Under the conditions described in the Protocol Text, increase in fluorescence is linear for greater than 60 min. The slope of the linear portion of the data corresponds to initial velocity. B) Serial dilution of a hydroxamate-based compound results in a dose-dependent increase in the amount of product formed over time. The concentrations shown are final concentrations of the compound in μM. C) Initial velocities (in RFU/min) are plotted against the inhibitor concentration, resulting in a sigmoidal dose-response curve. The concentration of an inhibitor required to reduce the initial velocity to 50% of its uninhibited value is the IC50 value. D) Representative image of a low-quality data set. Note the spikes that are present at t=30 min are due to insufficient mixing; these nonlinear data make determining the initial velocities challenging. Click here to view larger image.
Figure 2. Representative data output from a fluorescence plate reader. The image shows a heat-plot diagram of one time point measurement of a 384-well plate from a HTS campaign. The numbers in the wells represent fluorescence in RFU. Click here to view larger image.
The FRET-based BoNT/A LC assay described here represents an attractive method for identifying and characterizing small molecules that modulate BoNT/A LC activity. The solution-based nature of the assay makes this protocol amenable to high-throughput screening described in the Automated Operation section of the Protocol Text. The Z-factor is often used to determine the suitability of an assay for HTS campaigns18. Determined as Z = 1 – 3 (σp + σn)/(μp – μn), a Z-factor of 1 is ideal, between 0.5-1 is excellent, and less than 0.5 is marginal. The Z-factor of the HTS protocol as described here is 0.89 (standard deviation of the positive and negative controls are σp = 6.67 RFU/min and σn = 0.788 RFU/min, respectively; mean velocity of negative and positive controls are μp = 196.1 RFU/min and μn = -7.67 RFU/min, respectively), which indicates that this assay is well suited for HTS.
The assay tolerates up to 2% DMSO in the final volume without significant degradation of the assay signal or impact on enzyme function. Previous studies have also shown that 0.01% Tween 20 increases the mean velocity of the assay12. The velocity can be further increased by performing the assay at 30-32 °C, although typically the assay is performed at room temperature for convenience. The between-plate and day-to-day variation of the velocity of the negative control (uninhibited reaction) is less than 10% (coefficient of variation, CV = σp/μp).
It is important to note there are several critical parameters to consider when using this assay. Inherent properties of certain small-molecule compounds may interfere with the fluorescence signals. For example, some small molecules may exhibit auto-fluorescence and thereby impact the assay signal. The inclusion of the intrinsic compound fluorescence control (well 11 in the Manual Screening section of the Protocol Text) can determine whether a compound is auto-fluorescent. If a compound auto-fluoresces and interferes with the fluorescent readout of the assay, the signal in well 11 will appear higher than the signal for DMSO, and the compound may appear less potent. One way to compensate for compound auto-fluorescence is to subtract the signal from this control well (well 11) from the raw fluorescent kinetic data. On the other hand, a compound that quenches the fluorescence signal of the assay will decrease the apparent signal and the compound will consequently appear more potent. To determine if a molecule is quenching product signal, the FRET substrate can first be cleaved with a high concentration of BoNT/A LC (1-425) (e.g. 1 μM or higher) until no further increase in fluorescence is observed. Compound is then added to the fully cleaved substrate, and the change in fluorescence is recorded. A quenching compound will decrease the fluorescence signal, and such compounds can be analyzed in an alternative BoNT/A LC activity assay that does not utilize a fluorescent substrate or cell-based assay.
Although this FRET-based assay can be utilized to rapidly identify new scaffolds or compounds through HTS, or quickly compare the potency of several compounds, the assay cannot be utilized to determine more rigorous kinetic constants such as the Ki. Because the KM of the FRET substrate is well above 1 mM, any assay with this substrate is performed under substrate-limiting conditions19. To address these shortcomings, a more sensitive assay amenable to steady-state kinetics has been reported19. However, because this assay utilizes liquid chromatography coupled with mass spectrometry to detect cleavage products, it is not amenable for high-throughput applications. In our laboratory, we frequently use FRET-based screening as a primary triage of large compound libraries, with LC-MS-based peptide cleavage assays serving as secondary screens. Another FRET-based assay that is amenable to steady-state kinetics is based on a 66-mer FRET substrate and was shown to be effective for HTS (Z' values > 0.9)20. A clear advantage of the system is the ability to produce the FRET substrate by heterologous expression in E. coli, thereby reducing the operational costs, but to our knowledge has not been applied to a HTS campaign.
In summary, the protocol described here utilizes a FRET-based, commercial substrate to monitor botulinum neurotoxin light chain enzymatic activity. This assay allows for rapid, high-throughput screening of large libraries of small molecules to identify novel compounds that modulate enzymatic activity. This method can be used for both primary and secondary screening, as well as to compare the relative potency of identified compounds and rank compounds of interest based on their activity. Once initially characterized with this protocol, compounds can be further rigorously tested in the additional assays previously described (LC-MS, cell-based assays, etc.).
The authors have nothing to disclose.
This work was supported by a grant from the National Institutes of Health (AI082190 to T.J.D.) and the California Institute for Regenerative Medicine (TB1-01186 and CL1-00502).
HEPES | Teknova | H1021 | |
Tween-20 | Fisher Scientific | BP337-100 | |
Methanol (HPLC-grade) | Sigma-Aldrich | 34860 | |
Isopropanol (HPLC-grade) | Sigma-Aldrich | 650447 | |
96-well Black assay plate | Costar | 3915 | |
384-well Low-volume black assay plate | Greiner | 788076 | |
SNAPtide FITC/Dabcyl substrate | List Biological Laboratories | 521 | FRET-based BoNT/A LC substrate |
Pin cleaning solution | V&P Scientific | VP 110 | |
Lint-free blotting paper | V&P Scientific | VP 540DB | |
Biomek Seal and Sample Aluminum foil lids | Beckman Coulter | 538619 |