Genetically engineered liposwitch-based nanomaterials

Abstract

A fusion protein formed by a conditionally-activated lipid-binding domain that exposes a lipid moiety in response to a stimulus that is fused to a stimulus-responsive polypeptide domain. An exemplary conditionally-activate domain, a prototypical myristoyl switch, was fused with a thermo-responsive coil-protein. Biophysical characterizations confirmed the integrity and functionality of recoverin's myristoyl-switch within the fusion protein. Dynamic light scattering and cryo-TEM demonstrated that liposwitching modulates temperature-triggered phase separation and the hierarchical assembly of the fusion proteins. The fusion protein can therefore respond emergently to biologically relevant signals in a manner that mirrors the adaptability of riboswitches

Description

GENETICALLY ENGINEERED LIPOSWITCH-BASED NANOMATERIALS STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT

[0001] The present invention was made with government support under Grant No. 1R35GM142899 awarded by the National Institute of Health (NIH). The government has certain rights in the invention.BACKGROUND OF THE INVENTION1. FIELD OF THE INVENTION

[0002] The present disclosure relates to nanomaterials and, more particularly, to a fusion protein that combines the switching capabilities of a myristoyl-switch with a thermo- responsive coil protein so that switching occurs in response to a particular stimulus.2. DESCRIPTION OF THE RELATED ART

[0003] Gene fusion is a powerful strategy to merge functional properties of proteins with the versatility of nanomaterials. This approach enables seamless expression of fusion proteins with distinct domains for nano-assembly and functionality. Its major advantage lies in the genetic encoding that predetermines the nanoscale organization of proteins through sequence-specific interactions and domain characteristics like size, charge, and conformation. Moreover, recombinant expression of fusions is simpler and more cost-effective than conjugation of proteins to nanomaterials. These strengths have propelled the use of fusion proteins in various applications, including drug delivery, vaccine development, theranostics, tissue engineering, and enzyme engineering, particularly in scenarios where control of nanoscale structure is critical for function.

[0004] Nonetheless, the high specificity and control afforded by the genetic encoding also impose limitations on the dynamic alteration of protein structures after translation as the system’s architecture is largely fixed by its genetic blueprint. Consequently, a key focus area in the field is developing nanostructures capable of adapting to and interacting with environmental stimuli. Such adaptability would allow for programmable adjustments at the nano- / meso-scale, expanding their applications in biomedicine and biotechnology.

[0005] To enable adaptability, one strategy is to alter the properties of protein domains without changing their primary sequence. This has been successfully applied in fusions where one domain is a globular protein and the other a coil protein, such as elastinlike polypeptides (ELPs) that exhibit a lower-critical solution temperature to facilitate temperature-dependent aggregation. The utility of these systems lies in the design of the coilprotein to respond to environmental triggers, offering an external parameter like temperature to regulate domain interactions. However, the range of properties and the design space of these coil domains are still limited, necessitating new strategies to expand this class of adaptive nanomaterials.

[0006] Two nature-inspired strategies to overcome these limitations involve allosteric conformational changes and post- translational modifications (PTMs). Allostery, while increasingly employed in biosensing and catalysis, has seen limited integration into recombinant nanomaterials. On the other hand, PTMs represent a broader design space, with hundreds of identified modifications; however, their application in materials design has been constrained by the limited PTM capabilities of prokaryotic expression systems.BRIEF SUMMARY OF THE INVENTION

[0007] The present invention combines allosteric conformational changes and post- translational modifications (PTMs) that modify protein properties without altering their sequence to create programmable nanoparticles. These strategies are combined by using myristoyl-switches, which can be allosterically activated by a stimulus to reveals a sequestered lipid motif for programmable interactions with the cell membrane. This conditional exposure of lipids can dynamically regulate the assembly of fusion proteins and was used to form a fusion protein combining recoverin, a prototypical myristoyl switch, with a thermo-responsive coil-protein. Biophysical characterizations confirmed the integrity and functionality of recoverin’s myristoyl-switch within the fusion protein. Dynamic light scattering and cryo-TEM demonstrated that liposwitching modulates temperature -triggered phase separation and the hierarchical assembly of fusion proteins. The ability of these fusion proteins to respond emergently to biologically relevant signals open new frontiers in synthetic biology and cellular engineering, heralding a future that mirrors the adaptability of riboswitches.BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWING(S)

[0008] The present invention will be more fully understood and appreciated by reading the following Detailed Description in conjunction with the accompanying drawings, in which:

[0009] FIG. 1 is a schematic of a strategy for integrating liposwitching in protein nanomaterials. Diagram of the constitutive domains: recoverin (R-domain) as a myristoyl- switch, ELP (E-domain) as a thermo-responsive coil protein, and RE fusion. Recoverin transitions from a lipid-sequestered state to a lipid-exposed state upon Ca2+binding (PDB ids 1IKU and 1JSA), and ELP undergoes temperature-induced aggregation.

[0010] FIG. 2 is a characterization of purified recoverin-containing constructs using SDS-PAGE.

[0011] FIG. 3 is a characterization of recoverin controls using liquid chromatography and mass spectrometry, a) RP-HPLC confirms increased retention time for myristoylated R, consistent with the increased hydrophobicity. b,c) MALDI-TOF analysis shows a 210 Da increase in molecular weight of myristoylated constructs, aligning with myristoyl addition and water loss. Vertical lines are added to denote the expected m / z ratio for single- and double-charged ions.

[0012] FIG. 4 is a series of graphs of liquid chromatography and mass spectrometry of recoverin fusions, a) RP-HPLC indicates longer retention for myristoylated RE, due to lipidation-induced hydrophobicity. b,c) MALDI-TOF analysis detects molecular weight increases, with limited resolution for high molecular weight fusions, d) LC-MS trypsin digests reveal a distinct peak for myristoylated RE, notably at ~35 min. e) Isotope pattern of the novel peak matches myristoylated n-terminal peptide of RE, myr-GNSK. f) Extracted-ion chromatogram highlights the specific signal from trypsinized myristoylated RE.

[0013] FIG. 5 is a pair of graphs of emergent phase behavior of RE fusions compared to constituent domains. Variable-temperature turbidity assays of non-lipidated (a) and lipidated (b) RE fusions compared with their constitutive domains, i.e., R, m-R and ELP. The R-domain exhibits no temperature sensitivity, while the E-domain shows a distinct, calcium- independent phase transition. The fusion constructs exhibit unique behaviors with either single or dual transitions, influenced by myristoylation and calcium. Additionally, these factors modify the concentration dependencies of these transitions (detailed in FIG. 6).

[0014] FIG. 6 is a pair of graphs of the differential effects of myristoylation and calcium on concentration dependencies of RE Transitions. Four panels illustrate how various factors influence critical temperatures (Ttfor ELP, Tmfor Recoverin): a) Fusion effect [E vs. RE] without calcium or myristoylation (-m, -Ca2+). Gray line corresponds to data for E- domain; b) Impact of myristoylation (± m) without calcium (-Ca2+); c) Influence of calcium presence (± Ca2+) without myristoylation (-m); d) Combined effects of myristoylation and calcium (-m, -Ca2+) vs. (+m, +Ca2+).

[0015] FIG. 7 is a series of graphs of recoverin’s structural and functional integrity in RE fusions, a, b) CD spectra reveal an alpha-helical structure in recoverin (R) and a mixture of alpha-helical (from R) and random coil (from ELP) structures in RE, unaffected by myristoylation (±m). Error bars represent standard deviations from two replicates, c) Tryptophan fluorescence shift in both RE and R-domain, observed only with myristoylation(+m) and calcium (+Ca2+), indicates lipid release from recoverin’s core. The dashed line at 345 nm serves as a visual guide.

[0016] FIG. 8 is a pair of graphs of liposwitching-induced variations in the temperature-responsive phase behavior and nano-assembly of RE fusions, a) Variabletemperature turbidimetry demonstrates the influence of myristoylation (±m) and calcium (±Ca2+) on RE’s phase behavior, b) Variable temperature DLS elucidates that the nanoassembly of RE fusions is regulated by myristoylation and calcium, particularly above the ELP’s Tt. Error bars are standard deviations from two independent samples.

[0017] FIG. 9 is a graph of untransformed variable-temperature DLS data for RE fusions in presence or absence of myristoylation (± m) and calcium (± Ca2+). Temperaturedependent assembly of RE fusions is influenced by liposwitching.

[0018] FIG. 10 is a pair of graphs of variable-temperature DLS analysis of R-domain in presence or absence of myristoylation (± m) and calcium (± Ca2+). a) The untransformed DLS data shows that the size of R-domain assemblies is similar, regardless of myristoylation or calcium binding, b) The transformed DLS data further confirm that the myristoyl- switching of the R-domain does not significantly influence its self-assembly. These results emphasize that the distinct emergent assembly observed in RE fusions (FIG. 8b) is primarily driven by the interaction between the R-domain and ELP-domain.

[0019] FIG. 11 is a series of images of the impact of liposwitching on the assembly of RE fusion constructs. Cryo-TEM at 35 °C and confocal microscopy at 45 °C (insets) is used to visualize the nano- and meso-scale organization of RE assemblies, a) (-m, -Ca2+); b) (+m, - Ca2+); c) (-m, +Ca2+); and d) (+m, +Ca2+).

[0020] FIG. 12 is an image of the visualization of ELP-domain coacervates using confocal microscopy. ELP-domain forms canonical liquid-like coacervates with spherical morphology above its transition temperature.

[0021] FIG. 13 is a graph showing that modifying the composition of the E-domain, by increasing its hydrophilicity (E’) and decreasing its length (E”), can be used to modulate the liposwitching threshold.

[0022] FIG. 14 is a series of images showing that RE nanoclusters undergo efficient endocytosis in DU 145 cells when phase-separation is induced in the presence of myristoylation and calcium. DU145 cells were incubated with AlexaFluor488-labeled RE fusion proteins (1 pM) for 4 h in calcium-containing (top row) or calcium-free (1 mM EGTA, bottom row) media. Myristoylation was included (right column) or excluded (left column) to modulate nanocluster formation. Cells were fixed, stained for the nucleus (Hoechst) andmembrane (WGA), and imaged by confocal microscopy. Robust internalization (puncta) is only observed when REs form nanoclusters (bottom right), indicating enhanced endocytic uptake under conditions that favor multivalent assembly DETAILED DESCRIPTION OF THE INVENTION

[0023] Referring to the figures, wherein like numerals refer to like parts throughout, there is seen in FIG. 1, a schematic demonstrating the potential of combining allosteric conformational changes with post-translational modifications (PTMs) to form a conditionally activated liposwitch that responds to diverse biological triggers. In addition to the calcium switching used as an example here, the present invention includes pH-responsive mechanisms, such as Hisactophilin from Dictyostelium discoideum, lipid composition- responsive proteins, such as the HIV-1 matrix protein (Gag), which is activated by PI(4,5)Pa, ligand-binding responsive systems, such as ADP-ribosylation factors (ARFs), which expose their lipid anchor upon binding to GTP.

[0024] The present invention may be accomplished by fusing a conditionally- activated lipid-binding domain that exposes a lipid moiety in response to a stimulus to a stimulus-responsive polypeptide domain. The stimulus may be any one or more of the calcium ion concentration, the pH, the presence of a specific lipid species, or the binding of a specific molecular ligand.

[0025] The conditionally-activated lipid-binding domain may include a myristol switch, e.g., recoverin, as well as other proteins with similar mechanisms. While recoverin is used as a prototypical example herein, other myristoyl switches, including but not limited to members of the Neuronal Calcium Sensor (NCS) family such as Hippocalcin, Neurocalcin, Visinin-like protein 1, and Guanylate Cyclase Activating Proteins (GCAPs), can be used in the present invention.

[0026] The stimulus-responsive polypeptide domain may comprise an elastin-like polypeptide as well as other functionally equivalent thermo-responsive domains, such as Silk- Like Proteins (SLPs), and Collagen-Like Peptides (CLPs). These domains can provide, among other things, a thermo-responsive polypeptide domain that undergoes a temperaturedependent phase transition as well as a polypeptide domain that exhibits a lower critical solution temperature (LCST) or an upper critical solution temperature (UCST).

[0027] Myristoyl-switches, such as recoverin, are allosteric proteins bearing a myristoyl (m) lipid PTM. In the absence of calcium, recoverin hides its lipid anchor in a hydrophobic pocket; upon calcium binding the lipid anchor is exposed, as seen in FIG. 1. Harnessing lipid mo ie ties’ conditional exposure allows for dynamic regulation of fusionprotein assembly, a concept referred to herein as ‘liposwitching’. Calcium binding triggers allosteric conformational changes in recoverin, causing a 45-degree rotation of the N-terminal domain. When myristoylated, recoverin’s allosteric shift exposes the hydrophobic lipid, bringing it near the C-termini fusion and thereby influencing adjacent protein hydration, conformation, and the energy landscape. ‘Liposwitching’ can therefore effectively combine and transduce allosteric conformational shifts and PTMs, enabling dynamic regulation of the fusion protein’s assembly and function.

[0028] To evaluate the present invention, a fusion protein combining recoverin with a thermo-responsive ELP-domain, designed to leverage recoverin’s liposwitching mechanism — unveiling hydrophobic motifs in response to Ca2+— for modulating the temperature-sensitive aggregation of the fusion protein was developed. Initially, the functional integrity of recoverin’s myristoyl-switch within the fusion protein was established, followed by demonstrations of how its liposwitching influences phase separation, and hierarchical assembly of the RE fusions. Diverging from traditional globular-coil fusions that respond to single stimuli, the fusion protein of the present invention uniquely integrates multiple responsive elements: myristoylation, calcium binding, and temperature sensitivity. This intricate design enables dynamic adaptability and emergent response to external and biologically relevant signals.

[0029] The design of the fusion protein — the orientation and identity of domains — was informed by the stability and functional requirements of recoverin. Recoverin’s myristic acid must be located at the N-terminus to function as a liposwitch, dictating its placement at the fusion’s N-terminus, leaving the thermo-responsive ELP-domain to be positioned at the C-terminus. An ELP composed of 80 repeats of GVGVP (SEQ ID NO: 1) was selected, ensuring a transition temperature below 42 °C to maintain recoverin’s stability at physiological temperature. The fusion construct, recoverin-ELP (RE) and single-domain controls, recoverin (R) and ELP, were expressed in E. coli, and characterized using SDS- PAGE (FIG. 2), RP-HPLC, mass-spectrometry (FIG. 3 and FIG. 4).

[0030] To verify that the RE fusion retains recoverin’s myristoyl-switching behavior — critical for dynamic nano-assembly regulation — its biophysical properties was compared with a recoverin-only control (R) using circular dichroism (CD) and tryptophan autofluorescence. CD showed the RE fusion combines R-domain’s a-helical structure with random coil and [l-turn structures of the ELP-domain, a composition reflective of each domain’s size and unaffected by myristoylation (±m). For tryptophan autofluorescence, on a specific tryptophan residue (W31) in recoverin was the focus, which is part of thehydrophobic pocket that sequesters the lipid. Calcium binding leads to the lipid’s exposure, causing a red shift in fluorescence. The experiments demonstrated this shift in autofluorescence for both R and RE, but crucially, only when the myristoyl group and calcium were both present (+m, +Ca2+), indicating functional calcium binding and consistent conformational changes. These results confirm that recoverin within the RE fusion maintains its secondary structure and calcium responsiveness, a pre-requisite for modulating the temperature-responsive assembly of the ELP-domain.

[0031] After biophysical characterizations, variable-temperature turbidimetry was used to investigate how recoverin’s liposwitching, mediated by myristoylation and triggered by calcium binding, affects the ELP-domain’s temperature-triggered phase separation in RE fusions. Turbidity profiles for 10 pM RE solutions (FIG. 8a), with or without calcium (± Ca2+) and myristoylation (± m), revealed a distinctive phase separation behavior different from ELP or R domains (FIG. 5). The RE fusion exhibited two transitions: the first reflecting the ELP-domain’s lower-critical solubility temperature (Tt), which myristoylation reduced but calcium alone did not alter; the second transition, linked to recoverin’s thermal unfolding (Tm), was increased by both myristoylation and more notably, calcium binding (FIG. 6). These findings elucidate the influence of myristoylation and calcium on RE’s phase behavior, thus validating that liposwitching can control the fusion’s phase separation.

[0032] After confirming that recoverin’s liposwitching alters the RE fusion’s phase behavior, dynamic light scattering (DLS) was employed to assess its influence on nanoassembly at various temperatures. The DLS analysis (FIG. 8 and FIG. 9) not only highlighted how myristoylation and calcium binding modulate nano-assembly, but it also uncovered complex temperature-dependent interactions between these factors. This interplay was analyzed by plotting the natural log ratio of hydrodynamic sizes, Log Rh [(+m) / (-m)], across temperatures with and without calcium (±Ca2+). Positive values signify larger sizes for myristoylated constructs compared to non-myristoylated ones. Deviations from zero on each curve reflect the impact of myristoylation, while the difference between the curves for ±Ca2+indicates their combined effect. This analysis yielded two crucial insights: first, the RE fusion exhibits unique behavioral patterns across specific temperature regimes, patterns that are absent in the single-domain control (FIG. 10). Second, these distinct behavioral patterns align closely with the ELP-domain’s Ttand recoverin’s Tm, supporting our hypothesis that liposwitching in the R-domain effectively regulates the E-domain’s temperature-triggered aggregation in fusion constructs.

[0033] In the absence of calcium, the temperature-dependent aggregation of RE shows minimal impact from myristoylation alone, as indicated by the Log Rh [(+m) / (-m)] ratio staying near zero for most temperatures. However, distinct deviations at 24-28 °C and 44-48 °C highlight myristoylation’s subtle influence on the ELP and recoverin domains. Specifically, a slight reduction of ELP’s Ttand an increase in recoverin’s Tmdue to myristoylation are observed. At 24-28 °C, an increase in this size ratio signifies formation of larger aggregates due to ELP dehydration when the temperature surpasses its myristoylation- lowered Tt. This effect diminishes above 28 °C whereas the ELP in non-myristoylated fusion also transitions, resulting in similar aggregate sizes. In contrast, at 44 °C, a decrease in the ratio reflects the formation of larger aggregates due to the unfolding of non-myristoylated recoverin, a process that myristoylation delays until temperatures exceed 48 °C. Conversely, in the presence of calcium, the effects of calcium and myristoylation are minimal below the ELP’s Tt. However, beyond this threshold, the synergistic interaction between calcium and myristoylation significantly alters the size and stability of the assemblies, evidenced by the divergence in the ±Ca2+curves. This pattern aligns with the concept that calcium-induced changes in recoverin exert their maximal effect on the second protein when recoverin is myristoylated.

[0034] Building on the DLS analysis, a range of microscopy techniques was employed to elucidate the influence of liposwitching on the nano- and meso-scale structures of RE assemblies (FIG. 11). At 15 °C (T < Tt), microscopy showed no significant assemblies in RE, indicating that without myristoylation or calcium binding, the amphiphilicity of the hydrated ELP-domain and R-domain is not sufficiently different to drive assembly. However, as the temperature increases beyond the ELP-domain’s Tt, its increased hydrophobicity becomes a key factor in promoting the assembly of RE proteins. At 35 °C (Tt< T < Tm), cryo-TEM was utilized to examine the nanoscale assembly. For observations at 50 °C (T > Tm), where vitrification poses challenges, confocal imaging was used.

[0035] Cryo-TEM images (FIG. 11) reveal how lipidation (± m) and calcium binding (± Ca2+) modulate the temperature-dependent assembly of the fusion protein, affecting contrast, morphology, and cluster distribution. Calcium influences nanoscale clustering and higher-order organization of clusters irrespective of lipidation, while lipidation’s effect on nanoscale organization is subtler and calcium-dependent, underscoring the interplay between calcium and the lipid.

[0036] At 35 °C, RE (-m, -Ca2+) forms a diffusive network of interconnected globules (FIG. I la), but at temperatures above Tm, it transitions to large, fractal clusters distinct fromELP-domain coacervates (FIG. 12). In contrast, myristoylated RE (+m, -Ca2+) at 35 °C shows minimal change in nanoscale clustering (FIG. 1 lb), but forms more extended structures, possibly due to altered recoverin surface hydration post-myristoylation. At 50 °C, this results in larger, less interconnected aggregates (FIG. 1 lb, inset). With calcium, RE (-m, +Ca2+) forms worm-like micelles at 35 °C (FIG. 11c), maintaining small structures even at 50 °C. In the presence of both myristoylation and calcium, RE (+m, +Ca2+) assembles into densely packed clusters at 35 °C (FIG. l id), demonstrating the synergistic effect of lipidation and calcium on system assembly. These calcium-limited interactions yield aggregates larger than those in RE, yet at 50 °C remain below confocal resolution.

[0037] In summary, the present invention includes the design for a fusion protein that provides a conditionally activated liposwitch that responds to diverse biological triggers. For example, the fusion protein of the above Example successfully combines the liposwitching capabilities of a natural myristoyl-switch with thermo-responsive aggregation of a coil protein. The fusion protein phase behavior and nanoclustering can be effectively regulated by two biochemical signals, myristoylation and calcium. Given the widespread occurrence and precise regulation of these signals in living cells, the RE fusion emerges as a versatile platform poised to harness cellular biochemistry for biotechnological and biomedical applications.

[0038] The intricate interplay uncovered between lipidation, the calcium-dependent allosteric switch, and the thermo-responsive domain underscores the need for a deeper biophysical understanding to refine and advance this emerging class of biomaterials. Looking ahead, myristic acid analogues may be explored, such as the unsaturated fatty acids found in recoverin isoforms, to fine-tune the response of recoverin fusions. Additionally, alternative myristoyl-switches may be explored, aspiring to develop nanobiomaterials responsive to a diverse range of physiological triggers like pH and membrane composition. Envisioning a future that mirrors the adaptability of riboswitches, these liposwitch-based platforms could dynamically interact with and modulate cellular signals, thereby opening new frontiers in synthetic biology and cellular engineering.EXAMPLE

[0039] MATERIALS AND METHODS

[0040] Materials. Restriction enzymes, ligase, NEBuilder ® HiFi DNA Assembly cloning kit, Monarch ® DNA Gel Extraction Kits, corresponding buffers, and chemically competent Eb5alpha and BL21(DE3) cells were purchased from New England Biolabs (Ipswich, MA). DNA oligonucleotides and gene fragments were synthesized by IntegratedDNA Technologies (Coralville, Iowa). PureLink™ Quick Plasmid Mini-Prep Kits, high- performance liquid chromatography-(HPLC) grade acetonitrile, liquid chromatography mass spectrometry (LC-MS) grade water, acetonitrile, and formic acid, HisPur™ Cobalt Resin Spin Columns, Zeba™ desalting columns (3k MWCO), sinaptic acid, AlexaFluor-488 NHS Ester, mass spectroscopy grade Pierce™ trypsin protease, myristic acid, agarose, tris-HCl, imidazole, sodium chloride, ethylene glycol-bis(?-aminoethyl ether)-N,N,N?,N?-tetraacetic acid (EGTA), Calcium Chloride (CaC12+), ammonium bicarbonate, sodium bicarbonate, chloramphenicol, ampicillin sodium salt, and anhydrous DMSO was purchased from Thermo Fisher Scientific (Rockford, IL). Mini-PROTEAN® TGX Stain-Free™ Precast Gels, Precision Plus Protein™ Unstained Protein Standards, and 4x Laemmli buffer were purchased from BioRad Laboratories, Inc (Hercules, CA). Polyethersulfone (PES, 0.22 pm) syringe filters, ZipTip™ Pipette filters, Amicon® Ultra- 15 Centrifugal Filter (3k MWCO), ProteoMass™ Albumin and Aldolase MALDI-MS standard, sinaptic acid, ammonium bicarbonate, and trifluoro acetic acid (TFA), phenylmethylsulfonyl fluoride (PMSF) were purchased from Sigma-Aldrich (St. Louis, MO). Isopropyl 7-D-l -thiogalactopyranoside (IPTG) was purchased from GoldBio (St. Louis, MO). Kanamycin sulfate, sodium chloride, yeast extract, and tryptone was purchased from Bio Basic (Markham, ON, CA). p-Slide 8 Well Glass Bottom microscope slide was purchased from Abidi. The carbon-coated grid (CF300-Cu) was purchased from Electron Microscopy Sciences (Hatfield, PA). Deionized water was obtained from a Milli-Q® system (Millipore SAS, France). Simply Blue™ SafeStain was purchased from Novex (Carlsbad, CA). All chemicals were used as received without further purification.

[0041] Cloning. The copy DNA (cDNA) for Bos taurus recoverin (Uniport Accession # P21457) was obtained using Sanger sequencing from a template plasmid kindly provided by Dr. James Ames (University of California, Davis). Modifications were made to the cDNA sequence to remove internal restriction enzyme sites necessary for assembly with the ELP gene, including Acul, BseRI, and Bgll. These modifications were achieved through silent mutations and the addition of 20-40 base pair homology regions compatible with the recipient vector. The modified gene fragment was synthesized by IDT DNA Technologies and subsequently cloned into a linearized pET-24 plasmid using the NEBuilder® HiFi DNA Assembly according to manufacturer protocol.

[0042] The fusion protein’s (RE) gene was synthesized through two rounds of "plasmid reconstruction by recursive directional ligation (PRe-RDL)" technique. First, plasmids carrying the recoverin and ELP genes underwent double digestion with Bgll, andeither Acul or BseRI. Following this, DNA fragments containing either Rec or ELP genes were separated via agarose gel electrophoresis and then purified using the Monarch gel extraction kit, in accordance with the manufacturer’s guidelines. These purified fragments were ligated together using the Quick Ligase Kit (New England Biolabs). The ligated products were subsequently introduced into chemically competent NEB5a cells, and colonies were selected based on kanamycin resistance. Following DNA sequencing, this procedure was replicated to incorporate the C-terminal octa-histidine tag (8x-His). Finally, nanopore sequencing was utilized to confirm the sequence of all the plasmids.

[0043] Protein Expression. Protein expression was conducted using E. coli BL21(DE3) strain in 2x yeast extract and tryptone (2xYT) broth. For constructs requiring myristoylation, an orthogonal plasmid encoding yeast N-myristoytransferase (NMT) was cotransformed. This plasmid is characterized by a pl5A origin and includes a chloramphenicol resistance marker. To accommodate the secondary plasmid, growth media for expression of lipidated constructs was supplemented with both kanamycin (45 pg / mL) and chloramphenicol (25 pg / mL). On the other hand, the unmodified proteins were expressed in 2xYT supplemented with only kanamycin (45 pg / mL).

[0044] A freshly transformed colony was used to inoculate a 50 mL seed culture, which was incubated at 37 °C with shaking at 250 rpm until the culture achieved the optical density (OD600) of 0.6. The seed culture was used to inoculate 1 -liter expression cultures, using a 1:100 inoculum ratio. These expression cultures were cultivated at 37 °C with shaking until they reached an optical density OD600 of 0.6-0.8. Subsequently, the temperature of the incubator was lowered to 28 °C. For the myristoylated constructs, 100 pM myristic acid was added to the expression media. After an incubation period of 15 minutes, protein expression was induced by supplementing the media with 0.5 mM IPTG. Six hours post-induction, the cells were harvested using centrifugation (5000 xg, 20 min, 4 °C). The cell pellets were then resuspended in 5 mL of Tris-buffered saline (TBS; 50 mM Tris-HCl, 137 mM NaCl, pH 7.4) per liter of expression media, and subsequently stored at -80 °C before purification.

[0045] Protein Purification. Protein purification was conducted in two stages: initial purification using immobilized metal affinity chromatography (IMAC) and subsequent polishing with reverse-phase high-performance liquid chromatography (RP-HPLC). Before cell lysis, the frozen suspensions were thawed and treated with 1 mM phenylmethylsulfonyl fluoride (PMSF). Lysis was conducted on ice using sonication (Fisher, Model 505 Sonic Dismembrator), consisting of two cycles (1:30 min each, 10 s on, 60 s off, 60W), followed byclarification through centrifugation (21000 xg, 10 min, 4 °C). The cell lysate was supplemented with 10 mM Imidazole and incubated with HisPurTM Cobalt resin at a ratio of 10 mL lysate per 3 mL resin for 30 minutes. The resin was subsequently washed three times with two volumes of wash buffer (50 mM Tris-HCl, 10 mM imidazole, 300 mM NaCl, pH 7.4). His-tagged proteins were eluted by incubating the resin five times with one volume of elution buffer (50 mM Tris-HCl, 300 mM imidazole, 300 mM NaCl, pH 7.4). The identity and purity of the elution fractions were confirmed by SDS-PAGE before combining. The pooled fractions were concentrated using a centrifugal filter unit (AMICON® Ultra-15, 3k MWCO), following the manufacturer’s instructions. The concentrated eluates were further purified to homogeneity (>95% purity) using preparative reverse-phase HPLC (RP-HPLC), with details provided below. The HPLC fractions were flash-frozen in liquid nitrogen, lyophilized, and the resulting protein powder was stored at -20 °C.

[0046] RP-HPLC. HPLC analysis was performed using Shimadzu Prominence HPLC systems, with an LC2030 i-series for analytical and a Prominence modular HPLC for preparative purposes. Flow rates were maintained at 1 mL / min for analytical HPLC and 4.2 mL / min for preparative HPLC. Both systems employed Phenomenex Jupiter C18 columns (5 pm, 300 A), sized 250 x 4.6 mm for analytical and 250 x 10 mm for preparative applications. The mobile phase consisted of a linear gradient, starting with solvent A (H2O + 0.1% TFA) and transitioning to solvent B (Acetonitrile + 0.1% TFA), following this gradient profile: 0 min at 0% solvent B, 45 min at 90% solvent B. Background correction of HPLC traces was achieved by subtracting a blank trace acquired under identical conditions.

[0047] ELP Purification. The ELP control was purified by Inverse Transition Cycling (ITC) according to previously published protocols. Briefly, ITC utilized the lower-critical solubility temperature (LCST) property of the ELP, wherein the protein is soluble at low temperatures and insoluble at high temperatures. Through isothermal switching, the ELP can be separated from protein contaminants. Lysis was performed via sonication and clarified by centrifugation (parameters above). Protein contaminants sensitive to high temperatures were removed through a "bakeout," where the lysis supernatant was heated at 50 °C for 20 minutes, cooled on ice for 10 minutes, and clarified by centrifugation. The ELP was precipitated from solution through the addition of up to 2.5 M crystalline NaCl. The solution, now turbid due to insoluble ELP, was centrifuged at high-speed and the supernatant decanted. The resulting ELP pellet was resuspended in cold ddH2O and centrifuged at 4 °C to remove any final contaminants. Purity was assessed through analytical RP-HPLC

[0048] Fluorescent Labeling. Lyophilized proteins were reconstituted at 1 mg / mL in 100 rnM sodium carbonate buffer (pH 9), with stirring at 4 °C. Alexa Fluor™ 488 NHS Ester dye was prepared in anhydrous DMSO at 0.5 mM concentration. To the protein solution, 30 nmol of the dye stock was added and mixed under stirring for 3 hours at 4 °C. The reaction mixture was subsequently concentrated 4-fold using an Amicon® ultra- 15 centrifugal filter (3k MWCO). Excess dye was removed by passing the concentrated mixture through a Zeba™ desalting column (Thermo Fisher Scientific). The efficiency of labeling was determined by measuring the fluorophore concentration at 488 nm.

[0049] Sample Preparation. Lyophilized protein was dissolved in TBS at 4 °C to prepare the stock solution. Separate stock solutions of CaC12 and EGTA were also prepared in TBS at a concentration of 100 mM. The protein stock solution’s concentration was confirmed using UV-Vis spectroscopy at 280 nm, employing an extinction coefficient (?) of 25440 M-l.cm-1. This concentrated stock was diluted to the required protein concentration for each characterization technique (details provided below) and supplemented with either CaC12 (+Ca2+) or EGTA (-Ca2+) to the final concentration of 1 mM. All protein samples in this study were prepared using this method, except for those used in Circular Dichroism (CD) analysis. For CD, samples were prepared in Tris-buffer rather than TBS to minimize chloride ion interference.

[0050] Sodium dodecyl-sulfate polyacrylamide gel electrophoresis. The purity and molecular weight of the purified proteins were assessed using 4-20% Mini-PROTEAN® TGX Stain-Free™ Precast Gels. Initially, the gels were visualized under UV light with a BioRad Gel Doc EZ Imager, followed by staining with SimplyBlue™ (Coomassie Blue) to detect any residual protein contaminants. All samples for protein gel analysis were prepared following the Laemmli method.

[0051] MALDI-TOF-MS. Matrix-assisted laser desorption / ionization time-of-flight mass spectrometry (MALDI-TOF-MS) was conducted on a Brunker micro flex LRF with a microScout ion source. Samples for analysis were prepared by mixing an equal volume of sample and synaptic acid (SA) matrix (SA was prepared at a concentration of 10 mg / mL in 70% acetonitrile + 0.1% TFA). 2uL of this mixture was spotted on a steel target plate. All spectra were calibrated against an aldolase (39,211.28 Da) and albumin standard (66,429.09 Da).

[0052] LC-MS Analysis of Trypsin Digest. For LC-MS analysis, a 2.5 pL aliquot of the protein solution (2 mg / mL) was mixed with 100 mM ammonium bicarbonate (pH 8). The mixture underwent trypsin digestion (1:50 wt / wt ratio) for 8 hours at room temperature. Thedigested sample was desalted using ZipTip® columns. Chromatographic separation was performed using an Ultimate 3000 nano-LC pump (Dionex, Mountain View, CA, USA) and a Cl 8 column (Acclaim PepMap™ 100, 75 pm x 25 cm, nanoViper Cl 8, 3 pm, 100 A). This system was connected to an LTQ-Orbitrap-ETD XL mass spectrometer (ThermoFisher Scientific, San Jose, CA, USA) equipped with a nanospray ion source (New Objective, Wobum, MA, USA). The loading pump used a mobile phase of 0.1% formic acid in water at 200 pL / min, while the nano-LC pump used 80% acetonitrile with 0.1% formic acid (solvent B). Peptides were eluted at 300 pL / min using a linear gradient from 4% to 55% solvent B over 30 minutes, then ramped from 55% to 90% solvent B over 5 minutes.

[0053] Circular Dichroism. Spectra were recorded on an Aviv Model 420 CD spectrometer at 15 °C and processed using Aviv software v3. Protein solutions prepared at concentrations of 2.5 pM (for RE and m-RE) and 5 pM (for R and m-R) were analyzed in a 1 mm path length quartz cell across the 190-320 nm wavelength range. The background spectrum, obtained using the buffer alone, was subtracted from each protein spectrum prior to converting the data into mean molar residue ellipticity (in deg.cm2.dmol-l). Data deconvolution was performed using Dichroweb with the CONTIN model.

[0054] Tryptophan Autofluorescence. Fluorescence spectra were recorded using a Cary 100 eclipse spectrofluorometer (Agilent) at room temperature. Protein solutions were prepared at a concentration of 1 pM. The fluorescence emission spectra were recorded from 300 to 400 nm, while exciting the samples at 290 nm.

[0055] Variable-Temperature Turbidimetry. The temperature-dependent phase behavior of proteins was investigated using a Cary 300 UV-Vis Spectrophotometer (Agilent) equipped with a Peltier temperature controller. Turbidity was monitored by recording the absorbance at 350 nm across a temperature range of 15-65 °C, with a ramp rate of 1 °C / minute. Fusion protein (RE) and ELP samples were tested at concentrations ranging from 1-100 pM. Due to its limited solubility, recoverin was analyzed at concentrations of 1-50 pM. Transition temperatures (e.g., Tt or Tm) were determined by analyzing the first derivative of the turbidity plots and identifying the inflection points of the turbidimetry profdes.

[0056] Dynamic Light Scattering. Variable-temperature dynamic light scattering (VT- DLS) experiments were performed using a NanoLab 3D™ instrument (LSI Instrument) equipped with a 90° detector in a 3D cross-modulated geometry. Prior to DLS analysis, protein solutions were filtered through a 0.22 pm PES filter (Millipore Millex) chilled at 4 °C. VT-DLS measurements were carried out over a temperature range of 16-54 °C,increasing in 2 °C increments with a 1 -minute equilibration period at each temperature. Each measurement involved three 20-second repetitions, with the instrument’s automatic attenuation feature adjusting laser intensity to achieve a scattering intensity of around 250 kHz. The acquired correlation functions were analyzed using the cumulant algorithm in LSLab Software.

[0057] Confocal Microscopy. Protein structures were visualized using a Zeiss LSM 980 with Airyscan 2 confocal microscope, an inverted and fully motorized system, equipped with a 480 nm laser light source and a 20x objective. The microscope featured a heated stage and an environmental chamber to ensure consistent temperature control. Protein samples (30 pM, with 10% labeled) were placed in an Abidi 8-well glass microscope slide and incubated at each designated temperature for 10 minutes prior to imaging. The solution’s temperature was continuously monitored with a thermocouple (TCI Electronics) placed in a buffer-filled adjacent well. Image stacks obtained from confocal microscopy were analyzed and merged using Fiji software’s Z-stack function, employing the maximum intensity projection algorithm for image reconstruction.

[0058] Cryo-Transmission Electron Microscopy. Fresh protein solutions were prepared in TBS at 4 °C and incubated at 35 °C for 10 minutes. Post-incubation, the samples were deposited onto freshly plasma-cleaned Quantifoil grids (Quantifoil Micro Tools GmbH, Germany) maintained in a Mk IV Vitrobot (Thermo Fisher Scientific) with 100% humidity. Vitrification was achieved by plunging the grids into liquid ethane, with subsequent storage under liquid nitrogen. Imaging was conducted on a Tecnai BioTwin 120kV transmission electron microscope, equipped with a Gatan SC1000A CCD camera, and operated at liquid nitrogen temperature. Low-dose imaging was performed using Gatan 626 or 910 holders. Image processing was carried out in ImageJ, initially applying the despeckle plugin (median filter) for denoising. For calcium-containing samples requiring enhanced clarity, additional processing was done using a fast Fourier transform (FFT) band-pass filter (up to 10 pixels, down to 400, no stripe suppression, with autoscale and saturation).

[0059] DNA AND PROTEIN SEQUENCES

[0060] Recoverin DNA sequence:

[0061] The underlined sequence is the homology area with pET24a plasmid. The recognition sequences of BseRI and BamHI are highlighted as italics and double underline, respectively.

[0062] CTAGAAATAATTTTGTTTAACTTTAAGAAGGHGU4GTACATATGGG CAACAGCAAGAGTGGGGCCCTGTCCAAGGAGATCCTGGAAGAGCTGCAGCTGAACACCAAGTTCACGGAGGAAGAGCTGAGCTCCTGGTACCAGTCCTTCCTGAAAGA GTGTCCCAGTGGTCGGATCACCCGGCAGGAGTTCCAGACCATCTACTCCAAGTTC TTCCCCGAGGCCGACCCCAAGGCCTATGCCCAGCACGTGTTCCGAAGCTTTGATG CCAACAGCGATGGCACCTTGGACTTCAAGGAGTATGTCATCGCCTTACACATGACCAGCGCGGGCAAGACCAACCAGAAGCTGGAGTGGGCCTTCTCCCTCTATGACGT GGATGGCAATGGGACCATCAGCAAGAACGAGGTGCTGGAGATTGTCACGGCTAT CTTCAAAATGATCAGCCCTGAGGACACAAAGCATCTCCCAGAAGACGAGAACACTCCGGAAAAGCGAGCAGAGAAGATCTGGGGATTCTTTGGCAAGAAGGATGATGA TAAACTTACAGAGAAAGAATTCATCGAAGGGACCCTGGCCAATAAGGAAATTCT GCGACTGATTCAATTCGAGCCTCAAAAAGTGAAGGAGAAACTGAAAGAAAAGAAACTCGGCTGATAATGATAATCTTCAGGATCCGAATTCGAGCTCCGTCGACAAGCTTGCGGCCGCAC (SEQ ID N0:2)

[0063] Sequence of proteins

[0064] Recoverin-His (R)

[0065] GNSKSGALSKEILEELQLNTKFTEEELSSWYQSFLKECPSGRITRQEFQTIYSKFFPEADPKAYAQHVFRSFDANSDGTLDFKEYVIALHMTSAGKTNQKLEWAFSL YDVDGNGTISKNEVLEIVTAIFKMISPEDTKHLPEDENTPEKRAEKIWGFFGKKDDDKLTEKEFIEGTLANKEILRLIOFEPQKVKEKLKEKKLGHHHHHHHHG (SEQ ID NO:3)

[0066] Recoverin-ELP(V,80)-His (RE)

[0067] GNSKSGALSKEILEELQLNTKFTEEELSSWYQSFLKECPSGRITRQEFQTIYSKFFPEADPKAYAQHVFRSFDANSDGTLDFKEYVIALHMTSAGKTNQKLEWAFSL YDVDGNGTISKNEVLEIVTAIFKMISPEDTKHLPEDENTPEKRAEKIWGFFGKKDDDKLTEKEFIEGTLANKEILRLIQFEPQKVKEKLKEKKLGVGVPGVGVPGVGVPGVGVPG VGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVP GVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVG VPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGV GVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVP GVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGVPGVGV PGHHHHHHHHG (SEQ ID NO:4).

[0068] The phase behavior of chimeric polypeptides is altered by recoverin’s lipidation or calcium binding. The results were presented across multiple panels for a thorough evaluation. Initially, ELP fusion’s impact was assessed by comparing RE to E in theabsence of calcium (FIG. 6a). These results suggest that fusion of R and E alters the properties of both protein domains: 1) by lowering the Ttof ELP-domain and reducing its concentration dependence (gray arrow); 2) rendering the R-domain more susceptible to unfolding and aggregation at higher temperatures. The reduction in ELP’s Ttis attributed to the presence of a hydrophobic patch in recoverin near the fusion site that alters the hydration of ELP. As the temperature is increased above the Tt, the dehydration and aggregation of the ELP-domain may facilitate inter-molecular aggregations of recoverin at elevated temperatures (i.e., reducing the Tmof R-domain).

[0069] FIG. 6b compares the behavior of RE with m-RE to highlight the impact of myristoylation in the absence of calcium. First, even though the myristoyl group is attached to Recoverin, and not to the ELP, even when it is sequestered inside recoverin, it still alters the hydration and phase separation of the appended ELP, presumably by altering the hydration and surface properties of recoverin (red arrow). Second, in the absence of calcium, myristoylation increases recoverin’s aggregation temperature and alters its concentration dependence (gray arrow). This indicates that m-RE is more thermally stable than RE, and its unfolding likely occurs through a different mechanism, such as a cooperative intramolecular event.

[0070] Unlike lipidation that affected both domains, binding of calcium to RE increases the thermal stability of the R-domain only as the second transition disappears (FIG. 6c, green arrow), while having minimal effect on the ELP-domain. When myristoylation and calcium binding are combined (FIG. 6d), their effect on phase separation is combined as well. Myristoylation reduces the Ttof the ELP-domain (red arrow). On the other hand, both factors stabilize recoverin and increase its thermal stability (green arrow). However, calcium binding appears to play a more significant role in stabilizing the structures compared to myristoylation. Together, these experiments demonstrate that the liposwitching capability of recoverin modulates the temperature-dependent phase behavior of the fusion protein. EXAMPLE 2

[0071] Modulation of Liposwitching by E-domain Engineering

[0072] Based on the observed dependence of RE nanoassembly on the interaction between myristoylation and calcium that occurs above TE, it was hypothesized that this behavior can be further modulated by altering the E-domain's composition. To investigate this, two additional variants, RE', which displays increased hydrophilicity by substituting 20% of the valines with alanine, and RE", which combines increased hydrophilicity with decreased length of the E-domain (from 80 to 40 pentapeptide repeats), were engineered.Both modifications were anticipated to raise the transition temperature of the ELP domain, leading to increased TE values for RE' and RE", following the trend TE(RE) < TE(RE') < TE(RE"). The transformed DLS results, shown in FIG. 19, confirmed that these compositional changes resulted in changes to liposwitching threshold, as evidenced by the shifts in the plots to higher temperatures. This outcome is significant because it demonstrates the feasibility of manipulating the responsiveness to calcium-triggered lipid exposure by strategically modifying the phase boundaries of the disordered domain. Consequently, this strategy could be extended to other intrinsically disordered proteins that undergo phase separation under isothermal conditions, by adjusting factors such as pH, ionic strength, or specific molecular interactions.

[0073] The cellular uptake of RE fusion proteins in DU145 prostate cancer cells was assessed under varying conditions of myristoylation and calcium concentration. Confocal imaging showed that robust cytoplasmic internalization of the green fluorescent RE signal occurred only when both calcium was present and REs were myristoylated — conditions that promote nanocluster formation. In contrast, when either calcium or myristoylation was absent, uptake was markedly reduced, with the fluorescent signal remaining primarily at the cell surface or in the extracellular space.

[0074] These results indicate that RE nanocluster formation significantly enhances endocytic uptake in mammalian cells. This uptake correlates with the ability of RE fusions to undergo membrane-associated phase separation, driven by calcium binding and myristoylation. Nanocluster formation may increase avidity for membrane receptors or induce local curvature that promotes vesicle budding. The ability to toggle endocytosis through simple environmental inputs positions RE scaffolds as versatile platforms for intracellular delivery of biologies — such as enzymes, imaging agents, or therapeutic payloads. This approach offers a promising strategy for developing stimuli-responsive delivery systems for applications in targeted drug delivery and synthetic biology.

[0075] In one aspect, the invention is a method for delivering a therapeutic agent to a target population of cells, comprising the steps of: (a) providing a non-naturally occurring protein formed by a conditionally-activated lipid-binding domain fused with a stimulus- responsive polypeptide domain that is coupled to a therapeutic agent; and (b) increasing the temperature of the composition to a predetermined temperature between 37°C and 42°C to induce the formation of nanoparticles, thereby facilitating uptake of the therapeutic agent by the cells.

[0076] In another aspect, the invention is a non-naturally occurring protein, comprising: (a) a myristoyl switch; (b) an elastin-like polypeptide domain; and (c) a therapeutically active polypeptide fused to the protein.

[0077] In an additional aspect, the invention is a therapeutic composition comprising nanoparticles formed from a myristoyl switch and an elastin-like polypeptide domain fused to the myristol switch; wherein a small-molecule therapeutic agent is non-covalently encapsulated within or covalently conjugated to said nanoparticles

[0078] In a further aspect, the invention is a non-naturally occurring protein comprises of a conditionally-activated lipid-binding domain fused with a stimulus-responsive polypeptide domain that is formed into nanoclusters in the presence of calcium ions and at a temperature above a transition temperature of the stimulus-responsive polypeptide, wherein a therapeutic agent is attached to the non-naturally occurring protein.

Claims

CLAIMSWhat is claimed is:

1. A non-naturally occurring protein, comprising: a conditionally-activated lipid-binding domain that will expose a lipid moiety in response to a stimulus; and a stimulus-responsive polypeptide domain fused to the conditionally-activated lipid- binding domain.

2. The non-naturally occurring protein of claim 1, wherein the conditionally- activated lipid-binding domain is a neuronal calcium sensor switch domain.

3. The non-naturally occurring protein of claim 2, wherein the neuronal calcium sensor switch domain is selected from the group consisting of a myristoyl switch, hippocalcin, neurocalcin, visinin-like protein 1, and a guanylate cyclase activating protein.

4. The non-naturally occurring protein of claim 1, wherein the stimulus- responsive polypeptide domain is selected from the group consisting of a thermo-responsive domain, a pH responsive domain, a lipid composition responsive domain, and a ligand binding responsive domain.

5. The non-naturally occurring protein of claim 1, wherein the stimulus- responsive polypeptide domain is thermo-responsive domain.

6. The non-naturally occurring protein of claim 5, wherein the thermo-responsive domain is an elastin-like polypeptide domain.

7. The non-naturally occurring protein of claim 1, wherein the conditionally- activated lipid-binding domain is a myristoyl switch comprising SEQ ID NO: 3.

8. The non-naturally occurring protein of claim 1, wherein the stimulus- responsive polypeptide domain comprises a plurality of repeats of SEQ ID NO: 1.

9. The non-naturally occurring protein of claim 1, wherein the non-naturally occurring protein comprises SEQ ID NO: 4.

10. The non-naturally occurring protein of claim 1, further comprising a therapeutic agent attached to one of the stimulus-responsive polypeptide domain and the stimulus-responsive polypeptide domain.

11. A method of forming a non-naturally occurring protein, comprising the steps of: a conditionally-activated lipid-binding domain that will expose a lipid moiety in response to a stimulus; andfusing a stimulus-responsive polypeptide domain to the conditionally-activated lipid- binding domain.

12. The method of claim 11, wherein the stimulus is selected from the group consisting of a calcium ion concentration, a pH, a specific lipid species, and a specific molecular ligand.

13. The method of claim 12, wherein the conditionally-activated lipid-binding domain is a myristol switch and the stimulus-responsive polypeptide domain is an elastin-like polypeptide domain.

14. The method of claim 11, further comprising the step of modifying at least one of hydrophilicity or length of the stimulus-responsive polypeptide domain to achieve a desired transition temperature.

15. The method of claim 11, wherein the stimulus-responsive polypeptide domain comprises between 40 and 120 repeats of a responsive peptide unit.

Images