WO2000018905A1 - Method of producing permuteins by scanning permutagenesis - Google Patents

Abstract

A method of producing circularly-permuted proteins (permuteins) by scanning permutagenesis comprises making and inserting a series of circularly-permuted genes into a display vector, expressing these genes such that the gene products are localized to the surface of the display vector, generating a library of display vectors presenting the permuted protein, affinity-selecting the display vectors with a target protein that can bind the permuted protein, isolating and analyzing clones of selected display vectors to identify the circularly-permuted protein. The invention further discloses methods of expressing and uses of permuteins.

Description

Method of producing permuteins by scanning permutagenesis

Priority

The present application claims priority under Title 35, United States Code, § 119 of United States Provisional Application Serial No. 60/101,908, filed

September 25, 1998.

Field of the invention

A method of producing circularly-permuted proteins (permuteins) by scanning permutagenesis comprises making and inserting a series of circularly- permuted genes into a display vector, expressing these genes such that the gene products are localized to the surface of the display vector, generating a library of display vectors presenting the permuted protein, affinity-selecting the display vectors with a target protein that can bind the permuted protein, isolating and analyzing clones of selected display vectors to identify the circularly-permuted protein. The invention further discloses methods of expressing and uses of permuteins.

Background of the invention

Protein permutagenesis

Circularly permuted proteins are made by reordering the primary sequence of a parent protein. The amino and carboxy terminal ends of the parent protein are joined by a peptide linker and new amino and carboxy terminal ends are generated at other positions in the sequence. This technique of generating variants has been applied to a wide variety of proteins (Table 1).

Circularly permuted proteins, in many cases, are structurally and functionally similar to their non-permuted parent molecule after they undergo refolding. The information necessary to direct the folding of proteins into tertiary structures is present in secondary structural domains. Vectorial folding of proteins from their native amino to carboxy ends is not often observed. The ability of permuteins to retain structural and functional properties is remarkable, extending earlier observations on the plasticity of proteins with respect to amino acid substitutions (Olins P.O. et al., J. Biol. Chem. 270: 23754-23760, 1995; Lowman and Wells, J. Mol. Biol. 234: 564-578, 1993) and short amino acid insertions (Sondek, J. and D. Shortle, Proteins 7: 387-393, 1990; Shortle, D. and J. Sondek, Curr. Opin. Biotechno 6: 299-305).

Protein sequence reorganization

Rearrangements of DNA sequences serve an important role in evolution by generating a diversity of new proteins differing in structure and function. Gene duplication and exon shuffling, for example, generate diversity and provide organisms with a competitive advantage since the basal mutation rate is low (Doolittle, Protein Science 1: 191-200, 1992).

Recombinant DNA methods have facilitated studies on the effect of sequence transposition on protein folding, structure, and function. The first rearrangement of proteins using this approach was described by Goldenberg and Creighton (J. Mol Biol. 165:407-413, 1983). A new N-terminus is selected at an internal site (breakpoint) of the original sequence, the new sequence having the same order of amino acids as the original from the breakpoint until it reaches an amino acid that is at or near the original C-terminus. At this point the new sequence is joined, either directly or through an additional portion of sequence (linker), to an amino acid that is at or near the original N-terminus, and the new sequence continues with the same sequence as the original until it reaches a point that is at or near the amino acid that was N-terminal to the breakpoint site of the original sequence, this residue forming the new C-terminus of the chain. Similar approaches have also been used in other studies (Cunningham et al., Proc. Natl. Acad. Sci. U.S.A. 76:3218-3222, 1979; Teather & Erfle, J. Bacteriol. 172: 3837- 3841. 1990; Schimming et al., Eur. J. Biochem. 204: 13-19, 1992; Yamiuchi and

Minamikawa, FEES Lett. 260:127-130, 1991: MacGregor et al., FEBS Lett. 378:263-266, 1996).

These general approaches have been applied to proteins which range in size from 58 to 462 amino acids (Goldenberg & Creighton, J. Mol. Biol. 165:407- 413, 1983: Li & Coffino, Mol. Cell. Biol. 13:2377-2383, 1993). The proteins represent a broad range of structural classes, including proteins that contain predominantly alpha helix (interleukin-4; Kreit an et al., Cytokine 7:311-318, 1995), beta sheet (inteτleukin-1; Horlick et al., Protein Eng. 5:427-431. 1992), or mixtures of the two types of secondary structures (yeast phosphoribosyl anthranilate isomerase; Luger et al., Science 243:2,06-210, 1989). Although broad categories of protein function are represented in these sequence reorganization studies, the results of these studies have been highly variable. In many cases substantially lower activity, solubility, or thermodynamic stability were observed (E. coli dihydrofolate reductase, aspartate transcarbamoylase, phosphoribosyl anthranilate isomerase, glyceraldehyde-3- phosphate dehydrogenase, ornithine decarboxylase, ompA, yeast phosphoglycerate dehydrogenase). In other cases, the sequence rearranged protein appeared to have ^~" many nearly identical properties as its natural counterpart (basic pancreatic trypsin inhibitor, T4 lysozyme, ribonuclease Tl, Bacillus β-glucanase, interleukin- lβ, α-spectrin SH3 domain, pepsinogen, interleukin-4). In exceptional cases, an unexpected improvement over some properties of the natural sequence was observed, e.g., the solubility and refolding rate for rearranged α-spectrin SH3 domain sequences, and the receptor affinity and anti-tumor activity of transposed interleukin-4-Psezzdomonαs exotoxin fusion molecule (Kreitman et al., Proc. Natl Acad. Sci. U.S.A. 91:6889-6893, 1994; Kreitman et al, Cancer Res. 55:3357-3363,

1995).

The primary motivation for reorganization studies has been to study the role of short-range and long-range interactions in protein folding and stability. Sequence rearrangements of this type convert a subset of interactions that are long-range in the original sequence into short-range interactions in the new sequence, and vice versa. The fact that many of these sequence rearrangements are able to attain a conformation with at least some activity is persuasive evidence that protein folding occurs by multiple folding pathways (Viguera et al., J. Mol Biol. 247:670-681, 1995). In the case of the SH3 domain of alpha-spectrin, choosing new termini at locations that corresponded to beta hairpin turns resulted in proteins with slightly less stability, but which were nevertheless able to fold.

The positions of the internal breakpoints used in the studies cited above are found exclusively on the surface of proteins, and are distributed throughout the linear sequence without any obvious bias towards the ends or the middle (the variation in the relative distance from the original N-terminus to the breakpoint is ca. 10 to 80% of the total sequence length). The linkers connecting the original N- and C-termini in these studies have ranged from 0 to 9 residues. In one case (Yang & Schachman, Proc. Natl. Acad. Sci. U.S.A. 90:11980-11984, 1993), a portion of sequence has been deleted from the original C-terminal segment, and the connection made from the truncated C-termmus to the original N-terminus.

Flexible hydrophilic residues such as Gly and Ser are frequently used in the linkers. Viguera et al.(J. Mol. Biol. 247:670-681, 1995) compared joining the original N- and C- termini with 3- or 4-residue linkers; the 3-residue linker was less thermodynamically stable. Protasova et al. (Protein Eng. 7:1373-1377, 1994) used 3- or 5-residue linkers in connecting the original N-termini of E. coli dihydrofolate reductase; only the 3-residue linker produced protein in good yield.

Protein permutagenesis can be used to optimize the activity of fusion proteins or proteins conjugated to other molecules. A fusion between interleukin-4 _ (IL-4) and Pseudomonas exotoxin has been permuted resulting in a protein that has the first amino acid of the IL-4 domain at position 38 and the new carboxy end occurs at amino acid position 37 (Kreitman, R. J. et al., Proc Natl Acad Sci USA 91: 6889-6893, 1994). The permuted fusion has increased affinity for the IL-4 receptor, increased cytotoxicity to IL-4 receptor bearing renal carcinoma cells, and increased anti-tumor activity in a murine model, compared to the non-permuted parent fusion protein (Kreitman, R. J. et al., Proc Natl Acad Sci USA 91: 6889-6893, 1994; Kreitman, R. J. et al., Cancer Res. 55:3357-3363, 1995; Puri, R. K. et al., Cellular Immunol. 171: 80-86, 1996). Increased potency of the permuted molecule is believed to result from a reduction in steric interference between the IL-4 domain in the parent molecule and its receptor.

Steric hindrance is likely to be a concern for other chimeric proteins which interact with receptors through a relatively large area of their surface. The same issue also arises with bioconjugates, containing relatively small chemicals conjugated to proteins or other molecules in complex polymers (Rose, K. et al.,

Molecular Immunology 32: 1031-1037, 1995).

Phage display methods

Display methods allow affinity selection of protein variants from a library of displayed proteins or peptides (Clackson, T. and J.A. Wells, Tibtech 12: 173-184,

1994; Winter, G., Drug Development Res. 33: 71-89, 1994). Many biological entities can be used in display methodologies (so-called "genetic packages" for presentation, including bacterial and eukaryotic cells, various eukaryotic and prokaryotic viruses, and spores), but the moβt commonly used vehicles used for display are filamentous bacteriophage, as used herein. We envision the possibility that a genetic package other than the particular phage used here could be used to present libraries of permuteins, and if so, constitute essentially the same invention. Foreign proteins are presented on the surface of a phage particle, and the gene encoding the foreign protein is encapsulated in the virion. Because they are linked by the phage particle, affinity isolation of the presented protein also leads to affinity isolation of the corresponding genes. Extremely large libraries of phage presented proteins are constructed and affinity screened very rapidly. From the standpoint of how quickly mutant proteins can be made and screened for activity, phage display is the most efficient mutagenesis technique currently available.

Functional properties of permuteins

Permuteins can have improved biological properties by acting through several mechanisms. The permutein acting on the same type of cell as its parent molecule, may have increased binding, or other action, by virtue of increased avidity. Dimers or higher order multimers of these proteins with themselves or other chemical groups, including proteins, can have increased efficacy or potency, or both.

Permuteins can also have improved therapeutic properties through a variety of mechanisms such as: (1) alterations in the overall on- or off-rates or K_a or K_d of the ligand(s) on the target cell; (2) activation or blockade of complementary receptor signaling pathways; and/or (3) more specific targeting of to the cell of interest. The permuteins may also possess a unique pharmacokinetic distribution and clearance profile (Deh er et al., Circulation, 91, 2188-2194, 1995; Tanaka et al.,. Nature Medicine, 3, 437-442, 1997).

Permuteins can also have improved properties in vivo, compared to the two components individually, as a result of alterations in biodistribution or half-life.

The improved properties can also result from the binding of the permutein to one or more of the receptors, pharmacokinetics, or uptake of the permutein is altered in a favorable manner.

Molecular biology approaches have traditionally been used to permute proteins (Horlick, R.A. et al., Protein Engineering 5: 427-431, 1992) although chemical approaches have been used to make small permuted proteins (Goldenberg, D. P. and T. E. Creighton, J. Mol. Biol. 165: 407-413, 1983). These approaches are relatively labor intensive, limiting the number of permuteins that can be generated and efficiently screened for the desired biological activities. Rapid methods of generating permuteins. coupled with efficient methods for screening are needed that will result in the identification of novel active molecules. Summary of the invention

The present invention is an improved method for generating permuteins

(scanning permutagenesis) based on the display of proteins on bacteriophage surface proteins. Phage display is a powerful, yet convenient tool, traditionally used for mutagenesis and screening (Clackson, T. and J.A. Wells, Tibtech 12: 173-

184, 1994). Improvements to this technology allow the rapid generation and - screening of libraries of permuteins. Variables, such as position of the new termini and the length and composition of peptide linkers can easily be varied to generate libraries of the desired diversity.

The present invention relates to methods of producing biologically-active circularly permuted proteins of the formula C'-L'-N¹, derived from a parent protein of the formula N'-C, wherein C¹ is comprised of a segment derived from the carboxy portion of said parent protein; N¹ is comprised of a segment derived from the amino terminal portion of said parent protein; and L¹ is a chemical bond or a linker, linking C to the amino terminus of L¹ and carboxy terminus of L¹ to the amino terminus of N¹; comprising the steps of: (a) making a series of circularly- permuted genes; (b) inserting said circularly-permuted genes into a display vector; (c) expressing said circularly-permuted genes such that the proteins encoded by said genes are presented on the surface of the display vector; (d) generating a library of display vectors presenting the expressed circularly permuted protein; (e) affinity-select the presenting display vectors with a target protein that can bind a biologically-active circularly-permuted protein; (f) isolate and analyze clones of selected display vectors to identify the presented circularly-permuted protein.

Preferably the method of making a series of circularly-permuted genes is selected from the group consisting of making a tandemly-repeated intermediate, total synthesis of a synthetic gene, assembly of a gene from synthetic oligonucleotides, DNA amplification, and limited digestion of a circular intermediate.

Preferably, the display vector is selected from the group consisting of bacteriophage display vectors, bacteria, and baculovirus vectors. Even more preferably the presentation vector is a bacteriophage. Even more preferably, the presentation vector is bacteriophage M13. Most preferably, the presentation vector is a bacteriophage M13 gene III vector. Preferably the method of making a series of circularly permuted genes is a method of making a tandem repeat intermediate. Even more preferably circularly permuted genes are amplified from the repeat by gene amplification.

Preferably the method of affinity selection comprises the steps consisting of (a) binding said presentation display vectors to a target protein; (b) eluting said display vectors; (c) amplifying said display vectors; and (d) biopanning a pool of _ said amplified display vectors.

Preferably, the length of C in the permutein is longer than the length of C¹ in said parent protein. More preferably, the length of C in the permutein is shorter than the length of C in said parent protein. Most preferably, the length of

C in the permutein is the same length as the length of C in said parent protein.

Preferably, the length of N¹ in the permutein is longer than the length of N^l in said parent protein. More preferably, the length of N¹ in the permutein is shorter than the length of N¹ in said parent protein. Most preferably, the length of N¹ in the permutein is the same length as the length of N¹ in said parent protein.

The invention also contemplates circularly permuted proteins of the formula C'-L'-N¹ made by the method of scanning permutagenesis. Preferably, the DNA sequence encoding said linker L¹ is selected from the group consisting of SEQ ID NO: 1 through SEQ ID NO: 368.

Preferably, the circularly-permuted protein is the G-CSF receptor agonist domain of a species of mylepoietin (MPO). MPO is one member of a family of novel dual cytokine receptor agonists (McKearn, J.P., Myelorestorative activities of synthokine and myelopoietin. In Proceedings of the 1996 IBC Conference on Therapeutic Applications of Cytokines, pp. 4.3.1-4.3.18, 1996) which are amenable to manipulation by phage display (Merlin, S. et al., Applied Biochemistry and

Biotechnology 67: 15-29, 1997; Lee, S.C., Phage presentation for cytokine engineering. IBC's Second International Conference on Display Technologies, 1997).

Brief description of the figures

Figure 1. Schematic depiction of scanning permutagenesis

Plate A of Figure 1 shows the strategy to generate a scanning permutagenesis phage display library. A plasmid containing directly-repeated tandem copies of the hG-CSF gene, for example, is constructed by standard methods. The tandem repeat plasmid is used as the template for PCR amplification of genes encoding permuted proteins. Each copy of the G-CSF gene is indicated in light gray (turquoise), and a DNA segment encoding a peptide linker is indicated in dark gray (red).

In individual PCR reactions, oligonucleotide primers that initiate PCR polymerization at the first nucleotide of a chosen codon of G-CSF, and directing polymerization to the end of the tandem construct specifying the carboxy end of the protein encoded on the template is annealed to the tandem template. Also, a second specific primer is also annealed to the template that initiates polymerization at the last nucleotide of the codon encoding the amino acid immediately preceding the codon where polymerization begins with first primer, and which directs polymerization in the opposite direction from that first primer. Amplification between these two primers produces a DNA segment encoding a permuted protein. For example, amplification between the primer indicated by a black arrow initiating at codon 2 and the primer indicated by the blue arrow and initiating at the codon before 2 (codon 1) produces an amplified gene encoding a permuted protein whose amino terminal residue is amino acid 2 of the native protein, and whose final amino acid is amino acid 1 of the native protein.

A linker peptide is present between the first and final amino acids of the parent protein (residues 1 and 174 in this example). A total of 174 individual amplifications would produce a complete collection of all permuted proteins of this example. More limited collections containing only a selected set of permuteins can be made, as well as more extensive collections made from multiple tandem template plasmids, each containing a different linker sequence between the first and last residues of the two directly repeated tandem gene sequences. The collection of amplified segments can then be inserted into a phagemid presentation vector by standard methods. Phagemid particles produced from these presentation constructs are the scanning permutagenesis phage display library.

Plate B of Figure 1 shows the affinity screening of a phage display library

(See Clackson, T. and J.A. Wells, Tibtech 12: 173-184, 1994; Winter, G., Drug Development Res. 33: 71-89, 1994). In this example, a hG-CSF scanning permutagenesis library as described in Figure 1A is screened using the hG-CSF receptor expressed on mammalian cells as the affinity reagent. In Figure IB, individual presented proteins are indicated by the shaded circles or diamonds and the affinity reagent is indicated by the light gray (pink) rectangles. Presentation library particles are exposed to affinity reagent, unbound particles are washed away, and receptor-bound particles are eluted. The eluted particles are amplified in E. coli, and the screening cycle is repeated. During any round of the screening cycle, the genes encoded (in the present example encoding permuted proteins) by the selected particles can be expressed and evaluated.

Figure 2. Permuteins presented in the scanning permutagenesis library

Human G-CSF (serl7) protein is depicted as a string of circles, each circle corresponding to a single amino acid residue. Amino and carboxy ends of the protein are indicated. The amino acids of helical regions are indicated by medium gray balls, while the amino acids of inter-helical loops are indicated in light gray balls (See Hill et al., Proc. Natl. Acad. Sci. USA 90: 5167-5171, 1993). Amino ends of the permuteins made for presentation in the library are indicated in dark gray. Asterisks indicate the breakpoints of the presented permuteins which were isolated by affinity screening with cells expressing hG-CSF receptor as illustrated in IB.

Figure 3. Bioactivity of permuteins identified by affinity screening of the scanning permutagenesis library

Individual permuteins were expressed transiently in mammalian ςells.

Permeation molecules in the culture supernatants were quantitated by ELISA, and the proliferative activity of clones was determined using BAF-3-cells dependent on

G-CSF for growth. The horizontal axis indicate concentration of protein and the vertical axis indicate incorporation of tritiated thymidine.

Definitions

The following is a list of abbreviations and the corresponding meanings as used interchangeably herein: g = gram(s) mg = milligram(s) ml and mL = milliliter(s)

RT = room temperature ug and μg = microgram(s) uL and μl = microliter(s)

The following is a list of definitions of various terms used herein: The term "permutein" means a circularly-permuted protein: a protein in which the amino and carboxy ends of the parent protein are joined together by a peptide linker sequence of zero or more amino acids. The amino and carboxy ends of the permuted protein occur at amino acids within the parental sequence.

The terms "chemical ligation" and "conjugation" mean a chemical reaction which covalently links two similar or dissimilar functional groups together _ intramolecularly or intermolecularly.

The term "peptide linker' means a compound which forms a carboxamide bond between two groups having one or more peptide linkages (CONH-) and serves as a connector for the propose of amelioration of the distance or space orientation between two molecules.

The term "native sequence" refers to an amino acid or nucleic acid sequence which is identical to a wild-type or native form of a gene or protein.

The terms "mutant amino acid sequence," "mutant protein", "variant protein", "mutein", or "mutant polypeptide" refer to a polypeptide having an amino acid sequence which varies from a native sequence due to amino acid additions, deletions, substitutions, or all three, or is encoded by a nucleotide sequence from an intentionally-made variant derived from a native sequence.

Detailed description of the invention

Determination of the amino and carboxyl termini of permuteins

The present invention encompasses circularly permuted-proteins of the formula C'-L'-N¹ prepared by phage display techniques. The polypeptide can be joined either directly or through a linker segment. The term "directly" defines permuteins in which the polypeptide ends are joined without a linker. Thus L¹ represents a chemical bond or a linker, preferably a polypeptide segment to which both C¹ and N¹ are joined, wherein C¹ is comprised of a segment derived from the carboxy portion of the parent protein and N¹ is comprised of a segment derived from the amino terminal portion of a parent protein represented by the general formula N'-C¹. Preferably, N¹ and C¹ in the permuted protein C'-L'-N¹ are the same length as in the parent protein N'-C, but each may be independently shorter or longer depending on the desired structural characteristics of the permutein. Most commonly L' is a linear peptide in which C and N' are joined by amide bonds, linking C to the amino terminus of L' and carboxy terminus of L' to the amino terminus of N¹.

Additional peptide sequences may also be added to facilitate purification or identification of permuteins (e.g., poly-His). A highly antigenic peptide may also be added that would enable rapid assay and facile purification of the permuteins by a specific monoclonal antibody. _

Determination of the linker

The linking group (L¹) is generally a polypeptide of between 1 and 500 amino acids in length. The linkers joining the two molecules are preferably designed to (1) allow the two molecules to fold and act independently of each other,

(2) not have a propensity for developing an ordered secondary structure which could interfere with the functional domains of the two proteins, (3) have minimal hydrophobic characteristics which could interact with the functional protein domains and (4) provide steric separation of C and N¹ such that C¹ and N' could interact simultaneously with their corresponding receptors on a single cell.

Typically surface amino acids in flexible protein regions include Gly, Asn and Ser. Virtually any permutation of amino acid sequences containing Gly, Asn and Ser would be expected to satisfy the above criteria for a linker sequence. Other neutral amino acids, such as Thr and Ala, may also be used in the linker sequence. Additional amino acids may also be included in the linkers due to the addition of unique restriction sites in the linker sequence to facilitate construction of the multi-functional proteins.

Preferred L¹ linkers of the present invention include sequences selected from the group of formulas: (SEQ ID NO : 1 ) through SEQ ID NO : ₂68 )

Other linkers are also contemplated by the invention. The present invention is, however, not limited by the form, size or number of linker sequences employed. The only requirement of the linker is that it does not functionally interfere with the folding and function of the individual molecules of the multi- functional protein.

Utility of permuteins

Permuteins of the present invention may exhibit useful properties such as having similar or greater biological activity when compared to a single factor or by having improved half-life or decreased adverse side effects, or a combination of these properties.

Permuteins which have little or no activity maybe useful as antigens for the production of antibodies for use in immunology or immunotherapy, as probes or as intermediates used to construct other useful permuteins.

The permuteins of the present invention may have an improved therapeutic profile as compared to their parent molecules. For example, some permuteins of the present invention may have a similar or more potent activities relative to other compounds or proteins without having a similar or corresponding increase in side-effects. This is particularly true of multifunctional or fusion protein therapeutics, where permutation may relieve steric and other hindrances that impair the activity of the parent fusion molecules (see Kreitman, R. J. et al., Proc Natl Acad Sci USA 91: 6889-6893, 1994; Kreitman, R. J. et al., Cancer Res. 55:3357-3363, 1995, for examples).

A general utility of permuteins is in the area of nanoscale devices described alternatively as "nanobiological" or "nanobiotechnological." These are nanoscale devices containing both precise structure nanomaterials and biological functional components (such as proteins). Nanodevices have been the subject of several reviews (Lee, S.C., Trends in Biotechnology, 16: 239-240, 1998).

Nanobiological/nanobiotechnological devices generally contain proteins covalently coupled to polymers or other non-biological precise structure materials. Issues of steric and other interferences with protein activity are applicable to proteins in nanobiological/nanobiotechnological devices and are highly analogous to the issues with multifunctional/fusion proteins discussed above. Protein permutation is fully expected to offer a viable approach to deal with these considerations, just as it does in the case of fusion proteins (Kreitman, R. J. et al., Proc Natl Acad Sci USA 91: 6889-6893, 1994; Kreitman, R. J. et al.,., Cancer Res. 55:3357-3363, 1995).

Examples

The following examples will illustrate the invention in greater detail, although it will be understood that the invention is not limited to these specific examples. Various other examples will be apparent to the person skilled in the art after reading the present disclosure without departing from the spirit and scope of the invention. It is intended that all such other examples be included within the scope of the appended claims.

General Materials and Methods

General methods of cloning, expressing, and characterizing proteins are found in T. Maniatis et al., Molecular Cloning, A Laboratory Manual, Cold Spring

Harbor Laboratory, 1982, and references cited therein, incorporated herein by reference; and in J. Sambrook et al., Molecular Cloning, A Laboratory Manual, 2^nd edition, Cold Spring Harbor Laboratory, 1989, and references cited therein, incorporated herein by reference.

Unless noted otherwise, all specialty chemicals were obtained from Sigma

Chemical Company (St. Louis, MO). Restriction endonucleases and T4 DNA ligase were obtained from New England Biolabs (Beverly, MA) or Boehringer Mannheim (Indianapolis, IN).

Strains, plasmids, and bacteriophage

The bacterial strains used in these studies are listed in Table 1. Plasmids and bacteriophage used or constructed in this study are listed in Tables 2 and 3, respectively.

Phage and phagemid stocks were made and manipulated as described (Kay, B.K., Winter, J., and McCafferty, J., Phage Display of Peptides and Proteins, Academic Press, San Diego, California, 1996; Merlin, S. et al., Applied

Biochemistry and Biotechnology 67: 15-29, 1997).

Transformation of E. coli strains

E. coli strains (Table 1), such as DH5α™ (Life Technologies, Gaithersburg, MD) and TGI (Amersham Corp., Arlington Heights, IL) are used for transformation of ligation reactions and are the hosts used to prepare plasmid

DNA for transfecting mammalian cells. E. coli strains, such as JM101 (Yanisch- Perron et al., Gene, 33: 103-119, 1985) and MON105 (Obukowicz et al., Appl and Enυir. Micr., 58: 1511-1523, 1992) can be used for expressing the multi-functional proteins of the present invention in the cytoplasm or periplasmic space.

DH5α™ Subcloning efficiency cells are purchased as competent cells and are ready for transformation using the manufacturer^'s protocol, while both E. coli strains TGI and MON105 are rendered competent to take up DNA usinp a CaCl₂ method. Typically, 20 to 50 mL of cells are grown in LB medium (1% Bacto- tryptone, 0.5% Bacto-yeast extract, 150 mM NaCl) to a density of approximately 1.0 optical density unit at 600 nanometers (OD600) as measured by a Baush & Lomb Spectronic spectrophotometer (Rochester, NY). The cells are collected by centrifugation and resuspended in one-fifth culture volume of CaCl₂ solution (50 mM CaCl₂, 10 mM Tris-Cl, pH7.4) and are held at 4°C for 30 minutes. The cells are again collected by centrifugation and resuspended in one-tenth culture volume of - CaCl₂ solution. Ligated DNA is added to 0.2 mL of these cells, and the samples are held at 4°C for 30-60 minutes. The samples are shifted to 42°C for two minutes and 1.0 mL of LB is added prior to shaking the samples at 37°C for one hour. Cells from these samples are spread on plates (LB medium plus 1.5% Bacto-agar) containing either ampicillin (100 micrograms/mL, ug/mL) when selecting for ampicillin-resistant transformants, or spectinomycin (75 ug/mL) when selecting for spectinomycin-resistant transformants. The plates are incubated overnight at 37°C.

Colonies are picked and inoculated into LB plus appropriate antibiotic (100 ug/mL ampicillin or 75 ug/mL spectinomycin) and are grown at 37°C while shaking.

DNA isolation and characterization

DNA constructs were made and propagated in E. coli using standard molecular biology techniques (Sambrook, J. et al., Molecular Cloning, A Laboratory Manual, 2^nd edition, Cold Spring Harbor Laboratory, 1989).

Plasmid DNA can be isolated by a number of different methods and using commercially available kits known to those skilled in the art. Plasmid DNA is isolated using the Promega Wizard™ Miniprep kit (Madison, Wl), the Qiagen

QIAwelt Plasmid isolation kits (Chatsworth, CA) or Qiagen Plasmid Midi or Mini kit. These kits follow the same general procedure for plasmid DNA isolation. Briefly, cells are pelleted by centrifugation (5000 x g), the plasmid DNA released with, sequential NaOH/acid treatment, and cellular debris is removed by centrifugation ( 10000 x g). The supernatant (containing the plasmid DNA) is loaded onto a column containing a DNA-binding resin, the column is washed, and plasmid DNA eluted. After screening for the colonies with the plasmid of interest, the E. coli cells are inoculated into 50-100 ml of LB plus appropriate antibiotic for overnight growth at 37°C in an air incubator while shaking. The purified plasmid DNA is used for DNA sequencing, further restriction enzyme digestion, additional subcloning of DNA fragments and transfection into E. coli, mammalian cells, or other cell types.

Sequence confirmation

DNA sequence analysis was performed using the Genesis 2000 DNA analysis system using standard methods (Prober et al., Science 238: 336-341,

1987).

Purified plasmid DNA is resuspended in dH₂0 and its concentration is determined by measuring the absorbance at 260/280 nm in a Bausch and Lomb Spectronic 601 UV spectrometer. DNA samples are sequenced using ABI PRISM™ DyeDeoxy™ terminator sequencing chemistry (Applied Biosystems Division of

Perkin Elmer Corporation, Lincoln City, CA) kits (Part Number 401388 or 402078) according to the manufacturer's suggested protocol usually modified by the addition of 5% DMSO to the sequencing mixture. Sequencing reactions are performed in a DNA thermal cycler (Perkin Elmer Corporation, Norwalk, CT) following the recommended amplification conditions. Samples are purified to remove excess dye terminators with Centri-Sep™ spin columns (Princeton Separations, Adelphia, NJ) and lyophilized. Fluorescent dye labeled sequencing reactions are resuspended in deionized formamide, and sequenced on denaturing 4.75% polyacrylamide-8M urea gels using ABI Model 373A and Model 377 automated DNA sequencers. Overlapping DNA sequence fragments are analyzed and assembled into master DNA contigs using Sequencher DNA analysis software (Gene Codes Corporation, Ann Arbor, MI).

Expression of permuted proteins in mammalian cells

To obtain sufficient protein for analysis of the bioactivity of individual MPO moiecuJes containing cphG-CSFs, DNA segments containing individual affinity-selected MPO: cphGCSFs were subcloned into a mammalian expression vector, and expressed transiently in BHK cells as described below.

The BHK-21 cell line can be obtained from the ATCC (Rockville, MD). The cells are cultured in Dulbecco's modified Eagle media (DMEM high-glucose), supplemented to 2 mM (mM) L-riutamine and 10% fetal bovine serum (FBS). This formulation is designated BHK growth media. Selective media is BHK growth media supplemented with 453 units/mL hygromycin B (CalBiochem, San Diego, CA). The BHK-21 cell line was previously stably transfected with the HSV transactivating protein VP16, which transactiva es the IE110 promoter found on the plasmid pMON3359 and pMON3633 and the IE 175 promoter found in the plasmid pMON3360B (Hippenmeyer, P.J. and Pegg, L.E., Curr. Opin. Biotechnol. 6: 548-552, 1995). The VP16 protein drives expression of genes inserted behind the IE110 or IE175 promoter. BHK-21 cells expressing the transactivating protein VP16 are designated BHK-VP16. The plasmid pMON1118 expresses the hygromycin resistance gene from the SV40 promoter (Highkin et al., Poultry Sci., 70: 970-981, 1991). A similar plasmid, pSV2-hph, is available from ATCC.

BHK-VP16 cells are seeded into a 60 millimeter (mm) tissue culture dish at 3 x 10⁵ cells per dish 24 hours prior to transfection. Cells are transfected for 16 hours in 3 mL of "OPTIMEM"™ (Gibco-BRL, Gaithersburg, MD) containing 10 ug of plasmid DNA containing the gene of interest, 3 ug hygromycin resistance plasmid, pMONlllδ, and 80 ug of Gibco-BRL "LIPOFECTAMINE"™ per dish. The media is subsequently aspirated and replaced with 3 mL of growth media. At 48 hours post-transfection, media from each dish is collected and assayed for activity (transient conditioned media). The cells are removed from the dish by trypsin-EDTA, diluted 1:10, and transferred to 100 mm tissue culture dishes containing 10 mL of selective media. After approximately 7 days in selective media, resistant cells grow into colonies several millimeters in diameter. The colonies are removed from the dish with filter paper (cut to approximately the same size as the colonies and soaked in trypsin/EDTA) and transferred to individual wells of a 24 well plate containing 1 mL of selective media. After the clones are grown to confluence, the conditioned media is re-assayed, and positive clones are expanded into growth media.

Affinity selection and screening of phagemids

Affinity reagent used for the identification of functional MPO molecules containing cphG-CSF (MPO: cphG-CSF) species from the library were BHK cells expressing the hG-CSF receptor on their surface. The library pool was subjected to iterative affinity selection (four rounds) against BHK cells expressing the h-GCSF receptor using previously described techniques (Merlin, S. et al., Applied Biochemistry and Biotechnology 67: 15-29, 1997). Between rounds of selection, phage eluted from the affinity reagent were amplified in E. coli (Kay, B.K. J. Winter, and J. McCofferty, Phage Display of Peptides and Proteins, Academic Press, San Diego, California. 1996). Expression of proteins in E. coli

When large-scale quantities of recombinant protein are desirable for structure-function studies, DNA segments containing individual affinity-selected MPO phGCSFs are subcloned into any of a variety of bacterial plasmid expression vectors, and expressed as a cytoplasmic product or as a secreted protein in E. coli.

E. coli strain MON105 or JM101 harboring the plasmid of interest are grown at 37°C in M9 plus casamino acids medium with shaking in an air incubator Model G25 from New Brunswick Scientific (Edison, NJ). Growth is monitored at OD_grø until it reaches a value of 1.0 at which time nalidixic acid (10 mg/mL) in 0.1 N NaOH is added to a final concentration of 50 μg/mL, for cultures containing plasmids with the E. coli recA promoter driving expression of the recombinant gene. IPTG is used in place of nalidixic acid, as a chemical inducer to facilitate expression from plasmids containing the lac promoter or hybrid lac promoters. The cultures are then shaken at 37°C for three to four additional hours. A high degree of aeration is maintained throughout the culture period in order to achieve maximal production of the desired gene product. The cells are examined under a light microscope for the presence of inclusion bodies (IB). One mL aliquots of the culture are removed for analysis of protein content by boiling the pelleted cells, treating them with reducing buffer and electrophoresis via SDS-PAGE (see Maniatis et al., "Molecular Cloning: A Laboratory Manual", 1982). The culture is centrifuged (5000 x g) to pellet the cells.

Isolation of Inclusion Bodies

The cell pellet from a 330 mL E. coli culture is resuspended in 15 mL of sonication buffer (10 mM 2-amino-2-(hydroxymethyl) 1,3-propanediol hydrochloride (Tris-HCl), pH 8.0 + 1 mM ethylenediaminetetraacetic acid (EDTA).

These resuspended cells are sonicated using the microtip probe of a Sonicator Cell Disruptor (Model W-375, Heat Systems-Ultrasonics, Inc., Farmingdale, New York). Three rounds of sonication in sonication buffer followed by centrifugation are employed to disrupt the cells and wash the inclusion bodies (IB). The first round of sonication is a 3 minute burst followed by a 1 minute burst, and the final two rounds of sonication are for 1 minute each.

Purification

The folded proteins can be affinity -purified using affinity reagents such as monoclonal antibodies or receptor subunits attached to a suitable matrix. Purification can also be accomplished using any of a variety of chromatographic methods such as: ion exchange, gel filtration or hydrophobic chromatography or reversed phase HPLC. These and other protein purification methods are described in detail (Methods in Enzymology, Volume 182 "Guide to Protein Purification" edited by Murray Deutscher, Academic Press, San Diego, California, 1990).

Protein Characterization _

The purified protein is analyzed by RP-HPLC, electrospray mass spectrometry, and SDS-PAGE. The protein quantitation is done by amino acid composition, RP-HPLC, and Bradford protein determination. In some cases tryptic peptide mapping is performed in conjunction with electrospray mass spectrometry to confirm the identity of the protein.

Baf-3/G-CSF receptor assay

Briefly, the mouse lymphoid cell line Baf3 was transfected with human granulocyte colony stimulating factor receptor (hG-CSFR) cDNA. Stable clones of Baf3 which expressed the G-CSFR and proliferated in the presence of hG-CSF were isolated and used to investigate the activity of human G-CSF receptor agonists without the influence of other human cytokine receptor responses.

The cDNA encoding hG-CSFR (a gift from Dr. Daniel C. Link (Washington University, St. Louis, MO) was released from the plasmid pEMCV.Sralpha as a .HmdIII/EcoRI (5' to 3') fragment, gel-purified, and inserted into the mammalian cell expression plasmid pcDNA3 (Invitrogen, San Diego, CA). This plasmid contains enhancer-promoter sequences from the immediate early gene of the human cytomegalovirus (CMV), a bovine growth hormone polyadenylation signal and transcription termination sequences, a neomycin resistance gene is present for the selection of G418 stable cell clones, and an ampicillin resistance gene for selection in E. coli. Ligation mixtures were transformed into E. coli strain TGI [delta (lac-pro), supE, thi, hsdΩ5fF'(traO36, proAΕ^*, lacF, ZαcZdeltaMlδ] and plasmid DNA was purified using a Qiagen Midiprep Plasmid Kit. The structure of plasmid DNAs containing hG-CSFR were confirmed by restriction enzyme analysis and by automated DNA sequence analysis using an ABI sequencing machine. One of several plasmids with the correct structure was selected and given the designation pMON30298.

Baβ ceils, maintained in complete growth medium (RPMI 1640 supplemented to 10% FBS and 10% Wehi 3B supernatant as a source for mouse IL- 3), were seeded at a subconfluent cell density of 10^Λ5 cells/ml in growth media (RPMI 1640 5% FBS; 2 mM L-glutamine) the day prior to the electroporation. The cells were collected and rinsed twice in 10 ml serum-free RPMI 1640. The cells were diluted to 10^Λ6/ml in serum-free RPMI and 1 ml was placed into each electroporation chamber (Gibco/BRL #1608AJ). 50 ug of plasmid DNA was added to each chamber and the chamber were incubated on ice for 30 minutes prior to electroporation. The cells were electroporated on ice at a capacitance of 800 uF, - 400V, fast charge, and low ohms in a BRL CellPorator. The cells were immediately removed from the chambers and placed into 10 cm dishes containing 10 ml of growth medium. The cells were allowed to recover for 48 hr in growth media prior to selection.

After the recovery period, the cells were pelleted at 1000 rpm for 10 minutes, and resuspended into 10 ml of selection medium (growth medium containing 800 ug/ml G418 sulfate (Gibco/BRL). The cells were kept in selection media, being passaged twice weekly, until only a few viable cells could be seen in the mock transfected control cell dishes (approximately 2 weeks). After an additional 2 weeks in selection media, the cells which had been electroporated with the hG-CSFR cDNA had grown to a cell density which allowed them to be tested for proliferation in the presence of hG-CSF (Fukunaga, R. et al., EMBO J. 10 (10): 2855-2865, 1991).

The cell proliferation assay conditions are as follows: Briefly, 25,000 cells were plated in a microtiter 96 well plate with or without cytokine in IMDM medium supplemented with BSA (50 ug/ml), human transferrin (100 ug/ml), lipid (50 ug/ml) 2-mercaptoethanol (50 uM final concentration). Each well was incubated with 0.5 uCi of ³H-thymidine (16 hours) and the incorporated radioactivity was measured. Triplicate wells containing Baf3 cells were set up with 4 nM hG-CSF, 4 nM mIL-3 or media only control. Samples of different permuted proteins were tested in each assay.

Example 1: Construction of a permutein library without a linker region

Figure 1 shows a schematic of scanning permutagenesis. A plasmid construct comprising a tandem repeat of the modified human granulocyte colony stimulating factor (hG-CSF with a serine for amino acid 17) gene joined by a sequence (GCCGG, termed a zero order linker) was generated and subcloned into the plasmid pACYC177 (Chang, A.C.Y. and S.N. Cohen, J Bacteriol. 134: 1141- 1156, 1978) using standard molecular biology methods (Sambrook, J. et al., Molecular Cloning, A Laboratory Manual, 2^nd edition, Cold Spring Harbor Press, New York, 1989). The resultant plasmid construct (pMON15978) was linearized by restriction digestion (Smal) and used as a template for PCR amplification of circularly permuted hG-CSF (cphG-CSF) genes, following the method of Horlich (Horlick, R.A. et al., Protein Engineering 5: 427-431, 1992. For purposes of this demonstration of the scanning permutagenesis technique, we chose to make a limited permutein library rather than one containing every possible cphG-CSF. Figure 2 shows the position of the new amino termini for each new cphG-CSF.

Individual cphG-CSF genes were inserted into phagemid presentation vector pCANTAB 5E (Pharmacia Biotech,) such that they were expressed as a part of a MPO species (Feng, Y., N. R. Staten, C. M. Baum, N. L. Summers, M. Caparon, S. C. Bauer, L. Zurfluh, J. P. McKearn, B. K. Klein, S. C. Lee, C. A. McWherter. 1997. Multi-functional hematopoeitic receptor agonists. World Patent Application WO 97/12985) which was in turn fused to the amino end of the phage genelll product. The presented fusion protein contained, starting from its amino terminus, a hIL-3 receptor agonist, cphG-CSF, and the phage gene III product. The juncture between the presented protein and the gene III product was as previously described (Merlin, S. et al., Applied Biochemistry and Biotechnology 67: 15-29, 1997).

After confirmation of the structure of each phagemid construct, phagemid particles were produced for each individual cphG-CSF-presenting species (Merlin et al., 1997). Some of these lots of particles were used to individually define the affinity properties of specific presented cphG-CSF species in analytical biopanning experiments (Caparon, M. H. et al., Molecular Diversity 1: 241-246, 1996; Merlin, S. et al., Applied Biochemistry and Biotechnology 67: 15-29, 1997), but all of the phage particle lots were titered and equivalent numbers of transducing units of each particle preparation were pooled together to form the scanning permutagenesis library for hG-CSF in an MPO background. Figure 2 shows the MPO: cp hG-CSF species present in the library.

MPO : cphG-CSF 38/37 is an example of the nomenclature used to specify the identity of individual permuted proteins. It describes a MPO molecule containing a circularly permuted human G-CSF module (with the serine 17 substitution). The first amino acid of the cphG-CSF domain is amino acid 38 of the parent protein, and the last amino acid is residue 37 of the parent.

Example 2: Presentation and Affinity screening of the MPO: cphGCSF library

MPO: cphG-CSF 38/37, is a full hG-CSF receptor agonist (McKearn, J.P., Myelorestorative activities of synthokine and myelopoietin. In Proceedings of the 1996 IBC Conference on Therapeutic Applications of Cytokines, pp. 4.3.1-4.3.18, 1996). It was presented on filamentous phage as a positive control to demonstrate that permuted proteins can be presented on the surface of phage particles and - affinity selected. After phagemid particles were produced from this construct, they were subjected to analytical biopanning using cells expressing the hG-CSF receptor as affinity reagent.

Table 1 shows that phage presented MPO: cphG-CSF 38/37 was affinity selected by cells expressing the hG-CSF. MPO: cphGCSF 38/37-GPIII fusion was expressed, secreted and assembled into phagemid particles, and could be affinity selected by the hG-CSF receptor. Permutagenesis of a protein does not appear to impair its successful presentation.

Relative to typical phage display libraries, the complexities of cp libraries are low, containing perhaps hundreds to thousands of individuals. The demonstration library here contained about 50 distinct clones, as opposed to more typical phage libraries containing more than 10⁵ individuals (reviewed in Clackson, T. and J.A. Wells, Tibtech 12: 173-184, 1994).

37 randomly chosen selectants from round 1, and fewer from subsequent rounds (17, 11 and 14 were picked from rounds 2, 3 and 4, respectively) were chosen for sequence analysis. The identity of the MPO: cphG-CSFs identified in each round is shown in Figure 2.

A total of 14 MPO: cphGCSF species were identified from the output of affinity selection (Figure 2). Most of the MPO: cphGCSF species identified from the library had new carboxy and amino termini in loop segments (9 of 14 permuteins identified), rather than in clearly defined secondary structures (See Hill et al., 1993 for the hG-CSF structure). Five selectants had termini within helical domains of hG-CSF (MPO: cphG-CSFs 13/12, 19/18, 71/70, 123/122 and 159/158). For three of these molecules (MPO: cphG-CSFs 13/12, 71/70 and

123/122) their new ends lie at the outermost ends of helices, and therefore perturbation of secondary structure caused by these permuteins may be minimal. However, MPO: cphG-CSF 19/18 and MPO: cphG-CSF 159/158 have new termini well within helix 1 and helix 4 of hG-CSF, respectively. These data parallel the observations of Graf and Schachman, who developed a limited DNase I digestion method for "random" permutagenesis (Graf, R. and H. K. Schachman, Proc Natl Acad Sci USA 93:11591-11596, 1996). They identified two permutein species of aspartate transcarbamoylase that introduced new amino and carboxy ends into secondary structural domains and that retained biological activity. In their work, the majority of permuteins introducing ends into secondary structures (5/7 identified) were significantly diminished in activity. In ~ contrast, we found a several permuteins that introduced helical breaks retained activity (See Below). The method used by Graf and Schechman frequently introduces point mutations, small insertions and deletions into the permuted proteins, potentially complicating the analysis of the effects of permutagenesis.

Example 3: Biological activity of MPO: cphG-CSFs selected from the cp phage library

To obtain sufficient protein for analysis of the bioactivity of individual MPO molecules containing cphG-CSFs, DNA segments containing individual affinity-selected MPO: cphGCSFs were subcloned into a mammalian expression vector, and expressed transiently in BHK cells as described above.

The MPO: cphG-CSFs isolated from biopanning were all expressed transiently in mammalian cells and the amount of MPO: cphG-CSF in each supernatant was determined by sandwich hIl-3 ELISA (Olins P.O. et al., J. Biol.

Chem. 270: 23754-23760, 1995). The quantitated supernatants were then assayed for G-CSF receptor agonist activity in a Baf-3/G-CSF receptor assay (Figure 3, Table 4).

All but one of the transiently expressed MPO: cphG-CSF proteins exhibited G-CSF activity equivalent to or slightly better than that of the parent MPO molecule, including those MPO: cphG-CSFs with new carboxy and amino ends within helixes. The permutein encoded by pMON 16021 with a breakpoint between positions 48 and 49 did not exhibit activity in the G-CSF-dependent proliferation assay. These data suggest that most of the proteins isolated from the library are competent to bind the hG-CSF receptor and produce a proliferation signal.

All references, patents, or applications cited herein are incorporated by reference in their entirety, as if written herein. References

Buchwalder, A. et al., Biochemistry 31:1621-1630, 1992.

Caparon, M. H. et al., Molecular Diversity 1: 241-246, 1996.

Chang, A.C.Y. and S.N. Cohen, J Bacterial. 134: 1141-1156, 1978.

Clackson, T. and J.A. Wells, Tibtech 12: 173-184, 1994.

Feng et al., J. Mol. Biol. 259: 524-551, 1996.

Feng, Y., N. R. Staten, C. M. Baum, N. L. Summers, M. Caparon, S. C. Bauer, L. Zurfluh, J. P. McKearn, B. K. Klein, S. C. Lee, C. A. McWherter. 1997. Multifunctional hematopoeitic receptor agonists. World Patent Application WO 97/12985.

Fukunaga, R., Ishizaka-Ikeda, E., Pan, C. X., Seto, Y., and Nagata, S. Functional Domains of the Granulocyte Colony-Stimulating Factor Receptor. EMBO J. 10 (10):2855-2865, 1991.

Goldenberg, D. P. and T. E. Creighton, J. Mol. Biol. 165: 407-413, 1983.

Graf, R. and H. K. Schachman, Proc Natl Acad Sci USA 93:11591-11596, 1996.

Hahn, M. et al., Proc Natl Acad Sci USA 91: 10417-10421, 1994.

Highkin et al., Poultry Sci., 70: 970-981, 1991.

Hill et al., Proc. Natl. Acad. Sci. USA 90: 5167-5171, 1993.

Hippenmeyer, P.J. and L.E. Pegg, Curr. Opin. Biotechnol. 6: 548-552, 1995.

Horlick, R.A. et al., Protein Engineering 5: 427-431, 1992.

Jelinski, L.W. Biologically related aspects of nanoparticles, nanostructured materials and nanodevices. In "WTEC workshop on Global Assessment of R &D Status and Trends in Nanoparticles, Nanostructured Materials and Nanodevices", International Technology Research Institute, Loyola College, Baltimore, MD., S. C, 1998.

Johnson. J. and F. M. Raushel, Biochemistry 35: 10223-10233, 1996. Kay, B.K. J. Winter, and J. McCofferty, Phage Display of Peptides and Proteins, Academic Press, San Diego, California, 1996.

Koebnik, R. and L. Kramer, J. Mol. Biol. 250: 617-626, 1995.

Kreitman, R. J. et al. Cytokine 7(4): 311-318, 1995.

Kreitman, R. J. et al., Proc Natl Acad Sci USA 91: 6889-6893, 1994.

Kreitman, R. J. et al., Cancer Res. 55:3357-3363, 1995.

Lee, S.C. Biotechnology for Nanotechnology. Trends in Biotechnology, 16: 239-240, 1998.

Lee, S.C, Phage presentation for cytokine engineering. IBCs Second International Conference on Display Technologies, 1997.

Lin, X. et al., Protein Science 4: 159-166, 1995.

Lowman and Wells, J. Mol. Biol. 234: 564-578, 1993.

Luger et al., Protein Engineering 3: 249-258, 1990.

Luger et al., Science 243: 206-210, 1989.

Maniatis, T. et al., Molecular Cloning, A Laboratory Manual, Cold Spring Harbor

Laboratory, 1982.

McKearn, J.P. Myelorestorative activities of synthokine and myelopoietin. In Proceedings of the 1996 IBC Conference on Therapeutic Applications of Cytokines, pp. 4.3.1-4.3.18, 1996.

Merlin, S. et al., Applied Biochemistry and Biotechnology 67: 15-29, 1997.

Mullins, L. S. et al., J Am. Chem Soc. 116: 5529-5533, 1994.

Murray Deutscher (ed), Methods in Enzymology, Volume 182 "Guide to Protein Purification," Academic Press, San Diego, California, 1990.

Obukowicz, M. et &\., Appl. and Envir. Micr., 58: 1511-1523, 1992.

Olins P.O. et al., J. Biol. Chem. 270: 23754-23760, 1995.

Prober. J.M. et al., Science 238: 336-341, 1987. Protosova, N. Y. et al., Protein Engineering 7: 1373-1377, 1994.

Puri, R. K. et al., Cellular Immunol. 171: 80-86, 1996.

Rose, K. et al., Molecular Immunology 32: 1031-1037, 1995.

Sambrook, J. et al., Molecular Cloning, A Laboratory Manual, 2^nd edition, Cold Spring Harbor Press, New York, 1989.

Shortle, D. and J. Sondek, Curr. Opin. Biotechnol 6: 299-305.

Smith, G. P. , Curr. Opin. in Biotechnol. 2: 668-673, 1991.

Smith, G. P., Science 228: 1315-1317, 1985.

Sondek, J. and D. Shortle, Proteins 7: 387-393, 1990.

Thomas, J. W. et al., Proc Natl Acad Sci USA 92: 3779-3783, 1995.

Winter, G., Drug Development Res. 33: 71-89, 1994.

Yang, Y. R. and H. K. Schachman, Proc Natl Acad Sci USA 90: 11980-11984, 1993.

Yanisch-Perron et al., Gene, 33: 103-119, 1985.

Zhang, T. et al., Biochemistry 32: 12311-12318, 1993.

Tables

Table 1: Circularly permuted proteins

Protein Reference

Enzymes

T4 lysozyme Zhang et al., Biochemistry 32:12311-12318 (1993); Zhang et al., Nature Struct. Biol. 1:434-438 (1995) dihydrofolate reductase Buchwalder et al., Biochemistry 31:1621-1630

(1994); Protasova et al., Prot. Eng. 7:1373-1377 (1995) ribonuclease Tl Mullins et al., J. Am. Chem. Soc. 116:5529-5533

(1994); Garrett et al, Protein Science 5:204-211 (1996)

Bacillus β-glucanase Hahn et al., Proc. Natl. Acad. Sci. U.S.A. 91:10417-

10421 (1994) aspartate transcarbamoylase Yang and Schachman, Proc. Natl Acad.

Transcarbamoylase Sci. U.S.A. 90:11980-11984

(1993) phosphoribosyl anthranilate Luger et al., Science 243:206-210 (1989); isomerase Luger et al., Prot. Eng. 3:249-258 (1990) pepsin/pepsinogen Lin et al., Protein Science 4:159-166 (1995) glyceraldehyde-3-phosphate Vignais et al., Protein Science 4:994-1000 (1995) dehydrogenase ornithine decarboxylase Li & Coffino, Mol. Cell Biol. 13:2377-2383 (1993) yeast phosphoglycerate Ritco-Vonsovici et al., Biochemistry 34:16543- dehydrogenase 16551 (1995)

Enzyme Inhibitor basic pancreatic trypsin inhibitor Goldenberg & Creighton, J. Mol. Biol. 165:407-413

(1983)

Cytokines interleuk -l|^*ι Horlick et al., Protein Eng. 5:427-431 (1992) inter eukin-4 Kreitman et al., Cytokine 7:311-318 (1995)

Tyrosine Kinase Recognition Domain α-spectrin SH3 domain Viguera et al., J. Mol. Biol 247:670-681 (1995)

Transmembrane Protein omp A Koebnik & Kramer, J. Mol. Biol. 250:617-626

(1995) Chimeric Protein interleukin-4-Pseurfomonαs Kreitman et al., Proc. Natl. Acad. Sci. U.S.A. _, exotoxin fusion molecule 91 :6889-6893 (1994)

Table 2: Strains

Designation Description or Genotype Reference/Source

DH5α™ F , p/u80 dZαcZdeltaM15, Life Technologies, Rockville, delta(ZαcZYA-αrgF)U169, c eoR, Maryland recAl, endAl, hsdRll (rk ,mk^*), phoA, supE44, lambda-, thi-1, gyrA96, relAl

JM101 delta (pro lac), supE, thi, Yanisch-Perron et al., Gene, 33:

(ATCC# F'(trαD36, proAΕ^*Jαcr, 103-119, 1985

33876) ZαcZdeltaMlδ)

MON105 F , lambda-,IN (rrnD, rrnE)l, Obukowicz et al. , Appl. and Envir.

(ATCC# rpoD^*, rpoH358 Micr., 58: 1511-1523, 1992

55204)

MON208 W3110 rpoH358, Zαcl°, ompT::kan Alan Easton

TGI delta(lac-pro), supE, thi-1, Amersham Corp., Arlington hsdO5/F'(traO36, proA^*B Zαcl^q, Heights, Illinois ZαcZdeltaMlδ)

W3110 IN (rrnD-r E)l, rphl Lab collection

Table 3: Plasmids

Plasmid SEQ Selectable Description Source

ID Marker

NO.

pACYC177 Kan^R Plasmid with multiple cloning Chang, sites and two selectable A.C.Y. and

Amp markers S.N. Cohen, J Bacteriol. 134: 1141- llδ6, 1978

pMON 15978 Amp R Plasmid construct comprising a This work tandem repeat of the modified human granulocyte colony stimulating factor (hG-CSF with a serine for amino acid 17) gene joined by a sequence (GCCGG, termed a zero order linker), subcloned into the plasmid pACYC177

pCANTAB 5E _jjjpR Phage display vector containing Pharmacia lac promoter operably linked to Biotech, fd gene 3 signal sequence, a Piscataway, linker region, an E tag, and an NJ fd gene 3 structural gene all cloned into the vector backbone of pUC119 containing ColEl ori, the beta lactamase resistance gene, and an M13 ori.

PMON16016 A_jj R Phagemid presentation vector This work pCANTAB δE derivation containing inserted individual cphG-CSF gene such that it was expressed as a part of an MPO species, fused in turn to the amino terminus end of the phage genelll product. The first amino acid of the cphGCSF domain is amino acid 1 of the parent, and the last amino acid is residue 174 of the parent. The zero order linker is attached at the carboxyl end of amino acid 174.

pMON16017 π_jpR Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 3 of the parent, and the last amino acid is residue 2 of the parent.

pMON 16029 Amp^R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 7 of the parent, and the last amino acid is residue 6 of the parent.

pMON16030 _jnpR Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 9 of the parent, and the last amino acid is residue 8 of the parent.

pMON16018 π R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 11 of the parent, and the last amino acid is residue 10 of the parent. pMON16019 Amp R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 13 of the parent, and the last amino acid is residue 12 of the parent.

pMON16031 A_jnpR Identical to pMONl6016 except This work the first amino acid of the cphG-CSF domain is amino acid lδ of the parent, and the last amino acid is residue 14 of the parent.

pMONl6020 A_jnpR Identical to pMONl6016 except This work the first amino acid of the cphG-CSF domain is amino acid 19 of the parent, and the last amino acid is residue 18 of the parent.

pMON16032 Amp^R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 22 of the parent, and the last amino acid is residue 21 of the parent.

pMON 16033 Amp^R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 27 of the parent, and the last amino acid is residue 26 of the parent.

pMON 16034 Amp^R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino last amino acid is residue 30 of the parent.

pMON1603δ Amp^R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 3δ of the parent, and the last amino acid is residue 34 of the parent.

pMON 16036 AmpR Identical to pMONl6016 except This work the first amino acid of the cphG-CSF domain is amino acid 37 of the parent, and the last amino acid is residue 36 of the parent.

pMON16037 Amp^R Identical to pMONl6016 except This work the first amino acid of the cphG-CSF domain is amino acid 38 of the parent, and the last amino acid is residue 37 of the parent.

pMON 16038 Amp^R Identical to pMONl6016 except This work the first amino acid of the cphG-CSF domain is amino acid 39 of the parent, and the last amino acid is residue 38 of the parent.

pMON 16039 _jnpR Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 43 of the parent, and the last amino acid is residue 42 of the parent.

pMON16040 AmpR Identical to pMONl6016 except This work the first amino acid of the cphG-CSF domain is amino acid 4δ of the parent, and the last amino acid is residue 44 of the parent.

pMON16041 Amp R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 47 of the parent, and the last amino acid is residue 46 of the parent.

pMON16022 Amp R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 49 of the parent, and the last amino acid is residue 48 of the parent.

pMON16042 Amp R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid δl of the parent, and the last amino acid is residue δO of the parent.

pMON 16043 Amp R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid δ3 of the parent, and the last amino acid is residue δ2 of the parent.

pMON 16044 Amp R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 56 of the parent, and the last amino acid is residue 5δ of the parent. pMON 16023 AmpR Identical to pMONl6016 except This work the first amino acid of the cphG-CSF domain is amino acid 60 of the parent, and the last amino acid is residue δ9 of the parent.

pMON1604δ Amp^R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 64 of the parent, and the last amino acid is residue 63 of the parent.

pMON16024 Amp^R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 67 of the parent, and the last amino acid is residue 66 of the parent.

pMON16046 A_jnpR Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 69 of the parent, and the last amino acid is residue 68 of the parent.

pMON1602δ A_jnpR Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 71 of the parent, and the last amino acid is residue 70 of the parent.

pMON16047 AmpR Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 73 of the parent, and the last amino acid is residue 72 of the parent.

pMON16048 Amp Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 84 of the parent, and the last amino acid is residue 83 of the parent.

pMON16049 Amp^R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 98 of the parent, and the last amino acid is residue 97 of the parent.

pMON160δO Amp^R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 100 of the parent, and the last amino acid is residue 99 of the parent.

pMON160δl Amp^R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 102 of the parent, and the last amino acid is residue 101 of the parent.

pMON160δ2 AmpR Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 112 of the parent, and the last amino acid is residue 111 of the parent.

pMON160δ3 mp Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 121 of the parent, and the last amino acid is residue 120 of the parent.

pMON16026 Amp^R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 123 of the parent, and the last amino acid is residue 122 of the parent.

_PMON16027 Amp^R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 12δ of the parent, and the last amino acid is residue 124 of the parent.

pMON160δ4 A p Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 133 of the parent, and the last amino acid is residue 132 of the parent.

pMON16055 AmpR Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 142 of the parent, and the last amino acid is residue 141 of the parent.

pMON160δ6 Amp^R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 143 of the parent, and the last amino acid is residue 142 of the parent. pMON160δ7 Amp R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 147 of the parent, and the last amino acid is residue 146 of the parent.

pMON 16028 Amp^R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 159 of the parent, and the last amino acid is residue 158 of the parent.

pMON16058 Amp Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 168 of the parent, and the last amino acid is residue 167 of the parent.

pMON 16059 Amp^R Identical to pMON16016 except This work the first amino acid of the cphG-CSF domain is amino acid 170 of the parent, and the last amino acid is residue 169 of the parent.

Table 4: Analytical biopanning

Before receptor* After receptor* Enrichment

l/6.6xl0⁴ 1/6.5x10 ' 990-fold

* Amp^R/Kan^R resistant colonies

Analytical biopanning shows that MPO molecules containing permuted hG- CSF domains can be presented and affinity selected in a hG-CSF receptor dependent fashion. A mixture of phagemids presenting MPO: cphG-CSF 38/37 (ampicillin resistant) and M13k07 (kanamycin resistant) were exposed to BHK cells with or without the hG-CSF receptor on their surface, washed and eluted from the cell surface. Eluted phage were introduced into E. coli and the transfected cells were plated on media containing kanamycin or ampicillin. The ratio of ampicillin resistant to kanamycin resistant particles were determined prior to and following exposure to receptor by counting resistant colonies.

Table 5: Activity of selected permuteins

Plasmid Permutein breakpoint Activity in G-CSF- in G-CSF amino acid dependent proliferation sequence assay

pMON16017 3/2

pMON16018 11/10

pMON16019 13/12

PMON16020 19/18

pMON16021 49/48

pMON 16022 60/59

pMON 16023 67/66

pMON16024 69/68

pMON16025 71/70

pMON16026 123/122

pMON16027 125/124

pMON 16028 159/158

Table 6: SEQ ID Number/SEQ ID Name Correlation

1 TCT ACA CCA TTG GGC CCT GCC AGC

2 TCT CCA TTG GGC CCT GCC AGC TCC

3 TCT TTG GGC CCT GCC AGC TCC CTG

4 TCT GGC CCT GCC AGC TCC CTG CCC

5 TCT CCT GCC AGC TCC CTG CCC CAG

6 TCT GCC AGC TCC CTG CCC CAG AGC

TCT AGC CC CTG CCC CAG AGC TTC

8 TCT TCC CTG CCC CAG AGC TTC CTG

9 TCT CTG CCC CAG AGC TTC CTG CTC

10 TCT CCC CAG AGC TC CTG CTC AAG

11 TCT CAG AGC TTC CTG CTC AAG TCT

12 TCT AGC TC CTG CTC AAG TCT TTA

13 TCT TTC CTG CTC AAG TCT TTA GAG

14 TCT CTG TC AAG TCT TTA GAG CAA

15 TCT CTC AAG TCT TTA GAG CAA GTG

16 TCT AAG CT TTA GAG CAA GTG AGG π TCT TCT TTA GAG CAA GTG AGG AAG

18 TCT TTA GAG CAA GTG AGG AAG ATC

19 TCT GAG CAA GTG AGG AAG ATC CAG

20 TCT CAA GTG AGG AAG ATC CAG GGC

21 TCT GTG AGG AAG ATC CAG GGC GAT

22 CT AGG AAG ATC CAG GGC GAT GGC

23 TCT AAG ATC CAG GGC GAT GGC GCA

24 TCT ATC CAG GGC GAT GGC GCA GCG

25 TCT CAG GGC GAT GGC GCA GCG CTC

26 TCT GGC GAT GGC GCA GCG CTC CAG

27 TCT GAT GGC GCA GCG CTC CAG GAG

28 TCT GGC GCA GCG CTC CAG GAG AAG

29 TCT GCA GCG CTC CAG GAG AAG CTG

30 TCT GCG CTC CAG GAG AAG CTG TGT

31 TCT CTC CAG GAG AAG CTG TGT GCC

32 TCT CAG GAG AAG CTG TGT GCC ACC

33 TCT GAG AAG CTG TGT GCC ACC TAC

34 TCT AAG CTG TGT GCC ACC TAC AAG

35 TCT CTG TGT GCC ACC TAC AAG CTG

36 TCT TGT GCC ACC TAC AAG CTG TGC

37 TCT GCC ACC TAC AAG CTG TGC CAC

38 CT ACC TAC AAG CTG TGC CAC CCC

39 TCT TAC AAG CTG TGC CAC CCC GAG

40 TCT AAG CTG TGC CAC GAG GAG

41 TCT CTG TGC CAC CCC GAG GAG CTG

42 TCT TGC CAC CCC GAG GAG CTG GTG

43 TCT CAC CCC GAG GAG CTG GTG CTG

44 TCT CCC GAG GAG CTG GTG CTG CTC

45 TCT GAG GAG CTG GTG CTG CTC GGA

46 TCT GAG CTG GTG CTG CTC GGA CAC

47 CT CTG GTG CTG CTC GGA CAC TCT

48 TCT GTG CTG CTC GGA CAC TCT CTG

49 TCT CTG CTC GGA CAC TCT CTG GGC

50 TCT CTC GGA CAC TCT CTG GGC ATC

51 TCT GGA CAC TCT CTG GGC ATC CCC

52 TCT CAC TCT CTG GGC ATC CCC TGG

53 TCT TCT CTG GGC ATC CCC TGG GCT

54 TCT CTG GGC ATC CCC TGG GCT CCC

55 TCT GGC ATC CCC TGG GCT CCC CTG

56 TCT ATC TGG GCT CCC CTG AGC

57 TCT CCC TGG GCT CCC CTG AGC TCC

58 TCT TGG GCT CCC CTG AGC TCC TGC

59 TCT GCT CCC CTG AGC TCC TGC CCC

TCT CCC CTG AGC TCC TGC CCC AGC

TCT CTG AGC TCC TGC CCC AGC CAG

62 CT AGC TCC TGC CCC AGC CAG GCC

63 TCT TCC TGC CCC AGC CAG GCC CTG

64 TCT TGC CCC AGC CAG GCC CTG CAG

C5 TCT CCC AGC CAG GCC CTG CAG CTG

TCT AGC CAG GCC CTG CAG CTG GCA

67 ΓGCGCGC TCT CAG GCC CTG CAG CTG GCA GGC

68 -GCGCG: TCT GCC CTG CAG CTG GCA GGC TGC

69 TCT CTG CAG CTG GCA GGC TGC TTG

"0 TCT CAG CTG GCA GGC TGC TTG AGC

TCT CTG GCA GGC TGC TTG AGC CAA

Fes'¹ TCT GCA GGC TGC TTG AGC CAA CTC

- TCT GGC TGC TTG AGC CAA CTC CAT

74 TCT TGC TTG AGC CAA CTC CAT AGC

75 TCT TTG CAA CTC CAT AGC GGC

76 ^*~rτ CTC CAT AGC GG: CTT

TCT CAA T CAT AGC GGC CT~ TC

TCT CTC GAT AGC GGC CTT ~rc CTC

:" CAT GGC CTT TTC CT: TAC 80 FGS80 TCT AGC GGC CTT TC CTC TAC CAG

81 FGS81 TCT GGC CTT TTC CTC TAC CAG GGG

82 FGS82 TCT CTT TTC CTC TAC CAG GGG CTC

83 FGS83 TCT TTC CTC TAC CAG GGG CTC CTG

84 FGS84 TCT CTC TAC CAG GGG CTC CTG CAG

85 FGS85 CT TAC CAG GGG CTC CTG CAG GCC

86 FGS86 TCT CAG GGG CTC CTG CAG GCC CTG

87 FGS87 TCT GGG CTC CTG CAG GCC CTG GAA

88 FGS88 CT CTC CTG CAG GCC CTG GAA GGG

89 FGS89 TCT CTG CAG GCC CTG GAA GGG ATA

90 FGS90 TCT CAG GCC CTG GAA GGG ATA TCC

91 FGS91 TCT GCC CTG GAA GGG ATA TCC CCC

92 FGS92 TCT CTG GAA GGG ATA TCC CCC GAG

93 FGS93 TCT GAA GGG ATA TCC CCC GAG TTG

94 FGS94 TCT GGG ATA TCC CCC GAG TTG GGT

95 FGS95 TCT ATA TCC CCC GAG TTG GGT CCC

96 FGS96 TCT TCC CCC GAG TTG GGT CCC ACC

97 FGS97 TCT GAG TTG GGT CCC ACC TTG

98 FGS98 TCT GAG TTG GGT CCC ACC TTG GAC

99 FGS99 TCT TTG GGT CCC ACC TTG GAC ACA

100 FGS100 TCT GGT CCC ACC TTG GAC ACA CTG

101 FGS101 TCT CCC ACC TTG GAC ACA CTG CAG

102 FGS102 TCT ACC TTG GAC ACA CTG CAG CTG

103 FGS103 TCT TTG GAC ACA CTG CAG CTG GAC

104 FGS104 TCT GAC ACA CTG CAG CTG GAC GTC

105 FGS105 CT ACA CTG CAG CTG GAC GTC GCC i06 FGS106 TCT CTG CAG CTG GAC GTC GCC GAC

107 FGΞ107 TCT CAG CTG GAC GTC GCC GAC TTT

108 FGS108 TCT CTG GAC GTC GCC GAC TTT GCC

109 FGS109 TCT GAC GTC GCC GAC TT GCC ACC

110 FGS110 TCT GTC GCC GAC TTT GCC ACC ACC

111 FGS111 TCT GCC GAC TTT GCC ACC ACC ATC

112 FGS112 TCT GAC TTT GCC ACC ACC ATC TGG

113 FGS113 TCT TTT GCC ACC ACC ATC TGG CAG

114 FGS114 TCT GCC ACC ACC ATC TGG CAG CAG

115 FGS115 TCT ACC ACC ATC TGG CAG CAG ATG

116 FGS116 TCT ACC ATC TGG CAG CAG ATG GAA

117 FGS117 TCT ATC TGG CAG CAG ATG GAA GAA

118 FGS118 TCT TGG CAG CAG ATG GAA GAA CTG

119 FGS119 TCT CAG CAG ATG GAA GAA CTG GGA

120 FGS120 TCT CAG ATG GAA GAA CTG GGA ATG

121 FGS121 CT ATG GAA GAA CTG GGA ATG GCC

122 FGΞ122 TCT GAA GAA CTG GGA ATG GCC CCT

123 FGS123 TCT GAA CTG GGA ATG GCC CCT GCC

124 FGS124 TCT CTG GGA ATG GCC CCT GCC CTG

125 FGS125 TCT GGA ATG GCC CCT GCC CTG CAG

126 FGS126 TCT ATG GCC CCT GCC CTG CAG CCC

127 FGS127 TCT GCC CCT GCC CTG CAG CCC ACC

128 FGS128 TCT CCT GCC CTG CAG CCC ACC CAG

129 FGS129 TCT GCC CTG CAG CCC ACC CAG GGT

130 FGS130 TCT CTG CAG CCC ACC CAG GGT GCC

131 FGS131 TCT CAG CCC ACC CAG GGT GCC ATG

FGS132 TCT CCC ACC CAG GGT GCC ATG CCG

133 FGS133 TCT ACC CAG GGT GCG ATG CCG GCC

134 FGS134 TCT CAG GGT GCC ATG CCG GCC TTC

135 FGS135 TCT GGT GCC ATG CCG GCC TTC GCC

136 FGS136 TCT GCC ATG CCG GCC TTC GCC TCT

137 FGS1 7 TCT ATG CCG GCC TTC GCC TCT GCT

FGS138 TCT CCG GCC TTC GCC TCT GCT TTC

FGS139 TCT GCC TTC GCC CT GCT TTC CAG

140 FGS140 TCT TTC GCC CT GCT TTC CAG CGC

141 FGS141 TCT GCC TCT GCT TC CAG CGC CGG

1.42 FGS142 CT TCT GCT TTC CAG CGC CGG GCA

143 F"S143 CT GCT TTC CAG CGC CGG GCA GGA r« FGS144 TCT TTC CAG CGC CGG GCA GGA GGG

145 FGS145 TCT CAG CGC CGG GCA GGA GGG GTC

146 FGS146 TCT CGC CGG GCA GGA GGG GTC CTG

147 FGS147 TCT CGG GCA GGA GGG GTC CTG GTT

148 FGS148 TCT GCA GGA GGG GTC CTG GTT GCT

149 FGS149 TCT GGA GGG GTC CTG GTT GCT AGC

150 FGS150 TCT GGG GTC CTG GTT GCT AGC CAT

151 FGS151 TCT GTC CTG GTT GCT AGC CVT CTG

152 FGS152 TCT CTG GTT GCT AGC CAT CTG CAG

;^•> > F3S153 TCT GTT GCT AGC C T CTG CAG AGC i:4 FGS1 TCT GCT AGC CAT C-G CAG AGC TTC

"GSISS TCT AGC CAT CTG CAG AGC TTC CTG

-GS156 TCT CAT CTG CAG AGC TTC CTG GAG

_-~ TGE b^" TCT CTG CAG AGC TC CTG GAG GTG

158 FGS158 TCT CAG AGC TTC CTG GAG GTG TCG

159 FGS159 TCT AGC TTC CTG GAG GTG TCG TAC

160 FG=16C TCT TTC CTG GAG GTG TCG TAC CGC

161 , ATX, τ~τ CTG GAG GTG TCG TAC CGC GTT

161 T-: GAG GTG TCG TAC CGG GTT CTA

:«: TCT GTC TCG TAC CGG GTT CTA CGC

154 CT TCG TAC c— CTA CG" CAC

165 Tcc σz TCT TAC CGC GTT CTi CGC r_Λc CTT 166 -CT CGC GTT CTA CGC CAC CTT GCG

167 TCT GTT CTA CGC CTT GCG CAG

168 TCT CTA CGC CAC CTT GCG CAG CCC

169 TCT CGC CAC CTT GCG CAG CCC GA'C

170 TCT CAC CTT GCG CAG CCC GA'C ATG

171 TCT CTT GCG CAG CCC GA C ATG GCT

172 TCT GCG CAG CCC ATG GCT ACA

173 TCT CAG CCC ATG GCT ACA CCA

174 TCT CCC ATG GCT ACA CCA TTG

175 TCT CGC CAC CTT GCG CAG CCC A CT

176 TCT CAC CTT GCG CAG CCC A CT AGT

177 TCT CTT GCG CAG CCC A CT AGT CAT

178 TCT GCG CAG CCC A CT AGT CAT CCA

179 TCT CAG CCC AGT CAT CCA CCT

180 TCT CCC AGT CAT CCA CCT ATG

181 TCT CGC CAC CTT GCG CAG CCC GGC

182 TCT CAC CTT GCG CAG CCC GGC GGC

183 TCT CTT GCG CAG CCC GGC GGC GGC

184 TCT GCG CAG CCC GGC GGC GGC TCT

185 TCT CAG CCC GGC GGC GGC TCT GA'C

186 TCT CCC GGC GGC GGC TCT GA C ATG

187 AGC CAT GTC GGG CTG CGC AAG

188 AGC CAT GTC ACG CGT ACG ATT

189 AGC CAT GTC AGA GCC GCC GCC

190 TGT AGC CAT GTC GGG CTG CGC

191 TGT AGC CAT GTC ACG CGT ACG

192 TGT AGC CAT GTC AGA GCC GCC

193 TGG TGT AGC CAT GTC GGG CTG

194 TGG TGT AGC CAT GTC ACG CGT

195 TGG TGT AGC CAT GTC AGA GCC

196 CAA TGG TGT AGC CAT GTC GGG

197 CAA TGG TGT AGC CAT GTC ACG

198 CAA TGG TGT AGC CAT GTC AGA

199 GCC CAA TGG TGT AGC CAT GTC

200 AGG GCC CAA TGG TGT AGC CAT

201 GGC AGG GCC CAA TGG TGT AGC

202 GCT GGC AGG GCC CAA TGG TGT

203 GGA GCT GGC AGG GCC CAA TGG

204 CAG GGA GCT GGC AGG GCC CAA

205 GGG CAG GGA GCT GGC AGG GCC

206 CTG GGG CAG GGA GCT GGC AGG

207 GCT CTG GGG CAG GGA GCT GGC

208 GAA GCT CTG GGG CAG GGA GCT

209 CAG GAA GCT CTG GGG CAG GGA

210 GAG CAG GAA GCT CTG GGG CAG

211 CTT GAG CAG GAA GCT CTG GGG

212 AGA CTT GAG CAG GAA GCT CTG

213 TAA AGA CTT GAG CAG GAA GCT

214 CTC TAA AGA CTT GAG CAG GAA

215 TTG CTC TAA AGA CTT GAG CAG

216 CAC TTG CTC TAA AGA CTT GAG

217 CCT CAC TTG CTC TAA AGA CTT

218 CTT CCT CAC TTG CTC TAA AGA

219 GAT CTT CCT CAC TTG CTC TAA

220 CTG GAT CTT CCT CAC TTG CTC

221 GCC CTG GAT CTT CCT CAC TTG

222 ATC GCC CTG GAT CTT CCT CAC

223 GCC ATC GCC CTG GAT CTT CCT

224 TGC GCC ATC GCC CTG GAT CTT

225 CGC TGC GCC ATC GCC CTG GAT

226 GAG CGC TGC GCC ATC GCC CTG

227 CTG GAG CGC TGC GCC ATC GCC

228 CTC CTG GAG CGC TGC GCC ATC

229 CTT CTC CTG GAG CGC TGC GCC

230 CAG CTT CTC CTG GAG CGC TGC

231 ACA CAG CTT CTC CTG GAG CGC

232 GGC ACA CAG CTT CTC CTG GAG

233 GGT GGC ACA CAG CTT CTC CTG

234 GTA GGT GGC ACA CAG CTT CTC

235 CTT GTA GGT GGC ACA CAG CTT

236 CAG CTT GTA GGT GGC ACA CAG

C37 FGSil GCA CAG CTT GTA GGT GGC ACA

GTG GCA CAG CTT GTA GGT GGC

229 GGG GTG GCA CAG CTT GTA GGT

240 CTC GGG GTG GCA CAG CTT GTA

:n CTC CTC GGG GTG GCA CAG CTT z.~ CAG CTC CTC GGG GTG GCA CAG

-i: CAC CAG CTC CTC GGG GTG GCA

244 CAG CAC CAG CTC CTC GGG GTG

245 GAG CAG CAC CAG CTC CTC GGG

246 TCC GAG CAG CAC CAG CTC CTC

24- GTG TCC GAG CAG CAC CAG CTC

248 AGA GTG TCC GAG CAG CAC CAG

249 CAG AGA GTG TCC GAG CAG CC

250 GCC CAG AGA GTG CC GAG C^G

251 GAT GCC CAG AGA GTG -r~ GAG 252 PGS57 TATATAT GCGGCCGC GGG GAT GCC CAG AGA GTG TCC

253 RGS58 TATATAT GCGGCCGC CCA GGG GAT GCC CAG AGA GTG

254 RGS59 TATATAT GCGGCCGC AGC CCA GGG GAT GCC CAG AGA

255 RGS60 TATATAT GCGGCCGC GGG AGC CCA GGG GAT GCC CAG

256 RGS61 TATATAT GCGGCCGC CAG GGG AGC CCA GGG GAT GCC

257 RGS62 TATATAT GCGGCCGC GCT CAG GGG AGC CCA GGG GAT

258 RGS63 TATATAT GCGGCCGC GGA GCT CAG GGG AGC CCA GGG

259 RGΞ64 TATATAT GCGGCCGC GCA GGA GCT CAG GGG AGC CCA

260 RGS65 TATATAT GCGGCCGC GGG GCA GGA GCT CAG GGG AGC

261 PGS66 TATATAT GCGGCCGC GCT GGG GCA GGA GCT CAG GGG

262 RGS67 TATATAT GCGGCCGC CTG GCT GGG GCA GGA GCT CAG

263 RGS68 TATATAT GCGGCCGC GGC CTG GCT GGG GCA GGA GCT

264 RGS69 TATATAT GCGGCCGC CAG GGC CTG GCT GGG GCA GGA

265 RGS70 TATATAT GCGGCCGC CTG CAG GGC CTG GCT GGG GCA

266 RGS71 TATATAT GCGGCCGC CAG CTG CAG GGC CTG GCT GGG

267 RGS72 TATATAT GCGGCCGC TGC CAG CTG CAG GGC CTG GCT

268 RGS73 TATATAT GCGGCCGC GCC TGC CAG CTG CAG GGC CTG

269 RGS7 TATATAT GCGGCCGC GCA GCC TGC CAG CTG CAG GGC

270 RGS75 TATATAT GCGGCCGC CAA GCA GCC TGC CAG CTG CAG

271 RGS76 TATATAT GCGGCCGC GCT CAA GCA GCC TGC CAG CTG

272 RGS77 TATATAT GCGGCCGC GTT GCT CAA GCA GCC TGC CAG

273 RGS78 TATATAT GCGGCCGC GAG GTT GCT CAA GCA GCC TGC

274 RGS79 TATATAT GCGGCCGC ATG GAG GTT GCT CAA GCA GCC

275 RGS80 TATATAT GCGGCCGC GCT ATG GAG GTT GCT CAA GCA

276 RGS81 TATATAT GCGGCCGC GCC GCT ATG GAG GTT GCT CAA

277 RGS82 TATATAT GCGGCCGC AAG GCC GCT ATG GAG GTT GCT

278 RGS83 TATATAT GCGGCCGC GAA AAG GCC GCT ATG GAG GTT

279 RGS84 TATATAT GCGGCCGC GAG GAA AAG GCC GCT ATG GAG

280 RGE85 TATATAT GCGGCCGC GTA GAG GAA AAG GCC GCT ATG

281 RGS86 TATATAT GCGGCCGC CTG GTA GAG GAA AAG GCC GCT

262 RGSS" TATATAT GCGGCCGC CCC CTG GTA GAG GAA AAG GCC

283 RGS88 TATATAT GCGGCCGC GAG CCC CTG GTA GAG GAA AAG

284 ROS89 TATATAT GCGGCCGC CAG GAG CCC CTG GTA GAG GAA

285 RGS90 TATATAT GCGGCCGC CTG CAG GAG CCC CTG GTA GAG

286 RGS91 TATATAT GCGGCCGC GGC CTG CAG GAG CCC CTG GTA

287 RGS92 TATATAT GCGGCCGC CAG GGC CTG CAG GAG CCC CTG

288 RGS93 TATATAT GCGGCCGC TTC CAG GGC CTG CAG GAG CCC

289 RGS94 TATATAT GCGGCCGC CCC TTC CAG GGC CTG CAG GAG

290 RGS95 TATATAT GCGGCCGC TAT CCC TTC CAG GGC CTG CAG

291 RGS96 TATATAT GCGGCCGC GGA TAT CCC TTC CAG GGC CTG

292 RGS97 TATATAT GCGGCCGC GGG GGA TAT CCC TTC CAG GGC

293 RGS98 TATATAT GCGGCCGC CTC GGG GGA TAT CCC TTC CAG

294 RGS99 TATATAT GCGGCCGC CAA CTC GGG GGA TAT CCC TTC

295 RGS100 TATATAT GCGGCCGC ACC CAA CTC GGG GGA TAT CCC

296 RGS101 TATATAT GCGGCCGC GGG ACC CAA CTC GGG GGA TAT

297 RGS102 TATATAT GCGGCCGC GGT GGG ACC CAA CTC GGG GGA

298 RGS103 TATATAT GCGGCCGC CAA GGT GGG ACC CAA CTC GGG

299 RGS104 TATATAT GCGGCCGC GTC CAA GGT GGG ACC CAA CTC

300 RGS105 TATATAT GCGGCCGC TGT GTC CAA GGT GGG ACC CAA

301 RGS106 TATATAT GCGGCCGC CAG TGT GTC CAA GGT GGG ACC

302 RGS107 TATATAT GCGGCCGC CTG CAG TGT GTC CAA GGT GGG

303 RGS108 TATATAT GCGGCCGC CAG CTG CAG TGT GTC CAA GGT

304 RGS109 TATATAT GCGGCCGC GTC CAG CTG CAG TGT GTC CAA

305 RGS110 TATATAT GCGGCCGC GAC GTC CAG CTG CAG TGT GTC

306 PGS111 TATATAT GCGGCCGC GGC GAC GTC CAG CTG CAG TGT

307 RGS112 TATATAT GCGGCCGC GTC GGC GAC GTC CAG CTG CAG

308 RGS113 TATATAT GCGGCCGC AAA GTC GGC GAC GTC CAG CTG

309 RGΞ114 TATATAT GCGGCCGC GGC AAA GTC GGC GAC GTC CAG

310 RGS11S TATATAT GCGGCCGC GGT GGC AAA GTC GGC GAC GTC

311 RGS116 TATATAT GCGGCCGC GGT GGT GGC AAA GTC GGC GAC

312 RGS117 TATATAT GCGGCCGC GAT GGT GGT GGC AAA GTC GGC

313 RGS118 TATATAT GCGGCCGC CCA GAT GGT GGT GGC AAA GTC

314 RGS119 TATATAT GCGGCCGC CTG CCA GAT GGT GGT GGC AAA

315 RGS120 TATATAT GCGGCCGC CTG CTG CCA GAT GGT GGT GGC

316 RGS121 TATATAT GCGGCCGC CAT CTG CTG CCA GAT GGT GGT

317 RGS122 TATATAT GCGGCCGC TTC CAT CTG CTG CCA GAT GGT

318 RGS123 TATATAT GCGGCCGC TTC TTC CAT CTG CTG CCA GAT

319 PGS124 TATATAT GCGGCCGC CAG TTC TTC CAT CTG CTG CCA

320 RGS125 TATATAT GCGGCCGC TCC CAG TTC TTC CAT CTG CTG

321 RGS1 6 TATATAT GCGGCCGC CAT TCC CAG TTC TTC CAT CTG

322 RGS127 TATATAT GCGGCCGC GGC CAT CC CAG TTC TC CAT

323 PGS128 TATATAT GCGGCCGC AGG GGC CAT TCC CAG TTC TTC

324 RGS129 TATATAT GCGGCCGC GGC AGG GGC CAT TCC CAG TTC

325 PGS130 TATATAT GCGGCCGC CAG GGC AGG GGC CAT TCC CAG

326 PGS131 TATATAT GCGGCCGC CTG CAG GGC AGG GGC CAT TCC

2:^π RGS132 TATATAT GCGGCCGC GGG CTG CAG GGC AGG GGC CAT

--_" PGS133 TATATAT GCGGCCGC GGT GGG CTG CAG GGC AGG GGC

329 RGS134 TATATAT GCGGCCGC CTG GGT GGG CTG CAG GGC AGG

330 RGS135 TATATAT GCGGCCGC ACC CTG GGT GGG CTG CAG GGC

331 PGS136 TATATAT GCGGCCGC GGC ACC CTG GGT GGG CTG CAG

332 PGS13^*1 TATATAT GCGGCCGC CAT GGC ACC CTG GGT GGG CTG

333 RGS.,- TATATAT GCGGCCGC CGG CAT GGC ACC CTG GGT GGG

334 PGC." - TATATAT GCGGCCGC GGC CGG CAT GGC ACC CTG GGT

PCC_4G TATATAT GCGGCCGC GAA GGC CGG CAT GGC AGG CTG

PGS14.L TATATAT GCGGCCGC GGC GAA GGC CGG CAT GGC ACC

-37 RGS142 TATATAT GCGGCCGC AGA GGC GAA GGC CGG CAT GGC 338. RGΞ143 TATATAT GCGGCCGC AGC AGA GGC GAA GGC CGG CAT

339. RGΞ144 TATATAT GCGGCCGC GAA AGC AGA GGC GAA GGC CGG

340. RGΞ145 TATATAT GCGGCCGC CTG GAA AGC AGA ^~ GGC GAA GGC

341. RGS146 TATATAT GCGGCCGC GCG CTG GAA AGC AGA GGC GAA

342. RGS147 TATATAT GCGGCCGC CCG GCG CTG GAA AGC AGA GGC

343. RGS148 TATATAT GCGGCCGC TGC CCG GCG CTG GAA AGC AGA

344. RGS149 TATATAT GCGGCCGC TCC TGC CCG GCG CTG GAA AGC

345. RGS150 TATATAT GCGGCCGC CCC TCC TGC CCG GCG CTG GAA

346. RGS151 TATATAT GCGGCCGC GAC CCC TCC TGC CCG GCG CTG

347. RGS152 TATATAT GCGGCCGC CAG GAC CCC TCC TGC CCG GCG

348. RGS153 TATATAT GCGGCCGC AAC CAG GAC CCC TCC TGC CCG

349. RGS154 TATATAT GCGGCCGC AGC AAC CAG GAC CCC TCC TGC

350. RGΞ155 TATATAT GCGGCCGC GCT AGC AAC CAG GAC CCC TCC

351. RGS156 TATATAT GCGGCCGC ATG GCT AGC AAC CAG GAC CCC

352. RGS157 TATATAT GCGGCCGC CAG ATG GCT AGC AAC CAG GAC

353. RGS158 TATATAT GCGGCCGC CTG CAG ATG GCT AGC AAC CAG

354. RGS159 TATATAT GCGGCCGC GCT CTG CAG ATG GCT AGC AAC

355. P.GS160 TATATAT GCGGCCGC GAA GCT CTG CAG ATG GCT AGC

356. RGS161 TATATAT GCGGCCGC CAG GAA GCT CTG CAG ATG GCT

357. RGS162 TATATAT GCGGCCGC CTC CAG GAA GCT CTG CAG ATG

358. RGS163 TATATAT GCGGCCGC CAC CTC CAG GAA GCT CTG CAG

359. RGS164 TATATAT GCGGCCGC CGA CAC CTC CAG GAA GCT CTG

360. RGS165 TATATAT GCGGCCGC GTA CGA CAC CTC CAG GAA GCT

361. RGS166 TATATAT GCGGCCGC GCG GTA CGA CAC CTC CAG GAA

362. RGS167 TATATAT GCGGCCGC AAC GCG GTA CGA CAC CTC CAG

363. RGS168 TATATAT GCGGCCGC TAG AAC GCG GTA CGA CAC CTC

364. RGS169 TATATAT GCGGCCGC GCG TAG AAC GCG GTA CGA CAC

365. RGS170 TATATAT GCGGCCGC GTG GCG TAG AAC GCG GTA CGA

366. RGS171 TATATAT GCGGCCGC AAG GTG GCG TAG AAC GCG GTA

367. RGS172 TATATAT GCGGCCGC CGC AAG GTG GCG TAG AAC GCG

368 P.GS173 TATATAT GCGGCCGC CTG CGC AAG GTG GCG TAG AAC GCG

Claims

What is claimed is:

1. A method for making a biologically-active circularly-permuted protein of the formula C¹-L¹-N¹, derived from a parent protein of the formula N ϋ¹, 5 wherein

C¹ is comprised of a segment derived from the carboxy portion of said parent protein;

N¹ is comprised of a segment derived from the amino terminal portion of said parent protein; and

10 L¹ is a chemical bond or a linker, linking C¹ to the amino terminus of L¹ and carboxy terminus of L¹ to the amino terminus of N¹;

comprising the steps of:

(a) making a series of circularly-permuted genes;

(b) inserting said circularly-permuted genes into a display vector;

lδ (c) expressing said circularly-permuted genes such that the proteins encoded by said genes are presented on the surface of the display vector;

(d) generatmg a library of display vectors presenting the expressed circularly permuted protein;

20 (e) affinity-selecting the presenting display vectors with a target protein that can bind a biologically-active circularly-permuted protein;

(f) isolating and analyzing clones of selected display vectors to identify the presented circularly-permuted protein.

5 2. The method of claim 1 wherein the method of making a series of circularly-permuted genes is selected from the group consisting of making a tandemly-repeated intermediate, total synthesis of a synthetic gene, assembly of a gene from synthetic oligonucleotides, DNA amplification, and limited digestion of a circular intermediate.

3. The method of claim 1 wherein said display vector is selected from the group consisting of bacteriophage display vectors, bacteria, and baculovirus vectors.

5 4. The method of claim 3, wherein said presentation vector is a bacteriophage.

δ. The method of claim 4, wherein said presentation vector is bacteriophage M13.

6. The method of claim δ, wherein said

10 presentation vector is a bacteriophage M13 gene III vector.

7. The method of claim 1 wherein said method of making a series of circularly permuted genes is a method of making a tandem repeat intermediate.

lδ 8 The method of claim 7 wherein said circularly-permuted genes are amplified from the repeat by gene amplification.

9. The method of claim 1 wherein said method of affinity selection comprises the steps consisting of:

(a) binding said presentation display vectors to a target 20 protein;

(b) eluting said display vectors;

(c) amplifying said display vectors; and

(d) biopanning a pool of said amplified display vectors.

10. The method of claim 1 wherein L¹ is a linear peptide linker.

δ 11. The method of claim 1 wherein said the DNA sequence encoding said linker L¹ is selected from the group consisting of SEQ ID NO: 1 through SEQ ID NO: 368.

12. The method of claim 1 wherein the length of the C¹ in said permutein is shorter than the length of C¹ in said parent protein.

13. The method of claim 1 wherein the length of the N¹ in said permutein is shorter than the length of N¹ in said parent protein.

14. A circularly-permuted protein prepared by the method of claim 1.

lδ. A circularly-permuted protein of claim 14 comprising the G- CSF receptor agonist domain of a species of mylepoietin (MPO).