Fluorescence Lifetime

Fluorescence lifetime (FLT) is a recent addition to the portfolio of readouts for biochemical protease assays. Lifetime is an intrinsic property that corresponds to the average time fluorophore electrons remain in the excited state before relaxing to the ground state. The FLT assay principle is described in Figure 2.7. In the case of a single emitting species, the probability of observing a photon at a certain time point after the fluorophore is excited follows an exponential decay (Moger et al., 2006; Eggeling et al., 2003). Changes in the physico-chemical environment of the fluorophore can lead to changes in the fluorescence lifetime. This effect can be utilized for monitoring proteolyses by introducing a specific fluorophore-quencher interaction—incorporation of the fluorophore and quencher on either side of the scissile bond of the substrate. Cleavage of the substrate leads to a separation of the interaction partners, resulting in an increase in fluorescence lifetime of the fluorophore. The use of this parameter as readout in biological assays is more beneficial compared to conventional optical readouts such as absorption, luminescence, and fluorescence intensity and may even replace current readout technologies used in drug discovery (Jäger et al., 2003; Turconi et al., 2001a).

In most applications the fluorescence lifetime is determined in the time domain, i.e., the time-dependent decay of the fluorescence emission after employing repetitive brief excitation pulses, for example, time-correlated single photon counting (TCSPC; O'Connor and Phillips, 1984). Digital electronics correlate the arrival of the fluorescence photons at the detector in relation to the excitation pulses. In a smaller number of applications, the fluorescence lifetime is calculated from the phase domain, i.e., the phase shift and demodulation of sinusoidal modulated light (Clegg and Schneider, 1996). The sensitivity of the fluorescence lifetime readout and its intrinsic nature (independent of the set-up or adjustment of the instrument or overall fluorophore concentration) result in high statistical accuracy and make FLT a valuable tool for HTS applications. The fluorescence lifetimes of the different fluorescent species present in a sample can be determined and quantified taking background and autofluorescence signals into account (Jäger et al., 2003). Similar to fluorescence anisotropy, the FLT determined for a mixture of different species is also the mean value of the

AAAA

FIGURE 2.7 Fluorescence lifetime readout principle. In the intact peptidic substrate (amino acids symbolized by X, Y and Z) labeled with a fluorophore (dye) and a fluorescence quencher (Q) at the opposite sites of the scissile bond, a short fluorescence lifetime is observed (light gray dotted arrow) after excitation with a pulsed laser (dark gray dotted arrow) due to an energy transfer to the quencher (ET). After cleavage of the substrate between amino acids X and Y by a protease, the energy transfer is disrupted and a longer fluorescence lifetime is observed (light gray continuous arrow). An increase in fluorescence lifetime dependent on the enzymatic velocity is recorded.

FIGURE 2.7 Fluorescence lifetime readout principle. In the intact peptidic substrate (amino acids symbolized by X, Y and Z) labeled with a fluorophore (dye) and a fluorescence quencher (Q) at the opposite sites of the scissile bond, a short fluorescence lifetime is observed (light gray dotted arrow) after excitation with a pulsed laser (dark gray dotted arrow) due to an energy transfer to the quencher (ET). After cleavage of the substrate between amino acids X and Y by a protease, the energy transfer is disrupted and a longer fluorescence lifetime is observed (light gray continuous arrow). An increase in fluorescence lifetime dependent on the enzymatic velocity is recorded.

lifetimes of the individual species in the mixture. Accordingly, the fraction of the formed product can be calculated as:

where tsub is the fluorescence lifetime of the substrate and tprod is the fluorescence lifetime of the product. As with fluorescence anisotropy, the factor Q is necessary to correct for differences in FI of the fluorophore in the parent substrate (FIsub) and in the product (FIprod): Q = FIprod/FIsub. However, the applicability of the FLT for biological reactions is difficult to predict because FLT changes are not always predictable due to environmental quenching, solvent, or polarity effects. The lifetimes of most fluorophores are typically in the range of 0.1 to 10 ns when determined in the red region of the spectrum. Moger et al. (2006) determined the lifetimes of DY-633 at 0.2 ns, EVOblue at 0.65 ns, and MR-121 at 1.85 ns.

The same publication indicated that fluorescence lifetimes of compounds from the compound collection of Pfizer classified as problematic due to their autofluorescence characteristics resulted in false positive results in many FI-based assays. For most compounds, the fluorescence lifetimes were below 1 ns. Thus, FLT measurements with a reporter fluorophore displaying a lifetime significantly longer than 1 ns are suitable for application in protease assays for compound testing. However, for applications under initial velocity conditions with a substrate turnover below 20%, fluorophores with lifetimes of a few nanoseconds are still critical because the dynamic range of the assay is then too low with lifetimes below 1 ns.

Under confocal settings S/N ratios sufficient for high-throughput applications can be achieved with these fluorophores (Moger et al., 2006). Under standard optical settings, dynamic ranges far above 1 ns are required for robust screening assays.

Especially attractive for protease assays among the fluorophores suitable for fluorescence lifetime measurements is a group of dyes that can be quenched by natural amino acids in the peptide substrate. MR121, ATTO651, 2,3-diazabicyclo[2.2.2]oct-2-ene (DBO), and Puretime™ 14 (PT 14) belong to this class of dyes (Marme et al., 2004; Hennig et al., 2006, 2007). The long lifetime of DBO allows a time-gated intensity measurement, separating the probe signal from any unwanted, short-lived fluorescence intensity arising from compounds or other assay components, thereby increasing assay robustness. The DBO-based FLT assay strategy was applied to both endopeptidases and carboxypeptidases. Hennig et al. (2007) describe an assay for carboxypeptidase A using peptidic substrates, with DBO included into the non-primed-site sequence and a tryptophan as a C terminal amino acid in the S1' position serving as a quencher.

Smith et al. (2004) described the synthesis and the fluorescence properties of acridones and quinacridones. First pilot studies of protease (and kinase) activity assays demonstrated that the fluorescence lifetimes of certain acridone dyes are significantly reduced by a tyrosine residue in close proximity to the dye (Graves et al., 2005, 2006). Recently, the application of the PT14 acridone dye in a kallikrein 7 assay was described (Doering et al., 2009). The authors showed that the fluorescence lifetime of PT 14 was quenched by a tyrosine and even more efficiently by a tryptophan in close proximity to the dye. This allows the reduction from two labels within the peptide substrate (as in FRET assays) to one label plus one natural amino acid on either side of the scissile bond. As a result, artifacts in the interaction between protease and substrate on the one hand and between substrate and compound on the other hand, created by the introduction of two artificial dyes into the natural recognition sequence of a protease, are reduced. The enzymatic cleavage leads to a spatial separation of the probe and the tryptophan, resulting in an increase in the FLT of PT14. This increase correlates with the progress of substrate turnover.

Protease assays based on the FLT of PT14 are especially attractive due to the long lifetime of the dye of 14 ns that allows the discrimination of the lifetimes of short-lived fluorescent compounds from that of PT14. Thus, the rate of false-positive and -negative results can be reduced. The major

FIGURE 2.8 Dose-response curves for a protease inhibitor with autofluorescence characteristics. (a) Result of standard fluorescence intensity-based assay employing an EDANS/DABCYL-labeled substrate at a concentration of 2 |M. The profiling data (IC50 of 6 nM and Hill coefficient of 0.9) could be obtained only after excluding the four data points for compound concentrations above 0.3 pM from the curve fit (excluded data are marked with brackets; the point at 30 pM and -128% inhibition is not shown). (b) Result obtained with the novel fluorescence lifetime-based assay employing a PT14-labeled substrate at a concentration of 1 |M. The profiling data (IC50 of 4 nM and Hill coefficient of 0.9) were obtained without manipulation of data.

FIGURE 2.8 Dose-response curves for a protease inhibitor with autofluorescence characteristics. (a) Result of standard fluorescence intensity-based assay employing an EDANS/DABCYL-labeled substrate at a concentration of 2 |M. The profiling data (IC50 of 6 nM and Hill coefficient of 0.9) could be obtained only after excluding the four data points for compound concentrations above 0.3 pM from the curve fit (excluded data are marked with brackets; the point at 30 pM and -128% inhibition is not shown). (b) Result obtained with the novel fluorescence lifetime-based assay employing a PT14-labeled substrate at a concentration of 1 |M. The profiling data (IC50 of 4 nM and Hill coefficient of 0.9) were obtained without manipulation of data.

advantage of this novel assay format over the previously used standard fluorescence intensity assays employing EDANS and DABCYL as a fluorophore and quencher pair can be illustrated by the direct comparison of representative dose-response curves obtained for a proprietary inhibitor with autofluorescence characteristics. Upon excitation at 350 nm, the compound autofluorescence interferes with the intensity readout (Figure 2.8). In the case of the FRET quench assay, the four data points obtained for the compound concentrations above 0.3 |jM were excluded manually prior to the curve fitting to yield IC50 data. For the FLT-based assay, no interference was observed and the curve fit could be conducted automatically without exclusion of data points. For the example shown here, the data sets from both assay formats yield similar IC50 values in the range of 4 to 6 nM. The FLT-1 to FLT-5 examples in Table 2.2 illustrate the dynamic ranges that can be achieved with FLT assays based on PT14.

PT14 is comparatively cost efficient and thus affordable for high-throughput applications. We know of only one disadvantage of FLT as a readout for biochemical protease activity assays: the requirement for a detection device equipped with a pulsed laser diode and TCSPC capabilities. According to our knowledge, such readers are currently offered by only two companies (Tecan and Edinburgh Instruments).

0 0

Post a comment