One-pot method for recombinant geranylgeranylated protein production

Abstract

A method for the one-pot production of geranylgeranylated proteins in Escherichia coli. E. coli is engineered to express geranylgeranyl pyrophosphate synthase, an enzyme that catalyzes production of geranylgeranyl pyrophosphate, and a target protein. By co-expressing GGS with a geranylgeranyltransferase, efficient geranylgeranylation of model protein substrates was achieved, including intrinsically disordered elastin-like polypeptides (ELPs) and the globular protein mCherry. This modification affects the phase-separation and nano-assembly of ELPs, and lipid bilayer engagement of mCherry. The method offers a scalable, versatile and cost-effective strategy for producing geranylgeranylated proteins, which will facilitate future biochemical and biophysical studies of lipid-modified proteins, and their applications in drug delivery and tissue engineering.

Description

TITLEONE-POT METHOD FOR RECOMBINANT GERANYLGERANYLATED PROTEIN PRODUCTIONSTATEMENT REGARDING FEDERALLY SPONSORED RESEARCH AND DEVELOPMENT

[0001] This invention was made with government support under Grant No. 2105193 awarded by the National Science Foundation and Grant No. 5R35GM142899-04 awarded by the National Institutes of Health. The government has certain rights in the invention INCORPORATION BY REFERENCE

[0002] The XML plain text file entitled 156P713.xml dated December 4, 2025 and having a size of 18,517 bytes containing a Sequence Listing is hereby incorporated by reference.BACKGROUND OF THE INVENTION1. FIELD OF THE INVENTION

[0003] The present disclosure relates to post-translational protein modifications and, more particularly, an approach for one-pot production of geranylgeranylated proteins in Escherichia coli.2. DESCRIPTION OF THE RELATED ART

[0004] Geranylgeranylation is a post-translational modification (PTM) involving the addition of a geranylgeranyl (20-carbon) isoprenoid group to proteins via a thioether linkage to a cysteine residue within the CaaX box motif near the C-terminus of protein substrates. The substrates for this PTM include many small GTPases, including members of the Rho, Rab and Rap families, and y subunits of hetero trimeric G proteins. Given its non-polar nature, the geranylgeranylation is critical for membrane localization and function of the substrate proteins, especially their interactions with effector proteins and downstream cell signaling transduction. Therefore, dysregulation of geranylgeranylation and its biosynthesis has been linked to a wide range of human diseases, including cancer, type II diabetes, liver disorders, neurodegeneration, and others. Despite its significance, studying the role and mechanisms of this PTM, and elucidating structure-function relationships in geranylgeranylated proteins (GG-proteins), has been hindered by the technical challenges of producing these lipid- modified proteins.

[0005] Like other lipidated proteins, current methods for generating GG-proteins are labor-intensive, costly, and technically challenging. The primary challenge is that Escherichia coli, the workhorse organism for protein expression, lacks the enzymatic122656343. v1 -12 / 4 / 25machinery required for GG-protein production. While eukaryotic or cell-free expression systems can overcome this limitation, they are expensive, difficult to scale, and often produce heterogeneously modified proteins in low yields. Chemoenzymatic approaches offer alternative routes but typically require the use of surfactants, organic solvents, and extensive reaction optimization and refolding protocols,

[0006] To address these limitations, two primary strategies have been developed for geranylgeranyl modification of proteins. The first strategy involves semi-synthetic approaches, in which recombinantly expressed proteins lacking the lipidated domain are conjugated to synthetic geranylgeranylated peptides using expressed protein ligation (EPL) or chemoenzymatic coupling. Despite the complementary strengths of these approaches, both require significant reaction optimization and the use of organic solvents, detergents, denaturants, and an oxygen-free environment to balance the solubility and activity of proteins, reagents, and catalysts on a case-by-case basis. The high cost and limited stability of geranylgeranyl-based lipids further complicate efforts to efficiently produce GG-proteins at scale. Accordingly, there is a need in the art for an approach that can address these obstacles to facilitate a deeper exploration of the biophysical implications of geranylgeranylation and enable the development of GG-modified proteins for advanced biomedical technologies and engineered materials.BRIEF SUMMARY OF THE INVENTION

[0007] The present invention provides a one-pot method for producing GG-proteins in E. coli. To achieve this, endogenously produced isopentenyl diphosphate (IPP) and farnesyl pyrophosphate (FPP) - key intermediates in the biosynthesis of geranylgeranyl pyrophosphate (GGPP) - are employed. E. coli was engineered to express geranylgeranyl pyrophosphate synthase (GGS), the enzyme that catalyzes GGPP formation, and geranylgeranyltransferase (GGT), which transfers GGPP to the CaaX motif of substrate proteins. The approach of the present invention was tested using model proteins fused to a CaaX motif: intrinsically disordered elastin-like polypeptides (ELPs) and the globular protein mCherry. The one -pot method of the present invention demonstrated robust performance, producing GG-modified proteins at yields sufficient for biophysical and biomaterials studies. More specifically, the engineered strains according to the present invention achieved efficient GG-modification (90-95%) without compromising the native folding or function of the proteins. The approach of the present invention is highly compatible with standard cultivation technologies and can be easily adopted by the scientific community, thereby broadening access to GG-modified proteins.222656343. v1 -12 / 4 / 25

[0008] An exemplary method of producing a geranylgeranylated protein according to the present invention involves the step of modifying an amount of Escherichia coli bacteria to express a geranylgeranyl pyrophosphate synthase, to express a geranylgeranyltransferase, and to express a target protein having a cysteine residue within a CaaX box motif near a C- terminus of the target protein, then inducing protein expression in the amount of Escherichia coli bacteria to produce the geranylgeranylated target protein, and then isolating the geranylgeranylated target protein from the amount of Escherichia coli bacteria. The geranylgeranyl pyrophosphate synthase may comprise Deinococcus radiodurans geranylgeranyl pyrophosphate synthase. The geranylgeranyl pyrophosphate synthase may comprise SEQ ID NO: 11. The geranylgeranyltransferase may comprise Rattus norvegicus geranylgeranyltransferase. The geranylgeranyltransferase may comprise SEQ ID NO: 9 and SEQ ID NO: 10. The target protein may comprise an elastin-like polypeptide. The elastin-like polypeptide may have a sequence selected from the group consisting of SEQ ID NO: 12, SEQ ID NO: 13, and SEQ ID NO: 14. The target protein may comprise a globular protein. The globular protein may comprise SEQ. ID NO. 15. The target protein may comprise a GTPase. The GTPase may be selected from the group consisting of RhoA and Rap IB. The GTPase may be SEQ ID NO: 16. The GTPase may be SEQ ID NO: 17. The method may comprise the step of incubating the amount of Escherichia coli bacteria while inducing protein expression. In another embodiment, the present invention is a geranylgeranylated protein formed by the present method.BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWING(S)

[0009] The present invention will be more fully understood and appreciated by reading the following Detailed Description in conjunction with the accompanying drawings, in which:

[0010] FIG. 1 is a series of graphs of the characterization of D. radiodurans Geranylgeranyl Pyrophosphate Synthase (GGS). (a) Sequence alignment of D. radiodurans CrtE with human and E. coli prenyl synthase, highlighting similarities in the first aspartic- rich motif (FARM, red) and the chain length-determining region (CLD, yellow), (b) CD spectra of CrtE, revealing prominent a-helical content after deconvolution, (c) Representative structure of CrtE, from MD simulations, highlighting its alpha-helical rich structure, with the FARM and CLD regions highlighted, (d) Schematic of the reaction between IPP and FPP to form GGPP and pyrophosphate, with quantification of the released phosphate using Bimol Green reagent. Combining FPP and IPP in the presence of CrtE results in a significant increase in absorbance at 620 nm, indicating the release of phosphate, (e) RP-HPLC322656343. v1 -12 / 4 / 25chromatogram of reactions between FPP and IPP in the absence and presence of CrtE, showing a decrease in FPP peak intensity (black arrow) and the appearance of a new peak (red arrow) corresponding to GGPP. The asterisk denotes an impurity in FPP. (f) ESI-MS spectra of peaks labeled with black and red arrows, consistent with the molecular weight of FPP and GGPP, with mass shift indicating the addition of a C5 isoprenoid group. Error bars in b and d are std. dev. of n=3. Statistical significance in d was determined using one-way ANOVA with Dunnett's post hoc test, p < 0.05 (*), p < 0.01 (**), p < 0.001 (***).

[0011] FIG. 2 is a schematic of the chain elongation reaction catalyzed by mediumchain (E)-prenyl diphosphate synthases. The schematic illustrates the stepwise addition of isopentenyl diphosphate (IPP, black) to farnesyl pyrophosphate (FPP, blue), resulting in the elongation of the chain by one isoprene unit per step, with geranylgeranyl pyrophosphate (GGPP) as the final product. FARM: First aspartic-rich motif; CLD: Chain-length determining region.

[0012] FIG. 3 is a characterization of recombinantly expressed N-terminal His-tagged putative GGS. (a) SDS-PAGE analysis of elutions following IMAC purification, visualized using stain-free technology, confirming the presence of the purified protein, (b) MALDI- TOF-MS of purified proteins, with dotted and dashed lines indicating [M+2H]2+and [M+H]+molecular ion peaks, respectively.

[0013] FIG. 4 is a simulated quaternary structure of D. radiodurans GGS and its alignment with homologous synthases, (a) Homodimeric structure of GGS with chain A shown in cyan and chain B in red. (b) Aligned structures of homologous synthases (Table 7) with GGPPs, colored according to their Qres values. The GGS chain is highlighted in cyan to clearly demonstrate the structural overlap with the homologous enzymes.

[0014] FIG. 5 is a graph of the root mean square deviation (RMSD) of the two GGS production runs. A 10-point rolling average is applied to smooth fluctuations over production time (1.1 ps).

[0015] FIG. 6 is (a) RMSF plot for chain A and B of the GGS during the production simulation, (b) The RMSF projected on the structure of the protein.

[0016] FIG. 7 is (a) a histogram of hydrogen bonds for the GGS production run. (b) H-bond between Tyr25 and Glul41 in both chains. Chain A is represented in cyan and chain B is red.

[0017] FIG. 8 is SDS-PAGE analysis of protein isoforms obtained from expression of a model ELP in (±GGS / ±GGT) strains. Consistent with HPLC results, a significant amount of ELP dimers (apparent Mw ~ 40 kDa) are observed in -GGT strains. The addition of beta-422656343. v1 -12 / 4 / 25mercaptoethanol (PME) to the gel loading buffer leads to the disappearance of these bands, indicating disulfide bond formation between the chains.

[0018] FIG. 9 is a series of graphs of the plasmid architecture and analysis of geranylgeranylated model proteins, (a) Schematic of the plasmid architecture used in this study, featuring two orthogonal plasmids for co-expression of the protein of interest with GGS and the two subunits of GGT. Control plasmids are utilized to evaluate the effects of GGS and GGT in a 2x2 factorial experiment, (b) RP-HPLC chromatogram of a model protein co-expressed in four strains (± GGS / ± GGT). In the absence of GGT, only unmodified protein and its disulfide-bonded dimer are detected. Expression of GGT results in the modification of the protein with a hydrophobic motif, depending on whether GGS is coexpressed. (c) MALDI-TOF spectra of unmodified and hydrophobically modified protein isoforms, showing differences in molecular weight that suggest modification with one farnesyl group in the (-GGS / +GGT) strain and one geranylgeranyl group in the (+GGS / +GGT) strain, (d) Mass spectra of trypsin-digested proteins confirm the proposed modifications and establish the site of lipidation to the CaaX box sequence.Carboxymethylacetamide (Cam) is used for alkylating free thiols of the unmodified constructs to prevent dimerization.

[0019] FIG. 10 a series of graphs of MS / MS analysis of prenylated C-terminal peptide fragments, (a) MS / MS spectrum of Fr-modified peptide fragment obtained using collision-induced dissociation (CID); (b) MS / MS spectrum of Fr-modified peptide fragment using higher-energy collisional dissociation (HCD); (c) MS / MS spectrum of GG-modified peptide fragment using CID; (d) MS / MS spectrum of GG-modified peptide fragment using HCD.

[0020] FIG. 11 is a series of graphs of RP-HPLC traces of purified ELP isoforms used in this study, (a) ELPV40; (b) ELP(V8 / A2)go; (c) ELPA40. In each panel, the unmodified isoform is represented by a dotted line, the Fr-modified isoform by a dashed line, and the GG-modified isoform by a solid line. Arrows indicate the retention times corresponding to each isoform.

[0021] FIG. 12 is a series of graphs of MALDI-TOF-MS spectra of the ELP isoforms in this study, (a) ELPV40; (b) ELP(V8 / A2)go; (c) ELPA40. Vertical dotted lines indicate the theoretical molecular weight ([M + H]+) for each construct.

[0022] FIG. 13 is a study of the biophysical impact of GG-modification on model disordered proteins, (a) Representative temperature-composition phase diagrams for unmodified, Fr-, and GG-modified ELPs. Prenylation alters the phase boundaries, with the522656343. v1 -12 / 4 / 25extent of this alteration dependent on the nature of the attached lipid, (b) Variabletemperature DLS confirms that GG-modified ELPs form stable nanoparticles across all tested variants, with a sharp micelle-to-coacervate transition at higher temperatures. Blue, white, and red regions represent unimers, micelles, and coacervates, respectively, (c) Cryo-TEM visualization of GG-modified ELP nanoparticles, with the inset displaying the nanoparticle diameter histogram.

[0023] FIG. 14 is a graph of the representative turbidimetry plots for geranylgeranylated and unmodified ELPs. Geranylgeranylation increases the propensity of ELPs to phase-separate by lowering their transition temperature (Tt). Protein concentration = 50 pM in PBS..

[0024] FIG. 15 is a pair of temperature-composition plots for unmodified and prenylated ELPs. (a) ELPV40: The transition temperature shows an inverse relationship with protein concentration. While the behavior of unmodified (dotted line) and GG-modified (solid line) ELPs is adequately described by linear regression, Fr-modified ELPs exhibit intermediate behavior, fitting a sigmoidal model (dashed line), (b) ELPA40: Due to the hydrophilicity of these constructs, transition temperatures of unmodified and Fr-modified isoforms were outside the experimental range at concentrations below 50 pM. Data for GG- modified isoforms are fitted to a linear model. Error bars represent the standard deviation of three measurements.

[0025] FIG. 16 is a series of graphs of the DLS intensity-size distributions of unmodified and geranylgeranylated ELP isoforms, (a) ELPV40; (b) ELP(V8 / A2)go; (c) ELPA40. In each case, GG-modification induces the self-assembly of ELPs into nanoparticles, as evidenced by increase in the diameter of the major peaks.

[0026] FIG. 17 is a study of the Cryo-TEM visualization of nanoparticles formed by ELP(V8 / A2)go-GG. (a) Representative cryo-TEM image of nanoparticles from samples vitrified at 25°C. (b) Histogram of nanoparticle diameter distributions.

[0027] FIG. 18 is a study of one-pot GG-modification of mCherry and Visualization of its interaction with model membranes, (a) RP-HPLC traces of mCherry expressed in ±GGS / ±GGT strains, demonstrating the applicability of the method to globular proteins, (b- d) Confocal microscopy images showing interactions between unmodified and GG-modified mCherry with model GUVs. Geranylgeranylation enhances mCherry's interaction with GUVs, promoting colocalization into lipid-disordered (Id) regions.

[0028] FIG. 19 is a graph of a MALDI-TOF-MS analysis of mCherry isoforms. Vertical dotted lines denote the theoretical molecular weights for each construct.622656343. v1 -12 / 4 / 25

[0029] FIG. 20 is a graph of RP-HPLC traces of control 6xHis-mCherry constructs. a)Retention time of 6xHis-mCherry-AVLL expressed in (+GGS / +GGT) strains is identical to that of 6xHis-mCherry-CVLL expressed in (-GGS / +GGT) strains (b), indicating that geranylgeranylation requires the presence of cysteine in the CaaX box motif.

[0030] FIG. 21 is a pair of graphs of LC-MS characterization of the C-terminal fragments of mCherry-GG. (a) Extracted Ion Chromatogram (XIC) of trypsinized mCherry expressed in (+GGS / +GGT) strains shows a late-eluting peak at 51.19 minutes, with a molecular weight consistent with the geranylgeranylated 'GCVLL' peptide, (b) MS spectrum of this species displays mass accuracy within 0.7 ppm of the theoretical mass for the GCGGVLL peptide.

[0031] FIG. 22 is a pair of graphs of the excitation and emission spectra of unlipidated and GG-modified mCherry. (a) Excitation and emission spectra of unmodified isoform (mCherry-CVLL). (b) Excitation and emission spectra of geranylgeranylated isoform (mcherry-GG). Geranylgeranylation does not alter the fluorescence of mCherry.

[0032] FIG. 23 is a series of graphs of LC-MS analysis of the reduced and oxidized forms of the lipidated C-terminal peptide fragment, (a) Extracted Ion Chromatogram (XIC) of trypsin-digested ELP(V8 / A2)80 expressed in (+GGS / +GGT) strains shows a major peak at 52 min (red trace), corresponding to the geranylgeranylated GCVLL peptide, and two minor peaks at 49.3 and 49.9 min (maroon trace), which correspond to the oxidized (sulfoxide) form of the GG-modified peptide, (b) MS spectrum of the 52-min peak confirms the expected mass of the geranylgeranylated peptide, (c, d) MS spectra of the earlier-eluting peaks at 49.3 and 49.9 min reveal a +16 Da mass shift, consistent with sulfoxide formation. This oxidation introduces a new chiral center, resulting in the formation of two diastereomers, which are separated as distinct peaks after trypsin digestion. However, in the context of the full-length protein, this modification likely results in a single peak due to reduced chromatographic resolution (see peaks marked with asterisks in Figure 2b for representative examples).

[0033] FIG. 24 is a graph of the overlaid mass spectra of the reduced and oxidized forms of the famesylated (Fr)-modified C-terminal lipid fragments. A +16 Da mass shift is consistent with sulfoxide formation, as seen in the GG-modified peptide (FIG. 23).

[0034] FIG. 25 is a series of graphs of the experimental workflow and analysis of recombinant production of GG-modified RhoA. (a) Schematic representation of the plasmids used for co-expression of RhoA with or without GGT. In both cases, RhoA is co-expressed with GGS. Since E. coli lacks endogenous GGT, a single plasmid is used for the -GGT condition. After expression, the soluble fraction was purified using immobilized metal722656343. v1 -12 / 4 / 25affinity chromatography and analyzed by RP-HPLC. (b) RP-HPLC chromatogram comparing RhoA isoforms expressed with (solid line) and without (dashed line) GGT. In the absence of GGT, only unmodified RhoA is detected. Co-expression with GGT yields 96% geranylgeranylation (73.5% reduced, 22.5% sulfoxide). See Figures 27 and 29b for LC-MS analysis of each isoform.

[0035] FIG. 26 is a series of graphs of the experimental workflow and analysis of recombinant production of GG-modified RaplB. (a) Schematic representation of the plasmids used for co-expression of RaplB with or without GGT. In both cases, RaplB is coexpressed with GGS. Since E. coli lacks endogenous GGT, a single plasmid is used for the - GGT condition. After expression, inclusion bodies were solubilized in 6M guanidine hydrochloride, purified by IMAC under denaturing condition, and analyzed by RP-HPLC analysis, (b) RP-HPLC chromatogram comparing RaplB isoforms expressed with (solid line) and without (dashed line) GGT. In the absence of GGT, only unmodified RaplB is detected. Co-expression with GGT yields 87% geranylgeranylation (82% reduced, 5% sulfoxide). See Figures 28 and 29a for LC-MS analysis of each isoform.

[0036] FIG. 27 is a series of graphs of the LC-MS analysis of the unmodified and geranylgeranylated C-terminal peptide fragment of RhoA. Extracted Ion Chromatogram (XIC) and mass spectra of trypsinized RhoA expressed in (a) -GGT (+GGS / -GGT) and (b) +GGT (+GGS / +GGT) strains. In the absence of GGT (panel a), only the unmodified C- terminal peptide fragment is detected at 23.81 min (red trace), while no GG-modified peptide is observed (blue trace). With GGT (panel b), a strong signal for the GG-modified fragment is detected (51.36 min), with minimal unmodified peptide present, consistent with RP-HPLC results (Figure S20). Free thiols in the unmodified constructs were alkylated using carboxymethylacetamide (Cam) to prevent dimerization. Peaks marked with (*) in the XIC trace were identified as noise based on the isotopic pattern analysis.

[0037] FIG. 28 is a series of graphs of the LC-MS analysis of the unmodified and geranylgeranylated C-terminal peptide fragment of RaplB. Extracted Ion Chromatogram (XIC) and mass spectra of trypsinized RaplB expressed in (a) -GGT (+GGS / -GGT) and (b) +GGT (+GGS / +GGT) strains. In the absence of GGT (panel a), only the unmodified C- terminal peptide fragment is detected at 21.01 min (red trace), with no GG-modified peptide observed (blue trace). Consistent with RP-HPLC results (Figure 26), under +GGT conditions (panel b), both unmodified (21.05 min) and GG-modified (48.52 min) peptides are detected. Free thiols in the unmodified constructs were alkylated with carboxymethylacetamide (Cam)822656343. v1 -12 / 4 / 25to prevent dimerization. Peaks marked with (*) in the XIC trace were identified as noise based on isotopic pattern analysis.

[0038] FIG. 29 is a series of graphs of the LC-MS analysis of the oxidized forms of the GG-modified C-terminal peptide fragments of Rap IB and RhoA. Chemical structures, extracted ion chromatograms, and mass spectra of C-terminal peptide fragments from model proteins expressed in (+GGS / +GGT) strain are shown for (a) Rap IB and (b) RhoA. In both cases, oxidation of the thioether bond in the GG-modified peptide results in two peaks eluting approximately 2-3 minutes earlier than the reduced GG-modified peak. The MS spectra confirm a +16 Da mass shift compared to the reduced form, consistent with sulfoxide formation. This oxidation introduces a new chiral center, resulting in the separation of two diastereomers as distinct peaks after trypsin digestion.DETAILED DESCRIPTION OF THE INVENTION

[0039] Referring to the figures, wherein like numerals refer to like parts throughout, there is seen in FIG. 1, an analysis of D. radiodurans Geranylgeranyl Pyrophosphate Synthase (GGS) for use with a one-pot method for producing GG-proteins in E. coli. To facilitate heterologous expression in E. coli, a prokaryotic GGS enzyme that would provide a better compatibility with bacterial expression systems was identified. The enzyme identified was from Deinococcus radiodurans, an extremophilic bacterium known for producing a diverse array of carotenoids. The sequence of this protein (CrtE) showed a high degree of homology with the prenyl synthase family (Table 1 below), including a known FPP synthase (IspA in E. coli) and a human GGS.

[0040] Interestingly, the D. radiodurans CrtE shares features with both eukaryotic GGPPS (the DDXXDD “first aspartic-rich motif’) and prokaryotic enzymes (hydrophobic aromatic residues upstream of the FARM), which typically favor synthesis of shorter prenyl chains like FPP due to steric hindrance (FIG. 1, FIG. 2). Given these mixed features and the lack of direct characterization, we prioritized in vitro studies to confirm the enzyme's capacity to synthesize GGPP from £. coil's endogenous FPP and IPP.922656343. v1 -12 / 4 / 25

[0041] First, a codon-optimized gene for the D. radiodurans GGS, incorporating an N-terminal his-tag for ease of purification, was synthesized. The protein was expressed and purified as detailed below as seen in Table 2 below and FIG. 3.

[0042] Circular Dichroism (CD) spectroscopy was used to assess the secondary structure of the purified enzyme, revealing a predominantly alpha-helical content (FIG. lb). This is in line with the structure of other members of the prenyl synthase family and with predictions of AlphaFold3 and MD simulations (FIG. 1c, FIGS. 4-7). Next, we validated that the purified protein catalyzes formation of GGPP, following the reaction scheme depicted in FIG. Id. Reaction progress was monitored using two methods: a colorimetric end-point assay to detect the release of pyrophosphate (PPi), FIG. Id; and reverse-phase high-performance liquid chromatography (RP-HPLC) with mass spectrometry to directly detect GGPP formation (FIG. le,f). A significant increase in absorbance at 620 nm was observed, indicating the release of phosphate, only when all three components of the reaction (FPP, IPP and CrtE) were present. RP-HPLC further confirmed the formation of GGPP, as indicated by the appearance of a new peak corresponding to GGPP, benchmarked against chemically synthesized GGPP (bottom trace; FIG. le). Finally, electrospray ionization mass spectrometry (ESI-MS) further verified the reaction product, revealing a mass increase of 68.09 Da, consistent with the addition of a prenyl group. The observed mass of 449.1844 Da aligns with the theoretical mass for [GGPP-H]'1(449.1858 Da). Together, these experiments validate that D. radiodurans CrtE is capable of synthesizing GGPP from FPP and IPP, providing a solid foundation for its use as the GGS enzyme in our one-pot method for producing geranylgeranylated proteins.1022656343. v1 -12 / 4 / 25

[0043] Building on the validation of GGS, a minimalistic yet efficient system for geranylgeranylation in E. coli was established. This system necessitated the co-expression of two critical components, GGS and GGT. A type-I GGT enzyme from Rattus norvegicus was used due to its broad substrate scope and ability to modify proteins with a CaaX box motif at their C terminus. Since GGT is a heterodimer, two orthogonal plasmids were used to coexpress four polypeptide chains: the a and 0 subunits of GGT, GGS, and the protein substrate fused to the model CaaX box motif, CVLL (FIG. 9a). To address the aggregation tendency of the a subunit, we employed a translationally coupled expression system, while an orthogonal bicistronic plasmid was used to express GGS and the substrate in commonly used E. coli BL21(DE3) strains. Control vectors lacking either GGT or GGS were also designed to systematically evaluate the efficiency of GG-protein production (Table 3). Pairwise combinations of these vectors enabled systematic assessment of our one-pot GG-protein production system (Table 4).1122656343. v1 -12 / 4 / 25

[0044] As the first model protein, an elastin-like polypeptide (ELP), an artificial intrinsically disordered protein derived from the consensus sequence of tropoelastin. ELP’s disordered and uniquely hydrophobic nature allows facile isolation by treating the cells with isopropanol, which results in selective partitioning of the ELP from the rest of the proteome. To evaluate the efficacy of the engineered E. coli strains in facilitating protein geranylgeranylation, the model protein from each strain was isolated and its isoforms analyzed using RP-HPLC (FIG. 9b) and matrix-assisted laser desorption ionization-time of flight-mass spectrometry (MALDI-TOF-MS, FIG. 9c). ELP expressed in the absence of GGT, i.e. in -GGS / -GGT and +GGS / -GGT strains, had similar retention times on the HPLC traces, showing two prominent peaks corresponding to the unmodified protein (tR = 29.0 ± 0.1 min) and the disulfide-bonded dimer (tR = 29.4 ± 0.1 min), FIG. 8. MALDI-TOF-MS analysis confirmed these findings, as observed m / z for these constructs closely matched the molecular weight of the unmodified ELP. Introduction of GGT alone (-GGS / ±GGT strain) led to the appearance of a new species with an elution time of 33.9 ± 0.1 min (blue trace, FIG. 9b), that was confirmed to be a farnesylated (Fr) ELP product based on the results of MALDI-TOF-MS (blue line, FIG. 9c). Given that GGT can accept FPP in the absence of GGPP, this reaction conditions leverages E.coli's endogenous FPP supply, albeit with low yield as only 50% of the expressed ELPs were farnesylated (FIG. 9b). Importantly, the strain co-expressing both GGS and GGT (±GGS / ±GGT) produced a distinct peak at 36.5 min on the RP-HPLC chromatogram (red trace, FIG. 9b), that was confirmed to be GG-protein by MALDI-TOF (±272.5 peak, FIG. 9c). Under these conditions, 90% of the expressed ELPs were modified with GG, and no sign of farnesyl-modified ELP was observed on the HPLC trace (FIG. 9b). Finally, trypsin digest and subsequent LC-MS / MS analyses provided robust evidence that the modifications — whether famesyl or geranylgeranyl — were specifically localized to the cysteine residue of the CaaX box (FIG. 9d, FIG. 10, Table 5). Additionally, LC-MS analysis detected a small fraction of oxidized (sulfoxide) peptide, consistent with aerobic oxidation of the thioether moiety (Figures 23 and 24). Collectively, these results indicate that the engineered E. coli system produces high yields of GG-protein.1222656343. v1 -12 / 4 / 25

[0045] To showcase the versatility of our one-pot method for GG-p rotein production, it was used to rapidly generate a diverse library of prenylated ELPs (FIG. 11, 12, Table 6, 7).1322656343. v1 -12 / 4 / 251422656343. v1 -12 / 4 / 251522656343. v1 -12 / 4 / 25

[0046] In contrast to traditional chemoenzymatic approaches, which require laborious reaction optimization for each protein, our genetically encoded process simplifies the production by merely controlling the producer strain, making it highly accessible to standard cultivation technologies. This streamlined process accelerates the production of modified proteins at scale. Using our engineered strains, we efficiently produced nine proteins, encompassing three ELP variants with different lengths and hydrophobicity, each in unmodified, famesyl-modified (-Fr), and geranylgeranyl-modified (-GG) isoforms. This capability allowed us to systematically examine how lipid modifications influence the phase behavior of intrinsically disordered ELPs, revealing significant effects of prenylation on their biophysical properties.

[0047] Both Fr- and GG-modification increased the propensity of ELPs to phaseseparate by lowering their phase boundaries compared to unmodified isoforms, with the degree of reduction dependent on the physicochemistry of the lipid (i.e., its length) (FIG. 13, FIG. 14, FIG. 15). The transition temperature versus the natural log of concentration for both unmodified and GG-modified ELPs followed a linear model, with GG-modified constructs exhibiting a significantly lower slope and intercept compared to unmodified proteins (Table 8, Table 9).

[0048] In contrast, Fr-modified proteins followed a sigmoidal model, where their behavior at high concentrations was similar to GG-modified proteins, but at lower concentrations, they displayed intermediate behavior between GG-modified and unmodified isoforms. These findings are important for two reasons: First, they demonstrate that prenylation can modulate the phase behavior of proteins, adding to an emerging body of literature showing that lipidation can regulate the phase separation and material properties of protein condensates. Given the prevalence of prenylation, this offers new insights into how such modifications affect protein behavior in cellular environments. Second, the distinct behavior of GG-modified ELPs, with reduced variations in transition temperature across a broad concentration range, suggests that the increased hydrophobicity of the GG lipid leads to a more stable and pronounced alteration in the quaternary organization of the ELP chains.

[0049] Consistent with the turbidimetry, dynamic light scattering (DLS) revealed that geranylgeranylation leads to the formation of stable, thermo-responsive nanoparticles across all tested ELP variants (FIG. 13b, FIG. 16). The size of these GG-modified nanoparticles remained consistent, with hydrodynamic radii of X ± Y nm, regardless of the ELP composition or length — a 40-mer with Valine, a 40-mer with Alanine, or a mixed 80-mer with 80% Valine and 20% Alanine as guest residues, Table 10. This behavior was further1622656343. v1 -12 / 4 / 25supported by cryo-TEM (FIG. 13 c, FIG. 17), which confirmed the formation of spherical nanoparticles of similar dimensions. Importantly, while the nanoparticle size remained independent of ELP composition, all GG-modified nanoparticles underwent a sharp increase in size at elevated temperatures, consistent with coacervation. This micelle-to-coacervate transition temperature was tunable by altering the composition of ELPs. This decoupling of nanoparticle size from the transition temperature represents a significant advantage over systems modified with saturated fatty acids, where changes in ELP composition not only alter their LCST but also significantly impact the size and shape of the assemblies. The ability to maintain a stable nanoparticle size while fine-tuning the transition temperature improves the programmability of GG-modified constructs, enhancing their potential for drug delivery and tissue engineering applications.

[0050] To demonstrate the generalizability of our recombinant method for protein lipidation, it was applied to a globular fluorescent protein, monomeric red fluorescent protein (mCherry). Our mCherry construct included an N-terminal His-tag for purification and a C- terminal CVLL peptide for lipidation. Consistent with results obtained with ELPs, no lipidation was observed in the absence of GGT. When GGT was present, the attached lipid was determined by the co-expression of GGS. Without GGS, the expressed protein was primarily famesylated; with GGS, more than 95% was geranylgeranylated FIG. 18a,b, FIG. 19). A control variant lacking the cysteine in the CaaX box (C245A) was not geranylgeranylated, confirming the specificity of the modification to this site (FIG. 20), as further verified by LC-MS characterization of mCherry-GG (FIG. 21). Notably, mCherry-GG remained soluble and did not strongly associate with E. coli membrane fractions, and its fluorescence spectrum was indistinguishable from that of the unlipidated construct (FIG. 22), which confirms that recombinant geranylgeranylation did not impair the protein's folding or function. This result is significant because conventional methods for geranylgeranylation often require surfactants or denaturants, which necessitate extensive downstream processing to remove these agents and refold the proteins, often with uncertain recovery of function.

[0051] Further analysis revealed that geranylgeranylation enhanced mCherry's ability to associate with lipid bilayers, with the anchored mCherry showing a preference for lipid- disordered domains. Incubation of unmodified mCherry with giant unilamellar vesicles (GUVs) prepared from 1 ,2-dioleoyl-sn-glycero-3 -phosphocholine (18:1 (A9-Cis) PC, DOPC), did not result in substantial interaction between the protein and the GUVs (FIG. 18d). In contrast, mCherry-GG readily associated with DOPC GUVs, exemplified by the colocalization between the fluorescence channel (red) and the GUVs (FIG. 18e) with no1722656343. v1 -12 / 4 / 25detectable changes in GUV morphology. When GUVs were prepared with a mixture of saturated and unsaturated phospholipids (e.g., DPPC and DOPC), the mCherry-GG interacted with GUVs but preferentially colocalized with the regions containing unsaturated lipid (DOPC) (FIG. 18f). This behavior aligns with the physical properties of the geranylgeranyl lipid, which contains unsaturated bonds and favors interactions with lipid-disordered regions.

[0052] Finally, to demonstrate the applicability of our approach for producing natively geranylgeranylated proteins, we cloned two small GTPases, RhoA (a prototypical member of the Rho family involved in cytoskeletal dynamics) and Rap IB (a Ras-related protein that counteracts oncogenic Ras mutants). As shown in Figures 25 and 26, our strains efficiently geranylgeranylated both proteins, achieving >95% modification for RhoA and 87% for Rap IB as determined by RP-HPLC. These findings are particularly noteworthy because, consistent with previous reports, even the non-lipidated recombinant Rap IB accumulated in inclusion bodies. Achieving high modification efficiency under these suboptimal conditions highlights the robustness and broad applicability of the approach.

[0053] LC-MS analysis confirmed modification of the C-terminal peptide fragments (Figures 27 and 28) and identified oxidation at the thioether bond (Figure 29), affecting 22% of RhoA and 5% of Rap IB as determined by HPLC. Thioether oxidation, similar to methionine oxidation, is well-documented and has been observed in vitro for prenylated peptides. While the precise source of this oxidation remains unclear in our system, we attribute it primarily to air exposure during purification. To mitigate this, the purification protocol may be optimized, such as degassing buffers, to minimize oxidation. Additionally, to address the presence of residual unmodified protein, hydrophobic interaction chromatography or surfactant-based phase separation may be used to selectively isolate lipidated proteins from their unmodified counterparts, providing scalable solutions for producing high-purity geranylgeranylated proteins suitable for biochemical and structural studies.

[0054] The challenging synthesis of lipidated proteins remains a key barrier to understanding how lipidation affects protein structure and function at the molecular level, despite the well-established role of this PTM in regulating cellular processes. This challenge also limits lipidation's potential as a tool for engineering protein behavior and designing functional materials. Here, these obstacles were addressed with a robust, user-friendly system for producing GG-modified proteins in E. coli. The one-pot method of the present invention enables efficient geranylgeranylation of proteins without compromising their structural integrity or function. The ability to geranylgeranylate a structurally demanding protein like mCherry, while preserving its function, highlights the robustness and versatility of our1822656343. v1 -12 / 4 / 25approach. This method offers significant advantages over conventional lipidation techniques by eliminating the need for harsh chemicals and complex refolding steps. The isolated yields for geranylgeranylated proteins in this study was 5 mg / L of culture (except for ELPA40-GG which yielded ~ 1 mg / L), even without optimization of expression conditions, codon usage, or plasmid design. These results highlight the baseline efficiency of our system, and we anticipate that targeted refinements — such as tuning induction parameters or enhancing translational efficiency — could substantially improve the production yield of GG-proteins.

[0055] The complete genetic encoding of our platform unlocks exciting avenues for future research such as the evolution of E. coli strains capable of synthesizing and transferring non-natural isoprenoid analogues — lipid molecules not typically found in nature. The unique physiochemistry of these non-canonical lipids can then be exploited to precisely control lipidated protein behavior, such as membrane affinity, subcellular localization, and protein-protein interactions. Additionally, many prenylated proteins undergo further post- translational modifications after prenylation — such as endoproteolytic processing of the CaaX motif, followed by carboxyl methylation — before reaching their mature, functional forms. Recognizing the critical role of these modifications in proper protein function and localization, we are engineering strains to support these processing steps within a prokaryotic system, enabling efficient production of fully mature, post-translationally modified proteins.

[0056] Ongoing research may investigate how the unique physicochemical properties of GG influence the structure and function of biologically active proteins. By altering pharmacokinetics, internalization, and intracellular distribution of lipid-modified proteins, these lipids could significantly impact protein therapeutics. Overall, the recombinant platform of the present invention provides a reliable and scalable approach for producing prenylated proteins, overcoming previous production limitations and opening new opportunities for lipo- engineering in synthetic biology, biomedical engineering and materials science.

[0057] EXAMPLE

[0058] Materials

[0059] The chemically competent NEb5alpha, BL21(DE3) and T7 Express cells, restriction enzymes, ligase, and corresponding buffers and DNA extraction kits were purchased from New England Biolabs (Ipswich, MA). Isopropyl J3-D-1- thiogalactopyranoside (IPTG), Apomyoglobin, Cytochrome C, Aldolase, sinapinic acid, zinc sulfate, and Trifluoroacetic acid (TFA) were purchased from Sigma Aldrich (St. Louis, MO). Tryptone, yeast extract, sodium chloride, ampicillin, kanamycin, phosphate buffer saline (PBS), DMSO, isopropanol, acetonitrile, ethanol, 6x-His Tag Monoclonal Antibody1922656343. v1 -12 / 4 / 25(HIS.H8); Alexa Fluor™ 488, NBD Cholesterol [22-(N-(7-Nitrobenz-2-Oxa-l,3-Diazol-4-yl) Amino)-23,24-Bisnor-5-Cholen-3P-Ol] and bovine serum albumin (BSA) were obtained from ThermoFisher Scientific (Waltham, MA). High performance liquid chromatography (HPLC) grade acetonitrile was purchased from Fisher Scientific (Fair Lawn, NJ). Chloramphenicol was purchased from BIOBASIC (Markham, Canada). Mini-PROTEIN TGX stain free precast gels, precision plus protein unstained protein standards, 0.2 pm nitrocellulose membrane, Trans-Blot Turbo Transfer pack and EveryBlot Blocking Buffer were purchased from BioRad Laboratories, Inc. (Hercules, CA). Deionized water was obtained from Milli-Q system (Millipore SAS, France). Simply BlueTM Safe stain was purchased from Novex (Van Allen Way Carlsbad, CA). l-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) and 1,2- dioleoyl-sn-glycero-3 -phosphocholine (DOPC) were purchased from Avanti Polar Lipids (Alabaster, AL, USA). All chemicals were used as received without further purification. All oligonucleotides and gene fragments were purchased from Integrated DNA Technologies.

[0060] Cloning

[0061] General procedures

[0062] The genes and plasmids used in this study were constructed using standard molecular biology techniques, as detailed in the sections below. Non-repetitive genes (e.g., GGS, mCherry) were ordered as gene fragments from IDT DNA and assembled into vectors using NEBuilder® HiFi DNA Assembly according to the manufacturer's instructions. Gene amplification was performed using Q5 High-Fidelity Hot Start polymerase, with annealing temperatures optimized by gradient PCR (± 5 °C of the predicted annealing temperature by NEB Tm Calculator). Site-directed mutagenesis was performed using the Q5 Site -Directed Mutagenesis Kit with primers designed via NEB BaseChanger, and gradient PCR was applied as needed to optimize reaction conditions. Repetitive sequences (e.g., ELP variants), which are difficult to amplify due to degenerate codons, were assembled using recursive directional ligation by plasmid reconstruction (PRe-RDL) and cloned into final vectors using restriction enzyme-based ligation. All constructs were verified by Sanger sequencing (Genewiz), and final plasmids were fully sequence-verified using nanopore sequencing (Plasmidsaurus).

[0063] Cloning of codon-optimized GGS in pET24a

[0064] For the in vitro validation of Deinococcus radiodurans crtE (i.e., GGS), the codon-optimized cDNA was ordered as gene fragments from IDT, in frame with an N- terminal 6xHis tag and flanked by ~20 bp homology regions matching the pET24a vector. The pET24a vector (1 pg) was linearized by double digestion with Ndel and BamHI-HF,2022656343. v1 -12 / 4 / 25followed by purification through 1% agarose gel electrophoresis (130 V, 40 min) and gel extraction using the NEB Monarch Kit.

[0065] The linearized vector (50 ng) and synthetic gblock insert were combined at a 1:2 vector-to-insert ratio and assembled using NEBuilder HiFi master mix (50°C, 1 hour). After assembly, 2 pL of the reaction mixture was transformed into chemically competent E. coli Eb5a cells using heat shock (42°C, 30 seconds) and plated on LB agar containing 45 pg / mL kanamycin. Colony PCR was performed on selected colonies using standard T7 promoter primers, and positive constructs were confirmed by DNA sequencing.

[0066] Cloning of ELP-CVLL variants in pET24a

[0067] Fusion of ELP variants to the CVLL peptide, a substrate for geranylgeranyl transferase (GGT), was accomplished using a previously established procedure.1 Phosphorylated sense and antisense oligonucleotides encoding the CVLL sequence were annealed at 95 °C for 5 minutes, followed by cooling to room temperature in annealing buffer (50 mM Tris-HCl, 10 mM MgC12, 1 mM ATP, 10 mM DTT, pH 7.5). The annealed product was ligated into a modified pET24a vector, which had been linearized by double digestion with BseRI and BamHI and purified by agarose gel extraction. The ligation reaction, carried out using a 1:3 vector-to-insert ratio with a quick ligation kit, was incubated at room temperature for 15 minutes. The ligation products were transformed into chemically competent E. coli Eb5a cells using the heat shock method (42°C, 30 seconds). Transformed cells were plated on LB agar containing 45 pg / mL kanamycin, and positive constructs were identified by sequencing with the T7-term primer.

[0068] Next, the recursive directional ligation by plasmid reconstruction (PRe-RDL) method was employed to fuse the CVLL peptide sequence to the C-termini of ELP variants.2 The ELP plasmid (1 pg) was digested with Acul and Bgll, while the CVLL plasmid (1 pg) was digested with BseRI and Bgll for 3 hours at 37°C. DNA fragments were separated on a 1% agarose gel (130 V, 40 minutes), and the desired ELP and CVLL fragments were excised based on their predicted sizes, followed by purification using the Monarch Gel Extraction Kit (NEB).

[0069] Construction of pACYCDuet vectors

[0070] For in vivo geranylgeranylation studies, the pACYCDuet vector was used to co-express GGS and model proteins fused to the CVLL sequence. The construction of these vectors involved two steps: (1) subcloning of the GGS gene into the first multiple cloning site of the pACYCDuet vector, and (2) subcloning of POI-CVLL fusion genes into the second multiple cloning site.2122656343. v1 -12 / 4 / 25

[0071] Representative cloning method used for ELPs

[0072] First, the GGS gene was amplified from the pET24a plasmid using GGS-F and GGS-R primers (Table 1) in a PCR reaction containing 3 ng of template DNA, 500 nM of each primer, and Q5 High-Fidelity Hot Start master mix (NEB). The PCR conditions were initial denaturation at 98°C for 30 s, followed by 30 cycles of 98°C for 10 s, annealing at 70 °C for 30 s, and extension at 72°C for 30 s, with a final extension at 72°C for 2 minutes. The PCR products were purified using the Monarch PCR & DNA Cleanup Kit (NEB). The pACYCDuet-1 recipient vector was linearized by double digestion with Asci and NcoI-HF, and the linearized vector was gel-purified. The purified GGS PCR product and linearized vector were combined at a 1 :2 molar ratio and assembled using NEBuilder HiFi Master Mix (50°C, 30 minutes). The assembled plasmid (2 pL) was transformed into 25 pL of chemically competent NEB5a cells by heat shock and plated on LB agar containing 25 pg / mL chloramphenicol. Positive colonies were identified by colony PCR using ACYCDuetUPl and DuetDown-1 primers and confirmed by sequencing.

[0073] Next, the ELP-CVLL gene was excised from the donor pET24a plasmid using Ndel and Xhol and gel-purified. The recipient pACYCDuet GGS vector was linearized by digestion with Ndel and Xhol, and the linearized vector and ELP-CVLL insert were ligated in a 1:3 molar ratio using a quick ligation kit at room temperature for 10 minutes. The ligation mixture (3 pL) was transformed into 25 pL of NEB5a cells by heat shock and plated on LB agar containing 25 pg / mL chloramphenicol. Positive constructs (pACYDuet_GGS_POI-CVLL) were confirmed by sequencing.

[0074] For the construction of control pACYCDuet plasmids (lacking GGS), the first step of the procedure was skipped, and ELP-CVLL was directly cloned into an unmodified pACYCDuet vector (resulting in pACYDuet_()_ELP-CVLL plasmids).

[0075] Modified procedure for mCherry

[0076] To facilitate the purification of mCherry isoforms using immobilized metal affinity chromatography (IMAC), the following modifications to the cloning workflow were implemented: First, Q5 site-directed mutagenesis (Q5-SDM) was used to generate a pACYCDuet GGS plasmid lacking the N-terminal His-tag, allowing the His-tag to be used for the purification of mCherry instead. Second, NEBuilder HiFi Assembly was used for subcloning of mCherry-CVLL due to its higher efficiency compared to restriction-based ligation.

[0077] First, site-directed mutagenesis was performed using the Q5 Site -Directed Mutagenesis Kit (NEB) to remove the N-terminal 6xHis tag from the GGS gene. Primers2222656343. v1 -12 / 4 / 25(AHis F and AHis R, Table 1) were designed using the NEBaseChanger tool. A PCR reaction was set up with 10 ng of template DNA, 500 nM of each primer, and Q5 Hot Start High-Fidelity 2X Master Mix. The following PCR conditions were used: initial denaturation at 98°C for 30 s, followed by 25 cycles of 98°C for 10 s, 63°C for 30 s (annealing), and 72°C for 98 s (extension), with a final extension at 72°C for 2 minutes. After PCR, 1 pL of the product was incubated with the KLD reaction mix for 5 minutes at room temperature to phosphorylate and ligate the amplified DNA and degrade the template DNA via DpnI. Following incubation, 3 pL of the reaction mixture was transformed into 25 pL of NEB5a competent cells, and colonies were selected using the standard transformation protocol. Positive colonies were confirmed by DNA sequencing.

[0078] Next, the modified pACYCDuet GGS plasmid (1 pg) was linearized by restriction digestion using Ndel and Xhol. A synthetic gene fragment encoding mCherry fused to an N-terminal 6xHis tag and a C-terminal CVLL peptide was synthesized by Integrated DNA Technologies (IDT). The linearized backbone and synthetic gBlock were combined at a 1:2 molar ratio and incubated with NEBuilder HiFi Master Mix for 30 minutes at 50°C. Following assembly, 2 pL of the reaction mixture was transformed into 25 pL of chemically competent NEB5a cells using the heat shock method, and colonies were selected on LB agar plates supplemented with chloramphenicol (25 pg / mL). Positive colonies were confirmed by DNA sequencing.

[0079] For the construction of control pACYCDuet plasmids (lacking GGS), the first step of the procedure was omitted, and the mCherry-CVLL sequence was directly cloned into an unmodified pACYCDuet vector, resulting in the pACYCDuet mCherry-CVLL plasmid. Additionally, a control plasmid with a mutated AVLL peptide (pACYCDuet mCherry- AVLL) was constructed by replacing the CVLL sequence with AVLL using Q5 SDM kit using primers (C254A-F and C254A-R, Table 1).

[0080] Construction of pET23 vectors

[0081] pET23_GGTp_GGTa

[0082] A translationally coupled system was used for the co-expression of the alpha and beta subunits of Rattus norvegicus GGTase-I. The construction of this plasmid has been previously described.3

[0083] pET23_Control

[0084] As a control for 2x2 experiments, a plasmid with a similar backbone but lacking the alpha and beta subunits of Rattus norvegicus GGTase-I was synthesized. The parent plasmid, pET23a_GGT0_GGTa (1 pg), was linearized by restriction digestion with2322656343. v1 -12 / 4 / 25Xbal and Notl-HF, which excised the coding regions of the GGT subunits. The digest was separated by agarose gel electrophoresis, and the fragment corresponding to the pET23 backbone was excised and gel-purified. Phosphorylated sense and antisense oligonucleotides (DGGT-F and DGGT-R, Table 1) with compatible overhangs were annealed at 95 °C for 5 minutes, followed by cooling to room temperature in annealing buffer (50 mM Tris-HCl, 10 mM MgC12, 1 mM ATP, 10 mM DTT, pH 7.5). The annealed oligonucleotides were ligated into the pET23 backbone using a 1:3 vector-to-insert ratio with a quick ligation kit, incubated at room temperature for 15 minutes. The ligation products were transformed into chemically competent E. coli Eb5a cells using the heat shock method (42°C, 30 seconds) and plated on LB agar containing 150 pg / mL ampicillin. Positive colonies were identified by colony PCR using standard T7 primers and confirmed by DNA sequencing.

[0085] SDS-PAGE and Western Blot Analysis

[0086] SDS-PAGE was conducted using the Laemmli method on a 4-20% gradient Mini-PROTEIN Tris-glycerol (TGX) stain-free SDS-PAGE gel (Bio-Rad). Electrophoresis was performed at 200 V for 30 minutes. For proteins containing tryptophan, the gels were first visualized using the Gel-DOC EZ system with the stain-free protocol. Following visualization, the gels were stained with Simply Blue® (Coomassie-based dye) according to the manufacturer’s protocol and imaged.

[0087] For Western blot analysis, proteins were transferred from the SDS-PAGE gel to a 0.2 pm nitrocellulose membrane using a Trans-Blot Turbo Transfer pack (Bio-Rad) at 25V for 7 minutes. The membrane was washed three times with IX Tris-buffered saline (TBS) and blocked with EveryBlot Blocking Buffer (Bio-Rad) for 1 hour at room temperature with gentle agitation. Membranes were incubated overnight at 4°C in 10 mL of blocking buffer containing 6x-His Tag Monoclonal Antibody (HIS.H8), Alexa Fluor™ 488 (1:2000 dilution). After incubation, the membranes were washed three times for 10 minutes with Tris-buffered saline containing 0.1% Tween 20 (TBS-T) and imaged using ChemiDoc MP imager (Bio-Rad).

[0088] RP-HPLC

[0089] Analytical reverse-phase high-performance liquid chromatography (RP-HPLC) was performed on a Shimadzu LC-2030 instrument equipped with a Phenomenex Jupiter Cl 8 column (5 pm, 300 A, 250 x 4.6 mm), maintained at room temperature.Detection was carried out using a UV-Vis detector at 190 nm or 210 nm for proteins, and at 214 nm for lipids (in the GGS validation assay). All samples were filtered through a 0.2 pm2422656343. v1 -12 / 4 / 25PVDF filter (Durapore) before analysis. The flow rate was 1 mL / min, and the injection volume was 50 pL.

[0090] Due to the differences in solubility and stability of various compounds analyzed in this study (e.g., GGPP vs. geranylgeranylated proteins), different combinations of mobile phases and gradient programs were applied, as described below.

[0091] GGS validation

[0092] Mobile phase A: 25 mM ammonium bicarbonate (pH 8.0); Mobile phase B (organic): acetonitrile;

[0093] Gradient: 5-minute isocratic run with 0% B, followed by a linear increase to 90% B at the rate of 2.25%B / min.

[0094] Separation of ELP Isoforms

[0095] Mobile phase A: water + 0.1% TFA; Mobile phase B: acetonitrile + 0.1% TFA.

[0096] Gradient: 5-minute isocratic run with 0% B, followed by a linear increase to 90% B at the rate of 2.25%B / min.

[0097] Separation of mCherry Isoforms

[0098] Phenomenex Jupiter C4 column (5 pm, 300 A, 250 x 4.6 mm), Mobile phase A: water + 0.1% TFA; Mobile phase B: acetonitrile + 0.1% TFA. Gradient: 5-minute isocratic run with 0% B, followed by a linear increase to 90% B at the rate of 2.25%B / min. After baseline correction, the chromatograms were normalized using min-max normalization (0-1 range) in GraphPad Prism.

[0099] Separation of RhoA and Rap IB Isoforms

[0100] Phenomenex Jupiter C18 column (5 pm, 300 A, 250 x 4.6 mm), Mobile phase A: water + 0.1% TFA; Mobile phase B: acetonitrile + 0.1% TFA. Gradient: 5-minute isocratic run with 0% B, followed by a linear increase to 90% B at the rate of 2.25%B / min. After baseline correction, the chromatograms were normalized using min-max normalization (0-1 range) in GraphPad Prism.

[0101] MALDI-TOF

[0102] MALDI-TOF analysis was performed on a Bruker microflex LP equipped with a micro scout ion source and a nitrogen laser (337 nm), operated in linear positive mode. A solution of sinapinic acid in 70% acetonitrile with 0.1% trifluoroacetic acid (TFA) was used as the matrix. Protein samples were prepared by mixing 5 pL of protein (50 pM in water) with 5 pL of the matrix. Three serial dilutions were performed by a factor of 2 in the2522656343. v1 -12 / 4 / 25matrix solution to obtain a range of protein-to-matrix concentrations. The prepared solutions were applied to a MALDI sample plate and allowed to dry at room temperature.

[0103] The instrument was calibrated using dual standards flanking the molecular weight of the protein being analyzed. These standards included Cytochrome C ([M+H]+ = 12,362 Da), Apomyoglobin ([M+H]+ = 16,952.27 Da), and Aldolase ([M+H]+ = 39,212 Da) from the ProteoMass™ Protein MALDI-MS Calibration Kit (Sigma Aldrich). Each MALDI mass spectrum was acquired by averaging 256 laser shots to improve the signal-to-noise ratio. Data were analyzed by Bruker flex control 3.4 software. The theoretical molecular weight and experimentally observed m / z values are reported in Table 7.

[0104] Trypsin digestion and LC-MS

[0105] To confirm the selectivity of prenylation to the C-termini CaaX box, trypsin digest followed by LC-MS was employed.

[0106] Trypsin digest

[0107] All protein samples were digested using a standardized trypsin digestion protocol with minor alterations to account for differences in the solubility of the lipidated peptide fragments. Each sample (50 pg) was reduced with 10 mM TCEP and alkylated with 40 mM chloroacetamide for 30 minutes. Trypsin platinum (Promega) was added at an enzyme-to-protein ratio of approximately 1:50 (w / w), and digestion was allowed to proceed overnight at 37°C. After digestion, samples were acidified with trifluoroacetic acid (TFA, Sigma) and cleaned using stage tips (3M, 2241).4 The stage tips were activated with acetonitrile followed by 3% acetonitrile with 0.1% TFA. The eluted peptides were dried, resuspended in a mixture of water, acetonitrile, and formic acid, and analyzed by LC-MS.

[0108] Sample-specific alterations:

[0109] Unmodified: No acetonitrile was added to the digest reaction, cleaned using a 3 -punch Cl 8 stage tip and eluted with 100 pL of 75% acetonitrile with 0.1% TFA.

[0110] Prenylated (Fr- / GG-modified): Acetonitrile was added to the digest reaction to the final concentration of 20%. Cleaned using a 2-punch MCX stage tip and eluted with 75 pL of 85% acetonitrile with 5% NH4OH (Sigma).

[0111] LC condition

[0112] Samples (2 pL) were injected onto a pulled tip nano-LC column with 75 pm inner diameter packed to 25 cm with 3 pm, 120 A, C18AQ particles (Dr. Maisch). The peptides were separated using a 60 min gradient consisting of mobile phase A: water + 0.1% formic acid and mobile phase B: acetonitrile + 0.1% formic acid.

[0113] Gradient program:2622656343. v1 -12 / 4 / 25

[0114] MS condition

[0115] The LC column was connected in line with an Orbitrap Lumos via a nanoelectrospray source operating at 2.5 kV. The mass spectrometer was operated using alternating MSI and targeted MS2 scans. The MSI scans were collected at 120,000 resolution with a maximum injection time of 50 ms. MS2 scans were performed on the theoretical mass of the lipidated peptide and its oxidized counterpart, in the Orbitrap at 30,000 resolution, following each CID and HCD activation (four MS2 scans per cycle).

[0116] Protein expression and purification

[0117] Expression and purification of 6xHis-GGS

[0118] Expression

[0119] The plasmid encoding the N-terminal His-tagged variant of GGS (pET24a_6xHis-GGS) was transformed into E. coli BL21(DE3) cells using the heat shock method, followed by selection on LB agar plates containing 90 pg / mL kanamycin. A single colony was inoculated into 50 mL of 2xYT medium with kanamycin (90 pg / mL) and incubated at 37°C with shaking at 200 rpm. After the culture reached an OD600 of 0.7, the cells were harvested by centrifugation (4000 x g, 15 min, 4°C), and the pellet was resuspended in 5 mL of PBS buffer. One milliliter of the resuspended pellet was used to inoculate 1 L of 2xYT medium supplemented with kanamycin (90 pg / mL), i.e., 1 : 100 inoculum. The cultures were incubated at 37°C with shaking (200 rpm), and upon reaching mid-log phase (OD600 = 0.7), protein expression was induced with 1 mM IPTG. After 16 h, the cells were harvested by centrifugation (3745 x g, 20 min, 4°C).

[0120] Purification

[0121] For GGS validation studies, cells were resuspended in lysis buffer and lysed by sonication (3 min, 55-75 Watts, Pulse program: 10 s on, 15 s off). The lysate was clarified by centrifugation (22000 x g, 15 min, 4°C), and the His-tagged protein was purified using immobilized metal affinity chromatography (IMAC) on HisPur™ Cobalt Resin2722656343. v1 -12 / 4 / 25(ThermoFisher Scientific) following the manufacturer's protocol, with minor alterations for compatibility with downstream assays as described below.

[0122] For secondary structure determination using circular dichroism (Section 8), the bacterial pellet was resuspended in PBS (5 mL per liter of culture). The wash buffer contained 50 mM sodium phosphate, 300 mM NaCl, and 10 mM imidazole (pH 7.2), while the elution buffer contained 50 mM sodium phosphate, 300 mM NaCl, and 150 mM imidazole (pH 7.2). Elution fractions were pooled and dialyzed (MWCO 3 kDa) overnight against 10 mM phosphate buffer (pH 7.4) to remove chloride ion interference during far-UV circular dichroism data collection. Protein concentration in the dialyzed retentate was determined by measuring absorbance at 280 nm, using an extinction coefficient of 55920 M 'em1(calculated with ProtParam).

[0123] For in-vitro GGPP synthesis assay (Section 10), HEPES-based buffers were used to eliminate the interference of phosphate ions with the BIOMOL® Green assay for pyrophosphate quantification. Cells were lysed in HEPES buffer (100 mM, 137 mM NaCl, pH 7.2). The buffers used for IMAC purification were wash buffer (100 mM HEPES, 300 mM NaCl, 10 mM imidazole, pH 7.2) and elution buffer (100 mM HEPES, 300 mM NaCl, 150 mM imidazole, pH 7.2). Elution fractions were pooled, the protein concentration was determined, and the protein was used for the GGPP synthesis assay.

[0124] Expression and purification of ELP Isoforms

[0125] Expression

[0126] A pairwise combination of plasmids listed in Table 3 was used to co-express ELP in the absence or presence of GGS and GGT. A representative protocol for coexpression in the +GGS / +GGT strain (resulting in production of ELP-GG isoforms) is provided below.

[0127] The pET23a_GGTp_GGTa and pACYCDuet_GGS_ELP-CVLL plasmids were co-transformed into E. coli BL21(DE3) competent cells using heat shock (42°C, 30 seconds). Transformed cells were plated on LB agar containing 150 pg / mL ampicillin and 25 pg / mL chloramphenicol for selection. A freshly transformed colony was used to inoculate a 50 mL starter culture of 2xYT medium supplemented with ampicillin (150 pg / mL) and chloramphenicol (25 pg / mL). The starter culture was incubated in an orbital shaker (37°C, 250 rpm) until an OD600 of 0.7 was reached. The cells were harvested by centrifugation (5000 x g, 10 min, 4°C), and the supernatant was discarded. The pellet was resuspended in sterile, chilled PBS (4°C) and used to inoculate 1 L of 2xYT medium supplemented with ampicillin (150 pg / mL) and chloramphenicol (25 pg / mL) at a 1:100 inoculum ratio. The2822656343. v1 -12 / 4 / 25larger cultures were incubated at 37°C with shaking (250 rpm) until the OD600 reached 0.7. At this point, the temperature was reduced to 28°C, and 1 mM ZnSO4 (cofactor for GGT, lOOOx stock solution) was added to the cultures. After 10 minutes, protein expression was induced by adding IPTG to a final concentration of 0.5 mM. After 16 hours of incubation, the cells were harvested by centrifugation (5000 x g, 35 min, 4°C).

[0128] Sample-specific alterations: Hydrophilic ELP isoforms (ELPA40) were expressed in T7 Express (NEB cells). A higher concentration of IPTG (1 mM instead of 0.5 mM) was used for protein induction.

[0129] Purification

[0130] The ELP isoforms were isolated following an established procedure.5 Briefly, cells were lysed with isopropanol, which resulted in selective partitioning of the ELP into the organic layer. After separating insoluble fractions, the ELPs were precipitated by adding acetonitrile (a non-solvent) to the alcohol layer. The precipitated ELPs were resuspended and purified by RP-HPLC to achieve >95% homogeneity for self-assembly studies. A representative protocol is provided below.

[0131] Cell pellets were resuspended in 4 volumes of isopropanol (4 mL / g of wet pellet) by vortexing for 10 minutes, followed by sonication in an ultrasonic bath (VEVOR 15L Ultrasonic, 40 kHz, 25°C). The lysate was centrifuged (15000 x g, 10 minutes, 25°C) to separate the supernatant from the insoluble debris. ELPs were precipitated by adding acetonitrile to the supernatant to achieve a final composition of 70% acetonitrile in isopropanol (v / v), followed by centrifugation (15000 x g, 10 minutes, 4°C). The resulting ELP pellet was resuspended in 50% (v / v) ethanol in water (4 mL / g of ELP pellet) and centrifuged (15000 x g, 5 minutes, 25°C) to remove insoluble impurities. The supernatant was concentrated under reduced pressure (40 °C, 280 mm Hg) to approximately 10 mL, then diluted with deionized water to achieve a 20% ethanol (v / v) solution. The mixture was flash- frozen using liquid nitrogen and subsequently freeze-dried (Labconco FreeZone 4.5 -80°C Benchtop Freeze Dryer).

[0132] The lyophilized cake was resuspended in 50% ethanol (v / v), centrifuged (21000 x g, 10 min, 25°C), and the supernatant was filtered (0.22 pm PVDF) before preparative HPLC. Semi-preparative reverse-phase HPLC (RP-HPLC) was performed using a Prominence HPLC system (Shimadzu) with a PDA detector, equipped with a Phenomenex JUPITER® C18 column (5 pm, Cl 8, 300 A, 250 x 10 mm), at a flow rate of 4.2 mL / min. A mobile phase consisting of a gradient of water (solvent A) and acetonitrile (solvent B), both supplemented with 0.1% TFA, was used to elute the proteins. The gradient consisted of a 5-2922656343. v1 -12 / 4 / 25minute isocratic run at 0% B, followed by a linear increase to 90% B at a rate of 2.25% B / min. Eluted protein fractions were pooled, flash-frozen in liquid nitrogen, and lyophilized. Lyophilized proteins were stored at -20°C.

[0133] Expression and purification of mCherry Isoforms

[0134] Expression

[0135] A pairwise combination of plasmids listed in Table 3 was used to co-express mCherry in the absence or presence of GGS and GGT. A representative protocol for coexpression in the +GGS / +GGT strain (resulting in production of mCherry-GG isoform) is provided below.

[0136] The pET23a_GGTp_GGTa and pACYCDuet_GGS_6x-His-mCherry-CVLL plasmids were co-transformed into E. coli BL21(DE3) cells via heat shock and plated on LB agar containing 150 pg / mL ampicillin and 25 pg / mL chloramphenicol. A single colony was inoculated into 50 mL of 2xYT medium with the same antibiotics and incubated at 37°C with shaking (250 rpm) until an OD600 of 0.7. The cells were harvested (5000 x g, 10 min, 4°C), resuspended in chilled PBS (4°C), and used to inoculate 1 L of 2xYT medium (1 : 100 inoculum) supplemented with antibiotics. The culture was incubated at 37°C (250 rpm) until an OD600 of 0.7, after which the temperature was lowered to 28°C, and 1 mM ZnSO4 was added. After 10 minutes, protein expression was induced with 1 mM IPTG, and the culture was incubated for 16 hours. The cells were harvested by centrifugation (5000 x g, 35 min, 4°C), resuspended in PBS (5 mL / L of culture), and stored at -80°C before purification.

[0137] Purification

[0138] Cells were mixed with 4 volumes of bacterial protein extraction reagent (B- PER) supplemented with lysozyme (50 mg / mL), DNase I (2514 U / mL), and IX Halt protease inhibitor cocktail. After incubating at room temperature for 15 minutes, the lysate was chilled on ice for 5 minutes and then centrifuged at 21000 x g for 20 minutes at 4°C to remove insoluble fractions. The supernatant was supplemented with 10 mM imidazole and incubated with HisPur™ Cobalt Resin (ThermoFisher Scientific) on an end-to-end rotator for 30 minutes at 4°C. After the lysate was removed, the resin was washed three times with 2 volumes of wash buffer (50 mM Tris-HCl, 300 mM NaCl, 40 mM imidazole, pH 7.4). His- tagged proteins were eluted by incubating the resin with 1 volume of elution buffer (50 mM Tris-HCl, 300 mM NaCl, 300 mM imidazole, pH 7.4) for a total of four elution steps. Elution fractions were pooled, concentrated, and buffer-exchanged into PBS using an Amicon Ultra centrifugal concentrator (MWCO 3 kDa). The protein solutions were stored at -80°C.

[0139] Expression and purification of RhoA Isoforms3022656343. v1 -12 / 4 / 25

[0140] Expression

[0141] The schematic of plasmids used for RhoA expression in +GGS / ±GGT strains is shown in Figure 25. A representative protocol for expression in the +GGS / +GGT strain (resulting in the production of the RhoA-GG isoform) is provided below. The control experiment (expression of RhoA in +GGS / -GGT strain) was performed in parallel and under identical conditions, except that ampicillin was omitted, as it was only necessary for maintenance of GGT vector.

[0142] The pET23a_GGTp_GGTa and pACYCDuet_GGS_6xHis-RhoA plasmids were co-transformed into E. coli BL21(DE3) cells via heat shock and plated on LB agar containing 100 pg / mL ampicillin and 25 pg / mL chloramphenicol. A single colony was inoculated into 50 mL of 2xYT medium with the same antibiotics and incubated at 37°C with shaking (250 rpm) until an ODeoo of 0.7. The cells were harvested (5000 x g, 10 min, 4°C), resuspended in chilled PBS (4°C), and used to inoculate 1 L of 2xYT medium (1 : 100 inoculum) supplemented with antibiotics. The culture was incubated at 37°C (250 rpm) until an ODeoo of 0.7, after which the temperature was lowered to 28°C, and 1 mM ZnSCL was added. After 10 minutes, protein expression was induced with 1 mM IPTG, and the culture was incubated for 16 hours. The cells were harvested by centrifugation (5000 x g, 35 min, 4°C) and stored at -80°C before purification.

[0143] Purification

[0144] Cells were lysed with 4 volumes of bacterial protein extraction reagent (B- PER) supplemented with lysozyme (50 mg / mL), DNase I (2514 U / mL), 5mM MgCh, 5mM P-mercaptoethanol (BME), 10% glycerol, and ImM phenylmethylsulfonyl fluoride (PMSF). After incubating at room temperature for 15 minutes, the lysate was centrifuged at 21000 x g for 10 minutes at 4°C to remove insoluble fractions. The supernatant was supplemented with wash buffer (25mM Tris, 300 mM NaCl, 5mM MgCh, 5mM BME, 10% glycerol, ImM PMSF and 25 mM imidazole, pH 7.4) and incubated with HisPur™ Ni-NTA Resin (Thermo Fisher Scientific) on an end-to-end rotator for 30 minutes at 4°C. After the lysate was removed, the resin was washed three times with 2 volumes of wash buffer. His-tagged proteins were eluted by incubating the resin with 1 volume of native elution buffer (25mM Tris, 300 mM NaCl, 5mM MgCh, 5mM BME, 10% glycerol, ImM PMSF and 300 mM Imidazole, pH 7.4) for a total of three elution steps. Elution fractions were analyzed by RP- HPLC.

[0145] Expression and purification of RaplB Isoforms

[0146] Expression3122656343. v1 -12 / 4 / 25

[0147] The schematic of plasmids used for Rap IB expression in +GGS / ±GGT strains is shown in Figure 26. A representative protocol for expression in the +GGS / +GGT strain (resulting in the production of the RaplB-GG isoform) is provided below. The control experiment (expression of RaplB in +GGS / -GGT strain) was performed in parallel and under identical conditions, except that ampicillin was omitted, as it was only necessary for maintenance of GGT vector.

[0148] The pET23a_GGTp_GGTa and pACYCDuet_GGS_6x-His-RaplB plasmids were co-transformed into E. coli BL21(DE3) cells via heat shock and plated on LB agar containing 100 pg / mL ampicillin and 25 pg / mL chloramphenicol. A single colony was inoculated into 50 mL of 2xYT medium with the same antibiotics and incubated at 37°C with shaking (250 rpm) until an ODeoo of 0.7. The cells were harvested (5000 x g, 10 min, 4°C), resuspended in chilled PBS (4°C), and used to inoculate 1 L of 2xYT medium (1 : 100 inoculum) supplemented with antibiotics. The culture was incubated at 37°C (250 rpm) until an ODeoo of 0.7, after which the temperature was lowered to 28°C, and 1 mM ZnSCL was added. After 10 minutes, protein expression was induced with 1 mM IPTG, and the culture was incubated for 16 hours. The cells were harvested by centrifugation (5000 x g, 35 min, 4°C) and stored at -80°C before purification.

[0149] Purification

[0150] Cells were lysed with 4 volumes of bacterial protein extraction reagent (B- PER) supplemented with lysozyme (50 mg / mL), DNase I (2514 U / mL), 5mM MgCh, 5mM P-mercaptoethanol (BME), 10% glycerol and ImM phenylmethylsulfonyl fluoride (PMSF). After incubating at room temperature for 15 minutes, the lysate was centrifuged at 21000 x g for 10 minutes at 4°C. After washing the pellet with equal volume of lysis buffer, the inclusion bodies were resuspended in the resolubilization / wash buffer (6M guanidium-HCl, 25mM Tris, 300 mM NaCl, 5mM MgCl2, 5mM BME, 10% glycerol, ImM PMSF, and 25 mM imidazole, pH 7.4). The solution was briefly sonicated, and then incubated at room temperature on an end-to-end rotator for one hour. After centrifugation, the resolubilized inclusion bodies were incubated with HisPur™ Ni-NTA Resin (Thermo Fisher Scientific) for 30 minutes at room temperature. The resin was washed three times with 2 volumes of guanidinium wash buffer. His-tagged proteins were eluted by incubating the resin with 1 volume of guanidinium elution buffer (6M guanidium-HCl, 25mM Tris, 300 mM NaCl, 5mM MgCh, 5mM BME, 10% glycerol, ImM PMSF and 300 mM Imidazole, pH 7.4) for a total of three elution steps. Elution fractions were analyzed by RP-HPLC

[0151] Circular dichroism3222656343. v1 -12 / 4 / 25

[0152] Circular dichroism was used to evaluate the secondary structure of recombinantly expressed and purified GGS (Section 7.1) for comparison with MD simulation results. CD spectra were recorded on an AVIV Model 420 CD spectrometer (AVIV Biomedical, Lakewood, NJ, USA). Measurements were performed in the far-UV region (190-250 nm) using a 1.0 mm path length quartz cuvette at 25 °C. Each wavelength was scanned for 15 seconds, and the spectra were averaged over three scans. The protein solution was prepared at 5 pM in phosphate buffer (10 mM, pH 7.4). Baseline correction was performed by subtracting the CD spectra of the buffer.

[0153] The baseline-corrected CD spectra were converted to mean molar residue ellipticity (MRE) using the following equation:0 6mre~ 10 x C x Ar x ZWhere[0]MfiB= mean molar residue ellipticity (deg. cm2. mol-1)0 = observed ellipticity in mdegC = molar concentration of the proteinNr = number of residuesI = pathlength in cm

[0154] The CD signal was deconvoluted using the BeStSel Server6 for secondary structure analysis. The results are reported as the mean ± standard deviations of three measurements.

[0155] Computational modelling

[0156] System building

[0157] The crystal structure for Deinococcus radiodurans crtE has not been determined yet, but the amino acid sequence is available. The AlphaFold 3.07 webserver was used to generate the structure of this protein. The webserver yielded 5 similar models as seen by the small RMSD difference between them (<2 A). All the models showed the same ranking score of 0.93, which implies a good prediction and hence we selected one of the structures arbitrarily for our usage. The protein structure is shown in FIG. 6a with chain A and B in different colors.

[0158] The protein structure was further validated against other similar synthases whose crystal structures are available. The multiseq8 plugin in VMD9 was used to align a single chain of the following synthases: Iwmw, Iwyo, 2f7m, 3oyr, and 5djp as shown in FIG.3322656343. v1 -12 / 4 / 256b. The GGPP synthase chain A, colored in cyan, was also aligned with the rest of the synthases. The sequence identity of the homologues with respect to the GGS is shown in Table 7.

[0159] MD simulation

[0160] All molecular dynamics (MD) simulations were performed in NAMD with the CHARMM force field for the GGPP synthase protein, and TIP3P model was used for water. The Gaussian accelerated molecular dynamics (GaMD) used the GPU-offload mode of NAMD 2.14, while the production runs of conventional MD used the GPU resident mode of NAMD 3.0. The temperature was maintained at 310 K by a Langevin thermostat, and the pressure was kept at 1 atm using Langevin piston barostat. The Particle Mesh Ewald (PME) was used to calculate long range electrostatic interaction at every time step with a 12 A cutoff for non-bonded interactions. All bonds involving H-atoms were constrained using the SHAKE algorithm.

[0161] Following a 40000-step minimization, a two-step equilibration simulation is performed in which the protein is successively relaxed. The first equilibration was run for 100 ps where the GGPP synthase was constrained, and the second equilibration, where everything was released, was run for 10 ns. Both equilibration steps used a timestep of 1 fs / step. This was followed by a production run of 1.1 ps (2 replicates) with a timestep of 2 fs / step.

[0162] GaMD simulation

[0163] Following the equilibration simulation, we also ran GaMD simulations in two steps, boosting only boosted the dihedral energy and not the total potential energy. The first step was a 2ns conventional MD and 50 ns of GaMD equilibration, followed by a 200 ns GaMD production simulation with the upper limit of dihedral boost potential fixed at 15 kcal / mol.

[0164] GGS validation assay

[0165] BIOMOL green assay

[0166] Chemically synthesized famesyl pyrophosphate (FPP) and isopentenyl pyrophosphate (IPP) were obtained from Isoprenoids, LC (United States). Stock solutions ofFPP (5 mM) and IPP (10 mM) were prepared by dissolving the compounds in 30% (v / v) ammonium hydroxide in methanol and stored at -80 °C. Recombinant his-tagged Deinococcus radiodurans crtE was purified as described in section 7.1, and the concentration of the enzyme was determined by measuring absorbance at 280 nm.3422656343. v1 -12 / 4 / 25

[0167] The enzymatic activity of GGS was assessed in vitro using FPP (200 pM) and IPP (366 pM) as substrates. Reactions were carried out in a total volume of 50 pL in an assay buffer composed of 100 mM HEPES (pH 7.5), 5 mM MgC12, and 10 mM KC1. Recombinant crtE was used as the catalyst at a final concentration of 0.5 pM. Control reactions were performed in parallel by omitting one or both substrates (FPP, IPP) or the enzyme (crtE), and replacing them with appropriate diluents to maintain a consistent solvent composition across all samples. After the enzyme (or diluent for negative controls) was added, reactions were incubated at 28 °C for 2 hours. To terminate the reaction, 100 pL of BIOMOL GREEN reagent (Enzo Life Sciences) was added to each tube. After 20 minutes, the absorbance at 620 nm was measured using a Nanodrop One spectrophotometer (ThermoFisher Scientific).

[0168] Each assay was performed in triplicate, and the results are reported as the mean ± standard deviation. Statistical analysis was conducted using one-way ANOVA, followed by Dunnett’s multiple comparison test to compare the positive control reaction (FPP + IPP + crtE) to the control groups.

[0169] Characterization of reaction products by RP-HPLC and mass spectrometry

[0170] To further characterize the reaction products, the enzymatic reactions were scaled up to a total volume of 300 pL, using the same buffer as described above. Following incubation with enzyme, the reaction mixture containing all components (FPP, IPP, and crtE) became slightly turbid, hypothesized to be due to the limited solubility of the product (GGPP), in the reaction buffer. Therefore, the reaction mixture was centrifuged (21,000 x g, 5 min, RT). After aspiration of the supernatant, the pellet was dissolved in 25 mM ammonium bicarbonate buffer (pH 8), which is a more suitable solvent for GGPP. Both the supernatant and resolubilized pellet fractions were analyzed by RP-HPLC, as described in section 4. GGPP was detected exclusively in the resolubilized pellet fraction, while no significant product was observed in the supernatant (data not shown).

[0171] Liquid chromatography-mass spectrometry analysis was conducted to confirm the identity of the reaction products. Reaction mixtures were analyzed using a Thermo Vanquish binary pump LC coupled with a Thermo Quantis triple quadrupole MS. Chromatographic separation was achieved on a Waters XBridge C18 column (3.5 pm, 2.1 x 100 mm). The mobile phase consisted of 25 mM ammonium bicarbonate in water (solvent A) and acetonitrile (solvent B). The gradient program was set as follows:3522656343. v1 -12 / 4 / 25

[0172] The MS was operated in negative ion mode using heated electrospray ionization (H-ESI) with a spray voltage of 4500 V. The ion transfer tube temperature was maintained at 300°C, and the vaporizer temperature was 275°C. The scan range was set from 200 to 500 m / z with a resolution (FWHM) of 0.7. Data acquisition and analysis were performed using standard instrument software.

[0173] VT-Turbidimetry

[0174] The temperature-triggered phase-separation behavior of proteins was monitored using a UV-Vis spectrophotometer (Cary 100, Agilent) equipped with a Peltier temperature controller. The absorbance of protein solutions at 350 nm was measured across six concentrations (3.1, 6.2, 12.5, 25, 50, and 100 pM in PBS). The temperature was increased at a rate of l°C / min from 15°C to a maximum of 65°C, 80°C, or 97°C, depending on the hydrophobicity of the ELP construct. The transition temperature (Tt) at each concentration was defined as the inflection point of turbidimetry plots, corresponding to the maximum of the first derivative of absorbance vs. temperature. Variations in the final temperature did not impact the observed Tt as the phase-separation was complete before reaching these temperatures.

[0175] The concentration dependencies of Tt were fitted to two models:

[0176] Linear model: Tt=Td - mxlog([ELP]), where Td is the transition temperature of ELP under dilute condition (also known as reference temperature which is a measure of intrinsic hydrophobicity of the sequence) and m represents the concentration dependence of Tt (slope). This empirical model has been shown to predict the behavior of ELPs. The data for both unmodified ELP and GG-modified constructs fit well to this model, as indicated by the R2values.

[0177] Sigmoidal curve fit: The concentration dependencies of Fr-modified ELPs across the full concentration range did not fit well to the linear model. Visual inspection suggested a sigmoidal trend, so a four-parameter logistic (4PL) model was used to fit the data:3622656343. v1 -12 / 4 / 25Tc = Transition temperature at high concentration,Td = Transition temperature at dilute conditionCm = concentration at which Ttis halfway between transition temperature at high and low concentrations s = (Hill) slope of the curve at Cm.

[0178] Each experiment was conducted in triplicate, and the results are reported as mean ± standard deviation. The results are summarized in Table 9 and 10.3722656343. v1 -12 / 4 / 25

[0179] Dynamic light scattering

[0180] Dynamic light scattering was performed using a Zetasizer NanoZS (Malvern Panalytical, UK) equipped with a 173° backscattering detector. Protein samples (50 pM in PBS) were prepared at 4°C and fdtered into the DLS cuvette using a pre-chilled 0.22 pm PVDF (Durapore) filter. Measurements were conducted across a temperature range of 15- 65°C at 1°C increments. Each sample was equilibrated for 2 minutes at each temperature before recording intensity fluctuations over 11 runs, each lasting 60 seconds.

[0181] Autocorrelation functions were analyzed using Zetasizer software (Version 7.11) with the cumulants method to derive the average hydrodynamic diameter (Z-average, Zavg) and poly dispersity index (PDI). Intensity distributions were generated using the general-purpose algorithm with default settings. Each DLS run was performed with three technical replicates at each temperature, and the technical replicates were averaged to determine the average hydrodynamic parameters for sample at each temperature. Three results are reported as mean ± standard deviation of three independent samples (Table 11)

[0182] Cryo-TEM

[0183] Protein solution (4 pL, 100 pM in PBS) was deposited onto freshly plasma- cleaned Quantifoil grids (Quantifoil Micro Tools GmbH, Germany. The grids were stored in an environmentally controlled chamber at T < Tt for 5 minutes — 15 °C for ELPV40-GG and 25 °C for both ELP(V8 / A2)80-GG and ELPA40-GG — under 100% humidity. After blotting the excess solution, the grids were vitrified by plunging them into liquid ethane using a Vitrobot Mk IV (Thermo Fisher Scientific) and stored under liquid nitrogen until imaging.

[0184] Imaging was performed on a Tecnai BioTwin transmission electron microscope operating at 120 kV, equipped with a Gatan SC1000A CCD camera, and3822656343. v1 -12 / 4 / 25maintained at liquid nitrogen temperature. Images were acquired under low-dose conditions using a Gatan 626 cryo-holder.

[0185] For image analysis, ImageJ was used to quantify nanoparticle sizes across multiple images (n=77) for ELP(V8 / A2)80-GG and (n=73) for ELPV40-GG, and histograms of the results were generated using OriginPro. Due to the high hydrophilicity of the ELPA40 construct, visualization using cryo-TEM was challenging; however, based on DLS data, we anticipate that these constructs exhibit similar spherical morphologies.

[0186] Membrane interaction studies

[0187] GUV preparation

[0188] Giant unilamellar vesicles (GUVs) were prepared using the electroformation method with two distinct lipid compositions. The first lipid mixture consisted of DOPC and NBD-Cholesterol at a weight ratio of 150:1, while the second contained DOPC, DPPC, and NBD-Cholesterol at a weight ratio of 75:75: 1. The fluorescent probe NBD-Cholesterol is known to preferentially partition into the disordered regions of lipid membranes, specifically the DOPC-rich domains in mixtures containing both DOPC and DPPC. Both mixtures were dissolved in chloroform at a concentration of 9 mg / mL.

[0189] A volume of 10 pL of the lipid mixture was spread on the conductive surface of an indium tin oxide (ITO)-coated glass slide. After overnight evaporation of the solvent, 280 pL of buffer solution (1 mM HEPES, 300 mM glucose, pH = 7.4) was added to the lipid- coated surface within an O-ring chamber, which was then sealed with another ITO-coated slide (with the conductive surfaces facing each other). The electroformation chamber was connected to the Nanion Vesicle Prep Pro setup (Nanion / Vision-Tek), and GUV formation was induced using a three-step electroformation protocol: (1) the AC voltage was linearly increased from 0 to 3 V peak-to-peak at a frequency of 10 Hz over 15 minutes; (2) the voltage was maintained at 3 V and 10 Hz for 2.5 hours; (3) the frequency was linearly decreased to 4 Hz over 15 minutes. All steps were conducted at 37 °C. The resulting vesicles were collected, diluted 1:4 in the storage buffer (10 mM HEPES, 150 mM NaCl, pH= 7.4), and stored at room temperature for use within 24 hours.

[0190] GUV-b inding assays

[0191] GUV-binding assays were performed in a p-Slide 8 Well glass chamber (ibidi). The glass surface was passivated by incubating with 200 pL of 2 mg / mL BSA for 1 hour at room temperature. After incubation, the chamber was rinsed with distilled water (3 rinses, 200 pL each) to remove unbound BSA. Next, 100 pL of the GUV solution was added3922656343. v1 -12 / 4 / 25to the wells, followed by the addition of protein solutions to achieve a final protein concentration of 1 pM.

[0192] After a 15-minute incubation, the mixtures were imaged using a Zeiss LSM 980 Airyscan 2 confocal microscope equipped with a 63xoil immersion objective (NA 1.4). NBD-Cholesterol-containing GUVs were excited at 488 nm, and emission was detected from 491 to 585 nm. mCherry was excited at 561 nm, and emission was detected from 591 to 705 nm.4022656343. v1 -12 / 4 / 25

Claims

CLAIMSWhat is claimed is:

1. A method of producing a geranylgeranylated protein, comprising the steps of: providing an amount of Escherichia coli bacteria that has been modified to express a geranylgeranyl pyrophosphate synthase, to express a geranylgeranyltransferase, and to express a target protein having a cysteine residue within a CaaX box motif near a C-terminus of the target protein; and inducing protein expression in the amount of Escherichia coli bacteria; and isolating any geranylgeranylated target protein from the amount of Escherichia coli bacteria.

2. The method of claim 1, wherein the geranylgeranyl pyrophosphate synthase comprises Deinococcus radiodurans geranylgeranyl pyrophosphate synthase.

3. The method of claim 2, wherein the geranylgeranyl pyrophosphate synthase comprises SEQ ID NO: 11.

4. The method of claim 1, wherein the geranylgeranyltransferase comprises Rattus norvegicus geranylgeranyltransferase.

5. The method of claim 1, wherein the geranylgeranyltransferase comprises SEQ ID NO: 9 and SEQ ID NO: 10.

6. The method of claim 1, wherein the target protein comprises an elastin-like polypeptide.

7. The method of claim 7, wherein the elastin-like polypeptide has a sequence selected from the group consisting of SEQ ID NO: 12, SEQ ID NO: 13, and SEQ ID NO: 14.

8. The method of claim 1, wherein the target protein comprises a globular protein.

9. The method of claim 8, wherein the globular protein is SEQ. ID NO. 15.

10. The method of claim 1, wherein the target protein comprises a GTPase.

11. The method of claim 10, wherein the GTPase is selected from the group consisting of RhoA and Rap IB.

12. The method of claim 11, wherein the GTPase is SEQ ID NO: 16.

13. The method of claim 11, wherein the GTPase is SEQ ID NO: 17.

14. The method of claim 1, further comprising the step of incubating the amount of Escherichia coli bacteria while inducing protein expression.

15. A geranylgeranylated protein formed by the method of claim 1.4122656343. v1 -12 / 4 / 25

Images