Multimerization studies

Although aggregation can lead to artifacts in other binding experiments, we hypothesized that our assay, with its long lifetime of the luminescent label, should be beneficial for investigating proteins prone to multimerization. The addition of an excess of unlabeled MIA protein to MIA-Ru(bpy)₃ does not change FP (Figure 4A), indicating that the size of the multimers does not change and that there are no aggregates consisting of ≥10 molecules. We estimate this from the lifetime of the label and the molecular weight of the protein by the Perrin equation. To demonstrate the existence of smaller aggregates, we coated wells with a MIA-biotin conjugate. Indeed, a large increase in FP was detected, indicating the presence of direct MIA-MIA interactions. The formation of multimeric structures of MIA protein was also confirmed by Western blot analysis as shown in Figure 4B. These aggregates appear to be extraordinarily stable since they can even be observed after treatment with denaturing and reducing Laemmli buffer at 70°C.

Figure 4. HTFP assay investigation of MIA protein aggregation. (Click to enlarge)

Discussion

Several methods have been developed for the investigation of protein interactions. While surface plasmon resonance (SPR) (25) can also be used for small-molecule interactions with the help of antibodies, it is still costly because proprietary chips are generally required. Far-Western blotting is time consuming and not suitable for high-throughput applications (26). Furthermore, this method relies on the refolding of the protein to its native conformation on the membrane, which may not always be successful. Immuno-precipitation and pull-down experiments are also far more time-consuming than fluorescence-based investigations (27). Binding experiments using the 1-anilino-8-naphthalene sulfonate (ANS) probe (28), while compatible with multiwell plate–based assays, suffer from short excitation and emission wavelengths. Automated isothermal calorimetry measurements offer the advantage of label-free detection, but still require relatively large amounts of reagents.

Methods capable of handling high-throughput screening include various types of microarrays using enzymes, isotopes, or fluorescent labels. However, these techniques require special safety precautions, antibodies, and washing steps that may lead to cross-contamination and other artifacts.

FP detection is both high-throughput–capable and self-referenced, meaning that no washing steps are required. This is clearly an advantage over the traditional enzyme-linked immunosorbent assay (ELISA) platform. However, traditional homogeneous FP assays are limited by the molecular weight of the interaction partner to be investigated, due to the short lifetime of the required organic fluorophores. We extend this range by immobilizing a known interaction partner of the protein of interest, which is itself labeled with a long-lifetime luminescent transition metal chelate. The maximum acceptable molecular weight of the interaction partner obviously depends on the decay time of the label on the target. Given the decay time of Ru(bpy)₃, we estimate from the Perrin equation (29) that interactions with binding partners of up to 500 kDa should still be observable; however, that limit could be raised by using a transition metal with a longer decay time.

The results show that our HTFP assay allows the investigation of protein–small molecule and protein–protein interactions. In contrast to traditional homogeneous FP assays, interactions with both high-and low–molecular weight compounds can be investigated. As presented for the interaction of MIA protein with AR54, this FP assay should also be amenable for the screening of libraries of potential drug candidates. Additionally, our HTFP assay is suitable for the investigation of protein aggregation and compounds cleaving these aggregates. This tolerance of the HTFP assay for aggregation makes it unique and should allow for the investigation of proteins that show aggregation-related artifacts in other assays. However, one limitation of our assay is that it will be difficult to estimate aggregate sizes from the polarization values due to the long decay time.

Since our assay format is applicable to a variety of situations, it is conceivable that it might also be used for other analytical or diagnostic applications. For instance, the HTFP assay would be expected to enable the investigation of protein complexes, such as cell signaling molecules, transport proteins, or transcription factors. It could also be used for the identification of an initiator or regulator of polymerization reactions (i.e., for actin or tubulin subunits within the dynamic processes of the cytoskeleton). This assay may also serve to identify activators or co-activators for enzymatic reactions as well as for the design of immunoassays in the field of serology and diagnostics. Since the HTFP assay is based on a luminescent transition metal complex label, it benefits from all the associated advantages over organic fluorophores. While the inherent photostability is obviously convenient, the large Stokes shift increases the signal-to-noise ratio and allows a broader selection of suitable emission filters for the spectrometer. Furthermore, complex biological matrices in the samples are also tolerable because the autofluorescence of biological material has a very short lifetime and can thus be eliminated. Finally, the long lifetime of transition metal complex labels opens the possibility of time-gated measurements. This may be employed for multi-label experiments with different transition metal complexes of different lifetimes, allowing these labels to be resolved regardless of spectral overlap.