CHIMERIC PROTEIN SWITCH FOR THE OPTOGENETIC CONTROL OF AMYLOIDOGENESIS

Abstract

The present invention provides an optogenetic chimeric fusion polypeptide comprising an optimized amino acid sequence of the plant phototropin domain LOV2 fused to an amino acid sequence of the bacterial amyloidogenic effector RepA-WH1. Optimized LOV2 enables navigation through the folding landscape of RepA-WH1 from solubility to its aggregation as oligomers or amyloid fibres. Thus, this polypeptide assembles as hydrogels and amyloid fibres in the darkness, while under blue light illumination forms oligomeric particles that are proteotoxic for cells, preferably bacteria. This polypeptide is therefore proposed for inducing the formation of cytotoxic amyloid oligomers in cells that are targeted for killing.

Claims

1. A fusion polypeptide comprising the mutated amino acid sequence of the LOV2 domain shown in SEQ ID NO: 4 fused by its C-terminal end to the N-terminal end of the RepA-WH1 protein shown in SEQ ID NO: 5, wherein said fusion polypeptide comprises the amino acid sequence of SEQ ID NO: 1.

2. The fusion polypeptide according to claim 1, wherein the amino acid sequence of SEQ ID NO: 1 is fused, in its C-terminal end, to the amino acid sequence of the mCherry fluorescent protein probe.

3. The fusion polypeptide according to claim 2, wherein the amino acid sequence of the mCherry fluorescent protein probe is SEQ ID NO: 2.

4. The fusion polypeptide according to claim 3, which comprises the amino acid sequence SEQ ID NO: 3.

5. The fusion polypeptide according to claim 4, which consists of the amino acid sequence SEQ ID NO: 3.

6. A nucleic acid sequence encoding the fusion polypeptide according to any one of claims 1 to 5.

7. A genetic construct comprising the nucleic acid sequence according to claim 6, preferably wherein the genetic construct is a phage.

8. A cell comprising the fusion polypeptide according to any one of claims 1 to 5, the nucleic acid sequence according to claim 6 or the genetic construct according to claim 7.

9. The cell according to claim 8 which is a prokaryotic cell, preferably a bacterial cell.

10. Use of the cell according to claim 8 or 9 for the production of the fusion polypeptide according to any one of claims 1 to 5.

11. In vitro use of the fusion polypeptide according to any one of claims 1 to 5, the nucleic acid sequence according to claim 6 or the genetic construct according to claim 7 for inducing the formation of cytotoxic amyloid oligomers in a cell, preferably in a bacterial cell.

12. In vitro use according to claim 11, which comprises exposing the cell to blue light.

13. The fusion polypeptide according to any one of claims 1 to 5, the nucleic acid sequence according to claim 6, the genetic construct according to claim 7 or the cell according to claim 8 or 9 for use as a medicament.

14. The fusion polypeptide according to any one of claims 1 to 5, the nucleic acid sequence according to claim 6, the genetic construct according to claim 7 or the cell according to claim 8 or 9 for use in the treatment or prevention of microbial infections, preferably bacterial infections.

15. An in vitro method for inducing the formation of cytotoxic amyloid oligomers in a cell, preferably in a bacterial cell, which comprises: a. expressing the fusion polypeptide according to any one of claims 1 to 5 in the cell, and b. exposing the cell of step (a) to blue light.

Description

DESCRIPTION OF THE DRAWINGS

[0085] FIG. 1. Assessing the solubility of distinct helical phases for the Jα-α1 linker in LOV2-RepA-WH1 chimeras. (a) Schematic representation of the arrangement of the three chimeras assayed. The linker helix increases its length by a helical turn between LOV540 (closest position related to WH1) and LOV543 (farthest to WH1), displaying both domains on opposite faces of the helix, whereas in LOV542 both are found on the same face and intermediate position. (b) Fractionation of bacterial cells (C) expressing the chimeras into the soluble (S) and insoluble (P) fractions (top panel), plus their detection by Western blotting (anti-His tag antibody; bottom). Only LOV543-WH1 generates a major soluble fraction, with no major differences when cultures were carried out in the darkness or under blue light illumination. (c) These results are compatible with steric hindrance due to close apposition of the domains by a short linker (LOV540-WH1) and/or their mutual interference for folding (LOV542-WH1), whereas only LOV543-WH1 has the right distribution of both domains to fold independently.

[0086] FIG. 2. Limited proteolysis maps the Jα-α1 helical linker as the major light-altered structural element in LOV543-WH1. (a) SDS-PAGE analysis of LOV543-WH1 digested, in the darkness or under blue light illumination, by three alternative proteases. A major cleavage site for chymotrypsin and V8 protease is made accessible in the illuminated state, generating two fragments with the sizes expected for LOV2 (≈17 kDa) and WH1 (≈15 kDa). (b) Sequence of the LOV543-WH1 fusion protein. Residue numbers are the originals from the AsLOV phototropin and RepA. N-terminal sequencing of the proteolytic fragments identified unambiguously the cleavage sites (arrows) for chymotrypsin (Ch: I.sub.532KK and E.sub.409RI) and V8 protease (V8: G.sub.528VM). The N-terminal Met residue appears to have been removed in vivo (g.sub.−23ss; tag peptide in lower case letters). Box outlines the linker Jα-α1 helix. The three amino acid substitutions included in the LOV543m3-WH1 (SEQ ID NO: 1) mutant are also indicated.

[0087] FIG. 3. In vitro cross-seeding of the assembly of the hyper-amyloidogenic protein RepA-WH1(A31V) (light grey) by sub-stoichiometric amounts (1:100) of the LOV543 wt/m3-WH1 optogenetic switches. (a) Schematic overview of the seeding experiment. (b) Electron micrographs of the aggregates generated during the incubation of RepA-WH1(A31V) as substrate and the indicated optogenetic seeds, either in the darkness or under blue light illumination. Control samples, un-seeded or adding as seeds RepA-WH1(A31V)-mCherry aggregates purified from E. coli (S), are also displayed. LOV543 wt-WH1 templates on RepA-WH1(A31V) the assembly of large fibres (F), irrespective of being illuminated or not. LOV543m3-WH1 differentially seeds the growth of short fibres (darkness) or an ensemble of needle-shaped spikes (N) and spherical/drop-like oligomers (D) (blue light). Magnification: 20,000× (insets: 50,000×). (c) Analysis of the fluorescence emitted by ThS (left) and optical birefringence of Congo Red (CR, right) when incubated with samples in (a). Both amyloidotropic molecules bind to RepA-WH1(A31V) aggregates when templated by control seeds or by LOV543 wt-WH1 (whether it was incubated in the darkness or under blue light), thus they are amyloids. Particles assembled by LOV543m3-WH1, especially when incubated under blue light, bind weaker to ThS and CR. (d) Dot-blot analyses. Serial dilutions of samples assembled in the darkness (D) or under blue light (B) are displayed. B3h7, an antibody specific for the conformation of amyloidogenic RepA-WH1 oligomers, shows that these are more abundant when seeded with LOV543m3-WH1 (left). A replica incubated with α-WH1, a polyclonal antibody against RepA-WH1, is included as a control (right). Overall, LOV543m3-WH1 generates an operational switch of amyloidogenesis, discriminating between the assembly of mature amyloid fibres (darkness) and amyloidogenic oligomeric particles (blue light).

[0088] FIG. 4. Optogenetic regulation of a phase transition (liquid-hydrogel) in LOV543m3-WH1-mCherry. (a) The hydrogel formed in the darkness, when examined by TEM, is made of a mixture of tightly packed fibres (F) and large spherical, drop-like oligomers (D), whereas the liquid formed under blue light contains discrete annular oligomers (o). Magnification: 20,000× (insets: 50,000×). (b) Epifluorescence visualization of a time course gelation experiment (top row). Samples were assembled under blue light and, after 1 h, transferred to the darkness for up to 24 h. Images are the superposition of DIC and TRICT (mCherry) channels. The boundary separating the liquid (L) and the expanding hydrogel (H) phases is clearly seen. Bottom row: zoom (5×) of the sectors boxed above.

[0089] FIG. 5. Expression of LOV543m3-WH1-mCherry in E. coli results in the formation of large intracellular amyloid particles in the darkness, and smaller cytotoxic aggregates when grown under blue light. (a) DIC and epifluorescence microscopies, showing the intrinsic mCherry fluorescence and the extrinsic ThS staining. Arrows point to intracellular aggregates. Experiment was independently repeated three times. 454 cells grown in the darkness and 1,534 under blue light were counted in total. (b) Fractionation of cell lysates of bacteria grown in (a). After centrifugation, proteins were analysed by SDS-PAGE (left) and then Western-blotted with an anti-mCherry antibody (middle). (−), non-induced cells control. P: pellets, S: supernatants. D: darkness, B: Blue light. Quantitation of the three culture replicas (one way ANOVA with Tukey's comparison test to 95% confidence; ***: p<0.005, **: p<0.05) indicated that blue light illumination during protein expression increases the solubility of the triple chimera. (c) Serial dilutions of cultures on agar plates show no significant inhibition of bacterial growth by blue light itself. Expression of LOV543m3-WH1-mCherry in the darkness results in much smaller colony sizes and in ≥10.sup.2-fold inhibition in the case of blue light illumination, an indication for cytotoxicity of the oligomers generated by the optogenetic switch when excited with blue light.

[0090] FIG. 6. Conformational pathways regulated by blue light in the LOV2-WH1 optogenetic device. LOV2 allosterically selects between the assembly of the prion-like protein RepA-WH1 into large, non-cytotoxic hydrogels made of amyloid fibrils (right; darkness) and cytotoxic oligomers (left; blue light). In both scenarios, the Jα-α1 linker would act on the WH1 domain differentially, either as a rigid helical lever in amyloidogenic unfolding (darkness-promoted) or through the local destabilization of α1 (blue light-elicited).

[0091] FIG. 7. Site-directed mutagenesis to generate LOV543m3-WH1. DNA sequencing chromatographic profiles for the wt and the m3 versions of the mutagenized stretch in LOV543-WH1. SEQ ID NO: 12 to SEQ ID NO: 15.

[0092] FIG. 8. Three-dimensional model of LOV543m3-WH1 in its dark state. (a) Overview of a high score (0.86) LOV543m3-WH1.sub.11 model (top), as generated by the ROBETTA fully automated protein structure prediction server (Kim, D. E., Chivian, D. & Baker, D., 2004, Nucleic Acids Res., 32, W526-W531) on the structures of RepA-WH1 (1 HKQ.pdb) and AsLOV2 in its dark state (2V0U.pdb). ROBETTA uses the ROSETTA fragment insertion method. Models were generated from structures detected by PSI-BLAST or HHSEARCH and aligned by HHSEARCH and SPARKS. Loops were built from fragments and optimized to fit the template structures. Bottom: Expanded view of the region of LOV543m3-WH1 comprising the chimeric Jα-α1 helix and its backing LOV2 β-sheet. Mutant side chains are displayed as sticks. (b) A Ramachandran plot, generated by PROCHECK, showing the stereochemical compatibility of the generated model. It is noteworthy that the first N-terminal residue modelled is Leu404 in LOV2, to which ROBETTA assigns the first position.

[0093] FIG. 9. Biophysical characterization of LOV543 wt-WH1 (left column) and LOV543m3-WH1 (right). (a) Absorption spectra showing a time course of the return of the blue light-excited chimeras to the ground (dark) state. Spectra were acquired with 15 s intervals after switching the blue LEDs off and then plotted overlaid. Inset: kinetics of the return to the ground state, measured as the value of the absorption band at 447 nm for each spectrum. Curves display double exponential function fittings to the data points. τ½: half-life of the lit state. (b) Gel filtration elution profiles of the purified chimeras, run in the darkness or under continuous blue light illumination. The major species correspond to protein dimers, in spite of having been illuminated or not, as it was confirmed, for LOV543m3-WH1, through sedimentation velocity analyses (inset). Arrows point to minor oligomeric species. MW standards: aprotinin (APR), RNase A (RNA), ovoalbumin (OVO) and alcohol dehydrogenase (ADH). (c) CD spectra (left) and thermal denaturation profiles (right) of the purified chimeras.

[0094] FIG. 10. Biophysical characterization of the LOV2m3-WH1-mCherry chimera. (a) Absorption spectra displaying the return of the blue light-excited chimera to the ground (dark) state. Spectra were acquired with 15 s intervals after switching the blue LEDs off. Inset: kinetics of the recovery of the dark state, measured as the value of the absorption band at 447 nm for each spectrum. Curve shows a double exponential fitting for the data points. τ½: half-life of the lit state. (b) Gel filtration elution profiles of the purified chimeras, run in the darkness or under continuous blue light illumination. The major species corresponds to a protein dimer, as confirmed by sedimentation velocity profiles (inset; single asterisk). Arrows and double asterisks point to minor oligomeric aggregated species. (c) CD thermal denaturation profile. Inset: CD spectrum.

EXAMPLES

Example 1. Results

[0095] LOV2-WH1 chimeras: design of a suitable helical linker. The key determinant of RepA-WH1 (in short, WH1) stability is the formation of a helical latch by locking the C-terminal helix α5 in between the V-shaped N-terminal helices α1-α2. The possibility to manipulate the stability of WH1 by straining this domain at its N-terminus was explored constructing a chimeric continuous α-helix between the C-terminal Jα helix in the Avena sativa phototropin domain LOV2 and α1 in WH1. It is well established that upon absorption of blue light photons (λmax 447 nm) by the FMN chromophore in LOV2, Jα unfolds and detaches from the core of the domain, thus unconstraining the conformation of any sequence to which this helix had been intentionally linked. Three different helical phases in the Jα-α1 linker, and thus three distinct relative geometrical arrangements of the LOV2 and WH1, were constructed by PCR. The RepA-WH1 wild-type domain was used in the chimeras due to its higher solubility compared with some mutant variants (e.g., A31V, which is intrinsically hyper-amyloidogenic). The constructs (FIG. 1a) displayed both domains either at opposites sides (LOV540-WH1 and LOV543-WH1) or at the same side (LOV542-WH1) of the Jα-α1 helix. The difference between the former two chimeras is the length of the linker, being about a helical turn longer (4 amino acid residues) in LOV543-WH1 than in LOV540-WH1. The three chimeras were expressed in Escherichia coli with N-terminal His.sub.10 tags, either in the darkness or under illumination with blue light (using a custom-built LED device) and, upon cell lysis, their sedimentation behaviour was tested (FIG. 1b). It turned out that only the LOV543-WH1 chimera differentially increased its solubility upon blue light illumination. The unresponsiveness to light of the other two constructs may well reflect either blocking of the switch by a too tight package of the domains, imposed by the reduced length of the linker (in LOV540-WH1), or steric hindrance by the N-terminal LOV2 domain to proper folding of the C-terminal WH1 (most likely in LOV542-WH1) (FIG. 1c). All subsequent experiments were thus performed with the LOV543-WH1 chimera.

[0096] The LED device used was made of 27×8 blue AlGaInP LED bulbs (5 mm ø, 20 deg cone, 3 V, λ.sub.max=468 nm, 8 cd each; TheLEDLight company), distributed in six independently switchable sectors and with a relay wheel that allows regulation of the light emission intensity between 1,070 (min.) and 30,000 (max) Lux. These intensities were used, respectively, for the in vivo and the in vitro experimental settings.

[0097] Addressing light-responsiveness of LOV2-WH1 through limited proteolysis. Proteolysis is a useful test for the accessibility to the solvent of target peptide sequences, as well as for the stability of folded protein domains. Three of the specific proteases that were used to probe the full length RepA protein, thus uncovering the existence of its two tandem WH domains, were this time assayed on LOV543-WH1, either in the dark or under blue light illumination (FIG. 2a). Peptides were separated by SDS-PAGE, revealing that blue light enhanced the cleavage by chymotrypsin and V8 protease, which yielded two main protein bands whose sizes roughly corresponded to the expected for the individual LOV2 and WH1 domains. This suggested increased accessibility of the linker due to unfolding of the Jα-α1 helix upon blue light absorption. Then N-terminal peptide sequencing was performed on the chymotrypsin and V8 digestions that had been incubated under blue light, allowing for the identification of the major cleavage sites for each protease in the LOV543-WH1 sequence (FIG. 2b). Specific cleavage sites were found precisely at the joint between Jα-α1, confirming that the helical linker became unfolded, as designed, upon illumination of the LOV543-WH1 chimera with blue light.

[0098] Improving the LOV2-WH1 switch by mutagenesis. A major concern in the design of any synthetic switch through protein fusion is how this affects the dynamic range of the device, i.e., the net ratio between the response of a chimera to the ON and the OFF stimuli, which in optogenetics ultimately depends on the balance between the fraction of molecules that remain in the OFF (pseudo-dark) state upon illumination and the fraction of molecules that stay in the ON (pseudo-lit) state in the darkness. Three mutations were introduced in LOV543 wt-WH1, to generate LOV543m3-WH1 (FIG. 7): I532A, G528A, and L531E, which improve the packing of the N-terminus of Jα with the backing β-sheet in LOV2, leading to a more tightly bound Jα helix in the dark state. The structure of the LOV543m3-WH1 chimera was modelled on the PDB coordinates of its two component domains using the Rosetta de novo structure prediction method. FIG. 8a displays a high-score in silico model supporting the structural feasibility of the three mutations within the fully folded LOV2 and WH1 frames, which showed no stereochemical violations (FIG. 8b).

[0099] The LOV543 wt-WH1 and LOV543m3-WH1 chimeras were expressed in E. coli, purified and characterized through biophysical approaches (FIG. 9). Measurements of the return to the dark state after saturating blue-light stimulation, by following the evolution of the absorption spectra of the FMN prosthetic group (band at 447 nm), indicated a three-fold increase in the half-life of the excited state for the m3 mutant (16.7 s) compared with its parental wt protein (5.6 s) (FIG. 9a). No significant differences between the dark and lit sates, or between the wt and m3 chimeras, were observed regarding the association state of the proteins, as indicated by gel filtration analyses independently performed in the darkness and under blue light illumination (FIG. 9b). The higher solubility of the LOV543m3-WH1 chimera allowed confirmation of its dimeric state by sedimentation velocity analysis performed under both illumination conditions (FIG. 9b, right, inset). This observation suggests that, in the chimeras, dimerization continues to be dictated by the stable antiparallel β-sheet interface in RepA-WH1. Strikingly, LOV543 wt-WH1 aggregated massively on any attempt to either freeze or concentrate it further than 6 μM, while LOV543 m3-WH1 withstood both storage in a frozen state and at least 10-fold higher concentrations. Circular dichroism spectra of both chimeras in the darkness showed the typical α-helical profile, with a more pronounced band at 208 nm for the m3 chimera (FIG. 9c, left). This is a signature for increased α-helical content, as expected from the stabilization of Jα in the dark state by the three mutations engineered in LOV543m3-WH1. Thermal denaturation analyses (FIG. 9c, left) indicated that both chimeras were stable, showing a single transition between two states (folded and unfolded) with a Tm (50%) value of 57° C. (FIG. 9c, right).

[0100] Light modulates the capacity of LOV2-WH1 to cross-seed RepA-WH1 (A31 V) amyloidogenesis. Seeding, i.e., the ability of a pre-formed amyloid aggregate to template and nucleate amyloid growth from soluble molecules of the same (or a closely related) protein, is a hallmark of amyloidogenesis. To test the capacity of LOV543-WH1 to act as a light-controlled switch in amyloid nucleation, substoichometric amounts of the purified chimeras (either the wt or m3 variant) were supplied to an excess of soluble RepA-WH1(A31V) (FIG. 3a). This is a hyper-amyloidogenic RepA-WH1 variant that efficiently assembles as fibres under standard in vitro conditions provided that a nucleation agent, such as purified RepA-WH1(A31V) aggregates preformed in vivo, are supplied as seeds. The formation of RepA-WH1(A31V) fibres was thus explored either under continuous blue light illumination or in the darkness. Fibrillation was monitored by transmission electron microscopy (TEM) (FIG. 3b), thioflavin-S (ThS) fluorescence emission, Congo red (CR) birefringence under polarized light (FIG. 3c) and probing with B3h7, a monoclonal antibody specific for oligomers of RepA-WH1 with an amyloidogenic conformation, but showing reduced affinity for the fibres (FIG. 3d). The results of these assays indicated that just spare amorphous aggregates were found in the absence of any supplied seed, while fibres were the product of nucleation by the intracellular RepA-WH1(A31V) aggregates. Interestingly, similar fibrilar aggregates were generated upon the addition of LOV543-WH1 in its wt version. However, no differential fibrillation response to the darkness/blue light regimes was observed for this chimera. On the contrary, the LOV543-WH1m3 variant nucleated on RepA-WH1(A31V) the formation of defined, short fibrils in the darkness, but of drop-like oligomeric particles and needles under blue light (FIG. 3b). These oligomeric species showed much reduced amyloid character, when compared to the mature fibres, according to both ThS and CR staining (FIG. 3c). The higher affinity of the B3h7 antibody in dot-blot assays for the reactions seeded with LOV543m3-WH1 supports their enrichment in amyloidogenic RepA-WH1 oligomers (FIG. 3d). Because only the LOV543m3-WH1 chimera was able to modulate the fibrilar/oligomeric assembly of RepA-WH1 in vitro under switch conditions (i.e., darkness/blue light), this protein fusion was selected to explore the capability to build by itself complex supramolecular assemblies, both in vitro and in vivo.

[0101] LOV543m3-WH1-mCherry enables optogenetic control on phase transitions. To visually follow the effect of darkness/blue light illumination on LOV543m3-WH1 aggregation, both in vitro and in vivo, the red fluorescent reporter protein mCherry was fused to the C-terminus of the WH1 domain. The LOV543m3-WH1-mCherry protein was then purified and characterized through several biophysical approaches (FIG. 10). The half-life of the excited state, measured as the recovery of the absorption band at 447 nm after transferring pre-illuminated samples to the darkness, was improved by 2.3-fold (to 38.3 s) compared to the LOV543m3-WH1 fusion, with the mCherry band (at 586 nm) dominating the spectra and showing no temporal variation (FIG. 10a). As it was shown for its parental double chimera (FIG. 9b, right), according to gel filtration and sedimentation velocity analyses, LOV543m3-WH1-mCherry remained as a dimer either in the darkness or under blue light (FIG. 10b). The CD spectrum (FIG. 10c) was much dominated by the β-sheet structure of mCherry, and the triple chimera showed enhanced thermal stability (85.5° C.) compared to the double chimera (57° C.; FIG. 9c, right). However, as for the LOV543 wt/m3-WH1 fusions, it is noteworthy that these assays were carried out under ideal (diluted) buffer conditions, not attempting to mimic the crowded environment of the bacterial cytoplasm.

[0102] The assembly potential of LOV543m3-WH1-mCherry either in the darkness or under blue light illumination was then tested in vitro at a very high protein concentration (0.25 mM), in a low salt buffer and in the presence of polyethylene glycol (PEG) 4000, a crowding agent that enhances RepA-WH1 fibrillation (FIG. 4). In the darkness, the triple chimera decanted to the bottom of the test tubes with the visual appearance of a hydrogel, which clearly separated from the buffer supernatant. TEM revealed the coexistence within the hydrogel of spherical particles and tightly packed fibril bundles (FIG. 4a). On the contrary, when samples were illuminated with blue light, they stayed as a translucent red solution, in which just ring-like oligomers were evident by TEM (FIG. 4a). Experiments carried out to explore such phase transitions under an epifluorescence microscope revealed that samples remained liquid as long as they were illuminated with blue light, with some de-mixing into drops containing LOV543m3-WH1-mCherry. However, when turned into the darkness these drops readily fused and the protein solution evolved within hours into a thick hydrogel with the appearance of a sponge (FIG. 4b). As expected for amyloidogenesis, darkness-promoted gelation of LOV543m3-WH1-mCherry was not reversible.

[0103] LOV543m3-WH1-mCherry is an optogenetic switch for bacterial proliferation. Expression in the darkness of the LOV543m3-WH1(WT)-mCherry triple chimera resulted in filamentation of bacterial cells (FIG. 5a), a phenotype characteristic of RepA-WH1(A31V)-mCherry expression but not of RepA-WH1(WT)-mCherry, suggesting that LOV2 indeed destabilizes the fold of the native RepA-WH1 turning it into amyloidogenic. Although quantitation is difficult due to clumping of a substantial fraction of the cells, 35% of them exhibited intense ThS-stainable protein aggregates, an indication for their amyloid nature. On the contrary, if expression of LOV543m3-WH1-mCherry was carried out under blue light illumination, just 5% of the cells showed ThS fluorescence emission, and this was of much lower intensity than when bacteria were cultured in the darkness. Biochemical analysis of the solubility of the triple chimera upon cell lysis and subsequent sedimentation plus Western-blotting (FIG. 5b) confirmed that the fraction of LOV543m3-WH1-mCherry in the supernatant increased in cultures illuminated with blue light (59%) compared with those grown in the darkness (23%).

[0104] To survey a possible effect of the optogenetically-regulated expression of LOV543m3-WH1-mCherry on bacterial proliferation, serial dilutions of exponential phase cultures that had been grown in the darkness were plated on LB-agar including (or not) the inducer IPTG (FIG. 5c). While in the absence of expression of the triple chimera no difference was appreciated between drops incubated in the darkness or under blue light illumination, LOV543m3-WH1-mCherry expression reduced the growth of bacteria under blue light by a least two logs when compared to dark-state incubation conditions. Since under blue light illumination both LOV543m3-WH1 (FIG. 3) and LOV543m3-WH1-mCherry (FIG. 4) form amyloidogenic oligomers in vitro, and RepA-WH1(A31V)-mCherry oligomers are the most cytotoxic molecular species of in vivo, oligomerization of the triple chimera likely is the basis for the observed decrease in E. coli viability upon blue light illumination.

[0105] In summary, the LOV2-WH1 chimeras were first optimized by modulating the phase and length of the linker Jα-α1 helix (LOV543-WH1.sub.11; FIG. 1), plus the inclusion of mutations to stabilize the dark state conformation of Jα (FIGS. 7-9). In vitro studies (FIG. 3) suggest that, in the darkness, the stiff Jα-α1 chimeric helix connecting LOV2 and WH1 (FIG. 2) would act as a lever that, by unleashing the three-helix bundle (α1-α2-α5) that locks the fold of WH1, would enable the protein to template, on a hyper-amyloidogenic RepA-WH1 variant, the assembly of amyloid fibrils (FIG. 6a, right). On the contrary, under blue light illumination the partial unfolding of the Jα-α1 linker (FIG. 2) would compromise the α1-α2-α5 latch in a milder way, resulting in a conformational ensemble that templates the assembly of distinct oligomers (FIG. 3), either needle-shaped or drop-like, that show a weaker amyloid nature (FIG. 6a, left). Such optogenetic destabilization of the N-terminus of RepA-WH1 is a new way to trigger amyloidogenesis in this protein.

[0106] Therefore, light-modulated conformational remodelling of the LOV543m3-WH1 chimeras is a reliable approach to gain control on RepA-WH1 amyloidogenesis and toxicity, either in vitro or in vivo. Here it is shown that an optogenetic switch achieving control over protein amyloidosis, both in vitro and in vivo, can be built by combining in a chimera a light-responsive LOV2 domain with a versatile amyloidogenic bacterial module, RepA-WH1.

[0107] Amyloid opto(epi)genetics remained yet unexplored. The LOV2-WH1 chimeras presented herein enrich and expand the catalogue of available optogenetic tools with a novel way to guide the conformational landscape of proteins towards amyloidogenesis. Their potential applications may include controlling the assembly of amyloid nanoscaffolds to engage enzymes in sequential reaction steps; building transcriptional switches for synthetic gene expression circuits and light-controlled plasmid replication cassettes; the selective elimination of particular bacteria within a consortium, once they have fulfilled their task in a bioprocess; or the development of a completely new kind of antimicrobials based on triggering amyloidosis by, e.g., bacteriophages encoding light-switchable cytotoxic amyloids.

Example 2. Methods

Construction of the LOV2-WH1 Chimeras

[0108] The AsLOV2 gene (SEQ ID NO: 6) was custom-synthesized at ATG:biosynthetics (Merzhausen, Germany), with its codon composition optimized to the usage in E. coli (SEQ ID NO: 7), and delivered as a pUC18 derivative. The template source of repA-WH1 was pWH1(WT). Both genes were independently amplified by PCR, using Pfu DNA polymerase, in such a way that the primers at 3′ end of LOV2 and at the 5′end of repA-WH1 hybridize in the next step through their 5′ ends. Three alternative pairs of these linker primers were designed to generate three distinct transitions in the fusion between both domains: LOV.sub.540-WH1.sub.12, LOV.sub.542-WH1.sub.12 and LOV.sub.543-WH1.sub.11. A second PCR round on an equimolar mixture of both amplicons, and with the primers annealing at the 5′ end of LOV2 and the 3′ end of repA-WH1, yielded the three distinctly phased chimeras. The amplified chimeric fragments were then cloned into pRG-Ptac-His.sub.10-ORC4, by replacing the resident ORC4 gene through restriction with SacII and HindIII plus ligation (T4 DNA ligase).

[0109] For the site-directed mutagenesis to generate LOV543m3-WH1 the following oligonucleotides were used: G528A_I532A-F (SEQ ID NO: 8): 5′ GTGATGCGGCGGAACGTGAAGCCGTGATGCTGGCTAAAAAAACCGCAGAAAA CATTGAT and G528A_I532 A-R (SEQ ID NO: 9): 5′ ATCAATGTTTTCTGCGGTTTTTTTAGCCAGCATCACGGCTTCACGTTCCGCCG C ATCAC, to build the G528A and I532A double mutant. On this double mutant the following primers were used to build the triple mutant (including the L531E mutation): G528A_L531E_I532 A-F (SEQ ID NO: 10): 5′ CGGAACGTGAAGCCGTGATGGAGGCTAAAAAAACCGCAGAAAACA and G528A_L531E_I532A-R (SEQ ID NO: 11): 5′ TGTTTTCTGCGGTTTTTTTAGCCTCCATCACGGCTTCACGTTCCG.

[0110] The LOV2m3-WH1-mCherry chimera was built in an analogous way, but using pRG-Ptac-His.sub.10-LOV543m3-WH1 (see above) and mCherry (from pRG-Ptac-His.sub.6-mCherry) as the templates for the two PCR amplifications at the initial step. The product of the second PCR amplification round, once digested with SpeI and HindIII, was cloned into pRK2-Ptac-His.sub.10+/ac/.sup.q, a derivative of the low copy-number vector pSEVA121. All constructs were verified through DNA sequencing (Secugen, Madrid).

Protein Expression and Purification

[0111] The RepA-WH1(A31V) protein used in the fibrillation studies was purified as described (Giraldo, R., 2007, Proc. Natl. Acad. Sci. USA, 104, 17388-17393). The LOV2-WH1 chimeras (H.sub.10-LOV543 wt/m3-WH1 and H.sub.10-LOV543m3-WH1-mCherry) were expressed in the E. coli strain BL21, in the presence of a helper plasmid providing T7 lysozyme to facilitate cell lysis. 0.75 L of Terrific Broth medium supplemented with ampicillin (Ap) to 100 μg.Math.mL.sup.−1 was inoculated with colonies from overnight LB agar plates with Ap.sub.100 and chloramphenicol (Cm) to 30 μg.Math.mL.sup.−1 and grown at 37° C. to an OD.sub.600 nm≈0.8. Then, IPTG was supplied to 1.0 mM and the flasks covered with aluminium foil. Expression proceeded for 5 h at room temperature (RT). Cells were harvested, washed with cold 0.9 NaCl and resuspended in 15 mL of lysis buffer (0.5 M NaCl, 0.05 M imidazole pH 8.0, 1% Brij-58, 10% glycerol plus 1 pill of EDTA-free Roche protease inhibitors). Cell suspension was frozen at −70° C.

[0112] Cell lysis was enabled by thawing the cell suspension to RT and a clarified lysate was obtained by ultracentrifugation at 62,000×g for 1 h at 4° C. Supernatant was distributed in two aliquots and each one was independently loaded into an ABT Ni-affinity 5 mL cartridge wrapped in aluminium foil and coupled to an ÄKTA basic 10 FPLC (GE Healthcare). After an extensive wash with column buffer A (0.5 M NaCl, 0.05 M imidazole pH 7.8, 10% glycerol), a 25 mL linear gradient was run between this buffer and column buffer B (0.5 M NaCl, 0.75 M imidazole pH 7.8, 10% glycerol). Peak fractions were pooled and stored at 4° C. Further purification plus buffer exchange, to eliminate imidazole which inhibits the transition of LOV2 to the lit state, was achieved by size-exclusion chromatography (SEC) in a Superdex HR-200 column (GE Healthcare) equilibrated and run at 0.4 mL.Math.min.sup.−1 flow in SEC buffer (0.05 M Na.sub.2SO.sub.4, 0.010 M Hepes.NaOH pH 7.6, 0.1 mM EDTA). Peak elution profiles were monitored at A.sub.280, A.sub.447 and (for the chimera including mCherry) A.sub.590 nm.

[0113] The concentration of the purified proteins was determined by absorption at 280 (RepA-WH; ε=11,548 M.sup.−1.Math.cm.sup.−1), 447 (H.sub.10-LOV543 wt/m3-WH1; ε=13,800 M.sup.−1.Math.cm.sup.−1) or 590 (H.sub.10-LOV543m3-WH1-mCherry; ε=70,700 M.sup.−1.Math.cm.sup.−1) nm. Protein chimeras were stored at 4° C. in the darkness for up to a week.

Protein Solubility Assays

[0114] Solubility of the chimeras, expressed in the E. coli K-12 reduced genome strain MDS42, was assayed in whole cell lysates from 15 mL of cultures grown at 37° C. in LB plus Ap.sub.100. When bacterial cultures reached OD.sub.600 nm=0.2, IPTG was added to 0.5 mM and they were split into two aliquots, to be grown either in the darkness or under blue light illumination (1,070 Lux). After 4 h of induction cells were harvested and resuspended in 0.2× lysis buffer (see above), at a ratio of 0.33 mL per each unit of optical density (1.5×10.sup.9 cells). EDTA (to 1 mM) and lysozyme (to 1 ρ.Math.mL.sup.−1) were supplemented and incubation proceeded for 15 min at RT. Cell lysates were centrifuged at 16,100×g for 1 h at 4° C. and the supernatant and pellet fractions were carefully separated.

[0115] Both fractions were then analysed by SDS-PAGE (12.5% polyacrylamide gels), loading equal volumes of each supernatant and its corresponding resuspended pellet. Samples were run in duplicate: one set for Coomassie blue staining (whole protein detection) and the other for Western blotting. Transference to PVDF membranes was carried out by semi-dry blotting, followed by blocking in TTBS plus powder milk. Primary antibodies were used at 1:20,000 dilution, either mouse anti-His (Sigma) or rabbit anti-mCherry (Abcam), then incubated with HRP-conjugated secondary anti-mouse/rabbit antibodies at 1:20,000 (Sigma). Antibody binding was detected using the ECL 2 substrate (Pierce-Thermo) and X-ray films (AGFA Curix RP2 plus).

Limited Proteolysis

[0116] Three μg aliquots of purified H.sub.10-LOV543 wt-WH1 were displayed in 15 μL of SEC buffer and trypsin (0.025 units), chymotrypsin (0.004 u.) or V8 (0.05 u.) proteases (Sigma) were supplied on ice. Digestions were left to proceed, either in the darkness or under blue light illumination (30,000 Lux), for 1 and 2 h at RT. Reactions were stopped by adding SDS-PAGE loading buffer and immediately boiling for 5 min, followed by electrophoretic separation in 12.5% polyacrylamide gels and Coomassie blue staining. Replicated samples of the chymotrypsin and V8 digestions (1 h, blue light) were transferred frozen to the Protein Chemistry facility at CIB-CSIC for Edman's N-terminal sequencing (5 cycles in a Procise 494 sequencer, Applied Biosystems).

Optogenetic Seeding of RepA-WH1(A31V) Amyloidogenesis In Vitro

[0117] Amyloidogenesis assays in vitro were carried out as described (Molina-García, L., et al., 2018, Methods Mol. Biol., 1779, 289-312), by setting on ice in 2 mL Eppendorf tubes 50 μL aliquots made of: RepA-WH1(A31V) (25 μM) in fibril assembly buffer (0.1M Na.sub.2SO.sub.4, 4 mM MgSO.sub.4, 20 mM Hepes. NaOH pH 8.0, 14% PEG4000, 6% MPD), but using as seeds sub-stoichiometric amounts (1:100) of the purified H.sub.10-LOV543 wt/m3-WH1 chimeras. As controls, both un-seeded samples and others including RepA-WH1(A31V)-mCherry aggregates (0.1 μg) as seeds were also casted. Samples were then incubated under continuous shaking, at 300 rpm and 25° C. for 3 h, in a thermomixer (Eppendorf), either in the darkness or illuminating with blue light (30,000 Lux). The tubes were pre-covered with aluminium foil or a thin PVC film, respectively, to avoid or allow the passage of light while preventing evaporation.

[0118] Two μL aliquots of the samples were then immediately diluted (1:10) in water, blotted to carbon-coated copper grids (400 mesh; EM Sciences) for negative staining with 2% uranyl acetate and subsequent visualization in a JEOL JEM-1230 electron microscope operated at 100 kV.

[0119] For amyloid detection with amyloidotropic molecules, five μL aliquots of the samples were incubated with 200 μL of a thioflavin-S (ThS) solution (0.05% w/v in 12.5 ethanol) for 30 min at RT. The suspension was then centrifuged (16,100×g for 1 h at 4° C.) and the pellets washed twice with 200 μL of PBS buffer. Final aggregates were resuspended gently in 5 μL PBS and displayed on glass slides until drying. Samples were inspected in a Nikon Eclipse 90i epifluorescence microscope using 40× Plan Fluor objective (NA=0.75) and a FITC filter (λ.sub.ex=482/35; λ.sub.em=536/40; exposure: 1 s). In parallel, 2 μL aliquots were displayed on glass slides, left to dry and incubated with 2 μL of a saturated solution of Congo red (CR) in 70% ethanol. Samples were then inspected for green-apple birefringence in a stereomicroscope (Leica MZ12.sub.5) working at 3.2× magnification and with its two linear polarizers crossed at 90°.

[0120] For immunodetection of amyloidogenic protein species, two μL sample aliquots were also serially diluted in fibril assembly buffer, but with no PEG4000 or MPD added, and spotted on a nitrocellulose membrane (Hybond-C extra; GE Healthcare) casted in a Bio-Dot vacuum-blotter (Bio-Rad). Membranes were then blocked as for Western blots (see above) and incubated with B3h7, a mouse conformational antibody specific for amyloidogenic oligomers of RepA-WH1 (1:5,000), or the rabbit polyclonal antibody α-WH1 (1:1,000). Secondary antibodies incubations and luminescence detection were performed as above.

Assessing H.sub.10-LOV543m3-WH1-mCherry phase transitions (liquid-hydrogel) in Vitro

[0121] 40 μL samples were assembled in Eppendorf tubes under blue light (30,000 Lux), including LOV543m3-WH1-mCherry (250 μM) in 0.5×SEC buffer, 0.008% ThS (see above), 9% PEG4000 and 4% MPD. After 2 h illumination, pictures were taken and the tubes were transferred to the darkness and incubated overnight at room temperature before taking a second set of photos. Two μL of each type of samples were picked-up by pipetting, stained with uranyl acetate and observed by TEM (see above).

[0122] In parallel, 10 μL drops with identical composition were casted on glass slides under blue light illumination, and then immediately layered with a cover slip. Specimens, sealed with nail polish to avoid drying, were kept under blue light illumination at the microscope setting for 1 h, and then light was switched-off and incubation proceeded for up to 24 h at room temperature. Epifluorescence was examined at the indicated time intervals in a Nikon Eclipse 90i microscope, using a 60× Plan Apo oil immersion objective (NA=0.95) and TRITC (λ.sub.ex=543/22; λ.sub.em=593/40; exposure: 0.5 s) and cyan (λ.sub.ex=438/24; λ.sub.em=483/32; exp.: 0.3 s) filters. Differential interference contrast (DIC) images were also acquired (exp.: 0.3 s).

Optogenetic Switching of H.SUB.10.-LOV543m3-WH1-mCherry Amyloidogenesis In Vivo

[0123] E. coli MDS42 cultures carrying the pRK2-LOV543m3-WH1-mCherry plasmid were grown in 100 mL of LB medium plus Ap.sub.100, in the darkness at 37° C., to OD.sub.600 nm=0.2, when cultures were split into two 45 mL aliquots, then placed in sterile bottles that included a magnetic bar. The rest of the cultures were left to grow to OD.sub.600 nm=1.0, when serial dilutions (7 μL drops) were spotted on LB-agar plus Ap.sub.100+/−1.0 mM IPTG for testing viability of bacteria, upon incubation with blue light (1,070 Lux) or in the darkness, for 30 h at 37° C. To the two 45 mL culture aliquots, IPTG was added (to 0.5 mM) and one of the flasks was covered with aluminium foil while the other was left unwrapped. Incubation under blue light (as above) or in the darkness proceeded at 37° C. with stirring (150 rpm) for up to 4 h. Cells from 15 mL of the cultures were harvested and washed twice with PBS. Pellets were resuspended in 1 mL of PBS and, while 0.9 mL were centrifuged and the cells stored at −70° C. for solubility tests (see above), bacteria from the other 0.1 mL were fixed with 4% paraformaldehyde (Sigma) for microscopy. Fixed cells were then stained with ThS, as indicated above. Bacteria were observed in a Nikon Eclipse 90i microscope, as in the previous section, but using a 100× Plan Apo oil immersion objective (NA=1.4).

Biophysical Characterization of the H.sub.10-LOV543 wt/m3-WH1 and H.sub.10-LOV543m3-WH1-mCherry Chimeras

[0124] Purified protein chimeras were characterized for their response to darkness and blue light illumination regarding their photocycle, association state, secondary structure and stability in SEC buffer.

[0125] The photocycle of the LOV2 moiety in the H.sub.10-LOV543 wt/m3-WH1 chimeras was studied by saturating with blue LEDs illumination (30,000 Lux, for 10 min) 600 μL protein solutions (3 μM wt and 5 μM m3) that were displayed in quartz cuvettes (1 cm path length) placed into the sample holder of an Ultrospec 3300pro spectrophotometer (GE Healthcare). To measure the return of excited flavin chromophore to the dark state, immediately after switching the light off time-lapsed spectra acquisition started under the control of the Swift II software, with the following parameters: 300-600 nm wavelength interval (0.5 nm/data point); 2,649 nm/min scan speed; 24 accumulated spectra (i.e., one every 15 s). For the H.sub.10-LOV543m3-WH1-mCherry chimera (8 μM), wavelength acquisition interval was extended to 650 nm to get the full band from excitation of mCherry. Return to dark state was analysed by fitting (MATLAB, The MathWorks Inc., release 2010a) a bi-exponential Levenberg-Marquardt function (R.sup.2=0.9941 wt, 0.9771 m3, 0.9929 mCherry) to the A.sub.447 nm data points, corresponding to the main dark absorption band in LOV2.

[0126] For determining the association state of the chimeras, besides SEC (100 μL samples, 50 μM; see above), sedimentation velocity experiments were performed in a Beckman-Coulter Optima XL-I analytical ultracentrifuge, at 48,000 rpm and 20° C., with a protein concentration (in SEC buffer) of 5 μM (H.sub.10-LOV543m3-WH1) and 8 μM (H.sub.10-LOV543m3-WH1-mCherry). Each sample was distributed between two centrifuge cells: one of them was scanned at 275 nm (for the double chimera) or 590 nm (for the triple chimera) as the dark state, and the other was illuminated at 450 nm as the lit state. Sedimentation coefficients distributions were calculated with SEDFIT.

[0127] Circular dichroism spectroscopy was performed with the chimeras in a Jasco 720 spectropolarimeter, with 150 μL (2.5 μM) of the protein samples in SEC buffer. Proteins were set in 0.1 cm path length quartz cuvettes hold at 20° C., and 7 spectra were acquired, in the darkness, at 50 nm.Math.min.sup.−1 and accumulated for signal averaging. Protein stability was estimated by thermal denaturation, measuring the variation of ellipticity (θ) at 220 nm with the increase of temperature (20-90° C.).