Novel Bioluminescence-triggered Multivalent Fluorescence Amplification-based (BtMFAb)-reporter
Recently, novel Fluorophore-NanoLuc reporters were designed allowing to eliminate significant auto-fluorescence associated with ectopic excitation of fluorescent reporters in vivo. Fluorophore-NanoLuc reporters, coined LumiFluors, do not require an external global excitation due to intramolecular energy transfer in the form of bioluminescence from NanoLuc® (Promega) luciferase to fluorescent proteins like enhanced Green Fluorescent Protein (eGFP) or LSSmOrange. The photon flux (ℎν) generated during NanoLuc substrate-triggered bioluminescence excites the fluorescent proteins locally, which generates the brightest fluorescent signal known to date1. To test the sensitivity of GpNLuc (eGFP-NanoLuc) LumiFluor in vivo, mice were injected with NanoLuc substrate and bioluminescent imaging (BLI) was used (IVIS® Spectrum machine [PerkinElmer], acquisition times <1 min). The signal-to-noise ratio obtained for 500 cells expressing GpNLuc was 2–3 orders of magnitude higher than signal generated in control mice demonstrating that far fewer cells can be detected easily and most importantly non-invasively1.
Here, we propose to use Fluorophore-NanoLuc reporters to image the protein-protein interactions in the mouse brain longitudinally. We will use GpNLuc as an example, however, the approach is generalizable and can be used with various fluorophores allowing multiplexing. It is known that the interaction of huntingtin (HTT), tau, amyloid-β, or α-synuclein protein molecules in abnormal conformation results in aggregates accumulation, so-called aggregated fibrils. Thus, to be able to detect such pathological interaction of huntingtin (HTT), tau, amyloid-β, or α-synuclein it is required to have corresponding probes capable of recognizing such interaction. Obvious candidates for such probes are single-chain variable fragments (scFv) containing the epitope-binding regions consisting of variable light (VL) and heavy (VH) chains, which can be successfully expressed in soluble form intracellularly. Thus, it is required to fuse the scFv probe N-terminally to the GpNLuc. The N-terminal covalent fusion of scFv with GpNLuc LumiFluor will represent an example of Bioluminescence-triggered Multivalent Fluorescence Amplification-based (BtMFAb)-reporter class (including other fluorophores distinct from eGFP) (Figure). In particular, novel BtMFAb-reporter will consist of two moieties: (i) the probe recognizing the pathological proteins interaction; and (ii) modified LumiFluor-type structure. The probe will be typically scFv, however, it can be substantially reduced in size. The modified LumiFluor-type structure, for example, can be represented by super-folder GFP (sfGFP) and NanoLuc.
The mechanism of BtMFAb-reporter action will be as follows. Both Nano-Glo® (Promega) and ViviRen™ (Promega) formulas containing furimazine, a substrate for NanoLuc, penetrate the blood-brain barrier2 and thus can be used for BtMFA reporter non-invasive longitudinal detection in live mouse brain by BLI. In the mouse brain, either extracellularly or intracellularly the NanoLuc component of BtMFAb-reporter will convert furimazine into furimamide and will generate the flux of photons, which will excite the sfGFP component triggering its fluorescence (Figure). This mechanistic concept is not novel and was reported by Schaub and colleagues (see above). The novelty of the current Insight is represented by the scFv moiety of the BtMFAb-reporters. The scFv moiety of BtMFAb-reporter will make possible the usage of fluorophores’ unique property.
The principle of Multivalent Fluorescence Amplification (MFA)
It has been demonstrated that local increase in the fluorophore protein concentration triggers fluorescent signal amplification3–5. This phenomenon has been very recently coined multivalent fluorescence amplification (MFA)5. Localization and copy number imaging of DNA sequences in the genome3, imaging translation4 and transcription5 of single molecules in real-time in live cells were performed using MFA. For example, to image the translation in real-time, multiple identical epitopes in a protein being synthesized were bound by scFv-sfGFP fusion proteins expressed in the same cell. In particular, 24 copies of the above-mentioned epitope tag present in a single molecule of a synthesized protein and bound by scFv-sfGFP molecules yielded 18-fold brighter signal, than a single molecule of scFv-sfGFP4. These are pioneering and very important methods highlighting the power of MFA—almost linear signal amplification. The elegant methods described in the literature and applying the MFA, however, suffer from its artificial nature. In particular, the proteins of interest should be modified, which is inconvenient from an experimental point of view and might lead to perturbation in modified protein function and as a result render studies non-physiological as well as clinically irrelevant.
BtMFAb-PEST Reporter: the mechanism of action
In the current Insight, the shortcomings mentioned above will be overcome. The interaction of abnormally-folded proteins like tau, amyloid-β, or α-synuclein leads to their aggregation. As a result, aggregated HTT-, tau-, amyloid-β-, or α-synuclein-fibrils composed of abnormally folded proteins are formed. There are specific scFvs recognizing such aggregated fibrils. At the same time, the scFvs recognizing such pathological interactions do not recognize the normal conformation of respective functional proteins. Thus, it is not required to modify the above-mentioned proteins to study their interaction. The scFv moieties of BtMFAb-reporters specific for aggregated HTT-, tau-, amyloid-β-, or α-synuclein-fibrils will promote the local reporter-molecules accumulation, which will lead to MFA phenomenon. The BtMFAb-reporters should be brighter than LumiFluor-reporters due to the MFA phenomenon. Moreover, using BtMFAb-reporters will allow quantitation of the extent of pathologic interaction, as MFA is characterized by linear signal amplification. Thus, the longer the aggregated pathological fibril—the higher the bioluminescent/fluorescent signal (Figure).
To make the system more sensitive, it is required to remove the BtMFAb-reporters that are not engaged in the interaction with aggregated pathological fibrils. For this purpose, the BtMFAb-reporters should be C-terminally fused with the PEST region of mouse ornithine decarboxylase (mODC)—the proteasome-targeting peptide sequence rich in proline (P), glutamic acid (E), serine (S), and threonine (T). Such BtMFAb-reporter with C-terminal mODC PEST region will be termed BtMFAb-PEST-reporter. It has been demonstrated that mODC C-terminal PEST-region fusion to scFvC4, the scFv specifically recognizing normal HTT, led to ~90% decrease in the HTT levels due to proteasome targeting6. Butler and Messer also demonstrated that fusion of PEST-peptide to anti-fibrillar scFv did not reduce the aggregated HTT levels due to the fact that aggregated HTT is inefficiently degraded by the proteasome6. Even if ~10% of unreacted BtMFAb-PEST-reporters will be present in the system, the much brighter MFA fibril-specific signal can be detected using shorter acquisition times.
The only modification to the in vivo (living) system will be the non-invasive delivery of the BtMFAb-reporter into the mouse brain target cells. The BtMFAb-PEST-reporters can be genetically encoded and delivered using modified Adeno-Associated Virus (AAV) vectors. Our NextGenRnD Gene therapy targeting platform solution should allow significant AAV vector dose reduction by improving its transduction specificity. Moreover, this solution contains a strategy for significant minimization of scFv moiety size.