WO2008024311A2 - Purification of low-abundant recombinant proteins from cell culture - Google Patents

Abstract

A method of purifying a fusion protein from a cell culture containing the same by inverse transition cycling (ITC), said fusion protein comprising a protein or peptide of interest coupled to a bioelastic polymer, is carried out by adding free bioelastic polymer as a co-aggregant to the cell culture solution so that the mass of fusion protein captured from the soluble cell lysate is increased.

Description

PURIFICATION OF RECOMBINANT PROTEINS FROM CELL CULTURE

Ashutosh Chilkoti, Kimberly T. Carlson, Trine Christensen, Carlos Filipe, and Xin Ge

This application claims benefit of U.S. Provisional Patent Application Serial No. 60/839,270, filed August 22, 2006, the content of which is incorporated herein by reference in its entirety. This invention was made with Government support under grant no.

RO1GM061232 from the National Institutes of Health. The US Government has certain rights to this invention.

Field of the Invention This invention concerns the purification of recombinant proteins from cell culture, including purification at the single molecule per cell limit.

Background of the Invention

A crucial and challenging problem in proteomics is the purification, identification, and characterization of proteins, some of which are expressed at very low levels. The preferred method for purification of low abundance proteins exploits multiple affinity purification tags on a single recombinant protein, whereby recombinant proteins can be purified from complex mixtures using sequential, orthogonal affinity chromatography (Schimanski,et al. (2005) Eukaryotic Cell 4:1942- 1950; Graumann, et al. (2004) MoI. Cell. Proteomics 3:226-237; Honey, et al. (2001)

Nucl. Acids Res. 29:e24; Drakas, et al. (2005) Proteomics 5:132-137). One of the more widely used schemes is tandem affinity purification (TAP) (Drakas, et al. (2005)

Proteomics 5:132-137; Jao & Chen (2006) J. Cell. Biochem. 97:583-598; Rigaut, et al. (1999) Nature Biotech. 17:1030-1032; Elbing, et al. (2006) Biochem. J. 393:797- 805; Lichty, et al. (2005) Protein Expr. Pur. 41:98-105), in which recombinant proteins are first purified using an IgG affinity column followed by a column of immobilized calmodulin. The purified product is eluted by the addition of EGTA, which chelates the calcium, releasing the calmodulin binding peptide affinity tag. This technique is lengthy, requiring many sequential binding, washing and elution steps, as well as costly, requiring two different and expensive resins to recover the purified recombinant protein, so that processing large amounts of cell lysate makes it prohibitively expensive.

Inverse transition cycling (ITC) provides an alternative means for the purification of recombinant proteins without the need for specialized equipment or expensive resins. By fusing a recombinant protein to a bioelastic polymer whose solubility can be reversibly modulated by the manipulation of environmental variables including solution temperature, ionic strength, pH, ionic composition, and pressure, the target recombinant protein can be efficiently purified from complex biological mixtures using successive steps of triggering the phase transition to aggregate the bioelastic fusion protein, separating the bioelastic fusion proteins from the mixture, and resuspending the bioelastic fusion proteins in low ionic strength buffer. This process can be repeated as needed to increase the concentration and purity of the bioelastic fusion protein in the purified product. It is an inexpensive and easily implemented technique demonstrated to result in purified protein yields at least equal to if not greater than those produced with conventional for well-expressed recombinant proteins (Trabbic-Carlson, et al. (2004) Protein Sci. 13:3274-3284). Here, a methodology is described to implement ITC where target proteins are expressed at low levels.

Summary of the Invention

A first aspect of the invention is a method of purifying a fusion protein from a cell culture containing the same by inverse transition cycling (ITC), said fusion protein comprising a protein or peptide of interest coupled to a bioelastic polymer, the improvement comprising: adding free bioelastic polymer as a co-aggregant to said cell culture solution (e.g., a culture solution containing less than 1 mg/Liter of said fusion protein) so that the mass of fusion protein captured from the soluble cell lysate is increased. Alternatively stated, in some embodiments the cell culture contains less than 1000, 2000, 3000, 4000, or 5000 molecules of said fusion protein per cell in said cell culture. Alternatively stated, in some embodiments the cell culture contains less than 1, 2, 3, 4, or 5 x 10^"16 grams of protein per cell in said cell culture. In some embodiments, the culture contains less than 100, less than 10, or less than 1 micrograms/Liter of said fusion protein.

Any suitable cell culture may be used, including but not limited to plant, animal, yeast or bacterial (e.g., Escherichia colϊ) cell cultures.

ITC purification may be carried out by any suitable technique. Typically ITC comprises a separating step where aggregated fusion protein (induced by adding any suitable precipitant, such as inorganic salts, organic solvents, polymers, or combinations thereof) is separetd form soluble matter by any suitable technique, (including but not limited to centrifugation and microfϊltration) and a resuspension step where the fusion protein is redissolved in buffer. Additional bioelastic polymer may be added in any suitable manner, such as by combining a solution of bioelastic polymer with the cell culture solution (e.g. lysate, spent media, cell suspension) to produce final bioelastic polymer concentrations in the cell culture solution of 1 to 10 μM, 10 to 100 μM, or 100 to 1000 μM (or greater). Note that bioelastic polymer concentrations reflect the final concentration in the cell culture, and not the concentration of the stock that is added to the cell culture solution.

The present invention is explained in greater detail in the following non- limiting Examples.

Brief Description of the Drawings

Figure 1. Schematic presentation showing protein purification with free bioelastic polymer as a co-aggregant. Co-aggregation with free ELP results in purification of ELP fusion protein at concentrations which are otherwise unrecoverable using traditional ITC conditions.

Figure 2. Percentage recovery of 10 μg (~20 nM fusion protein; Figure 2A) and 1 μg (~2 nM fusion protein; Figure 2B) ELP fusion proteins from 10 mL of PBS solution with and without the use of 5 μM free ELP. Error bars reflect the first standard deviation for at least three replicates of each sample. One-way ANOVA analysis indicates that the amount of ELP fusion protein captured with the use of free ELP is statistically different from the amount captured using NaCl alone to induce the ELP inverse phase transition (p=2.3xlθ^"4, 5.IxIO^"11, 2.5x10^"'° for 10 μg of thioredoxin - A -

(Trx), blue fluorescent protein (BFP), and chloramphenicol acetyltransferase (CAT), respectively and p=1.7xlθ^"9, 6.OxIO^'2, 3x10^~10 for 1 μg Trx, BFP₇ CAT respectively).

Figure 3. Silver-stained SDS-PAGE gel showing the difference in recovery of fusion protein using the free ELP co-purification technique and the traditional NaCl purification. Five rounds of ITC were carried out after 10 μg Trx-ELP has been added to soluble lysates. The fusion protein was cleaved with thrombin overnight before loading on the gel. Lane 1, molecular weight markers; lane 2, (control) previously purified Trx-ELP (10 μg) cut with thrombin; lane 3, (control) previously purified Trx- ELP (5 μg) cut with thrombin; lane 4, (control) previously purified Trx-ELP (1 μg) cut with thrombin; lane 5, recovered Trx-ELP purified with NaCl and cut with thrombin; lane 6, recovered Trx-ELP purified with ELP/NaCl and cut with thrombin.

Figure 4. Percent recovery of 1 μg ^uC-labeled Trx, BFP, and CAT ELP fusion proteins from soluble E. coli lysate using 5 μM free ELP to facilitate fusion protein capture. Error bars reflect the first standard deviation from three replicates of each fusion, protein. The mass of fusion protein added to soluble lysate (1 μg) is reflective of the concentration expected for 2-3 expressed molecules per E. coli cell.

Figure 5. Schematic presentation showing purification of an anti-GFP antibody.

Figure 6. Capture of anti-GFPhrp antibody with GFP/ELP with (+) and without (-) free ELP.

Detailed Description of the Preferred Embodiments

"Transition temperature" or "T_t" as used herein, refers to the temperature above which a polymer that undergoes an inverse temperature transition is insoluble in an aqueous system (e.g., water, physiological saline solution, blood plasma), and below which such a polymer is soluble in an aqueous system.

A "bioelastic polymer" is, in general, a polypeptide that exhibits an inverse temperature transition. Bioelastic polymers are discussed in greater detail below. Such bioelastic polymers are typically elastin-like peptides.

The purification of a fusion protein comprising a protein or peptide of interest and a bioelastic polymer by Inverse Transition Cycling (ITC) is known and described in, for example, U.S. Patent No. 6,852,834 to Chilkoti, the disclosure of which is incoφorated herein by reference. For example, ITC may be carried out by fusing a recombinant protein to a bioelastic polymer whose solubility can be reversibly modulated by the manipulation of environmental variables (including but not limited to solution temperature, ionic strength, pH, ionic composition, and pressure). The target recombinant protein can the be efficiently purified from complex biological mixtures using successive steps of triggering the phase transition to aggregate the bioelastic fusion protein, separating the bioelastic fusion proteins from the mixture (e.g. by centrifugation or filtration), and resuspending the bioelastic fusion proteins (e.g., in low ionic strength buffer). This process can be repeated as needed to increase -the concentration and purity of the bioelastic fusion protein in the purified product.

"Cell culture solution" as used herein encompasses a variety of solutions suitable to recombinant protein purification, including but not limited to a suspension of cultured cells, the suspension of soluble and insoluble material following sonication or other lysing methods (e.g. french press) of a cell suspension, the soluble fraction of a soluble lysate, spent media in which cells where cultured (applicable in the case of secreted expression), etc.

Bioelastic Polymers. In general, the fusion protein comprises a protein of interest and a bioelastic polymer. Bioelastic polymers are known and described in, for example, U.S. Patent No. 5,520,672 to Urry et al. In general, bioelastic polymers are polypeptides comprising elastomeric units of bioelastic pentapeptides, tetrapeptides, hexapaeptides and/or nonapeptides (that is, "elastin-like peptides"). Thus, in some embodiments the elastomeric unit is a pentapeptide, in other embodiments the elastomeric unit is a tetrapeptide, in other embodiments the elastomeric unit is a hexapeptide and in still other embodiments the elastomeric unit is a nonapeptide. Bioelastic polymers that can be used to carry out the present invention are set forth in U.S. Patent No. 4,474,851, which describes a number of tetrapeptide, pentapeptide and hexapeptide repeating units that can be used to form a bioelastic polymer. Specific bioelastic polymers that can be used to carry out the present invention are also described in U.S. Patent. Nos. 4,132,746; 4,187,852; 4,500,700; 4,589,882; and 4,870,055. Still other examples of bioelastic polymers are set forth in U.S. Patent No. 6,699,294 to Urry; U.S. Patent No. 6,753,311 to Fertala and Ko; and U.S. Patent No. 6,063,061 to Wallace. In one embodiment, the bioelastic polymers used to carry out the present invention are polypeptides of the general formula (Val-Pro-Gly-Xaa-Gly)_m (SEQ ID NO:1), where Xaa is any amino acid other than proline (e.g., Ala, Leu, Phe) and m is any suitable number such as 2, 3 or 4 up to 60, 80 or 100 or more. The frequency of the various amino acids as the fourth amino acid can be changed, as well as the identity of Xaa.

For example, bioelastic polymers used to carry out the present invention may comprise repeating elastomeric units selected from the group consisting of bioelastic pentapeptides and tetrapeptides, where the repeating units comprise amino acid residues selected from the group consisting of hydrophobic amino acid and glycine residues and where the repeating units exist in a conformation having a beta-turn of the formula:

wherein R₁-R₅ represent side chains of amino acid residues 1-5, and m is 0 when the repeating unit is a tetrapeptide or 1 when the repeating unit is a pentapeptide. Nonapeptide repeating units generally consist of sequential tetra- and pentapeptides. Preferred hydrophobic amino acid residues are selected from the group consisting of alanine, valine, leucine, isoleucine, proline, phenylalanine, tryptophan, and methionine. In many cases, the first amino acid residue of the repeating unit is a residue of valine, leucine, isoleucine or phenylalanine; the second amino acid residue is a residue of proline; the third amino acid residue is a residue of glycine; and the fourth amino acid residue is glycine or a very hydrophobic residue such as tryptophan, phenylalanine or tyrosine. Particular examples include the tetrapeptide Val-Pro-Gly- GIy (SEQ ID NO:2), the tetrapeptide Gly-Gly-Val-Pro (SEQ ID NO:3), the tetrapeptide Gly-Gly-Phe-Pro (SEQ ID NO:4), the tetrapeptide Gly-Gly-Ala-Pro (SEQ ED NO:5), the pentapeptide Val-Pro-Gly-Val-Gly (SEQ ID NO:6), the pentapeptide Gly-Val-Gly-Val-Pro (SEQ ID NO:7), the pentapeptide Gly-Lys-Gly- Val-Pro (SEQ ID NO:8), the pentapeptide Gly-Val-Gly-Phe-Pro (SEQ ID NO:9), the pentapeptide Gly-Phe-Gly-Phe-Pro (SEQ ID NO: 10), the pentapeptide Gly-Glu-Gly- Val-Pro (SEQ ID NO: 11), the pentapeptide Gly-Phe-Gly-Val-Pro (SEQ ID NO: 12), and the pentapeptide Gly-Val-Gly-Ile-Pro (SEQ ID NO: 13), the nonapeptide Val-Pro- Gly-Val-Gly-Val-Pro-Gly-Gly (SEQ ID NO: 14), and the hexapeptides Val-Ala-Pro- Gly-Val-Gly (SEQ ID NO:15), Ala-Pro-Gly-Val-Gly-Val (SEQ ID NO:16), Pro-Gly- Val-Gly-Val-Ala (SEQ ID NO:17), Gly-Val-Gly-Val-Ala-Pro (SEQ ID NO: 18), VaI- Gly-Val-Ala-Pro-Gly (SEQ ID NO: 19), Gly-Val-Ala-Pro-Gly (SEQ ID NO:20) See, e.g., U.S. Patent No. 6,699,294 to Urry.

Proteins and Peptides of Interest. In general, the fusion protein also comprises a protein or peptide of interest. The protein of interest is preferably a biologically active protein. For example, suitable proteins include those of interest in medicine, agriculture or other scientific or industrial fields. Examples of suitable proteins include antibodies; enzymes utilized in replacement therapy; hormones for promoting growth in animals, or cell growth in cell culture; and active proteinaceous substances used in various applications, e.g., in biotechnology or in medical diagnostics. Specific examples include, but are not limited to, superoxide dismutase, interferon, asparaginase, glutamase, arginase, arginine deaminase, adenosine deaminase ribonuclease, trypsin, chrornotrypsin, papain, insulin, calcitonin, ACTH, glucagon, somatosin, somatropin, somatomedin, parathyroid hormone, erthyropoietin, hypothalamic releasing factors, prolactin, thyroid stimulating hormones, endorphins, enkephalins, interleukins, tumor necrosis factor, angiostatin, endostatin, growth factors, interleukin-1 receptor antagonist, ghrelin, leptin, magainin, and vasopressin.

Other components of the fusion protein may be a spacer element comprising one or more amino acid residues. The spacer element may be inert, may be an additional bioactive protein such as, an intein, may be a cleavable site, and may be combinations thereof. In one embodiment, the spacer is an amino acid sequence recognizable by a specific protease. Examples include sequences cleavable by serine, cysteine (thiol), aspartyl (carboxyl) or metallo-proteases. Such separation permits the phase transition component of the fusion protein to be enzymatically cleaved to enable isolation and/or partial purification of the protein of interest. In another embodiment, the spacer is inert and not uniquely susceptible to enzymatic cleavage by serine, cysteine, aspartyl, or metallo-proteases. In another embodiment, the spacer is a bioactive intein sequence allowing the bioelastic molecule to be separated from the target recombinant protein by the introduction of a reducing agent or by a change in the solution pH, which thereby induces the intein self-cleavage.

Methods of Culturing and Purifying. The fusion protein can be engineered to comprise a signal sequence that causes the fusion protein to be directed to the cell surface or excreted from a recombinant organism that is used to produce the fusion protein. The fusion protein may be cleaved at the cell surface or may be enzymatically cleaved in solution.

The fusion protein may also contain a sequence that permits separate purification by affinity chromatography, commonly referred to as affinity tags. Examples include (His) tag, FLAG, s-tag, etc.

The fusion protein may also contain a "detection tag," i.e., a sequence that is retained on the protein of interest after cleavage of the phase transition component and which by virtue of binding to a reporter molecule can be used to detect the protein of interest (e.g., antibody epitopes for western blot analysis).

Also included within the scope of the invention are derivatives comprising fusion proteins, which have been differentially modified during or after synthesis, e.g., by benzylation, glycosylation, acetylation, phosphorylation, amidation, PEGylation, derivatization by known protecting/blocking groups, proteolytic cleavage, linkage to an antibody molecule or other cellular ligand, etc. In one embodiment, the fusion proteins are acetylated at the N-terminus and/or amidated at the C-terminus. In another embodiment, the fusion proteins are conjugated to polymers, e.g., polymers known in the art to facilitate oral delivery, decrease enzymatic degradation, increase solubility of the polypeptides, or otherwise improve the chemical properties of the therapeutic polypeptides for administration to humans or other animals. The polymers may be joined to the fusion proteins by hydrolyzable bonds. For example, in one aspect where the fusion proteins are therapeutically active, the polymers are joined to the fusion proteins by hydrolyzable bonds, so that the polymers are cleaved in vivo to yield the active therapeutic fusion proteins.

Fusion proteins can be prepared by known recombinant expression techniques or variations thereof that will be apparent to those skilled in the art. To recombinantly produce an fusion protein, a nucleic acid sequence encoding a fusion protein is operatively linked to a promoter such that the fusion protein is produced from the sequence. Preferred promoters are those useful for expression in E. coli, such as the T7 promoter. In a preferred embodiment, the nucleic acid is DNA. Any commonly used cell expression system may be used, e.g., eukaryotic or

•prokaryotic systems. Specific examples include microbial, bacterial, fungal, plant, animal (avian, insect, mammalian) cells, yeast, Pichia, and bacterial systems, such as E. coli, and Caulobacter.

A vector comprising the nucleic acid sequence can be introduced into a cell for expression of the fusion protein. The vector can remain episomal or become chromosomally integrated, as long as it can be transcribed to produce the desired RNA. Vectors can be constructed by standard recombinant DNA technology methods. Vectors can be plasmid, viral, or other types known in the art, used for replication and expression in eukaryotic or prokaryotic cells. It will be appreciated by one. of skill in the art that a wide variety of components known in the art may be included in the vectors of the present invention, including a wide variety of transcription signals, such as promoters and other sequences that regulate the binding of RNA polymerase to the promoter. The operation of promoters is well-known in the art. . Any promoter known to be effective in the cells, in which the vector will be expressed, can be used to initiate expression of the fusion protein. Suitable promoters may be inducible or constitutive. Examples of suitable promoters include the SV40 early promoter region, the promoter contained in the 3' long terminal repeat of Rous sarcoma virus, the HSV-I (herpes simplex virus-1) thymidine kinase promoter, the regulatory sequences of the metallothionein gene, etc., as well as the following animal transcriptional control regions, which exhibit tissue specificity and have been utilized in .transgenic animals: elastase I gene control region which is active in pancreatic acinar cells; insulin gene control region which is active in pancreatic beta cells, immunoglobulin gene control region which is active in lymphoid cells, mouse mammary tumor virus control region which is active in testicular, breast, lymphoid and mast cells, albumin gene control region which is active in liver, alpha-fetoprotein gene control region which is active in liver, alpha 1 -antitrypsin gene control region which is active in the liver, beta-globin gene control region which is active in erythroid cells, myelin basic protein gene control region which is active in oligodendrocyte cells in the brain, myosin light chain-2 gene control region which is active in skeletal muscle, and gonadotropin releasing hormone gene control region which is active in the hypothalamus. In one aspect of the invention, a mammal is genetically modified to produce the fusion protein in its milk. Techniques for performing such genetic modifications are described in U.S. Patent No. 6,013,857, issued Jan. 11, 2000, for "Transgenic Bovines and Milk from Transgenic Bovines." The genome of the transgenic animal is modified to comprise a transgene comprising a DNA sequence encoding an fusion protein operably linked to a mammary gland promoter. Expression of the DNA sequence results in the production of fusion protein in the milk. The fusion protein peptides may then be isolated by phase transition from milk obtained from the transgenic mammal. The transgenic mammal is preferably a bovine.

When expressed in cell culture, the fusion protein may be isolated and/or partially purified by lysing the cells of the cell culture and isolating the fusion protein from solution by inverse transition cycling. Where the fusion protein is secreted from live cells, it will not be necessary to lyse the cells.

The fusion proteins of the invention can be separated from other contaminating proteins to high purity using a phase transition cycling procedure. Methods of isolation can employ the temperature-dependent solubility of the fusion protein. It has been surprisingly discovered that soluble fusion protein can be selectively aggregated by raising the solution temperature above the T_t with no effect on other soluble proteins present in the cell lysate. Successive inverse phase transition cycles may be used to obtain a higher degree of purity (see, e.g., U.S. Patent No. 6,852,834 to Chilkoti).

Other purification techniques may also be employed in conjunction with the inverse phase transition. For example, recombinant cells may be designed to secrete the fusion protein; the cells may be cultured in a cross-flow filter system that permits the secreted fusion protein to diffuse across a membrane. The fusion protein may then be purified from other contaminants by inverse phase transition.

Inverse phase transition can also be induced by depressing the T_t by manipulating other solution conditions. For example, the T_t can be adjusted so that soluble fusion protein can be isothermally aggregated at room temperature, for example, by the addition of a variety of inorganic salts. Because this process is reversible, altering the solution conditions back to the original conditions results in the recovery of soluble, pure, and functionally-active fusion protein. Aggregated fusion protein capture above T_t can be accomplished by centrifugation or filtration. 5. The inverse transition of the ELP also provides a simple method for purifying the ELP tag from the protein of interest after cleavage at a protease recognition site encoded in the primary amino acid sequence between the protein of interest and the ELP carrier. After cleavage, the protein of interest is easily separated from free ELP by another round of inverse transition cycling. 0 In addition to temperature and ionic strength, other environmental variables useful for modulating the inverse transition of fusion proteins include pH, the addition of organic solutes and solvents, side-chain ionization or chemical modification, and pressure.

Although purification of recombinant proteins is the most obvious and5 immediate application of the fusion proteins of the invention, the invention provides other applications in biotechnology and medicine.

The purification technique of the invention can be scaled down and multiplexed for concurrent, parallel laboratory scale purification from numerous cell cultures, to achieve simultaneous purification of proteins from multiple cultures. Such0 high-throughput purification application of the invention can be utilized, for example, to expedite both structure-function studies of proteins and the screening of proteins in pharmaceutical studies.

The present invention is explained in greater detail in the following non- limiting examples. 5

EXAMPLE 1

General Approach of Using ELPs in the Purification of Proteins

A new and simple strategy to purify soluble recombinant proteins from E. coli

^• at a protein concentration that approaches the limit of a single protein molecule per0 cell is disclosed herein. This method utilizes the unique aggregation properties of elastin-like polypeptides (ELPs) to capture recombinant fusion proteins composed of a protein of interest and an ELP tag from cell lysate. ELPs are artificial, genetically encodable polypeptides that are composed of repeating pentapeptides with the sequence Val-Pro-Gly-Xaa-Gly (SEQ ID NO:1), where the guest residue Xaa can be any naturally occurring amino acid except Pro (Urry (1997) J. Phys. Chem. B 101 :11007-11028; Urry (1992) Prog. Biophys. Molec. Biol. 57:23-57; Urry, et al. (1985) Biopolymers 24:2345-2356). ELPs exhibit unique inverse phase transition behavior; below a critical transition temperature (T₁), ELPs are highly soluble in aqueous solution, but at temperatures even a few degrees Celsius above its T_t, an ELP will undergo a solubility-insolubility phase transition, leading to aggregation of the polypeptide (Urry, et al. (1985) Biopolymers 24:2345-2356; Urry (1988) J. Protein Chem. 7:1-34; McPherson, et al. (1996) Protein Expres. Purif. 7:51-57). This behavior is reversible, so that upon lowering the solution temperature below the Z¹,, the ELP completely recovers its solubility. The T₁ of an ELP is a function of a number of variables including the identity and stoichiometry of the guest residue (Xaa), the ELP molecular weight, and the ELP concentration, as well as the type of salt and its concentration in aqueous solution (Meyer & Chilkoti (2004) Biomacromolecules 5:846-851; Urry, et al. (1988) J. Am. Chem. Soc. 110:3303-3305; Meyer & Chilkoti (2002) Biomacromolecules 3:357-367; Urry, et al. (1985) Biochemistry 24:5182- 5189; Girotti, et al. (2004) Macromolecules 37:3396-3400; Li & Daggett (2003) Biopolymers 68:121-129; Urry, et al. (1995) Chem. Phys. Lett. 239:67-74).

The environmental sensitivity and reversible solubility of ELPs are retained when expressed as recombinant fusions with proteins, and this feature can be exploited for simple, non-chromatographic purification of recombinant proteins by inverse transition cycling (ITC) (Meyer & Chilkoti (1999) Nat. Biotechnol. 17:1112- 1115).. In ITC, an ELP fusion protein is selectively separated from other contaminating biomolecules in cell lysate by the sequential and repeated steps of aggregation, centrifugation, and resolubilization of the fusion protein (Meyer & Chilkoti (1999) Nat. Biotechnol. 17:1112-1115; Meyer, et al. (2001) Biotechnol. Prog. 17:720-728). Since the original demonstration of ITC (Meyer & Chilkoti (1999) Nat. Biotechnol. 17:1112-1115), a number of different proteins have been purified by this method using either centrifugation (Meyer, et al. (2001) Biotechnol. Prog. 17:720-728; Trabbic-Carlson, et al. (2004) Protein ScL 13:3274-3284; Trabbic- Carlson, et al. (2004) Protein Eng. Des. SeI. 17:57-66; Stiborova, et al. (2003) BiotechoL Bioeng. 82:605-611; Sun, et al. (2005) J. Proteome Res. 4:2355-2359; Kim, et al. (2005) Anal. Chem. ll-J3\%-2i22; Kim, et al. (2005) Biotechnol. Bioeng. 90:373-379; Banki, et al. (2005) Nat. Methods 2:659-661) or micro-filtration (Ge, et al. (2005) J. Am. Chem. Soc. 127:11228-11229). The direct purification of ELPs has also been extended to the capture of native- proteins by ELP -tagged capture reagents (Stiborova, et al. (2003) Biotechol. Bioeng. 82:605-611; Sun, et al. (2005) J. Proteome Res. 4:2355-2359; Kim, et al. (2005) Anal. Chem. 77:2318-2322; Kim, et al. (2005) Biotechnol. Bioeng. 90:373-379). However, in all of these studies, the ELP fusion proteins were purified from cells at expression levels of 10-200 mg/L culture. At these levels of expression, recombinant proteins can also be effectively purified by other competing techniques such as affinity chromatography (Lichty, et al. (2005) Protein Expr. Pur. 41:98-105; Terpe (2003) Appl. Microbiol. Biotechnol. 60:523-533; Waugh (2005) Trends Biotechnol. 23:316-320).

In contrast, proteins expressed at the level of micrograms per liter of culture are difficult to purify by chromatography because of the losses associated with nonspecific and irreversible adsorption of the protein of interest to chromatography resins as well as the relative increase in contamination from host proteins that are non- specifically adsorbed to the chromatography resin are subsequently eluted with the protein of interest. In this study, a new variant of ITC is provided that solves this problem by the addition of excess free ELP to cell lysate to efficiently drive the phase transition at low concentrations of ELP fusion proteins (Figure 1). This simple modification of ITC enables the purification of ultra-low concentrations of ELP fusion proteins (defined as < 100 μg soluble expressed protein per liter of culture) from E. coli lysate with high purity and good yield.

EXAMPLE 2 Materials and methods

Materials. Phosphate-buffered saline (PBS; 10 mM phosphate, 140 mM NaCl, 3 mM KCl, pH 7.4) was obtained from from Calbiochem (San Diego, CA), TB Dry Powder Growth Media from Mo Bio Laboratories, Inc. (Carlsbad, CA), ACRODISC syringe filters with a 0.2 μm SUPOR membrane from Pall Corporation (Ann Arbor, MI), Millex-HV hydrophobic PVDF syringe filters with a 0.45 μm membrane and vacuum steriflip filters with a DURAPORE PVDF 0.22 μm membrane from Millipore (Billerica, MA), precast Mini-Protean SDS-PAGE gels (4-20 % Tris-HCl) and silver stain kit from Bio-Rad (Richmond, CA), thrombin from EMD Biosciences (San Diego, CA), NCS-II tissue solubilizer and ACS-II aqueous counting scintillant were both obtained from Amersham Biosciences (Piscataway, NJ).

Preparation of Free ELP[V₅A₂G₃^O] and Fusion Proteins. The free ELP used in this study was ELP[VsA₂Gs^O]. The construction of its synthetic gene, its expression in E. coli, and the purification of this ELP has been described elsewhere (Meyer & Chilkoti (2002) Biomacromolecules 3:357-367). The same ELP was fused to thioredoxin (Trx), blue fluorescent protein (BFP), and chloramphenicol acetyltransferase (CAT) at the C-terminal end of the protein. Gene synthesis and protein expression of the three fusion proteins has been described elsewhere (Meyer & Chilkoti (1999) Nat. Biotechnol. 17:1112-1115; Trabbic-Carlson, et al. (2004) Protein ScL 13:3274-3284). AU three fusion proteins have a thrombin cleavage site between the protein and the ELP. The concentrations of the free ELP and the fusion proteins were measured by UV- visible spectrophotometry at 280 nm using the molar extinction coefficients of 5690 M^"1 cm^"1, 1.987xlO⁴ M^"1 cm^""1, 2.442xlO⁴ M^"1 cm^*"1, and 4.882xlO⁴ M^"1 cm^"1 for free ELP, Trx-ELP, BFP-ELP, and CAT-ELP, respectively. The molecular weights of Trx-ELP, BFP-ELP, and CAT-ELP are 49,982.6 Da, 65,489.6 Da, and 64,254.6 Da, respectively.

Biosynthesis of^l4C-Labeled Fusion Proteins. Labeling of ELP fusion proteins with ¹⁴C was carried out using established methods (Liu, et al. (2006) J. Control. Release 114:184-192). Briefly, E. coli were adapted from rich nutrient medium (TB Dry) to a low nutrient medium (a modified M9 medium consisting of 6 g/L Na_JHPO₄, 3 g/L KH₂PO₄, 1 g/L NH₄Cl, 0-.5 g/L NaCl, 4 g/L glucose) by sequential steps of growth in mixtures of TB and the modified M9 in which the amount of TB Dry in the medium was reduced at each step. Once fully adapted to the M9 medium, E. coli were inoculated into 50 mL of modified M9 medium supplemented with 100 μg/mL ampicillin and were cultured overnight. Five mL of the overnight culture were centrifuged. The pelleted cells were resuspended in 5 mL of the modified M9 medium and were used to inoculate 1 L of modified M9 medium supplemented with 0.4 % (w/v) [U-¹⁴C]-D-glucose (Cambridge Isotope Laboratories Inc., Andover, MA) so that ¹⁴C-labeled glucose was the sole carbon source. E. coli were cultured until ODβoo of 2.0 was reached (~6 hours), at which time expression was induced by the addition of IPTG (final concentration 0.5 mM). Induced cells were cultured for 2 hours, and were then were harvested by centrifugation. The ^I4C-labeled fusion protein was purified by ITC (Meyer & Chilkoti (2002) Biomacromolecules 3:357-367; Meyer & Chilkoti (1999) Nat. Biotechnol. 17:1112-1115). The purity of the fusion proteins was determined by SDS-PAGE and the concentrations were determined spectrophotometrically. Fusion Protein Capture Analysis. ¹⁴C-labeled calibration standards for beta- counting of purified ELP fusion proteins samples consisted of 10, 5, 1, 0.5, or 0.1 μg of each ¹⁴C-labeled fusion protein in 1 mL PBS in triplicate, to which to 1 mL of tissue solubilizer was added. The samples were incubated overnight at 50⁰C with gentle shaking and neutralized with 30 mL glacial acetic acid. Fifteen mL of scintillation cocktail was added to each vial, and the samples were incubated in the dark for at least 6 hours before the radioactivity was quantified on a beta-counter (1214 Rackbeta; LKB/Wallac, Gaithersburg, MD). The measured counts per minute (CPM) for each fusion protein mass were averaged, and standard curves were generated by plotting the logarithm of the average CPM versus the logarithm of the mass of each fusion protein, which was independently quantified by UV- visible spectrometry using known extinction coefficients. The percent fusion protein captured from PBS or from soluble E. coli lysate was calculated from the logarithm of their •measured radioactivity in CPM.

Statistical Analysis. Statistical analysis of the ^uC-labeled fusion protein capture data was performed with SPSS software (SPSS, Chicago, IL). One-way ANOVA with post hoc Tukey HSD tests were performed on all the fusion protein capture data from PBS (both with added ELP and without) to determine the significance that free ELP has on the capture of fusion proteins. Following ANOVA a Kolmogorov-Smirnov test was used to test for normality in the data. To determine the effects that different fusion proteins and their concentration have on the capture by free ELP, two-way ANOVA with post hoc Tukey HSD tests were performed on only the fusion proteins capture from PBS where free ELP was added. Finally, one-way ANOVA with post hoc Tukey tests were performed on the fusion protein capture from E. coli lysate data. Capture of ¹⁴C-Labeled Fusion Proteins from Buffer. For all three ¹⁴C-labeled fusion proteins, a known mass of each fusion protein (1 μg or 10 μg) was added to two different 10 mL aliquots of PBS. To one aliquot, free ELP was added to a final concentration of 5 μM,- while free ELP was omitted from the control group. Next, 1.8 g NaCl, corresponding to a final concentration of 3 M NaCl was added to each aliquot. The salt was dissolved by gentle mixing, and the protein solutions were filtered through 0.2 μm ACRODISC syringe filters with a SUPOR membrane at room temperature. Each filter was washed with 10 mL PBS containing 3 M NaCl, 0.5 mL cold PBS was added to each filter, the PBS-wetted filters were incubation at 4°C for 10 minutes, and samples were eluted from the filters with an additional 1.5 mL cold PBS. All pair-wise fusion protein capture experiments were performed in triplicate, and the radioactivity in the eluted samples was measured by scintillation counting.

Capture of ELP Fusion Proteins from E. coli Soluble Cell Lysate. E. coli BLR(DE3) cells (Novagen, Madison, WI) were grown overnight from a stab of a frozen stock in 50 mL of TB Dry media lacking antibiotics with shaking at 300 rpm at 37°C. Ten mL of this culture were used to inoculate 1 L of TB Dry Media, which was incubated for 24 hours without antibiotics at 37°C with shaking at 250 rpm. The final cell yield was determined from culture optical density at 600 ran, assuming that ODβoo = 1 equals 8x10^s cells/mL. The cultures were harvested by centrifugation at 4°C, resuspended in 10 mL PBS buffer, and lysed by ultrasonic disruption (Fisher Scientific 550 Sonic Dismembrator) at 4°C. Two mL of polyethyleneimine (10% w/v) was added per liter of culture before the cell debris was removed by ultracentrifugation at -13000 rpm for 15 minutes at 4°C. The volume of the soluble cell lysate generated for each liter of culture ranged from 15rl8 mL. For experiments requiring lysate generated from more than 1 L of culture, the soluble lysate from multiple 1 L cultures was mixed prior to ELP-fusion protein addition to insure that all samples and controls for a single experiment (i.e., one SDS-PAGE gel) had the same cell lysate composition. To test the effect that added ELP has on the capture of fusion proteins from soluble cell lysate, lysate was prepared as stated above. To the volume of lysate representative of 1 L of E. coli culture a sample of fusion protein was added (10 μg of unlabeled fusion protein for SDS-PAGE analysis and 10 μg or 1 μg of ¹⁴C-labeled fusion protein for scintillation counting). NaCl was dissolved to a final concentration of 3 M, the solutions were filtered though 0.45 μm Millex-HV hydrophobic PVDF syringe filters to remove E. coli contaminants that denature in 3 M NaCl. Free, un- fused ELP (4.4 mM in PBS) was added to 5 mL PBS before combining this solution with the soluble cell lysate, both solutions were chilled on ice. The final concentration of free ELP was 5 μM in the lysate/PBS mixture. After adding free ELP, more NaCl was added to bring the NaCl concentration back to 3 M. For SDS-PAGE analysis, a control capture experiment was performed without the use of free ELP. The solutions were heated to room temperature to allow for the ELP phase transition to occur before filtration through a 0.22 μm vacuum STERIFLIP™ filter (Millipore). Each filter was washed with 150 mL PBS containing 3 M NaCl. The aggregated ELP/ELP fusion protein was eluted from each membrane with a total of 3 mL cold PBS after incubation at 4°C for about 10 minutes. Further purification was performed by five cycles of ITC as described above. After the final hot spin, ¹⁴C-labeled fusion protein samples were added to tissue solubilizer for subsequent scintillation counting as described above, but unlabeled samples for SDS-PAGE were resuspended in only 15 mL PBS. Samples for SDS-PAGE were digested with thrombin (1 μL of a 1 U/μL solution). The ELP phase transition was triggered by the addition NaCl, the free ELP was removed by centrifugation, and the supernatants were loaded onto the gel. To visually estimate from SDS-PAGE the % fusion protein captured from 10 μg of Trx- ELP from lysate, standards of 10 μg, 5 μg, and 1 μg Trx-ELP that had also been digested with thrombin were also run on the gel. The bands on the gel were visualized by silver stain.

It should be noted that filtration (as opposed to centrifugation) was used for the first round of ITC purely because of equipment limitations. A centrifuge capable of spinning ~20 mL radioactive samples was not available for these experiments. Because only microcentrifuges were available for work with radioactive samples, the first filtration step was used to reduce the ~20 mL volume of E. coli lysate to ~2 mL, so that microcentrifuges could be used for subsequent rounds of purification. However, it is contemplated that this technique will be amenable to centrifugation- based ITC for the full purification process, eliminating the need for filtration.

EXAMPLE 3 Purification of Trx. BFP and CAT To demonstrate the feasibility of purifying ultra-low levels of expressed proteins, the purification of three ELP fusion proteins was investigated, i.e., thioredoxin (Trx), blue fluorescence protein (BFP), and chloramphenicol acetyltransferase (CAT) that were each fused at their C-terminus to ELP[V₅A2G3-90], a 36 kDa polypeptide containing 50% valine, 20% alanine, and 30% glycine at the guest residue position (Xaa) of the pentapeptide repeat (Val-Pro-Gly-Xaa-Gly; SEQ ID NO:1). Both unlabeled and ¹⁴C-labeled ELP fusion proteins were separately expressed from E. coli at moderately high levels (20-200 mg/L culture), and purified to homogeneity (as assessed by SDS-PAGE) by ITC under standard conditions prior to the experiments reported in this study.

Effect of Free ELP on Capture of ELP Fusion Proteins from PBS. To determine if the addition of free ELP as a co-aggregant could efficiently capture and purify ELP fusion proteins from solution, a proof-of-principle experiment was performed by mixing 5 μM ELP with three different ¹⁴C-labeled, purified ELP fusion proteins (Trx-ELP, BFP-ELP, and CAT-ELP) at two different concentrations (1 μg and 10 μg fusion protein in 10 mL PBS). The concentration of free ELP of 5 μM was chosen because preliminary experiments indicated that this concentration optimally balanced recovery of Trx-ELP, as measured by thioredoxin reductase activity (Holmgren & Bjomstedt (1984) Methods Enzymol. 107:295-300), with ease of filtration. Concentrations higher than 5 μM increased the filtration pressure without increasing the recovery of Trx-ELP.

The inverse phase transition was triggered by adding NaCl to a final concentration of 3 M, which was chosen because it is high enough to drive the phase transition when this ELP is present at a concentration of 5 μM (Meyer & Chilkoti (2004) Biomacromolecules 5:846-851). In control experiments, the free ELP was omitted to investigate the effect of the ELP co-aggregant on recovery of the ELP fusion proteins. After triggering the ELP phase transition by the addition of NaCl, all fusion protein solutions were filtered through 0.2 μm syringe filters to capture the ELP aggregates, and were washed with 10 mL of a high-salt buffer. The aggregates were then dissolved by reversing the phase transition by injection of 2.0 mL low-salt PBS buffer and eluted from the filters. The radioactivity in each sample was measured by scintillation counting and is reported in counts per minute (CPM). The percent recovery for each fusion for all experimental conditions was calculated from a log-log calibration curve of the CPM versus the mass of fusion protein.

Figure 2 shows that the addition of free ELP results in a 7- to 50-fold greater capture of fusion protein as compared to the control (ITC performed without free ELP). The recovery of the fusion proteins with the addition of free ELP ranged from 45% ± 11% to 93% ± 20%. A one-way ANOVA performed on the logarithm of the percent fusion protein captured found significant differences in the data

vd=₃₀)=49, p=4.8xlθ^'16, R²=0.95), and a Kolmogorov-Smirnov test found the data to be normally distributed (p=0.165). Post hoc Tukey HSD tests confirmed that the amount of fusion protein captured by co-aggregation with free ELP was statistically significantly greater than the amount captured using only added salt to induce the inverse phase transition (p=2.3xlθ^"4, 5.1x10^"", 2.5xlO^"10 for the 10 μg level capture of Trx-ELP, BFP-ELP, and CAT-ELP respectively (Figure 2B) and p=1.7xlθ^"9, 6.0x10^{" 2}, 3x10^"10 for the 1 μg level capture of the same proteins (Figure 2A)). This significant increase in fusion protein capture likely results from more efficient retention of the mixed ELP/ELP fusion protein on the filter because the added ELP not only lowers the T₁ due to the increase in ELP concentration (Meyer & Chilkoti (2004) Biomacromolecules 5:846-851),. but also likely promotes the formation of large aggregates that are more easily retained by the membrane filter. To determine the overall effect of different fusion proteins and their concentration on the capture process with added free ELP, a two-way ANOVA was performed on the logarithm of the percent fusion protein captured by the addition of free ELP. Again, the model found significant differences in the data (F(vπ=₅,

P=S-SxIO^"4, R²=0.76); however, only the effect of fusion partner (i.e., Trx-ELP, BFP- ELP, or CAT-ELP) was significant

Neither. the fusion protein concentration (p=0.091) nor the interaction parameter for the combined effects of fusion protein concentration and identity (p=0.24) were found to be significant over the narrow concentration range investigated (~2 nM to 20 nM). No significant differences were found between the capture of Trx-ELP and BFP-ELP with free ELP (post hoc Tukey HSD p=0.91), but there were significant differences between CAT-ELP and both Trx-ELP and BFP-ELP (P=I^xIO^"4 and 1.7x10^ respectively). This is consistent with the qualitative observations that regardless of concentration, the amount of Trx-ELP and BFP-ELP captured were very similar and were 1.5- to 2-fold greater than the amount of CAT-ELP captured. As a side note, although it may appear that the amount of ELP fusion protein capture without the addition of free ELP increases with decreasing fusion protein concentration, this elevation in the apparent % capture is merely an artifact of the detection limit β- counting in that background radioactivity constitutes a higher fraction of the total radioactivity at the lower fusion protein concentration.

ELP Fusion Proteins Captured by Free ELP in Soluble Cell Lysate. To determine how the ELP co-aggregation technique would actually perform in purifying a poorly expressed ELP fusion protein from E. coli culture, the soluble E. coli lysate from 1 L of culture of (~17 mL) was spiked with 1 μg of each ELP fusion protein, which corresponds to a concentration of ~1 nM ELP fusion protein in the soluble lysate. By measuring the optical density of the E. coli culture at 600 nm prior to cell harvest, and assuming that ODeoo ⁼ 1 corresponds to 8xlO⁸ cells/ml, it was calculated that 1 μg of the fusion proteins corresponded to expression levels of 2-3 ELP fusion protein molecules per E. coli cell.

ELP fusion protein capture from lysate was carried out as described previously with the following minor modifications. First, because of the very high, concentration of E. coli biomolecules in the soluble lysate, a prefiltration step of the lysate was necessary to prevent clogging of the filter. After addition of the ELP fusion protein and NaCl (3 M final concentration), the lysate was raised to room temperature and was filtered through a 0.45 μm membrane to remove non-specifically aggregated protein. The filtered lysate was then chilled on ice, free ELP was added, and the lysate was warmed to room temperature to trigger the phase transition. The aggregates were captured by filtration through a 0.22 μm membrane, and were washed with 150 mL PBS + 3 M NaCl to remove contaminants. The aggregates were then dissolved by reversing the phase transition by injection of 3 mL PBS, and the soluble ELP fusion protein and free ELP were eluted from the membrane with PBS buffer. The eluted fusion protein and free ELP were then further purified by five rounds of ITC utilizing centrifugation (in place of filtration) to purify the fusion protein from any remaining E. coli proteins.

Figure 3 shows an SDS-PAGE gel demonstrating the enhanced recovery with the addition of 5 μM free ELP from lysate containing 10 μg of Trx-ELP. Lanes 5 and 6 show the amount of Trx recovered for each capture condition (with no addition of free ELP and with addition of 5 μM free ELP) after thrombin cleavage of the Trx from its fused ELP and after it had been subjected to an additional round of ITC to remove the ELP tag from solution. These additional purification steps were needed for visualization of Trx by silver staining because the amount of free ELP in the samples where ELP was added would have so overloaded the gel that ELP band would have obscured the observation of the Trx-ELP band.

•Comparison of lane 5 (without added ELP) and lane 6 (with free ELP) shows ■5 the enhanced recovery of Trx with the addition of free ELP. Qualitative assessment of the intensity of the Trx band in lane 6 relative to the Trx standards (lanes 2-4), which were created from the thrombin cleavage of 10 μg, 5 μg, and 1 μg Trx-ELP indicates that the amount of Trx captured was somewhere between 1 μg and 5 μg (between 10 and 50% recovery). Although the thrombin used to cleave Trx from the0 ELP tag was not removed from solution and is visible in air samples in Figure 3, biotinylated thrombin is commercially available (Novagen, Madison, WI), which

.would allow thrombin to be removed with the use of streptavidin beads during the final ITC purification step to remove the cleaved ELP and free ELP. Furthermore, the

. final ITC purification step to remove cleaved and free ELP from the protein of interest5 can be induced with only a minimum of added salt since the previous rounds of ITC concentrate the ELP/ELP fusion protein. In this case, the free and cleaved ELP was transitioned with the addition of 400-600 mM NaCl to achieve the samples shown in Figure 3; however this step was in not optimized and the ~2 mM solution of free and cleaved ELP achieved after five rounds of ITC should have transitioned at 34°C0 without the addition of NaCl. Independent measurement of recovery by scintillation counting confirmed that 39 ± 2% of the Trx-ELP was recovered with the addition of free ELP, which is consistent with the intensity of the Trx band in lane 6. Furthermore, SDS-PAGE demonstrates that this purification protocol results in very pure protein (>95% purity). 5 To examine the lower limit of protein purification by this methodology, the same purification protocol was performed to analyze the recovery of 1 μg of ^Relabeled Trx-ELP, BFP-ELP, and CAT-ELP from the volume of lysate generated from 1 L of E. coli culture. In these experiments, the recovery of the ELP fusion proteins was tracked using ¹⁴C-labeled protein and scintillation counting. Control experiments0 without added ELP were omitted as recovery from buffer (Figure 2) as well SDS- PAGE (Figure 3) indicated that capture without added ELP does not result in the recovery of detectable amounts of ELP fusion protein. Figure 4 shows that 62 ± 12, 31 ± 4, and 19 ± 7 % of the 1 μg of Trx-ELP, BFP-ELP, and CAT-ELP fusion proteins, respectively, were recovered with the addition of free ELP. One-way ANOVA performed on the logarithm of the % fusion protein captured from the soluble lysate of E. coli culture (Figure 4) showed that there were significant differences between the three proteins (F_(vn=₂, _Vd=_Of=^₅ p=4.7xlθ^~3, R²=0.83) and post hoc Tukey HSD test indicate, similar to their capture from PBS, that Trx and CAT were significantly different (p=3.8xlθ^"3), while the difference between BFP and CAT approached significance (p=0.051).

EXAMPLE 4 Purification of Antibodies

The method of the invention was also applied to the capture and concentration of small amounts of antibodies. As a proof of principle experiment, green fluorescent protein (GFP) fused to ELP[VaI₅AIa₂GIy₃]-! 80 was employed to bind and capture an anti-GFP antibody. Free un-fused ELP was added to the solution as a co-aggregant and purification was carried out by inverse transition cycling (ITC). Figure 5 depicts the purification scheme employed in the purification of an anti-GFP antibody.

The experiment was carried out by using anti-GFP antibody covalently linked to horseradish peroxidase. Anti-GFP was added to 10 mL PBS buffer so that the concentration of antr-GFP was 0.5 nM. GFP/ELP was added at two different concentrations, 2 μM or 1 μM. A final concentration of free ELP of 10 μM was added and the phase transition was triggered by adding NaCl. The obtained pellet was resuspended in 150 μL PBS buffer and the results are shown in Figure 6. The amount of captured antibody was measured by the horseradish peroxidase assay using 2,2'- azino-bis(3-ethylbenzthiazoline-6-sulphonic acid) (ABTS) as the substrate. The results of this analysis indicated that, even with small amounts of the ELP fusion protein, substantial amounts of antibody could be captured and concentrated.

EXAMPLE 5 Advantages of ELP Purification of Proteins It has now been demonstrated that the ELP fusion tag technique can be used to purify ultra-low levels of ELP fusion proteins (15-20 pmoles of protein in a liter of culture, which corresponds to 2-3 protein molecules per cell) by the addition of an excess of free ELP to facilitate the selective aggregation of the fusion protein from cell lysate. These results are notable for the following reasons. This facile and inexpensive purification of the ultra-low level of proteins purified by this methodology is unprecedented, and likely cannot be matched by competing techniques such as affinity chromatography. While 200 pM of streptavidin-binding peptide (SBP) fusion protein was captured utilizing immobilized streptavidin matrix in a single step of affinity chromatography, it was speculated that this tag may be advantageous for purification of 10-500 μg of SBP fusion proteins (Keefe, et al. (2001) Protein Expr. Pur. 23:440-446). However, this technique has some limitations including fusion protein loss due to overloading of the streptavidin matrix. While tandem affinity techniques have been used to purify a few micrograms of recombinant protein from a liter of culture (Schimanski,et al. (2005) Eukaryotic Cell 4:1942-1950; Elbing, et al. (2006) Biochem. J. 393:797-805), similar to the amounts of ELP fusion proteins reported herein, tandem affinity purification is significantly more costly and lengthy, resulting in final purified products which may require subsequent processing (e.g., desalting, concentration). In this regard, particular embodiments of the present invention relate to the direct use of the purified product without subsequent purification or concentration.

ELP fusion protein capture is also notable because 20-60% of the fusion protein was captured at the lower limit of 1 μg of protein (Figure 4), and the amounts of the recovered proteins are in the picomole range from a liter of culture, which provides enough protein for downstream analysis by mass spectrometry (Johnson, et al. (2005) Methods 35:223-236; Shevchenko, et al. (1996) Proc. Natl. Acad. Sd. USA 93:14440-14445; Smith (2002) Trends Biotechnol. 20:s3-s7; Savitski, et al. (2005) J. Proteome Res. 4:2348-2354; Chen, et al. (2005) Anal. Chem. 77:8179-8184; Roesli, et al. (2006) Curr.Opin. Chem. Biol. 10:35-41; Rappsilber, et al. (2003) Int. J. Mass Spectrom. 226:223-237; Bogdanov, et al. (2005) Mass Spectrom. Rev. 24:168-200). Furthermore, because of the detection limits of scintillation counting for ¹⁴C radioactivity, the lower limits of this technique were not tested; however, no inherent limitations are contemplated in purifying even lower concentrations of ELP fusion proteins by this methodology.

Finally, because the purified ELP fusion protein is concentrated during purification by virtue of the ITC process, this methodology has the useful ancillary benefits of desalting and concentrating the purified protein for subsequent downstream analysis. In contrast, the streptavi din-binding peptide (SBP)- and tandem affinity purification (TAP)-tagged fusion proteins must be eluted from the matrix in the final step of chromatography, diluting the purified protein and requiring subsequent purification steps to concentrate the protein and/or dialyze the protein to remove eluants such as biotin or EGTA.

ELP capture of low level ELP fusion proteins represents a significant step forward in the purification of poorly expressed proteins. The capture efficiency and protein purity produced by this technique are competitive with those of established chromatographic techniques while maintaining all the intrinsic advantages of conventional ITC purification of ELP fusion proteins; it is inexpensive, easy, requires no specialized equipment, and the purified protein is finally recovered at a desired concentration and in the buffer of choice. Moreover, it is contemplate that this technique will be useful for the purification of proteins secreted from mammalian cells, where low or moderately expressed proteins are diluted by their secretion into the surrounding medium (Hanania, et al. (2005) Biotech. Bioeng. 91:872-876; Wurm (2004) Nat. Biotechnol. 22:1393-1398), and as such should be of great utility to the study of the human proteome.

The foregoing is illustrative of the present invention, and is not to be construed as limiting thereof. The invention is defined by the following claims, with equivalents of the claims to be included therein.

Claims

THAT WHICH IS CLAIMED IS:

1. In a method of purifying a fusion protein from a cell culture containing the same by inverse transition cycling, said fusion protein comprising a protein or peptide of interest coupled to a bioelastic polymer, the improvement comprising: adding free bioelastic polymer as a co-aggregant to said cell culture solution so that the mass of fusion protein captured from the soluble cell lysate is increased.

2. The method of claim 1, said cell culture containing less than 1 milligram ofsaid fusion protein /Liter of culture.

3. The method of claim 1, said cell culture containing less than 100 micrograms of said fusion protein/Liter of culture.

4. The method of claim 1, said cell culture containing less than 1 microgram of said fusion protein/Liter of culture.

5. The method of claim 1, said cell culture containing less than 3000 molecules of said fusion protein per cell in said cell culture.

6. The method of claim 1, said cell culture containing less than 3 x 10^'16 grams of protein per cell in said cell culture.

7. The method of claim 1, wherein said cell culture is a plant, animal, yeast or bacterial cell culture.

8. The method of claim 1, wherein said cell culture is an Escherichia coli cell culture.

9. The method of claim 1, wherein said inverse transition cycling comprises a separating step.

10. The method of claim 9, wherein said separating step is carried out by centrifugation or microfiltration.

11. The method of claim 1 wherein said inverse transition step comprises the step of further adding a precipitant to induce the bioelastic polymer phase transition.

12. The method of claim 11, wherein said precipitant is selected from the group consisting of inorganic salts, organic solvents, and polymers.

13. The method of claim 1, wherein said adding step is carried out by combining a solution of bioelastic polymer with said cell culture solution to obtain a final concentration of 1-10 micromolar of bioelastic polymer.

14. The method of claim 1, wherein said adding step is carried out by combining a solution of bioelastic polymer with said cell culture solution to obtain a final concentration of 10-100 micromolar of bioelastic polymer.

15. The method of claim 1, wherein said adding step is carried out by combining a solution of bioelastic polymer with said cell culture solution to obtain a final concentration of 100-1000 micromolar of bioelastic polymer.

16. In a method of purifying a fusion protein from a cell culture containing the same by inverse transition cycling, said fusion protein comprising a protein or peptide of interest coupled to a bioelastic polymer, the improvement comprising: adding free bioelastic polymer as a co-aggregant to said cell culture solution so that the mass of fusion protein captured from the soluble cell lysate is increased; wherein said cell culture is a plant, animal, yeast or bacterial cell culture; wherein said inverse transition cycling comprises a separating step; and wherein said inverse transition step comprises the step of further adding a precipitant to induce the bioelastic polymer phase transition; and wherein said step of adding free bioelastic polymer is carried out by combining a solution of bioelastic polymer with said cell culture solution to obtain a final concentration of 1- 1000 micromolar of bioelastic polymer.

17. The method of claim 16, said cell culture containing less than lmilligram of said fusion protein /Liter of culture.

18. The method of claim 16, said cell culture containing less than 3000 molecules of said fusion protein per cell in said cell culture.

19. The method of claim 16, said cell culture containing less than 3 x 10^"16 grams of protein per cell in said cell culture.

20. The method of claim 16, wherein said cell culture is an Escherichia coli cell culture.

21. The method of claim 16, wherein said separating step is carried out by centrifugation or microfiltration.

22. The method of claim 16, wherein said precipitant is selected from the • group consisting of inorganic salts, organic solvents, and polymers.