AU627477B2 - Interleukin ii analogs - Google Patents

Description

ci i:.d I __LI OPI DATE 05/02/90 AOJP DATE 22/03/90 APPLN. ID 38776 89 PCT NUMBER PCT/US89/02917 PCrr INTERNATIONAL APPLICATION PUBLISHED UNDER THE PATENT COOPERATION TREATY (PCT) (51) International Patent Classification 4 (I1) International Publication Number: WO 90/00565 C07K 13/00, 17/00, 3/26

A

A61K 45/00, C12N 15/00 A (43) International Publication Date: 25 January 1990 (25.01.90) C07H 17/00, 21/04 (21) International Application Number: PCT/US89/02917 (74) Agents: SCOTT, Watson, T. et al.; Cushman, Darby Cushman, Eleventh Floor, 1615 L Street Washing- (22) International Filing Date: 5 July 1989 (05.07.89) ton, DC 20036-5601 (US).

Priority data: (81) Designated States: AT (European patent), AU, BE (Euro- 214,998 5 July 1988 (05.07.88) US pean patent), CH (European patent), DE (European patent), FR (European patent), GB (European patent), IT (European patent), JP, LU (European patent), NL (Eu- (71)Applicant: AMGEN INC. [US/US]; 1900 Oak Terrace ropean patent), SE (European patent).

Lane, Thousand Oaks, CA 91320 (US).

(72)Inventors: ALTROCK, Bruce, W. 115 Jerome Avenue, Published Newbury Park, CA 91320 BOONE, Thomas, C. With international search report.

3913 Elkwood, Newbury Park, CA 91320 GOLD- MAN, Robert, A. 12 Pinebroc.: Road, Boulder, CO 80301 KENNY, William, C. 2654 Castillo Circle, Thousand Oaks, CA 91360 STABINSKY, Yitzhak 82 Ahuza, 43 000 Raanana (IL).

(54)Title: INTERLEUKIN II ANALOGS Arg 81- Leu j.

d p: ii

I;

B' E Glu 106 1--Lys 8 'Asp 109 Pro 47 Thr 51 Ala 112 A Lys 48 LYS 54 R

R'

(57) Abstract Interleukin II analogs and DNA sequences comprising structural genes coding for such analogs which differ from the naturally-occurring forms in terms of the identity and/or location of one or more amino acids are disclosed.

/i j

I

I

PCT/US89/0291 7 WO 90/00565 1 INTERLEUKIN II ANALOGS i 1i

I

II

.1 1,8 9 The present invention relates generally to the manipulation of genetic materials and, more particularly, to the manufacture of specific DNA sequences useful in recombinant procedures to secure expression of Interleukin II analogs.

Background of the Invention Interleukin II a glycoprotein with a molecular weight of approximately 15,000, is a member of a group of proteins, called lymphokines, that control the body's immune response. IL-2 is produced by certain white blood cells, lectin- or antigen-activated T cells, and plays a central role in the body's immune system as a lymphocyte regulating molecule.

IL-2 has been reported to enhance thymocyte mitogenesis, to stimulate long-term in vitro growth of 20 activated T-cell clones, to induce cytotoxic T-cell reactivity, to modulate immunological effects on activated B cells and lymphokine activated cells, to induce plaque-forming cell responses in cultures of nude mouse spleen cells, and to regulate production of gamma interferon. It also augments natural killer cell activity and mediates the recovery of the immune function of lymphocytes in selected immunodeficient states.

Additionally, in the laboratory, IL-2 is used to maintain cultures of functional monoclonal T-cells to study the molecular nature of T-cell differentiation, and to help elicit the mechanism of differentiated T-cPll functions. Thus, IL-2 has application in both research and the treatment of neoplastic and immunodeficiency diseases.

PCT/US89/029j~7 WO 90/00565 2 IL-2 asserts its effect binding to a specific high affinity receptor on the surface of target cells; consequently, the IL-2 molecule has become a focal point for studying receptor-effector interactions that modulate cell proliferation in the immune response.

The high affinity (K D -10-11M) receptor reponsible for mediating the effect of IL-2 on target cells consists of two distinct membrane-bound proteins of size 55 kD (p55 or Tac) and 75 kD (p75); each of these two proteins can act by itself as an apparent low affinity (K D -10- 8 M) receptor for IL-2, and both are required for IL-2 activity. This suggests that IL-2 must bind both p55 and p75 to form a trimeric complex for activity, and by inference, that IL-2 must have two separate receptor binding sites.

The limited amount of purified native IL-2 obtainable from peripheral blood lymphocytes and tumor cell lines was an impediment to studies of the biological role of this lymphokine until the advent of recombinant production of IL-2.

Taniguchi, et al., Nature, 302: 305-310 (1983) described the sequence analysis, cloning, and expression of a complementary DNA coding for human IL-2, cloned from a cDNA library prepared from partially purified IL-2 mRNA from the Jurkat leukemia cell line.

IL-2 was proposed to comprise 133 amino acid residues and to have a calculated molecular weight of about 15,420. Taniguchi described the cloning procedures and the expression of the cDNA for IL-2 in cultured monkey COS cells. The publication states that expression of the IL-2 cDNA in E. coli had not yet been accomplished. See also European Patent Applications 118,617, published September 19, 1984; 118,977, published September 19, 1984; and 119,621, published September 26, 1984, and U.S. Patent 4,738,927.

I WO 90/00565 PCT/US89/02917 3 Rosenberg, et al., Science, 223: 1412-1415 (i984) reported the isolation of another cDNA clone of the IL-2 gene from the Jurkat tumor cell line and from normal human peripheral blood lymphocytes. These researchers inserted the gene into E. coli, purified the polypeptide product and assayed it for biological activity. See also, Wang, et al., Science, 224, 1431-1433 -(1984) referring to site-specific mutagenesis of a human IL-2 gene as well as European Patent Application 109,748, published May 30, 1984.

IL-2 modifications reported in the literature include: Ju et al., J. Biol. Chem. 262, 5723(1987); Liang et al., J. Biol. Chem. 261, 334(1986); and Miyaji et al., Agric. Biol. Chem., 51 1135(1987).

Considerable interest exists in the development of methods and materials for the production of large amounts of purified IL-2 analogs to replace IL-2-containing preparations currently employed in immunotherapy research.

It is an object of the subject invention to provide improved forms of IL-2.

It is a further object of the invention to provide IL-2 analogs having less toxicity than IL-2 preparations currently used.

It is a further object of the invention to provide IL-2 analogs which allow atVchment of a ligand, without affecting biological activity.

It is a still further object of the invention to provide a method of purifying IL-2.

Other objects, features and characteristics of the present invention will become apparent upon consideration of the following description and the appended claims.

I a W09 Ia Ui i1 i

I

14 0/00565 Summary c PCT/US89/029 17 4 of the Invention The subject invention relates to IL-2 analogs having modified receptor domains, and analogs having stabilized IL-2 structure. The subject invention also relates to IL-2 analogs which have been modified to permit the attachment of a ligand. More particularly the subject invention relates to a polypeptide product of the expression in a host cell of a manufactured gene, the polypeptide having an amino acid sequence represented by formula below wherein at least one of the 47th, 51st, 80th, 81st, 106th, 109th, 112th, 119th, 120th, 123rd, 127th, 129th, 131st, and 133rd original amino acid residues is replaced by a substitution amino acid residue, or wherein at least two of the 8th, 47th, 48th, 51st, 54th, 80th, 81st, 106th, 109th, 112th, 119th, 120th, 121st, 123rd, 127th, 129th, 130th, 131st, 132nd, and 133rd original amino acid residues are replaced by substitution amino acid residues, and/or an additional residue is attached at the carboxy terminus, and wherein X Cys, Ala, and Ala Gin Leu Tyr Thr Thr Glu Asn Arg Val Thr Thr Trp Thr is selected Ser: Pro Thr Ser Leu Gln Leu Gln Met lie Lys Asn Pro Phe Lys Phe Glu Leu Lys Glu Leu Lys Leu Ala Gln Pro Arg Asp Ile Val Leu Thr Phe Met Ala Thr Ile Ile Thr Phe Leu Thr from the group consisting of Ser Ser Glu His Leu Asn Lys Leu Tyr Met His Leu Pro Leu Ser Lys Leu lie Glu Leu Cys Glu Val Glu X Gin Thr Lys Leu Leu Gly Ile Thr Arg Pro Lys Gin Cys Glu Glu Asn Phe Ser Asn Lys Gly Tyr Ala Phe Leu Ser Ile Lys Thr Leu Asp Asn Asn Met Leu Lys Ala Leu Glu Val Leu His Leu Ile Asn Ser Glu Asp Glu Asn Arg Ile Ser The invention also encompasses: IL-2 analogs wherein one, two, three or more original amino acids in any helix or in any helices of IL-2 (advantageously at least two amino acids in helix A) have been replaced by substitution amino acids which maintain or reduce the amphiphilicity of the helix or helices; analogs wherein one, two, three or more original amino acids in helix A, B, B' and/or E, preferably one amino acid in helix E and/or at least two amino acids in helix A, B and/or B' have been replaced by substitution amino acids, each having a different charge from the original amino acid it replaces; and analogs wherein one, two, three or more preferably at least one original amino acid in helix C, D and/or E, and/or at least two original amino acids in helix AB, B' and/or F have been replaced by substitution amino acids, each having a greater preference for alpha-helical structure then the original amino acid it replaces. Also included are IL-2 analogs of formula I above wherein at least one original amino acid in helix A, B' C, D and/or E and/or at least two :original amino acids in helix B and/or F have been replaced by substitution amino acids having a greaiter preference for an alpha-helical structure than the original amino acid replaced, and each of the substitution amino acids altering the amphiphilicity of the helix containing it. The invention also relrtes to manufactured DNA sequences encoding such polypeptides, which genes optionally can be labelled. Further, the WO 90/00565 Pcr/US89/02917 I 5a invention relates to monoclonal antibodies specifically binding such peptides, and to methods of purifying IL-2 and IL-2 analogs.

Preferably, IL-2 analogs of this invention will be polypeptides of formula I, above, which are characterized by the presence of one or more of the following substitutions in amino acid sequence: Lyss Cys 8 47 47 Pro47Gly 7 Thr 5 1 4Asp51; Leuo*Cys 80 Arg 8Cys 80 Glu 0 Lys 106 Asp i9Lys 112 112 20 Ala Lys112 119 119 :Asn 9+Gln 19 120 120 Arg zoLys 2 0 Thr +Ala123 Ser 27 +Ala 127 or Gin 127 129 129 25 Ile 2 9 4Leu1 29 Ser 13oGln13 0 131 131 Thr1 +Ala Leu l 32 Cys32 Thr1 33 +Ala 1 33 addition of Cys134 Z© and optionally one or more of the following substitutions: ~s9 I ~t :C! r~ I C

I

5b Lys 4 8 4Gly48; Lys 5 4 +Glu54; Trp 212Phe 121 Brief Descrintion of the Drawinas

C

C

Figure l(a) represents the alpha carbon backbone of IL-2. Figure l(b) is a schematic stereo drawing of IL-2; helices are represented as cylinders.

Figure 2 is a schematic drawing showing a possible mode of interaction of IL-2 with its receptors.

Figure 3 shows the IL-2 structure and the positions of relevant amino acids.

Detailed Description of the Invention Novel polypeptide analogs of Interleukin II have been discovered. In a first embodiment of the invention, site specific modifications of the proposed receptor binding domains of naturally occurring IL-2 are made, and alterations which stabilize IL-2 ij i i i k jl

J--

I L -c~ PCT/ US89/02917 WO 90/00565 6 /US89/029 helix structure and the overall IL-2 structure, are made. In a second embodiment, an amino acid such as an odd cysteine is incorporated into IL-2 at a location far removed from the proposed receptor binding domains but accessible to chemical reaction with other molecules.

Also provided by the present invention are manufactured genes capable of directing synthesis, in selected microbial hosts bacteria, yeast and mammalian cells in culture), of the above noted IL-2 analogs. In preferred forms of manufactured genes, the base sequence includes one or more codons selected from among alternative codons specifying the same amino acid on the basis of preferential expression characteristics for the codon in a projected host microorganism, e.g., E. coli (see Alton et al., PCT application WO 83/04053).

Other preferred forms of manufactured genes include those wherein there is provided the nucleotide bases for a codon specifying an additional amino acid residue in the polypeptide coded for, which facilitates the direct expression in E. coli organisms an initial Met residue). In still other preferred forms of manufactured genes, the base sequence of codons specifying the desired polypeptide is preceded by and/or followed by and/or includes one or more sequences of bases facilitating formation of expression vectors or generation of new structural genes for polypeptide analogs, sequences of bases providing for selected restriction endonuclease cleavage sites on one or both 30 ends of the structural gene or at intermediate positions therein, and sequences providing a site for ribosome binding, e.g. CAA GGA GGT.

Also provided by the present invention are manufactured genes capable of directing the microbial expression of IL-2 analogs which differ from the naturally-occurring polypeptide in terms of the identity, and/or location of one or more amino acid residues.

W 90/06 T7 WO 90/00565 DyllL----~ WO 90/00565 PCT/US89/02917 7 In the practice of the invention, manufactured DNA sequences are inserted into viral or circular Splasmid DNA vectors to form hybrid vectors and the hybrid vectors are employed to transform microbial hosts such as bacteria E. coli), yeast cells, or mammalian cells in culture. The transformed microorganisms are thereafter grown under appropriate nutrient conditions and express the polypeptide products of the invention.

Also comprehended by the invention are pharmaceutical compositions comprising effective amounts of polypeptide products of the invention together with suitable diluents, adjuvants and/or carriers useful in IL-2 therapy.

As employed herein, the term "manufactured" as applied to a DNA sequence or gene shall designate a product chemically synthesized by assembly of nucleotide bases, synthesized by site-directed mutagenesis, or derived from the biological replication of a product thus synthesized. As such, the term is exclusive of products "synthesized" by cDNA methods or genomic cloning methodologies which involve materials which are of biological origin.

As employed herein the term "substitution amino acid" means an amino acid which replaces the naturally occurring ("original") amino acid, and which is different from the original amino acid.

In another embodiment of the invention, antibodies are provided which specifically bind the polypeptides of the subject invention but which do not cross-react with naturally occurring IL-2, These antibodies can be tagged using methods known to those skilled in the art.

The following abbreviations shall be employed herein to designate amino acids: Alanine, Ala; Arginine, Arg; Asparagine, Asn; Aspartic acid, Asp; a:i: e i I J ii r s

I/

i :a ji i-i a

II

rS

I

N

i i PcT/US89/02917' WO 90/00565 8 Cysteine, Cys; Glutamine, Gin; Glutamic acid, Glu; Glycine, Gly; Histidine, His; Isoleucine, Ile; Leucine, Leu; Lysine, Lys; Methionine, Met; Phenylalanine, Phe; Proline, Pro; Serine, Ser; Threonine, Thr; Tryptophan, Trp; Tyrosine, Tyr; Valine, Val. The following abbreviations shall be employed for nucleotide bases: A for adenine; G for guanine; T for thymine; U for uracil; and C for cytosine.

While not wishing to be constrained to any 10 particular theory of operation of the invention, the following detailed description is presented.

It has now been established that IL-2 is an alpha-helical protein (Fig. Brandhuber et al., Science, 238, 1707 (1987) hereby incorporated by reference. It has a short helical segment near the amino terminus (residues 11 to 19; helix A in Fig. 1), followed by an extended loop; residues 33 to 56 form a helix interrupted, or "bent," near the middle by Pro 47 (hence the two segments are referenced as B and following Cys 58 of the disulfide are helix C, residues 66 to 78, and D, residues 83 to 101; following Cys 105 is a short, apparently helical stretch E, residues 106 to 113, which leads into the carboxyl-terminal helix F, residues 117 to 133. There are no apparent segments of 8-secondary structure in the molecule. The overall helical content of about 65 percent is in good agreement with estimates based on circular dichroism. The disulfide between Cys 58 and Cys 105 links two extending loops that connect the helices across the "top" (in the orientation of Fig. 1) of the molecule.

Helices B, C, D, and F form an antiparallel alpha helical bundle which differs significantly from the classical four-helix bundle represented by cytochrome cytochrome b 562 and myohemerythrin. The packing regions of the helices are shorter, involving only three to four turns of helix, while classical four-

I

WO 90/00565 PCT/US89/02917 9 helix bundles usually have at least five turns in each helix. Further, the packing angles all fall in the range of 250 to 300, and hence are somewhat larger than the average of approximately 180 found in classical four-helix bundles.

Murine IL-2 is expected to have a similar structure to recombinant human IL-2 beginning with helix A and including the proline-induced bend in helix This is significant since recombinant human IL-2 shows activity on both human and murine T cells, and recombinant murine IL-2 is reported to have a low but measurable activity on human T cells. The murine and human IL-2 sequences have 64 percent overall homology.

The amino acid sequence of the mature murine protein is identical to the human sequence for the first seven residues, and then has one or more insertions, a total of 15 amino acids, relative to human, including a 12-residue poly (Gln) stretch, prior to Leu 14 of human IL-2; hence the amino terminal region of the murine protein may have significant structural differences from the human protein up to, and possibly including the first turn of, helix A. The only additional insertion in the murine sequence is between human IL-2 residues and 81, in the loop connecting helices C and D.

The current data on IL-2 receptor binding suggest that the molecule "bridges" two receptor molecules, p55 and p75, with two independent binding sites, when bound to its high affinity receptor.

Earlier work did not presage the presence of two receptor molecules; hence, modifications that affect IL-2 receptor binding do not discriminate between those involving a p55-(IL-2) interaction, a p75-(IL-2) interaction, or both.

Antibodies to peptides that cross-react with IL-2 have been used to map global regions in the IL-2 sequence likely to be important in receptor binding. In

I

WO 90/00565 PCT/US89/0291/ particular, Kuo and Robb have presented evidence suggesting regions within the residue bounds 8 to 27 and 33 to 54 are directly involved in receptor binding, L. Kuo and R. J. Robb, J. Immunol. 137, 1538 (1986), while Altman and colleagues found that antibodies against peptides of residues 59 to 72, 91 to 105, and 119 to 133 did not inhibit IL-2 receptor binding, A. Altman, et al., Proc. Natl. Acad. Sci. U.S.A. 81, 2176 (1984).

Ju et al. supra, have demonstrated that deletion of residues 1 to 10 of human IL-2 (the amino Sterminus to the beginning of helix A) reduces induction of proliferation of murine CTLL-2 cells by only 30 to percent, whereas deletion of residues 1 to 20 (the amino terminus including helix A) abolishes activity completely, Ju et al., supra. Deletion analysis of murine IL-2 shows a similar pattern of effects on proliferation activity of murine HT2 T cells, S. M. Zurawski et al., J. Immunol. 137, 3354 (1986).

Deletion of murine residues 1 to 11 or 1 to 13 (prior to helix A, assuming murine IL-2 is structurally similar to human IL-2) reduces activity by at most 50 percent.

Deletions of the murine poly (Gln) section, residues to 26, coupled with various changes in sequence in the first 37 amino acids, has resulted in mutant protein with as much as one-third the specific activity of the native protein. However, deletion through murine residue 30 (corresponding to human residue 16, in the middle of helix A) reduces activity to about 0.4 percent 30 that of the native protein, and deletion through residue 41 (corresponding to human residue 27) abolishes activity completely, Most of the other reported deletion; that abolish activity many of which would delete a significant fraction of an internal helix in the structure or the peptide connecting them are such that they may disrupt the overall tertiary structure of IL-2.

WO 90/00565 PCT/US89/02917 11 Data on site-specific amino acid substitutions suffer from lack of distinction between those mutations that affect activity by destabilizing the IL-2 structure and those that directly affect receptor binding. Except for alterations that. destroy the disulfide of IL-2 or modify Trp 121 (whose side chain is internal in the structure) all of the mutations shown to lower activity of human IL-2 are in sequence regions 3 to 17 and 36 to 54, Ju et al. supra, which corroborates the receptor binding regions suggested by antibody competition studies. The "down" point mutations, when placed on the IL-2 model, do not identify a specific receptor binding surface.

It is believed that Helices B. C, D, and F form a structural scaffold, and that helices A, B' and part of B, and E form the receptor binding sites of IL-2 (Fig. The involvement of helix E is suggested primarily by its spatial accessibility and its proximity to regions of the molecule probably involved in receptor interactions.

IL-2 binds, through a high affinity receptor, to T-cells but will also bind and activate other immune system cells through lower affinity receptors. It is believed that activation of these other cells contributes to the observed toxic side effects of IL-2.

Structural Variants of Human IL-2 Alterations of the receptor binding domains of IL-2 and alterations which stabilize the IL-2 structure, produce IL-2 analogs of altered specificity towards T-cells and results in an improved IL-2 molecule possessing altered activity and/or toxicity. Included in the subject invention are IL-2 analogs wherein amino acids, advantageously hydrophilic amino acids in helices in the receptor binding domain, are replaced by amino

I

i WO 90/00565 PCT/US89/02917 12 acids having a different charge replacing an amino acid having a positively charged side chain by an amino acid having a negatively charged side chain, or by an amino acid having an uncharged side chain, see below).

Also encompassed by the subject invention are analogs wherein one or more amino acids which have a preference for a-helical structure (see Chou and Fasman, Annu. Rev.

Biochem. 47, 251(1978), have been substituted into one or more of the helices of IL-2, particularly helices A, B, E and F, in order to stabilize the structure of the helix and of the analog as whole. For ease in understanding the present invention, the Chou and Fasman hierarchy is presented below: Preference for Forming a Helix Glu(-) 1.51 Met 1.45 H strong a former Ala 1.42 Leu 1.21 Lys(+) 1.16 Phe 1.13 Gln 1.11 ha a former Trp 1.08 Ile 1.08 Val 1.06 Asp(-) 1.01 Ia weak a former His(+) 1.00 Arg(+) 0.98 Thr 0.83 ia a indifferent Ser 0.77 Cys 0.70 Tyr 0.69 ba a breaker Asn 0.67 Pro 0.57 Ba strong a breaker Gly 0.57 f4 r

_I-

13 The helices of IL-2 are amphiphilic helices (see Kaiser et al., PNAS, 80, 1137-1143(1983) and Kaiser et al., Science, 223, 249-255(1984).

This amphiphilic helical structure is shared by several cytotoxic peptides such as mellitin, pardoxin, and maganins. More specifically, the F helix is very amphiphilic and some of the amino acids in the F helix do not have a strong preference for the a-helical structure, and it is believed that the interaction of the hydrophobic face of the F helix with the C and D helices provides the energy required to maintain the F helix sequence in its helical form. IL-2 analogs have been constructed containing altered helix sequences in which the amino acid replacements were selected to contain 15 residues with a greater preference for a a-helical structure Asn-Gln, Trp-Phe, Ser-Gln) and consequently the helices do not require strong amphiphilic interactions to maintain their helical structure, and thus the amphiphilicity of the helix can be maintained (for example an amino acid having a hydrophilic side chain being changed to a different amino acid having a hydrophilic side chain the three substitutions noted above maintain amphiphilicity) or altered (for example by replacement of an amino acid having a hydrophilic side chain with one having a hydrophobic side chain, or vice versa), by amino acid substitution, such.that the ratio of amino acids having hydrophobic side chains to amino acids having hydrophilic 0 13a side chains in the molecule is altered.

Analogs have been constructed to which other molecules can be covalently attached without damaging activity. These analogs are used to attach toxins, reporter groups, or antiviral or other therapeutic compounds which lead to the development of IL-2 conjugates of therapeutic importance as well as the production of specifically labelled IL-2 for the development of sensitive biological assays. The most convenient way to achieve these site specific conjugations is through the introduction of an odd I 4/ .i oo o* 2 l 1 o o

L

SWO 90/00565 PCT/US89/02917 14 cysteine residue which will present a unique reactive sulphydryl group. Since the naturally occurring odd cysteine (Cys 125), is buried and most likely unreactive, the strategy focuses on the incorporation of a chemically accessible cysteine into (Ala 1 25 )IL-2.

The subject invention includes alterations where an additional amino acid is inserted between existing amino acids as an alternative to replacement of an existing amino acid.

I. Alterations in the Receptor Binding Domain of IL-2 and Alterations which Stabilized the IL-2 Structure Hydrophilic amino acids in alpha helical regions (designated A, B and E in Figures 1, 2 and implicated in high affinity receptor interactions, are among the candidate locations for amino acid changes. Substitution of other amino acids (particularly amino acids which alter the charge on the helix surface which interacts with the receptor) at these sites alter the spectrum of activities of the molecule by affecting receptor binding properties. Such changes also alter the ability of the molecule to interact with receptors on different cell types. Those cells are then more or less susceptable to IL-2 depending on the character of their surface IL-2 receptors. For example, cells bearing only the p75 component of the IL-2 receptor do not respond to an IL-2 species wherein the p 7 5 binding domain has been altered such that the molecule can no longer bind to the receptor in a biologically meaningful way. The introduction of such selectivity to the IL-2 cell stimulation process allows for a greater therapeutic index fc: the material by a reduction in undesirable side effects which results from stimulation of one cell type while at the same time retaining the ability of IL-2 to stimulate an appropriate effector cell i

B

4 WO 90/00565 PCT/US89/029 i 7 1.5 type which, in turn, limits disease). For example, i certain alterations in helices in the receptor binding domain produce IL-2 analogs having a reduced capacity to induce induction of lymphokines such as INF-y, IL-1 and 5 TNF, but having equivalent biological activity, relative to (Ala25)IL-2 or natural IL-2.

For ease in understanding the present invention, the side chains of the following amino acids I are generally considered to be nonpolar (hydrophobic): 1 i0 Ile, Leu, Met, Phe, Pro, Trp, Tyr, and Val; the side chains of the following amino acids are polar (hydrophilic) but uncharged: Ala, Asn, Cys, Gln, Gly, Ser, and Thr; the side chains of the following amino acids are hydrophilic and positively charged: Arg, His, and Lys; and the side chains of the following amino acids are hydrophilic and negatively charged: Asp and Glu.

See also Hopp and Woods, PNAS 78 No. 6, 3824-3828(1981); Kyte and Doolittle, J. Molec. Biol., 157, 105-132(1982); and Parker et al., Biochemistry, 25 5425-5432(1986), A. Alterations of the E Helix The short E helix is involved in receptor binding. Therefore, the E helix is an excellent target for mutations that alter receptor binding. The side groups of three amino acids, Glu 106, Asp 109, and Ala 112, protrude out from the E helix for possible interaction with the receptor. Figure 3 shows the IL-2 structure, the proposed receptor binding domains, and positions of the relevant amino acids. Receptor effector IL-2) interactions are in part mediated by electrostatic interactions between charged groups on

I

t 4WO 90/00565 PCTr/US89/02917 -16 the receptor with oppositely charged groups on the effector. The E heJix has no positively charged amino acid side chains and three negatively charged side chains (Glu 106, Asp 109 and Glu ll0).Alterations of the type of charge in the E helix of IL-2 thus alters binding efficiency. An Asp 109 -Lys 109 substitution alters receptor binding with a single mutation. Asp 109 protrudes from the center of the E helix and is a conserved amino acid. A Glu 106, Asp 109, 110 Ala 112 -Lys 106, Lys 109, Lys 112 substitution radically changes the E helix from a negatively charged I to positively charged surface. Changing either Glu 106 or Asp 109 (acidic residues) to Lys results in a net change in charge of +2 of the E helix. Changing a neutral residue (Ala 112) to Lys gives a change of +1.

Changing all three residues to Lys results in a change in Pharge of These changes can be made individually or together. Further, substitutions can be made with amino acids which have a greater preference for the a-helical structure.

B. Alterations of the B and B' Helices Three amino acids, Lys 48, Thr 51, and Lys 54 protrude out from the B' helix for possilble interaction with receptor. The B' helix has one negatively charged amino acid side chain (Glu 52), and four positively charged side chains (Lys 48, Lys 49, Lys 54, and His 55). A Lys 48, Thr 51, Lys 54 .Glu 48, Asp 51, Glu 54 substitution radically changes the BI helix from a positively charged to negatively charged surface.

Similar changes in charge can be made in the helix.

As in the case with the E helix alterations, these changes can be done individually or together. As with the other helices of IL-2, substitutions can be made with amino acids which have a greater preference for the a-helical structure.

I

I I WO 90/00565 PCT/US89/02917 17 C. Pro 47 -Gly 47 The breakage of the B helix into the B and B' helices is a unique structural feature of IL-2. It is believed that Pro 47 rigidly holds B and B' helices at the optimum angle to allow the B' residues to interact properly with the receptor. A Pro 47 -Gly 47 substitution gives more flexibility to the hinge joint separating the B and B' helices which changes the positioning of the important B' residues and ultimately alters the IL-2-receptor interaction. Substitutions by other amino acids at position 47 are also encompassed by the subject invention.

D. Alterations of the A Helix The A helix is involved in receptor binding.

The A helix has one negatively cnarged amino acid side Schain (Glui15), and one positively charged side chain (His 16). As with the the other helices of IL-2, amino acids which have a preference for the a-helical structure can be substituted to strengthen the structure of the A helix. Alteration of charge in the A helix as in the E, B, and B' helices, is also encompassed by the subject invention. Lastly, amino acid changes which maintain or reduce the amphiphilicity of the A helix are also included in the subject invention.

E. Alteration of the F Helix The F helix is an amphiphilic helix. Some of the amino acids in the F helix do not have a strong preference for the a-helical structure and it is believed that the interaction of the hydrophobic face of the F helix with the C and D helices provides the energy required to maintain the F helix in its helical form.

t WO 90/00565 PC/US89/02917 31 1 WO 90/00565 WO 90/00565 PCT/US89/02917 18 IL-2 analogs have been constructed containing altered F helix sequences in which the amino acid replacements were selected to contain residues with a higher preference for an a-helical structure Asn Gln, Trp Phe. Ser Gln) and consequently the F helix does not require strong amphiphilic interactions to maintain its helical structure, and thus the amphiphilicity of the helix can be maintained or reduced.

II. Alterations in IL-2 Which Allow Attachment of a Ligand I The sites of substitutions which allow Sattachment of a ligand include the carboxyl terminus as well as portions of the molecule with surface exposure chosen such as to minimally perturb the structure of active IL-2. For example, addition, insertion or substitution of a cysteine residue provides a sulphydryl group which can be chemically conjugated to: a) radiolabeled moieties (for assay, imaging). Conjugation of reporter groups to IL-2 allow for the rapid and sensitive detection of the resulting active IL-2 analogs. These conjugations are used as the key component in the development of sensitive IL-2 biological assays; b) enzymatic moieties (for assay, directed therapeutic delivery); c) toxins (for selective cell killing, in vitro or in vivo). Conjugation of IL-2 to cytotoxic 30 agents should direct the toxins to cells which present IL-2 receptors. These conjugations are useful in treating certain leukemias, transplant rejections, autoimmune or altered immune states, or other cell populations of pathological significance; d) drugs (directed therapeutic delivery, e.g.

AZT for AIDS). Cells that are susceptible to HIV infection are, for the most part, the cells that carry *i WO 90/00565 PCT/US89/02917 19 IL-2 receptors. Conjugation of IL-2 to retrovirus inhibitors, such as AZT (a reverse transcriptase inhibitor) direct the inhibitor to the infected cells and provide a mechanism for the internalization of the inhibitor through the IL-2 receptor. These conjugations are useful in treating AIDS and other related diseases; e) antibodies or mitogen3 (selective cell targeting, e.g. helper T cells using OKT 4 or equivalent, and/or selective cell activation to an IL-2 responsive state); or the like.

Examination of the x-ray struct'-re of IL-2 reveals that it is possible to conjugate other molecules to IL-2 either at the carboxy terminus or in the region spanning amino acids 79 to 82, without interfering with the receptor binding domain of the IL-2 molecule (see Figure A free cysteine residue has been incorporated into (Ala 2 5 )IL-2 (an IL-2 analog containing one disulfide bond but no free cysteines) at the carboxy term-ius and/or the 79 to 82 region of IL-2. Incorporation of the free cysteine residue(s) at these positions, accomplished by modifying the recombinant IL-2 gene, allows for the specific chemical conjugation of other molecules to IL-2, by reaction with the free sulphydryl group(s), in a manner that does not affect the binding of IL-2 to its cell surface receptors.

A. Alterations in the Carboxy Terminus Region The carboxy terminus region of IL-2 is not involved in receptor binding and is a good location for the incorporation of an odd cysteine. Leu 132 was chosen because it is close to the C-terminus (next to last amino acid) and it is an unconserved residue when comparing human, bovine, and murine IL-2 sequences. An analog in which cysteine is simply added to the carboxy end of IL-2 ((Cys 134 )IL-2) also accomplishes the goal of incorporating an odd cysteine at the C-terminus.

J ,WO 90/00565 PCT/US89/02917 PCT/USt9/Q2 WO 90/00565 PCT/US9/ -20 S- 20 B. Alterations in the Region Between C and D Helices SThe alignment of human, bovine, and murine IL-2 sequences show that, relative to the human sequence, the bovine and murine sequences contain an insertion between amino acids 80 and 81. The amino acids are part of a four amino acid loop (amino acids to 82) that connects the C and D helices of IL-2. Th observation coupled with the fact that this region of 110 the molecule is far removed from the proposed recepto I binding domains and other Cys residues in the molecul makes this an ideal location for an insertion of an o cysteine residue. Substitutions such as Leu 80 -Cys I are also encompassed by the invention.

4 C. Alterations in the Amino-Terminal Region A Lys 8 -Cys 8 substitution at a nonconserv residue provides a reactive group near the amino terminus.

IV. Active or Competitive IL-2 Fragments 1917 79 is r e, dd ed ;i' Some structural component or combination of components of IL-2 have retained or lost biological properties of intact IL-2 together with the ability to bind IL-2 receptor(s). Such a peptide is useful in place of intact IL-2 or as an aitagonist of its action(s).

Peptides for this application include at least one of the 30 A, B, B' and E helices, for example, the invention includes A and E helical regions which preserve their own internal symmetry as species isolated and apart from the conformational constraints of the intact parent molecule. Such isolated structures bind to components of the IL-2 receptor and either do or do not have IL-2 biological activity. Such structures retain activity on WO 90/00565 PCT/US89/02917 -21 only a subset of IL-2 responsive cells allowing for greater precision in manipulating the IL-2 response.

Such isolated structures have lost biological activity but function as competitive inhibitors of IL-2 binding and are useful in antagonizing physiological states involving a stimulant effect of IL-2. Such structures function to up or down regulate IL-2 receptors and, thereby influence cellular receptivity tu IL-2.

V. Additional Alterations Additional analogs which are encompassed by the present invention include the analogs of IL-2 noted above further characterized by the presence of one or more of the following alterations in the amino acid sequence of naturally-occurring IL-2.

deletion and/or replacement of amino acid residues providing sites of intramolecular folding; deletion of terminal amino acid residues; addition of amino acid residues to terminal amino acid residues; deletion and/or replacement of amino acid residues providing sites of hydrolytic instability under highly acidic conditions; replacement of amino acid residues with glutamine residues; replacement of amino acid residues with phenylalanine residues; deletion and/or replacement of tryptophan residues; deletion and/or replacement of asparagine residues; deletion and/or replacement of cysteine residues; replacement of amino acid residues with serine residues; and I -i PcIr/US89/029i WO 90/00565 22 replacement of amino acid residues with alanine residues.

IL-2 Purification and Removal of Pyrogens IL-2 and IL-2 analogs are very hydrophobic V proteins and as such have a propensity to bind pyrogens which are also hydrophobic. These peptides can be formulated in a stable, monomeric form at acidic pH values. Under these conditions, pyrogens tend to form higher molecular weight aggregates, even though the monomeric molecular weight of pyrogens is comparable to that of IL-2. Thus, manufacturing procedures which fractionate proteins on the basis of size are used to separate monomeric IL-2 from aggregated forms of pyrogens. A suitable procedure for carrying out this step is by ultrafiltration through YM-30 (Amicon) membranes. Repeated dilution and ultrafiltration can be used to enhance the yield of IL-2. Glucose, mannitol, or another bulking agent, can be added as a toxicity modifier and the desired concentration of the IL-2 can be obtained by concentration by ultrafiltration or by dilution with an appropriate buffer. Pyrogens can also be separated from monomeric IL-2 by size exclusion chromatography, e.g. using Sephadex G-75. Detergents laurate, sarcosine, sodium dodecyl sulfate) render this method ineffective since, in the presence of detergents, pyrogens are reduced to lower molecular weight forms and have apparent molecular weights comparable to IL-2. Procedures which might be expected to remove pyrogens from IL-2 solutions, such as ion exchange chromatography and hydrophobic chromatography proved to be ineffective under the conditions examined. The method of the subject invention is easy to scale up and is very cost effective.

i f( 1 I I' I I

I

WO 90/00565 PCT/US89/02917 23 The following examples illustrate practice of the invention in the manufacture of the DNA sequences coding for microbial expression of IL-2 and polypeptide analogs thereof. Also illustrated is the construction of expression vectors for microbial expression of desired polypeptides EXAMPLE 1 Construction of Oligonucleotide Sequences This example is directed to the procedure employed in the synthesis of oligonucleotide sequences employed to manufacture the IL-2 analog genes according to the invention.

Oligonucleotide sequences were synthesized using a four-step procedure with several intermediate washes. Syntheses were performed on Applied Biosystems (ABI) Model 380 automated synthesizers using ABI supplied reagents. Polymer bound dimethoxyltrityl protected nucleoside in support columns was first stripped of its 5'-protecting group (dimethoxyltrityl) using 3% trichloroacetic acid in dichloromethane for one minute. The polymer was then washed with acetonitrile. The was' ed polymer was then rinsed with dry acetonitrile, placed under argon and then treated in the condensation step using tetrazole in acetonitrile with the protected nucleoside phosphorimidite in acetonitrile. This reaction was allowed to proceed for minutes. The reaccants were then removed by filtration. This was followed by capping the unreactEd '-hydroxyl groups using a solution prepared by mixing one part of a mixture containing acetic anhydride, 2,6-lutidine and tetrahydrofuran and one part dimethylaminopyridine in tetrahydrofuran. After one minute the capping solution was removed and the polymer was treated for 1.5 minutes with an oxidizing solution (0.1 M 12 in H 2 0/2,6-lutidine/THF, 1:10:40).

WO90/00565 PCT/US89/02917' 24 This was followed by an acetonitrile rinse. The cycle began again with a trichloroacetic acid/methylene chloride deprotection and was repeated until the desired oligonucleotide sequence was obtained.

The final oligonucleotide chain was treated with fresh concentrated ammonia at room temperature for hours. After decanting the solution from the polymer, the concentrated ammonia solution was heated at 0 C for 16 hours in a sealed tube.

Each oligonucleotide solution was extracted with 1-butanol and ethyl ether and the concentration of each solucion was determined with a spectrophotometer (260nm). 5.0 OD units of each oligonucleotide were dried down for preparative electrophoresis and loaded into a 15% polyacrylamide, 7 molar urea gel. After electrophoresis, the product band was visualized by UV shadowing, cut from the gel, extracted and then desalted on a Sephadex column to yield the purified oligonucleotide.

EXAMPLE 2 Construction of IL-2 Analogs by OligonucleDtide Site-Directed Mutagenesis This example relates to the use of recombinant methods to generate analogs of IL-2. Site directed mutagenesis procedures according to Souza, et al., published PCT Application No. WO 85/00817, published February 28, 1985, were carried out on the DNA sequence shown in Table 1 (which has E. coli preference codons), using the oligonucleotides shown in Table 2.

TABLE 1 MET A, Pro Thr Ser Ser Ser Thr s CTAGAAAAAAC CATGAGGGT IAATAAATA ATG GCT COT ACG AGC TCT TCT ACT AAG t to Lys Thr (Gin Leu Gin Leu Giu His Leu Leu Leul Asp Leu Gin MET Ilie Leu Asn AAA, ACC (ZAG CTG CAA CTG GAA CAT CTG CTT CTT GAG GTG CAA ATG ATG GTG AAC B Gly Ilie Asn Asn Tyr Lys In P-ro Lys Leu Thr Arg MET Leu Tht Phe Lys Phe GGT ATC AAC AAC TAG AAA AAC GGG AAG OTT AGO GGT ATG GTG ACT TTG AAA TTO B' Tyr METI fjp EY Lys Ala flgGlu Leu Lys His Leul Gin Cys Leu Giu Glu Giu TAO ATG CCG iX'Ai AAA GCA AGO GAA GIG AAA GAG CTG GAG TGT CTG GAA GAA GAA c Leu Lys Pro F~eu Glu Glu Val Leu Asn Leu Ala Gin Ser Lys Asn Phel His Leu i CTG AAA CCT CTG GAG GAA GTT TTA AAO CTG GOT CAA TOG AAG AAC TTT CAT CIG T7G A~ Pro J~g Aspteu lie Ser Asn lie Asn Val lie Val Leu Glu Leu Lys Gly OGT CCA CGI GAT CTG ATO AGC A AC ATT AACGOTT ATO GTA OTG GA.& OTT AAA GGG Ser Glu TMh- Thr Phe MET Gys f~Ghz Tyr Ala Asp Giu Thr Ala ThIq lie Val Glu TOT GAA ACT AGO TIC ATG TGO G AA TAT GOA GAG GAG AGO AGO ATO GIG GAA 120 Ah~f~125 120 InCC 130 133 jPhe Leu Asn Au-g Trp Ilie Thr Phe Gys Gin Ser lie lie Ser Thr Leu TMSTOP 4 TTT CTG AAT OGT rGG ATO ACT TTC TGT GAG TOO ATG ATO AGO ACT GIG ACC TAA

STOPI

Z

IL-2 ANALOGS LeuJ 80 _*Cys 80 Lysa *Cys8 Pro4 .Gly 4 Glu 106 *Lys 106 Asp' 9 4Lys 1 09 Ala"1 2 -Lys 112 Lys 48 .Glu 48 Thr 5 As51 Lys 54 *Glu 54 Cys 134 ACG TGG GCT CTT AAA TIC GTC TGC AGC GGT CCA CGA CTA CAT CGA AGA ACC GAA GGA TCC

ACG

CTA

TAC

ATA

CTL

TGG

GLCC

MAG

LIG

TAT

TABLE 2 SEQUENCES 3') GCA ATG AAA GTT CTT GIA MAA CCC ATG GGG MAG A

CTT

AGC

GLA AC Length 24 24 26 TTT GCA CAT TTT TGC ATA TTT TGG TCT GGA GAA AGL CAG ALG MLC GAA CAL CTG TAG LAG GTC

GAA

TTL

LG T

MLC

TGA

CAG

AGA

1 n WO 90/00565 PCT/US89/02917 27 Oligonucleotide site-directed mutagenesis was performed by cloning the IL-2 region from Xbal to BamHl, from expression vector pCFM 536 IL-2 into both M13mpl0 and M13mpll, and the single-stranded phage DNA was isolated as for DNA sequencing. Although pCFM536 (see U.S. Patent 4,710,473 hereby incorporated by reference) was used, any suitable expression vector could have been used. This DNA was mixed with the synthetic deoxynucleotides of Table 2. The DNA in these mixtures was allowed to anneal, by heating to 65 0 C and then slowly cooling to room temperature. The oligomers contained the appropriate base changes from the natural recombinant IL-2 sequence in the middle of their sequences. To the annealed DNAs were added ATP, dATP, dCTP, dGTP, TTP, T4 DNA ligase, and the Klenow fragment of E. coli DNA polymerase I. This reaction allowed the single-stranded primed phage DNAs to convert into covalently closed, double-stranded, circular DNAs. This DNA was transfected directly into E. coli strain JM103 without first purifying the in vitro synthesized double stranded DNA on alkaline sucrose gradients. Many of the plaques from the transfection contained phage DNA with the original recombinant IL-2 sequence, but some contained the IL-2 sequence with the desired base changes. These plaques were identified by lifting plaques onto nitrocellulose filters, and then hybridizing the filters with the synthetic deoxynucleotide end-labeled with ATP (y- 3 2 After hybridizing, the filters were washed at a temperature S& 30 0-3 0 C below the melting temperature of the synthetic e deoxynucleotide and its complementary DNA sequence.

These wash conditions selectively left strong autoradiography signals corresponding to plaques with phage containing the mutated sequence. Positive clones for each analog were confirmed by DNA sequencing, and these were cloned back into pCFM 536 from Xbal to BamHl.

T--C

k WO 90/00565 PCr/US89/02917 A manufactured gene according to Claim 14 PCT/US89/02917' WO 90/00565 28 Cultures of recombinant IL-2 analogs were grown in media cuntaining 10 g tryptone, 5 g yeast I extract, and 5 g NaCI per liter at 30°C with shaking until they reached an A 600 of 0.5 at which point they were rapidly heated to 42 0 C. The flasks were allowed to continue shaking at 42 0 C for three hours. Cells were harvested by centrifugation at 10,000 x G for 20 minutes at 4 0 C. Cell pellets were resuspended at 0.4 g wet weight/ml with 1 mM dithiothreitol (DTT) and were passed twice through a French Pressure Cell at 10,000 psi for minutes at 4 0 C, and the broken cell supernatants were discarded. The pellets were resuspended in 50 mM Tris, mM EDTA, 5 mM DTT, 0.5 M NaCI, 1% sodium deoxycholate (DOC), pH 9.0 at 0,25 g wet weight original pellet/ml and were allowed to mix for 30 minutes at room temperature. These mixtures were centrifuged at 10,000 x G for 15 minutes at 14 0 C and the supernatants were discarded. The pellets were resuspended in H 2 0 at 0.15 g wet weight original pellet/ml and centrifuged at 10,000 x G for 15 minutes at 4°C. The supernatants were discarded and the pellets were solubilized at room temperature in 4% sodium laurate, 50 mM Tris, ethanol, 50mM DTT, pH 8.7 at approximately 20-30 mg protein/ml. The solubilized protein was chromatographed on Sephadex G-75 in 2% sodium laurate, 50 mM Tris, ethanol, pH 8.7. Fractions were analyzed by SDS-PAGE and IL-2 containing fractions of greater than 95% purity were pooled.

Under certain circumstances it was desiraule 2 30 to have the IL-2 analogs essentially free of pyrogenic substances and endotoxins. This was accomplished by further purification of the molecule. The protein was oxidized in the presence of Cu 2 concentrated and chromatographed on Sephadex G-75 equilibrated with 1% laurate/25 mM Tris/ 5% ethanol, pH 8.7. Those fractions containing monomeric forms of IL-2 were pooled and the 1 WO 90/00565 PCT/US89/02917 29 protein was precipitated by addition of an equal volume of ethanol. The pellet was collected by centrifugation, washed with 50% ethanol, and then solubilized in acetic acid. The solution was diluted 50 fold to allow for refolding of the IL-2 analog, concentrated, then diafiltered against a sodium acetate buffer such that the final concentration and pH were 10 mM sodium acetate, pH 4.

EXAMPLE 3 Activity of the IL-2 Analogs This example relates to the activity of analogs generated in Example 2.

1| A number of IL-2 analogs were generated that differ from the native sequence of IL-2 by two amino acids. All analogs tested have Ala at position 125.

(Ala 125 )IL-2 (which differs from the native sequence of IL-2 by one amino acid) was used as a positive control for these experiments. (Asp 51 )IL-2 and (Glu 48 )IL-2 are molecules with amino acid changes in the putative receptor-binding B' domain of IL-2. Specifically, in (Asp 51 )IL-2 the neutral threonine at position 51 was replaced by a negatively charged aspartic acid, and in (Glu 48 )IL-2 the positively charged lysine was replaced by the negatively charged glutamic acid. In (Gly 47 )IL-2, the proline between the B and B' domains was replaced with a glycine. This change was predicted to have significant structural consequences for the 30 molecule that would result in a substantial loss of biological activity. These analogs were tested in several in vitro assays for IL-2 activity the incorporation of 3 H-thymidine into the murine T cell line CTLL-1 or into human peripheral blood leukocytes (hPBL), the generation of lymphokine activated killer cells (LAK cells) from hPBL cultures and for the ability It le r WO 90/00565 30 to induce IFN-gamma, IL-1 and TNF produc' cultures. These experiments were executi partially purified material. It is note separate CTLL-1 assays using highly puri the specific activity of (Ala 1 25 )IL-2 war 7.8 x 106 U/mg, that for (Glu 48 )IL-2 was 7.9 x 106 U/mg, and that for (Asp 51

IL-I

be 6.6 x 106 U/mg. The data from severa.

shown in Table 3 below.

I PCT/ US89/02917' tion in hPBL ed with d that in fied materials, s found to be found to be 2 was found to 1 experiments is mON...

43 TABLE 3

ASSAY

fip_ A13a~n I Ala(125) Asp( 51) 2 Ala(125) Asp(51) Glu( 48) Gly(47) 3 A18(125) -Asp(51) -Glu(48) Gly( 41) 4 Ala(125) Asp(51) Glu (48) G ly(41)

CTLL-I

(finitsmg 1.2xI0 6 20Ox10 6 1.4xI0 6 3x 10 4 3 Hdtr into hPRI (cm LAK cells I FN-g9a IL-lbeta INF 25,987 37,546 25,983 5.788 1.2010 5 1.2010 5 I~ U105 7.4xIO 42x 10 3 47xI0 3 49x 10 3 7.4xIO 97x10 3 128xIO BI103 19XI0 266x1io3 2656x iO3 375 x10 35XI0 3 4,640 5,600 2,080 720 11,400 10,000 6,650 220 WO 90/00565 PC/US89/02917' WO 90/00565 32 Taken together, these results indicate that (Asp 5 1 )IL-2 and (Glu 4 8 )IL-2 were equivalent to (Ala 1 2 5 )IL-2 as T cell mitogens and as inducers of LAK cells. However, (Glu 48 )IL-2 consistently induced less IFN-gamma production from hPBL cultures than did an equal concentration of (Ala 1 25 )IL-2. In two of three experiments, (Asp 51 )IL-2 was less effective than (Ala 1 2 5 )IL-2 in IFN-gamma induction. In one experiment, the production of IL-1 and TNF by the analogs was measured. The induction of these lymphokines by (Asp 51 )IL-2 was equal to that of (Ala 1 2 5 )IL-2 while induction by (Glu 4 8 )IL-2 was approximately 50% as effective. Depending on the assay, (Gly 47 )IL-2 had only 2% to 20% of the biological activity of (Ala 1 2 5 )IL-2 as predicted.

EXAMPLE 4 Alterations in the F-Helix to Include Residues with High Preference for Alpha-Helical Structure The analogs containing altered F-helix sequence were constructed by a two step procedure. The first step involved the introduction of an Eco R1 restriction site at a region in the gene corresponding to the Glu 11 6 Phe 1 1 7 sequence. This was accomplished by site directed mutagenesis using the primer shown below which codes for the desired change in DNA sequence while leaving the encoded amino acid sequence intact.

Mutagenesis Primer for the Introduction of Eco R1 Site Glu 11 6 Phe 117 CGTG GAA TTC CTGAATCGTT 3' Eco R1

I~

ICT/US89/02917 WO 90/00565 33 The presence of the Eco RI site allowed for the excision of the portion of the IL-2 gene coding for the F-helix by digestion with Eco RI and Bam H1. The second step involved the replacement of the excised portion of the gene with the synthetic DNA sequence coding for the altered F-helix amino acid sequences. Using this method, the following IL-2 analogs C4 and C5 were constructed: Natural Sequence 116 120 125 130 GAA TTT CTG AAT CGT TGG ATC ACT TTC TGT CAG TCC ATC ATC AGC ACT CTG ACC Glu Phe Leu Asn Arg Trp Ile Thr Phe Cys Gin Ser Ile Ile Ser Thr Leu Thr Analog C4 (Gln11 9 Lysl20Alal 23 Alal 27 Leul 29 Alal31)IL-2 125 GAA TTC CTG CAG AAA TGG ATC GCT TTC GCA CAG GCT ATC CTG AGC GCA CTG ACC Glu Phe Leu Gin Lys Trp Ile Ala Phe Ala Gin Ala Ile Leu Ser Ala Leu Thr Analog CS (Glnt 9 Lysl 20 Phel 21 Alal 23 G1nl 27 Leul 29 Gl1nl 30 Alal 31 Alal 33 IL-2 125 GAA TTC CTG CAG AAA TTC ATC GCT TTC GCA CAG CAG ATC CTG CAG Glu Phe Leu Gin Lys Phe Ile Ala Phe Ala Gin Gin Ile Leu Gin GCA C'TG GCT Ala Leu Ala Natural Sequence Preferred Structure (A=a helix, B=a-sheet, T=turn) Glu Phe Leu Asn Arg Trp Ile Thr Phe Cys Gin Ser lie lie Ser Thr Leu Thr T T T T B 8 B B B B B B 88 B Analog C4 Structure 125 Glu Phe Leu Gin Lys Trp Ile Ala Phe Ala Gin Ala Ile Leu Ser Ala Leu Thr A A A A A A A A A A A A A A A A A A Analog C5 Structure 125 Glu Phe Leu Gin Lys Phe Ile Ala Phe Ala Gin G1n Ile Leu Gin Ala Leu Ala A A A A A A A A A A A A A A A A A A

I

4t i I WO 90/00565 i PCT/US89/02917' 34 While the present invention has been described in terms of preferred embodiments, it is understood that variations and modifications will occur to those skilled in the art. Therefore, it is intended that the appended claims cover all such equivalent variations which come within the scope of the invention as claimed.

h1 i WO 90/00565 i PCT/US89/02917

Claims (23)

1. A polypeptide product of the expression in a host cell of a manufactured gene, said polypeptide having an amino acid sequence represented by formula [I] below wherein at least one of the 47tc., 51st, 81st, 106th, 109th, 112th, 119th, 120th, 123rd, 127th, 129th, 131st and 133rd original amino acid residue is replaced by a substitution amino acid residue, or wherein at least two of the 8th, 47th, 48th, 51st, 54th, 81st, 106th, 109th, 112th, 119th, 120th, 121st, 123rd, 127th, 129th, 130th, 131st, 132nd and 133rd original amino acid residues are replaced by substitution amino acid residues, and wherein X is selected from the group consisting of Cys, Ala, and Ser: Ala Pro Thr Ser Ser Ser Thr Lys5 Lys Thr Gin Leu Tyr Th r Tb r Glu As n Arg Va I' Th:. Th r Trp Leu Gin Lys Phe Giu Giu I e u Pro Ile Thr Ala Ile Gin Met As n Lys Leu Leu Ala Arg Val1 Phe Thr Thr Leu Ile Pro Phe Ly s Ly s G I n Asp Leu Met I le Phe Glu Leu Lys Ty r His Pro Se r Leu Glu Cys Val1 His Asn Leu Met Leu Leu Lys Ile Leu Glu Glu Gin Leu Gly Th r Pro Gin Glu Asn Ser Lys Tyr Phe Ser Leu Ile Arg Ly s Cy s Glu Phe As n Gly Ala Leu Ile Leu As n Met Lys Leu Val His Ile Se r Asp Asn Ile Asp Asn Leu Ala Glu Leu Leu As n Glu Glu Arg Se r II Thr Leu Thr. WO 90/00565 -36

2. A polypeptide a characterized by the presence following substitutions in am iLys8+Cys Pro 4 7 .Gl Thr 5 1 -Asj L 8 Leu 8 .Cy Argn 1 Cyi Glu 06 ~L' -Asp 0 L A112cL Asn119_G Argl 220 L 19 Th r A SEr27_A I129 L Serl 3 0 .G Thr1 31 .A Leu 13 2 C addition I PC[/US89/02917, ccording to Claim 2 of one or more of the ino acid sequence:

8. y47; p51P 581; 106; ys109; Y1121 ys ln-l.9; Y120, ys212 la1 23 ia1 27 or Gln. 2 7 eu 2 29 in 1 30 Y 232; la 2 33 of Cys134; and optionally one or more of the follo.loing substitutions: Lys 4 g.Gly 4 g Lys 5 4 .G,u 5 4 Trp 1 21 .Phel 21 *1 WVO 90/00565 PCT/1S89/02917 2/4 I ,WO 90/00565 PCT/US89/02917 37 3. A polypeptide according to Claim 2 selected from the group consisting of: (Cys 8 IL-2; (Gly 47 IL-2; (Asp 51 IL-2; (Cys 80 IL-2; (Cys 81 IL-2; (Lys 106 IL-2; (Lys 109 IL-2; (Lys 11 2 IL-2; (Cys 132 IL-2; (Cys 134 IL-2; 4. A polypeptide according to Claim 1 wherein said polypeptide also has one or more of the following alterations in the amino acid sequence: V(a) deletion and/or replacement of amino acid residues providing sites of intramolecular folding; deletion of terminal amino acid residues; addition of amino acid residues to terminal amino acid residues; deletion and/or replacement of amino acid residues providing sites of hydrolytic instability under highly acidic conditions; replacement of amino acid residues with glutamine residues; replaceme.. of amino acid residues with phenylalanine residues; deletion and/or replacement of tryptophan residues; v(h) deletion and/or replacement of asparagine residues; deletion and/or replacement of cysteine residues; replacement of amino acid residues with serine residues; and WO 90/00565 PCT/US89/02917 WO 90/00565 38 replacement of amino acid residues with alanine residues. A polypeptide product according to Claim 1 in which at least one of: the 8th, 80th and 132nd amino acid residues is replaced by a substitution amino acid residue which permits attachment of a ligand, and/or a 134th amino acid residue which permits attachment of a ligand, is added. 6i. A polypeptide as in Claim 5 wherein said substitution amino acid residue and said 134th amino acid residue are Cys. 7. A polypeptide as in Claim 5 coupled via said 8th, 80th, 81st, 132nd or 134th amino acid residue, to a label. 8. A polypeptide as in Claim 5 coupled via said 8th, 80th, 81st, 132nd or 134th amino acid residue, to an er.zymatic moiety.

9. A polypeptide as in Claim 5 coupled via said 8th, 80th, 81st, 132nd or 134th amino acid residue, to a toxin. 'l0. A polypeptide as in Claim 5 coupled via said 8th, 80th, 81st, 132nd or 134th amino acid residue, to a drug. /11. A polypeptide as in Claim 5 coupled via said 8th, 80th, 81st, 132nd or 134th amino acid residue, to an antibody or mitogen.

12. An antibody specifically binding the polypeptide product of Claim 1 but which does not cross react with naturally occurring human IL-2. I I pCTIUS89/0291 7 WO 90/00565 39

13. An antibody as in Claim 12 which is tagged. .14. the synthesis of Claim 1. the synthesis of Claim 2. A manufactured gene capable of directing in a selected host cell of the polypeptide A manufactured gene capable of directing in a selected host cell of the polypeptide v16. A manufactured gene according to Claim 14 wherein the base sequence includes one or more codons, selected from among alternative codons specifying the same amino acid, on the basis of preferential expression characteristics of the codon in a projected host cell. v17. A manufactured gene according to Claim 14 wherein the base sequence includes one or more codons, selected from among alternative codons specifying the same amino acid, on the basis of preferential expression characteristics of the codon in E. coli. f18. A manufactured gene according to Claim 14 wherein base codons specifying the polypeptide include initial and/or terminal codons respectively specifying additional initial and/or terminal amino acids in the polypeptide synthesized.

19. A manufactured gene according to Claim 18 wherein said initial codons specifying additional initial amino acids are codons specifying an initial methionine residue. "A- WO 90/00565 PC/US89/02917 40 A manufactured gene according to Claim 14 wherein the base codons specifying the polypeptide are preceded and/or followed by and/or include a sequence of bases comprising a portion of a base sequence which provides a recognition site for restriction endonuclease enzyme cleavage. /21. A manufactured gene according to Claim 14 wherein the base codons specifying the polypeptide are preceded by a sequence of bases comprising a portion of a base sequence which provides a site for ribosome binding.

22. A manufactured gene according to Claim 21 wherein said ribosome binding site is specified by the sequence 5'-CAA GGA GGT-3'.

23. A manufactured gene according to Claim 14 which is labelled. /24. A biologically functional DNP transformation vector including the manufactured gene of Claims 14 or /25. A transformed cell with a vector V including a manufactured gene of Claims 14 or S26. A method of removing pyrogens from an IL-2 solution containing pyrogens comprising the steps of: adjusting the pH of a detergent free IL-2 S- solution to a pH so that the pyrogens form aggregates of molecular weight greater than the molecular weight of monomeric IL-2; and separating the aggregates from the IL-2 solution by size. rA 41

27. A method as in Claim 26 wherein step comprises repeatedly i) diluting the IL-2 solution and ii) separating the aggregates from the IL-2 solution.

28. A method as in Claim 26 wherein, step comprises separating the aggregates from the IL-2 solution by ultrafiltration or size exclusion chromatography.

29. An IL-2 analog which is the product of the expression in a host cell of a manufactured gene, said analog having one or more of the biological properties of naturally occurring IL-2 and an amino acid sequence represented by formula I below, with X being Cys, Ala or Ser, wherein at least two original amino acids in helix A and/or helix F of the amphiphilic helical structure of said IL-2 have been replaced by substitution amino acids which alter the amphiphilicity of each helix containing a 10 substitution amino acid: a.* oooe a.* a a a a a a a. oo 2_ Ala Gln Leu Tyr 15 Thr Thr Glu Asn Arg Val Thr Pro Thr Leu Gln Gln Met Lys Asn Phe Lys Glu Leu Glu Leu Leu Ala Pro Arg Ile Val Thr Phe Ser Leu Ile Pro Phe Lys Lys Gln Asp Leu Met Ser Glu Leu Lys Tyr His Pro Ser Leu Glu Cys Ser His Asn Leu Met Leu Leu Lys Ile Leu Glu Thr Leu Gly Thr Pro Gln Glu Asn Ser Lys Tyr Lys Lys Thr Leu Leu Asp Ile Asn Asn Arg Met Leu Lys Lys Ala Cys Leu Glu Glu Val Leu Phe His Leu Asn Ile Asn Gly Ser Glu Ala Asp Glu I. Ji I i i ;ai; C 42 Thr Ala Thr Ile Val Glu Phe Leu Asn Arg Trp Ile Thr Phe X Gin Ser Ile Ile Ser Thr Leu Thr. ii *00 Ii 0 0 0 1? I An IL-2 analog as in claim 29 wherein the substitution of amino acids in helix A and/or helix F results in a replacement of an amino acid having a hydrophilic side chain with one having a hydrophobic side chain or an amino acid having a hydrophobic side chain, such that the ratio of amino acids having hydrophobic side chains to amino acids having hydrophilic side chains in the molecule is altered.

31. An IL-2 analog as in Claim 29 wherein said at least two original amino acids are amino acids having hydrophobic side chains and said substitution amino acids are amino acids having hydrophilic side chains.

32. An IL-2 analog as in Claim 29 wherein each of said substitution amino acids has i greater preference for an alpha-helical structure than the original amino acid it replaces.

33. An IL-2 analog which is the product of the expression in a host cell of a manufactured gene, said analog having one or more of the biological properties of naturally occurring IL-2 and an amino acid sequence represented by formula I below, with X being Cys, Ala or Ser, wherein at least one original amino acid in helix E, and/or at least two original amino acids in helix A, B :A and/or B' have been replaced by substitution amino acids -C I 43 which have a different charge than the original amino acid replaced: Ala Gln Leu Tyr Thr Thr Glu Asn Arg Val Thr Thr 15 Trp Pro Leu Gln Lys Phe Glu Glu Leu Pro Ile Thr Ala Ile Thr Gln Met Asn Lys Leu Leu Ala Arg Val Phe Thr Thr Ser Ser Ser Thr Lys Lys Leu Glu His Leu Leu Leu Ile Leu Asn Gly Ile Asn Pro Lys Leu Thr Arg Met Phe Tyr Met Pro Lys Lys Lys His Leu Gln Cys Leu Lys Pro Leu Glu Glu Val Gln Ser Lys Asn Phe His Asp Leu Ile Ser Asn Ile Leu Glu Leu Lys Gly Ser MeP Cys Glu Tyr Ala Asp Ile Val Glu Phe Leu Asn Phe X Gln Ser Ile Ile Thr Asp Asn Leu Ala Glu Leu Leu Asn Glu Glu Arg Ser I j I: 1 1. PA8 .0* t 4 f 0 Ki Thr Leu Thr.

34. An IL-2 analog as in Claim 33 wherein said original amino acids are hydrophilic amino acids. An IL-2 analog as in Claim 33 wherein said original amino acids are charged amino acids.

36. IL-2 analog as in Claim 33 wherein each of said substitution amino acids has a greater preference for an alpha-helical structure than the original amino acid it replaces.

37. An IL-2 analog which is the product of the expression in a host cell of a manufactured gene, said C Ir L i i i irr 43a analog having one or more of the biological properties of naturally occurring IL-2 and an amino acid sequence represented by formula I below, with X being C's, Ala or Ser, wherein at least one original amino acid in helix C, D and/or E, and/or at least two original amino acids in helix A, B, B' and/or F have been replaced by substitution amino acids which have a greater preference for an alpha-helical structure than the o:iginal amino acid replaced: S. S S S @0 Ala Pro Thr Gin Leu Gln Leu Gin Met Tyr Lys Asn Thr Phe Lys 15 Thr Glu Leu Glu Glu Leu Asn Leu Ala Arg Pro Arg Val Ile Val 20 Thr Thr Phe Thr Ala Thr Trp Ile Thr Ser Leu Ile Pro Phe Lys Lys Gin Asp Leu Met Ile Phe Ser Ser Glu His Leu Asn Lys Leu Tyr Met His Leu Pro Leu Ser Lys Leu Ile Glu Leu Cys Glu Val Glu X Gin Thr Leu Gly Thr Pro Gin Glu Asn Ser Lys Tyr Phe Ser Lys Lys Thr Leu Leu Asp Ile Asn Asn Arg Met Leu Lys Lys Ala Cys Leu Glu Glu Val Leu Phe His Leu Asn Ile Asn Gly Ser Glu Ala Asp Glu Leu Asn Arg Ile Ile Ser Thr Leu Thr.

38. An IL-2 analog as in Claim 37 wherein said original amino acids in helix A, B, E and/or F have been replaced by said substitution amino acids, each of said I I 43b substitution amino acids having a greater preference for an alpha-helical structure than the original amino acid it replaces.

39. An IL-2 analog as in claim 37 wherein said at least one amino acid in helix F has been replaced by an amino acid which has a greater preference for an alpha-helical structure. An IL-2 analog as in Claim 37 wherein said substitution amino acids are selected fron the group consisting of: Glu, Met, Ala, Leu, Lys, Phe, Gln, Trp, Ile, Val, Asp, and His.

41. An IL-2 analog which is the product of the expression in a host cell of a manufactured gene, said analog having one or more of the biological properties of naturally occurring IL-2 and an amino acid sequence represented by formula I below, with X being Cys, Ala or Ser, wherein at least one original amino acid in helix A, C, D and/or E, and at least two original amino acids in helix B and/or F have been replaced by substitution amino acids having a greater preference for an alpha-helical structure than the original amino acid replaced, and wherein each of said substitution amino acids alters the amphiphilicity of the helix containing it: f~ 0 0 00 Ala Pro Thr Ser Ser Ser Thr Lys Lys Thr Gln Leu Gln Leu Glu His Leu Leu Leu Asp Leu Gln Met Ile Leu Asn Gly Ile Asn Asn i ii i 1 i Tyr Thr Thr SGlu Asn Arg Val Thr Thr Trp Thr 43c Lys Asn Pro Lys Leu Thr Arg Phe Lys Phe Tyr Met Pro Lys Glu Leu Lys His Leu Gin Cys Glu Leu Lys Pro Leu Glu Glu Leu Ala Gin Ser Lys Asn Phe Pro Arg Asp Leu lie Ser Asn Ile Val Leu Glu Leu Lys Gly Thr Phe Met Cys Glu Tyr Ala Ala Thr lie Val Glu Phe Leu Ile Thr Phe X Gin Ser Ile Leu Thr. Met Leu Lys Ala Leu Glu Val Leu His Leu lie Asn Ser Glu Asp Glu Asn Arg Ile Ser CC C C C. 9 CC C C

42. An IL-2 analog as in Claim 41 wherein if the original amino acid is hydrophilic, the substitution amino acid is selected from the group consisting of: Met, Leu, Phe, Trp, Ile, and Val; and if the original 5 amino acid is hydrophobic the substitution amino acid is selected from the group consisting of: Glu, Ala, Lys, Gin, Asp, and His. 1~ WO 90/00565 44

43. A peptide whic Sexpression in a host cell of peptide being capable of bind including at least one of hel not including helices C, D, a )44. A peptide as i one original amino acid in he replaced by a substitution am preference for an alpha-helic original amino acid it replac A pharmaceutic an effective amount of a poly Claim 1 and a pharmaceuticall Sadjuvant or carrier. i j J- PCY/US89/02917 ;h is the product of the a manufactured gene, said ing the IL-2 receptor and ices A, B, and E, and *nd F. n Claim 43 wherein at least lices A, B, B' and E is lino acid which has greater al structure than the :es. :al composition comprising peptide according to .y acceptable diluent, 1