HK1117055A - Conjugates of insulin-like growth factor-1 and poly(ethylene glycol) - Google Patents

Description

Conjugates of insulin-like growth factor-1 (IGF-I) and polyethylene glycol

The present invention relates to conjugates of insulin-like growth factor-I (IGF-I) with polyethylene glycol (PEG), pharmaceutical compositions comprising such conjugates, and methods of making and using such conjugates.

Background

Alzheimer's Disease (AD) is an increasingly common form of neurodegeneration, accounting for approximately 50% -60% of all cases of dementia over the age of 65. It is estimated that it currently affects 1500 million people worldwide, and its prevalence may increase in the next 20 to 30 years due to the relative increase in aging population among the population. AD is a progressive disorder in which the average period from onset of clinical symptoms to death is about 8.5 years. Pyramidal neuronal death and neuronal synaptic loss in brain regions associated with higher mental functions produce typical symptoms characterized by severe and progressive impairment of cognitive function (Francis, P.T. et al, J.Neurol.Neurosurg: Psychiatry66(1999) 137-. AD is the most common form of senile and presenile dementia in the world and is clinically recognized as a crusted progressive dementia, with which there is a concomitant increase in memory, intellectual dysfunction and language deficits (Merritt, A Textbook of Neurology, 6 th edition, Lea & Febiger, Philadely, pp.484-489, 1979). Neuropathologically, the main hallmark of AD is the presence of two characteristic lesions: amyloid senile plaques and neurofibrillary tangles (NFTs). Although the plaque is deposited outside the neurons, tangles are observed inside the neurons in the brain after death. One of the major components in the amyloid plaque core is the pathologically deposited small amyloid- β -peptide (A β), which is cleaved by secretases from the Amyloid Precursor Protein (APP) (Selkoe, D.J., Physiol.Rev.81(2001) 741-. The 39-43 residue self-aggregating peptide A β (MW-4 kDa) is synthesized as part of a larger APP (110-120kDa) which is a type I membrane-integrated glycoprotein with a large N-terminal extracellular domain, a single transmembrane domain and a short cytoplasmic tail. The most common inference that APP is involved in neuronal cell death in AD is the amyloid hypothesis. This hypothesis postulates that amyloid deposits or partially aggregated soluble a β cause a neurotoxic cascade, resulting in neurodegeneration similar to AD pathology (Selkoe, d.j., physiol.rev.81(2001) 741-.

Insulin-like growth factor I (IGF-I) is a circulating hormone structurally related to insulin. IGF-I has traditionally been considered to be the major mediator of growth hormone action on peripheral tissues. IGF-I consists of 70 amino acids and is also known as somatomedin C and is defined as SwissProt No. P01343. The uses, activities and productions are mentioned in the following documents: for example, le Bouc, Y, et al, FEBS Lett.196(1986) 108-; de Pagter-Holthuizen, P.et al, FEBS Lett.195(1986)179- > 184; sandberg Nordqvist, A.C. et al, Brain Res.mol. Brain Res.12(1992) 275-; steenbergh, P.H. et al, biochem. Biophys. Res. Commun.175(1991) 507-514; tanner, J.M. et al, Acta Endocrinol. (Copenh.) (84 (1977)) 681-696; uthne, K. et al, J.Clin.Endocrinol.Metab.39(1974) 548-554; EP 0123228; EP 0128733; US5,861,373; US5,714,460; EP 0597033; WO 02/32449; WO 93/02695.

The regulation of IGF-I function is quite complex. In circulation, only 0.2% of IGF-I exists in free form, while the majority binds to IGF-binding proteins (IGFBPs), which have very high affinity for IGF and modulate IGF-I function. This factor may be released by mechanisms that release IGF-I, such as proteolytic local release of IGFBP by proteases.

IGF-I plays a paracrine role in the development and maturation of the brain (Werther, G.A. et al, mol. Endocrinol.4(1990) 773-778). In vitro studies have shown that IGF-I is a potent non-selective nutrient for several types of neurons in the CNS (Knuisel, B. et al, J.Neurosci.10(1990) 558. 570; Svrzic, D., and Schubert, D., biochem. Biophys. Res. Commun.172(1990)54-60), including dopaminergic neurons (Knuisel, B. et al, J.Neurosci.10(1990) 558. 570) and oligodendrocytes (Mcrris, F.A., and Dubois-Dalcq, M, J.Neurosci. Res.21(1988) 199. Mcrn.209; Morris, F.A. et al, Proc.Natl.Acad.Sci.USA 83(1986) 826; zernis, R.L., and Morrosci.382. J.1991. 10. Mcrzi. D.)). US5,093,317 mentions that survival of cholinergic neuronal cells is enhanced by administration of IGF-I. IGF-I is further known to stimulate peripheral nerve regeneration (Kanje, M. et al, Brain Res.486(1989)396-398) and to increase ornithine decarboxylase activity (US5,093,317). Methods for treating central nervous system injuries or diseases are mentioned in US5,861,373 and WO 93/02695, which affect glial and/or non-cholinergic neuronal cells primarily by increasing the active concentration of IGF-I and/or its analogs in the central nervous system of a patient. WO0232449 relates to a method of reducing or preventing ischemic injury in the central nervous system of a mammal by nasal administration of a pharmaceutical composition comprising a therapeutically effective amount of IGF-I, or a biologically active variant thereof, to the mammal. IGF-I or a variant thereof in an amount effective to reduce or prevent ischemic injury associated with an ischemic event is absorbed through the nasal cavity of the mammal and transported into the central nervous system. EP0874641 claims the use of IGF-I or IGF-II for the preparation of a medicament for the treatment or prevention of neuronal damage in the central nervous system due to AIDS-related dementia, AD, parkinson's disease, pick's disease, huntington's chorea, hepatic encephalopathy, cortico-basal ganglia syndrome, progressive dementia, familial dementia with spastic paraplegia, progressive supranuclear palsy, multiple sclerosis, Schilder brain sclerosis or acute necrotizing hemorrhagic encephalomyelitis, wherein said medicament is in a form for parenteral administration of said IGF in an effective amount outside the blood-brain barrier or the blood-spinal cord barrier.

The decline in brain and serum levels of free IGF-I is associated with the pathogenesis of sporadic and familial forms of AD. In addition, IGF-I protects neurons from A β -induced neurotoxicity (Niikura, T. et al, J. Neurosci.21(2001) 1902-1910; Dore, S. et al, Proc. Natl. Acad. Sci.USA 94(1997) 4772-4777; Dore, S. et al, Ann. NY Acad. Sci.890(1999) 356-364). Recently, peripheral administration of IGF-I has been shown to reduce brain A β levels in rats and mice (Carro, E. et al, nat. Med.8(2002) 1390-. Furthermore, studies demonstrated that prolonged IGF-I treatment significantly reduced the accumulation of brain amyloid plaques in a transgenic AD mouse model.

These data strongly support the view that IGF-I is able to reduce brain a β levels and alleviate plaque-associated brain dementia by clearing a β from the brain.

Covalent modification of proteins with polyethylene glycol (PEG) has proven to be a useful method for extending the in vivo circulating half-life of proteins (Hershfield, M.S. et al, N.Engl.J.Med.316(1987) 589-596; Meyes, F.J. et al, Clin.Pharmacol.Ther.49(1991) 307-313; Delgado, C et al, Crrag.Rev.Ther.Drug Carrier. 9(1992) 249-304; Katre, Advanced Drug Delivery Systems 10 1993 (91; EP-A0400472; Monfardini, C et al, Bioconjugate. m.6(1995) 62-69; Satake-Ishikawa, R. et al, Struct.17 (Funct. 17; Acad. Cel.C et al, Bioconjugate. 6(1995) 62-69; Cheju. J.42, Scien. J. 1987; Cheju. J. 1994, J. 1987) 14919; Cheugu.31, J. 1987-1994; Cheju. J. 1987, J. 1994).

Other advantages of pegylation are increased solubility and decreased immunogenicity of the protein (Katre, N.V., J.Immunol.144(1990) 209-213). A common method for pegylation of proteins is the use of amino-reactive reagents such as N-hydroxysuccinimide (NHS) -activated polyethylene glycol. Such agents are used to attach polyethylene glycol to proteins at the e-amino group of free primary amino groups, such as the N-terminal a-amino group and lysine residues. However, the main limitation of this approach is that proteins generally contain a large number of lysine residues and thus polyethylene glycol groups are linked to the protein in a non-specific manner at all free e-amino groups, resulting in a heterologous product mixture of randomly pegylated proteins. Therefore, many NHS-pegylated proteins are not suitable for commercial applications due to low specific activity. Inactivation results from covalent modification of one or more lysine residues or the N-terminal amino residue required for biological activity or covalent attachment of polyethylene glycol residues near or at the active site of the protein. For example, modification of human growth hormone with NHS-PEGylation reagent was found to reduce the activity of the protein by more than 10-fold (Clark, R. et al, J.biol.chem.271(1996) 21969-. Human growth hormone contains 9 lysines with the N-terminal amino acid. Some of these lysines are located in protein regions known to be critical for receptor binding (Cunningham, B.C. et al, Science 254(1991) 821-825). In addition, modification of erythropoietin by the use of amino-reactive polyethylene glycol reagents also results in a near complete loss of biological activity (Wojchowski, D.M. et al, Biochim. Biophys. acta 910(1987)224- "232). Covalent modification of interferon- α 2 with amino-reactive pegylation reagents resulted in a loss of biological activity of 40-75% (U.S. Pat. No. 5,382,657). Similar modifications to G-CSF resulted in greater than 60% loss of activity (Tanaka, H. et al, Cancer Res.51(1991)3710-3714) and similar modifications to interleukin-2 resulted in greater than 90% loss of biological activity (Goodson, R.J., and Katre, N.V., Biotechnology 8(1990) 343-346).

WO 94/12219 and WO 95/32003 claim polyethylene glycol conjugates of IGF comprising a PEG and an IGF or cysteine mutant, said PEG being attached to said mutein at the free cysteine of the N-terminal region of the mutein. N-terminally PEGylated IGF-I is described in WO 2004/60300.

Summary of The Invention

The present invention encompasses IGF-I variants characterized by having amino acid changes at amino acids 27, 37, 65 and/or 68 of the wild-type IGF-I amino acid sequence such that one or more of amino acids 37, 65, 68 is lysine (K) and amino acid 27 is a polar amino acid, but not lysine.

Preferably, the amino acid at position 27 is arginine.

Such IGF-I variants are useful as intermediates for the production of lysine-pegylated IGF-I (IGF-I intermediates).

It has surprisingly been found that lysine-pegylated IGF-I variants (amino-reactive pegylated IGF-I variants), preferably pegylated IGF-I variants of 20kDa to 100kDa and especially preferred mono-pegylated IGF-I variants, have superior properties in therapeutic applications.

Another embodiment of the invention are lysine-pegylated IGF-I variants and N-terminally pegylated IGF-I variant compositions of the invention. Preferably in a molar ratio of from 9: 1 to 1: 9. Also preferred is a composition wherein the molar ratio is at least 1: 1 (at least 1 part lysine-pegylated IGF-I variant per 4 parts N-terminally pegylated IGF-I variant), preferably at least 6: 4 (at least 6 parts lysine-pegylated IGF-I variant per 4 parts N-terminally pegylated IGF-I variant).

Preferably, both the lysine-pegylated IGF-I variant and the N-terminally pegylated IGF-I variant are mono-pegylated. Preferably, in the composition, the variants are identical in the lysine-pegylated IGF-I variant and the N-terminally pegylated IGF-I variant. The variant is preferably selected from the group consisting of RRKK, RRKR, RRRK, RKRR. PEG preferably has an average molecular weight of 30-45kDa, especially 30kDa or 40 kDa. Lysine-pegylated IGF-I variants and N-terminally pegylated IGF-I variants exhibit comparable affinities for binding of IGF binding proteins (e.g., BP4 and BP5), but show different activities for IGF-IR phosphorylation.

The present invention provides a conjugate consisting of an IGF-I variant (IGF-I conjugate or conjugate) and a polyethylene glycol group, characterized in that said IGF-I variant has an amino acid change at amino acid position 27, 37, 65 and/or 68 of the wild-type IGF-I amino acid sequence such that one or more of amino acids 37, 65, 68 is/are lysine (K), amino acid 27 is a polar amino acid but not lysine, and said PEG is conjugated to said IGF-I variant via a primary amino group, preferably via a primary amino group of lysine.

Preferably, the polyethylene glycol groups have a total molecular weight of at least 20kDa, more preferably about 20-100kDa and especially preferably 20-80 kDa.

The polyethylene glycol group is conjugated to the IGF-I variant via the primary amino group of the lysine at one or more of amino acids 37, 65, 68 (amino-reactive pegylation) and optionally pegylated at the N-terminal amino acid. The conjugates are preferably mono-or di-pegylated at one or more lysine residues at amino acids 37, 65, 68 and optionally pegylated at the N-terminal amino acid. It is further preferred that the conjugate is mono-pegylated at K65, K68 or K37 or di-pegylated at K65 and K68 and optionally pegylated at the N-terminal amino acid.

Preferably not more than 20% of the pegylated IGF-I variant is additionally pegylated at the N-terminus.

IGF-I variants are named as follows: k65 means that the 65 th amino acid is lysine, R27 means that the 27 th amino acid is arginine, etc. The IGF-I variant with amino acids R27, K37, K65, K68 was designated RKKK. The IGF-I wild type is designated KRKK.

Particularly preferred IGF-I variants and variants in conjugates are: RRKK, RRKR, RRRK, RKRR.

It is furthermore preferred that the (pegylated) IGF-I variant of the invention is one in which at most three, preferably all three, amino acids at the N-terminus are truncated. The corresponding wild-type mutant is called Des (1-3) -IGF-I and lacks the amino acid residues gly, pro and glu from the N-terminus (Kummer, a. et al, int.j. exp. diabetes res.4(2003) 45-57).

The polyethylene glycol groups are linear or branched.

The invention further comprises methods of making the conjugates of the invention using IGF-I intermediates. The process comprises preparing a conjugate comprising an IGF-I variant and one or two polyethylene glycol groups, preferably having a total molecular weight of at least 20kDa, more preferably from about 20 to 100kDa and especially preferably from 20 to 80kDa, by reacting an IGF-I intermediate with activated polyethylene glycol under conditions such that the polyethylene glycol is chemically bound to the IGF-I intermediate via the lysine primary amino group of the IGF-I variant.

The invention further comprises a pharmaceutical composition comprising a conjugate of the invention, preferably a conjugate of the invention with a pharmaceutically acceptable carrier.

The invention further comprises a method of producing a pharmaceutical composition comprising a conjugate of the invention.

The invention further comprises the use of a conjugate of the invention in the manufacture of a medicament for the treatment of AD.

The invention further comprises a method of treating AD, characterized in that a pharmaceutically effective amount of an amino-reactive pegylated IGF-I variant is administered to a patient in need of such treatment, preferably 1-2 times per week.

Detailed Description

As used herein, "pegylated IGF-I variant" or "amino-reactive pegylation" means that an IGF-I variant is covalently bound to one or two polyethylene glycol groups via amino-reactive coupling of one or two lysines of the IGF-I variant molecule. The PEG group is attached to the primary e-amino site of the lysine side chain of the IGF-I variant molecule. It is also possible that pegylation also occurs at the N-terminal alpha-amino group. Due to the synthetic method and the variants used, pegylated IGF-I variants may consist of a mixture of pegylated IGF-I variants at K65, K68 and/or K37 with or without N-terminal pegylation, whereby the pegylation sites may be different on different molecules or essentially homogeneous with respect to the amount of polyethylene glycol side chains per molecule and/or the pegylation sites of the molecules. Preferably, the IGF-I variant is mono-pegylated and/or di-pegylated and is especially purified from an N-terminally pegylated IGF-I variant.

Amino-reactive pegylation, as used herein, refers to a process of randomly linking a polyethylene glycol chain to a primary amino group of lysine of an IGF-I variant by using a reactive (activated) polyethylene glycol, preferably an N-hydroxysuccinimidyl ester, preferably methoxypolyethylene glycol. The coupling reaction links the polyethylene glycol to the reactive primary e-amino group of the lysine residue of IGF-I and optionally the alpha-amino group of the N-terminal amino acid of IGF-I. Conjugation of PEG to such amino groups of proteins is well known in the art.

An overview of such methods is provided, for example, by Veronese, f.m., Biomaterials22(2001) 405-. Conjugation of PEG to primary amino groups of proteins can be carried out by using activated PEG that alkylates primary amino groups as described by Veronese. To carry out such reactions, activated alkylated PEGs such as PEG aldehyde, PEG-trifluoroethanesulfonyl chloride or PEG epoxide may be used. Additional useful reagents are acylated PEGs such as hydroxysuccinimidyl esters of carboxylated PEG or PEGs with the terminal hydroxyl group activated by chloroformates or carbonylimidazole. Another useful PEG reagent is PEG with an amino acid arm. Such agents may comprise so-called branched PEG whereby at least 2 identical or different PEG molecules are linked to each other by a peptide spacer (preferably lysine) and e.g. bound to an IGF-I variant of an activated carboxylate as lysine spacer. Single-N-terminal coupling is also described in adv. drug Deliv. Rev.54(2002)477-485 by Kinstler, O.

"PEG or polyethylene glycol" as used herein means a water-soluble Polymer that is commercially available or can be prepared by ring-opening polymerization of ethylene glycol according to methods well known in the art (Kodera, Y. et al, Progress in Polymer Science 23(1998) 1233-. The term "PEG" is used broadly to include any polyethylene glycol molecule in which the number of ethylene glycol units is at least 460, preferably 460-. The upper value of the PEG unit is limited only by the solubility of the conjugate. PEG larger than PEG comprising 2300 units is not typically used. Preferably, the PEG used in the present invention has a hydroxyl group or a methoxy group (methoxy PEG, mPEG) at one end and is covalently linked to a linker moiety through an ether oxygen bond at the other end. The polymers are linear or branched. Branched PEGs are described, for example, in Veronese, F.M. et al, Journal of Bioactive and Compatible Polymers12(1997)196- "207. Useful PEG reagents are available, for example, from Nektar therapeutics (C.) (www.nektar.com)。

PEG of any molecular weight may be practically required, for example, from about 20 kilodaltons (Da) to 100kDa (n is 460-. The number of repeating units "n" in the PEG is approximate for molecular weights described in daltons. For example, if two PEG molecules are attached to the linker, wherein each PEG molecule has the same molecular weight of 10kDa (n is about 230 each), then the total molecular weight of the PEG on the linker is about 20 kDa. The molecular weight of the PEG attached to the linker may also be different, for example, one PEG molecule may be 5kDa and one PEG molecule may be 15kDa of the two molecules on the linker. Molecular weight always means average molecular weight.

In the examples below, certain preferred reagents for producing amino-reactive pegylated IGF-I variants are described. It is understood that modifications, for example based on the method described by Veronese, F.M., Biomaterials22(2001)405-417, can be made in the operating steps as long as the method can produce a conjugate of the invention.

The presence of up to three possible reactive primary amino groups (up to two lysines and one terminal amino acid) in the target protein results in a series of pegylated IGF-I variant isomers that differ at the point of attachment of the polyethylene glycol chain.

The present invention provides pegylated forms of IGF-I variants with improved properties. Such pegylated IGF-I variant conjugates contain one or two PEG groups, linear or branched, randomly attached thereto, whereby the total molecular weight of all PEG groups in the conjugate is preferably about 20-80 kDa. It will be apparent to those skilled in the art that small derivatized forms from this molecular weight range are possible as long as the pegylated IGF-I variant exhibits activity in reducing the level of Α β peptide in the brain. Furthermore, pegylation of IGF-I variants using PEG with a molecular weight greater than 80kDa resulted in higher bioavailability. However, such activity is expected to decrease with increasing molecular weight due to decreased IGF-I receptor activation and blood brain barrier transport. Therefore, the 20-100kDa PEG molecular weight range must be understood as the optimal range for conjugates of PEG and IGF-I variants for effective treatment of AD patients.

As used herein, "molecular weight" means the average molecular weight of PEG.

The following pegylated forms of IGF-I variants are contemplated embodiments of the conjugates of the invention:

monopegylated IGF-I variants in which the PEG group has a molecular weight of 20-80kDa

(460-1840 PEG units);

dimeric pegylated IGF-I variants with each PEG group having a molecular weight of about 10-50kDa

Quantum (230- > 1150 PEG units);

and mixtures thereof.

Especially preferred is a mono-pegylated IGF-I selected from RRKK, RRKR, RRRK and RKRR, wherein the branched PEG group has a molecular weight of 30-45, preferably 40-45kDa (about 920 PEG units). For example, based on a PEG average molecular weight of 44kDa and a molecular weight of 7.6kDa for IGF-I, such a mono-PEG-IGF-I has a calculated average molecular weight of about 51.6 kDa. Also preferred is mono-pegylated IGF-I selected from RRKK, RRKR, RRRK and RKRR, wherein the PEG has an average molecular weight of 30 or 40 kDa.

"PEG or PEG-based" according to the invention means a residue comprising polyethylene glycol as an essential moiety.

Such PEG may comprise additional chemical groups essential for the binding reaction; which results from the chemical synthesis of the molecule; or a spacer for optimal spacing of the molecular moieties from each other. In addition, such PEG may consist of one or more PEG side chains attached to each other. PEG groups with more than one PEG chain are referred to as multi-arm or branched PEGs. For example, branched PEGs can be prepared by adding polyethylene oxide to various polyols, including glycerol, pentaerythritol, and sorbitol. For example, a four-arm branched PEG can be prepared from pentaerythritol and ethylene oxide. Branched PEGs are generally provided with 2-8 arms and are described, for example, in EP-A0473084 and U.S. Pat. No. 5,932,462. PEG with two PEG side chains (PEG2) linked by a primary amino group of lysine (Monfardini, C et al, Bioconjugate chem.6(1995)62-69) is particularly preferred.

As used herein, "substantially homogeneous" means that the pegylated IGF-I variant molecules produced, contained, or used are only those molecules with one or two attached PEG groups. The preparation may contain small amounts of unreacted (i.e., lacking PEG groups) protein. As determined from peptide mapping and N-terminal sequencing, one example below provides a preparation of at least 90% PEG-IGF-I variant conjugate and up to 5% unreacted protein. Such homogeneous preparations of pegylated IGF-I variants can be isolated and purified by common purification methods, preferably size exclusion chromatography.

As used herein, "monopegylation" means that the IGF-I variant is PEGylated on only one lysine per IGF-I variant molecule, whereby only one PEG group is covalently attached at that site. Such mono-pegylated IGF-I may be additionally pegylated to some extent at the N-terminus. Pure mono-pegylated IGF-I variants (without N-terminal pegylation) comprise at least 80%, preferably 90%, and most preferably mono-pegylated IGF-I variants comprise more than 92% or 92% of the preparation, the remainder being e.g. unreacted (non-pegylated) IGF-I and/or N-terminally pegylated IGF-I variants. Thus, the homogeneity of the mono-pegylated IGF-I variant preparations of the present invention is sufficient to exhibit the advantages of homogeneous preparations, e.g. in pharmaceutical applications. The same applies to the di-pegylated molecules.

"activated PEG or activated PEG reagents" are well known in the current state of the art. Electrophilically activated PEGs such as alkoxybutyric acid succinimidyl esters of polyethylene glycol ("lower alkoxy-PEG-SBA") or alkoxypropionic acid succinimidyl esters of polyethylene glycol ("lower alkoxy-PEG-SPA") or N-hydroxysuccinimide activated PEGs are preferably used. Any conventional method of reacting an activated ester with an amine to form an amide may be used. In the reaction of activated PEG with IGF-I, a typical succinimidyl ester is a leaving group that results in amide formation. The use of succinimidyl esters in the production of protein-bearing conjugates is disclosed in U.S. Pat. No. 5,672,662.

The reaction conditions used have an effect on the relative amounts of the different pegylated IGF-I variants. By controlling the reaction conditions (e.g., ratio of reagents, pH, temperature, protein concentration, reaction time, etc.), the relative amounts of the different pegylated molecules can be varied. Preferably, the reaction is carried out in an aqueous buffer solution comprising 5-15% (v/v) ethanol and 0.5-4% (v/v) ethylene glycol at a pH of 8-10. If a mono-pegylated variant is produced, the protein to PEG ratio is preferably from 1: 0.5 to 1: 2, and if a di-pegylated variant is produced, the protein to PEG ratio is preferably from 1: 2.2 to 1: 5. The reaction temperature and reaction time may be varied according to the knowledge of the person skilled in the art, whereby high temperatures and long reaction times lead to an increased pegylation. If a mono-pegylated variant is produced, it is therefore preferred to carry out at 4 ℃ for up to 30 minutes. When the pegylation reagent was mixed with IGF-I variant in a reaction buffer consisting of 50mM sodium borate, 10% ethanol and 1% di (ethylene glycol) (DEG) at pH about 9.0 at a protein: PEG ratio of 1: 1.5 and the reaction temperature started at 4 ℃, a mixture of mono-, di-and micro-tri-pegylated molecules was generated. When the protein to PEG ratio is about 1: 3, mainly di-pegylated and oligo-pegylated molecules are produced.

IGF-I variant conjugates of the invention were prepared as follows: IGF-I variant conjugates are formed by covalently reacting a primary lysine amino group of an IGF-I variant with a bifunctional reagent to form an intermediate with an amide bond and covalently reacting the intermediate containing the amide bond with an activated polyethylene glycol derivative.

In the above process, the bifunctional reagent is preferably N-succinimidyl-S-acetylthiopropionate or N-succinimidyl-S-acetylthioacetate, and the activated polyethylene glycol derivative is preferably selected from the group consisting of iodo-acetyl-methoxy-PEG, methoxy-PEG-vinylsulfone and methoxy-PEG-maleimide.

Particular preference is given to using N-hydroxysuccinimide-activated branched PEG esters (mPEG2-NHS) with a molecular weight of 40kDa (Monfardini, C et al, Bioconjugate Chem.6(1995) 62-69; Veronese, F.M. et al, J.Bioactive Compatible Polymers12(1997) 197-207; U.S. Pat. No. 5,932,462).

IGF-I variant conjugates may be prepared by amino-reactive covalent attachment ("activation") of a thiol group to an IGF-I variant and coupling the resulting activated IGF-I variant to a polyethylene glycol (PEG) derivative. The first step comprises the modification of lysine NH by IGF-I₂-mercapto group to which group is covalently linked. This activation of IGF-I variants is performed using bifunctional reagents carrying a protected thiol group and an additional reactive group, such as active esters (e.g. succinimidyl esters), anhydrides, sulfonates, carboxylic acids and sulfonic acid halides, respectively. The thiol group is protected with a group known in the art, such as acetyl. These bifunctional reagents are capable of reacting with the epsilon-amino group of lysine by forming an amide bond. The preparation of the bifunctional reagents is well known in the art. Precursors of bifunctional NHS esters are described in DE 3924705, while derivatization of acetomercapto compounds is described in March, J., Advanced organic chemistry (1977) 375-. The bifunctional reagent SATA is commercially available (molecular probes, Eugene, OR, USA and Pierce, Rockford, IL) and is described in Duncan, R.J., anal.biochem.132(1983) 68-73.

For example, the reaction is carried out in an aqueous buffer solution of pH 6.5 to 8.0, for example, 10mM potassium phosphate, 300mM NaCl, pH 7.3. Bifunctional reagents may be added to DMSO. After completion of the reaction, preferably after 30 minutes, the reaction is terminated by addition of lysine. The excess bifunctional reagent may be separated by methods well known in the art, for example by dialysis or column filtration. The average number of thiol groups added to an IGF-I variant can be determined by photometric assays as described, for example, in gracetti, D.R,. and Murray, j.f., arch, biochem, biophysis.119 (1967) 41-49. The above reaction is followed by covalent coupling of an activated polyethylene glycol (PEG) derivative.

Activated PEG derivatives are well known in the art and are described, for example, in Morpurgo, M.et al, J.bioconjugate chem.7(1996)363-368 in the context of PEG-vinyl sulfone. Both straight-chain and branched PEG molecules are suitable for the preparation of the compounds of formula I. Examples of reactive PEG reagents are iodine-acetyl-methoxy-PEG and methoxy-PEG-vinyl sulfone. The use of these iodine-activated substances is well known in the art and is described, for example, in Hermanson, G.T., bioconjugate technology, Academic Press, San Diego (1996) pp.147-148.

Further purification of the compounds of the invention, including isolation of mono-pegylated and/or di-pegylated IGF-I variants and preferably isolation from the N-terminally pegylated forms, can be carried out by methods well known in the art, for example, column chromatography, preferably ion exchange chromatography, especially cation exchange chromatography.

The percentage of mono-PEG conjugates and the ratio of mono-PEG to di-PEG molecules can be controlled by mixing a broader range of fractions around the elution peak to reduce the percentage of mono-PEG in the composition or mixing a narrower range of fractions to increase the percentage of mono-PEG in the composition. About 90% of the mono-PEG conjugates are well balanced between yield and activity. Sometimes, for example, a composition in which at least 95% or at least 98% of the conjugate is a single PEG molecule may be desirable. In one embodiment of the invention, the percentage of mono-pegylated conjugates is between 90% and 98%.

"polar amino acid" as used herein means an amino acid selected from the group consisting of cysteine (C), aspartic acid (D), glutamic acid (E), histidine (H), asparagine (N), glutamine (Q), arginine (R), serine (S) and threonine (T). Lysine is also a polar amino acid, but is not included because lysine is substituted in accordance with the present invention. Arginine is preferably used as the polar amino acid.

Pharmaceutical preparation

The pegylated IGF-I of the invention provides improved circulatory stability, enabling sustained access to the IGF-I receptor throughout the body at low administration intervals.

Pegylated IGF-I variants can be separated as a mixture or as different polyethylenes separated by ion exchange chromatography or size exclusion chromatographyA glycolated species. The compounds of the present invention may be formulated according to methods for preparing pharmaceutical compositions well known to those skilled in the art. To produce such compositions, the pegylated IGF-I variants of the present invention can be admixed with a pharmaceutically acceptable carrier, preferably by dialysis against an aqueous solution containing the desired components of the pharmaceutical composition. For example, such acceptable carriers are described in Remington's Pharmaceutical Sciences, 18^thedition, 1990, Mack Publishing Company, Oslo et al (e.g., 1435-. Typical compositions comprise an effective amount of a substance of the invention, for example about 0.1-100mg/ml, together with an appropriate amount of a carrier. The composition may be administered parenterally. Preferably, the pegylated IGF-I of the invention is administered intraperitoneally, subcutaneously, intravenously, or intranasally.

The pharmaceutical formulations of the present invention may be prepared according to methods well known in the art. Solutions of pegylated IGF-I variants are typically dialyzed against the buffers specified for pharmaceutical compositions and the desired final protein concentration is adjusted by concentration or dilution.

Such pharmaceutical compositions may be for injection or infusion administration, preferably by intraperitoneal, subcutaneous, intravenous or intranasal administration, and comprise an effective amount of a pegylated IGF-I variant together with pharmaceutically acceptable diluents, preservatives, solubilizers, emulsifiers, adjuvants and/or carriers. Such compositions include diluents of various buffer contents (e.g., arginine, acetate, phosphate); pH and ionic strength agents; additives, such as detergents and solubilisers (e.g. Tween)^TM80/Polysorbate, Pluronic^TMF68) (ii) a Antioxidants (e.g., ascorbic acid, sodium metabisulfite); preservative (Timersol)^TMBenzyl alcohol) and bulking substances (e.g., sucrose, mannitol); the materials are incorporated into polymeric compounds, particulate preparations such as polylactic acid, polyglycolic acid, or into liposomes. Such compositions may affect the physical state stability rate of release and clearance of the mono-pegylated IGF-I of the present invention.

Dose and drug concentration

Generally, in standard treatment regimens, patients are treated with a dose ranging from 0.001 to 3mg, preferably 0.01 to 3mg, of pegylated IGF-I variant per kg/day for a period of time, for 1 day to about 30 days or even more. The drug is administered as a single daily subcutaneous or intravenous or intraperitoneal bolus injection or infusion of a pharmaceutical formulation comprising 0.1-100mg pegylated IGF-I/ml. This method of treatment is combined with any standard (e.g., chemotherapy) treatment by administering pegylated IGF-I before, during, or after said standard treatment. This results in improved efficacy compared to single standard treatment.

It was found that only 1 or 2 times a week administration of pegylated IGF-I of the present invention can be successfully treated. Another embodiment of the present invention is thus a method of treating Alzheimer's disease comprising administering to a patient in need thereof a therapeutically effective amount of a pegylated IGF-I of the present invention in a dose range of 0.001-3mg, preferably 0.01-3mg per kg and every 3-8 days, preferably every 7 days. When pegylated IGF-I is used, preferably, mono-pegylated IGF-I is preferably used as a composition of lysine-pegylated IGF-I variants of the invention with an N-terminally pegylated IGF-I variant wherein the molar ratio is at least 1: 1.

Another embodiment of the invention is the use of pegylated IGF-I of the invention for the manufacture of a medicament for the treatment of alzheimer's disease by administering to a patient in need thereof a therapeutically effective amount and 1 or 2 doses in the range of 0.001-3mg, preferably 0.01-3mg pegylated IGF-I variant per kg and every 6-8 days, preferably every 7 days. When pegylated IGF-I is used, preferably, mono-pegylated IGF-I is preferably used as a composition of lysine-pegylated IGF-I variants of the invention with an N-terminally pegylated IGF-I variant wherein the molar ratio is at least 1: 1.

The following examples, references and figures are provided to aid the understanding of the present invention, the exact scope of which is set forth in the appended claims. It is understood that modifications may be made to the operating steps described without departing from the spirit of the invention.

Sequence listing

SEQ ID NO: 1 is the amino acid sequence of human IGF-I (SwissProt P01343, amino acids 1-70: IGF-I; amino acids 71-105: propeptide).

Brief Description of Drawings

FIG. 1: IEC-HPLC of PEG-IGF. The pure positional isomers were separated from the PEGylation mixture using a preparative strong cation exchange column (TOSOH-BIOSEP, SP-5 PW).

5 different peak fractions (numbers 0-4) were isolated and processed for further analysis.

FIG. 2: SDS-PAGE analysis of the single PEG-IGF-I peak. The 5 purified peak fractions (numbers 0-4) were electrophoresed by 4-12% Tris-glycine SDS-PAGE under non-reducing (A) and reducing (B) conditions and proteins in the gel were stained with Coomassie blue. MW, molecular weight markers.

FIG. 3: peptide analysis of mono-pegylated IGF-I peaks 1, 2 and 3. The purified mono-PEGylation peaks 1, 2 and 3 were digested with Asp-N and the peptide fractions were separated by HPLC. The peptide sequence of the pegylated fragment was obtained by degradation of the Edman N-terminal peptide as described. A: HPLC analysis produced 6 different fragments of rhIGF-I peptide at 30-45 min retention time and major fragment 7 (and smallest fragment 7') of the pegylation peak peaked at-70 min retention time. Arrows indicate the major relative changes in peptide fragments in the different peaks compared to rhIGF-I. B: peptide sequence of 6 fragments obtained from Asp-N cleavage of rhIGF-I. Lysine (K) is marked in bold and appears in fragments 3 and 4. Fragment 5 exemplifies the N-terminal peptide. C: PEGylation of the peptide sequence of peptide fragment 7 following Edman degradation. Cysteine and pegylated lysine residues deliver fragments in the peptide sequence.

FIG. 4: hierarchical cluster-like analysis of AD from different global expression profiles of IGF-I and mono-pegylated isomers. The incubation time for IGF-I and pegylated variants was 24 hours; the mouse cell line NIH-3T3 was used. Data analysis conditions are described herein. (A) A down-regulated gene cluster. (B) An up-regulated gene cluster.

FIG. 5: IGF-IR is phosphorylated by rhIGF-I and mono-PEG-IGF-I. Human IGF-IR stably transfected NIH3T3 cells were serum starved overnight and incubated with increasing doses (0.01-10. mu.g/ml) of rhIGF-I or mono-PEG-IGF-I isoforms (Peak 1, 2, 3, Peak mixture, Des (1-3) Peak 3) for 30 minutes. The cells are then processed for western blotting or in-cell analysis as described in the methods. Dose response curves were obtained using rhIGF-I (A), Peak 1(B), Peak 2(C), Peak 3(D), Peak mixture (E) and Des (1-3) Peak 3(F), respectively. Phosphorylation signals were normalized to IGF-IR levels in each well. Data points represent the mean ± SEM of 6 measurements from 3 independent cultures.

FIG. 6: quantitative analysis of IGF-IR phosphorylation in NIH-3T3 cells. Dose response curves obtained from the experiments in example 7 were fitted using single point binding kinetics in order to obtain specific binding (B)_Max) Half maximal binding concentration (EC)₅₀) And Hill coefficient (n)_H). Data represent mean ± SEM of 6 measurements from 3 independent cultures.

FIG. 7: rhIGF-I reduces brain A β in PS2APP mice in vivo. 2-3 month old double transgenic PS2AAP mice (n ═ 10) were treated with rhIGF-I (50. mu.g/kg i.p.) for 2 hours. Cortical extracts were prepared and assessed for APP, C99, a β and control protein (albumin) levels as described in the methods. The C99/APP, a β/APP and C99/a β ratios were calculated from the values normalized to albumin and expressed as% of untreated control and as mean ± SEM. A: representative western blot of 2 month old cortex extracts. Note that a β levels were significantly reduced by rhIGF-I treatment, while other levels were unchanged. B: quantitative analysis of cortical extracts from PS2APP mice: (^＊＊P < 0.01 compared to untreated control).

FIG. 8: rhlGF-I decreased the time course of A β in PS2APP mice in vivo. With rhIGF-I (50. mu.g/kg)I.p.) 2-month old mice were treated for 2,6, or 24 hours. The A β/APP and C99/A β levels were then evaluated and the ratios calculated as described. Data are expressed as% of untreated control and as mean ± SEM. (n-10-13). A: a β/APP ratio at 2,6 and 24 hours post injection; b: C99/Abeta ratios at 2,6 and 24 hours post-injection (^＊＊P < 0.01 compared to untreated control).

FIG. 9: peaks 1-3 reduced a β in AAP and PS2APP mice in vivo. Experiments were performed in a mixed population of single transgenic APP and dual transgenic PS2APP mice. The a β/APP and C99/a β ratios of peaks 1, 2 and 3 together show a direct comparison of their relative effects on a β reduction and clearance (n-8-10). A: the A β/APP ratio represents the effect of peaks 1-3 on brain A β accumulation. B: the C99/A β ratio of peaks 1-3 indicates clearance of A β from the brain ((R))^＊，p＜0.05；^＊＊，p＜0.01；^＊＊＊P < 0.001, compared to untreated control).

FIG. 10: peak 3 decreased the time course of a β in AAP and PS2APP mice in vivo. A mixed population of 2-month old APP and PS2APP mice was treated with peak 3(50 μ g/kg i.p.) for 2,6, 24, 48 or 72 hours. APP, C99, a β and actin levels were then assessed and ratios calculated as described. Data are expressed as% of untreated control and as mean ± SEM.

(n-10-15). A: a β/APP at 6, 24 and 48 hours post-injection; b: C99/Abeta ratios at 6, 24 and 48 hours post-injection (^＊，p＜0.05；^＊＊，p＜0.01；^＊＊＊P < 0.001, all compared to untreated control).

FIG. 11: peaks 1-3 reduced a β in PS2APP mice in vivo. The a β/APP ratios of peaks 1, 2 and 3 together show a direct comparison of their relative effects on a β reduction and clearance (n ═ 3-13);^＊p < 0.05 compared to untreated control).

FIG. 12: peak 3 decreased the time course of a β in PS2APP mice in vivo. 2-month old PS2APP mice were treated with a single injection of peak 3(50 μ g/kg i.p.) for 2,6, 24, 48 or 72 hours.

The A β/APP ratio is then calculated as described. Data are expressed as% of untreated control and as mean ± SEM (n-4-13;^＊，p＜0.05；^＊＊p < 0.01, compared to untreated control).

Examples

Materials and methods

Recombinant human insulin-like growth factor (rhIGF-I) was purchased from PeproTech (Rocky Hill, NJ, USA) by Cell Concepts (Umkirch, Germany) and Des (1-3) -IGF-I was obtained from GroPep (Adelaide, Australia). Polyethylene glycol (PEG) reagent was from Nektar ltd. (SanCarlos, CA, USA). All other chemicals and solvents used in this study had the highest purity available. IGF-I variants can be produced by recombinant means according to the current state of the art, for example by using a combination of site-directed mutagenesis and preferably expression methods in e.coli (e.coli) as described, for example, in US 6,509,443 or US 6,403,764.

PEGylation of IGF-I

Producing IGF-I with PEGylation at a single site (mono-PEG-IGF-I). MonoPEG-IGF-I is prepared by conjugating the lysine epsilon-amino group to an activated branched PEG moiety of molecular weight 40kDa on the surface or N-terminus of the IGF-I molecule. The pegylation reaction mixture contained rhIGF-I and 40kDa PEG-NHS in a 1: 1.5 molar ratio in 50mM sodium borate buffer (10% ethanol and 1% DEG), pH 9.0. The reaction was allowed to proceed at 4 ℃ for 30 minutes. Based on the 44kDa average molecular weight of the PEG moiety used and the 7.6kDa molecular weight of IGF-I, the calculated average molecular weight of mono-PEG-IGF-I is expected to be about 51.6 kDa.

Separation of PEG-IGF-I positional isomers by ion exchange chromatography

Pure mono-PEG-IGF-I isoform was prepared using a purification protocol of a preparative strong cation exchange column (TOSOH-BIOSEP, SP-5PW, internal diameter 30mm and length 75 mm). The buffer system consisted of 7.5mM sodium acetate, 10% ethanol and 1% diethylene glycol adjusted to pH4.5 (buffer A) and 20mM potassium phosphate, 10% ethanol and 1% diethylene glycol adjusted to pH 6.5 (buffer B).

The column was pre-equilibrated with 25% buffer B. After loading the PEG-IGF-I sample, the column was washed with 35% buffer B, followed by a linear gradient up to 65% buffer B to separate the isomers. To elute non-pegylated IGF-I, the system was switched to buffer B with a change in ph.8.0. The flow rate was 8 ml/min and detection was performed at 218 nm. The resulting protein samples were collected manually and stored in aliquots at-20 ℃ for analysis by various protein chemical and biological assays (see below).

The purity of the isolated positional isomers was investigated using an analytical strong cation exchange column (TOSOH-BIOSEP, SP-NPR, particle size 2.5 μm, diameter 4.6mm, length 3.5 cm). For this analytical column we used the same mobile phase as the preparative analytical column, but with reduced flow rate and run time. Based on the absorption at 280nm of the mono-PEG-IGF-I protein fraction by spectrophotometry (E)^1mg/ml ₂₈₀0.584) the protein concentration of the single PEG-IGF-I isomer was determined.

Analysis of MonoPEG-IGF-I purity

Each single PEG-IGF-I isomer was analyzed by 4-12% Tris-glycine SDS-PAGE under non-reducing or reducing conditions. Proteins were fixed and stained using Simple Blue SaveStain (Invitrogen, Basel, Switzerland).

Mass spectrometric identification of mono-PEG-IGF-I isomers

Asp-N was used to cleave the purified mono-PEG-IGF isomer in order to identify 4 possible PEGylation sites at the N-terminus, K27, K65 or K68. The lysis buffer consisted of 100mM Tris/HClpH 8.0 and 0.04. mu.g/. mu.l Asp-N (Roche Diagnostics GmbH, DE). Mu.g of mono-PEG-IGF were incubated with 1. mu.g of Asp-N for 16 hours at 37 ℃ in lysis buffer.

After 16 hours, the reaction mixture was reduced by adding TCEP (10mM) at 37 ℃ for 1.5 hours. The reaction solution was then quenched by the addition of 1/20 volumes of 10% TFA to complete the cleavage reaction. The peptide mixture obtained was analyzed directly by on-line HPLC ESI mass spectrometry, HPLC or stored at-80 ℃.

For ESI-LC-MS analysis, the peptide mixture was separated on an Agilent 1100 HPLC system equipped with a Phenomenex Jupiter C18 reverse phase column (1X 250mm, 5 μm, 300 Å) at a flow rate of 40 μ l/min. The UV signal was also read at 220 nm. The Q-ToFII or LCT mass spectrometer (Micromass) was directly coupled to the HPLC system. ESI-ToF spectra were recorded at 1 scan/sec in the mass range of 200-2000 m/z. The UV and TIC spectra were evaluated and a single peak could be assigned to each peptide in the chromatogram.

For HPLC analysis, the peptide mixture was separated on an Agilent 1100 HPLC system equipped with a Phenomenex Jupiter C18 reverse phase column (1X 250mm, 5 μm, 300 Å) at a flow rate of 40 μ l/min. The UV signal was also read at 220 nm. PEG peptides were shifted to higher acetonitrile concentrations over retention time and they were collected manually and subjected to N-terminal Edman degradation.

HPLC assay conditions (solvent A: 0.1% TFA in water; solvent B: 0.1% TFA in acetonitrile) are as shown in Table 1:

TABLE 1 gradient

Time of day	％B
		0	0
10	0
		30	20
60	28
		70	48
80	100
		85	100
86	0
		96	0

N-terminal Edman degradation sequencing was performed on an Applied Biosystems Proteinsequencer Procise 492 with the recipe System control software and according to the manufacturer's instructions. 20 μ l/fraction collected from HPLC was applied directly to BioBrene Plus^TMconditioned microTFA filter. The filter was dried in an argon atmosphere sealed with a filter cartridge gasket and introduced into a protein sequencer. The polypeptides were subjected to sequence degradation using automated standard procedures. Analysis of the HPLC chromatograms from each degradation cycle using Applied Biosystems data evaluation software 610A revealed the position of each amino acid. Cysteine and modified amino acids, e.g.Pegylated lysine appeared as a gap in the chromatogram.

Oligonucleotide array transcription assay

For in vitro transcriptional profiling of different mono-PEG-IGF-I isoforms, NIH-3T3 cells stably transfected with IGF-IR were serum-starved for 2 hours and incubated with 0.1 or 1. mu.g/ml rhIGF and 1. mu.g/ml of the corresponding mono-PEG-IGF-I isoform or the obtained mixture of non-isolated isoforms in the absence of serum over 24 hours. The cultured cells were then harvested and total cellular RNA was extracted using RNA-BeeTM. 10 μ gRNA in each sample was reverse transcribed, labeled and processed using a commercial kit and according to the supplier's instructions (Invitrogen, Basel, Switzerland; Ambion, Huntingdon, UK). The method for alkaline thermal fragmentation and hybridization conditions for the MOE 430 AGEnChip array are standard procedures provided by the manufacturer (Affymetrix, US).

Fluorescence (cell intensity) of the array was recorded using a confocal laser scanner and the data was analyzed using MAS 5.0 software (Affymetrix, US). The expression level of each gene was calculated as the normalized average difference in fluorescence intensity compared to mismatched oligonucleotide hybridization (expressed as the Average Difference (AD)). Each experiment was performed in triplicate to account for biological variability.

The following two criteria were chosen for selecting differentially expressed genes: i) the mRNA level of the treated cells must be at least 5-fold higher or 5-fold lower than that of the untreated cells; ii) the standard deviation must be significantly less than the absolute change in the mean difference and the confidence level of the calculated gene is set at greater than 97% (p < 0.03).

IGF-IR phosphorylation assay

Human IGF-IR stably transfected NIH-3T3 cells between passages 2-4 were used for these experiments. Cells were cultured in uncoated 24-well or 96-well plates and grown to 70% confluence. The cells were subsequently serum-starved overnight and then incubated with rhIGF-I or the corresponding PEG-IGF-I peak (or mixture of peaks) for 30 minutes. Cells were then lysed in Laemmli buffer for western blotting or fixed with 4% paraformaldehyde for 30 min for intracellular IGF-IR phosphorylation analysis. For Western blot analysis, protein extracts were separated by 10% Bis-Tris SDS-PAGE and blotted on nitrocellulose membranes. The blot was co-incubated with mouse anti-phosphotyrosine (4G10, 1: 1000; Upstate) and rabbit anti-IGF-IR (C-20, 1: 1000; Santa Cruz) primary antibodies and labeled with anti-mouse Alexa680 (1: 10000; Molecular Probes) and anti-rabbit IRDye800 secondary antibodies (1: 10000; Jackson). For intracellular analysis, the fixed cells were blocked and permeabilized with 2% goat serum and Triton X-100 (0.1%) and incubated with the same primary and secondary antibodies. Protein bands were subjected to fluorescence detection using the Odyssey imaging system (Licor Biosciences). The pixel intensities of the protein bands were quantified from the digital images and dose response curves were analyzed using GraphPad Prism software. The phosphotyrosine levels were normalized using IGF-IR values obtained from the same regions of interest in order to obtain real-time IGF-IR activation changes. Experiments were performed in duplicate and repeated 3 times in order to obtain 6 independent study results for each dose. Data are expressed as mean ± SEM.

In vivo experiments Using rhIGF-I and PEG-IGF-I isomers

A mouse model of Alzheimer's disease consisting of single transgenic AAP and double transgenic PS2AAP mice (Richards, J.G. et al, J.Neurosci.23(2003)8989-9003) was used to study the effects of rhIGF-I (and mono-PEG IGF-I) on cerebral amyloidosis, which has recently been demonstrated in other mouse and rat models (Carro, E.et al, Nat. Med.8(2002) 1390-1397). It has been demonstrated that the PS2APP mouse model develops amyloidosis with measurable a β levels at 2 months and onset of plaque deposition at 8 months of age (Richards, j.g. et al, j.neurosci.23(2003) 8989-. Single transgenic APP mice show very similar a β levels at their young age and are therefore included in this group. We analyzed pre-plaque age (2-3 months) to investigate the effect of rhIGF-I and PEG-IGF-I isoforms on soluble brain A β levels. All experiments were performed according to the Swiss animal protection rights and damage to animals was kept to a minimum. Injection solutions in 0.9% NaCl in a solvent below 1% were prepared from stock solutions (mono-PEG-IGF-I isomers) in 1mM HCI (rhIGF-I) or PBS containing 10% glycerol. The control group was injected with 0.9% NaCl. Injections were given intraperitoneally under mild isoflurane anesthesia. Animals were sacrificed under isoflurane anesthesia at various time points (2, 6, 24, 48 or 72 hours) post-injection.

Mice were decapitated and brains were removed for isolation of telencephalon (including the cortex of the hippocampus). Cortical protein extracts were prepared in hypotonic lysis buffer containing 4mM Tris pH 7.4 and 320mM sucrose (both from Fluka) in admixture with protease and phosphatase inhibitors (both from Sigma).

The sample buffer was laemmli (fluka) containing 8M urea. Proteins were separated by 4-12% Bis-TrisSDS-PAGE and blotted on nitrocellulose membranes. Blots were co-incubated with mouse anti-Amyloid Precursor Protein (APP) antibodies (WO-2 clone, 1: 5000; The genetics company), fragment C99 and A β to detect APP and goat anti-actin antibodies (C-11, 1: 5000; Santa Cruz) primary antibodies and labeled with anti-mouse Alexa680 (1: 10000; Molecular Probes) and anti-goat IRDye800 secondary antibodies (1: 10000; Jackson). Protein bands were subjected to fluorescence detection using the Odyssey imaging system (Licor Biosciences). Pixel intensities for protein bands were analyzed from the digital images using GraphPad Prism software. Data are expressed as mean ± SEM.

All values were normalized to actin (or albumin as a control protein) and specific ratios (C99/APP, A β/APP, C99/A β) were calculated. The C99/APP ratio gives information about the activity status of β -secretase and γ -secretase, since C99 is the product of β -secretase and the substrate of γ -secretase; changes in this measurement may indicate a regulatory effect on one of these secretases independent of the late fate of a β. Since the C99/APP level is constant after a particular treatment, the C99/a β ratio can monitor the clearance of a β independent of its production, i.e. increasing with increasing clearance and decreasing with decreasing clearance. In addition, a β was normalized to APP as its production was dependent on transgenic APP expression that was altered between individual mice. Ratio calculations were performed for each individual animal. All data obtained are expressed as% of untreated controls included in each experiment.

Individual experiments using 2-5 animals per dose/time interval were repeated 2-4 times. Statistical differences were assessed by unpaired t-test according to mean ± SEM, where p < 0.05 was considered statistically significant.

Example 1

Chromatographic separation of mono-PEG-IGF-I positional isomers

IGF-I contains 4 amino groups as potential PEGylation sites and there are expected to be 4 possible mono-PEGylated IGF-I (mono-PEG-IGF-I) isomers. Other derivatives are expected to be oligomerized depending on the reaction conditions. Strong cation high pressure liquid chromatography (IEC-HPLC) was developed for the separation of PEG-IGF-I or Des (1-3) -IGF-I isomers based on their local charge differences. The preparative elution profile of 5mg PEG-IGF-I is shown in FIG. 1. The results of this method can be separated into 5 major peaks, 3 with baseline separation and 2 with partial separation. The decrease in baseline absorption to the end of the chromatogram suggests no additional mono-pegylated IGF-I molecule eluted at higher retention times. An additional peak appeared between peaks 3 and 4 due to switching to modified buffer B with a different pH. The purity of each isomer and contamination of other positional isomers in the IEC fractions were estimated by analytical IEC-HPLC. All mono-pegylated peaks had a purity of > 99%.

Example 2

SDS-PAGE analysis of Single PEG-IGF-I isomers

SDS-PAGE was performed under non-reducing and reducing conditions to assess potential unwanted cross-linking of different IGF-I molecules via intermolecular disulfide bonds. Both conditions produced similar results, indicating that there was no significant abnormal cross-linking of the protein (FIG. 2). SDS-PAGE analysis showed peak 0 to have an apparent molecular weight > 100kDa with major detection bands peaks 1-3 at-70 kDa; in addition, peak 4 was detected at-70 kDa, which is the expected size of non-pegylated IGF-I (FIG. 2). From this running profile and retention time obtained in HPLC, we concluded that peak 0 most likely consisted of di-PEG-IGF-I and oligo-PEG-IGF-I. In contrast, peaks 1-3 were named 3 different mono-PEG-IGF-I isomers.

There was a deviation between the predicted molecular weight for mono-PEG-IGF-I (51.6kDa) and the apparent size of-70 kDa; however, the higher apparent molecular weight observed can be explained by the larger hydrodynamic volume of PEG due to water binding and the increase in apparent molecular weight that makes the electrophoretic movement of PEG-IGF-I fairly slow (Foser, S. et al, Pharmacogenomic J.3(2003) 319). The IEC-HPLC and SDS-PAGE experiments together show that the purity of the IEC fractions can be considered sufficiently pure for further characterization.

Single PEG-Des (1-3) -IGF-I was PEGylated and isolated in relative alignment to rhIGF-I and similarly yielded 3 major single PEG-Des (1-3) -IGF-I peaks.

Example 3

Analysis of MonoPEG-IGF-I isomers

Asp-N was separated using HPLC at 30-45 min retention time to cleave 6 independent peptide fragments of rhIGF-I (FIG. 3A, upper panel). After splitting the purified peaks 1-3 (fig. 3A, lower panel), a different distribution of 6 fragments and the appearance of an additional fragment (fragment 7) were observed to elute at a retention time of-70 minutes. For peak 1, specific peptide fragment 4 decreased with increasing fragment 7. For peak 2, we observed a significant reduction in fragment 3 and a slight reduction in fragment 5 and the appearance of major and minor PEG fragments (7 and 7'). Similarly, HPLC analysis of peak 3 produced a significant reduction in fragments 3 and 5 with concomitant appearance of fragments 7 and 7'. FIG. 11B shows the peptide sequence of the corresponding fragment obtained by Asp-N cleavage of rhIGF-I. We analyzed the peptide sequence of the major fragment 7 obtained using pegylation peaks 1, 2 and 3 using Edman N-terminal peptide degradation. Whereby the cysteine and pegylated amino acid delivers a fragment of the peptide sequence. Peak 1 could be clearly mapped to pegylated rhIGF-I at K27 using this analysis (fig. 3C). For peak 2, the major fragment 7 (> 90%) corresponded to rhIGF-I pegylated at K65, since K68 was confirmed to be unmodified lysine (K) by Edman degradation. In contrast, the fraction of peak 3 was pegylation at K68 (gap in sequence), where K65 gave a signal in the Edman degradation HPLC chromatogram (fig. 3C). The minor 7' peak could not be sequenced by Edman degradation, indicating that the N-terminus is not readily reacted, most likely due to pegylation. Collectively, these data indicate that peak 1 consists of the K27 pegylated isomer, peak 2 is IGF-I pegylated at K65 and peak 3 is IGF-I pegylated at K68 and heavily N-terminally pegylated.

Example 4

Transcriptional profiling of rhIGF-I and MonoPEG-IGF-I isoforms

We used DNA microarrays to determine the transcription response of NIH-3T3 cells stably transfected with human IGF-IR for 24 hours treated with rhIGF-I or PEG-IGF-I isoforms. Therefore, we stimulated cells with 0.1 and 1. mu.g/ml IGF-I or 1. mu.g/ml mono-PEG-IGF-I peaks or a mixture of peaks obtained from the PEGylation reaction. We compared the overall transcriptional activity of cells stimulated with control cultures using a commercial chip type (MOE 430A; Affymetrix Inc.) that contained a probe set for approximately 14,000 mouse genes, including all known IGF-I responsive genes.

mRNA abundance is expressed as the Average Difference (AD) between a correctly matched oligonucleotide probe and the corresponding probe with a mismatch at the central position. We only considered genes with greater than 5-fold change and 97% reproducibility (p-value < 0.03) in three identical biological samples. This analysis yielded a total of 162 genes, 86 up-regulated by all IGF-I variants and 76 down-regulated by all IGF-I variants. A general correlation profile of the transcriptional activity of the different mono-PEG-IGF-I isoforms is explained in FIG. 4 in the form of hierarchical clusters of up-and down-regulated genes. Examination of the levels of induction of each gene by IGF-I at 0.1. mu.g/ml and 1.0. mu.g/ml showed that the selected clusters were very similar. Peak 3 produced a similar expression profile to non-pegylated IGF-I at the same concentration and was more efficient than the PEG-IGF mixture and other peaks. Interestingly, peak 2 elicited a similar transcription response to the PEG-IGF mixture. Consistent with biological activity, peak 1 showed the weakest transcription response, indicating that pegylation interfered with receptor interactions.

Examples 5 and 6

In vitro IGF-IR phosphorylation by rhIGF-I and MonoPEG-IGF-I

For in vitro analysis of IGF-IR activation, NIH-3T3 cells stably expressing human IGF-IR were used. After overnight serum starvation, cells were treated with increasing doses of rhIGF-I or the corresponding PEG-IGF-I isoform (0.003-10. mu.g/ml). Western blot analysis of phosphorylated IGF-IR as above and fitting the obtained dose response curves with Single Point binding kinetics, including Hill coefficient (n)_H) (ii) a The quantitative data for the binding curves are shown in the table in fig. 6. EC was derived from dose-response curves of rhIGF-I (FIG. 5A)₅₀6.3nM and occurs almost in an all-or-nothing manner, where n_HWas 2.27 (FIG. 6). In contrast, peak 1 of mono-PEG-IGF-I showed an apparent dose response with unsaturation, n_H0.34 and estimated EC₅₀Was 91.5nM (FIG. 5B, 6). Peaks 2 and 3 (FIGS. 5C and 5D) show similar binding affinities, with EC₅₀Values were 13.4 and 21.5nM, respectively (FIG. 6). In both cases, n_HAre regular, 1.27 and 1.19 respectively. Peak mixture showed similar IGF-IR activation pattern with slight decrease in affinity, EC₅₀N is 28.8nM and regular_HWas 1.28 (FIGS. 5E, 7). Finally, PEG-Des (1-3) -IGF-I Peak 3 has the highest affinity of all PEG isomers, with EC₅₀N is 10.8nM and regular_HWas 1.08 (FIGS. 5F, 7). This data indicates that all but peak 1 specifically activate human IGF-IR with a 2-5 fold loss of affinity compared to rhIGF-I.

Example 7

In vivo A beta reduction of 2-month old PS2APP mice by rhIGF-I

Double transgenic PS2APP mice were used to analyze the short-term effects of rhIGF-I on brain a β accumulation. We treated these mice intraperitoneally with 50 μ g/kg rhIGF-I and analyzed cortical a β 2 hours after injection. Figure 7A shows a western blot of brain extracts from 2 month old mice. While APP, C99 and control protein (albumin) levels were shown to be unchanged by rhIGF-I, a β levels decreased 2 hours after rhIGF-I injection. Quantitative analysis was performed and the corresponding pixel intensity ratios were calculated in order to obtain information about the potential effect of rhIGF-I on a β production. This analysis revealed that the APP/control protein (97.5 ± 5.7% of control) and C99/APP (87.2 ± 9.8% of control) ratios were unchanged, indicating that neither transgenic APP expression nor APP processing was altered 2 hours after treatment with rhIGF-I (fig. 7B). In contrast, a significant decrease in Α β/APP to 68.4 ± 7.1% (p < 0.01) and an increase in C99/Α β to 157.9 ± 16.6% (p < 0.01) of the control indicates that rhIGF-I reduces Α β. Together, these data indicate that treatment of young PS2APP mice with rhIGF-I for 2 hours primarily increases the clearance of a β from the brain.

Example 8

Time course of in vivo A β reduction of rhIGF-I on PS2APP mice

To evaluate the time course of this short-term effect of rhIGF-I on soluble brain a β, young PS2APP mice (2 months of age) without brain plaques were treated by intraperitoneal injection of 50 μ g/kg rhIGF-I and cortical APP, C99, a β and actin levels were measured 2,6 or 24 hours thereafter. The APP/albumin and C99/APP ratios were not significantly altered by rhIGF-I at any of the time points studied. A decrease in A β/APP by rhIGF-I was observed 2 hours post-injection, whereas this effect was absent after 6 and 24 hours (FIG. 8A). Similarly, an increase in a β reduction monitored by the C99/a β ratio was detectable only at 2 hours and disappeared at 6 and 24 hours (fig. 8B). This suggests that the effect of rhIGF-I on A.beta.clearance may be of short duration due to the short half-life of isolated IGF-I in the blood stream.

Example 9

Comparative analysis of Peak 1-3 in vivo Ass clearing efficiency in 2-month-old APP and PS2APP mice

In this example, data for PEG-IGF-I peaks 1-3 are shown together to directly compare the efficiency of A β reduction and A β clearance. Similar to rhIGF-I, APP/actin (actin is used herein as a control protein) or C99/APP ratio was not altered by peak 1, 2, or 3 at any concentration. Peak 3 had the highest efficiency in reducing a β/APP at 6 hours post-i.p. (figure 9A). In contrast, peak 1 was inactive throughout the tested concentration range and peak 2 only had a significant effect on reducing a β/APP at the highest concentration used (500 μ g/kg). Similarly, as shown in figure 9B, peak 3 is only a compound that is active at low doses (15-50 μ g/kg) in increasing the C99/Α β ratio (as representative of increased Α β clearance). Furthermore, in this evaluation, peak 1 was inactive and peak 2 exerted a significant effect only at 500 μ g/kg. Together, these data indicate that peak 3 is the most active mono-PEG-IGF-I isomer in this in vivo experimental paradigm. Figure 11 shows the results in a double transgenic PS2APP only mouse model.

Example 10

Peak 3 time course of in vivo A β reduction in PS2APP mice

Peak 3 at 50 μ g/kg plays a significant role in reducing brain a β levels 6 hours after i.p. injection. To test how long this effect was maintained, we analyzed brain extracts from PS2APP mice treated with 50 μ g/kg peak 3 for 2,6, 24, 48 and 72 hours. No significant change in APP/actin or C99/APP ratios was observed over the entire time period. In contrast, a β/APP levels were significantly reduced at 6, 24 and 48 hours post-i.p. (p < 0.05, p < 0.001 and p < 0.01, respectively), indicating that peak 3 had a β -lowering effect for at least 48 hours (fig. 10A). The C99/Α β ratio representing Α β clearance independent of Α β production (since C99/APP remained constant) increased significantly over a very similar time course, being sufficiently maintained for at least 24 hours post-injection (fig. 10B). Together, these data indicate that peak 3 is able to reduce brain Α β for PS2 APP. Figure 12 shows the results in a double transgenic PS2APP only mouse model.

Example 11

Binding to IGF binding proteins IGFBP4(BP4) and IGFBP5(BP5)

IGFBP4(SwissProt 22692) was identified and cloned by Shimasaki, s., mol. endocrinol.4(1990) 1451-. IGFBP5(SwissProt 2493) was identified and cloned by Kiefer, m.c., biochem. biophysis. res. commun.176(1991) 219-. Both binding proteins are produced recombinantly in E.coli.

For Surface Plasmon Resonance (SPR) analysis of protein interactions, a Biacore 3000 instrument was used. The assay and reaction buffer was HBS-P (10mM HEPES, 150mM NaCl, 0.005% poly-surfactant, ph 7.4) at 25 ℃. All samples were pre-cooled at 12 ℃. Amine coupling of IGFBP4 and IGFBP5 was performed at a concentration of 5. mu.g/ml. Coupling on a CM5 chip produced a loading signal of 700 RUs. Pegylated IGF-I samples (analytes) were injected at 5 concentrations between 1.23nM and 100nM for 5 minutes (binding period) and washed with HBS-P at a flow rate of 50. mu.l/min for 5 minutes. The chip surface was regenerated by a single injection of 100mM HCl for 1 minute.

The sequence of the chip, assay format and injection corresponded to that described in table 2. Data evaluation was performed by using a 1: 1Langmuir binding model.

Table 2:

chip and method for manufacturing the same	Ligands	Analyte	ka(l/Ms)	kd(l/s)	KD(M)
						CM5	BP4	IGF-1	4.0×106	7.2×10-4	1.8×10-10
CM5	BP4	40kDa/NT-RRRK	5.4×104	1.5×10-3	2.8×10-8
						CM5	BP4	40kDa/RRRK	7.1×104	6.8×10-4	9.5×10-9
CM5	BP4	Composition comprising a metal oxide and a metal oxide	6.9×104	5.3×10-4	7.7×10-9
						CM5	BP5	IGF-1	9.6×106	1.6×10-3	1.7×10-10
CM5	BP5	40kDa/NT-RRRK	1.1×105	2.1×10-3	2.0×10-8
						CM5	BP5	40kDa/RRRK	1.5×105	2.0×10-3	1.3×10-8
CM5	BP5	Composition comprising a metal oxide and a metal oxide	1.1×105	2.5×10-3	2.3×10-8

Abbreviations:

40 kDa: 40kDa branched PEG

40 kDa/NT-RRRK: RRRK PEGylated at the N-terminus with 40kDa PEG

40 kDa/RRRK: RRRK PEGylated at K68 lysine using 40kDa PEG

Composition (A): composition of 40kDa/NT-RRRK and 40kDa/RRRK (1: 1)

These results indicate that all pegylated IGF samples effectively bound IGFBP4 and IGFBP5 in a similar range. All pegylated samples showed significantly reduced binding rate constants compared to non-pegylated IGF.

Negative control data (e.g., buffer curve) is subtracted from the sample curve to calibrate baseline drift and fiduciary decline in the system.

Example 12

Ligand induced IGF-IR autophosphorylation

To determine the ability of the polypeptides of the invention to activate IGF-IR and induce phosphorylation of IGF-IR, cells overexpressing IGF-IR were stimulated with the polypeptides of the invention and the IGF-IR phosphorylation status was subsequently analyzed by ELISA.

For ELISA, 96-well streptavidin-coated polystyrene plates (Nunc) were coated with 100. mu.l of anti-human IGF-lR α (0.5mg/ml) monoclonal antibody (diluted 1: 350 in PBST containing 3% BSA). After incubation on a plate shaker for 1 hour at room temperature, the coating solution was removed and the plates were washed three times with 200 μ l PBST/well.

IGF-IR transfected NIH-3T3 cells were cultured in high glucose MEM Dulbecco's medium (DMEM) (PAA, catalog No. A15-771) supplemented with 2mM L-glutamine (Gibco, catalog No. 25030-024) and 0.5% heat-inactivated FCS (PAA, catalog No. A15-771)E15-009). To determine EC₅₀Value, at 37 ℃ and 5% CO₂Under the condition of 1.3X 10⁴Individual cells/well seeded 96 well plates were cultured for 2 days.

After 48 hours of culture using low serum medium, the medium was carefully removed and replaced with a different concentration of the polypeptide of the invention diluted with 50. mu.l of the corresponding medium. At 37 ℃ and 5% CO₂After incubation for 10 min under conditions, the medium was carefully removed by aspiration and 120. mu.l of cold lysis buffer (50mM Tris pH 7.5, 1mM EDTA, 1mM EGTA, 20% glycerol, 1% Triton-X100, 100mM NaF, 1mM NaVO) was added per well₄，Complete^TMProtease inhibitors). The plates were incubated with lysis buffer at 4 ℃ for 15 minutes on a plate shaker and thereafter 100. mu.l of the well contents were transferred to ELISA plates coated with monoclonal antibody against human IGF-I R alpha. The lysate was incubated on a plate shaker for 1 hour at room temperature to allow IGF-IR to bind to the capture antibody and carefully aspirated thereafter. Unbound material was removed by washing three times with 200 μ Ι PBST/well each time.

To detect bound phosphorylated IGF-IR, 100. mu.l of polyclonal IgG rabbit antibody against human IGF-IR α (diluted 1: 12650 in 3% BSA/PBST) was added to each well followed by incubation on a plate shaker at room temperature for an additional 1 hour. The well contents were again carefully removed and each well was washed three times with 200 μ Ι PBST/well. To detect polyclonal rabbit antibodies, 100. mu.l of HRP (CellSignaling Technology Inc. USA) conjugated anti-rabbit IgG polyclonal antibody (diluted 1: 6000 in 3% BSA/PBST) was added to each well. After incubation on a shaker for 1 hour at room temperature, unbound detection antibody was removed by washing the plate 6 times with 200 μ l PBST/well. As a substrate for antibody-conjugated HRP, 100. mu.l of 3,3 '-5, 5' -tetramethylbenzidine was added to each well, followed by further incubation for 0.5 hour at room temperature on a plate shaker.

After using 25. mu.l/well 1M H₂SO₄After the reaction was terminated, the amount was determined by measuring the absorption at a wavelength of 450 nm.

The obtained OD450 values of the samples were converted into percent activation: percent activation ═ sample-min)/(max-min by the following formula with 10nM IGF-I as 100% (max) and no IGF-I as 0% (min) control. The obtained EC₅₀The values (polypeptide concentration at half maximal activation of IGF-1R) are summarized in Table 3.

TABLE 3

Sample (I)	EC50[nM]	“STDEV”
			K68-RRRK 20kDa Linear	2.1	0.2
NT-RRRK 20kDa Linear	29.6	1.9
			Composition 20kDa Linear	7.6
K68-RRRK 30kDa Linear	4.7	0.9
			NT-RRRK 30kDa Linear	44.5	5.0
Composition 30kDa Linear	9.2	1.1
			Composition 40kDa branching	16.4	0.7
K68-RRRK 20kDa branched	1.5	1.1
			Composition 20kDa branching	5.1	0.5
NT-RRRK 40kDa branched	19.9	1.1

For abbreviations see example 11.

Reference to the literature

Bush，A.I.，and Tanzi，R.E.，Proc.Natl.Acad.Sci.USA 99(2002)7317-7319

Carro，E.，et al.，Nat.Med.8(2002)1390-1397

Chamow，S.M.，et al.，Bioconjugate Chem.5(1994)133-140

Clark，R.，et al.，J.Biol.Chem.271(1996)21969-21977

Cunningham，B.C.，et al.，Science 254(1991)821-825

de Pagter-Holthuizen，P.，et al.，FEBS Lett.195(1986)179-184；

DE 39 24 705

Delgado，C.，et al.，Crit.Rev.Ther.Drug Carrier Syst.9(1992)249-304

Dore，S.，et al.，Ann.NY Acad.Sci.890(1999)356-364

Dore，S.，et al.，Proc.Natl.Acad.Sci.USA 94(1997)4772-4777

Duncan，R.J.，Anal.Biochem.132(1983)68-73

EP 0 123 228

EP 0 128 733

EP 0 597 033

EP-A 0 400 472

EP-A 0 473 084

Foser，S.，et al.，Pharmacogenomic J.3(2003)319

Francis，G.E.，et al.，Int.J.Hematol.68(1998)1-18

Francis，P.T.，et al.，J.Neurol.Neurosurg.Psychiatry 66(1999)137-147

Goodson，R.J.，and Katre，N.V.，BioTechnology 8(1990)343-346

Grasetti，D.R，.and Murray，J.F.，Arch.Biochem.Biophys.119(1967)41-49

Hardy，J.，and Selkoe，D.J.，Science 297(2002)353-356

Hermanson，G.T.，Bioconjugate Techniques，Academic Press，San Diego(1996)pp.147-148

Hershfield，M.S.，et al.，N.Engl.J.Med.316(1987)589-596

Inoue，H.，et al.，J.Lab.Clin.Med.124(1994)529-536

Kanje，M.，et al.，Brain Res.486(1989)396-398

Katre，Advanced Drug Delivery Systems 10(1993)91

Katre，N.V.，et al.，Proc.Natl.Acad.Sci.USA 84(1987)1487-1491

Katre，N.V.，J.Immunol.144(1990)209-213

Kiefer，M.C.，Biochem.Biophys.Res.Commun.176(1991)219-225

Kinstler，O.，et al.，Adv.Drug Deliv.Rev.54(2002)477-485

Knusel，B.，et al.，J.Neurosci.10(1990)558-570

Kodera，Y.，et al.，Progress in Polymer Science 23(1998)1233-1271

Kummer，A.，et al.，Int.J.Exp.Diabesity Res.4(2003)45-57

le Bouc，Y.，et al.，FEBS Lett.196(1986)108-112；

March，J.，Advanced Organic Chemistry(1977)375-376

McMorris，F.A.，and Dubois-Dalcq，M.，J.Neurosci.Res.21(1988)199-209

McMorris，F.A.，et al.，Proc.Natl.Acad.Sci.USA 83(1986)822-826

Merritt，A Textbook of Neurology，6th edition，Lea & Febiger，Philadelphia，pp.484-489，1979

Meyers，F.J.，et al.，Clin.Pharmacol.Ther.49(1991)307-313

Monfardini，C.，et al.，Bioconjugate Chem.6(1995)62-69

Morpurgo，M.，et al.，J.Bioconjugate Chem.7(1996).363-368

Mozell，R.L.，and McMorris，F.A.，J.Neurosci.Res.30(1991)382-390

Niikura，T.，et al.，J.Neurosci.21(2001)1902-1910

Remington′s Pharmaceutical Sciences，18^th edition，1990，Mack PublishingCompany，edited by Oslo et al.(e.g.pp.1435-1712

Richards，J.G.，et al.，J.Neurosci.23(2003)8989-9003

Sandberg Nordqvist，A.C.，et al.，Brain Res.Mol.Brain Res.12(1992)275-277

Satake-Ishikawa，R.，et al.，Cell Struct.Funct.17(1992)157-160

Selkoe，D.J.，Physiol.Rev.81(2001)741-766

Shimasaki，S.，Mol.Endocrinol.4(1990)1451-1458

Steenbergh，P.H.，et al.，Biochem.Biophys.Res.Commun.175(1991)507-514；

Svrzic，D.，and Schubert，D.，Biochem.Biophys.Res.Commun.172(1990)54-60

Tanaka，H.，et al.，Cancer Res.51(1991)3710-3714

Tanner，J.M.，et al.，Acta Endocrinol.(Copenh.)84(1977)681-696

Tsutsumi，Y.，et a1.，Jpn.J.Cancer Res.85(1994)9-12

US 5,093,317

US 5,672,662

US 5,714,460

US 5,861,373

US 5,932,462

US 6,403,764

US 6,509,443

Uthne，K.，et al.，J.Clin.Endocrinol.Metab.39(1974)548-554

Veronese，F.M.，Biomaterials 22(2001)405-417

Veronese，F.M.，et al.，Journal of Bioactive and Compatible Polymers 12(1997)196-207

Werther，G.A.，et al.，Mol.Endocrino1.4(1990)773-778

WO 02/32449

WO 2004/60300

WO 93/02695

WO 94/12219

WO 95/32003

Wojchowski，D.M.，et al.，Biochim.Biophys.Acta 91O(1987)224-232

Sequence listing

<110> Fuffmann-Rarosch Co., Ltd (F.Hoffmann-La Roche AG)

<120> conjugates of insulin-like growth factor-1 (IGF-1) and polyethylene glycol

<130> 22904

<150> EP 04030415

<151> 2004-12-22

<160> 1

<170> PatentIn version 3.2

<210> 1

<211> 105

<212> PRT

<213> human (Homo sapiens)

<400> 1

Gly Pro Glu Thr Leu Cys Gly Ala Glu Leu Val Asp Ala Leu Gln Phe

1 5 10 15

Val Cys Gly Asp Arg Gly Phe Tyr Phe Asn Lys Pro Thr Gly Tyr Gly

20 25 30

Ser Ser Ser Arg Arg Ala Pro Gln Thr Gly Ile Val Asp Glu Cys Cys

35 40 45

Phe Arg Ser Cys Asp Leu Arg Arg Leu Glu Met Tyr Cys Ala Pro Leu

50 55 60

Lys Pro Ala Lys Ser Ala Arg Ser Val Arg Ala Gln Arg His Thr Asp

65 70 75 80

Met Pro Lys Thr Gln Lys Glu Val His Leu Lys Asn Ala Ser Arg Gly

85 90 95

Ser Ala Gly Asn Lys Asn Tyr Arg Met

100 105

Claims (19)

1. A conjugate consisting of an insulin-like growth factor-1 (IGF-I) variant and polyethylene glycol (PEG), characterized in that said IGF-I variant has an amino acid change at amino acid positions 27, 37, 65, 68 of the wild-type IGF-I amino acid sequence such that one or more of amino acids 37, 65, 68 is lysine (K), amino acid 27 is a polar amino acid but not lysine, and said PEG is conjugated to said IGF-I variant via a primary amino group.

2. The conjugate of claim 1, characterized in that said PEG has an overall molecular weight of 20-100 kDa.

3. The conjugate of claim 1 or 2, characterized in that the IGF-I variant is additionally pegylated on the N-terminal amino acid.

4. The conjugate of claim 1 or 3, characterized by being mono-pegylated at K65, K68 or K37 or di-pegylated at K65 and K68.

5. The conjugate of claims 1-4, characterized in that said IGF-I variant is R27, R37, K65, K68(RRKK), R27, R37, R65, K68(RRRK), R27, R37, K65, R68(RRKR), R27, K37, R65, R68 (RKRR).

6. The conjugate of claims 1-5 or the IGF-I variant of claim 5, characterized in that up to three N-terminal amino acids are truncated in the IGF-I variant.

7. The conjugate according to claims 1 to 6, characterized in that the polyethylene glycol group is a branched polyethylene glycol group.

8. The conjugate of claims 1-7, characterized in that the polyethylene glycol group has a total molecular weight of 20kDa to 100 kDa.

An IGF-I variant characterized by having amino acid alterations at amino acids 27, 37, 65, 68 of the wild-type IGF-I amino acid sequence such that one or more of amino acids 37, 65, 68 is lysine (K) and amino acid 27 is a polar amino acid, but not lysine.

10. Use of the IGF-I variant according to claim 9 as an intermediate for the production of pegylated IGF-I variants.

11. A process for preparing a conjugate comprising an IGF-I variant and one or two polyethylene glycol groups having an overall molecular weight of about 20 to about 100kDa, comprising reacting an IGF-I intermediate of claim 9 with activated polyethylene glycol under conditions such that said polyethylene glycol is chemically bound to said IGF-I intermediate via the primary lysine amino group of the IGF-I variant.

12. The method of claim 11, characterized in that said polyethylene glycol is additionally chemically bound to said IGF-I intermediate via the N-terminal amino group of the IGF-I variant.

13. A pharmaceutical composition comprising the conjugate of claims 1-8 and a pharmaceutically acceptable carrier.

14. Use of a conjugate according to claims 1-8 for the manufacture of a medicament for the treatment of alzheimer's disease.

15. A method of treating alzheimer's disease comprising administering to a patient in need thereof a therapeutically effective amount of the conjugate of claims 1-8.

16. A composition of a lysine-pegylated IGF-I variant having amino acid alterations at amino acids 27, 37, 65, 68 of the wild-type IGF-I amino acid sequence such that one or more of amino acids 37, 65, 68 is lysine (K), amino acid 27 is a polar amino acid but not lysine, and said PEG is conjugated to said IGF-I variant via a primary amino group, and an N-terminally pegylated IGF-I variant.

17. A pharmaceutical composition comprising a lysine-pegylated IGF-I variant having amino acid alterations at amino acids 27, 37, 65, 68 of the wild-type IGF-I amino acid sequence such that one or more of amino acids 37, 65, 68 is lysine (K), amino acid 27 is a polar amino acid, but not lysine, and said PEG is conjugated to said IGF-I variant via a primary amino group, and an N-terminally pegylated IGF-I variant and a pharmaceutically acceptable carrier.

18. Use of a composition of a lysine-pegylated IGF-I variant having amino acid alterations at amino acids 27, 37, 65, 68 of the wild-type IGF-I amino acid sequence such that one or more of amino acids 37, 65, 68 is lysine (K), amino acid 27 is a polar amino acid but not lysine, and said PEG is conjugated to said IGF-I variant via a primary amino group, and an N-terminally pegylated IGF-I variant for the manufacture of a medicament for the treatment of alzheimer's disease.

19. A method of treating alzheimer's disease comprising administering to a patient in need thereof a therapeutically effective amount of the composition of claim 16 or 17.