Proteases are essential for many physiological processes; thus, their dysregulation has been implicated in a variety of disease processes such as infectious disease, cardiovascular disease, cancer and others(1)
. To date, the FDA has approved six protease inhibitor drugs for clinical use, including ACE, HIV protease and proteasome inhibitors(2)
. Proteases constitute 5–10% of all pharmaceutical targets for small-molecule drug discovery(2)
. And, with over 560 proteases or protease homolog coding regions annotated from the human genome(3)
, this target class continues to be an important and active area for drug discovery.
Critical for studying novel proteases is first understanding how they bind to and process their substrates. A standard method for determining the consensus sequences for protease recognition is to use a synthetic combinatorial library of fluorescently labeled short polypeptides(4)
. While rigorous, synthesis of these libraries requires specialized equipment and expertise, which are not readily available to most life science researchers. Although peptides can be purchased from commercial sources, they can be expensive, and some sequences may not be commercially available. As an alternative, some researchers create fusion protein libraries. Sequence specificity is then ranked based on the percent of digested product measured by gel densitometry. This approach, in addition to being laborious and time-consuming, has a limited dynamic range and sensitivity. To address the need for simple, sensitive, and readily available tools for the elucidation of protease recognition sequence specificity, we developed the Protease-Glo™ Assay.
Assay Concept and Procedure
The Protease-Glo™ Assay takes advantage of the conformational change between the open inactive form and the closed active form of firefly luciferase (Photinus pyralis) (Figure 1; (7)
). The native N- and C-termini of the luciferase protein are covalently linked through the insertion of a short polypeptide sequence containing a protease recognition sequence. In this configuration, the luciferase activity is greatly reduced. After proteolytic cleavage by the cognate protease, the luciferase activity increases. This change in luciferase activity is detected by adding the luciferase detection reagent.
Linking the N and C termini creates a circular protein. Therefore, expression of the mutant luciferase used in this assay requires the creation of new N and C termini at positions tolerant to proper folding and function of the enzyme. In the Protease-Glo™ Assay, the new N- and C-termini are located at positions 234 and 233, respectively. The resulting circularly permuted form of firefly luciferase is called the GloSensor™ protein. The Protease-Glo™ Assay kit (Cat.# G9451) contains all the necessary reagents for cloning and expressing the GloSensor™ protein as well as the reagents for detecting the luminescent readout(7)
"The Protease-Glo™ Assay's open-ended format allows researchers to clone in any recognition sequences desired for their protease of interest…"
Protease Detection: Broad Applicability and High Sensitivity
Because proteases constitute a large and varied class of proteins having broad functional characteristics, an ideal assay should be broadly compatible with diverse reaction conditions, such as pH, yet use the same simple detection format. The assay should yield a sensitive, linear readout of protease activity over a wide dynamic range throughout these reaction conditions.
The Protease-Glo™ Assay's open ended format allows researchers to clone in any recognition sequences desired for their protease of interest (typically less than 14 amino acids in length). An examination of a wide variety of proteases indicates that the Protease-Glo™ Assay has broad applicability over several protease classes (Figure 2;(7)
). Importantly, proteases with and without P´ requirements (nomenclature of Schechter and Berger;(9)
) are both able to activate the GloSensor™ proteins.
When we digested the GloSensor™ proteins with the cognate protease, the fold activation ranged from 160 to 1,250 (Figure 2; (7)
). Similar results were seen with two additional proteases, the enterokinase and PreScission® proteases, resulting in fold activations of 185 and 2,600, respectively. Furthermore, the assay is functional over a pH range of 5.2–9.1. Thus, it can accommodate the diverse buffer requirements of different proteases(7)
Figure 2. Examples of GloSensor™ [protease site] activation by protease digestion.
Plasmid DNA encoding seven different GloSensor™ [protease site] proteins with protease recognition sequences for the indicated proteases were isolated. Plasmids encoding a GloSensor™ [42AA] protein containing a 42-amino-acid sequence or the full-length firefly luciferase protein were created for use as controls. Plasmid DNAs were transcribed and translated using TnT® SP6 High-Yield Protein Expression System for 2 hours at 25°C; a no-DNA control reaction was performed in parallel. Following expression, 25µl of the TnT® reaction was combined with 25μl of 2X protease buffer and the following cognate proteases: HIV-1 protease (30 units, AnaSpec), caspase-3 (1.7 units, UpState Biotechnology), caspase-8 (2.5 units, Chemicon), GST-SARS-CoV 3CL (6μg), ProTEV (3 units, Promega), granzyme B (112 units, BioMol), PSA (3.2μg, Sigma-Aldrich). For the GloSensor™ [42AA] protein and full-length firefly luciferase protein, 25µl of the TnT® reaction was combined with 25µl of 100mM HEPES (pH 7.5). Following incubation at 30°C for 1 or 2 hours, aliquots of each reaction were added to Bright-Glo™ Assay Reagent in triplicate and incubated for 5 minutes at room temperature. Luminescence was measured using a GloMax® 96 Microplate Luminometer and 1-second integration time. Error bars are the standard deviation of the mean.
To assess sensitivity and dynamic range of the Protease-Glo™ Assay for protease detection studies, we serially diluted the Tobacco Etch Virus (TEV) protease(8)
. The signal-to-noise ratio was calculated from the luminescent values and is defined as the (mean signal minus mean background) divided by the standard deviation of the background (Figure 3). The limit of detection was defined as the amount of TEV protease detected at a signal-to-noise ratio of 3 (i.e., 3 standard deviations higher than the background). Our results indicate that the Protease-Glo™ Assay is highly sensitive with a detection limit of 0.163mU TEV protease / sample and has a wide dynamic range, ≥1,000-fold linear range (0.163 to 167mU [the highest concentration tested in this experiment] Figure 3). These results are similar to what we have described for the caspase-3 protease, where the assay also had ≥1,000-fold linear range and was over 1,000-fold more sensitive than SDS-PAGE.
Protease Function: Evaluation of Substrate Specificity
Because GloSensor™ proteins are quickly and easily generated, multiple substrates may be interrogated with the same protease. Previously, we compared all 20 amino acids at the P1′ position of the TEV protease recognition sequence(8)
. The results were consistent with other methodologies but took only 10 days from cloning the 20 GloSensor™ proteins to the end results.
In just four days we were able to clone the GloSensor™ proteins and compare the Hepatitis C Virus (HCV) NS3/4A protease activity for the HCV junction sequences for 4A/4B: EFDEMEECSQHLPY and 5A/5B: EDVVPCSMSY. The 5A/5B GloSensor™ protein was activated more by the HCV NS3/4A protease than the 4A/4B GloSensor™ protein (206-fold versus 101-fold activation; Figure 4, Panel A). Importantly, the luminescent values correlate with the SDS-PAGE results, which show more digestion product for the 5A/5B GloSensor™ protein than the 4A/4B GloSensor™ protein (Figure 4, Panel B). The Protease-Glo™ Assay, however, is less labor-intensive than SDS-PAGE gels and provides a more sensitive and quantitative result over a wider dynamic range. These results are consistent with previous reports in the literature(11)
Figure 4. HCV activation by protease digestion.
Panel A. Plasmid DNA encoding two different GloSensor™ [HCV protease site] proteins, a GloSensor™ [42AA] protein containing a 42-amino-acid sequence and the full-length synthetic firefly coding region, were isolated and then transcribed and translated using TnT® SP6 High-Yield Protein Expression System for 2 hours at 25°C; a no-DNA control was performed in parallel. Following expression, 22µl of the TnT® reaction was combined with 22µl of 2X protease buffer and HCV NS3/4A protease (33 units, AnaSpec) or water. Following incubation at 30°C for 1 hour, aliquots of each reaction, in triplicate, were added to Bright-Glo™ Assay Reagent and incubated for 5 minutes at room temperature. Luminescence was then measured using a GloMax® luminometer (1-second integration time). Error bars are the standard deviation of the mean. Panel B. The GloSensor™ [HCV protease site] proteins, a GloSensor™ [42AA] protein containing a 42-amino-acid sequence, the full-length synthetic firefly protein and no-DNA control reaction were labeled using FluoroTect™ GreenLys tRNA, size-fractionated on a 4–12% Bis-Tris NuPAGE® gel (Invitrogen), then visualized. In the reaction containing the GloSensor™ [HCV protease site] proteins without HCV NS3/4A protease digestion (–lanes) only the uncut 61kDa proteins are visible. However, after HCV NS3/4A digestion, the smaller 36kDa and 25kDa expected-sized protein fragments also are visible (+lanes). Note that the amount of protease digestion products correlate with the fold activations calculated from the luminescent signals. Since the digestions were incomplete, the uncut 61kDa protein is also visible. Only the uncut 61kDa proteins and no digestion products are visible in the GloSensor™ [42AA] protein or the full-length synthetic firefly protein lanes with or without HCV NS3/4A protease digestion. No labeled proteins were observed in the no-DNA control reactions, with or without HCV NS3/4A protease digestion. The similar amount of free labeled tRNA visible in each lane indicates that a similar amount of lysate was added to each lane.
The Protease-Glo™ Assay offers an excellent tool for quickly and cost-effectively determining optimal sequence specificity without chemical synthesis. Once the optimal protease recognition sequence is determined, GloSensor™ proteins are ideally suited for many subsequent protease studies including protease detection, comparing sequence specificity between related proteases or viral strains, small molecule screening, and protease modulator characterization.
However, because the substrate (GloSensor™ protein) is made in a cell-free lysate it is difficult to quantify. This limits the utility of this assay for detailed protease enzymatic characterization. If detailed enzymatic characterization is required, the pre-optimized peptide sequence can first be quickly and easily determined using the Protease-Glo™ Assay and then ordered. Typically these sequences are provided either labeled with a fluorophor or as a modified aminoluciferin substrate. Aminoluciferin substrates have the added advantage of offering all the benefits afforded bioluminescent assays(12)
Protease Modulation: Identification and Characterization Modulators by High-Throughput Screening
Using the optimal protease recognition sequence, GloSensor™ proteins can be used to screen for protease small-molecule modulators. TEV protease is widely used as a tool for removing affinity tags after protein purification. In some cases, it is desirable to "turn off" the TEV protease after purification for the downstream applications(8)
. Unfortunately, when we tested many of the commonly used cysteine inhibitors, they either did not inhibit TEV protease, a cysteine protease, or they significantly interfered with downstream steps or, in the case of the CMK peptide inhibitor, they were too expensive. We next sought to find a novel small-molecule TEV inhibitor by screening the Library of Pharmacologically Active Compounds (LOPAC1280; Sigma-Aldrich). However, the standard method for assessing TEV protease activity, the protein fusion cleavage method, is not practical for a large-screening campaign.
Therefore, the LOPAC1280 was screened for TEV protease inhibitors in 384-well plates using the
Protease-Glo™ Assay with a pre-optimized TEV recognition sequence(8)
. We included a GloSensor™[42AA] protein containing a 42-amino-acid sequence in the screen to eliminate nonspecific GloSensor™ inhibitors.
The results from a representative 384-well plate are shown in Figure 5. We identified twelve compounds that inhibited TEV activity greater than 60% (out of 1280 compounds, 0.9%). Five of these were retested and IC50 potency determined using both the standard protein fusion cleavage assay and the Protease-Glo™ Assay (Figure 6). The results from the two assays were the same in terms of confirming three of the five hits as well as the relative IC50 ranking of two of the three hits. We could not determine the IC50 for one compound, aurintricarboxylic acid (ATA), using the Protease-Glo™ Assay because of color interference. We did not retest all 12 compounds, and so we could not determine an exact false-hit rate. However, the false-hit rate range was very low (2–9 compounds out of 1280, or 0.16–0.70%.)
Figure 6. Inhibitor potency titration curves.
Five compounds were retested at titrating concentrations. The five compounds were: aurintricarboxylic acid (ATA), 8,8′-[carbonylbis(imino-3,1-phenylenecarbonylimino)]bis(1,3,5-naphthalene-trisulfonic acid) hexasodium salt (NF-023), 4-[3-(4-acetyl-3-hydroxy-2-propylphenoxy)propoxy]phenoxyacetic acid (L-165,041), SCH-202676 hydrobromide or N-(2,3-diphenyl-1,2,4-thiadiazol-5-(2H)-ylidene)methanamine hydrobromide (SCH HBr), and 6-hydroxyl-DL-DOPA or 2,5-dihydroxy-DL-tyrosine (L-DOPA). Panel A. Using the Protease-Glo™ Assay, NF-023 and ATA were tested from 0–0.2mM, and the other three compounds were tested from 0–1mM. Data are the mean values, n=3. Error bars are the standard deviation of the mean. Panel B. Using the protein fusion cleavage assay, compounds were tested from 0–0.1mM. Data are the average of n=2.
Proteases are crucial for almost all aspects of biology. However, current methods for studying novel proteases are either expensive, laborious or require chemical synthesis expertise and equipment. In contrast, the Protease-Glo™ Assay allows you to rapidly generate protease substrates using molecular cloning and cell-free coupled transcription/translation expression enabling easy evaluation of protease function without the need for chemical synthesis or the expense of purchasing peptides(7)
The assay easily supports several protease classes and can be used with proteases with and without P´ requirements (Figure 2). Furthermore, the bioluminescent readout is easily quantitated, highly sensitive, and proportional to protease activity across a wide dynamic range. Applications for this assay include evaluating protease recognition sequence specificity and establishing an optimized assay to identify and characterize protease inhibitors using high-throughput screening methods.