Efficient purification of ribulose-1,5-bisphosphate carboxylase / oxygenase (rubisco) from complex cellular material using liquid-liquid phase separation

Abstract

Methods and systems are provided for the purification of native and recombinant Ribulose- 1,5- bisphosphate Carboxylase / Oxygenase (RuBisCO) from cellular lysates using liquid-liquid phase separation (LLPS). The invention enables scalable and cost-effective extraction of highly pure RuBisCO complexes from plant and bacterial sources while maintaining functional protein integrity. The methods and compositions disclosed herein address key obstacles to RuBisCO isolation and offer commercial advantages for protein production and utility in nutritional formulations.

Description

EFFICIENT PURIFICATION OF RIBULOSE-1,5-BISPHOSPHATE CARBOXYLASE / OXYGENASE (RUBISCO) FROM COMPLEX CELLULAR MATERIAL USING LIQUID-LIQUID PHASE SEPARATIONCROSS REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U. S. Provisional Patent Application No. 63 / 715,169, filed November 1, 2024, which is incorporated by reference herein in its entirety.REFERENCE TO AN ELECTRONIC SEQUENCE LISTING

[0002] The contents of the electronic sequence listing (24-1317-WO MIT26071 WO_ST26_Sequence_Listing.xml; Size: 20,121 bytes; and Date of Creation: October 30, 2025) is herein incorporated by reference in its entirety.BACKGROUND

[0003] Developing RuBisCO as a food protein is highly attractive due to its nutritional quality, digestibility, and abundance, but significant obstacles have historically limited its commercial use in food applications. RuBisCO comprises up to 50% of leaf protein and possesses a complete essential amino acid profile, outperforming many conventional plant and animal proteins for human nutrition.

[0004] However, significant challenges exist. RuBisCO is present in relatively low concentrations in plant leaves, requiring the processing of substantial biomass to recover useful amounts of protein. Further, extraction and purification are complicated by the need to maintain the native protein structure and function, removal of chlorophyll and bitter-tasting compounds, and avoidance of protein denaturation and browning reactions during processing. Current methods are labor-intensive, costly, and result in low yields and compromised product quality, limiting scalability and consumer acceptance.

[0005] There is urgent need for technically robust, scalable approaches to isolate highly pure, RuBisCO suitable for food use. Such methods should improve process yields, product taste, appearance, and nutritional value, overcoming the limitations of existing protein extraction technologies. The present disclosure addresses these long-standing obstacles by enabling efficient, selective LLPS-based RuBisCO purification, facilitating functional protein recovery for sustainable food and biotechnology markets.SUMMARY

[0006] The present disclosure generally relates to method for purifying or isolating RuBisCO, as well as compositions of matter related thereto.

[0007] In an aspect of the disclosure, a method for purifying Ribulose-l,5-bisphosphate Carboxylase / Oxygenase (RuBisCO) from plant cells is provided. This method comprises lysing a plant cell to make a plant cell lysate that contains a precipitate and supernatant, separating the precipitate from the supernatant, and removing the supernatant. Afterward, a RuBisCO linker protein is added to the supernatant, where it binds to at least two separate molecules of RuBisCO, forming a RuBisCO-linker complex that induces phase separation. The resulting phase-separated solution contains a RuBisCO-linker precipitate, which is then separated from the supernatant. The precipitate is resuspended to dissociate the RuBisCO and linker protein, and purified RuBisCO is isolated. The purified RuBisCO comprises eight large subunits (RbcL) and eight small subunits (RbcS), and either or both subunits contain one or more RuBisCO binding interfaces.

[0008] In another aspect of the disclosure, a method is provided for isolating recombinant RuBisCO expressed in a prokaryotic or eukaryotic cell. The method includes expressing recombinant RbcL and RbcS in a host cell, lysing the cell, adding a RuBisCO linker protein to bind at least two molecules of recombinant RuBisCO, thereby forming a phase-separated, turbid solution. The precipitate from this phase separation is isolated, resuspended, and the purified recombinant RuBisCO is obtained. The recombinant RbcL, RbcS, or both, comprise one or more RuBisCO binding interfaces, and these binding interfaces are present on separate RuBisCO molecules.

[0009] In yet another aspect of the disclosure, a method is provided for isolating recombinant RuBisCO produced in a cell-free expression system. Recombinant RbcL and RbcS are expressed in a cell-free system, combined with a RuBisCO linker protein that binds at least two separate molecules of RuBisCO. This interaction also drives phase separation, allowing for isolation of the precipitate, resuspension, and recovery of purified RuBisCO. The recombinant subunits contain one or more binding interfaces on separate RuBisCO molecules.

[0010] In various embodiments of these aspects:

[0011] The purified RuBisCO comprises eight large (RbcL) and eight small (RbcS) subunits, and is present in natively folded form.

[0012] The RbcL subunit may have one or more RuBisCO binding interfaces, and similarly, the RbcS subunit may contain these binding interfaces, either alone or in combination.

[0013] The RuBisCO linker protein may contain at least two RuBisCO recognition domains separated by a peptide linker, which can range from at least 20 residues up to about 250 residues, with specific embodiments including 25-45 residues.

[0014] The binding interfaces may comprise sequences at least about 80% identical to key residues of Synechococcus elongatus or tobacco RbcL and RbcS, and may involve shared or distributed binding sites among subunits.

[0015] The method may employ low salt conditions for linker binding and phase separation, and high salt conditions for dissociation of the linker complex. Low salt may be less than about 250 mM ionic strength, while high salt may be greater than about 500 mM ionic strength.

[0016] Additional purification steps before linker addition may be included, such as chromatography, electrophoresis, ultrafiltration, or dialysis.

[0017] Both wildtype and non-wildtype sequences of RbcL and RbcS are contemplated, with embodiments covering both natural and genetically modified plant, prokaryotic, and eukaryotic cells, or cell lines expressing non-native RuBisCO or subunits.

[0018] Expression vectors comprising TobL35S DNA or protein sequences, purified recombinant RuBisCO with wildtype or non-native subunits, and genetically modified cell lines expressing various RuBisCO forms are also included as aspects of the invention.

[0019] The formation of a RuBisCO-linker network, comprising interconnected complexes, is also described.

[0020] In another aspect of the disclosure, a RuBisCO-linker network composition is provided, including (i) a plurality of RuBisCO proteins; and (ii) a multivalent RuBisCO linker protein comprising at least two RuBisCO recognition domains, in which each RuBisCO protein is non-covalently bound to one or more RuBisCO linker proteins, and each RuBisCO linker protein is non-covalently bound to two or more RuBisCO proteins, thereby forming an interconnected network capable of liquid-liquid phase separation.

[0021] In some embodiments of the RuBisCO-linker network composition, the composition forms a phase-separated precipitate under low salt conditions; which, in embodiments, can dissociate under high salt conditions to release purified RuBisCO.

[0022] The methods and compositions disclosed herein have significant commercial utility for production of RuBisCO as a sustainable protein source for human consumption. RuBisCO represents an abundant, nutritionally complete plant protein with favorable functional properties for food formulations (e.g., foaming, gelation, emulsification). The disclosed LLPS-based purification enables scalable, cost-effective extraction from agricultural biomass (e.g., leaves, plant manufacturing byproducts) or recombinant production systems.BRIEF DESCRIPTION OF THE DRAWINGS

[0023] FIG. 1A to FIG. 1C show the biophysical and nutritional qualities of RuBisCO. FIG. 1A:Side and top view representation of RuBisCO from plants. The enzyme is a complex of eight large subunits (RbcL; 53 kDa) and eight small subunits (RbcS; 15 kDa) of approximately 540 kilodaltons (kDa). FIG. IB: The amino acid profile of RuBisCO (RbcL and RbcS) from 15 phylogenetically diverse crop varieties. FIG. 1C: Proportional of essential human dietary aminoacids found in RuBisCO compared to the main protein components of wheat grain (Glutenin), pea seeds (Vicilin), and egg white (Ovalbumin).

[0024] FIG.2A to FIG.2D shows the principles of LLPS and the mechanism of LLPS as it related to RuBisCO. FIG. 2A: Condensation of protein (dark color spheres) into a single phase excludes surrounding materials (light color spheres), a core need for RuBisCO purification from leaf and cellular impurities. FIG. 2B: LLPS leads to turbidity in solution with increasing condensate mass, enabling isolation of the heavier (higher molecular weight) condensate fraction. FIG. 2C:RuBisCO LLPS is mediated by network formation with linker proteins that drive the protein into a liquid condensate of much greater molecular weight (kiloDaltons; kDa) compared to the individual constituents (Adapted from Blikstad et al. 2022). FIG.2D: LLPS of 6301LS RuBisCO is driven by the interaction of the RuBisCO binding domain (SSUL) of linker protein (herein referred to as “M35”) with residues on three independent subunits of RuBisCO (Adapted from Wang et al. 2019).

[0025] FIG. 3A to FIG. 3D shows LLPS excludes other biomolecules and generates separatable native RuBisCO protein. FIG. 3A: SDS-PAGE analysis of protein from E. coli expressing Synechococcus sp. PCC6301 RuBisCO (6301LS). T - Total protein; S - Soluble protein; M - Size marker. Soluble E. coli protein when incubated with increasing M35 to the final specified micromolar (pM) concentration. Protein recovered from the pellet following centrifugation contains stoichiometrically similar amounts of both RbcL (52.4 kDa) and M35 (36 kDa). FIG.3B:Native PAGE analysis of protein recovered from the LLPS following incubation with M35 (0 - 4 pM) shows the major protein component migrating identically to pure Synechococcus RuBisCO at the lowest linker LLPS concentration, demonstrating the effectiveness of this new purification approach. FIG. 3C: Purified M35 mediates LLPS of purified Synechococcus sp. PCC6301 RuBisCO, but does not interact with purified Spinach (Spinacia oleracea) RuBisCO in vitro. FIG.3D: Residues at an interface (IF) region on RbcS (IF 1 ) and RbcL (IF2) are not conserved between RbcL and RbcS of Synechococcus (Syn RbcS: AGFDNIKQCQT (SEQ ID NO:1); and Syn RbcL: YTPDYTPKDTD (SEQ ID NO:2) and Spinach RuBisCO (labeled “Plant” RbcS: IGFNDKREVQC (SEQ ID NO:3); and “Plant” RbcL: YPTEYETLDTD (SEQ ID NO:4)).

[0026] FIG. 4A to FIG. 4E show that modified tobacco (Nicotiana tabacuni) Rubisco TobL3?S can be purified using LLPS with M35. FIG. 4A: Native-PAGE analysis of protein from E. coll expressing TobL33S incubated with increasing pure M35 (pM). FIG. 4B: Pellet formation was only observed for TobL35S following centrifugation of 1 mL total soluble bacterial protein incubated with 5 pM of pure M35. FIG. 4C: Pellet of precipitate from 6 mL of total soluble bacterial protein containing TobL33S produced as in (B) prior to solvation in buffer containing 0.5 M NaCl and subjection to size exclusion chromatography (SEC), elution profile shown. FIG. 4D:SDS-PAGE analysis of TobL3?S soluble (S) or pellet (P) protein fractions following incubation and with or without M35 and centrifugation at 5,000 g. RbcL (53 kDa), RbcS (15 kDa) and M35 (36 kDa). Nb leaf; Nicotiana benthamiana leaf used as a RuBisCO size control. FIG. 4E: Native-PAGE analysis of 2 pM purified TobL35S, following SEC purification shown in (C), and purified TobL35S incubated with M35 at a final concentration of 0.25-10 pM. Formation of higher molecular weight TobL35S-M35 as displayed including as retention in the gel well.

[0027] FIG. 5A to FIG. 5C show TobL33S CO2 fixation activity and photosynthetic modelling.FIG 5A: RuBisCO abundance in the total soluble protein of bacteria expressing TobL35S or wildtype TobLS as measured by [14C]-CABP binding. FIG 5B: Michaelis-Menten plot of the CO2 fixation activity of TobL3:>S and derived maximal CO2 fixation rate and affinity for substrate CO2 in an atmosphere containing 21% oxygen (air). FIG 5C: Modelling at 25 °C of the influence of TobL35S RuBisCO on leaf CO2 assimilation rates using a model of C3 photosynthesis. Modelling is performed for TobLS using the equations and parameters in Sharwood, 2017, and a leaf RuBisCO content set to 20 pmol active sites per nr2. Modelled performance of TobL35S at an equivalent content normally found in N. tabacum leaves is shown in the lowest curve, with performance shown under an elevated biogenesis scenario of 60 pmol active sites per nr2shown in the top curve. Modelling at 40 pmol active sites per nr2gave an estimated leaf CO2 assimilation rate similar to wild-type (plot not shown; WT shown as middle curve). Model assumes no change in specificity factor (Sc / o) for TobL35S.

[0028] FIG. 6 shows ranking of the disclosed binders in complex with wild type tobacco RuBisCO by AlphaFold3 (AF3). IPAE (interface predicted aligned error) is the score AF3 used to evaluate how confident AF3 is in the relative position of two interface residues within the predicted structure between binder and target. A lower IPAE score indicates a lower predicted error of therelative orientation of those interface residues. IPTM (interface predicted template modeling) is the score in AF3 that evaluates the accuracy of the predicted protein-protein interfaces. A higher IPTM score indicates a more confident prediction of the interface structure. Thus, on the graph, the better designs are predicted to be at the bottom right which have low IPAE score and high IPTM score. Indeed, our positive control, a crystal structure of Synechococcus sp. PCC6301 RuBisCO and the SSUL1 binding domain from M35 (PDB ID: 6HBC) was used as positive control and is located at the bottom right of the figure; whereas, tobacco RuBisCO (PDB ID: 4RUB) aligned onto the Synechococcus sp. PCC6301 RuBisCO-SSULl structure, which served as the negative control, is located at the top left.

[0029] FIG. 7A and FIG. 7B show a high throughput screening assay. FIG. 7A: The luciferase assay consisted of two components. The first component features a fusion between the RuBisCO small subunit and a smBiT tag. The second component features a fusion between a RuBisCO binding domain and LgBiT. The LgBiT and smBiT together complement to form NanoLuc, a shrimp luciferase enzyme that will emit light when exposed to furimazine. FIG.7B: The luciferase high throughput screening assay was initiated by expressing the two components and extracting the soluble lysate. Fluorofurimazine (5 pM) was added into the LgBiT-binder soluble lysate to incubate with LgBiT. A plate reader was used to measure the background, which shows the solubility of different LgBiT-binders. smBiT-TobLS-WT was then added into each well and a second read was taken. This bioluminescence read was further normalized to the solubility of different LgBiT-binders by subtracting the background reads and dividing by the background reads. RuBisCO-binder binding affinity is expressed as a fold change in luminescence.

[0030] FIG. 8 shows a validation of the top 12 binders found in the high throughput assay. The top 12 binders were sequenced and expressed in separate flasks in triplicate to eliminate the possible error in the high throughput assay. SSULl-TobL35S interaction was used as a positive control and SSULl-TobLS WT interaction was used as a negative control. Among the top 12 binders, binder 1737 demonstrated the highest average fold change of luminescence.

[0031] FIG.9A and FIG.9B show native PAGE gel shift analysis of 1737 linker efficiency. FIG.9A: The SynM35-TobL35S interaction is quite efficient with clear higher molecular weight complexes formed when 1 pM SynM35 linker was mixed with 2 pM TobL35S. FIG. 9B: 1737linker was less efficient with higher molecular weight complexes formed when 10 pM 1737 linker was mixed with 2 pM TobLS WT.

[0032] FIG. 10A and FIG. 10B show Luciferase assays for 1737 A29R R31H characterization.FIG. 10A: SSUL1, SSUL3 (negative control), 1737_WT and 1737_A29R_R31H with LgBiT interacting with smBiT-TobLS WT. 1737 A29R R31H demonstrated significantly higher fold change of luminescence to smBiT-TobLS WT than 1737 WT. FIG. 10B: shows SSUL1 (positive control), SSUL3 (negative control), 1737_WT and 1737_A29R_R31H with LgBiT interacting with smBiT-TobL35S. 1737 A29R R31H demonstrated essentially equivalent binding to smBiT-TobL35S as 1737_WT, both significantly stronger than the SSUL1 positive control. ***: p< 0.0001; ns: p > 0.5.

[0033] FIG. 11 shows SDS-PAGE analysis of Nicotiana benthamiana leaf lysate mixed with purified 3x1737 ARRH linker. Supernatant or pellet protein fractions following incubation with purified 3x1737 ARRH linker and centrifugation at 10,000 g. RbcL (53 kDa), RbcS (15 kDa) and 3x1737 ARRH linker (38 kDa).

[0034] FIG. 12 shows SDS-PAGE analysis of Nicotiana benthamiana leaf lysate mixed with purified 5x1737 ARRH linker. Supernatant or pellet protein fractions following incubation with purified 5x1737 ARRH linker and centrifugation at 10,000 g. RbcL (53 kDa), RbcS (15 kDa) and 5x1737 ARRH linker (64 kDa).

[0035] FIG. 13 shows the sequence location of residues substituted in TobL35S. BLOSUM62 global protein sequence alignment of RbcL from Synechococcus elongatus sp. PCC6301 (6301L) and Nicotiana tabacum (TobL, WT). Residues important to the interaction with M35 are shown by black filled triangles and those substituted in TobL35S to the residue found in 6301L are indicated with hollow triangles. The remaining two important residues (D93 and N94) are conserved in RbcS between these species. TobL35S is >98% identical to the naturally occurring form. RbcL secondary structure is indicated based on the crystal structure of N. tabacum RuBisCO (PDB code: 3RUB).DESCRIPTION

[0036] Disclosed herein are methods, compositions, and systems for efficient purification of native and recombinant Ribulose- 1,5 -bisphosphate Carboxylase / Oxygenase (“RuBisCO”; “Rubisco”) proteins from complex biological materials, including plant and bacterial cellular lysates. The methods utilize liquid-liquid phase separation driven by engineered linker proteins to selectively isolate intact RuBisCO complexes while preserving native folding and assembly. The disclosure provides scalable protocols for phase separation-based purification, linker engineering, and recombinant expression, enabling recovery of functional RuBisCO for applications in food, biotechnology, and protein manufacturing. Also, the disclosure provides methods for salt-mediated dissociation and further processing to achieve high-purity RuBisCO suitable for nutritional and industrial use.

[0037] A number of terms are introduced below, which are used to describe the invention of the present disclosure. In instances where a technical or scientific term is not specifically defined herein, they will have the same meaning as commonly understood by one of ordinary skill in the art. For example, any nomenclatures used in connection with, and techniques of, cell culture, molecular biology, microbiology, genetics, and protein and nucleic acid chemistry described herein are well known and commonly used in the art. In case of conflict, the present disclosure, including definitions, will control. Exemplary methods and materials are described below, although methods and materials similar or equivalent to those described herein can be used in practice or testing of the embodiments and aspects described herein.

[0038] The term “RuBisCO binding interface”, as used herein, refers to an area on the three-dimensional surface of a native RuBisCO protein (i.e., globular complex of sixteen protein chains composed of two subunits in equal stoichiometry (eight RbcL and eight RbcS subunits)) that is recognized and potentially bound by a RuBisCO recognition domain on a RuBisCO linker protein.

[0039] The term “RuBisCO recognition domain”, as used herein, refers to a domain or area on the three-dimensional surface of a RuBisCO linker protein that recognizes one or more RuBisCO binding interfaces on a RuBisCO protein. Binding of a RuBisCO recognition domain to a RuBisCO binding interface forms a RuBisCO-linker complex. Preferably, a RuBisCO linker protein has at least two RuBisCO recognition domains such that it can bind to at least two separate molecules of RuBisCO.

[0040] The term “valency”, as used herein, refers to the number of RuBisCO binding interfaces present on the surface of the native RuBisCO protein capable of occupancy by a RuBisCO recognition domain. Preferably, a native RuBisCO protein has a valency of two or greater to facilitate an interconnected network of RuBisCO and RuBisCO linker proteins that results in liquid-liquid phase separation of RuBisCO and the RuBisCO linker proteins.

[0041] The terms “RuBisCO linker” and “RuBisCO linker protein” refers to a multivalent linker molecule (e.g., protein) that is capable of binding to at least two RuBisCO through weak bonds, e.g., hydrogen bonds and van der Waals interactions.

[0042] The term “RuBisCO-linker network” refers to an interconnected matrix comprising a plurality of RuBisCO binding interfaces across a plurality of RuBisCO proteins bound by a plurality of RuBisCO recognition domains across a plurality of RuBisCO linker proteins through weak bonds, e.g., hydrogen bonds and van der Waals interactions.

[0043] The term “liquid-liquid phase separation” or “LLPS” refers to a process by which a uniform mixture of dissolved biomolecules spontaneously segregates into two or more distinct liquid phases with different compositions (most often an aqueous phase of low biomolecule concentration and a biomolecular condensate of high biomolecule concentration (FIG. 2A)), driven by weak multivalent interactions between proteins, nucleic acids, or other molecules. In biological systems, LLPS enables the formation of membraneless compartments (biomolecular condensates) such as nucleoli and stress granules, coordinately organizing cellular biochemical reactions, structural assemblies, or macromolecular complexes without lipid membranes. LLPS relies on thermodynamic principles of demixing, where changes in protein concentration, ionic strength, temperature, or sequence composition create localized environments that concentrate specific factors, often critical for regulatory, structural, or catalytic functions within cells.

[0044] Sufficiently large and numerous condensates will diffract visible light often turning a transparent single phase into a turbid mixture (FIG. 2B). Protein LLPS occurs when the local protein concentration becomes sufficiently high that other molecules, including the solvent, are largely excluded from intermolecular spaces.

[0045] The term "peptide", "polypeptide", "protein", and the like, as used herein, refers to a molecule that is formed using naturally occurring L-amino acids or analogs thereof, like D-amino acids, or N-alkylated amino acids, or the like. Preferred amino acids are selected from the group consisting of Ala, Arg, Asn, Asp, Cys, Glu, Gin, Gly, He, Leu, Lys, Met, Phe, Pro, Ser, Thr, Trp, Tyr, and Vai. Additionally, modifications such as fluorescence dyes or biotin are also contemplated.

[0046] "Functional derivatives" of proteins are also contemplated, in which a functional derivative refers to a "chemical derivative," "fragment," "polymorph" or "variant" of the polypeptide or nucleic acid of the invention. A functional derivative retains at least a portion of the function of the protein, which permits its utility in accordance with the invention. It is well known in the art that, due to the degeneracy of the genetic code, numerous different nucleic acid sequence can code for the same amino acid sequence. It is also well known in the art that conservative changes in amino acid can be made to arrive at a protein or polypeptide that retains the functionality of the original. In both cases, all permutations are intended to be covered by this disclosure.

[0047] Another functional derivative intended to be within the scope of the present invention is a "variant" polypeptide, which either lacks one or more amino acids or contains additional or substituted amino acids relative to the native polypeptide. Such variants having added, substituted and / or additional amino acids retain the functional portion of the original polypeptide. A functional derivative of a protein with deleted, inserted and / or substituted amino acid residues may be prepared using standard techniques well-known to those of ordinary skill in the art (e.g., site-directed mutagenesis). Alternatively, proteins with amino acid deletions, insertions and / or substitutions may be conveniently prepared by direct chemical synthesis, using methods well-known in the art.

[0048] The terms "identity", "identical", "similar", "similarity", "homology", "homologous", and the like, refer to the "likeness" or "sameness" of two or more sequences, e.g., between two nucleic acid sequences or two peptide sequences, often expressed as a percentage. Sequence identity is the amount of characters which match exactly between two different sequences, where gaps are not counted and the measurement is relational to the shorter of the two sequences. This has the effectthat sequence identity is not transitive, i.e. if sequence A=B and B=C then A does not necessarily equal C (in terms of the identity distance measure): Consider exemplary sequences A: AAGGCTT, B: AAGGC, and CAAGGCAT. Here identity(A, B)=100% (5 identical nucleotides / min(length(A),length(B))). Identity(B, C)=100%, but identity(A, C)=85% ((6 identical nucleotides / 7)). So 100% identity does not necessarily indicate two sequences are the same. Sequence similarity accounts for sequence identity and conservative substitutions with positive scores in substitution matrices. E.g., a leucine to isoleucine substitution would receive partial "similarity" credit while not for a sequence identity calculation. Methods for aligning sequences for comparison are well-known in the art, and a detailed consideration of sequence alignment methods and homology calculations can be found in, e.g., Altschul et al. (1990) J. Mol. Biol. 215:403-10.

[0049] The term "vector" is well known in the art, and as used herein refers to a nucleic acid molecule, e.g. double-stranded DNA, which may have inserted into it a nucleic acid sequence according to the present invention. A vector is suitably used to transport an inserted nucleic acid molecule into a suitable host cell. A vector typically contains all of the necessary elements that permit transcribing the insert nucleic acid molecule, and, preferably, translating the transcript into a polypeptide. A vector typically contains all of the necessary elements such that, once the vector is in a host cell, the vector can replicate independently of, or coincidental with, the host chromosomal DNA; several copies of the vector and its inserted nucleic acid molecule may be generated.

[0050] Included within the scope of the invention are functional equivalents of the herein-described isolated nucleic acid molecules. The degeneracy of the genetic code permits substitution of certain codons by other codons that specify the same amino acid and hence would give rise to the same protein. The nucleic acid sequence can vary substantially since, with the exception of methionine and tryptophan, the known amino acids can be coded for by more than one codon. The encoded amino acid sequence thereof would, however, be preserved.

[0051] In addition, the nucleic acid sequence may comprise a nucleotide sequence which results from the addition, deletion or substitution of at least one nucleotide to the 5’ end and / or the 3’ end, provided that its addition, deletion or substitution does not alter the amino acid sequence described herein, which is encoded by the nucleotide sequence. For example, the nucleic acid molecule ofthe present invention may have restriction endonuclease recognition sites added to its 5’ end and / or 3’ end.

[0052] Further, it is possible to delete codons or to substitute one or more codons with codons other than degenerate codons to produce a structurally modified polypeptide, but one which has substantially the same utility or activity as the polypeptide produced by the unmodified nucleic acid molecule. As recognized in the art, the two polypeptides are functionally equivalent, as are the two nucleic acid molecules that give rise to their production, even though the differences between the nucleic acid molecules are not related to the degeneracy of the genetic code.

[0053] The term "fragment" is used to indicate a polypeptide derived from another polypeptide having a length less than the full-length polypeptide from which it has been derived. Such a fragment may, for example, be produced by proteolytic cleavage of the full-length protein. Such a fragment may also be obtained recombinantly by appropriately modifying the DNA sequence encoding the proteins to delete one or more amino acids at one or more sites of the C-terminus, N-terminus, and / or within the native sequence. Such fragments retain the functional portion of the native protein.

[0054] All methods described herein can be performed in a suitable order unless otherwise indicated herein or otherwise clearly contradicted by context. The use of any and all examples, or exemplary language (e.g., "such as"), is intended merely to better illustrate the invention and does not pose a limitation on the scope of the invention unless otherwise claimed. No language in the specification should be construed as indicating any non-claimed element as essential to the practice of the invention as used herein. Unless defined otherwise, technical and scientific terms used herein have the same meaning as is commonly understood by one of skill in the art of this disclosure.

[0055] Furthermore, the disclosure encompasses all variations, combinations, and permutations in which one or more limitations, elements, clauses, and descriptive terms from one or more of the listed claims are introduced into another claim. For example, any claim that is dependent on another claim can be modified to include one or more limitations found in any other claim that is dependent on the same base claim. Where elements are presented as lists, e.g., in Markush groupformat, each subgroup of the elements is also disclosed, and any element(s) can be removed from the group.

[0056] Recitation of ranges of values herein are merely intended to serve as a shorthand method of referring individually to each separate value falling within the range, unless otherwise indicated herein, and each separate value is incorporated into the specification as if it were individually recited. For example, if a size range is stated as 1 nm to 100 nm (or concentrations, degrees, mass amounts, and the like), it is intended that values such as 2 nm to 90 nm, 10 nm to 70 nm, 30 nm to 95 nm, 75 nm to 100 nm, or 2 nm to 27 nm, etc., are expressly enumerated in this specification. These are only examples of what is specifically intended, and all possible combinations of numerical values between and including the lowest value and the highest value enumerated are to be considered to be expressly stated in this disclosure.

[0057] Furthermore, when "about", "approximately" and / or "substantially" is / are utilized to describe a value, this is meant to encompass minor variations (up to + / - 10%) from the stated value. Where no stated value is provided, an element described as "substantially" means at least about 60%, 70%, 80%, 90%, 95%, 99%, or more of the element, as is logically coherent within in the context. Unless specifically stated to the contrary, for ranges specified using "about" language, the about applies to both ends of the recited range whether specified or not. For example, "between about 10 mM and 10 pM" is equivalent to "between about 10 mM and about 10 pM".

[0058] As used herein, the terms "comprising", "having", "including", and "containing" are to be construed as open-ended terms (i.e., meaning "including, but not limited to") unless otherwise noted.

[0059] When introducing elements of the present disclosure or the aspects and embodiment thereof, the articles "a," "an," and "the" are intended to mean that there are one or more of the elements. Similarly, the adjective "another," when used to introduce an element, is intended to mean one or more elements. The term “exemplary” is not intended to be construed as a superlative example but merely one of many possible examples.

[0060] The phrase "and / or," as used herein in the specification and in the claims, should be understood to mean "either or both" of the elements so conjoined, i.e., elements that areconjunctively present in some cases and disjunctively present in other cases. Multiple elements listed with "and / or" should be construed in the same fashion, i.e., "one or more" of the elements so conjoined. Other elements may optionally be present other than the elements specifically identified by the "and / or" clause, whether related or unrelated to those elements specifically identified. Thus, as a non-limiting example, a reference to "A and / or B", when used in conjunction with open-ended language such as "comprising" can refer, in one embodiment, to A only (optionally including elements other than B); in another embodiment, to B only (optionally including elements other than A); in yet another embodiment, to both A and B (optionally including other elements); etc.

[0061] As used herein in the specification and in the claims, the phrase "at least one," in reference to a list of one or more elements, should be understood to mean at least one element selected from anyone or more of the elements in the list of elements, but not necessarily including at least one of each and every element specifically listed within the list of elements and not excluding any combinations of elements in the list of elements. This definition also allows that elements may optionally be present other than the elements specifically identified within the list of elements to which the phrase "at least one" refers, whether related or unrelated to those elements specifically identified. Thus, as a nonlimiting example, "at least one of A and B" (or, equivalently, "at least one of A or B," or, equivalently "at least one of A and / or B") can refer, in one embodiment, to at least one, optionally including more than one, A, with no B present (and optionally including elements other than B); in another embodiment, to at least one, optionally including more than one, B, with no A present (and optionally including elements other than A); in yet another embodiment, to at least one, optionally including more than one, A, and at least one, optionally including more than one, B (and optionally including other elements); etc.

[0062] The phrase "one or more," as used herein, means at least one, and thus includes individual components as well as mixtures / combinations of the listed components in any combination.

[0063] As used herein, "optional" or "optionally" means that the subsequently described event or circumstance may or may not occur, and that the description includes instances where said event or circumstance occurs and instances where it does not.

[0064] It should also be understood that, in certain methods described herein that include more than one step or act, the order of the steps or acts of the method is not necessarily limited to the order in which the steps or acts of the method are recited unless the context indicates otherwise. Furthermore, the terms first, second, etc., as used herein are not meant to denote any particular ordering, but simply for convenience to denote a plurality of, for example, layers.

[0065] Moving to the specifics of the subject matter, disclosed herein are methods of purifying RuBisCO utilizing Liquid-Liquid Phase Separation (LLPS). Such methods are described in more detail below.

[0066] Formation of Multivalent RuBisCO-Linker Networks

[0067] A key aspect of the disclosure relates to the formation of interconnected RuBisCO-linker networks that drive liquid-liquid phase separation (LLPS) for selective RuBisCO purification. These networks are supramolecular assemblies formed through non-covalent, multivalent binding interactions between RuBisCO linker proteins and (native or recombinant) RuBisCO complexes. Unlike traditional fusion-tag approaches that require covalent attachment and subsequent proteolytic cleavage, the RuBisCO-linker networks of the present disclosure maintain reversible, non-covalent interactions throughout the purification process, thereby preserving the native structure and function of RuBisCO.

[0068] Each RuBisCO linker protein comprises at least two RuBisCO recognition domains capable of simultaneously binding separate RuBisCO molecules. When added to a solution containing multiple RuBisCO complexes, the linker proteins act as molecular bridges, connecting individual RuBisCO proteins into higher-order assemblies. Because each RuBisCO complex (comprising eight large subunits and eight small subunits arranged as an L8S8structure, FIG. 1A) presents multiple potential binding sites, and each linker protein can engage multiple RuBisCO molecules, the system forms an extensive three-dimensional network.

[0069] This network architecture is fundamentally different from simple binary protein -protein complexes. Rather than forming discrete 1:1 or 2:1 stoichiometric complexes, the multivalent interactions generate a percolating network where RuBisCO molecules are interconnected through multiple linker proteins, and linker proteins simultaneously engage multiple RuBisCO molecules.The resulting supramolecular structure can be conceptualized as a mesh or lattice wherein RuBisCO complexes serve as multivalent nodes and linker proteins serve as multivalent edges.

[0070] Thermodynamic Basis for Liquid-Liquid Phase Separation

[0071] The formation of RuBisCO-linker networks creates conditions favorable for liquid-liquid phase separation, a thermodynamic process by which a homogeneous solution spontaneously demixes into two distinct liquid phases with different protein concentrations. LLPS is driven by the cumulative weak interactions within the network — including the specific binding interactions between linker recognition domains and RuBisCO binding interfaces, as well as potential nonspecific interactions between protein surfaces.

[0072] The multivalent nature of both components (multivalent linkers and multivalent RuBisCO binding sites) is necessary for achieving LLPS. Multivalency amplifies binding avidity, where multiple weak interactions act cooperatively to create effectively strong association. This principle is well-established in biological phase separation systems such as the formation of membraneless organelles and stress granules. In the context of RuBisCO purification, multivalency enables network formation at physiologically relevant protein concentrations and creates a sufficient thermodynamic driving force for phase separation.

[0073] Upon formation, the RuBisCO-linker network undergoes LLPS to generate two coexisting liquid phases: (1) a protein-dense phase (condensate) enriched in RuBisCO-linker networks, which appears as visible turbidity and ultimately sediments as a precipitate upon centrifugation; and (2) a protein-dilute phase (supernatant) depleted of RuBisCO but containing non-binding proteins and other cellular components. This phase separation effectively partitions RuBisCO into the dense phase while excluding contaminant proteins that do not interact with the linker, thereby achieving selective purification.

[0074] Structural and Compositional Features of RuBisCO-Linker Networks

[0075] RuBisCO-linker network compositions comprise at minimum: (i) a plurality of RuBisCO molecules, each comprising eight large subunits (RbcL) and eight small subunits (RbcS) in a LsSsquaternary structure; and (ii) multivalent RuBisCO linker proteins, each comprising at least two RuBisCO recognition domains connected by linkers, e.g. peptide linkers.

[0076] The stoichiometry and architecture of the network can vary depending on several factors including the concentration of RuBisCO, the concentration of linker protein, the valency (number of recognition domains) of the linker protein, the affinity of each recognition domain for RuBisCO, the spatial arrangement of binding sites on RuBisCO, and solution conditions such as ionic strength, pH, and temperature. In typical implementations, linker proteins are added at sub-stoichiometric to equimolar ratios relative to RuBisCO, allowing each linker to bridge multiple RuBisCO molecules.

[0077] The network exhibits dynamic properties wherein individual linker-RuBisCO interactions are reversible and exchange over time, but the overall network structure is stabilized by the collective avidity of multiple simultaneous interactions. This dynamic nature is advantageous for purification applications because it allows the network to reorganize, potentially excluding weakly-associated contaminant proteins while retaining RuBisCO through specific high-affinity interactions.

[0078] Molecular Basis of Linker-RuBisCO Binding

[0079] The binding interfaces between RuBisCO linker proteins and RuBisCO complexes are critical determinants of network formation and LLPS efficiency. In naturally-derived linker systems, such as those based on the cyanobacterial carboxysome shell protein CcmM, the recognition domains (e.g., M35, residues 185-219 of CcmM) bind specifically to the large subunit (RbcL) and small subunit (RbcS) of cyanobacterial RuBisCO. Structural studies have revealed that these interactions involve shape complementarity and specific amino acid contacts at the RuBisCO surface.

[0080] For plant RuBisCO purification, two strategies are disclosed to achieve linker binding. First, plant RuBisCO large subunits can be engineered to introduce binding interfaces compatible with natural cyanobacterial linkers. The TobL35S mutant, comprising six mutations (Q30T, T31P, S76D, R86H, D351E, F353H) in tobacco RbcL, successfully binds to the M35 domain fromSynechococcus sp. PCC6301 CcmM, enabling efficient LLPS-based purification of the mutant RuBisCO expressed in E. coli (Example 2, FIG. 4A-E, 5A-C).

[0081] Second, synthetic RuBisCO linkers can be computationally designed to bind wild-type plant RuBisCO without requiring RuBisCO engineering. These synthetic recognition domains can be generated through, e.g., machine learning-assisted protein design workflows that optimize binding interfaces for plant RuBisCO structures (FIG. 6-10). Synthetic binders such as the 1737 scaffold and optimized variants (e.g., 1737_A29R_R31H) successfully bind wild-type tobacco RuBisCO and, when configured as multivalent linkers with 3-5 tandem repeats, drive phase separation of RuBisCO from plant leaf lysates (FIG. 11-12).

[0082] The binding interfaces on RuBisCO are spatially distributed across the L8S8complex, with multiple potential binding sites accessible on the surface of each large subunit, and potentially on small subunits as well. This spatial arrangement is important for network formation, as it allows a single RuBisCO complex to be bound by multiple linker proteins simultaneously, and enables geometric configurations where linkers can bridge between separate RuBisCO molecules. Structural modeling indicates that the M35 binding sites are located on the equatorial region of RuBisCO, providing multiple symmetrically-arranged binding locations per complex (FIG. 13).

[0083] Salt-Dependent Reversibility and Network Dissociation

[0084] An important design feature of RuBisCO-linker networks is their salt-dependent reversibility. Network formation and LLPS are promoted under low ionic strength conditions (typically less than 250 mM ionic strength, and preferably about 50-150 mM), where electrostatic interactions between linker recognition domains and RuBisCO binding interfaces are favored. Under these conditions, the cumulative binding interactions stabilize the network and drive phase separation.

[0085] Conversely, the network dissociates under high ionic strength conditions (typically greater than 500 mM ionic strength, such as 500-1000 mM NaCl), where electrostatic screening weakens protein-protein interactions. This salt-dependent dissociation enables controlled release of RuBisCO from the linker network by simple buffer exchange, eliminating the need for proteolytic cleavage or harsh chemical treatments that might denature the protein. The released RuBisCOremains in its native L8S8quaternary structure, as confirmed by native PAGE (FIG. 3B, 4A, 4E) and size exclusion chromatography (FIG. 4C), and retains full enzymatic activity for CO2 fixation (FIG. 5B)

[0086] The salt-dependent reversibility also enables recycling of linker proteins for subsequent purification cycles, improving the economic feasibility of large-scale applications. After dissociation, linker proteins can be separated from RuBisCO through size exclusion chromatography or other standard protein separation methods, re-equilibrated in low-salt buffer, and reused.

[0087] Visualization and Characterization of RuBisCO-Linker Networks

[0088] RuBisCO-linker network formation and LLPS can be readily visualized and characterized using multiple complementary techniques. Visual inspection provides the most immediate confirmation: solutions containing RuBisCO and linker protein that were initially clear become visibly turbid or cloudy upon network formation, indicating light scattering by phase-separated droplets. Upon centrifugation (typically 3,000-20,000 xg), the phase-separated condensate sediments to form a visible pellet, while the depleted supernatant remains relatively clear.

[0089] Protein composition analysis by SDS-PAGE demonstrates enrichment of RuBisCO in the pellet fraction relative to the supernatant and starting material (FIG.3A, 4D, 11, 12). Characteristic RbcL bands (approximately 55 kDa) and RbcS bands (approximately 12-15 kDa) appear prominently in pellet lanes, often accompanied by visible linker protein bands at the expected molecular weight. Comparison of band intensities across fractions provides semi-quantitative assessment of purification efficiency.

[0090] Native PAGE analysis confirms that RuBisCO maintains its native quaternary structure throughout LLPS and purification (FIG.3B, 4A, 4E). Intact RuBisCO complexes migrate as bands at the expected position for L8S8assemblies, distinct from dissociated subunits or aggregated protein. The presence of intact complexes in the final purified RuBisCO (after dissociation) demonstrates that network formation and dissociation do not disrupt RuBisCO assembly.

[0091] Size exclusion chromatography (SEC) provides high-resolution assessment of oligomeric state and purity (FIG. 4C). Native RuBisCO elutes as a sharp peak at the elution volume corresponding to its approximately 540 kDa molecular weight, well-separated from smaller contaminant proteins and from large aggregates that would elute in the void volume. SEC analysis of purified RuBisCO following LLPS-based purification shows a single dominant peak, confirming both high purity and native assembly.

[0092] Functional activity assays provide the ultimate validation that purified RuBisCO retains native properties. CO2 fixation activity, measured using radiolabeled14CO2 incorporation assays or coupled enzyme assays monitoring NADH oxidation, confirms that the purified enzyme is catalytically competent (FIG. 5B). Activity levels approaching or exceeding those of wild-type RuBisCO demonstrate that the LLPS-based purification process does not cause denaturation or inactivation.

[0093] Applications and Advantages of RuBisCO-Linker Network Systems

[0094] The RuBisCO-linker network approach offers significant advantages over conventional protein purification methods. First, the method is rapid, typically requiring only minutes for network formation and phase separation, compared to hours or days for multi-step chromatography workflows. Second, the method is scalable, as LLPS can be induced in vessels ranging from microliters (for analytical applications) to liters or larger (for industrial production). Third, the mild conditions (physiological pH, moderate temperature, avoidance of organic solvents or denaturants) preserve protein integrity, making the purified RuBisCO immediately suitable for food applications or functional studies.

[0095] Fourth, the non-covalent nature of linker-RuBisCO interactions eliminates the need for fusion tag cleavage, avoiding potential issues with incomplete cleavage, residual tag sequences, or protease contamination. This represents a key distinction from elastin-like peptide (ELP) systems and other fusion tag approaches that require covalent attachment and post-purification processing. Fifth, the modular design of linker proteins enables straightforward optimization for different RuBisCO variants or even adaptation to other target proteins, providing a generalizable platform technology.

[0096] RuBisCO-linker network compositions represent a novel class of supramolecular protein assemblies that exploit multivalent binding and liquid-liquid phase separation for efficient, scalable purification of native or recombinant RuBisCO. The modular, tunable, and reversible nature of these networks provides a powerful platform for producing high-quality RuBisCO protein for diverse applications.

[0097] Thus, in a first aspect, the disclosure provides methods for purifying Ribulose- 1,5-bisphosphate Carboxylase / Oxygenase (RuBisCO) utilizing liquid-liquid phase separation including lysing a plant cell to make a plant cell lysate (PCL) comprising PCL precipitate and PCL supernatant; separating PCL precipitate in the PCL from PCL supernatant, e.g., by centrifugation; removing the PCL supernatant, e.g., by aspiration; adding to the PCL supernatant a RuBisCO linker protein that binds to at least two separate molecules of RuBisCO to form a RuBisCO-linker complex, and, due to the valency of two or greater on each RuBisCO protein, to form a network of interconnected RuBisCO-linker complexes, the presence of which forms a phase-separated solution (PSS) comprising a RuBisCO-linker precipitate, i.e., a RuBisCO-linker network precipitate; separating the RuBisCO-linker precipitate in the PSS from supernatant, e.g., by centrifugation; removing the supernatant, e.g., by aspiration; resuspending the RuBisCO-linker precipitate to dissociate the RuBisCO and the RuBisCO linker protein; and isolating purified RuBisCO; wherein the purified RuBisCO comprises eight RuBisCO large subunits (RbcL) and eight RuBisCO small subunits (RbcS); and wherein the RbcL, RbcS, or both, comprise one or more Rubisco binding interface.

[0098] In a second aspect, a method for isolating a recombinant RuBisCO is provided, the method including expressing recombinant RuBisCO large subunit (RbcL) and RuBisCO small subunit (RbcS) in a prokaryotic or eukaryotic cell; lysing the prokaryotic or eukaryotic cell to make a cell lysate; adding to the cell lysate a RuBisCO linker protein that binds to at least two separate molecules of RuBisCO to form a RuBisCO-linker complex, and, due to the valency of two or greater on each RuBisCO protein, to form a network of interconnected RuBisCO-linker complexes, the presence of which forms a turbid, phase-separated solution; separating precipitate in the turbid, phase-separated solution from supernatant, e.g., by centrifugation; removing the supernatant, e.g., by aspiration; resuspending the precipitate; and isolating purified RuBisCO; wherein the recombinant RbcL, recombinant RbcS, or both, comprise one or more RuBisCObinding interfaces, with a total valency of at least 2 for the natively folded RuBisCO protein; and wherein at least two RuBisCO binding interfaces are on separate molecules of RuBisCO.

[0099] In a third aspect, a method for isolating a recombinant RuBisCO from a cell-free system is provided, including: expressing recombinant RuBisCO large subunit (RbcL) and RuBisCO small subunit (RbcS) in a cell-free expression system; adding to the cell-free expression system a RuBisCO linker protein that binds to at least two separate molecules of RuBisCO to form a RuBisCO-linker complex, and, due to the valency of two or greater on each RuBisCO protein, to form a network of interconnected RuBisCO-linker complexes, the presence of which forms a turbid, phase-separated solution; separating precipitate in the turbid, phase-separated solution from supernatant, e.g., by centrifugation; removing the supernatant, e.g., by aspiration; resuspending the precipitate; and isolating purified RuBisCO; wherein the recombinant RbcL, recombinant RbcS, or both, comprise one or more Rubisco binding interfaces; and wherein the at least two RuBisCO binding interfaces are on separate molecules of RuBisCO.

[0100] In another aspect, the disclosure provides a RuBisCO-linker network composition that includes (i) a plurality of RuBisCO proteins; and (ii) a multivalent RuBisCO linker protein comprising at least two RuBisCO recognition domains, in which each RuBisCO protein is non-covalently bound to one or more RuBisCO linker proteins, and each RuBisCO linker protein is non-covalently bound to two or more RuBisCO proteins, thereby forming an interconnected network capable of liquid-liquid phase separation.

[0101] In some embodiments of the RuBisCO-linker network composition, the composition forms a phase-separated precipitate under low salt conditions; which, in embodiments, can dissociate under high salt conditions to release purified RuBisCO.

[0102] In some embodiments of the various aspects described herein, the purified RuBisCO includes eight RuBisCO large subunits (RbcL) and eight RuBisCO small subunits (RbcS), and in some embodiments, the purified RuBisCO is present in its natively folded form as shown, e.g., in FIG. 3, panel B. In some embodiments, the RbcL subunit comprises one or more RuBisCO binding interfaces, while in some embodiments the RbcS subunit includes one or more RuBisCO binding interfaces, and in some embodiments, one or more of the RbcL and RbcS subunits include one ormore RuBisCO binding interfaces. In some embodiments, the one or more RuBisCO binding interfaces are shared across at least one RbcL and at least one RbcS subunit in three-dimensional space. In some embodiments, the one or more RuBisCO binding interfaces comprise a sequence at least about 80% identical to the residues found at position 29, 30, 31, 32, 76, 79, 85, 86, 351, 353, and 355 of Synechococcus elongates sp. PCC6301 RbcL (Uniprot identifier: P00880) as numbered according to their corresponding position in Nicotiana tabacum RbcL (Uniprot identifier: P00876), or the residues at position 29, 30, 31, 32, 76, 79, 85, 86, 351, 353, and 355 of TobL35, and in some embodiments, the one or more RuBisCO binding interfaces comprise residues 93 and 94 of Synechococcus elongates sp. PCC6301 RbcS (Uniprot identifier: P04716) as numbered according to their native position, or the corresponding residues at position 162 and 163 of Nicotiana tabacum RbcS (Uniprot identifier: P69249) which corresponds to residues at position 104 and 105 following removal of the chloroplast-targeting peptide and renumbering of residue 58 of the pro-peptide to residue 1 and is the numbering of the RbcS form when expressed recombinantly and lacking the chloroplast targeting peptide.

[0103] In some embodiments, the RuBisCO linker protein comprises at least two RuBisCO recognition domains, each separated by a peptide linker, which can be, e.g., at least about 15 residues, about 20 residues, about 25 residues, about 30 residues, about 35 residues, about 40 residues, about 45 residues, about 50 residues, about 55 residues, about 60 residues, about 65 residues, about 70 residues, about 75 residues, about 80 residues, about 85 residues, about 90 residues, about 95 residues, about 100 residues, about 110 residues, about 120 residues, about 130 residues, about 140 residues, about 150 residues, about 160 residues, about 170 residues, about 180 residues, about 190 residues, about 195 residues, about 200 residues, about 210 residues, about 220 residues, about 230 residues, about 240 residues, about 250 residues, or between about 25 and 250 residues, between about 15 and 250 residues, between about 20 and 250 residues, between about 25 and 250 residues, between about 15 and 200 residues, between about 20 and 200 residues, between about 25 and 200 residues, between about 15 and 150 residues, between about 20 and 150 residues, between about 25 and 150 residues, or between about 15 and 50 residues, between about 20 and 50 residues, between about 25 and 50 residues, between about 30 and 50 residues, between about 15 and 45 residues, between about 20 and 45 residues, between about 25 and 45 residues, between about 30 and 45 residues, between about 15 and 40 residues, between about 20 and 40residues, between about 25 and 40 residues, between about 15 and 35 residues, or between about 20 and 35 residues in length. In some embodiments, the peptide linker comprises between about 25-45 residues.

[0104] Liquid-liquid phase separation (LLPS) is driven by multivalency, which is the ability of molecules to engage in multiple weak interactions. This is in contrast to other methods of the art employing elastin-like polypeptide (ELP) fusion tags that require covalent attachment to target proteins and subsequent proteolytic cleavage to obtain native protein, the methods disclosed herein utilize non-covalent, reversible multivalent interactions between engineered linker proteins and native RuBisCO complexes. This approach preserves the native structure and function of RuBisCO throughout purification, as demonstrated by CO2 fixation activity (FIG. 5B) and native gel electrophoresis (FIGs. 3B, 4A, and 4E), and permits facile dissociation and linker recycling through simple buffer exchange.

[0105] Under certain conditions, interactions among like molecules can overcome entropy, causing the molecules to become enriched and form a condensed phase along with a dilute phase. Conditions including concentration of proteins, pH, ionic strength, and temperature of the solution can influence LLPS. In some embodiments, differential salt concentrations are used to first induce the phase separation for the RuBisCO-linker complex and facilitate its separation from other molecules, and then to dissociate the RuBisCO and the RuBisCO linker protein to yield purified, natively folded RuBisCO. In some embodiments, the adding to the PCL supernatant or the CL supernatant of a RuBisCO linker protein that binds to at least two separate molecules of RuBisCO to form a RuBisCO-linker complex is carried out in low salt conditions. Low salt conditions can include, e.g., an ionic strength of less than about 200 mM, less than about 150 mM, less than about 100 mM, less than about 75 mM, less than about 50 mM, or less than about 25 mM. In some embodiments, low salt conditions comprise an ionic strength of less than about 250 mM.

[0106] In some embodiments higher salt conditions are used to dissociate the RuBisCO-linker complex into RuBisCO and the RuBisCO linker protein. In some embodiments, high salt conditions include an ionic strength of greater than about 500 mM, greater than about 600 mM, greater than about 700 mM, greater than about 800 mM, greater than about 900 mM, greater thanabout 1000 mM, greater than about 1250 mM, or greater than about 1500 mM. In some embodiments, high salt conditions comprise an ionic strength of greater than about 500 mM.

[0107] In some embodiments, protein purification from prokaryotic or eukaryotic cells, and plant cells in particular, may include one or more additional protein purification steps prior to the adding of the linker protein. For example, one or more purification steps including chromatography, extraction, electrophoresis, centrifugation, ultrafiltration, dialysis, differential solubility, affinity methods, and the like, may be performed prior to LLPS.

[0108] The RbcL and RbcS subunits that comprise RuBisCO can each be a wildtype sequence, e.g., wildtype tobacco (Nicotiana tabacum) RbcL and RbcS. In some embodiments, the subunits can be from the same species or different species. RbcL and RbcS subunits may be recombinant and expressed in a heterologous prokaryotic or eukaryotic cell type, e g., in an E. coli cell, in a yeast cell, or in an insect cell, or purified from a natural cell, e.g., wildtype alfalfa RuBisCO from alfalfa plant cells. In some embodiments, the RbcL and RbcS subunits are a wildtype sequence from a vascular plant, e.g., spinach or alfalfa.

[0109] In some embodiments, after LLPS the RuBisCO and the RuBisCO linker are separated, e.g., through exposure to high salt conditions or other technique to disrupt the weak bonding between RuBisCO the recognition domains on the RuBisCO linker protein and the RuBisCO binding interfaces.

[0110] In another aspect, an expression vector is provided that includes one or more TobL35DNA sequences as described herein. In some embodiments, the one or more TobL35DNA sequences is at least 90% identical, at least 95% identical, or at least 99% identical to the TobL35DNA sequence described herein, as follows:atgtcaccacaaacagagactaaagctagcgttggattcaaagctggtgttaaagagtacaaat tgacttattatactcctgagtacacgcccaaggatact at tattggcagcattccgagtaac tcctcaacctggagttccacctgaagaagcaggggccgcggtagctgccgaatcttctactggt acatggacaactgtatggaccgatggacttaccgaccttgatcgttacaaagggcgatgctacc acatcgagcgtgttgttggagaaaaagatcaatatattgcttatgtagcttaccctttagacct ttttgaagaaggttctgttaccaacatgtttacttccattgtaggtaacgtatttgggttcaaagccctgcgcgctctacgtctggaagatctgcgaatccctcctgcttatgttaaaactttccaag gtccgcctcatgggatccaagttgaaagagataaattgaacaagt tggtcgtcccctgttggg atgtactattaaacctaaattggggttatct ctaaaaactacggtagagctgtttatgaatgt cttcgcggtggacttgattttaccaaagatgatgagaacgtgaactcacaaccatttatgcgtt ggagagatcgtttcttattttgtgccgaagcactttataaagcacaggctgaaacaggtgaaat caaagggcattacttgaatgctact caggtacatgcgaagaaatgatcaaaagagctgtattt gctagagaattgggcgttccgatcgtaatgcatgactacttaacggggggattcaccgcaaata ctagcttggctcattattgccgagataatggtct cttcttcacatccaccgtgcaatgcatgc ggttattgatagacagaagaatcatggtatccacttccgggtattagcaaaagcgttacgtatg tctggtggagatcatattcactctggtaccgtagtaggtaaacttgaaggtgaaagagacataa ctttgggctttgttgatttactgcgtgaggatcatgttgaacaagatcgaagtcgcggtattta tttcactcaagattgggcctctttaccaggtgttctacccgtggcttcaggaggtattcacgtt tggcatatgcctgctctgaccgagatctttggggatgattccgtactacagttcggtggaggaa ctttaggacatccttggggtaatgcgccaggtgccgtagctaatcgagtagctctagaagcatg tgtaaaagctcgtaatgaaggacgt atctt ctcaggaaggtaatgaaattattcgcgaggct tgcaaatggagcccggaactagctgctgcttgtgaagtatggaaagagatcgtatttaattttg cagcagtggacgttttggataagtaa (SEQ ID NO:5).

[0111] In another aspect, a purified or expressed protein is provided comprising a TobL35protein sequence as described directly below and elsewhere herein, or analog translated from an RbcL DNA sequence and producing a protein that is at least 90% identical, at least 95% identical, or at least 99% identical to the TobL35protein sequence described directly below and elsewhere herein.MSPQTETKASVGFKAGVKEYKLTYYTPEYTPKDTDILAAFRVTPQPGVPPEEAGAAVAAESSTG TWTTVWTDGLTDLDRYKGRCYHIERWGEKDQYIAYVAYPLDLFEEGSVTNMFTSIVGNVFGFK ALRALRLEDLRIPPAYVKTFQGPPHGIQVERDKLNKYGRPLLGCTIKPKLGLSAKNYGRAVYEC LRGGLDFTKDDENVNSQPFMRWRDRFLFCAEALYKAQAETGEIKGHYLNATAGTCEEMIKRAVF ARE LGVP I VMHD YL TGG FTANT S LAH YCRDNGLLLH I HRAMHAVI DRQKNHG I H FRVLAKALRM SGGDHIHSGTWGKLEGERDITLGFVDLLREDHVEQDRSRGI YFTQDWASLPGVLPVASGGIHV WHMPALTE I FGDDSVLQFGGGTLGHPWGNAPGAVANRVALEACVKARNEGRDLAQEGNE I IREA CKWSPELAAACEVWKEIVFNFAAVDVLDK (SEQ ID NO:6).

[0112] Mutations in TobL35relative to wild-type A tabacum RbcL are as follows: Q30T, T31P, S76D, R86H, D351E, and F353H, which are indicated in bold in the preceding sequence.

[0113] Table 1 describes amino acid residues constituting the minimal set of interactors on RbcL for the RuBisCO-M35 binding interface. Residues and numbering for Synechococcus sp. PCC6301 RbcL and TobL35are indicated relative to the corresponding residues found in N. tabacum RbcL. The Uniprot identifier for each natural sequence is indicated in brackets. Residues altered in TobL35compared to N. tabacum RbcL are indicated in bold.Table 1

[0114] Table 2 describes amino acid residues constituting the minimal set of interactors on RbcS for the RuBisCO-M35 binding interface. Residues and numbering for Synechococcus elongatus PCC 6301 RbcS and TobS are indicated as found in the natural sequences. The Uniprot identifier for each natural sequence is indicated in brackets. The residue numbering as found for therecombinantly expressed, or the predominant form found in the chloroplast of N. tabacum following chloroplast leader peptide cleavage, are indicated. Modification of TobS in TobL35S was found to be unnecessary due to the conservation of these residues.Table 2

[0115] In some embodiments, a purified, recombinant RuBisCO is provided comprising wildtype RbcL subunits and wildtype RbcS subunits. In other embodiments, a purified, recombinant RuBisCO is provided that includes non-native RbcL subunits e.g., TobRbcL35, and native or nonnative RbcS subunits. In other embodiments, a purified, recombinant RuBisCO is provided that includes native RbcL subunits and non-native RbcS subunits. Further, a cell line genetically modified to express (i) a non-native RbcL subunit, e.g., TobRbcL35, and a native RbcS subunit to form a RbcL-non-native RuBisCO protein, or (ii) native RbcL subunits and non-native RbcS subunits to form a RbcS -non-native RuBisCO, or (iii) fully non-native RuBisCO protein comprising eight non-native RbcL subunits and eight non-native RbcS subunits is provided herein.

[0116] All statements herein reciting principles, aspects, and embodiments of the disclosure, as well as specific examples thereof, are intended to encompass both structural and functional equivalents thereof. Additionally, it is intended that such equivalents include both currently known equivalents as well as equivalents developed in the future, i.e., any elements developed that perform the same function, regardless of structure.

[0117] Various other components may be included and called upon for providing for aspects of the teachings herein. For example, additional materials, combinations of materials and / or omissionof materials may be used to provide for added embodiments that are within the scope of the teachings herein. Adequacy of any particular element for practice of the teachings herein is to be judged from the perspective of a designer, manufacturer, seller, user, system operator or other similarly interested party, and such limitations are to be perceived according to the standards of the interested party.

[0118] In the disclosure hereof any element expressed as a means for performing a specified function is intended to encompass any way of performing that function including, for example, a) a combination of circuit elements and associated hardware which perform that function or b) software in any form, including, therefore, firmware, microcode or the like as set forth herein, combined with appropriate circuitry for executing that software to perform the function. Applicants thus regard any means which can provide those functionalities as equivalent to those shown herein. No functional language used in claims appended herein is to be construed as invoking 35 U. S. C. §112(f) interpretations as “means-plus-function” language unless specifically expressed as such by use of the words “means for” or “steps for” within the respective claim.

[0119] The following examples further illustrate the present invention but should not be construed as in any way limiting its scope.EXAMPLES:

[0120] The inventors discovered that differences in the physical properties of the aqueous and RuBisCO condensate phase in LLPS preparations of the enzyme allow for efficient concentration and physical separation of the high molecular weight RuBisCO phase by centrifugation, thereby enabling rapid purification. Furthermore, interaction between linker proteins and RuBisCO are favored when salt concentrations are low, promoting ionic interactions between amino acid side chains on the sticker (binding) domains and RuBisCO surface. Therefore, phase-separated RuBisCO can be easily returned to a soluble form via addition of salt. As LLPS of proteins can be induced within minutes at room temperature, is reversible, requires and relies upon the native structure of the protein being preserved, and does not necessarily require any chromatography, it has been discovered to be a method for highly scalable RuBisCO extraction and purification in the proteins native form by a simple low speed centrifugation step, or even gravitational forces.

[0121] Known phase separation linker proteins do not have homologs in vascular plants and RuBisCO from vascular plants (the most abundant and accessible isoform) does not naturally exhibit LLPS. The best characterized RuBisCO linker protein is the M35 subdomain of CcmM from the cyanobacterium Synechococcus elongatus sp. PCC6301 which can be used to induce S. elongatus sp. PCC6301 RuBisCO (6301LS) LLPS in vitro. Cyanobacteria and vascular plants are separated by more than 1-2 billion years of evolution, and it was unknown and unexplored whether cyanobacterial linker proteins could productively function with plant RuBisCO proteins. Several other linker proteins have been documented and characterized within the peer-reviewed scientific literature and protein structural databases. These include CsoS2, EPYC1, PYCO1, and RCA. None of these linkers have been demonstrated to facilitate LLPS of a plant RuBisCO and analysis of their interaction interface similarly suggests they are not optimal for interaction with plant RuBisCO and will not already mediate LLPS without significant, non-trivial engineering and experimentation.

[0122] Example 1: Proof of concept utilizing Synechococcus elongatus PCC6301 (6301LS)

[0123] To test the foundation of the purification concept, the model phase separating RuBisCO from Synechococcus elongatus PCC6301 (6301LS) was expressed in Escherichia coli to examine if it could be enriched from the cellular protein content via LLPS with its natural linker protein M35. 6301LS is poorly soluble in E. coli and typically accumulates to -0.5% of the total soluble protein or -60* less than RuBisCO in leaf tissue. Despite these low levels, addition of purified M35 to the bacterial total soluble protein led to rapid LLPS and a white protein pellet upon centrifugation at 20,000 g for 5 minutes. The protein recovered from the pellet fraction contained highly concentrated RuBisCO only when M35 was present (FIG. 3A). Furthermore, the resulting pellet protein, when resolubilized in buffer containing 250 mM NaCl and subjected to Native polyacrylamide gel electrophoresis (Native PAGE), was found to contain RuBisCO in the native RbcLsRbcSs conformation (FIG.3B). Hence, for the first time it was demonstrated that LLPS can, even in preliminary and unoptimized conditions, be used to rapidly purify RuBisCO in its native state.

[0124] To examine if M35 was capable of productive interaction with RuBisCO from a vascular plant, both 6301LS and RuBisCO were purified from Spinacia oleracea (Spinach). Addition of M35 to 6301LS and Spinach RuBisCO in vitro followed by room temperature incubation for 10minutes lead to a change in turbidity for 6301LS, but not Spinach. Furthermore, M35 was found to sequester 6301LS into higher molecular weight aggregates that did not migrate under Native PAGE, whereas the plant RuBisCO remain largely unperturbed (FIG. 3C). Analysis of the structural interface between the RuBisCO binding domain of M35 and 6301LS followed by sequence alignment of Spinach RuBisCO and 6301LS highlighted critical residue differences between the two proteins at two interfaces essential for interaction (FIG.3D).

[0125] Example 2: Bacterial expression of vascular plant RuBisCO

[0126] To extend the invention to a form of RuBisCO found in vascular plants, a bacterial expression system was utilized to produce RuBisCO from Nicotiana tabacum (tobacco). Analysis of the protein-protein binding interface between the RuBisCO binding domain of M35 and the Synechococcus elongatus PCC 6301 RuBisCO, followed by a sequence homology alignment between Synechococcus elongatus PCC 6301 RuBisCO and N. tabacum RuBisCO subunits (FIG.13) indicated that 7 out of 13 M35 interacting residues were already conserved between Synechococcus and N. tabacum RbcL and RbcS (Table 1 and Table 2). Subsequently, it was discovered that engineering the substitutions (Q30T, T31P, S76D, R86H, D351E, and F353H) in N tabacum RbcL (TobL33) provided a productive binding interface with M35, which enabled the LLPS of N tabacum RuBisCO. The substitutions were integrated into a bacterial expression vector for N. tabacum RuBisCO pET16_NtL35S-Rca using site directed mutagenesis and expressed alongside required biogenesis chaperones in E. colt BL21(star).

[0127] TobL35S produced in 72 coli migrated on native PAGE consistent with RuBisCO assembled in a native configuration (FIG. 4A). The addition of purified M35 to extracts from bacteria expressing TobL35S rapidly became turbid in appearance suggesting the presence of phase-separated TobL35S-M35 condensates. Indeed, the addition of M35 at increasing amounts to bacterial lysates containing expressed TobL35S or wild-type N. tabacum RuBisCO showed a reduction of band intensity of TobL35S in the presence of 5 pM M35 (FIG. 4A), similar to the behavior of 6301LS when exposed to pure M35 (FIG. 3C). This band shift was not observed for the wild-type RuBisCO indicating that the interaction was specific to TobL35S. Centrifugation at 5,000 g of TobL35S extracts containing M35 added produced a pure white pellet which could be easily separated from the remaining liquid by aspiration, and which did not appear when M35 wasapplied to bacterial lysate containing wild-type N. tabacum RuBisCO (FIG. 4B). Resuspension of the pellet in a volume of low salt buffer (50 mM EPPS-NaOH pH 8.0, 15 mM MgCh, 1 mM EDTA) generated a pale cloudy solution. Analysis of these fractions by SDS-PAGE (FIG. 4D) revealed the resuspended pellet fraction consisted of only three protein species with molecular weights equivalent to RbcL, RbcS, and M35 confirming the pellet resulted from the LLPS of TobL35S RuBisCO and M35. Resuspension of a similarly produced pellet in high salt buffer (50 mM EPPS-NaOH pH 8.0, 15 mM MgCl₂, 1 mM EDTA, 500 mM NaCl) resulted in a clear solution with no discernable aggregate, consistent with a reversible LLPS condensates driven by ionic interactions between M35 and TobL35S and not by terminal aggregation. Gel filtration using Superdex 200 size exclusion chromatography column of the salt suspended pellet resolved a peak fraction consistent with the retention volume of native RuBisCO (FIG. 4C), which was then isolated and flash frozen in liquid nitrogen. Analysis of the purified TobL35S protein via native PAGE and upon incubation with purified M35 led to the formation of complexes of TobL35S and M35 that resolved as diverse higher molecular weight complexes, some of which were too large to migrate into the gel matrix and consistent with the large molecular weight of a large biomolecular condensate.

[0128] Analysis of the total abundance of RuBisCO present in the bacterial total soluble protein indicated that TobL35S is produced and assembled in E. coli at approximately 4-fold higher amounts relative to wild-type N. tabacum RuBisCO (FIG. 5A). Analysis of the CO2 fixation activity of purified TobL35S demonstrated that the enzyme fixation rate is approximately 50% (1.7 s-1) of wild-type N. tabacum RuBisCO (3.2 s’1)16(FIG. 5B). However, photosynthetic modelling indicates that if the biogenesis improvements displayed by TobL35S in E. coli translates to N. tabacum chloroplasts, the reduced CO2 fixing activity would be more than offset by the increased active site content, resulting in an overall improvement to photosynthetic CO2 assimilation (FIG.5B) Furthermore, plants producing an elevated amount of RuBisCO would constitute a more abundant source for extraction using the LLPS purification strategy demonstrated here.

[0129] Example 3: Synthetic Linker Design

[0130] We next sought to develop linker proteins that contained properties to elicit LLPS of wildtype RuBisCO as is found in leaf extracts. Synthetic binding domains were developed to achievethis goal. The interface between Synechococciis elongatus PCC 6301 RuBisCO and the SSUL1 binding domain from M35 (PDB ID: 6HBC) was utilized as a model scaffold. To adapt M35 to recognize plant RuBisCO, we aligned the structural coordinates of tobacco RuBisCO (PDB ID: 4RUB) onto the Synechococciis elongatus PCC 6301 cyanobacterial RuBisCO-SSULl structure, and then occluded the cyanobacterial enzyme. A hybrid scaffold was created in which the M35 binding domain was positioned against tobacco RuBisCO in the same geometry as the native interaction.

[0131] The M35 domain was then subjected to sequence redesign using ProteinMPNN, conditioned on the fixed back-bone geometry of the scaffold. The objective was to identify mutations that could potentially preserve the binding mode while improving compatibility with plant-like RuBisCO, in this case, tobacco RuBisCO. A total of 10,000 M35 sequence redesigns were generated using a combination of ProteinMPNN for sequence design on the fixed hybrid scaffold and SolubleMPNN to ensure solubility of the binder. To reduce the design space, Rosetta FastRelax was applied to all designs, and scoring was performed based on the interface AAG (ddg). The top 1,880 designs were selected for further analysis. The 1,880 candidate complexes were evaluated with AlphaFold3 (AF3) to assess recapitulation of the designed binding mode. Structures that failed to recapitulate the intended RuBisCO-SSUL1 interaction geometry were excluded. From the AF3-predicted models, ranking was performed using interface predicted Aligned Error (iPAE) at the interface and interface predicted TM-score (iPTM) as primary metrics. These two scores were used in combination to balance structural confidence and interface quality. The top 90 candidates, representing the highest-confidence binders, were obtained to characterize binding. Six further binders that failed selection were also obtained for use as putative negative controls for binding.

[0132] To screen binder designs for interaction affinity against wild-type tobacco RuBisCO, a high-throughput reporter assay for RuBisCO binders was devised using a commercial NanoBiT kit from Promega. The overall design consisted of two components. The first component featured a fusion between the RuBisCO small subunit and a smBiT tag. The second component featured a fusion between a RuBisCO binding domain and LgBiT. The LgBiT and smBiT together complement to form NanoLuc, a shrimp luciferase enzyme that will emit light when exposed to furimazine (FIG. 7A). The light emission is proportional to the relative proximity of LgBiT andsmBiT as the interaction affinity between these components is low (KD = 190 pM). Therefore, it was determined that stronger binding affinity of the synthetic binders could be identified from the assay through luminescence screening.

[0133] The synthetic binders were fused to LgBiT in batch ligation using golden gate cloning to generate a library of fused binder designs, along with controls featuring the wild-type SSUL1, SSUL2, and SSUL3 domains from M35. The library was transformed into E. coli and colonies inoculated into 96 well plates, grown to turbidity and lysed with lysozyme. To each well, a standard solution containing smBiT tagged wild-type tobacco TobLS or TobL35S was added followed by addition of furimazine. Luminescence was monitored using a plate reader (FIG. 7B). Material from the 12 wells exhibiting the greatest fold change in luminescence over background was collected and subjected to sequencing to identify the binder composition (FIG. 8).

[0134] The sequences of the top twelve binders:

[0135] Sequence 1783:MLLDEETIKLIKEWLEKGYSIKVEYADPKEHKQNIWKTAKTINYKDLEEVLKALKEIIAENKGK YIYLHAVDEEKEKWGKKLIQRPG (SEQ ID NO:7)

[0136] Sequence 807:MLLDEETIKLVKELLEKGYRIRTEYADPREHRQNIWTPCEPINSTDLEEVLKKLEEILEKNKGK YVRLLGYDEKTEQWYSKLIQRPG (SEQ ID NO:8)

[0137] Sequence 1555:MLLPPEWELVKQWLAQGYGIRTQAATPREHRVNLWRPCEPINSTDLDEVLAKLKEILETHKGE YVELLAYDPATERWGRQVIQRPG (SEQ ID NO:9)

[0138] Sequence 1450:MLLDEETKKIVEELLKKGYS IGVEYATPREHRENIWHPCEPINSTDLEEVLKQLEKILEKHKGK YVYLLGIDTKTNKWYRKLVQRPE (SEQ ID NO: 10)

[0139] Sequence 1763:MLLDEETKKLIEEYLKKGYLIGVEYANPREHRANIWRPATPINSTDLEEVIKKLEEILEKYKGK YVYLIAIDPEKNEWGKKLIQRPE (SEQ ID NO: 11)

[0140] Sequence 1662:MLLTPEQKALVEQLLAQGYAIRTEYADPKEHRANIWTPTKPIDTSDLDEVLAALEEILKENEGK YVYLYGFDPENEKWFKQLIQRPG (SEQ ID NO: 12)

[0141] Sequence 540:MLLTPETVKLIEEWLAKGYQIRVEYADPKEHRQNIWRPAKPINYTDLKEVLAALEKILAEHKGK YVYLLAYDPKTQRWGKKLIQRPG (SEQ ID NO: 13)

[0142] Sequence 503:MLLNEETIKLVKELLKKGYSIGLEYANPVEHRENVWHPCEPINSKDLEEVLKKIEEILEKNKGK YVYLLGIDEKTNTWYRKLIQRPE (SEQ ID NO: 14)

[0143] Sequence 1737:MLLTEETIKIIKELLEKGYVIRLEYATPARRRANIWTPAEPINSTDLEEVLAKVKEILEKNKGK YWLLGWDPEKDRIVYRKLIQRPG (SEQ ID NO: 15)

[0144] Sequence 1370:MLLDEETIEWIKEYLEKGYLIGLEYANEKEHRENIWTPCEPINSTDLEEVLKKLKEILEKHKGK YVYLLGIDPETNTWGRKLIQRPE (SEQ ID NO: 16)

[0145] Sequence 1282:MLLTEETIELIKKWLEEGYSIGLEYADPREHRENVWRPAKPINYTDLEEVLKKLEEILKENKGK YVYLLGIDTKTNTWGRKLIQKPG (SEQ ID NO: 17)

[0146] Sequence 783:MLLDEETIKKIKEWLEKGLLIGLEYADEKRHKQNIWRPAKPINSTDLEEVLKKLKEILEKNKGK YVKLLAIDPETNTWGEEI IQRPE (SEQ ID NO: 18)

[0147] The synthetic binding domain from 1737 was reconstituted as a duplicate repeat of domains (replaced the SSUL1 and SSUL2 domains in SynM35) with interdomain sequence and SSUL3identical to SynM35, to form the synthetic 1737 linker. The linker was his-tagged and purified then added to a solution containing 2 pM wild-type tobacco RuBisCO in stoichiometric increments. Phase separation was not observed but native PAGE analysis indicated that interaction with RuBisCO was generating higher molecular weight complexes like those observed with M35 and TobL35S, albeit at a 10-fold greater concentration of linker.

[0148] Further optimization of 1737 was performed using rational design of the interface. A potentially important residue interaction between F253 on SSUL1 and H353 on 6301L was identified. H353 is not conserved in TobL and TobL WT has F353 instead. To potentially assist the binding between TobL35S and SynM35, an F353H mutation was made on TobL to preserve this F-H interaction. To preserve such a F-H interaction, a designed binder can have an H residue at a position suitable to interact with F353 on the TobL. However, 1737 was found to have an Arg at the corresponding residue (R31) on the binding domain instead. Consequently, an R31H mutation was introduced into 1737. Also, R252 on the SSUL1 is close to D76 on 6301L from the structure. The two residues may form a strong interaction (salt-bridge) to help the binding. Thus, to further optimize 1737, an A to R mutant (A29R) was additionally installed. So, the modified 1737 construct was 1737_A29R_R31H, which was then expressed and run through the luciferase assay described above against TobLS WT, which shows increased binding affinity of 1737 A29R R31H over 1737 WT (FIG. 10A) and TobL35S which shows similar binding affinity between 1737_A29R_R31H and 1737 WT (FIG. 10B). These results confirm not only that preserving the F-H interaction is important but also show that the designed binder is binding to the TobLS WT at the position determined.

[0149] The optimized binding domain of 1737 A29R R31H was reconstituted as a triplicate repeat of domains with interdomain sequence identical to SynM35, to form the synthetic 3x1737 ARRH linker. The linker was his-tagged and purified then added to Nicotiana benthamiana leaf lysate. Phase separation was observed and further SDS page confirmed phase separation (FIG.11)

[0150] Sequence 1737_A29R_R31H:MLLTEETIKIIKELLEKGYVIRLEYATPRRHRANIWTPAEPINSTDLEEVLAKVKEILEKNKGK YWLLGWDPEKDRIVYRKLIQRPG (SEQ ID NO: 19).

[0151] To offset the apparent lower binder affinity for wild-type tobacco RuBisCO as compared to Synechococcus elongatus PCC6301 RuBisCO, the number of binding domain repeats was increased to 5x. The longer linker was found to increase the amount of tobacco RuBisCO accumulating in the pellet fraction of phase-separated leaf lysate (FIG. 12).

[0152] Among the synthetic linkers tested, the 5x1737 ARRH multivalent linker (comprising five tandem repeats of the 1737 A29R R31H binding domain) demonstrated superior performance for wild-type tobacco RuBisCO phase separation from plant cell lysate (FIG. 12), achieving greater RuBisCO accumulation in the pellet fraction compared to the 3x1737 ARRH linker (FIG. 11).

[0153] While phase separation of wild-type plant RuBisCO was successfully achieved using synthetic multivalent linkers (visible pellet formation prior to gels, FIG. 11-12; RuBisCO bands in pellet fraction), the purity achieved with current linker designs (Figures 11-12) is lower than that obtained with TobL35S mutant RuBisCO and M35 linker (Figure 4D). Co-purifying proteins appear in the pellet fraction. The 5x1737 ARRH linker (FIG. 12) showed improved RuBisCO accumulation relative to the 3x1737 ARRH linker (FIG. 11), demonstrating that increased valency enhanced phase separation efficiency. Nonetheless, these results establish the feasibility of using engineered synthetic linkers to selectively phase-separate wild-type plant RuBisCO from complex leaf lysates, representing the first demonstration of LLPS-based purification of native plant RuBisCO.

[0154] The work described herein conclusively shows that LLPS can be used as a high throughput method to purify RuBisCO in a ‘native state’ from complex biological solutions. Mutations Q30T, T31P, S76D, R86H, D351E, andF353H (TobL35S) in A. tabacum RuBisCO produce a functionally active enzyme capable of LLPS with the M35 subdomain of CcmM from Synechococcus elongatus sp. PCC6301. TobL35S displays a production advantage relative to the wild-type that if expressed in planta could enable the scalable production and purification of RuBisCO from leaf material using LLPS. In addition, the combination of binder design and rational protein engineering demonstrates the potential of machine learning-assisted approaches to generate synthetic binders capable of phase separating and purifying wild-type plant RuBisCO. The described methodology enables the design of binding proteins that bind to wild-type plant RuBisCO complexes and facilitate selective purification from plant extracts. Further optimization of binder affinity andspecificity is expected to enhance the practical utility of this approach. This strategy provides a foundation for developing scalable, cost-effective purification technologies for RuBisCO with potential applications in food production and biomanufacturing.

[0155] It should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific implementations described above. The specific implementations described above are disclosed as examples only.

Claims

In the claims:

1. A method for purifying Ribulose- 1,5-bisphosphate Carboxylase / Oxygenase (RuBisCO), comprising:lysing a plant cell to make a plant cell lysate (PCL) comprising PCL precipitate and PCL supernatant;separating PCL precipitate in the PCL from PCL supernatant, e.g., by centrifugation; removing the PCL supernatant, e.g., by aspiration;adding to the PCL supernatant a RuBisCO linker protein that binds to at least two separate molecules of RuBisCO to form a RuBisCO-linker complex, the presence of which forms a phase-separated solution (PSS) comprising a RuBisCO-linker precipitate; separating the RuBisCO-linker precipitate in the PSS from supernatant, e.g., by centrifugation;removing the supernatant, e.g., by aspiration;resuspending the RuBisCO-linker precipitate to dissociate the RuBisCO and the RuBisCO linker protein; andisolating purified RuBisCO;wherein the purified RuBisCO comprises eight RuBisCO large subunits (RbcL) and eight RuBisCO small subunits (RbcS); andwherein the RbcL, RbcS, or both, comprise one or more Rubisco binding interfaces.

2. A method for isolating a recombinant Ribulose- 1,5 -bisphosphate Carboxylase / Oxygenase (RuBisCO), comprising:expressing recombinant RuBisCO large subunit (RbcL) and RuBisCO small subunit (RbcS) in a prokaryotic or eukaryotic cell;lysing the prokaryotic or eukaryotic cell to make a cell lysate (CL);adding to the CL a RuBisCO linker protein that binds to at least two separate molecules of RuBisCO to form a RuBisCO-linker complex, the presence of which forms a turbid, phase-separated solution;separating precipitate in the turbid, phase-separated solution from supernatant, e.g., bycentrifugation;removing the supernatant, e.g., by aspiration;resuspending the precipitate; andisolating purified RuBisCO;wherein the recombinant RbcL, recombinant RbcS, or both, comprise one or more Rubisco binding interfaces; andwherein the at least two RuBisCO binding interfaces are on separate molecules of RuBisCO.

3. A method for isolating a recombinant Ribulose- 1,5 -bisphosphate Carboxylase / Oxygenase (RuBisCO), comprising:expressing recombinant RuBisCO large subunit (RbcL) and RuBisCO small subunit (RbcS) in a cell-free expression system;adding to the cell-free expression system a RuBisCO linker protein that binds to at least two separate molecules of RuBisCO to form a RuBisCO-linker complex, the presence of which forms a turbid, phase-separated solution;separating precipitate in the turbid, phase-separated solution from supernatant, e g., by centrifugation;removing the supernatant, e.g., by aspiration;resuspending the precipitate; andisolating purified RuBisCO;wherein the recombinant RbcL, recombinant RbcS, or both, comprise one or more Rubisco binding interfaces; andwherein the at least two RuBisCO binding interfaces are on separate molecules of RuBisCO.

4. The method of claim 2 or claim 3, wherein the purified RuBisCO comprises eight RuBisCO large subunits (RbcL) and eight RuBisCO small subunits (RbcS).

5. The method of any one of claims 1-4, wherein the purified RuBisCO is present in natively folded form.

6. The method of any one of claims 1-5, wherein the RbcL subunit comprises one or more RuBisCO binding interfaces.

7. The method of any one of claims 1-6, wherein the RuBisCO linker protein comprises at least two RuBisCO recognition domains, each separated by a peptide linker.

8. The method of claim 7, wherein the peptide linker comprises at least 20 residues.

9. The method of either of claims 7 or 8, wherein the peptide linker comprises between about 25 and 250 residues.

10. The method of any one of claims 7-9, wherein the peptide linker comprises between about 25-45 residues.

11. The method of any one of claims 1-10, wherein one or more RuBisCO binding interfaces comprise a sequence at least about 80% identical to residues 29, 30, 31, 32, 76, 79, 85, 86, 351, 353, and 355 of Synechococcus elongatus sp. PCC6301 RbcL, or at least about 80% identical to residues 29, 30, 31, 32, 76, 79, 85, 86, 351, 353, and 355 of TobL35.

12. The method of claim 11, wherein one or more RuBisCO binding interface comprise residues 29, 30, 31, 32, 76, 79, 85, 86, 351, 353, and 355 of Synechococcus elongatus sp.PCC6301 RbcL and residues 93 and 94 Synechococcus elongatus sp. PCC6301 RbcS in the native RuBisCO RbcLRbcS complex, or residues 29, 30, 31, 32, 76, 79, 85, 86, 351, 353, and 355 of TobL35and residues 104 and 105 of the N. tabacum RbcS in the native RuBisCO TobL35S complex.

13. The method of any one of claims 1-12, wherein the adding to the PCL supernatant or CL supernatant of a RuBisCO linker protein that binds to at least two separate molecules of RuBisCO to form a RuBisCO-linker complex is carried out in low salt conditions.

14. The method of any one of claims 1-13, wherein the resuspending of the RuBisCO-linker precipitate to dissociate the RuBisCO and the RuBisCO linker protein is carried out in high salt conditions.

15. The method of any one of claim 1-14, wherein low salt conditions comprise an ionic strength of less than about 200 mM, less than about 150 mM, less than about 100 mM, less than about 75 mM, less than about 50 mM, or less than about 25 mM.

16. The method of claim 13 15, wherein low salt conditions comprise an ionic strength of less than about 250 mM.

17. The method of any one of claim 1-16, wherein high salt conditions comprise an ionic strength of greater than about 500 mM, greater than about 600 mM, greater than about 700 mM, greater than about 800 mM, greater than about 900 mM, greater than about 1000 mM, greater than about 1250 mM, or greater than about 1500 mM.

18. The method of any one of claim 1-17, wherein high salt conditions comprise an ionic strength of greater than about 500 mM.

19. The method any one of claims 1 to 18, further comprising one or more additional purification steps prior to the adding of the linker protein.

20. The method of claim 19, wherein the one or more additional purification steps comprises chromatography, electrophoresis, centrifugation, ultrafiltration, and dialysis.

21. The method of any one of claims 1-20, wherein one or both of the RbcL and the RbcS subunits are a wildtype sequence.

22. The method of claim 21, wherein the wildtype sequence is from a vascular plant.

23. The method of any one of claims 1-20, wherein one or both of the RbcL and the RbcS subunits are not a wildtype sequence.

24. The method of any one of claims 1-20, wherein the RbcL subunits comprise both wildtype and non-wildtype sequences.

25. The method of any one of claims 1-20, wherein the RbcS subunits comprise both wildtype and non-wildtype sequences.

26. The method of any one of claims 2 or 4-25, wherein the prokaryotic cell comprises an E. coli.

27. The method of any one of claims 2 or 4-25, wherein the eukaryotic cell comprises a yeast cell, a mammalian cell, or an insect cell.

28. The method of any one of claim 1-27, wherein the RuBisCO and the RuBisCO linker protein are separated.

29. An expression vector comprising one or more TobL35S DNA sequences as described herein.

30. The expression vector of claim 29, wherein the one or more TobL35S DNA sequences is at least 90% identical, at least 95% identical, or at least 99% identical to the TobL35S DNA sequence described herein.

31. A purified protein comprising a TobL35protein sequence as described herein, or analog translated from a TobL35S DNA sequence that is at least 90% identical, at least 95% identical, or at least 99% identical to the TobL35S DNA sequence described herein.

32. Purified, recombinant RuBisCO comprising wildtype RbcL subunits and wildtype RbcS subunits.

33. Purified, recombinant RuBisCO comprising one or more non-native RbcL subunit and one or more non-native RbcS subunit.

34. Purified, recombinant RuBisCO comprising wildtype RbcL subunits and one or more non-native RbcS subunits.

35. Purified, recombinant RuBisCO comprising one or more non-native RbcL subunits and wildtype RbcS subunits.

36. A cell line genetically modified to express a non-native RbcL subunit, a non-native RbcS subunit, both non-native RbcL subunits and non-native RbcS subunits, or non-native RuBisCO protein comprising eight RbcL subunits and eight RbcS subunits.

37. The method of any one of claims 1-28, wherein when adding a RuBisCO linker protein to the plant cell lysate (PCL) supernatant, to the cell lysate (CL), or to the cell-free expression system, to form a RuBisCO-linker complex, a plurality of RuBisCO-linker complexes bind to each other to form a RuBisCO-linker network.

38. A RuBisCO-linker network composition comprising:(i) a plurality of RuBisCO proteins; and(ii) a multivalent RuBisCO linker protein comprising at least two RuBisCO recognition domains;wherein each RuBisCO protein is non-covalently bound to one or more RuBisCO linker proteins, and each RuBisCO linker protein is non-covalently bound to two or more RuBisCO proteins, thereby forming an interconnected network capable of liquid-liquid phase separation.

39. The composition of claim 18, wherein the composition forms a phase-separated precipitate under low salt conditions.

40. The composition of claim 45, wherein the phase-separated precipitate dissociates under high salt conditions to release purified RuBisCO.

41. The method of any one of claims 1-28, wherein the RuBisCO linker protein comprises one or more ofSEQ IDNO:7 to SEQ ID NO: 19.

42. The method of claim 41, wherein the RuBisCO linker protein comprises SEQ ID NO: 11.

43. The method of claim 41, wherein the RuBisCO linker protein comprises SEQ ID NO: 15.

44. The method of claim 41, wherein the RuBisCO linker protein comprises SEQ ID NO: 19.

Images