HK1038024A - Hydrophobically-modified protein compositions and methods - Google Patents

Description

Hydrophobically modified protein compositions and methods

Background

It is well known that certain proteins may have higher biological activity when linked to other moieties, either by forming multimeric complexes in which the proteins have the opportunity to act in conjunction, or by other changes in the physico-chemical properties of the proteins, such as absorption, biodistribution and half-life of the proteins. Thus, one of the major areas in biotechnology research today involves the development of methods to modify the physico-chemical properties of proteins so that they can be administered in smaller amounts, with fewer side effects, new routes and at lower cost.

For example, the binding affinity (e.g., the affinity of a ligand for its cognate receptor) of any single protein may be low. However, cells normally express hundreds of replicates of a particular surface receptor, and many receptor-ligand interactions occur simultaneously. When many surface molecules are involved in binding, the total effective affinity is higher than the sum of the binding affinities of the individual molecules. Conversely, if ligand proteins are removed from the cell surface and purified or isolated by recombinant DNA techniques for use in, for example, therapy, they act only as monomers, losing the advantage of synergy with many other replicators of the same proteins closely associated on the cell surface. Thus isolated, the low affinity of a protein for its receptor may become a serious drawback to its effectiveness as a therapeutic agent, blocking specific binding pathways, since it must compete with high affinity cell-cell interactions. Effective treatment may require constant administration and/or high doses. However, if a method of forming multimeric forms of the isolated proteins could be found, these drawbacks could potentially be avoided.

Similarly, it may be useful to modify other physico-chemical properties of a biologically active protein, for example, to induce association of the protein with a membrane, thereby localizing it at the site of administration and enhancing its ability to bind or otherwise interact with a particular target. Such changes may also affect the drug profile of the protein.

Several methods have been developed to produce conjugated proteins. Many of these methods are not highly specific, that is, they are not directed to coupling to any specific site on the protein. As a result, conventional coupling agents can attack the functional site or sterically block the active site, inactivating the coupled protein. In addition, the coupled product is oriented such that the active sites do not act synergistically, thereby rendering the product less potent than the monomeric protein alone.

As an additional motivation for finding new methods of protein modification, proteins with N-terminal cystine residues are susceptible to oxidation or other chemical modification, potentially compromising activity or half-life. In addition, some buffers commonly used in protein purification contain components or impurities that can modify the N-terminal cystine. Even when these buffers are avoided, the N-terminal cystine may change over time, possibly due to storage containers or chemicals in the air. As a consequence, the formulation buffer must contain a protective agent, such as dithiothreitol, to prevent cystine modification and/or oxidation. However, protective agents themselves also have significant biological activity, and they therefore have the potential to complicate experimentation and negatively impact the therapeutic application of the formulation.

Thus, there remains a need in the art to develop more efficient methods of obtaining derivatized products, or multimeric forms thereof, to alter the properties of the protein, affecting its stability, potency, pharmacokinetics, and pharmacokinetics.

Brief description of the invention

In one aspect of the invention, we have addressed a method for conveniently making modified forms of biologically active proteins. The methods of the invention can be used to derive multimeric forms of proteins and/or can be used to alter their physico-chemical properties. We have found that modifying a protein (e.g. adding or suspending a hydrophobic group to an existing amino acid or substituting an amino acid with a hydrophobic group) to introduce a hydrophobic group on the protein can increase the biological activity of the protein and/or its stability. For example, the N-terminal cystine may be used as a convenient "target" for attaching hydrophobic groups (e.g., lipids) and thereby modify biologically active proteins.

Alternatively, a hydrophobic group, such as a hedgehog protein, may be attached to the C-terminal residue of the biologically active protein to modify the activity of the protein. Hydrophobic groups may also be appended to internal amino acid residues to enhance protein activity, provided that the modification does not affect protein activity, e.g., the ability of the protein to bind to a receptor or co-receptor, or affect the steric structure of the protein. Preferably, the hydrophobic group is pendant from internal amino acid residues that are on the surface of the protein when the protein is in its native form. The hydrophobic modification of the present invention provides a versatile method for producing proteins with altered physico-chemical properties compared to the unmodified form.

The present invention stems from the following findings: we expressed the full-length Sonic hedgehog protein in insect and mammalian cells, and in addition to having cholesterol at its C-terminus, the mature form of the protein (residues 1-174 in the mature sequence) was derivatized with fatty acid at its N-terminus. It is evident that this form of hedgehog showed a 30-fold increase in potency in vitro experiments compared to soluble, unmodified hedgehog.

Accordingly, one aspect of the present invention is an isolated protein comprising an N-terminal amino acid and a C-terminal amino acid, wherein the protein is selected from the group consisting of: a protein having an N-terminal cystine with at least one hydrophobic group suspended from the cystine; a protein with an N-terminal amino acid that is not cystine appended with a hydrophobic group; and a protein in which a hydrophobic group replaces an N-terminal amino acid. The hydrophobic group may be a hydrophobic peptide or lipid or any other chemical group that is hydrophobic.

The protein may be modified at its N-terminal amino acid, and the N-terminal amino acid is preferably cystine or a functional derivative thereof. The protein may be modified at its C-terminal amino acid or at both the N-terminal amino acid and the C-terminal amino acid, or at least one amino acid within (i.e., intermediate to) the N-terminal and C-terminal amino acids, or a combination of the above changes. The protein may be an extracellular signaling protein, and in a preferred embodiment, the protein is a hedgehog protein, including Sonic, Indian and Desert hedgehog, obtained from vertebrates, more preferably from humans.

Another embodiment is an isolated protein of the form a-Cys- [ Sp ] -B-X, wherein a is a hydrophobic group, Cys is cystine or a functional equivalent thereof, [ Sp ] is an optional spacer peptide sequence, B is a protein comprising a plurality of amino acids, including at least one optional spacer peptide sequence; and X is an optional additional hydrophobic group attached to the protein.

The isolated protein may be an extracellular signaling protein, preferably a hedgehog protein. The protein may be modified with at least one hydrophobic group at least at one other amino acid. In another embodiment, the protein is contacted with a vesicle selected from the group consisting of a cell membrane, a micelle, and a liposome.

Another aspect of the invention is an isolated protein having a C-terminal amino acid and an N-terminal thioproline group formed by the reaction of an aldehyde with the N-terminal cystine of the protein. Yet another aspect of the invention is an isolated protein having a C-terminal amino acid and an N-terminal amide group formed by the reaction of a fatty acid thioester with an N-terminal cystine of the protein. Yet another aspect of the invention is an isolated protein having a C-terminal amino acid and an N-terminal maleimide group formed by reaction of the maleimide group with an N-terminal cystine of the protein. Yet another aspect of the invention is an isolated protein having a C-terminal amino acid and an N-terminal acetamide group. Yet another aspect of the invention is an isolated protein having a C-terminal amino acid and an N-terminal thiomorpholine group.

In these embodiments, the C-terminal amino acid of the protein may be modified with a hydrophobic group. The isolated protein may be an extracellular signaling protein, most preferably a hedgehog protein.

The methods of the invention include methods of producing a multivalent protein complex, comprising the steps of: the hydrophobic group is attached to the N-terminal cystine, or functional equivalent of the N-terminal cystine, of the protein in the presence of the vesicle. The linking step may comprise linking a lipid moiety, the lipid being selected from saturated and unsaturated fatty acids having 2-24 carbon atoms. The protein may be an extracellular signaling protein, preferably a hedgehog protein selected from Sonic, Indian and Desert hedgehog.

Another method of the invention is a method of modifying the physico-chemical properties of a protein comprising the steps of: at least one hydrophobic group is introduced on the N-terminal cystine or functional equivalent of the N-terminal cystine of the protein. The hydrophobic group may be a lipid moiety selected from saturated and unsaturated fatty acids having 2-24 carbon atoms. It may also be a hydrophobic protein. The protein modified using this method may be an extracellular signaling protein, preferably a hedgehog protein selected from Sonic, Indian and Desert hedgehog. Protein complexes made by these methods are also encompassed by the present invention.

Other extracellular signaling proteins besides hedgehog include gelsolin, interferons, interleukins, tumor necrosis factor, single cell colony stimulating factor, granulocyte macrophage colony stimulating factor, erythropoietin, platelet derived growth factor, growth hormone, and insulin.

Another method is a method of modifying proteins having an N-terminal cystine, such as extracellular signal proteins. The method comprises reacting an N-terminal cystine with a fatty acid sulfate to form an amide, wherein such modifications enhance the biological activity of the protein.

Yet another approach is a method of modifying a protein having an N-terminal cystine (such as an extracellular signaling protein) comprising reacting the N-terminal cystine with a maleimide group, wherein such modifications enhance the biological activity of the protein. Other embodiments of the method involve reacting the N-terminal cystine with an aldehyde, acetamide, or thiomorpholine group.

Yet another approach is a method of modifying a protein (e.g., an extracellular signaling protein) that includes suspending a hydrophobic peptide on the protein. The hydrophobic group may be pendant from an amino acid of the protein selected from the group consisting of: an N-terminal amino acid, a C-terminal amino acid, an intermediate amino acid between the N-terminal amino acid and the C-terminal amino acid, and combinations thereof. In one embodiment, the present invention provides hedgehog polypeptides modified with lipophilic groups. In certain embodiments, the hedgehog proteins of the invention may be modified with lipophilic groups at one or more interior sites of the mature treated extracellular region, and may or may not also be derivatized with lipophilic groups at the N-or C-terminal residues of the mature polypeptide. In other embodiments, the polypeptide is modified at the C-terminal residue with a hydrophobic group other than cholesterol. In another embodiment, the polypeptide is modified at the N-terminal residue with a cyclic (preferably polycyclic) lipophilic group. Various combinations of the above embodiments may also be made. The therapeutic method of the present invention is a method of treating a neurological disease in a patient, comprising administering to the patient a hydrophobically modified protein of the present invention.

Brief Description of Drawings

FIG. 1 is a chart depicting the palmitoylated form of Shh. The binding form of human Shh is defined by High Five^TMInsect cells were purified by immunoaffinity and analyzed by SDS-PAGE. The protein was pre-stained with Coomassie blue (lane a, Life technologies 1 grams, Inc, high molecular weight marker; lane b, soluble Shh (0.6. mu.g), lane c, bound Shh (0.6. mu.g), lane d, a mixture of soluble and bound Shh (0.6. mu.g)). The ability of Shh and Ihh (see lane h) to be modified with palmitic acid was analyzed using the cell-free system described in example 2. Soluble form of hedgehog protein (3. mu.g/sample) was combined with murine liver microsomes, ATP, coenzyme A, and³h-palmitic acid was incubated for 1 hour and then palmitoylation was analyzed by SDS-PAGE. The samples shown in lanes e-i were visualized by fluorescence visualization (lanes e, Shh; lane f, des1-10 Shh; lane g, Cys-1 to Ser Shh; lane h, Ihh; lane i, His-tagged Shh), while the samples in lanes j-k were visualized by Coomassie staining (lanes j, Shh; lanes k des1-10 Shh).

FIG. 2 shows the purified Shh analyzed by ESI-MS. Soluble human Shh (A) and bound human Shh (B) were analyzed by ESI-MS on a Micromass Quattro II triple quadrupole mass spectrometer equipped with an electron-jet ion source. All electrospray mass spectral data were obtained and stored in file format and then processed using the Micromass MassLynx data system. See molecular mass spectrometry (mass distribution generated by data system).

FIG. 3 is a reversed phase HPLC analysis of bound Shh. Soluble human Shh (A), by HighFive^TMConjugated human Shh (B) obtained from insect cells, conjugated human Shh (C) obtained from EBNA-293 cells, and cell-associated murine Shh (D) in narrow bore Vydac C₄Reverse phase HPLC analysis was performed on a column (2.1 mm ID. times.250 mm). The column was subjected to a gradient elution with 0-80% acetonitrile in 0.1% trifluoroacetic acid over a period of 30 minutes at a flow rate of 0.25 ml/min and the eluate was monitored at 200-300nm (data at 214 nm) using a light emitting diode array detector. The peak fractions were collected and then further characterized by SDS-PAGE and MS (data summarized in tables 3, 4 and 5).

FIG. 4 is a representation of Shh by LC-MS. Bound human Shh (A) and soluble human Shh (B) were alkylated with 4-vinylpyridine (1. mu.l/100. mu.l sample in 6M guanidine hydrochloride, 1mM EDTA, 100mM Tris HCl pH 8.0), ethanol precipitated and then digested with endoprotease Lys-C in 50mM Tris HCl pH 7.2, 2M urea at a 1: 5 enzyme to protein ratio as described (27). Digests were analyzed on-line using an electron-jet Micromass Quattro ii triple quadrupole mass spectrometer. Scans were performed over the entire range and then processed with a Micromass MassLynx data system (total ion chromatogram for the entire range is shown). Asterisks indicate the position of the N-terminal peptide confirmed by MALDI PSD or N-terminal Edman sequencing.

FIG. 5 is a sequence of N-terminal Shh peptide using MALDI PSD measurement. Voyager-DE of bound N-terminal endoprotease Lys-C peptide in human Shh in flight Mass Spectroscopy^TMMALDI PSD measurements were performed at STR time. The predicted fragmentation pattern and nomenclature (PA, palmitic acid; 4vp, 4-pyridylethyl) of the fragment ions detected are shown at the upper end of the figure. The rest of the figure shows the effect ofThe resulting molecular mass spectra were examined. The relevant ions are indicated by schematic nomenclature. b₁-b₈Are 447.3, 504.3, 601.4, 658.4, 814.5, 871.5, 1018.6 and 1075.6, respectively. For y₁-y₈Masses (Da) were 147.1, 204.1, 351.2, 408.2, 564.3, 621.3, 718.4, and 775.4, respectively. z is a radical of₈The calculated mass of (A) is 758.4 Da. b₈Due to the additional 18Da contained in the added water.

FIG. 6 is the increase in activity of bound Shh in the C3H10T 1/2 experiment. The relative potency of soluble and bound human Shh alone (A) or in the presence of anti-hedgehog neutralizing Mab5E1 (B) was assessed by measuring alkaline phosphatase induction on C3H10T 1/2 cells. The numbers shown reflect the average of two repeated measurements. Soluble (6) and bound (8) Shh (A) series of 2-fold dilutions were incubated with the cells for 5 days and the level of alkaline phosphatase activity achieved was measured at 405nm using the alkaline phosphatase chromogenic substrate p-nitrophenyl phosphate. Serial dilutions of Mab5E1 (B) were incubated with soluble Shh (5. mu.g/ml, black bars) or bound Shh (0.25. mu.g/ml, grey bars) or vehicle control without Shh addition (white bars) for 30 minutes before analysis in the C3HT 101/2 experiment.

FIG. 7 is an analysis of Shh in a receptor binding assay. The relative potency of soluble (6) and bound (8) Shh to bind Patched was assessed by FACS analysis on Patched transfected EBNA-293 cells. Serial dilutions of the test samples were incubated with EBNA-293 cells, washed, and percent binding was measured by the ability of the sample to compete with Shh-Ig for binding to the cells. Bound Shh-Ig was quantified by mean fluorescence as a read out using FITC labeled anti-Ig antibody probes. The data were fitted to a curve using non-linear regression.

FIG. 8 is an alignment of N-terminal fragments of human hedgehog protein. The 20kDa human hedgehog proteins (Sonic "Shh", Desert "Dhh" and Indian "Ihh") were aligned relative to their N-terminal cystine (Cys-1 in the mature sequence). This cystine is usually Cys-24 in the full length Shh precursor protein due to the presence of the native signal sequence, which is removed during secretion. The actual position of cystine varies slightly due to species differences.

FIG. 9 is a consensus sequence of an N-terminal fragment of the human hedgehog protein.

FIG. 10 is a graph of the effect of lipid chain length on human Sonic hedgehog activity. A series of fatty acid-modified hedgehog proteins were synthesized according to the invention, and the effect of fatty acid chain length on hedgehog activity was tested using the C3H10T 1/2 alkaline phosphatase induction assay described herein. The results are shown in block diagram form.

FIG. 11 is a C3H10T 1/2 experiment with human Sonichehghog palmitoylated, myristoylated, laurylated, decylated, and caprylated. Prepared in 5mM Na on C3H10T 1/2 cells by measuring alkaline phosphatase induction₂HPO₄Human Sonic hedgehog with pH 5.5, 150mM NaCl, 10% octyl glucoside, palmitoylation, lauroylation, decanoylation, and octanoylation in 0.5 mM DTT, and myristoylation in 150mM NaCl, 0.5 mM DTT. Palmitoylation (. smallcircle.), myristoylation (●), lauroylation (□), decanoylation (■), octanoylation (. DELTA.) and unmodified (. tangle-solidup and X) human Sonichehehehigh were incubated with the cells for 5 days and the level of alkaline phosphatase activity was measured at 405nm using the chromogenic substrate p-nitrophenyl phosphate of alkaline phosphatase. Palmitoylated, myristoylated, lauroylated and decanoylated proteins were analyzed in one experiment with unmodified protein indicated at a, whereas octanoylated proteins were analyzed in another experiment with unmodified protein indicated at x. The arrow on the y-axis indicates the background level of alkaline phosphatase without the addition of hedgehog protein.

FIG. 12 is the genetic structure of various hydrophobically modified hedgehogs. (A) A fatty acid derivative, wherein R = the hydrocarbon chain of the fatty acid; (B) a thiazolidine derivative wherein R = hydrocarbyl; (C) amino acid substitutions, wherein R = hydrophobic amino acid side chain; (D) maleimide derivatives, wherein R = hydrocarbyl; (E) SH = free on the N-terminal cystine of wild-type hedgehogA thiol; (F) iodoacetamide derivative wherein R₁= hydrocarbyl, and R₂H or a hydrocarbyl group; and (G) thiomorpholinyl derivatives, wherein R = hydrocarbyl. HH = hedgehog for all structures.

FIG. 13 is a graph of the relative potency of various hydrophobically modified hedgehogs in C3H10T 1/2 experiments. EC of unmodified wild-type human Sonic hedgehog₅₀(2. mu.g/ml) is represented by 1X. Potency of other proteins EC of wild-type protein was used₅₀EC divided by modified protein₅₀Is expressed by the ratio of (A) to (B). Unless otherwise indicated, the modification is at the N-terminus of the protein.

Figure 14 is a graph of the relative potency of unmodified, myristoylated, and criii mutated human Sonic hedgehog in malonic acid induced murine striatal injury experiments. The numbers indicate the reduction in malonate-induced lesion volume resulting from administration of unmodified, myristoylated, and caiii mutated human Sonic hedgehog to the murine striatum.

FIG. 15 shows specific activity of maleimide-modified and unmodified hedgehog polypeptides.

Detailed description of the invention

The present invention is based in part on the following findings: human Sonic hedgehog, expressed in full length structures in insect or mammalian cells, has a hydrophobic palmitoyl group pendant from the alpha-amine of the N-terminal cystine. This is the first example found by the inventors of the present invention, which is also an extracellular signaling protein modified in this way, and the modified linkage is very easily reversed as opposed to the thiol-linked palmitic acid modification, while the novel N-linked palmitoyl group is very stable like the myristic acid modification.

As a direct result of this first finding, the inventors of the present invention found that increasing the hydrophobicity of the signaling protein increases the biological activity of the protein. Specifically, the present inventors have found that suspending a hydrophobic group on a signal protein, such as a hedgehog protein, enhances the activity of the protein. The present inventors have also found that the N-terminal cystine of a biologically active protein not only provides a convenient site for hanging hydrophobic groups, thereby altering the physico-chemical properties of the protein, but that modification of the N-terminal cystine can also increase the stability of the protein. In addition, the addition of hydrophobic groups to internal amino acids on the surface of the protein structure can enhance the activity of the protein. We use hydrophobically modified hedgehog proteins (such as lipids and hydrophobic amino acids) as an example.

One aspect of the present invention relates to the discovery that: in addition to those effects seen by the addition of cholesterol to the C-terminus of the extracellular segment of the protein, at least some of the biological activity of the hedgehog gene product is enhanced by modifying the protein with lipophilic groups at other sites on the protein and/or with groups other than cholesterol. Certain aspects of the invention relate to the preparation of hedgehog polypeptides that are modified at sites other than the N-terminal or C-terminal residues of naturally processed proteins, and/or modified at such terminal residues with lipophilic groups other than C-terminal cholesterol or N-terminal fatty acids.

Hedgehog polypeptides are commonly used to repair and/or modulate the functional properties of many cells, tissues and organs in vitro and in vivo, and have the following therapeutic effects, as described in PCT publications WO 95/18856 and WO 96/17924 (which are incorporated herein by reference): neuroprotection, nerve regeneration, enhancement of nerve function, bone regulation and cartilage formation and repair, regulation of sperm production, regulation of lung, liver and other organs produced by primitive gut, regulation of hematopoietic function, and the like. Thus, the methods and compositions of the invention include the use of derivatized hedgehog polypeptides for applications for hedgehog proteins. Alternatively, the method may be performed on cells in culture (in vitro) or on cells of whole animals (in vivo).

In one aspect, the invention provides pharmaceutical formulations comprising a hedgehog polypeptide derivatized with one or more lipophilic groups as described herein as an active ingredient.

Host hedgehog treatment is effective in both humans and animals. Animal hosts that can be used in the present invention include domestic animals and livestock raised as pets or for commercial purposes. Examples include dogs, cats, cattle, horses, sheep, pigs and goats.

hedgehog proteins are a class of extracellular signaling proteins that regulate various aspects of embryonic development in both vertebrates and invertebrates (see 1, 2). The most characteristic hedgehog protein is Sonic hedgehog (Shh), which is involved in anteroposterior formation, apical ectodermal ridge formation, posterior intestinal mesoderm, spine, distal limb, rib development, and lung development, and induces anterior cell types in the spinal cord, hindbrain, and forebrain (3-8). Although the mechanism of action of hedgehog proteins is not fully understood, recent biochemical and genetic data suggest that receptors for Shh are products of tumor suppressor genes, patched (9, 10) and other proteins; smoothened (10, 11), Cubitus interrupt (12, 13) and fused (14) are involved in the hedgehog signal path.

Human Shh is a synthesized 45kDa precursor protein that is produced by autocatalytic cleavage: (I) a 20kDa N-terminal fragment corresponding to all known hedgehog signaling activities (SEQ ID NOS.1-4); and (II) a 25kDa C-terminal fragment which contains self-processing activity (15-17). The N-terminal fragment consists of amino acid residues 24-197 of the full-length precursor sequence.

The N-terminal fragment remains associated with the membrane upon addition of cholesterol at its C-terminus (18, 19). This cholesterol is critical to limit tissue localization of hedgehog signaling. The addition of cholesterol is catalyzed by the C-terminal region during the treatment step.

All documents cited herein are incorporated by reference unless otherwise indicated. I, definition

The invention will now be further described with reference to the detailed description, which includes the definitions below.

"amino acids" -monomeric units of a peptide, polypeptide or protein. There are 20 amino acids found in natural peptides, polypeptides and proteins, all of which are L-isomeric. The term also includes analogs of amino acids and D-isomers of proteinogenic amino acids and their analogs.

"protein" -A polymer consisting essentially of any of the 20 amino acids. Although "polypeptide" is generally used to refer to relatively larger polypeptides and "peptide" is generally used to refer to smaller polypeptides, the use of these terms in the art is cross-over and variable. The term "protein" as used herein refers to peptides, proteins and polypeptides, unless otherwise indicated.

"N-terminal" -refers to the first amino acid (amino acid 1) of the mature protein.

"N-terminal cystine" -refers to the amino acid sequence set forth in SEQ ID NOS: 1 to 4 (1). It also refers to any cystine or functional equivalent of such cystine at position 1 of any other protein (see section iv).

"spacer" sequence refers to a short sequence that can be as large as a single amino acid, which can be inserted between the amino acid to be hydrophobically modified (e.g., the N-terminal cystine or functional equivalent thereof) and the rest of the protein. The spacer is used to separate the hydrophobic modification (e.g., the modified N-terminal cystine) from the rest of the protein to prevent the modification from interfering with the function of the protein and/or to make it easier to modify (e.g., the N-terminal cystine) to attach a lipid, vesicle, or other hydrophobic group. Thus, if the protein is modified at its N-terminal cystine and other sites of amino acids, there may be two or more spacer sequences.

"associated" protein-refers to a protein that has been hydrophobically modified according to the present invention.

"multivalent protein complex" -refers to a plurality of proteins (e.g., one or more). A lipid or other hydrophobic group is attached to at least one of the plurality of proteins. Lipids or other hydrophobic groups may optionally be contacted with the vesicles. If a protein lacks lipids or other hydrophobic groups, it is possible for the protein to be cross-linked or bound to a protein having lipids or other hydrophobic groups. Each protein may be the same or different, and each lipid or other hydrophobic group may also be the same or different.

"vesicle" -refers to an aggregate of any lipophilic molecule. Vesicles may be derived from biological material (e.g. lipid bilayers such as cell membranes or bile acid derived detergent formulations) or non-biological material (non-biological detergent vesicles as described in section iv). The shape, type and configuration of the blisters are not intended to limit the scope of the present invention.

"functional equivalent" of an amino acid residue (e.g., an N-terminal cystine) refers to (i) an amino acid having similar reactive properties as the amino acid residue substituted with a functional equivalent; (ii) an amino acid of a ligand of a polypeptide of the invention, which amino acid has similar hydrophobic (e.g. lipid) group binding properties as the amino acid residue replaced by a functional equivalent; (iii) non-amino acid molecules having hydrophobic (e.g. lipid) group binding properties similar to those of amino acid residues substituted with functional equivalents.

"Gene fusion" -refers to the formation of co-linear covalent bonds between two or more proteins or fragments thereof via their individual peptide backbones by gene expression of polynucleotide molecules encoding these proteins.

By "chimeric protein" or "fusion protein" is meant a fusion of a first amino acid sequence encoding a hedgehog polypeptide with a second amino acid defining a domain that is foreign or substantially non-homologous to any domain of an hh protein. Chimeric proteins may represent a foreign domain present in an organism that also expresses the first protein (but may be in a different protein), or an "interspecies", "intergenic", etc. fusion of protein structures expressed by different species of organisms. In general, the fusion protein can be represented by the general formula (X)_n-(hh)_m-(Y)_n(wherein hh represents all or part of a hedgehog protein, X and Y each independently represent a polypeptide chain which is not a naturally occurring amino acid sequence, e.g., contiguous with a hedghog sequence, m is an integer greater than or equal to 1, and each n is independently 0 or an integer greater than or equal to 1, n and m preferably being no greater than 5 or 10).

"mutant" -any change in the biological genetic material, particularly any change (such as a deletion, substitution, addition or alteration) in the wild-type polynucleotide sequence or any change in the wild-type protein.

"wild-type" -a natural polynucleotide sequence of a protein or a portion thereof, or of an exon of a protein sequence or a portion thereof, as it normally exists in vivo.

"Standard hybridization conditions" -salt and temperature conditions substantially equal to 0.5 XSSC to about 5XSSC and 65 ℃ during hybridization and washing. The term "standard hybridization conditions" as used herein refers to operational limitations and includes a range of hybridization conditions. See also "Current Protocols in molecular Biology", John Wiley & Sons, Inc. New York, Sections 6.3.1-6.3.6 (1989).

"expression control sequences" -polynucleotide sequences that, when operably linked to genes, can control and regulate the expression of those genes.

"operably linked" -a polynucleotide sequence (DNA, RNA) is operably linked to an expression control sequence when the expression control sequence controls and regulates the transcription and translation of the polynucleotide sequence. The term "operably linked" as used herein includes having a suitable initiation signal (e.g., ATG) before the polynucleotide sequence to be expressed and maintaining the correct reading frame under the control of an expression control sequence for expression of the polynucleotide sequence and production of the desired polypeptide encoded by the polynucleotide sequence.

"expression vector" -a polynucleotide, such as a DNA plasmid or phage (among other examples), which allows the expression of at least one gene when the expression vector is introduced into a host cell. The vector may or may not be capable of replicating in the cell.

"isolated" (used interchangeably with "substantially pure") -when used with a nucleic acid, such as a polynucleotide sequence encoding a polypeptide, represents an RNA or DNA polynucleotide, a partial genomic polynucleotide, a cDNA, or a synthetic polynucleotide that, due to origin or operating conditions (i), is not associated with all polynucleotides with which it is naturally associated (e.g., is present in a host cell as an expression vector or a portion thereof); or (ii) attached to the nucleic acid or other chemical group other than that to which it is naturally attached; or (iii) is not naturally occurring. "isolated" further represents the following polynucleotide sequences: amplified in vitro, for example, by the Polymerase Chain Reaction (PCR); (ii) chemically synthesized; (iii) produced by clonal recombination; or (iv) purified by cleavage and gel separation.

Thus, a "substantially pure nucleic acid" is a nucleic acid that is not immediately contiguous with one or both of the coding sequences with which it is normally contiguous in the natural genome of the organism from which it is derived. Substantially pure DNA also includes recombinant DNA that is part of a hybrid gene encoding other hedgehog sequences.

"isolated" (used interchangeably with "substantially pure") -when used with a polynucleotide, represents a polynucleotide or a portion thereof that, due to origin or operating conditions: an expression product present in a host cell as part of an expression vector; or (ii) attached to a protein or other chemical group other than that to which it is naturally attached; or (iii) a protein which is not naturally occurring, e.g. chemically modified by hanging or adding at least one hydrophobic group to the protein, such that the protein is not in its natural form. "isolated" further represents the following polynucleotide sequences: chemically synthesized; or (ii) expressed in a host cell and purified from associated and contaminating proteins. The term generally denotes polynucleotides isolated from other proteins and their naturally-occurring nucleic acids. Preferably, the polynucleotide is also isolated from, for example, antibodies or a gel matrix (polyacrylamide) used for the purification of the nucleotide.

"heterologous promoter" -as used herein refers to a promoter that is not naturally associated with a gene or a purified nucleic acid.

"homologous" -which is identical to "herein is an synonym and refers to sequence similarity between two polypeptides, molecules, or two nucleic acids. When a position in two aligned sequences is occupied by the same base or amino acid monomer subunit (e.g., a position in two DNA molecules is occupied by adenine, or a position in two polynucleotides is occupied by lysine), then the corresponding molecules are homologous at that position. Partial homology between two sequences is a function of the number of positions at which the two sequences match or are homologous, divided by the number of positions compared, x 100. For example, if 6 of 10 positions in two sequences are matched or homologous, then the two sequences are 60% homologous. For example, the DNA sequences CTGACT and CAGGTT are 50% homologous (3 of the 6 total positions are matched). Typically, two sequences are placed side by side for comparison to give maximum homology. Such alignments can be carried out, for example, using the method of Needleman et al (J.mol Bio 1.48: 443-. Homologous sequences have identical or similar amino acid residues, wherein similar residues are conservative substitutions or "allowed point mutations" of corresponding amino acid residues in the reference sequence of the alignment. In this regard, a "conservative substitution" of a residue in a reference sequence is one that is physically or functionally similar to the corresponding reference residue, e.g., of similar size, shape, charge, chemical properties, including the ability to form covalent or hydrogen bonds, and the like. Particularly preferred conservative substitutions are those which meet the criteria defined by Dayhoff et al (5: map of Protein sequences and structures), 5: suppl.3, Chapter 22: 354-.

The "hedgehog protein" or "hedgehog polypeptide" in the present invention are terms used interchangeably and have at least the amino acid sequence defined by the consensus amino acid sequence SEQ ID NO: 4 is part of a building block. The term also represents a hedgehog polypeptide having biological activity, or a functional variant of a hedgehog polypeptide, or a homolog of a hedgehog polypeptide, or a functional variant. In particular, the term encompasses preparations of the hedgehog protein and peptide fragments thereof, both in the form of agonists and antagonists, as the particular context will make it more clear. The term "biologically active fragment of a hedgehog protein" as used herein refers to a fragment of a full-length hedgehog polypeptide, wherein the fragment specifically agonizes or antagonizes an inducible event mediated by the wild-type hedgehog protein. The hedgehog biologically active fragment is preferably the soluble extracellular portion of the hedgehog protein, and the solubility is with reference to a physiologically compatible solution. Exemplary biologically active fragments are described in PCT publications WO 95/18856 and WO 96/17924. In a preferred embodiment, the hedgehog polypeptide of the invention binds to a patched protein.

The term "corresponding" when referring to a particular polypeptide or nucleic acid sequence, means that the sequence of interest is identical or homologous to the corresponding reference sequence.

The terms "peptide", "protein" and "polypeptide" are used interchangeably herein. The terms "polynucleotide sequence" and "nucleotide sequence" are also used interchangeably herein. The terms "hedgehog fragment" and "hedgehog N-terminal fragment" are used interchangeably herein with "hedgehog".

A hedgehog molecule is "biologically active" if it has at least one of the following properties: a molecule that meets the hedgehog consensus criteria defined herein (SEQ ID NO: 4) and has the ability to bind its receptor, patched, or which encodes a polypeptide having such a characteristic when expressed; (ii) the molecule meets the hedgehog consensus criterion defined herein, or encodes a polypeptide having such a characteristic when expressed; and (iii) it induces alkaline phosphatase activity in C3H10T 1/2 cells. Generally, any protein has "biological activity" if it has an in vitro effect, property, or characteristic that one of ordinary skill in the art would consider representative of, consistent with, or reasonably expected to be an in vivo effect of the protein.

The term "hydrophobic" refers to having a tendency for non-polar atomic chemical groups to interact with each other rather than with water or other polar atoms. The hydrophobic substances are mostly insoluble in water. Natural substances having hydrophobicity include lipids, fatty acids, phospholipids, sphingolipids, acylglycerols, waxes, sterols, steroids, terpenes, prostaglandins, thromboxanes, leukotrienes, isoprenoids, retenoids, biotin, and hydrophobic amino acids such as tryptophan, phenylalanine, isoleucine, leucine, valine, methionine, alanine, proline, and tyrosine. Chemical groups may also be hydrophobic or have hydrophobic properties if their physical properties are determined by the presence of non-polar atoms. The term includes lipophilic groups.

"lipophilic group" in connection with a polypeptide refers to a group having a high hydrocarbon content and thus a high affinity of the group for the lipid phase. The lipophilic group may, for example, comprise a relatively long chain alkyl or cycloalkyl (preferably n-alkyl) group having from about 7 to 30 carbon atoms. The alkyl group may have a hydroxyl group or a primary amine "tail" at the terminus. To further illustrate, lipophilic molecules include natural and synthetic aromatic and non-aromatic groups such as fatty acids, esters and alcohols, other lipid molecules, cage structures such as adamantanes and fullerenes (buckminsterfullerenes), and aromatic hydrocarbons such as benzene, north, phenanthrene, anthracene, naphthalene, pyrene, chrysene and tetracene.

"internal amino acids" represent any amino acid in the peptide sequence that is neither the N-terminal nor the C-terminal amino acid.

"surface amino acid" means any amino acid that is exposed to a solvent when folded into its native form.

An "extracellular signaling protein" refers to any protein that is secreted from a cell or binds to the outside of a cell and that elicits a response in a target cell when bound to a receptor for the protein of the target cell.

An "effective amount" of a hedgehog polypeptide, relative to a particular treatment, refers to an amount of the polypeptide in a formulation that controls changes in cell proliferation rate and/or cell differentiation status and/or cell viability when administered at a fraction of the desired dose, depending on the clinical criteria or cosmetic purpose for the disease to be treated.

The "patient" or "host" to be treated by the methods of the invention may be a human or non-human animal. The "growth state" of a cell refers to the proliferation rate of the cell and the differentiation state of the cell.

The practice of the present invention will employ, unless otherwise indicated, conventional techniques in cell biology, cell culture, molecular biology, microbiology, recombinant DNA, protein chemistry, and immunology, which are within the skill of the art. Such techniques are all described in the literature. II Total Properties of isolated hedgehog proteins

The polypeptide portion of the hedgehog compositions of the methods of the invention are produced according to a variety of techniques, including purification of natural proteins, recombinantly produced proteins, and synthetic chemistry. The hedgehog therapeutic in the form of a polypeptide is preferably derived from a vertebrate hedgehog protein, e.g., having a sequence corresponding to a native hedgehog protein of a vertebrate organism, or a fragment thereof. However, it will be appreciated that the hedgehog protein may correspond to a hedgehog protein (or fragment thereof) in any metazoa.

Isolated hedgehog proteins useful in the invention are naturally occurring or recombinant hedgehog proteins and can be obtained from invertebrates or vertebrates (see references below). Members of the vertebrate hedgehog protein, which is encoded by the Drosophila hedgehog (hh) gene, share homology (33). Currently, combinatorial screening of the murine genome and cDNA literature has identified three mammalian hhs, termed Sonic hedgehog (Shh), indianhedgehog (ihh), and Desert hedgehog (Dhh), which are also present in other mammals, including humans, but also in fish and birds. Other members include Moontra hedgehog (Mhh), and chicken Sonic and zebrafish Sonic.

The murine and chicken Shh and murine Ihh genes encode fragmented polynucleotides, yielding an amino-terminal fragment of about 20kDa (see fig. 8) and a carboxy-terminal fragment of about 25 kDa. The most preferred 20kDa fragment has the consensus sequence SEQ ID NO: 4 and includes SEQ ID NOS: 1-3. Other various fragments including the 20kDa portion are also considered to be within the scope of the invention. Documents disclosing these sequences and their chemical and physical properties include (34-38), PCT patent application WO 95/23223 (Jessell, Dodd, Roelink and Edlund), WO 95/18856 (Ingham, McMahon and Tabin), and WO 96/17924 (Beachy et al).

Members that can be used in the methods of the invention include any native hedgehog protein, including allelic, germline counterparts, or other variants, whether native or chemically formed, including muteins, as well as recombinant forms and novel active members. Particularly useful hedgehog polypeptides include seq id NOS: 1-4.

The isolated hedgehog polypeptides used in the invention are biologically active. The polypeptide comprises at least 60%, 80%, 90%, 95%, 98% or 99% homologous to the amino acid sequence of SEQ ID NOS: 1-4. The polypeptide may further comprise a sequence substantially identical to the amino acid sequence of SEQ ID NOS: 1-4, and the like. The polypeptide has a length of at least 5, 10, 20, 50, 100, or 150 amino acids, and includes at least 5, preferably at least 10, more preferably at least 20, most preferably at least 50, 100, or 150 amino acids that are identical to SEQ ID NOS: 1-4 contiguous amino acids.

Preferred polypeptides of the invention include hedgehog polypeptide sequences as well as other N-and/or C-terminal amino acid sequences, or they may include all hedgehog amino acid sequences or fragments thereof. The isolated hedgehog polypeptide can also be a recombinant fusion protein having a first hedgehog portion and a second polypeptide portion, e.g., a second portion polypeptide having an amino acid sequence unrelated to hedgehog. The second polypeptide moiety can be, for example, a histidine tag, a maltose binding protein, a glutathione S-transferase, a DNA binding domain, or a polymerase activation domain.

The polypeptides of the invention include those that result from the presence of multiple genes, alternative transcription events, alternative RNA splicing, and alternative translation and posttranslation events. The polypeptide may be made entirely synthetically, or expressed in a system, such as a cultured cell, that produces substantially the same post-translational modifications when the protein is expressed in the native cell, or that does not produce post-translational modifications when expressed in the native cell.

In a preferred embodiment, the isolated hedgehog is a hedgehog polypeptide having one or more of the following characteristics:

and the amino acids SEQ ID NOS: 1-4 have a sequence that is at least 30, 40, 42, 50, 60, 70, 80, 90, or 95% identical;

(ii) it has a cystine or functional equivalent as the N-terminus;

(iii) it induces alkaline phosphatase activity in C3H10T 1/2 cells;

(iv) its sequence in comparison with SEQ ID NOS: 1-4 have at least 50%, preferably at least 60%, more preferably at least 70, 80, 90 or 95% total sequence identity;

(v) it may be isolated from a natural source such as a mammalian cell;

(vi) it may bind or interact with patched; and

(vii) it is hydrophobically modified (i.e., it has at least one hydrophobic group attached to the polypeptide). III, other proteins

Because techniques exist for engineering cystine residues (or functional equivalents thereof) into the original sequence of a polypeptide, virtually any protein can be converted to a hydrophobically modified form using the methods described herein.

Viral receptors, cellular receptors, and cellular ligands are useful because they typically bind to cells or tissues that have many receptor replicons. Useful virus-cell protein receptors that can be complexed using the methods of the invention include ICAM1 (a rhinovirus receptor), CD2(Epstein-Ban virus receptor), and CD4 (a receptor for Human Immunodeficiency Virus (HIV)). Other proteins include cell adhesion molecules such as ELAM-1 and VCAM-1 and VAM-1b and their lymphocyte counterparts (ligands); lymphocytes are associated with antigens LFA1, LFA2(CD2) and LFA3(CD58), CD59 (a second ligand for CD2), members of CD 11/CD 18 and very late antigens such as VLA4 and their ligands.

Immunogens produced by various pathogens, such as bacteria, fungi, viruses and other eukaryotic parasites, may also be used as polypeptides in the methods of the invention. Bacterial immunogens include, but are not limited to, bacterial sources directed to bacterial pneumonia and pneumocystis pneumonia. The source of parasites includes the Plasmodium malaria parasite. Viral sources include poxviruses (e.g., vaccinia, herpes simplex, cytomegalovirus); an adenovirus; papovavirus (e.g., papilloma virus); parvoviruses (e.g., adeno-associated viruses); retroviruses (e.g., HTLV I, HTLV II, HIV I and HIV II) and other viruses. Immunoglobulins or fragments thereof may also be used as polypeptides modified according to the invention. One can use conventional methods (49) to generate monoclonal Fab fragments that recognize specific antigens and use the individual Fab domains as functional parts in the multimeric structure according to the invention. Other useful proteins include gelsolin (50); cytokines, including various interferons (interferon- α, interferon- β, and interferon- γ); various interleukins (e.g., IL-1, -2, -3, -4, etc.); tumor necrosis factor-alpha and-beta; single cell stimulating factor (M-CSF), granulocyte colony stimulating factor (G-CSF), granulocyte macrophage colony stimulating factor (GM-CSF), erythropoietin, Platelet Derived Growth Factor (PDGF), and human and animal hormones, including growth hormone and insulin.

In general, the structure of the modified proteins of the invention has the following general formula: A-Cys- [ Sp ] -B- [ Sp ] -X, wherein A is a hydrophobic group; cys is cystine or a functional derivative thereof; [ Sp ] is an optional spacer peptide sequence; b protein (which optionally has other spacer peptide sequences as shown herein); and X is a hydrophobic group attached (optionally via a spacer peptide) to the C-terminus of the protein or to other surface sites of the protein, and the derivatized protein includes at least one of a or X. If X is cholesterol, B may or may not be a hedgehog protein. As mentioned above, the purpose of the spacer is to form a space between the hydrophobic group and the rest of the protein, so that the hydrophobic group (e.g. a modified N-terminal cystine) is more easily attached to the rest of the protein, which may be a lipid or vesicle. Another purpose of the spacer is to make it more difficult for the modification step to interfere with the function of the protein. The spacer may be as small as a single amino acid in length. In general, proline and glycine are preferred. Particularly preferred spacer sequences are derived from Sonic hedgehog and are represented by the amino acid sequence: G-P-G-R. IV, preparation of recombinant polypeptide

The isolated polypeptides described herein can be prepared by any suitable method known in the art. Such methods include direct protein synthesis methods up to the construction of DNA sequences encoding the isolated polypeptide sequences and expression of these sequences in a suitable transformed host.

In one embodiment of the recombinant method, the DNA sequence is constructed by isolating or synthesizing a DNA sequence encoding the wild-type protein of interest. Optionally, the sequence can be mutagenized by site-specific mutagenesis to form a functional analog thereof. See, for example, U.S. patent nos. (40) and 4,588,585. Another method for constructing a DNA sequence encoding a polypeptide of interest is chemical synthesis using oligonucleotide synthesizers. Such oligonucleotides may preferably be designed according to the amino acid sequence of the desired polypeptide and preferably those encoding sequences which favor the host cell in which the recombinant polypeptide of interest is produced.

Standard methods are available for synthesizing an isolated polynucleotide sequence encoding an isolated polypeptide of interest. For example, a reverse-translated gene can be constructed using the complete amino acid sequence. See Maniatis et al, supra. In addition, DNA oligomers can be synthesized that comprise nucleotide sequences that encode particular isolated polypeptides. For example, several small oligonucleotides encoding portions of the desired polypeptide can be synthesized and then ligated. Individual oligonucleotides typically contain 5 'or 3' overhangs for compensatory assembly.

Once assembled (by synthesis, site-directed mutagenesis, or by other means), a mutant DNA sequence encoding a particular isolated polypeptide of interest is inserted into an expression vector and operatively linked to expression control sequences suitable for expression of the protein in the desired host. Proper assembly can be confirmed by nucleotide sequencing, restriction mapping, and expression of the biologically active polypeptide in a suitable host. As is well known to those skilled in the art, in order for a transfected gene to produce high levels of expression in a host, the gene must be operably linked to transcriptional and translational expression control sequences that are functional in the expression host of choice.

The choice of expression control sequences and expression vectors depends on the choice of the host. Various expression host/vector combinations may be used. Expression vectors useful for eukaryotic hosts include vectors containing expression control sequences derived, for example, from SV40, bovine papilloma virus, adenovirus and cytomegalovirus. Expression vectors useful for bacterial hosts include known bacterial plasmids such as those of Escherichia coli, including pCR1, pBR322, pMB9, and derivatives thereof, a broader host range of plasmids such as M13, and filamentous single stranded DNA phages. Preferred E.coli vectors include pL vectors containing the pL promoter of lambda phage (U.S. Pat. No. 4,874,702), pET vectors containing the T7 polymerase promoter (Studier et al, Methods in Enzymology 185: 60-89, 19901), and pSP72 vectors (Kaelin et al, supra). Expression vectors useful for yeast cells include, for example, 2T and centromeric plasmids.

In addition, various expression control sequences can be used in these vectors. Such useful expression control sequences include those related to the structural genes of the expression vectors described above. Examples of useful expression control sequences include, for example, the early and late promoters of SV40 or adenovirus, the lac system, the trp system, the TAC or TRC system, the major operator and promoter regions of phage lambda such as the control regions for pL, fd coat protein, the promoters of 3-phosphoglycerate kinase or other glycolytic enzymes, the promoters of acid phosphatase such as Pho5, the promoters of the yeast alpha-junction system, and other sequences known to control gene expression in prokaryotic or eukaryotic cells and their viruses, and various combinations thereof.

Any suitable host can be used to quantitatively produce the isolated hedgehog polypeptides described herein, including bacterial, fungal (yeast), plant, insect, mammalian or other suitable animal cells or cell lines, as well as transgenic animals or plants. More specifically, these hosts include known eukaryotic and prokaryotic hosts,e.g.coli, Pseudomonas, Bacillus, Streptomyces strains, fungi, yeasts (e.g.Hansenula), insect cells such as Spodoptera frugiperda (SF9), and High Five^TM(see example 1), animal cells such as Chinese Hamster Ovary (CHO), murine cells such as NS/O cells, African green monkey cells COS1, COS7, BSC1, BSC40 and BMT10, and human cells and plant cells.

It is understood that not all vectors and expression control sequences will have the same effect on the expression of a particular isolated polypeptide. Nor do all hosts function identically for the same expression system. However, one skilled in the art can select among these vectors, expression control systems, and hosts without undue experimentation. For example, in order to produce an isolated polypeptide of interest in a large scale animal culture, the number of replications of the expression vector must be controlled. Amplifiable vectors are known in the art. See, for example, (41) and U.S. patent nos. 4, 470, 461 and 5, 122, 464.

Operably linking the DNA sequence to an expression control sequence comprises providing a translation initiation signal in the correct reading frame upstream of the DNA sequence. If the specific DNA sequence expressed does not start with methionine, the initiation signal will generate an additional amino acid (methionine) which is located at the N-terminus of the product. If the hydrophobic group is attached to a protein comprising an N-terminal methionine, the protein may be used directly in the composition of the present invention. However, since the preferred N-terminus of the protein is composed of cystine (or a functional analog thereof), methionine must be removed prior to use. There are many methods in the art for removing such N-terminal methionine from polypeptides expressed using them. For example, certain hosts and fermentation conditions allow for the removal of substantially all of the N-terminal methionine in vivo. Other hosts require in vitro removal of the N-terminal methionine. Such in vitro and in vivo methods are well known in the art.

The protein produced by the transformed host may be purified according to any suitable method. Such standard methods include chromatography (e.g., ion exchange, affinity, and size column chromatography), centrifugation, differential solubility, or by any other standard technique in protein purification. For immunoaffinity chromatography (see example 1), proteins such as Sonic hedgehog can be isolated by binding them to an affinity column containing an anti-Sonic hedgehog antibody or related protein, which is then immobilized on an immobilization carrier. Alternatively, affinity tags such as hexa-histidine, maltose binding domain, influenza coating sequence, and glutathione S-transferase are attached to the protein to allow for easier purification from a suitable affinity column. Isolated proteins may also be physically characterized using techniques such as proteolysis, nuclear magnetic resonance, and X-ray crystallography. A. Preparation of fragments and analogs

Fragments of an isolated protein (e.g., fragments of SEQ ID NOS: 1-4) can also be efficiently prepared by recombinant methods, proteolytic digestion, or by chemical synthesis using methods known to those skilled in the art. In recombinant methods, an internal or terminal fragment of a polypeptide can be generated by removing one or more nucleotides from one end (for the terminal fragment) or both ends (for the internal fragment) of a DNA sequence encoding an isolated hedgehog polypeptide. Expression of the mutagenized DNA produces polypeptide fragments. Digestion with "end-cutting" endonucleases can also produce DNA encoding a series of fragments. DNA encoding protein fragments may also be generated by random shearing, restriction digestion, or a combination thereof. Protein fragments can be produced directly from the intact protein. Peptides can be specifically cleaved with proteolytic enzymes, including but not limited to plasmin, thrombin, trypsin, chymotrypsin, or pepsin. These enzymes are specific for the type of peptide bond they attack. In the peptide bond of the trypsin catalyzed hydrolysis, the carbonyl group is of a basic amino acid, usually arginine or lysine. The carbonyl groups that pepsin and chymotrypsin catalyze the hydrolysis are aromatic amino acids, such as tryptophan, tyrosine and phenylalanine. Preventing cleavage of sites susceptible to proteolytic enzyme attack, whereby additional cleaved protein fragments may be generated. For example, the epsilon amino acid group of lysine reacts with ethyltrifluoroacetate in moderately alkaline solution to form a blocked amino acid residue, the adjacent peptide bond of which is no longer susceptible to hydrolysis by trypsin. Proteins may be modified to increase peptide bonds susceptible to proteolytic enzymes. For example, alkylation of cystine residues with β -haloethylamine results in peptide bonds that are hydrolyzed by trypsin (51). In addition, chemical agents that cleave peptide chains at specific residues can be used. For example, cyanogen bromide cleaves peptides at methionine residues (52). Thus, treatment with various combinations of modifiers, proteolytic enzymes and/or chemicals can break the protein into fragments of desired length, where there is no overlap of the fragments, or into overlapping fragments where overlap is desired.

Fragments can also be chemically synthesized by using techniques known in the art, such as Merrifield solid phase Fmoc or t-Boc chemistry (Merrifield, Recent Progress in Hormone Research 23: 451 (1967)).

Examples of prior art methods for generating and testing fragments and analogs are discussed below. These and similar methods can be used to prepare and screen fragments and analogs of isolated polypeptides (e.g., hedgehog) for biological activity. Exemplary methods for testing hedgehog fragments and analogs for biological activity are described in example 3. B. Preparation of altered DNA and peptide sequences: stochastic method

Amino acid sequence variants of a protein (e.g., variants of SEQ ID NOS: 1-4) can be prepared by random mutagenesis of DNA encoding the protein or a specified portion thereof. Useful methods include PCR mutagenesis and saturation mutagenesis. A series of random amino acid sequence variants can also be formed by synthesizing a degenerate set of oligonucleotide sequences. Methods of using altered DNA and peptides to generate amino acid sequence variants of a given protein are known in the art. The following examples of such methods are not intended to limit the scope of the present invention, but are merely illustrative of representative techniques. One of ordinary skill in the art will recognize that other methods may be used in this regard. PCR mutagenesis: briefly, random mutations are introduced into cloned fragments of DNA using Taq polymerase (or other polymerases) (42). The PCR conditions are selected by using a dGTP/dATP ratio of, for example, 5 and adding Mn to the PCR reaction²⁺While the use of Taq DNA polymers reduces the fidelity of DNA synthesis. Amplified DNA fragmentThe pool is inserted into a suitable cloning vector to form a random pool of mutated genes. Saturation mutagenesis: (43) the general method is described in (1). Briefly, this technique involves generating mutations by chemically treating or irradiating single-stranded DNA in vitro, followed by synthesis of the cDNA strand. The frequency of mutation is adjusted by the degree of treatment, and essentially all possible base substitutions are available. Degenerate oligonucleotide mutagenesis: libraries of homologous peptides can be generated from a series of degenerate oligonucleotide sequences. Chemical synthesis of degenerate sequences can be carried out in an automated DNA synthesizer and the synthesized genes then ligated into suitable expression vectors. See, for example, (44, 45) and Itakura et al (Recombinant DNA, Proc. 3)^rdCleveland Symposium on Macromolecules, pp.273-289 (edited by A.G. Walton), Elsevier, Amsterdam, 1981). C. Preparation of altered DNA and peptide sequences: direct process

Non-random or direct mutagenesis may generate specific sequences or mutations in specific portions of a polynucleotide sequence encoding an isolated peptide to form variants comprising deletions, insertions, or substitutions of residues of a known amino acid sequence of the isolated peptide. The mutation sites may be changed individually or in series as follows: (1) first with conservative amino acids, and then with more radical selection, depending on the result to be achieved; (2) deleting the target residue; or (3) insertion of residues of the same or different classes at adjacent sites, or a combination of options 1-3 above.

For clarity, such site-directed methods are those in which an N-terminal cystine (or functional analog thereof) may be introduced into a given polypeptide sequence to form an attachment site for a hydrophobic group. Alanine scanning mutagenesis: the method is limited to those residues or regions of the desired protein that are preferred sites for mutagenesis (46). In alanine screening, the residue or group of residues of interest can be selected and replaced with alanine. The substitutions may affect the interaction of the amino acid with adjacent amino acids and/or with the surrounding aqueous or membrane environment. Those with functional sensitivity to substitution can be refined by introducing other variants at other sites. Oligonucleotide-mediated mutagenesis: one version of the method can be used to prepare substitution, deletion, and insertion variants of DNA (47). Briefly, the desired DNA is altered by hybridizing oligonucleotide primers encoding DNA mutations to a DNA template, typically a single-stranded plasmid or phage, which contains an unaltered or wild-type DNA sequence template for the desired protein (e.g., hedgehog protein). After hybridization, a second and complementary template DNA strand is prepared with a DNA polymerase, which will insert an oligonucleotide primer and then encode the selected alteration in the desired DNA sequence. Usually, oligonucleotides with a chain length of at least 25 nucleotides are used. The optimal oligonucleotide has 12-15 nucleotides which completely compensate for the template on one side of the mutation. This ensures that the oligonucleotide will hybridize properly to a single-stranded DNA template molecule. Box type mutagenesis: this method (48) requires a plasmid or other vector containing the protein subunit DNA to be mutated. The codens in the protein subunit DNA are identified and then unique restriction endonucleases are inserted on each side of the identified mutation sites using the above-described mutagenesis method involving oligonucleotides. The plasmid is then linearized by cleaving at these sites. Double-stranded oligonucleotides encoding DNA sequences between the restriction sites but containing the desired mutation are synthesized using standard methods. The two strands are synthesized separately and then hybridized together using standard methods. The double-stranded oligonucleotide is a "cassette" and has 3 'and 5' ends that fit into the ends of a linear plasmid, enabling it to be directly ligated thereto. The plasmid now contains the mutated desired protein subunit DNA sequence. Combination transformation: in one version of this method (Ladner et al, WO 88/06630), the amino acid sequences of homologues or other related proteomes are compared, preferably to promote the highest homology possible. All amino acids present at a given position in the compared sequences can be selected to produce a degenerate combined sequence. Combinatorial mutagenesis is performed at the nucleic acid level, thereby generating different gene pools. For example, a mixture of synthetic oligonucleotides can be enzymatically ligated into a gene sequence such that a degenerate potential sequence can be expressed as a single peptide or as a protein comprising the entire degenerate sequence. D. Additional variants of isolated Polypeptides

The isolated molecules included in the present invention are: allelic variants, natural mutants, induced mutants, and proteins encoded by DNA that hybridizes under high or low stringency conditions to a nucleic acid encoding a polypeptide such as an N-terminal fragment of Sonic hedgehog (SEQ ID NO: 1) and a polypeptide that specifically binds by antisera directed to a hedgehog peptide, particularly by antisera directed to the active or binding site of hedgehog. All variants described herein contemplate that: (1) the biological function of the original protein is reserved; and (2) retain the ability to attach to hydrophobic groups such as lipids.

The methods of the invention also feature the use of isolated fragments, preferably biologically active fragments or analogs, of peptides such as hedgehog. In particular, a biologically active fragment or analog is one that has in vivo or in vitro activity as set forth in SEQ ID NOS: 1-4 or other naturally isolated hedgehog. More preferably, the hydrophobically modified fragment or analog has a sonic hedgehog activity of at least 10%, preferably 40% or more, or most preferably at least 90% in an in vivo or in vitro assay (see example 3).

Analogs differ from the native protein isolate in either or both of the amino acid sequence and the unrelated sequence. The most preferred polypeptides of the invention have preferred non-sequence modifications, which include chemical derivatization (e.g., of their N-terminus) in vivo or in vitro, as well as possible changes in acetylation, methylation, phosphorylation, amidation, carboxylation, or glycosidation.

Other analogs include proteins such as Sonic hedgehog or biologically active fragments thereof that differ in sequence from the wild-type consensus sequence (e.g., SEQ ID NO: 4) by one or more conservative amino acid substitutions or one or more non-conservative amino acid substitutions, or deletions or insertions that do not eliminate the biological activity of the isolated protein. Conservative substitutions typically include the replacement of one of the amino acids with another having similar characteristics, such as substitutions in the following groups: valine, alanine and glycine; leucine and isoleucine; aspartic acid and glutamic acid; asparagine and glutamine; serine and threonine; lysine and arginine; and phenylalanine and tyrosine. Non-polar hydrophobic amino acids include alanine, leucine, isoleucine, valine, proline, phenylalanine, tryptophan, and methionine. Polar neutral amino acids include glycine, serine, threonine, cystine, tyrosine, asparagine, and glutamine. Positively charged (basic) amino acids include arginine, lysine and histidine. Negatively charged (acidic) amino acids include aspartic acid and glutamic acid. Other conservative substitutions are known to those skilled in the art. For example, for alanine, conservative substitutions can be selected from D-alanine, glycine, beta-alanine, L-cystine and D-cystine. For lysine, the substitution may be any of D-lysine, arginine, D-arginine, homoarginine, methionine, D-methionine, ornithine, or D-ornithine.

In general, the substitutions that can be used to induce a change in the functional properties of an isolated polypeptide are as follows: polar residues, such as serine or threonine, for substitution or replacement by hydrophobic residues, such as leucine, isoleucine, phenylalanine or alanine; (ii) cystine residues are used for substitution or replaced by any other residue (see example 10); (iii) a residue having an electropositive side chain such as lysine, arginine or histidine, for substitution or substitution with a residue having an electronegative side chain such as glutamic acid or aspartic acid; or (iv) residues with large side chains, such as phenylalanine, are used instead of or instead of residues without such side chains, such as lysine.

Other analogs useful in the methods of the invention are those having modifications that increase the stability of the peptide. Such analogs may include, for example, one or more non-peptide bonds (which replace peptide bonds) in the peptide sequence. It still includes: analogs comprising residues other than natural L-amino acids, e.g., D-amino acids or unnatural or synthetic amino acids, such as beta or gamma amino acids and cyclic analogs. The insertion of a D-amino acid in place of an L-amino acid in an isolated hedgehog polypeptide can increase its tolerance to proteases. See U.S. patent No. 5,219, 990 (supra).

The term "fragment" as applied to an isolated hedgehog analog can be as large as a single amino acid, provided that the biological activity is retained. It may be at least about 20 residues, more usually at least about 40 residues, preferably at least about 60 residues in length. Fragments may be generated by methods known to those skilled in the art. The ability of a candidate fragment to isolate the biological activity of hedgehog can also be assessed using methods known in the art as described herein. V. preparation of hydrophobic derivatives

The inventors of the present invention have discovered that increasing the overall hydrophobicity of a signaling protein, such as a hedgehog protein, increases the biological activity of the protein. The potency of signaling proteins such as hedgehog can be increased as follows: (a) chemical modification, e.g. addition of hydrophobic groups to the thiol group and/or alpha-amine of the N-terminal cystine (examples 8 and 9); (b) replacement of the N-terminal cystine with a hydrophobic amino acid (example 10); or (c) replacing the N-terminal cystine with a different amino acid, and then chemically modifying the replaced residue to add a hydrophobic group at the replacement site.

In addition, by (a) replacing internal residues with hydrophobic amino acids; or (b) replacing internal residues with different amino acids and then chemically modifying the replaced residues to add hydrophobic groups at the replacement sites (see example 10), thereby modifying a protein, such as a hedgehog protein, with hydrophobic groups at internal residues on the surface of the protein, which will maintain or enhance the biological activity of the protein.

Further, by (a) replacing the C-terminal residue with a hydrophobic amino acid; or (b) replacing the C-terminal residue with a different amino acid, and then chemically modifying the replaced residue to add a hydrophobic group at the substitution site, thereby modifying a protein, such as a hedgehog protein, with a hydrophobic group at the C-terminus, which will maintain or enhance the biological activity of the protein.

There are many lipophilic groups that can be used to derive hedgehog polypeptides. Lipophilic groups are, for example, relatively long alkyl or cycloalkyl groups (preferably n-alkyl) having from about 7 to 30 carbon atoms. Alkyl groups may be terminated with a hydroxyl or primary amine "tail". To further illustrate, lipophilic molecules include natural and synthetic aromatic and non-aromatic groups such as fatty acids, esters, and alcohols, other lipid molecules, cage structures such as adamantane and fullerene, and aromatic hydrocarbons such as benzene, north, phenanthrene, anthracene, naphthalene, pyrene, chrysene, and tetracene.

Particularly useful lipophilic molecules are aliphatic cyclic hydrocarbons, saturated and unsaturated fatty acids and other lipid and phospholipid groups, waxes, cholesterol, isoprenoids, terpenes and multi-aliphatic cyclic hydrocarbons such as adamantane and fullerene, vitamins, polyethylene or oligoethylene glycols, (C)₁-C₁₈) -alkylphosphoric acid diester, -O-CH₂-CH(OH)-O-(C₁₂-C₁₈) -alkyl groups, and in particular conjugates with pyrene derivatives. Lipophilic groups may be lipophilic dyes suitable for use in the present invention including, but not limited to, diphenylhexatriene, Nile red, N-phenyl-1-naphthylamine, Prodan, Laurodan, pyrene, north, rhodamine B, tetramethylrhodamine, Texas red, sulforhodamine, 1 '-didodecyl-3, 3, 3' -tetramethylindocarbocyanine perchlorate, octadecylrhodamine B, and BODIPY dyes available from Molecular Probes inc.

Other exemplary lipophilic groups include aliphatic carbonyl groups including 1-or 2-adamantylacetyl, 3-methyladamantan-1-ylacetyl, 3-methyl-3-bromo-1-adamantylacetyl, 1-decalactonyl acetyl, camphoracetyl, camphanylacetyl, noradamantylacetyl, norbornaneacetyl, bicyclo [ 2.2.2. ] -oct-5-eneacetyl, 1-methoxybicyclo [ 2.2.2 ] -oct-5-ene-2-carbonyl, cis-5-norbornene-endo-2, 3-dicarbonyl, 5-norbornen-2-ylacetyl, (1R) - (-) -myrtenanyl acetyl, 2-norbornaneacetyl, anti-3-oxo-tricyclo [ 2.2.1.0 <2, 6> ] -heptane-7-carbonyl, decanoyl, dodecanoyl, dodecenoyl, tetradecadienoyl, decynoyl or dodecenoyl.

Exemplary hydrophobically modified structures are shown in fig. 12. If there is no suitable amino acid at a particular position, site-directed mutagenesis is used to place a reactive amino acid at that site. Reactive amino acids include cystine, lysine, histidine, aspartic acid, glutamic acid, serine, threonine, tyrosine, arginine, methionine, and tryptophan. Mutagenesis can be used to place reactive amino acids at the N-or C-terminus or at internal positions.

For example, we have found that the N-terminal cystine of a biologically active protein, such as a hedgehog protein, can be chemically modified, or completely eliminated and still retain the biological activity of the protein, provided that the modified or replaced reactive group is hydrophobic. The inventors have found that the enhanced biological activity of hedgehog is roughly related to the hydrophobicity of the modifying group. In addition to enhancing the activity of the protein, modification or substitution of the N-terminal cystine may eliminate undesirable cross-reactivity and/or modification of cystine during preparation, purification, formulation and storage of the protein. The thiol of the N-terminal cystine is very reactive because it is close to the a-amine, lowering the pKa of the cystine and increasing the proton dissociation and formation of the reactive thiol ion at neutral or acidic pH.

We have demonstrated that replacement of the N-terminal cystine of hedgehog with a hydrophobic amino acid can result in proteins with increased potency in cell-based signaling experiments. By replacing cystine, the method eliminates the problem of inhibiting undesirable other modifications of cystine that occur during preparation, purification, formulation, and storage of the protein. The generality of the method is supported by our following findings: three different hydrophobic amino acids, phenylalanine, isoleucine and methionine, each produce a more active hedgehog. Thus, substitution of cystine with any other hydrophobic amino acid should result in an active protein. In addition, since we have found a correlation between the hydrophobicity of an amino acid or active modifying group and the potency of the corresponding modified protein in C3H10T 1/2 experiments (e.g., Phe > Met, long chain fatty acid > short chain fatty acid), it is believed that adding more hydrophobic amino acids to the hedgehog sequence increases the potency of the protein beyond that achieved by adding a single amino acid. Indeed, in the C3H10T 1/2 experiment, the addition of two consecutive isoleucine residues at the N-terminus of human Sonic hedgehog resulted in an increase in potency compared to the mutant with only a single isoleucine (see example 10). Thus, the addition of hydrophobic amino acids to the N-or C-terminus of the hedgehog protein, or some combination of these positions, in the surface loop structure can be expected to result in a more active protein. The substituted amino acid need not be one of the 20 common amino acids. Methods have been reported for replacing unnatural amino acids at specific sites in proteins (78, 79), and it would be advantageous if the amino acids were more hydrophobic, more resistant to enzymatic attack, or could be used to further target specific sites in the hedgehog protein in vivo, which would make the activity stronger or more specific. Unnatural amino acids can be inserted at specific sites in proteins during in vitro translation, and methods have been reported to produce systems for forming such modified proteins on a larger scale in vivo.

Unexpectedly, proteins such as hedgehog proteins can retain their biological activity after modification according to the invention. First, the N-terminal cystine is retained in all known hedgehog protein sequences, including fish, frogs, insects, birds, and mammals. Therefore, it is reasonable to believe that the thiol group of the N-terminal cystine is very important for the structure or activity of the protein. Second, hedgehog proteins lacking the N-terminal cystine are inactive in cell-based C3H10T 1/2 experiments, which are described in example 3, due to enzymatic cleavage or mutation to hydrophilic amino acids (e.g., aspartic acid or histidine).

There are many ways to modify the N-terminal cystine to protect the thiol and add hydrophobic groups. These modifications will be discussed in detail below. Those skilled in the art will be able to determine which modification is most appropriate for a particular therapeutic application. Factors that influence such choices include cost and ease of manufacture, purification and formulation, solubility, stability, efficacy, pharmacokinetics and pharmacokinetics, safety, immunogenicity and tissue targeting. A. Chemical modification of original amino acid sequence

The chemical modification of the N-terminal cystine to protect the thiol and to activate it simultaneously with the hydrophobic group can be carried out in a number of ways known to those skilled in the art. Sulfhydryl groups, of which the thiol ion is the most reactive functional group in proteins, have many reagents that can react with thiols more rapidly than any other group. See: chemistry concerning Protein binding and crosslinking (S.S. Wong, CRCPress, Boca Raton, FL, 1991). The thiol of the N-terminal cystine can be found in all hedgehog proteins and is believed to be more reactive than the internal cystine in the sequence. This is because approaching alpha-amines will lower the pKa of the thiol, allowing a greater degree of dissociation of the reactive thiol ion at neutral or acidic pH. In addition, the cystine at the N-terminus of the structure is more easily exposed than the other two cystines in the hedgehog sequence, which are found buried in the protein structure. We have shown that the N-terminal cystine is modified only after 1 hour reaction with N-ethylmaleimide at pH 5.5 (see example 9) and 18 hours reaction with N-isopropyl iodoacetamide at pH 7.0 (see example 9). Other examples of this process are reaction with other alpha-haloacetyl compounds, organomercury, disulfide reagents, and other N-substituted maleimides. Most hydrophobic derivatives of these active species are commercially available (e.g., ethyl iodoacetate (Aldrich, Milwaukee Wis.), diphenyl disulfide (Aldrich), and N-pyrene maleimide (Molecular Probes, Eugene OR)) OR can be readily synthesized (e.g., N-alkyl iodoacetamide (84), N-alkyl maleimide (85), and organomercury compounds (86)). We have shown that the N-terminal cystine of human Sonic hedgehog is specifically modified with N-isopropyl iodoacetamide, and that the hydrophobically modified protein is 2-fold more potent than the unmodified protein in the C3H10T 1/2 experiment (see example 9). It is believed that modification of Shh with long chain alkyl iodoacetamide derivatives may result in modified proteins with higher potency. Such N-alkyl iodoacetamides can be readily synthesized by one of ordinary skill in the art using commercially available starting materials.

Other aspects of the reactivity of the N-terminal cystine are that it can participate in reaction chemistries that are unique to its 1, 2-aminothiol configuration. One example is the reaction with thioester groups to form N-terminal amide groups by rapid S-N migration of the thioester. This reaction chemistry can be coupled with synthetic peptides and can be used to add single or single, natural or unnatural amino acids or other hydrophobic groups through appropriate activation peptides. Another example shown herein is the reaction with an aldehyde to form a thiazolidine adduct. Many hydrophobic derivatives of thiol esters (e.g., C2-C24 saturated and unsaturated fatty acyl-coa esters (sigmal chemical co., st. louis MO)), aldehydes (butyraldehyde, n-decanal, and n-dodecanal (Aldrich)), and ketones (e.g., 2-, 3-, and 4-decanone (Aldrich)) are commercially available or can be readily synthesized (87, 88). In a similar manner, thiomorpholine derivatives, exemplified by 1-bromo-2-butanone chemistry described in example 9, can be prepared from a variety of α -halo ketone starting materials (88). Since alternative methods of modifying the N-terminal cystine or the thiol of cystine in any protein can be readily found, we do not wish to be limited to the specific examples described herein.

The alpha-amine of a protein is preferably modified relative to other amines in the protein because its lower pKa at neutral or acidic pH produces a greater amount of the reactive unprotonated form. We have shown that modification of the N-terminal amine with a long chain fatty amide group can activate hedgehog proteins to two orders of magnitude while retaining free cystine thiol groups (see example 8). Thus, chemistry directed at preferentially reacting with the N-terminal amine can also be used to increase the efficacy of hedgehog proteins. Aryl halides, aldehydes and ketones, anhydrides, isocyanates, isothiocyanates, imidoesters, acid halides, N-hydroxysuccinimidyl (e.g., thio-NHS-acetate), nitrophenyl esters, acylimidazoles, and other activated esters are known to be capable of reacting with amine functionality.

When replacing the N-terminal cystine of hedgehog with some other amino acid, other chemistries can be used to add a hydrophobic group at the N-terminus. One example is the addition of a serine or threonine at the N-terminus, oxidation of the amino acid to form an aldehyde, and then attachment of a chemical group containing a1, 2 aminothiol structure (e.g., cystine) to the protein. A second example is the addition of histidine at the N-terminus to couple to a C-terminal thiocarboxylic acid. Chemical modification of other amino acids

There are specific chemical methods for modifying many other amino acids. Thus, another approach for synthesizing the more active form of hedgehog is to chemically attach a hydrophobic group to an amino acid of hedgehog other than the N-terminal cystine. If there is no suitable amino acid at the desired position, site-directed mutagenesis can be used to add a reactive amino acid at that site of the hedgehog structure, whether at the N-or C-terminus or other position. Reactive amino acids include cystine, lysine, histidine, aspartic acid, glutamic acid, serine, threonine, tyrosine, arginine, methionine, and tryptophan. Thus, the goal of increasing the hydrophobicity of hedgehog can be achieved by a number of chemical methods, and we do not wish to limit with specific chemical or modification sites, as our results support the versatility of the method. The hedgehog polypeptide can be attached to the hydrophobic moiety by a variety of methods, including chemical coupling or genetic engineering.

For purposes of illustration, a wide variety of chemical crosslinking agents are known to those of ordinary skill in the art. For the present invention, the preferred crosslinking agent is a heterobifunctional crosslinking agent that can be used to step-link the hedgehog polypeptide and the hydrophobic group. Heterobifunctional cross-linkers have the ability to design more specific coupling methods for binding proteins and thereby reduce the likelihood of undesirable side reactions such as homoprotein polymers. There are many heterobifunctional crosslinking agents known in the art. They include: succinimidyl 4- (N-maleimidomethyl) cyclohexane-1-carboxylate (SMCC), m-maleimidobenzoyl-N-hydroxysuccinimide ester (MBS), N-succinimidyl (4-iodoacetyl) aminobenzoate (SIAB), succinimidyl 4- (p-maleimidophenyl) butyrate (SMPB), 1-ethyl-3- (3-dimethylaminopropyl) carbodiimide hydrochloride (EDC), 4-succinimidyloxycarbonyl-a-methyl-a- (2-pyridyldithio) -toluene (SMPT), N-succinimidyl 3- (2-pyridyldithio) propionate (SPDP), succinimidyl 6- [3- (2-pyridyldithio) propionic acid ] hexanoate (LC) -SPDP). Those cross-linking agents having N-hydroxysuccinimide groups are available as N-hydroxysulfosuccinimide analogs, which typically have greater water solubility. In addition, those cross-linkers having disulfide bridges in the linking chain may be synthesized in the form of alkyl derivatives to reduce the amount of cleavage of the linker in vivo.

In addition to heterobifunctional crosslinkers, there are many other crosslinkers, including homobifunctional and photoreactive crosslinkers. In the present invention, disuccinimidyl suberate (DSS), Bismaleimidohexane (BMH), and Dimethylpimelimidate.2HCl (DMP) are examples of useful homobifunctional crosslinkers, while bis- [ beta- (4-azidosalicylamido) ethyl ] disulfide (BASED) and N-succinimidyl-6- (4 '-azido-2' -acylphenyl-amino) hexanoate (SANPAH) are examples of useful photoreactive crosslinkers. For a review of recent protein coupling techniques, see Means et al (1990) bioconjugate chemistry 1: 2-12, which are incorporated herein by reference.

A particularly useful class of heterobifunctional crosslinking agents includes those described above, which contain a primary amine reactive group, N-hydroxysuccinimide (NHS), or a water-soluble analog thereof, N-hydroxysulfosuccinimide (sulfo-NHS). At basic pH, the primary amine (lysine epsilon group) is unprotonated and can react by nucleophilic attack on NHS or sulfo-NHS esters. This reaction results in the formation of an amide bond and releases NHS or sulfo-NHS as a byproduct.

Other reactive groups that are part of the heterobifunctional crosslinking agent are thiol-reactive groups. Typical thiol-reactive groups include maleimide, halogen, and pyridyl disulfide. Maleimides react specifically with free thiol groups (cystine residues) under slightly acidic to neutral conditions (pH 6.5-7.5) within a few minutes. Halogen (iodoacetyl functional group) reacts with-SH groups under physiological pH conditions. These reactive groups result in the formation of stable thioether bonds.

The third component of the heterobifunctional crosslinker is a spacer or bridge. The bridge is a structure connecting two reactive ends. The most obvious contribution of bridges is their effect on steric hindrance. In some cases, a longer bridge may more easily span the distance necessary to connect two complex bimolecules. For example, SMPB has a span of 14.5 angstroms.

The use of heterobifunctional reagents for the preparation of protein-protein conjugates is a two-step process involving an amine reaction and a thiol reaction. For the amine reaction of the first step, the selected protein should contain a primary amine. This may be lysine epsilon amines or primary alpha amines present on the N-terminus of most proteins. The protein should not contain free sulfhydryl groups. If both proteins to be bound contain free thiols, one of the proteins may be modified with, for example, N-ethylmaleimide so that all thiols are blocked (see Partis et al (1983) J.Pro.chem.2: 263, which is incorporated herein by reference). Ellman's reagent can be used to calculate the amount of sulfhydryl groups in a particular protein (see, e.g., Ellman et al (1958) Arch. biochem. Biophys.74: 443 and Riddles et al (1979) anal. biochem.94: 75, which are incorporated herein by reference).

The reaction buffer should be free of exogenous amines and thiols. The pH of the reaction buffer should be 7.0-7.5. This pH range prevents the reaction of the maleimide group with the amine, protecting the second reaction of the maleimide group with the thiol.

Crosslinking agents comprising NHS esters have limited water solubility. Before introducing the cross-linking agents to the reaction compounds, they should be dissolved in a minimum amount of organic solvent (DMF or DMSO). The crosslinker/solvent forms an emulsion, which allows the reaction to take place.

sulfo-NHS analogs are more soluble in water and can be added directly to the reaction buffer. High ionic strength buffers should be avoided because they have a tendency to "salt out" the sulfo-NHS ester. To avoid loss of reactivity due to hydrolysis, the cross-linking agent is added to the reaction mixture immediately after the protein solution is dissolved.

The reaction is more efficient in concentrated protein solutions. The more alkaline the pH of the reaction mixture, the faster the reaction rate. The hydrolysis rates of NHS and sulfo-NHS esters also increased with increasing pH. Higher temperatures will increase the reaction rate of hydrolysis and acylation.

Once the reaction is complete, the first protein is activated with a thiol reactive group. The activated protein can be isolated from the reaction mixture by simple gel filtration or dialysis. To perform the second step of crosslinking, i.e., the thiol reaction, the lipophilic group selected for reaction with maleimide, activated halogen, or pyridyl disulfide must contain a free thiol. Alternatively, primary amines can be modified by the addition of mercapto groups.

In all cases, the buffer should be degassed to prevent oxidation of the thiol groups. EDTA may be added to chelate any oxidizing metals present in the buffer. The buffer should not contain any thiol containing compounds.

In the slightly acidic-neutral pH range (pH 6.5-7.5) maleimide reacts specifically with-SH groups. A neutral pH is sufficient for reactions involving halogen and pyridine disulfide. Under these conditions, the maleimide typically reacts with the-SH group within a few minutes. Longer reaction times are required for halogen and pyridyl disulfides.

The first thiol-reactive protein produced in the amine reaction step is mixed with thiol-containing lipophilic groups under suitable buffering conditions. The conjugate is isolated from the reaction mixture by methods such as gel filtration or dialysis.

For conjugation reactions, exemplary activated lipophilic groups include: n- (1-pyrene) maleimide, 2, 5-dimethoxystilbene-4 '-maleimide, eosin-5-maleimide, fluorescein-5-maleimide, N- (4- (6-dimethylamino-2-benzofuranyl) phenyl) maleimide, benzophenone-4-maleimide, 4-dimethylaminophenylazophenyl-4' -maleimide (DABMI), tetramethylrhodamine-5-maleimide, tetramethylrhodamine-6-maleimide, rhodamine red TM C2 maleimide, N- (5-aminopentyl) maleimide trifluoroacetate, N- (2-aminoethyl) maleimide trifluoroacetate, N- (5-dimethoxystilbene-4 '-maleimide, N- (4-dimethylamino-2-benzofuranyl) phenyl) maleimide, N- (4-benzophenone-4' -maleimide, and mixtures thereof, Oregon Green TM488 maleimide, N- (2- ((2- (((4-azido-2, 3, 5, 6-tetrafluoro) benzoyl) amino) ethyl) disulfide) ethyl) maleimide (TFPAM-SSl), 2- (1- (3-dimethylaminopropyl) -indol-3-yl) -3- (indol-3-yl) maleimide (diindolylmaleimide; GF 109203X), BODIPY _ FL N- (2-aminoethyl) maleimide, N- (7-dimethylamino-4-methylcoumarin-3-yl) maleimide (DACM), AlexaTM C5 maleimide, AlexaTM 594C 5 maleimide sodium salt, N- (1-pyrene) maleimide, N- (2-fluoro) maleimide, N- (1-fluoro) maleimide, N- (2-methyl-4-methylcoumarin-3-yl) maleimide, N- (2-dimethylaminopropyl) maleimide, N- (2-methyl-3-yl) maleimide, N- (, 2, 5-dimethoxystilbene-4 '-maleimide, eosin-5-maleimide, fluorescein-5-maleimide, N- (4- (6-dimethylamino-2-benzofuranyl) phenyl) maleimide, benzophenone-4-maleimide, 4-dimethylaminophenylazophenyl-4' -maleimide, 1- (2-maleimidoethyl) -4- (5- (4-methoxyphenyl) oxazol-2-yl) pyridium methanesulfonate, tetramethylrhodamine-5-maleimide, tetramethylrhodamine-6-maleimide, rhodamine red TM C2 maleimide, N- (5-aminopentyl) maleimide, N- (2-dimethylamino-2-benzofuranyl) phenyl, N- (N-methyl-phenyl) maleimide, N- (N-methyl-2-methyl-phenyl) pyridium methanesulfonate, N- (N-phenyl) pyridium, N- (2-aminoethyl) maleimide, N- (2- ((2- (((4-azido-2, 3, 5, 6-tetrafluoro) benzoyl) amino) ethyl) disulfide) ethyl) maleimide, 2- (1- (3-dimethylaminopropyl) -indol-3-yl) -3- (indol-3-yl) maleimide, N- (7-dimethylamino-4-methylcoumarin-3-yl) maleimide (DACM), 11H-benzo [ a ] fluorene, benzo [ a ] pyrene.

In one embodiment, the hedgehog polypeptide may be derivatized using pyrene maleimide, which is commercially available from Molecular Probes (Eugene, Oreg.), such as N- (1-pyrene) maleimide or 1-pyrene methyl iodoacetate (PMIA ester). As shown in FIG. 1, the activity curve of pyrene-derived hedgehog protein showed 2 orders of magnitude higher activity than that of the unmodified protein. B. Preparation of hydrophobic peptide derivatives

According to the invention, hydrophobic peptides can also be used to modify proteins. The term "peptide" as used herein includes a sequence of at least one amino acid residue. Preferably, the length of the peptide is between 1 amino acid and 18-26 amino acids, the latter being typical of the membrane spanning segment of the protein. To increase the hydrophobicity of the protein, the amino acids are selected mainly from the following hydrophobic amino acids: phenylalanine, isoleucine, leucine, valine, methionine, tryptophan, alanine, proline, and tyrosine. Hydrophobic peptides may also comprise non-natural amino acid analogs having hydrophobic characteristics or other characteristics of D-amino acids, peptide-like bonds, N-terminal acetylation, or other characteristics that reduce the susceptibility of the peptide to enzymatic cleavage. Methods for substituting unnatural amino acids at specific sites in proteins are known (78, 79).

Typically, hydrophobic peptides are suspended at various sites on the protein. One site may be the N-terminal residue. Alternatively, the hydrophobic peptide is substituted at the N-terminal residue. In another embodiment, the hydrophobic peptide is pendant from the C-terminus of the protein. Alternatively, the hydrophobic peptide is substituted at a C-terminal residue. The C-terminus may be the natural C-terminal amino acid, but may also be the C-terminus of the sheared protein, to suspend the hydrophobic peptide from the final C-terminal amino acid of the sheared protein, which is also referred to as the "C-terminus". The sheared hedgehog protein retains activity when up to 11 amino acids are deleted from the native C-terminal sequence. The hydrophobic peptide may also be inserted between the N-terminal residue and an internal residue immediately adjacent to the N-terminal residue, or between the C-terminal residue and an internal residue immediately adjacent to the C-terminal residue, or between two internal residues.

In a certain embodiment, the lipophilic group is an amphiphilic polypeptide, such as magainin, cecropin, attacin, maltesin, gramicin S, alpha-toxin of staph. Fusion coating proteins of viral particles may also be a conventional source of the amphipathic sequence of the hedgehog proteins of the invention. C. Preparation of lipid derivatives

Another form of protein encompassed by the present invention is a protein derivatized with various lipid groups. In general, a "lipid" is a member of hydrophobic substances of various origins, which are characterized byOrganic solvents have various solubilities, but are mostly insoluble in water. The main lipid classes encompassed by the present invention are fatty acids and sterols (such as cholesterol). The derived proteins of the invention comprise cyclic, acyclic (e.g., linear), saturated or unsaturated, monocarboxylic fatty acids. Exemplary saturated fatty acids have the general formula: CH (CH)₃(CH₂)_nCOOH. The following table lists examples of some fatty acids that can be derivatized using conventional chemical methods. Table 2: exemplary saturated and unsaturated fatty acids saturated fatty acids: CH (CH)₃(CH₂)_nCOOH

Value of N	Common name
		2	Butyric acid
4	Hexanoic acid
		6	Octanoic acid
8	Capric acid
		10	Lauric acid
12	Myristic acid*
		14	Palmitic acid*
16	Stearic acid*
		18	Arachidic acid*
20	Behenic acid
		22	Tetracosanoic acid

Unsaturated fatty acid:

CH3CH=CHCOOH	crotonic acid
		CH3(CH2)3CH=CH(CH2)7COOH	Myristicin acid*
CH3(CH2)5CH=CH(CH2)7COOH	Palmitoleic acid*
		CH3(CH2)7CH=CH(CH2)7COOH	Oleic acid*
CH3(CH2)3(CH2CH=CH)2(CH2)7COOH	Linoleic acid
		CH3(CH2CH=CH)3(CH2)7COOH	Linolenic acid
CH3(CH2)3(CH2CH=CH)4(CH2)3COOH	Arachidonic acid

Asterisks indicate the fatty acids we found in the recombinant hedgehog protein secreted by the soluble structure.

Other lipids that can be attached to proteins include branched chain fatty acids as well as phospholipid-based fatty acids such as phosphatidylinositol (e.g., phosphatidylinositol 4-monophosphate and phosphatidylinositol 4, 5-diphosphate), phosphatidylcholine, phosphatidylethanolamine, phosphatidylserine, and isoprenoids such as farnesyl or geranyl.

We have demonstrated that lipid-modified hedgehog proteins can be further purified from natural sources, or can be obtained by chemically modifying soluble and unmodified proteins. For proteins purified from natural substances, we show that if full-length human sonic hedgehog (Shh) is expressed in insect cells and membrane-bound Shh is purified from detergent-treated cells using a combination of SP-Sepharose chromatography and immunoaffinity chromatography, the purified protein moves in one distinct band on a reducing SDS-PAGE gel with an apparent mass of 20kDa (see example 1). Soluble, membrane-bound Shh proteins are readily discernible by reverse phase HPLC, while the bound form is subsequently eluted in an acetonitrile gradient (see example 1 and fig. 3). We also show that human sonic hedgehog binds to the cell membrane in two forms, one containing cholesterol and thus resembling the data previously reported for Drosophila hedgehog (18), while the second novel form contains cholesterol and a palmitic acid modifying group. The soluble and bound form of Shh was analyzed by electron-jet mass spectrometry using a triple quadrupole mass spectrometer equipped with an electron-jet ion source (example 1) and also by liquid chromatography-mass spectrometry (see example 1). The identification of the N-terminal peptide in bound Shh digested by endoproteinase Lys-C can be confirmed by measuring MALDI time of a flight mass spectrometer by MALDI PSD mass spectrometry. The site of palmitoylation was identified by a combination of peptide mapping and sequence analysis, and was at the N-terminus of the protein (residue 1 of the mature protein sequence in SEQ ID NOS: 1-4). The activity of both binding forms in the C3H10T 1/2 alkaline phosphatase assay was identical, but interestingly, both were about 30-fold more potent than soluble human Shh lacking the binding group. Lipid modification had no significant effect on the apparent binding affinity of Shh to its receptor patched (figure 7).

We next examined the utility of the derivative forms by assaying soluble and bound Shh on C3H10T 1/2 cells for alkaline phosphatase induction alone or in the presence of anti-hedgehog neutralization Mab5E 1. In addition, the relative potency of soluble and bound Shh to bind patched was assessed by FACS analysis on patched transfected EBNA-293 cells (example 3).

For lipid-modified proteins obtained by chemically modifying soluble and unmodified proteins, we have shown that palmitic acid and other lipids can be added on soluble Shh to produce lipid-modified forms with greater potency in the C3H10T 1/2 experiment (example 8). We have shown (examples 1, 2 and 8) that thiol and a-amine on the N-terminal cystine contribute to the lipid derivatization reaction. Without wishing to be bound by any particular theory, lipid modification on proteins begins with the formation of a thioester intermediate, and the lipid group is then transferred to the N-terminal α -amine by forming a cyclic intermediate. Typically, the reactive lipid group may thus be a thioester form of a saturated or unsaturated carboxylic acid, such as a coenzyme a thioester. Such substances and their derivatives may include, for example, commercially available coenzyme A derivatives such as palmitoyl oleoyl coenzyme A, eicosanoyl coenzyme A, eicosatetraenoyl coenzyme A, lauroyl coenzyme A, and the like. These materials are readily available from Sigma chemical company (St. Louis, MO., 1998 catalog pp.303-306).

The effect of different lipid groups on the functional activity of hedgehog proteins has been analyzed (see example 8 and figures 10 and 11). Similarly, the effect of different lipid groups on the functional activity of other proteins as described in section iii above can be conveniently tested using methods known to those skilled in the art. For example, functional testing of gelsolin (50), various interferons (interferon- α, interferon- β, and interferon- γ), various self-interleukins (e.g., IL-1, -2, -3, -4, etc.), tumor necrosis factors- α and- β, and other growth factors, which are lipid modified according to the present invention, have been accomplished using methods known in the art.

While we have established a chemical method by which fatty acids can be attached to the N-terminal cystine of the hedgehog protein, it is contemplated that lipids can be attached at the same or other sites using enzymatic reactions. In vivo palmitoylation of proteins is mediated by a class of enzymes called palmitoyl-coa: the enzyme of protein S-palmitoyltransferase. In vitro acylation of protein substrates has been demonstrated using purified enzymes (80, 81). The substrate specificity of palmitoyl transferase has not been completely defined; analysis of the palmitoylation sites of cellular and viral proteins was almost absent in the consensus sequence surrounding the modified cystine residues, but indicated the co-presence of lysine or arginine residues within the two amino acids of cystine as well as large and hydrophobic amino acids close to cystine. The amino-terminal sequence CGPGRGFG of Shh matches this consensus sequence and serves as a recognition site for palmitoylation.

As an alternative, myristoylation of the amino terminus of the hedgehog protein can be performed using N-myristoyl transferase (NMT), many of which have been characterized in mammals (82) and yeast (83). The recognition site for N-myristoyl acyltransferase may be introduced into the hedgehog N-terminal sequence to facilitate recognition by the enzyme. These strategies require the use of fatty acyl-coa derivatives as substrates, as described in example 8 for non-enzymatic lipoylation of human sonic hedgehog. In addition, proteins with introduced recognition sequences can be myristoylated when expressed in a suitable cell line. Other methods of modifying proteins such as hedgehog with hydrophobic groups are to add recognition sites to add isoprenoid groups to the C-terminus of the protein. The recognition sites for farnesyl and geranyl-geranyl additions are known and proteins can be modified when expressed in eukaryotic cells (Gelb et al, Cur. Opin. chem. biol. 2: 40-48 (1998)). VI, polymeric protein compound

The hydrophobically modified proteins described herein are particularly suitable for formulation into multimeric protein complexes. The multimeric protein complexes of the invention comprise proteins attached to vesicles, optionally via their hydrophobic groups (e.g. lipids). The vesicles may be naturally occurring biofilms, purified from natural substances, or synthetic structures. Preferred vesicles are spherical structures made substantially of amphiphilic substances such as surfactants or phospholipids. The lipids of these spherical vesicles are typically organized in a lipid form having one or more structural layers, such as multilamellar vesicles (lipid bilayers of onion-shaped shells, with aqueous volumes contained between the bilayers) or micelles.

In particular, liposomes are small spherical vesicles consisting essentially of various types of lipids, phospholipids, and auxiliary lipophilic components. These components are typically arranged in a bilayer format, similar to the lipid arrangement of biological membranes.

Typically, the polar end of a constituent lipid or lipoid molecule is contacted with a surrounding solution, usually an aqueous solution, while the non-polar hydrophobic end of the lipid or lipoid molecule is contacted with the non-polar hydrophobic end of another lipid or lipoid molecule. The bilayer membrane formed (e.g., vesicles) is selectively permeable to molecules having certain sizes, hydrophobicity, shapes, and net charges. Most vesicles are lipid or lipoidal in nature, but other liposomal bilayer structures exist, including surfactants as well as lipids or cholesterol.

Liposomal vesicles are particularly preferred because they are useful in many therapeutic, diagnostic, industrial, and commercial applications. They can be used to transport molecules that are not readily soluble in water or when timed release is desired. Because they are selectively permeable to many chemical compounds, liposomes are very useful as drug carriers or biomaterials. Thus, lipid-derived proteins such as hedgehog can be made multimeric by incorporation into the lipid bilayer of the liposome vesicles. Upon reaching the target site, the liposomes may be degraded (e.g., by enzymes in the gastrointestinal tract) or they may fuse with the membrane of the cell.

Several methods are known for preparing vesicles such as liposomes. Methods for making phospholipid vesicles are well known (53). For a review of the methods of conventional use see (54). The more commonly used method is (1) sonication of a lipid-containing solution, sometimes followed by evaporation/lyophilization and rehydration (see, e.g., Stryer, Biochemistry, pp. 290-291, Freeman & Co., New York, (1988), and (55)); (2) homogenizing the lipid solution, sometimes under high pressure or high shear (see, e.g., U.S. patent No. 4,743,449 issued 5/10 1988 and U.S. patent No. 4,753,788 issued 6/28 1988); (3) hydrating and sometimes sonicating the vesicles to form a dry film of lipids, wherein the lipid film is made by evaporating a lipid solution dissolved in an organic solvent (see, e.g., U.S. patent nos. 4, 452, 747 issued on 5.6.4.198, 4, 895, 719 issued on 23.1.1990, and 4, 946, 787 issued on 7.8.1990); (4) lyophilization or evaporation and rehydration (see, e.g., U.S. patent No. 4,897,355 issued at 30.1.1990, EP267 published at 5.11.1988, 050, U.S. patent No. 4,776,991 issued at 11.10.1988, EP172,007 published at 19.2.1986, and australian patent application No. AU-a-48713/85 published at 24.4.1986); (5) solvent injection or infusion of lipid solutions in aqueous media or vice versa (see, e.g., (56), U.S. Pat. No. 4,921,757 issued 5/1 1990, U.S. Pat. No. 4,781,871 issued 11/1 1988, WO 87/02396 issued 3/24 1988, and U.S. Pat. No. 4,895,452 issued 1/23 1990); (6) spray drying (see, e.g., australian patent application No. AU-a-48713/85, published 24.4.1986, and U.S. patent No. 4,830,858, issued 16.5.5.1989); (7) filtration (see, e.g., WO 85/01161); (8) reverse phase evaporation (see, e.g., (57)); and (9) combinations of the above methods (see, e.g., (58) and (59)).

When used to prepare vesicles, preferred lipid or lipid components include phospholipids, mixtures of phospholipids, basic lipids, neutral lipids, fatty acids, and derivatives thereof. Preferred lipids have the following characteristics: when dispersed alone in water, they are lipid emulsion phases at temperatures above the lipid transition temperature. In certain embodiments, the lipid is a single fatty chain of more than about 12 carbon atoms, and may be saturated or unsaturated, or substituted. Suitable lipids include esters, alcohols, and acid forms of the following fatty acids: stearic acid, oleic acid, linoleic acid, arachidic acid, arachidonic acid, and other single chain fatty acids. Additional candidates are esters, alcohols, and acid forms of retinoids, particularly retinol and retinoic acid. Other preferred lipids include Phosphatidylcholine (PC), Phosphatidylglycerol (PG), and derivatives thereof, which may be synthetically produced or derived from a variety of natural substances.

In certain embodiments, the vesicles may be sterically commonly incorporated with polyethylene glycol (PEG) or stabilized by a synthetic PEG head group attached to Distearoylphosphatidylethanolamine (DSPE), see, e.g., (61). Preferred surfactants are those with good miscibility, such as Tween^TM、Triton^TM. Sodium Dodecyl Sulfate (SDS), sodium lauryl sulfate, or sodium octyl glycoside.

Preferred surfactants form micelles when added to the aqueous solution above the phase inversion temperature of the surfactant. The surfactant may be comprised of one or more fatty chains. These fatty chains may be saturated, unsaturated, or otherwise substituted, such as ethoxylated, with typical fatty chains containing greater than about 12 carbon atoms. Other suitable surfactants include the following: lauryl-, myristyl-, linoleyl-, or stearyl-sulfobetaine; lauryl-, myristyl-, linoleyl-, or stearyl-sarcosine; linoleyl-, myristyl-, or cetyl-betaine; lauramidopropyl-, cocamidopropyl-, linoleamidopropyl-, myristamidopropyl-, palmitoamidopropyl-, or isostearamidopropyl-betaine (such as lauramidopropyl); myristamidopropyl-, palmitoamidopropyl-, or isostearamidopropyl-dimethylamine; methyl cocoyl or methyl oleyl-sodium taurate; and the MONAQUAT series (Mona Industries, inc., Paterson, n.j.). See example 4.

Preferred sterols and sterol esters suitable for use in preparing the multimeric protein complex include cholesterol, cholestanols, cholesterol sulfates, and other cholesterol analogs and derivatives. The fact that vesicles may include many different lipids and detergents allows for great variability in engineering binding protein-vesicle complexes with desirable properties. For example, one can make vesicles that bind different types of proteins by changing the composition of the lipids in the starting material to produce larger vesicles, or by increasing the percentage of phosphatidylinositol lipids in the vesicles. VII, practicality

In general, the modified proteins described herein are useful in the treatment of some conditions that can be treated with the unmodified protein. However, the hydrophobically modified proteins described herein have several significant improvements over the unmodified form. First, their increased potency allows for treatment with smaller amounts of protein in a shorter time. This is very important for both systemic and CNS applications. Second, the replacement of the N-terminal cystine with a less chemically reactive amino acid will make the protein easier to manufacture, formulate and store for clinical use. Third, the pharmacokinetics of the protein will be altered by hydrophobic modifications which will localize the protein to the vicinity of the site of administration, thereby increasing safety by minimizing systemic exposure and increasing efficacy by increasing local concentration. The proteins of the invention may also be used in diagnostic compositions and methods for detecting their corresponding receptors.

As an example of the first point, it has been found that hedgehog has a very short half-life following systemic application and requires multiple injections to achieve a strong response to the protein. The higher potency of the modified form and the possibility of formulation into a formulation in liposomes allow the response to be achieved with less treatment. For CNS applications, higher potency can provide sufficient amounts of protein in a small volume necessary for direct injection into the CNS.

The importance of the second point is illustrated by the following facts: we have found that the N-terminal cystine of hedgehog is very susceptible to chemical attack, either by forming other chemical adducts or by oxidative dimerization with other hedgehog proteins. To prevent this, special buffers and procedures are required during purification and dithiothreitol is required in the final formulation. These measures make it necessary to carefully evaluate the action of the formulation buffer in a substrate model.

As a third example, the more limited the extent of outward diffusion of the protein from the site of administration, the higher the local concentration. This higher local concentration can therefore produce a more specific clinical response after direct injection into the desired brain or spinal area during treatment of neurological diseases.

Similarly, the modified proteins may be administered topically to the fracture site to aid in the healing of the fracture, to the gonads to treat fertility disorders, intraocularly to treat eye disorders, and subcutaneously to treat skin conditions and stimulate local hair growth. The hydrophobically modified proteins are localized to the site of administration and thus reduce the likelihood of undesirable systemic contact with other tissues and organs.

For therapeutic applications, the hydrophobically modified proteins of the present invention are placed in a pharmaceutically acceptable sterile isotonic formulation and may optionally be administered by standard methods known to those skilled in the art. The formulation is preferably in liquid form, or may be a lyophilized powder. It is envisioned that therapeutic administration of multimeric protein complexes may include liposomes incorporating derivatized proteins as described herein.

It will be appreciated by those of ordinary skill in the art that the particular administration, dosage, and clinical use of the hydrophobically modified proteins of the present invention will vary depending on the particular protein and its biological activity.

As an example of therapeutic applications of the proteins of the present invention, therapeutic hydrophobically modified hedgehog proteins can be administered to patients suffering from various neurological disorders. The ability of hedgehog proteins to regulate neuronal differentiation during nervous system development and presumably in adults suggests that it is reasonable to believe that hydrophobically modified hedgehog is beneficial in controlling maintenance, functional performance of human neurons, and the aging of normal cells; repair and regeneration of damaged cells; and to prevent degeneration and premature death due to loss of differentiation in certain pathological conditions. Thus, the hydrophobically modified hedgehog compositions of the present invention can prevent and/or reduce neurological disorders resulting from: acute, subacute, or chronic damage to the nervous system, including trauma, chemical injury, vascular injury, and ischemia (e.g., ischemia resulting from stroke), as well as infection and tumor-induced injury; (ii) aging of the nervous system, including Alzheimer's disease; (iii) chronic neurodegenerative diseases of the nervous system including Parkinson's disease, Huntington's disease, amyotrophic lateral sclerosis, etc.; and (iv) immune diseases of the nervous system, including multiple sclerosis. The hydrophobically modified proteins can also be injected into the cerebral medullary fluid, for example, against deficiencies in brain cells, or for example, into the lymphatic system or blood as required for other tissue or organ system specific diseases.

The hedgehog compositions of the invention are useful in rescuing, for example, various neurons from injury-induced death and in guiding these neurons to regenerate following such injury. The damage may be attributed to (but is not limited to) the following: CNS traumatic infarction, infection, metabolic disease, nutritional deficiencies, and toxic drugs (e.g., cisplatin treatment). Certain hedgehog proteins can cause neoplastic or proliferative deformed cells to become post-mitotic or apoptotic. Thus, such compositions are useful for treating, for example, malignant tumors, medulloblastoma and neuroectodermal tumors.

The protein may be linked to a detectable label, such as a fluorescent or radio-opaque substance, and administered to a host to discover tissues that may express the hedgehog receptor. The protein may also be bound to a substance, such as horseradish peroxidase, which can be used as an immunocytochemical dye to visualize the area of hedgehog ligand-positive cells on a tissue portion. The hydrophobically modified proteins of the invention can be used alone or in the form of multivalent protein complexes for the medical treatment of cancers and tumors specifically directed against receptors expressing the protein. These substances can be made more effective as cancer therapeutic agents by using them as transport carriers for antitumor drugs, toxins, and cytotoxic radionuclides such as yttrium 90.

Toxins may also bind to hydrophobically modified hedgehog (or vesicles containing multivalent complexes thereof) to selectively target and kill hedgehog-responsive cells, such as tumors that express hedgehog receptors. Other toxins are also useful as would be known to one skilled in the art. Such toxins include, but are not limited to: pseudomonas exotoxin, dipheria toxin, and saporin. This approach has proven successful because hedgehog receptors are expressed in a very limited number of tissues. Other methods for such treatments are the use of radiolabeled hydrophobically modified proteins (or multivalent complexes thereof). Such radiolabeled compounds will preferably target the radioactive site in cells expressing the protein receptor, sparse normal tissue. Depending on the radioisotope used, radiation emitted by a radiolabeled protein bound to tumor cells may also kill nearby malignant tumor cells that do not express protein receptors. Various radionuclides may be used. Radioactive iodine (e.g. iodine)¹³¹I) A monoclonal antibody against CD20 present on B cell lymphomas has been successfully used (63).

The protein compositions used in treatment may be formulated into various formulations and dosages depending on the medical practice in consideration of the disease to be treated, the condition of the individual patient, the site of administration of the isolated polypeptide, the method of administration, and other factors known to those skilled in the art. Therapeutic agents may be prepared by mixing a protein, protein-containing vesicle, or derivatized complex of the desired purity with a physiologically acceptable carrier, such as one that is non-toxic to recipients at the dosages and concentrations employed.

It is envisioned that topical administration to a site is the primary route of therapeutic administration of the proteins of the present invention. Intravenous administration or administration via a cannula or other surgical tube is also conceivable. Other routes include tablets and the like, commercially available sprays for liquid formulations, and inhalation lyophilized or aerosolized formulations. The liquid formulation may be used after reconstitution from a powder formulation.

The dosage administered depends on the nature of the vesicles and proteins used, such as the binding activity and in vivo half-life, the concentration of vesicles and proteins in the formulation, the route of administration, the site and rate of administration, the clinical tolerability of the patient, the pathological conditions the patient is exposed to, and other factors known to those skilled in the art. In general, it is preferred that each administration be at a dose of about 5X 10 per patient^-7To 5X 10^-9M, but the dose depends on the nature of the protein. Different doses may be used during a series of consecutive administrations.

The invention also relates to a pharmaceutical formulation comprising a hedgehog protein modified according to the invention and a pharmaceutically acceptable carrier. In one embodiment, the formulation may also include a vesicle.

The hydrophobically modified hedgehog proteins of the invention are also useful in gene therapy methods.

Several animal models are known to have clinical predictive value for neurodegenerative diseases. For parkinson's disease, the model involves protection or restoration in rodents or primates in which the substantia nigra-striatum dopaminergic channels are destroyed by systemic MPTP administration or by topical (intracranial) administration of 6-hydroxydopamine (6-OHDA), which are two selective dopaminergic toxins. The concrete model is as follows: MPTP-treated murine model (64); MPTP-treated primate (marmoset or Rhesus) model (65); and a unilateral 6-OHDA lesion murine model (66). For ALS (amyotrophic lateral sclerosis), the model involves the treatment of several murine lines that exhibit spontaneous motor neuron degeneration, including the wbler (67) and pmn mice (68), and the treatment of transgenic mice expressing the human mutant superoxide dismutase (hSOD) gene linked to familial ALS (69). For spinal cord injury, the most common model involves contusion injury in mice, either by standard weight loss or fluid (hydrodynamic) injury (70). For huntington's disease, the model involves protection of the mouse striatum from damage by excitotoxins (NMDA, quinolinic acid, kainic acid, 3-nitro-propionic acid, APMA) (71, 72). Recently, transgenic mouse models have also been described that overexpress human trinucleotides that are repeatedly expanded in the huntingtin gene (73). For multiple sclerosis, EAE in mice and rats is induced by immunization with MBP (myelin basic protein) or by passive transfer of T cells activated with MBP (74). For alzheimer's disease, the relevant murine model is an assay for protection against umbrella-fornix lesions (septal lesions) in mice, the use of the main nerve bundle supplying the cholinergic innervation of the hippocampus (75) and transgenic mice overexpressing the human beta-starch gene. For peripheral neuropathy, the relevant model is the loss of peripheral nerve conductivity in protected mice and rats caused by chemotherapeutic agents such as paclitaxel, vincristine, and cisplatin (76).

The invention will now be described in more detail with reference to the following non-limiting examples. Example 1: lipid-modified A of human Sonic hedgehog when expressed in insect cells, expression of human Sonic hedgehog

The cDNA of full-length human Sonic hedgehog (Shh) is subcloned in pBluescript SK⁺(20) (supplied by David Bumcrot, Cambridge MA, oncogene, inc.) 1.6 kb ecori fragment. 5 ' and 3 ' NotI sites were added by unique site-eliminating mutagenesis using the Pharmacia kit according to the manufacturer's recommendations, which site flanked by Shh open reading frames. The 1.4 kb NotI fragment carrying the full length Shh cDNA was then subcloned into the insect expression vector pFastBac (Life Technologies, Inc.). Recombinant baculovirus was produced using the procedure provided by Life Technologies, inc. For the resulting virusTo produce a high titer virus feed solution. The method for preparing and purifying Shh is described below. The presence of membrane associated Shh was checked by FACS and Western blot analysis. Peak expression occurred 48 hours post infection. For Western blot analysis, supernatants and cell lysates from Shh infected or uninfected cells were subjected to SDS-PAGE under reducing conditions on 10-20% gradient gels, electrophoretically transferred to nitrocellulose, and Shh was detected using murine polyclonal antiserum raised against N-terminal Shh15 polypeptide-keyhole limpet hemocyanin conjugates. Cell lysate was purified by passage over 20mM Na₂HPO₄pH6.5, 1% Nonidet P-40 and 150mM NaCl or 20mM Tris-HCl pH8.0, 50mM NaCl, 0.5% Nonidet P-40 and 0.5% sodium deoxycholate in 25 ℃ temperature incubation of cells for 5 minutes to prepare, then in Eppendorf 13000rpm and 4 ℃ centrifugal 10 minutes to precipitate particles. B. Purified membrane bound human Sonic hedgehog

In High Five^TMA recombinant baculovirus encoding full-length Shh as described above was used in insect cells (Invitrogen) to make a membrane-bound form of Shh. High Five^TMCells were grown in sf900 II serum medium (Life Technologies, Inc.) in a 10L bioreactor with controlled oxygen content at 28 ℃. Cells were cultured at approximately 2X 10 in the late log phase⁶Cells/ml were infected with virus at an MOI of 3 and then harvested 48 hours post-infection (cell survival at harvest > 50%). Cells were harvested by centrifugation and incubated at 10mM Na₂HPO₄pH 6.5, 150mM NaCl pH + 0.5 mM PMSF. The resulting cell pellet (150g wet weight) was suspended in 1.2L of: 10mM Na₂HPO₄pH6.5, 150mM NaCl, 0.5 mM PMSF, 5. mu.M pepstatin A, 10. mu.g/ml leupeptin and 2. mu.g/ml E64, and then 120ml of a 10% solution of Trition X-100 was added.

After incubation on ice for 30 min, the pellet was removed by centrifugation (1500 × g, 10 min). All subsequent steps were carried out at 4-6 ℃. The supernatant was adjusted to pH 5.0 with 0.5M MES pH 0.5 stock solution (50mM final), and then loaded on 150ml SP-Sepharose Fast Flow column (Pharmacia). The column was packed with 300ml of 5mM Na₂HPO₄pH5.5, 150mM NaCl, 0.5 mM PMSF, 0.1% Nonidet P-40, and then 200ml of 5mM Na₂HPO₄Washed at pH 5.5, 300mM NaCl, 0.1% Nonidet P-40, and then with 5mM Na₂HPO₄Bound hedgehog was eluted at pH 5.5, 800mM NaCl, 0.1% Nonidet P-40.

Shh was followed by immunoaffinity chromatography on 5E1-Sepharose resin prepared by binding 4mg of antibody per ml of CNBr-activated Sepharose 4B resin. The SP-Sepharose elution pool was diluted with 2 volumes of 50mM HEPES pH 7.5, and the material was then loaded on 5E1 resin (1 hour). The resin was collected on a column, washed with 10 column volumes of PBS containing 0.1% hydrogenated Trition X-100(Calbiochem), and then with 25mM NaH₂PO₄pH3.0, 200mM NaCl, 0.1% hydrogenated Triton X-100. The eluted fractions were neutralized with 0.1 volume of 1M HEPES pH7.5, and then analyzed for total protein content by absorbance measurement at 240-340nm and purity by SDS-PAGE. The components were stored at-70 ℃.

The peak fractions from the three affinity steps were pooled, diluted with 1.3 volumes of 50mM HEPES pH 7.5, 0.2% hydrogenation Trition X-100, and the material was then loaded onto 5E1 resin. The resin was collected on a column, washed with 3 column volumes of PBS pH 7.2, 1% octyl glucoside (USBiochemical Corp.), then 25mM NaH₂PO₄pH3.0, 200mM NaCl, 1% octyl glucoside elution. The eluted fractions were neutralized and analyzed as described above, pooled, filtered through a 0.2 micron filter, aliquoted and stored at-70 ℃.

When the full-length human sonic hedgehog (Shh) is expressed in High Five^TMIn insect cells, more than 95% of the N-terminal fragments were detected in cell-bound form by Western blotting. The purified protein moved in a single distinct band by SDS-PAGE, with an apparent mass of 20kDa (FIG. 1, lane c). This protein moves about 0.5 kDa faster than the soluble protein produced in E.coli (FIG. 1, lanes b-d), which is consistent with the previously published data-Thus (19). Similar as described in (19), soluble and membrane-bound Shh proteins can also be readily distinguished by reverse phase HPLC, where the bound form is eluted later in an acetonitrile gradient. The concentration of acetonitrile required to elute the membrane bound form was 60%, while for the soluble form it was only 45%. This indicates a significant increase in the hydrophobicity of the protein. C. Mass spectrometric analysis of Membrane-bound human Sonic hedgehog

Equal portions of Shh at C₄Reverse phase HPLC was performed on a column (Vydac, Cat. No. 214TP104, column size 4.6 mm ID. times.250 mm) at room temperature. The bound fractions were eluted with a 0-80% gradient of acetonitrile in 0.1% trifluoroacetic acid for 30 minutes at a flow rate of 1.4 ml/min. The column eluate was monitored at 280nm and fractions of 0.5 min were collected. A25 μ L aliquot of the protein-containing fraction was dried in a Speed Vac concentrator, dissolved in electrophoresis sample buffer and then analyzed by SDS-PAGE. The hedgehog containing fractions were pooled, concentrated 4-fold in a Speed Vac concentrator and analyzed for protein content by absorbance at 280nm with a Shh solution of 1 mg/ml using an extinction coefficient of 1.33. Samples were subjected to ESI-MS on a Micromass Quattro II triple quadrupole mass spectrometer equipped with an electron-jet ion source. A volume of 6. mu.l of HPLC-purified hedgehog was directly diffused in the ion source at a flow rate of 10. mu.l/min using 50% water, 50% acetonitrile, 0.1% formic acid as solvent in a syringe pump. Scanning was performed across the sample spread. All electrospray mass spectral data were obtained and stored in file format and then processed using the micromass masslynx data system.

The peptides generated by digestion of pyridylethylated Shh with endoproteinase Lys-C were analyzed on-line by reverse phase HPLC with a Micromass Quattro II triple quadrupole mass spectrometer. In Reliasil C₁₈Using Michrom on column^TMDigesta were separated on an ultrafast micro protein analyzer system with a flow rate of 50 μ l/min and eluted with a gradient of 5-85% acetonitrile in 0.05% trifluoroacetic acid. The whole range was scanned by m/z 400-2000 and processed as described above.

Sequencing of the N-terminal peptide from binding to Shh in Voyager-DE^TMSTR (PerSeptiveBiosystems, Framingham, MA) time of flight (TOF) mass spectrometry was performed by Post Source Decay (PSD) measurement using α -cyano-4-hydroxycinnamic acid as matrix (22, 23). Precisely 0.5. mu.l of the HPLC purified endoproteinase Lys-C peptide was mixed with 0.5. mu.l of matrix on the target plate. To cover the entire spectral range of the fragment ions, the mirror voltage was reduced from 20 to 1.2 kV in 11 steps.

Electrospray ionization mass spectrometry data for Shh in both soluble and membrane-bound form showed masses of 19560 and 20167Da, respectively, for the starting material (fig. 2). The measured mass of 19560Da matched the expected mass of Shh starting with Cys-1 and ending with Gly-174 (calculated mass 19560.02 Da). In contrast, the 20167Da mass is neither consistent with what is predicted, nor the difference 607Da between the bound and soluble form masses can be explained by any known modification or enzymatic process. Previously, Porter et al (18) demonstrated that Drosophila hedgehog contains cholesterol groups and therefore the mass difference in the human system is likely due at least in part to cholesterol (the calculated mass of esterified cholesterol is 368.65 Da). The presence of a minor component in the bound Shh mass spectrum at 19796Da, which differs by 371Da from the main peak, supports this conclusion.

Further evidence for cholesterol is obtained by treating the bound Shh with a medium base under conditions that can cleave the cholesterol linkage without breaking peptide bonds (18), and then reanalyzing the reaction products by Mass Spectrometry (MS). Briefly, insect cell-derived Shh was treated with 50mM KOH, 95% methanol at room temperature for 1 hour, then analyzed with ESI-MS or digested with endoproteinase Lys-C, and LC (liquid chromatography) -MS performed on a Micromass Quattro II triple quadrupole mass spectrometer. For samples subjected to LC-MS, the protein was first treated with 4-vinylpyridine. The alkaline treatment shifted the mass by 387Da, which is consistent with the loss of cholesterol with water (see table 3). The quality of soluble Shh was not affected by the alkali treatment. Together these observations suggest that membrane-bound human Shh contains two modifications, cholesterol and a second group with a mass of 236 Da. The similarity between this value and the mass of the added palmitoyl group (238Da) suggests that eggs are producedWhite may be palmitoylated. More accurate mass calculations indicate correlations within the expected mass of 0.1 Da palmitoyl groups, as described below. Table 3: binds to the MS characteristic of Shh. The calculated mass values were determined using the average residue mass in part a and the monoisotopically protonated mass in part b.

Protein	Mass (Da)
			Computing	Measuring
	KOH treated Shh
Unbound (-treatment)				19560．02	19560
Unbound (+ treatment)	19560．02	19561
Combined (-treatment)	20167．14	20167
Combination (-treatment)	19798．49	19780
			N-terminal Endonuclease Lys-C peptide (MH)+)*
Unbound	983．49	983．50
Combined with	1221．72	1221．79

All mass values described herein are protonated masses.

We subsequently determined that it is possible to bind Shh by HPLC fractionation in subspecies with varying elution gradients and developed a simple HPLC analysis to quantify the various forms. The results of these analyses are shown in FIG. 3. In this analysis, unmodified Shh elutes first (peak 1), then cholesterol-modified Shh elutes (peak 2), and finally the product comprising cholesterol and palmitic acid-modified Shh elutes (peak 3). The complex shape of peak 3 reflects the presence of the palmitoyl modified form, which was identified by sequencing by maldi psd measurement. The change was 2Da, less than predicted, and therefore likely to contain unsaturated bonds (data not shown). D. Localization of palmitic acid modifications in human Sonic hedgehog sequences

The site of palmitoylation in the human sequence was identified by a combination of peptide mapping and sequence analysis. FIG. 4B shows the results obtained from peptide mapping analysis of soluble proteins using LC-MS readings. Mass data for more than 98% of soluble Shh sequences can be calculated from this analysis. The peaks marked with an asterisk correspond to the N-terminal peptide (residues 1-9 plus 4-vinylpyridine, observed mass 983.50 Da, calculated mass 983.49 Da, Table 3). In the analysis of the corresponding binding product (FIG. 4A), the peptide did not, instead, be a more hydrophobic peptide with a mass of 1221.79 Da (marked with an asterisk).

The 1221.79 Da group is consistent with the presence of a modified form of the N-terminal peptide, i.e., 983.49 Da corresponds to the peptide composition plus 238.23 Da. The 1221.79 Da peptide was then subjected to sequence analysis by MALDI PSD measurement. The resulting PSD spectrum is shown in fig. 5. Detected to correspond to b1, b2, b4, b5, b8+ H₂O, y8, y7, y5, y4, y3, y2, y1 fragments. In addition, the b1 and b2 ions indicate that the pyridylethylated Cys-1 adduct is palmitoylated. Only the ions containing Cys-1 contained the added mass of 238.23 Da.

Since cystine is the normal site of palmitoylation in proteins, it is not surprising to find a novel adduct attached to the N-terminal cystine. However, two lines of evidence suggest that lipids are attached to the amino group of cystine, rather than to the thiol. First, in peptide mapping studies, we used 4-vinylpyridine as a spectroscopic marker to monitor free thiol groups (72). Pyridylethylation is highly specific for cystine thiols and adds a 105Da adduct, which can be detected by MS. The Cys-1 containing fragment observed in the PSD spectrum included both palmitoyl and pyridylethyl modifications, indicating the presence of free thiol groups. Second, automated N-terminal Edman sequencing was performed in conjunction with Shh, but no sequence was obtained, suggesting a block at the N-terminal α -amine. Conversely, the corresponding soluble form of Shh can be readily sequenced. Example 2: human Sonic hedgehog can be modified with palmitic acid in a cell-free system

By using³H-palmitic acid labels soluble Shh in a cell-free system using a modified version of the published procedure (24). The liver homogenate was centrifuged at 3000 Xg for 10 minutes, 9000 Xg for 20 minutes, and 100000 Xg for 30 minutes in this order, thereby separating a crude microsomal fraction from the liver of mice. 100000 Xg of the pellet was suspended in 10mM HEPES pH 7.4, 10% sucrose and centrifuged again at 100000 Xg for 20 minutes. The final pellet (from 10g liver) was suspended in 3ml of 20mM Tris-HCl pH 7.4, 150mM NaCl, 1mM EDTA, 10. mu.g/ml leupeptin, 0.15% Trition X-100, aliquoted and stored at-70 ℃. The reaction contained 3. mu.g of Shh, 1. mu.l of mouse microsomes, 50 ng/ml of coenzyme A (Sigma), 0.3 mM ATP, 20mM Tris-HCl pH 7.4, 150mM NaCl, 1mM EDTA, 10. mu.g/ml leupeptin, and 0.5. mu. Ci- [9, 10-³H]Palmitic acid (50 Ci/mmol; New England Nuclear), the reaction being carried out at room temperature for 1 hour. As the run-down buffer reaction stopped, SDS-PAGE was performed on 10-20% gradient gels followed by visual inspection with fluorescence.

As shown in FIG. 1 (lane e), Shh is readily labeled with a radioactive tracer. None of about 1 hundred other proteins in the reaction mixture was labeled (see the corresponding coomassie blue stained gel pattern in lane j), indicating a high specificity of the palmitoylation reaction. Further evidence for the specificity of the palmitoylation reaction was that we tested two Shh variants in which the site of palmitoylation had been eliminated. FIG. 1 (lane f) shows the results of an analysis of a truncated form of soluble Shh that lacks the first 10 amino acid residues of the mature sequence, while lane g is a mutant of Shh that contains a single Cys-1 to Ser point mutation at the N-terminus. None of these variants is labeled.

The importance of the N-terminal cystine as a site for lipid derivatization has been further confirmed by the fact that: wild-type soluble Shh is easily labeled, while N-terminal cystine to serine mutants are not labeled. The inability to label the N-terminal serine mutant led to the debate on the simple reaction mechanism in which the palmitoyl group was directly linked to the N-terminal alpha-amine, since serine should be substituted for cystine under the conditions tested.

We also examined the effect of the free N-terminus using soluble Shh with an N-terminal histidine (His) -tag protrusion. The soluble Shh used in these studies was first made as a His-tagged fusion protein with an enterokinase cleavage site at the junction of the mature sequence, and then treated with enterokinase to remove the His-tag. His-tagged Shh was not palmitoylated despite the presence of free thiol groups of cystine (see FIG. 1, lane i). Although we cannot exclude the possibility that the N-terminal overhang sterically inhibits palmitoylation, Cys-1 is in the PI, position of the enterokinase cleavage site and is readily accessible for enzymatic processing. Thus, it appears that both the thiol and the α -amine of Cys-1 contribute to the palmitoylation reaction. Since all known palmitoylation reactions are directed to the side chain of a Cys, Ser or Thr residue, we infer that the modification on hedgehog starts with the formation of a thioester intermediate, and that the palmitoyl group is transferred to the N-terminus by forming a cyclic intermediate. The above hypothesis was confirmed by a study using palmitoyl-coa to modify human Sonic hedgehog (see example 8). Example 3: confirming that the natural fatty acylated human Sonic hedgehog has increased potency in cell-based (C3H10T 1/2) experiments

Shh was tested for function in cell-based experiments in which alkaline phosphatase induction was measured on C3H10T 1/2 cells (25) for a total of 5 days. The experiment was performed in 96-well plates. The samples were run in duplicate. For bound Shh (100. mu.g/ml), the samples were first diluted 200-fold with normal growth medium, followed by serial 2-fold dilutions in the plate. Each well was normalized for the potential effect of added octyl glycoside, which was 0.005% addition of octyl glycoside to the medium. Blocking studies using neutralizing murine mAb5E 1(26) were performed as follows: shh and a series of antibody dilutions were mixed in culture medium for 30 minutes at room temperature before adding the test sample to the plate.

In this experiment, soluble human Shh produced a dose-dependent response in which IC₅₀At 1. mu.g/ml, the maximum signal appeared at 3. mu.g/ml (FIG. 6A). In bound human Shh, attachment of cholesterol at the C-terminus and palmitoyl at the N-terminus, which similarly produced a dose-dependent response in the experiment, but IC₅₀At 0.03. mu.g/ml, the maximum signal occurred at 0.1. mu.g/ml, indicating that it is about 30 times as potent as soluble Shh. To confirm that the observed activity is hedgehog specific, we also tested whether this activity could be inhibited by anti-hedgehog neutralizing mAb5E 1. Treatment with 5E1 inhibited both soluble and bound Shh (fig. 6B). One tenth of the inhibition of binding Shh was required for 5E1, which is consistent with the increased activity in its experiments.

Shh binding was tested in a receptor binding assay to monitor its ability to bind patched using a modified version of the recently published assay (10). Binding to Shh showed dose-dependent binding to patched expressing cells with apparent IC₅₀400 ng/ml (FIG. 7). In the same experiment, soluble Shh binds to the apparent IC of patched₅₀At 150 ng/ml, this indicates that the bound form binds only slightly tightly to its receptor. Example 4: analysis of bound human Sonic hedgehog after reconstitution into liposomes

This example illustrates that liposomes reconstituted to positive and negative charges by detergent dilution in a broad range of lipid: protein ratios (w/w) from 1: 1 to 100: 1 had no effect on Shh-binding activity in C3H10T 1/2 experiments.

The reconstitution into phospholipid-containing liposomes provides a useful formulation for lipid-containing proteins because it enables the lipid-containing proteins to be present in near normal sedimentation. To test whether such formulations are viable for binding Shh, we used detergent dilution to incorporate the protein into liposomes (60), where the liposomes used were mixed with the octyl glycoside and the protein of interest, and then the detergent was diluted below its critical micelle concentration, thereby driving recombination. While any of a number of pure or lipid mixtures can be utilized, we have chosen two commercially available mixtures as models: negatively charged liposome kit containing egg L- α -phosphatidylcholine, dicetylphosphate and cholesterol (cat. No. L-4262; Sigma, st. louis, MO); and a positively charged liposome kit consisting of egg phosphatidylcholine, stearylamine and cholesterol (cat. No. l-4137, Sigma).

Briefly, the lipids were transferred to Pyrex tubes, dried in a nitrogen stream, and the residual solvent was lyophilized. The lipids were suspended in 10mM HEPES ph 7.5, 100mM NaCl, 2.0% octyl glycoside, stirred, and then sonicated until the appearance of the suspension became clear. The lipids were then filtered through a 0.2 micron filter. High Five infected with baculovirus was treated with 400, 1000, 5000 and 20000 fold excess lipid (w/w) in octyl glycoside^TMAliquots of bound Shh obtained in insect cells were then diluted after a 15 minute pre-incubation and assayed for activity in the C3H10T 1/2 assay.

Neither positive or negative charge liposome treatment had an effect on hedgehog activity, indicating that lipid carriers are viable formulations. To confirm that hedgehog has indeed been reconstituted, centrifugation is performed with parallel samples, provided that binding to Shh normally precipitates, while liposomes float to the surface of the sample. In this case, bound Shh floats on the surface, indicating that recombination has occurred. Example 5: membrane-bound human Sonic hedgehog characteristics obtained from mammalian (EBNA-293) cells

To assess that palmitoylation is a routine modification pathway for Sonic hedgehog or that it is specific for insect cell formation, proteins were also made in EBNA-293 in mammalian systems. To express full-length Shh in mammalian cells, a 1.4 kb NotI fragment containing full-length Shh (see example 1) was cloned into a derivative of the vector CH269 pCEP4(Invitrogen, San Diego, CA (21)). Constructs were transfected into EBNA-293 cells using transfection amines (Life Technologies, Inc.) and cells were harvested 48 hours after transfection. Expression of surface Shh was confirmed by FACS and Western blot analysis.

In the narrow hole C₄Bound Shh from EBNA-293 was fractionated on a column (see example 3) using reverse phase HPLC. The individual peaks were analyzed by ESI-MS (sections a and b in Table 4) or MALDI-TOF MS on a Finnigan LaserMat mass spectrometer using α -cyano-4-hydroxycinnamic acid as matrix (section c in Table 4). The protein moved slightly faster than soluble Shh, as demonstrated by SDS-PAGE, which lags behind C in reverse phase HPLC analysis₄On the column and confirmed by mass spectrometry to contain ions corresponding to the modification of palmitic acid plus cholesterol. However, in contrast to insect cell-derived products in which more than 80% of the products contained palmitic acid and cholesterol modifications, HPLC elution profile and mass spectrometry data indicated that most mammalian cell-derived proteins lacked palmitoyl groups (see table 4 and fig. 3C). That is, in peak 2 of EBNA-293 cells, the proportion of MS signal of trimmed (des-1-10) to intact protein was 50%, while only about 10% of Shh in peak 1 was trimmed. Interestingly, both insect cell and mammalian cell-derived products showed comparable activity in the C3H10T 1/2 experiment, suggesting that cholesterol as well as cholesterol plus palmitic acid modifications are functionalized. Whether to simply use a second lipid linkage to further stabilize protein binding to the membrane or whether it plays a more important role and affects its formation or protein-protein contact still needs to be determined.

Fatty acid derivatization of proteins is a common post-translational modification that occurs late in the maturation process (28, 29). For cystine derivatives, the process is dynamic, involving a separate enzyme that adds and removes modifications on the thiol group. The most common function of such derivatizations (e.g., palmitoylation) is to alter the physico-chemical properties of the protein, such as targeting the protein to its functional site, facilitating protein-protein interactions, and mediating protein-membrane interactions (30). For hedgehog, although it is surprising that there is a difference in the degree of palmitoylation in insect and mammalian cell-derived preparations (80% for insect cells and 30% for mammalian cells), we do not know whether it is physiologically significant or whether it simply reflects the difference in the cellular mechanisms of the two test systems for the addition and removal of palmitic acid. The difference in the degree of modification in insect and mammalian cells is unlikely to be species dependent because the bound Drosophila hedgehog produced in insect cells is devoid of palmitic acid (19), despite having an independent N-terminal sequence.

Table 4: mass spectrometric analysis of EBNA-293-derived conjugated human Sonic hedgehogExample 6: lipid modification of murine Sonic hedgehog

This example illustrates that various lipids can be attached to soluble murine hedgehog if the murine Shh gene encoding residues 1-174 is at High Five^TMExpressed in insect cells, essentially the same as the full-length human Shh described in example 1. Lipid modification makes the moiety membrane-bound. The N-terminal fragment (residues 1-174 of untreated murine Sonic hedgehog) differs from the N-terminal fragment of human Sonic hedgehog by only 2 amino acid residues. In the murine Sonic hedgehog N-terminal fragment, threonine replaces serine at position 44 and aspartic acid replaces glycine at position 173. If at High-Five^TMThe murine Sonic hedgehog lacking the autohandling domain is expressed in an insect cell/baculovirus expression system and the vast majority of the protein is secreted in the culture medium due to its lack of ability to attach cholesterol groups to the C-terminus. This soluble form has specific biological activity (measured using the C3H10T 1/2 alkaline phosphatase induction assay of example 3) similar to the soluble N-terminal fragment of human Sonic hedgehog expressed and purified from E.Coli.

However, a small fraction of total protein is still associated with insect cells. Cell-bound murine sonic hedgehog protein was purified essentially as described in example 1 and was found to be comprised of e.coli and High Five, respectively, in alkaline phosphatase experiments (data not shown)^TMThe soluble N-terminal fragment of human or mouse Sonic hedgehog purified by insect cell/baculovirus expression system has more remarkable effectActivity of (2). Subsequent studies of murine Sonic hedgehog N-terminal fragments (as described in example 1) using HPLC and electrospray mass spectrometry analysis suggested that the protein was lipid-modified and that there were more than 1 type of lipid modification. Supporting evidence includes the following facts:

1. cell-bound form-specific soluble N-terminal fragments of human and murine sonic hedgehog at C₄Eluted more slowly in a reverse phase HPLC column (Vydac cat No. 214TP104) eluting linearly with a 0-70% acetonitrile gradient in 0.1% trifluoroacetic acid for 30 min;

2. as shown in table 5, the quality of the cell-bound form is consistent with the predicted value of the lipid-modified protein.

Table 5: mass of various lipid-modified forms of murine Sonic hedgehog

Protein	Adduct compounds	Expected mass*(MH+)	Observation quality (MH)+)
				Unmodified	Is free of	19632．08	19632
Myristoyl radical	CH3(CH2)12CO-	19842．50	19841
				Palmitoyl radical	CH3(CH2)14CO-	19870．55	19868
Stearoyl radical	CH3(CH2)16CO-	19898．60	19896
				Eicosanyl group	CH3(CH2)18CO-	19926．66	19925

Experimental mean mass in calculating expected mass

The position of the lipid group was determined using a combination of sequence analysis and peptide mapping. Automated N-terminal Edman sequencing of lipid-modified forms indicated that the N-terminus was blocked, suggesting that the lipid was attached to the alpha-amine of the N-terminal cystine. Endo-Lys-C peptide mapping, MALDI-TOF mass spectrometry and MALDI PSD analysis (as described in example 1) of the 4-vinylpyridine alkylated lipid-modified form were used to confirm the location of lipid modifications and to determine their precise mass.

As shown in Table 6, the mass of the N-terminal peptide carrying the lipid modification (including residues 1-9 plus 4-vinylpyridine attached to the thiol side chain of the N-terminal cystine) is consistent with the expected value for the lipid-modified peptide in the 0.1 Da range.

Table 6: isolation from various lipid-modified murine Sonic hedgehogQuality of the N-terminal peptide of (1)

Protein	Adduct compounds	Expected mass*(MH+)	Observation quality (MH)+)
				Myristoyl radical	CH3(CH2)12CO-	1193．69	1193．76
Palmitoyl radical	CH3(CH2)14CO-	1221．72	1221．65
				Stearoyl radical	CH3(CH2)16CO-	1249．75	1249．71

Experimental mean mass in calculating expected mass

In addition to the lipid-modified peptides shown in table 6, peptides with masses of 1191.74, 1219.84, and 1247.82 were also detected. These masses are consistent with the unsaturated forms of myristate, palmitate and stearate, respectively, but the position of the double bond in the alkyl chain cannot be determined. These facts indicate that saturated as well as unsaturated fatty acids can be covalently linked to the N-terminal cystine. For both saturated and unsaturated lipid-modified peptides, MALDI PSD analysis as described in example 1 confirmed that the lipid is covalently linked to the N-terminal cystine residue. Example 7: lipid modification of Indian hedgehog

To assess whether the palmitoylation reaction was unique to human Shh or whether it occurred on other hedgehog proteins, we tested whether human Indian hedgehog (expressed as a His-tagged fusion protein in e.coli, with an enterokinase cleavage site immediately adjacent to the beginning of the mature sequence, and purified exactly as recombinant human Sonic hedgehog (example 9)) could be palmitoylated using the experiment described in example 2. Human Indian hedgehog was modified (see figure 1, lane h), indicating that palmitoylation is likely to be a common feature of hedgehog proteins. The ability to directly label Shh and Ihh with radioactive palmitic acid in a cell-free system allows for simple screening of the amino acids involved in the modification reaction. In addition, Indian hedgehog palmitoylated using the procedure described in example 8 was significantly more potent in the C3H10T 1/2 experiment than unmodified Ihh. Example 8: lipid modification of Sonic hedgehog using acyl-CoA

In vitro acylation of proteins containing an N-terminal cystine can be accomplished by a two-step chemical reaction with a fatty acid-thioester donor. In the first step, the acyl group of the thioester donor is transferred to the thiol group of the N-terminal cystine of the protein by spontaneous transesterification. Subsequently, the acyl group undergoes S to N migration to transfer to the α -amine of the N-terminal cystine, forming a stable amide bond. Direct acylation of amine functions on proteins also occurs with prolonged incubation with thioesters, but the presence of cystine on proteins will accelerate the reaction and control at the acylation site. In this example, a commercially available coenzyme A derivative (Sigma Chemical Company, St. Louis MO) was used, but other thioester groups could achieve the same result. In fact, certain thioester leaving groups, such as thiobenzyl ester, are likely to react more rapidly. Internal cystine residues may also facilitate acylation to adjacent lysines (e.g., in internal cystine-lysine pairs), and this can be conveniently tested by using synthetic peptides. The second acylation that occurs on the protein during the reaction with the thioester can be prevented by controlling buffer composition, pH, or by site-directed mutagenesis adjacent to the lysine.

In pre-analyzing the effect of acylation on the ability of human Sonic hedgehog to induce alkaline phosphatase in C3H10T 1/2 cells, the reaction mixture contained 1 mg/ml human Sonic hedgehog (51. mu.M), 500. mu.M of a particular commercially available acyl-CoA (compounds tested included acetyl-CoA (C2: 0), butyryl-CoA (C4: 0), hexanoyl-CoA (C6: 0), octanoyl-CoA (C8: 0), decanoyl-CoA (C10: 0), lauroyl-CoA (C12: 0), myristoyl-CoA (C14: 0), palmitoyl-CoA (C16: 0), palmitoyl-CoA (C16: 0), stearoyl-CoA (C18: 0), eicosyl-CoA (C20: 0), docosanyl-CoA (C22: 0), tetracosanyl-CoA (C24: 0), succinyl-CoA and benzoyl-CoA (C24: 0), 25mM DTT, and 50mM Na₂HPO₄pH 7.0. The reaction was incubated under the organisms for 3 hours and then immediately analyzed (without purification) for biological activity in the C3H10T 1/2 experiment as described in example 3. Samples analyzed in reverse phase HPLC and other physical methods are typically stored at-70 ℃. HPLC analysis in Vydac C₄Reverse phase column (4.6 mm ID. times.250 mm, 5 micron particle) with a gradient elution of 5-85% acetonitrile in 0.1% TFA for 40 min at a flow rate of 1 ml/min. The eluate was monitored at 280nm, fractions were collected in some experiments, and hedgehog protein was analyzed on SDS-PAGE with coomassie staining and Western blot detection.

Comparison of the activities of the various reaction mixtures (FIG. 10) shows that chain lengths between 12 and 18 carbon atoms are optimal in inducing alkaline phosphatase activity compared to the unmodified protein. Increasing chain length further leads to an apparent decrease in activity, and the double bond in unsaturated palmitoyl CoA (C16: 1) makes the activity the same as that of fully saturated palmitoyl CoA (C16: 0). When the reaction mixture was analyzed by reverse phase HPLC, we observed that many shorter chain acyl-coa derivatives did not react with hedgehog protein, and thus the dependence of biological activity shown in fig. 10 did not fully reflect acyl chain length.

To obtain data on the true activity of the modified protein, and on the dependence of activity on acyl chain length, we developed a method to synthesize and purify the individual N-terminal acylated forms. Palmitoylation, myristoylation, lauroylation, decanoylation, and octanoylation of human Sonic hedgehog proteins carrying a single acyl chain linked to the alpha-amine of the N-terminal cystine were prepared in a reaction mixture comprising: 0.80 mg/ml (41. mu.M) human Sonichehdge, 410. mu.M (10-fold molar excess) palmitoyl-CoA, myristoyl-CoA, or lauroyl-CoA, or 4.1 mM (100-fold molar excess) decanoyl-CoA or octanoyl-CoA, 25mM DTT (for reaction mixtures comprising palmitoyl-CoA, myristoyl-CoA, or lauroyl-CoA) or 0.5 mM DTT (for reaction mixtures comprising decanoyl-CoA or octanoyl-CoA), and 40mM Na₂HPO₄pH 7.0. The reaction mixture was incubated at 28 ℃ for 24 hours. The reaction of the N-terminal cystine with the acyl thioester results in the transfer of the acyl group to the sulfhydryl group by spontaneous transesterification, and then to the α -amine by S to N migration, forming a stable amide bond. The free sulfhydryl group is then subjected to a second transesterification reaction to produce a protein bearing a fatty acyl group attached to the sulfhydryl group by a thioester bond. The thioester-linked acyl group is then removed as follows: sequentially adding 0.11 volume of 1M Na₂HPO₄pH9.0 and 0.11 volume of 1M hydroxylamine (0.1M final concentration), then at 28 ℃ temperature in 18 hours, this will only remain connected to the protein acyl amide (62). Then 0.25 volume of 5% octyl glycoside (1% final concentration) was added and the mixture was incubated at room temperature for 1 hour. Then, the protein was purified by cation exchange chromatography using SP-Sepharose Fast Flow (Pharmacia) and Bio Scale S (Biorad) in the presence of 1% octyl glycoside. Purified protein relativeAt 5mM Na₂HPO₄Dialyzed at pH 5.5, 150mM NaCl, 1% octyl glycoside, 0.5 mM DTT, and then stored at-70 ℃. The presence of octyl glycoside was required to maintain all solubility, and removal of the detergent by dilution and dialysis resulted in the loss of 75%, 41% and 15% of palmitoylated, myristoylated and lauroylated proteins, respectively. ESI-MS of HPLC purified proteins confirmed their integrity: palmitoylated Sonic hedgehog, measured mass =19798, calculated mass = 19798.43; myristoylated Sonic hedgehog, measured mass =19770, calculated mass = 19770.33; lauroylated Sonic hedgehog, measured mass =19742, calculated mass = 19742.33; sonic edgehog decanoylated, measured mass =19715, calculated mass = 19714.28; caprylylated Sonic hedgehog, measured mass =19686, calculated mass = 19686.23.

Analysis of the various acylated forms of human Sonic hedgehog (FIG. 11) in the C3H10T 1/2 assay showed that the activity of the protein was chain length dependent. Palmitoylated, myristoylated, and lauroylated proteins show nearly identical activity with EC₅₀Values of 5-10 ng/ml (a 100-fold increase in potency compared to unmodified protein) were obtained. EC of decycylated human Sonic hedgehog₅₀Values of 60-70 ng/ml (15-30 fold increase in potency compared to unmodified protein) were less active than palmitoylated, myristoylated, and lauroylated proteins, while the octanoylated form showed the lowest activity, its EC₅₀At 100-200 ng/ml (10-fold increase in potency compared to unmodified protein). All acylated forms are more potent than the unmodified protein, the latter EC₅₀1000-. Except EC₅₀In addition to the reduction, palmitoylated, myristoylated, and lauroylated proteins induced about 2-fold higher alkaline phosphatase activity compared to unmodified proteins, whereas decanoylated and octanoylated proteins induced 1.5-fold more.

In addition to the observed increased potency of myristoylated human Sonic hedgehog in the C3H10T 1/2 experiment, myristoylated forms were significantly more potent than unmodified proteins in inducing forebrain neurons in explants of the telencephalon of E11 mice at the embryonic stage. Incubation of E11 telencephalon explants with various concentrations of unmodified or myristoylated Sonic hedgehog followed by staining of explants with the products of the dlx and islet-1/2 genes (markers for forebrain neurons) showed that induction of the unmodified protein was first observed at 48nM, but induction of the myristoylated form was first observed at 3 nM. Furthermore, the unmodified protein induced restricted expression at 3070nM, while the myristoylated protein induced a broad range of expression at 48 nM. A similar increase in potency was observed when the telogen at embryonic stage E9 was expected to be incubated with unmodified or myristoylated protein. Explants stained for the nkx 2.1 gene product (an early marker of forebrain neurons) indicated that at 384nM unmodified protein induced nkx 2.1 first, whereas myristoylated protein at 12nM expression of nkx 2.1 was first observed. Furthermore, the expression of Nkx2.1 was expanded at 48nM myristoylated Sonic hedgehog, but not at this concentration using the unmodified form.

In addition, myristoylated human Sonic hedgehog has been shown to be significantly more protective than the unmodified protein, reducing the volume of lesions resulting from malonate administration to the striatum of the murine brain (see example 16). Example 9: chemical derivative A of N-terminal cystine of human Sonic hedgehog, alkylation of the total process protein. In 50. mu.l of 6M guanidine hydrochloride, 50mM Na₂HPO₄Samples containing about 20. mu.g of protein were treated with 0.5. mu.l of 4-vinylpyridine at pH 7.0 for 2 hours at room temperature. 40 volumes of cold ethanol were added, thereby precipitating the S-pyridylethylated protein. The solution was stored at-20 ℃ for 1 hour and then centrifuged at 14000 Xg and 4 ℃ for 8 minutes. The supernatant was discarded and the precipitate was washed with cold ethanol. The protein was stored at-20 ℃. Peptide mapping. Alkylated protein (0.4 mg/ml in 1M guanidine hydrochloride, 20mM Na₂HPO₄pH6.0) was digested with endo Lys-C (Wako Pure Chemical Industries, Ltd.) at a ratio of 1: 20. Digestion was performed at room temperature for 30 hours. The reaction was stopped by acidification with 5. mu.l of 25% trifluoroacetic acid. In thatThe digests were analyzed on a Waters 2690 Separation Module, which was equipped with a 996 model photodiode matrix detector. Guanidine hydrochloride in solid form was added to the digest to a concentration of 6M prior to injection to dissolve insoluble material. Using reversed phase Vydac C for separation₁₈(2.1 mm ID. times.250 mm) column, eluting with a 0.1% trifluoroacetic acid/acetonitrile gradient for 90 minutes at a flow rate of 0.2 ml/min 0.1% trifluoroacetic acid/acetonitrile. Each peak was collected manually for mass spectrometry. And (4) measuring the mass. Molecular mass of intact proteins was determined on a Micromass Quattro II triple quadrupole mass spectrometer using electron spray ionization mass spectrometry (ESI-MS). Reliasil C for ultrafast micro-protein analyzer using on-line Michrom₄(1mm ID. times.50 mm) column the sample was desalted. The flow rate was 20. mu.l/min. All electron jet mass spectral data were processed using the Micromass MassLynx data system. The molecular mass of the peptides was determined by matrix assisted laser desorption ionization time of flight mass spectrometry (MALDI-TOF-MS) on a Voyager-DETM STR (PerSeptive Biosystems, Framingham, Mass.). Sequencing of the modified peptide was performed by post-source decay (PSD) measurement on the same instrument. Alpha-cyano-4-hydroxycinnamic acid was used as a base. N-terminal sequencing. Proteins were sequenced by Edman degradation on a Perkin-Elmer Applied Biosystems model 477A pulsed liquid protein sequencer. PTH-thioproline was prepared on-line by directly adding thioproline (thiazolidine-4-carboxylic acid) to a sample addition column of a sequencer. Bacterial expression and purification of chemically modified wild-type soluble human Sonic hedgehog N-terminal fragments. The bacterial pellet from Shh-expressing cells, which was 4-5% of total protein, was thawed and resuspended in lysis buffer (25mM Na) at a ratio of 1: 4 (w/v)₂HPO₄pH8, 150mM NaCl, 1mM EDTA, 1mM PMSF, 0.5 mM DTT), and then passed twice by a Gaulin compressor (manufactured by APV Rannie, Copenhagen, Denmark) of 12000 p.s.i. for dissolution. All subsequent purification steps were carried out at 2-8 ℃ unless otherwise stated. The homogenate was centrifuged at 19000 Xg for 60 minutes, and then MES 0.5 m pH5 was added to the resulting lysate in a ratio of 1: 10 (v/v). The lysate (pH 5.5) was added to SP Sepharose Fast Flow (P)Harmacea, Piscataway, NJ) column (4g E.Coli wet weight/ml resin) over 25mM Na₂HPO₄pH5.5, 150mM NaCl equilibrium. Washing with 4 Column Volumes (CV) of equilibration buffer followed by 3CV of 25mM Na₂HPO₄The His-tagged Shh was eluted with 800mM NaCl in the same buffer, after washing at pH 5.5, 200mM NaCl, 0.5 mM DTT. The eluted fractions were analyzed by absorbance at 280nm and SDS-PAGE. Imidazole (1M feed solution pH7) and NaCl (5M feed solution) were added to the pool with peak Shh containing the fraction from the SP Sepharose eluate to give final concentrations of 20mM and 1M respectively. This material was then added to a column (20 mg/ml resin) of NTA-Ni agarose (Qiagen, Santa Clara, Calif.) packed with 25mM Na₂HPO₄pH8, 1M NaCl, 20mM imidazole, 0.5 mM DTT equilibrium. The column was washed with 5CV of the same buffer and then 3CV of 25mM Na₂HPO₄His-tagged Shh was eluted at pH8, 1M NaCl, 200mM imidazole, 0.5 mM DTT. The protein content in the elution pool eluted from the NTA-Ni column was determined by absorbance at 280 nm. The cell was warmed to room temperature and then an equal volume of 2.5M sodium sulfate was added. Phenyl Sepharose step was performed at room temperature. The material was loaded onto a Phenyl Sepharose Fast Flow (Pharmacia, Piscataway, NJ) column (20 mg/ml resin) using 25mM Na₂HPO₄pH8, 400mM NaCl, 1.25M sodium sulfate, 0.5 mM DTT. With 25mM Na₂HPO₄His-tagged Shh was eluted at pH8, 400mM NaCl, 0.5 mM DTT. In general, we recovered 2-3g of His-tagged Shh from 0.5 kg of bacterial slurry (wet weight). The product was filtered through a 0.2 μm filter, aliquoted and stored at-70 ℃. The purity of the His-tagged Shh was determined to be 95% by SDS-PAGE. The purified product was further characterized and the sample was analyzed by electron-jet ionization mass spectrometry (ESI-MS). About 50% of the proteins lost the N-terminal methionine.

To cleave the hexahistidine tag, the enzyme was used at a ratio of 1: 1000 (w/w): shh ratio enterokinase (Biozyme, San Diego, Calif.) was incubated with His-tagged Shh at 28 ℃ for 2 hours. Passing the digest through a second NTA-Ni SepharoseThe column (20mg Shh/ml resin) was passed through, thereby removing the uncleaved His-tagged Shh and the free His-tag. Imidazole (1M feed solution, pH7) and NaCl (5M feed solution) were added to the digest to final concentrations of 20mM and 600mM, respectively, prior to loading. The material was loaded on an NTA-Ni column at 25mM Na₂HPO₄pH8, 600mM NaCl, 20mM imidazole, 0.5 mM DTT, and fractions were collected. The column was washed with 1CV of the same buffer and fractions were collected. MES (0.5M stock solution, pH5) was added to the NTA-Ni agarose unbound fraction to a final concentration of 50mM, followed by 2 volumes of water. This material was loaded onto a second SP Sepharose fast Flow column (20 mg/ml resin) using 5mM Na₂HPO₄pH5.5, 150mM NaCl, 0.5 mM DTT equilibrium. The column was washed with 3CV of equilibration buffer, 1CV of the same buffer containing 300mM NaCl. With 5mM Na₂HPO₄Shh was eluted at pH 5.5, 800mM NaCl, 0.5 mM DTT. Atomic absorption data indicate that Shh contains 0.5 mol equivalents of Zn at this stage²⁺. Adding equimolar ZnCl into Shh eluent₂Then relative to 5mM Na₂HPO₄pH5.5, 150mM NaCl, 0.5 mM DTT dialysis protein. SDS-PAGE, Size Exclusion Chromatography (SEC) and ESI-MS confirmed that the purity of the Shh obtained was greater than 98%, while atomic absorption confirmed that it contained 0.9-1.1 Zn²⁺／Shh。

The data for His-tagged Shh and ESI-MS of the product obtained after removal of the His tag are summarized in Table 7.

Table 7: ESI-MS characterization of Shh

Protein	Mass (Da)
			Computing	Measuring
	His-tagged Shh (-Met)	21433．82	21434
(complete)	21565．01	21565
Enterokinase-disrupted Shh	19560．02	19560

B. Specific chemical modification human Sonic hedgehog was modified with N-ethylmaleimide. At 5mM Na₂HPO₄purified Shh at pH 5.5, 150mM NaCl, 0.5 mM DTT was treated with 10mM N-ethylmaleimide on ice for 1 hour, then 5mM Na₂HPO₄Dialyzed against 150mM NaCl at pH 5.5. MALDI-TOF-MS data indicate that the mass of the N-ethylmaleimide (NEM) modified protein is increased by 126Da, which indicates that only one cystine residue in Shh is modified by the reagent. N-terminal sequencing data indicated that the protein was sequenceable and that an unusual peak was detected in the first cycle (data not shown), possibly associated with PTH-NEM-Cys. Mass spectrometry of pyridylethylated NEM-Shh under denaturing conditions showed that only two cystine residues in the protein were alkylated,it was confirmed that only the thiol group of the N-terminal cystine residue was modified by NEM under native conditions (table 8). The other two cystine residues are apparently buried in the hydrophobic core of the protein and are not modified prior to denaturation.

Table 8: MS characterization of NEM-modified Shh

Protein	Quality (calculation)	Quality (measurement)
			Pyridylethylated NEM Shh		19895Da
If it contains 2 free cystine residues	19895Da
			If it contains 3 free cystine residues	20000Da

When tested in the C3H10T 1/2 assay (see example 2), the N-ethylmaleimide-modified hedgehog protein was identical in activity to the unmodified protein. This confirms that the free thiol group at the N-terminus of hedgehog is not essential for activity, and that the hydrophobicity of N-ethylmaleimide is sufficient to confer some activity on hedgehog compared to other more hydrophobic modifying groups, such as conversion of Cys-1 to His or Asp, which would reduce activity. Human Sonic hedgehog is modified with formaldehyde to form N-terminal thioproline, and treated with acetaldehyde and butyraldehyde to form the N-terminal thioproline derivative.For formaldehyde modification, at 5mM Na₂HPO₄purified Shh at 3 mg/ml in 0.5 mM DTT, pH 5.5, 150mM NaCl, was treated with 0.1% formaldehyde at room temperature for 1-6 hours, with or without 10% methanol. Protein relative to 5mM Na₂HPO₄Dialyzed against 150mM NaCl at pH 5.5, or purified on a CM-Poros column (Perseptive Biosystems) as described below, and then purified against 5mM Na₂HPO₄Dialyzed at pH 5.5 against 150mM NaCl. For modification with acetaldehyde or butyraldehyde, at 5mM Na₂HPO₄purified Shh at pH 5.5, 150mM NaCl, 3 mg/ml in 0.5 mM DTT was treated with 0.1% acetaldehyde or butyraldehyde at room temperature for 1 hour, and then the protein was purified on a CM-Poros column. ESI-MS data for formaldehyde, acetaldehyde and butyraldehyde treated proteins showed that they had mass 13Da, 27Da and 54Da respectively higher than the unmodified protein (see Table 9). Table 9: expected and observed quality of human Sonic hedgehog treated with formaldehyde, acetaldehyde and butyraldehyde

Protein	Expected mass*(MH+)	Quality of observation*(MH+)
			Unmodified	19560．02	19560
Of formaldehyde treatment	19572．03	19573
			Of acetaldehyde	19586．06	19587
Butyraldehyde treated	19614．11	19614

Using the average mass for calculating the expected mass.

For the formaldehyde-treated protein, as described above, peptide mapping confirmed that the site of modification occurred in the peptide spanning the first 9N-terminal residues, and that the precise mass gain was 12 Da. MALDI-PSD MS studies of this peptide showed that modification occurred at Cys-1 and could be explained by the modification of the N-terminal alpha-amine and Cys-1 thiol side chains to form thioproline (see FIG. 12). The structure of thioproline was confirmed by automated N-terminal Edman sequencing using "on-line" prepared PTH-thioproline as a standard. For acetaldehyde and butyraldehyde treated proteins, the ESI-MS data are consistent with modification by the same chemical occurrence as the reaction with formaldehyde, but the site of modification has not been identified. The formaldehyde, acetaldehyde and butyraldehyde-modified proteins were approximately 8, 2 and 3 times more potent than unmodified Shh, respectively, when tested in a C3H10T 1/2 cell-based experiment. Human Sonic hedgehog was modified with N-isopropyl iodoacetamide. This example shows that modification of human Shh with a hydrophobic derivative of iodoacetamide enhances the efficacy of the protein compared to unmodified Shh. Purified Shh (1 mg/ml in 5mM Na)₂HPO₄ph 7.0, 150mM nacl, 0.1 mM DTT) was incubated with 1mM N-isopropyl iodoacetamide (NIPIA) at 4 ℃ for 18 hours. DTT was added to 10mM final concentration, then the sample was compared to 5mM Na₂HPO₄Extensive dialysis was performed at pH 5.5, 150mM NaCl, 0.5 mM DTT. Samples were purified on SP Sepharose Fast Flow resin and compared to 5mM Na₂HPO₄The mixture was further dialyzed against 150mM NaCl, 0.5 mM DTT, pH 5.5. ESI-MS data indicate complete conversion to a substance of mass 19660, corresponding to a single modificationThe expected quality value of the protein of (19659). Specific modification of the N-terminal cystine was confirmed by peptide mapping of the proteolytic fragments. When tested in a C3H10T 1/2 cell-based experiment, NIPIA-modified human Shh was approximately 2-fold more potent than the unmodified protein. Although modification of the protein resulted in only a moderate increase in potency, it is expected that modification of the protein with a long chain alkyl iodoacetamide derivative would result in a hydrophobically modified protein with a higher increase in potency, perhaps 200-fold similar to that observed for palmitoylated, myristoylated, and lauroylated Shh proteins (see example 8). Human Sonic hedgehog was modified with 1-bromo-2-butanone to form a 6-membered hydrophobic ring at the N-terminus. Thiomorpholinyl- (tetrahydrothiazinyl-) derivatives of Shh were prepared as follows: human Shh-N (3 mg/ml in 5mM Na) was allowed to stand at room temperature₂HPO₄pH 5.5, 150mM NaCl, 0.15 mM DTT) with 11mM 1-bromo-2-butanone for 60 minutes, followed by 5mM NaCl BH₃Reduction was carried out at room temperature for 60 minutes. The reaction product was purified on a CM-Poros column (Perseptive biosystems) as described below, and the concentration of the purified product was 5mM Na₂HPO₄Dialyzed against 0.5 mM DTT at pH 5.5 and 150mM NaCl. The ESI-MS and proteolytic peptide mapping data indicate that the product is the desired thiomorpholinyl derivative (calculated mass =19615, observed mass =19615) and a mixture of two protein forms, both of which have 16 additional mass units. One of these forms is an acyclic keto-thioether intermediate. The mixture was tested in the C3H10T 1/2 experiment and showed 5-fold higher potency than the unmodified protein. Example 10: genetic engineering mutagenesis of human Sonic hedgehog A, N-terminal cystine

In this example, we show that the specific substitution of the N-terminal cystine (Cys-1) of human Sonic hedgehog with mono-and poly-hydrophobic amino acid residues results in an increase in potency compared to the wild-type protein in a C3H10T 1/2 cell-based experiment as described in example 3. Construction of Shh Cys-1 mutants. The 584 bp NcoI-XhoI restriction fragment carrying the His-tagged wild-type ShhN-terminal fragment generated by p6H-SHH was subcloned into the pUC wild-type cloning vector pNN05 to construct plasmid pEAG 649. According to manufacturer's constructionThe protocol used the Pharmacia kit to perform unique site-directed mutagenesis of the pEAG649 plasmid template, thereby generating the Cys-1 mutant of soluble human Shh. In designing the mutagenic primers, if the desired mutation does not produce a restriction site change, silent mutations are introduced in adjacent codons that produce a restriction site change to facilitate identification of the mutated clone after mutagenesis. To avoid the use of aberrant coders, substituted coders were selected from those that occurred at least once elsewhere in the human Shh cDNA sequence. The following mutagenesis primers were used: (1) for C1F: 5 'GGCGAT GAC GAT GAC AAA TTC GGA CCG GGC AGG GGG TTC 3' (SEQ ID NO:), which introduces an ApoI site to make pEAG 837; (2) for C1I: 5 'GGC GAT GAC GAT GAC AAA ATA GGA CCG GGC AGG GGG TTC 3' (SEQ ID NO: SEQ ID NO), which has lost the RsrII site to make pEAG 838; and (3) for C1M: 5 'GGC GAT GAC GAT GAC AAA ATG GGC CCG GGC AGGGGG TTC GGG 3' (SEQ ID NO:), which lost the RsrII and AvaII sites to make pEAG 839. The mutation was confirmed by DNA sequencing by carrying a 180 bp NcoI-BglII restriction fragment from the N-terminus of the mutated SHH protein in plasmid pEAG 837-839. The 180 bp NcoI-BglII fragment of each mutant plasmid and the 404 bp BglII-XhoI fragment from pEAG649 were subcloned into the phosphatase-treated 5.64 kb XhoI-NcoIpET11d vector backbone of p6H-SHHH, thereby constructing expression vectors. The presence of the introduced restriction site changes was re-confirmed in the expression vectors for each of the Cys-1 mutants (C1F in pEAG840, C1I in pEAG841, and C1M in pEAG 842). The expression vectors were transformed into competitive E.coli BL21(DE3) pLysS (Stratagene) according to the manufacturer's recommendations and selection was performed on LB agar plates containing 100. mu.g/ml ampicillin and 30. mu.g/ml chloramphenicol. Individual colonies were selected and transformed bacteria were grown to an A of 0.4-0.6₆₀₀Then, it was induced with 0.5 mM IPTG for 3 hours. Bacterial pellets were analyzed for expression of muteins by reducing SDS-PAGE and Westem blot.

Preparation of a plasmid with multiple N-termini by unique site-directed mutagenesis Using the Pharmacia kit according to the manufacturer's recommendationsSoluble human Shh mutants with hydrophobic substitutions (ClII). In designing the mutagenic primers, if the desired mutation does not produce a restriction site change, silent mutations are introduced in adjacent codons that produce a restriction site change to facilitate identification of the mutated clone after mutagenesis. To avoid the use of aberrant coders, substituted coders were selected from those that occurred at least once elsewhere in the human Shh cDNA sequence. For ciii, the following mutagenesis primers were used on the C1F template plasmid pegg 837: 5 'GCG GCG ATGACG ATG ACA AAA TCA TCG GAC CGG GCA GGG GGT TCG GG 3' (SEQ ID NO:), which removes the ApoI site to make pEAG 872. The mutation was confirmed by DNA sequencing through a 0.59 kp NcoI-XhoI restriction fragment carrying the mutant ClII Shh. The NcoI-XhoI fragment of the mutant plasmid was subcloned into the phosphatase-treated 5.64 kb XhoI-NcoI pET11d vector backbone of p6H-SHHH, thereby constructing an expression vector. The presence of the introduced restriction site changes was confirmed anew in expression vector pEAG875 for the ciii mutant. The expression vectors were transformed into competitive E.ColiBL21 (DE3) pLysS (Stratagene) according to the manufacturer's recommendations and selection was performed on LB agar plates containing 100. mu.g/ml ampicillin and 30. mu.g/ml chloramphenicol. Individual colonies were selected and transformed bacteria were grown to an A of 0.4-0.6₆₀₀Then, it was induced with 0.5 mM IPTG for 3 hours. The bacterial pellet was analyzed as described above to confirm the expression of the mutant Shh protein. Purification of Cys-1 mutant of human Sonic hedgehog. His-tagged mutant hedgehog protein was purified from bacterial pellets as described above for the wild-type protein, but with two changes. (1) The Pheny Sepharose step was eliminated and instead in preparation for the enterokinase cleavage step, the protein pool from the first NTA-Ni column was dialyzed into 25mM Na₂HPO₄pH8, 400mM NaCl, 0.5 mM DTT. (2) The final ion exchange step was varied from elution on SP-Sepharose Fast Flow to gradient elution on a CM-Poros column (Perseptive biosystems). This was at 50mM Na₂HPO₄performed at pH 6.0, with a gradient of 0-800mM NaCl over 30 column volumes. From the peak fractions pooled in this step, dialysis was performed into 5mM Na₂HPO₄pH 5.5, 150mM NaCl and then stored at-80 ℃. Mass spectra of the purified proteins gave the expected mass ions for each purified form. Cys-1 mutant activity of human Sonic hedgehog. As shown in Table 10, in the C3H10T 1/2 experiment, the mutation of the N-terminal cystine had a significant effect on the potency of the resulting hedgehog protein. For a single change, potency is usually related to the hydrophobicity of the substituted amino acid, that is, phenylalanine and isoleucine produce the greatest activation, methionine is less activated, and histidine and aspartic acid eliminate the activity compared to wild-type cystine. Substitution of two isoleucine for cystine resulted in a greater increase in activity than a single isoleucine substitution. If 9 amino acids are classified as more hydrophobic than cystine (Proteins: structures and molecular properties), 2 nd edition, 1993, t.e.creighton, w.h.freeman co. p.154), the substitutions tested above are clearly not exhaustive studies of possible mutations at the N-terminus that can promote hedgehog activity. However, this result demonstrates that activation is not limited to a single amino acid structure, and that substitution of more than 1 amino acid can further increase efficacy. Thus, one skilled in the art would be able to substitute other amino acids at the N-terminus to yield a hedgehog form with greater potency than the wild-type protein. Table 10: relative potency of human Sonic hedgehog amino acid modification in C3H10T 1/2 experiments

N-terminal	Relative potency
		C (wild type)	1X
M	2X
		F	4X
I	4X
		II	10X

B. Construction of ClII/A169C mutant by genetic engineering of internal residues. Soluble human Shh mutant ClII/A169C (in which cystine is substituted for the optionally missing C-terminal residue A169 that is predicted to have high separation solvent access) was prepared by unique site-eliminating mutagenesis using the Pharmacia kit according to the manufacturer's suggested protocol and using the mutagenesis oligo design principle described above. The following mutagenesis primers were used on the ClII Shh template, pegg 872: 5 'GAG TCA TCA GCC TCC CGA TTTTGC GCA CAC CGA GTT CTC TGC TTT CAC C3' (SEQ ID NO:), to add the FspI site to make pSYS 049. The ClII/A169C mutation was confirmed by DNA sequencing through a 0.59 kb NcoI-XhoI restriction fragment. The NcoI-XhoI fragment was subcloned into the phosphatase-treated 5.64 kb XhoI-NcoI pET11d vector backbone of p6H-SHHH, thereby constructing an expression vector. The presence of the introduced restriction site changes was reconfirmed in the expression vector. The expression vector was transformed into competitive e.coli BL21(DE3) pLysS as described above, colonies were selected, and expression for Shh was induced and screened. Purification of the ClII/A169C mutant. The ClII/A169C mutant was purified as described in example 9 for wild-type Shh, with the following changes. (1) EDTA was removed in the lysis buffer. (2) The order of NTA-Ni and SP Sepharose steps was switched and the Phenyl Sepharose step was omitted. (3) After centrifugation to clarify the lysed bacteria, additional sodium chloride was added to the supernatant to a final concentration of 300 mM. (4) Elution buffer for NTA-Ni column contains 25mM Na₂HPO₄pH8.0, 200mM imidazole, 400mM NaCl. (5) The elution pool of the NTA-Ni column was diluted with 3 volumes of 100mM MES pH 5.0 and loaded on the SP Sepharose column. (6) The SP Sepharose elution pool was half-volume with 50mM Na before addition of enterokinase₂HPO₄And (4) diluting at pH8.0. And (7) the DTT in the buffer of the final SP Sepharose column contained 0.2 mM DTT, and the elution pool of this step was aliquoted and stored at-70 ℃. Hydrophobic modification and Activity of ClII/A169C mutant. For the modification with N- (1-pyrene) maleimide (Sigma), purified ClII/A169C (4.6 mg/ml in 5mM Na) was diluted with an equal volume of 50mM MES pH 6.5₂HPO₄ph 5.5, 800mM NaCl, 0.2 mM dtt), one twentieth volume of pyrene maleimide, derived from 2.5 mg/ml starting solution in DMSO, was added to the mixture. The samples were incubated for 1 hour at room temperature in the dark. At this point, additional DTT was added to 0.5 mM and the sample was further incubated at room temperature for 1 hour. The activity of the modified proteins was tested directly in the C3H10T 1/2 experiment as described in example 3. Before modification, the specific activity of the protein is EC₅₀= 0.22. mu.g/ml, whereas the specific activity increased to EC after treatment with pyrene maleimide₅₀= 0.08 μ g/ml. A 3-fold increase in specific activity of the modified product was often observed, indicating that the addition of a hydrophobic group close to the C-terminus of Shh can further increase activity compared to the ciii starting material. Compared with the wild type unmodified Sonicheohedgehog protein, the N- (1-pyrene) maleimide modified ClII protein has about 30 times higher efficiency. While pyrene maleimide provides a simple test system for assessing modification at this site, other hydrophobic maleimides or other cystine targeting chemicals may also be used. Example 11: comparison of the efficacy of various hydrophobically modified human Sonicheoghehog in the C3H10T 1/2 experiment

The activity of various hydrophobically modified human Sonic hedgehog (prepared using the chemistry and genetic engineering described in section V) was tested in a C3H10T 1/2 assay as described in example 3.

Each derivative was tested over the concentration range as described in example 3. The concentration of hedgehog derivative that produced 50% of the maximal response in the experiment was compared to the wild type concentration. The relative activities are shown in table 11 below and fig. 13.

Table 11: in the C3H10T 1/2 experimentRelative potency of hedgehog derivatives

Modification of	EC50 (a factor of higher potency than wild-type Shh)
		C: 16 palmitoyl radical	100
C: 14 myristoyl radical	100
		C: 12 lauroyl group	100
C: 10 decanoyl	33
		Isoleucyl-isoleucyl with A169C pyrenyl	30
C: 8 Octanyl radical	10
		Isoleucyl-isoleucyl	10
C：Othiaprolyl	8
		Thiomorpholinyl	5
Phenylalanyl radical	4
		Isoleucyl group	4
N-isopropyl acetamido group	2
		Egycinyl	2
N-ethylmaleimido group	1
		Cystinyl (wild type)	1
Aspartyl radicals	＜1
		Histidinyl radical	＜1

C3H10T 1/2 experiments demonstrated that various hydrophobic modifications of hedgehog increased protein activity compared to wild-type unmodified protein. The hydrophilic modification (aspartic acid and histidine) did not have this effect. Example 12: evaluation of efficacy of Hydrophobically modified human Sonic hedgehog in murine malonate-induced striatal injury experiments

Injection of malonic acid (an inhibitor of the mitochondrial enzyme succinate dehydrogenase) in the murine striatum (a rodent equivalent to primates and putamen) can result in degeneration of spinal cord neurons in the striatum. In humans, degeneration of the central spinal cord neurons in both caudate and putamen is a major pathological feature of huntington's disease. Thus, malonate-induced striatal injury in mice can be used as a model to test whether hydrophobically modified hedgehog proteins can prevent neuronal death that is denatured in huntington's disease.

Various concentrations of hydrophobically modified human Sonic hedgehog were injected into the striatum of Sprague-Dawley mice using stereotactic techniques. Stereotactic injections (2 μ l) were performed under sodium pentobarbital anesthesia (40 mg/kg) and placed at the following sites of synergy: 0.7 mm in front of the chimney, 2.8 mm laterally from the midline, and 5.5 mm vertically from the skull surface at the chimney. Mice were anesthetized with isoflurane at various times (typically 48 hours) after injection of the hydrophobically modified protein and stereotactic injections of malonic acid (2 μmol in 2 μ l) were performed at the same synergistic site of the striatum. 4 days after malonic acid injection, mice were sacrificed and brains were removed for histological analysis. Coronal sections were excised from the striatum at a thickness of 25 μm and stained for cytochrome oxidase activity to distinguish between damaged and undamaged tissue. The lesion volume in the striatum was measured using an image analysis system.

The effect of hydrophobically modified human Sonic hedgehog protein in the malonic acid-induced murine striatal injury model is shown in figure 14. Unmodified Sonic hedgehog (as prepared in example 9), myristoylated Shh (as prepared in example 8), and ciii mutants of Shh (as prepared in example 10) all reduced lesion volume to a similar extent in this model. However, hydrophobically modified proteins (myristoylated Shh and ciii Shh) showed increased potency compared to unmodified Sonic hedgehog. Example 13: sHh-N derivatization with N-octylmaleimide for a final concentration of 1 mg/ml

(1) A solution of 20mM octyl maleimide (m.w. =209) in DMSO (about 4.2 mg/ml) was prepared.

(2) Stock solutions of 10 mg/ml sHh-N (in 5mM NaPO) were diluted 10-fold with PBS (Gibco product #20012-027, pH 7.2)₄pH 5.5, 150mM NaCl, 0.5 mM DTT) to form a1 mg/ml (or 50. mu.M) solution of sHh-N. (Note: DTT competes with sHh-N for maleimide in the subsequent reaction, which is also 50. mu.M in this solution)

(3) To 1 mg/ml sHh-N was added 1/200 volumes of octylmaleimide immediately (i.e., 5. mu.l/ml). This gave a molar ratio of octylmaleimide to sHh-N of 2: 1 (100. mu.M: 50. mu.M).

(4) The tube was gently shaken to mix the above solutions and incubated at room temperature for 1 hour.

(5) Finally, 1/1000 volumes of 0.35M DTT were added to each tube to remove any residual octylmaleimide and act as a reducing agent.

(6) For the vehicle control, the vehicle solution (5mM NaPO) was mixed at a ratio of 1: 10₄pH5.5, 150mM NaCl, 0.5 mM DTT) and PBS (Gibco product # 20012-. 1/400 volumes of 20mM octylmaleimide in DMSO and 1/400 volumes of DMSO were added to form a final concentration of 50 μ M N-octylmaleimide and 0.5% DMSO. Finally, 1/1000 volumes of 0.35M DTT were added. Approximate composition of 1 mg/ml N-octylmaleimide sHh solution: PBS (ca. pH 7.2) 50. mu.M sHh-N50. mu.M DTT 350. mu.M DTT 0.5% DMSON-octylmaleimide carrier solution bound to N-octylmaleimide: PBS (ca. pH7.2) 50. mu.M DTT bound to N-octylmaleimide 350. mu.M DTT 0.5% DMSO to a final concentration of 3 mg/ml

(1) A solution of 60mM octyl maleimide (m.w. =209) in DMSO (about 12.6 mg/ml) was prepared.

(2) Stock solutions of 10 mg/ml sHh-N (in 5mM NaPO) were diluted 10-fold with PBS (Gibco product #20012-027, pH 7.2)₄pH 5.5, 150mM NaCl, 0.5 mM DTT) to form a3 mg/ml (or 150. mu.M) solution of sHh-N. (Note: DTT competes with sHh-N for maleimide in the subsequent reaction, which is also 150. mu.M in this solution)

(3) To 3 mg/ml sHh-N was added 1/200 volumes of octylmaleimide immediately (i.e., 5. mu.l/ml). This gave a molar ratio of octylmaleimide to sHh-N of 2: 1 (300. mu.M: 150. mu.M).

(4) The tube was gently shaken to mix the above solutions and incubated at room temperature for 1 hour.

(5) Finally, 1/1000 volumes of 0.35M DTT were added to each tube to remove any residual octylmaleimide and act as a reducing agent.

(6) For the vehicle control, the vehicle solution (5mM NaPO) was mixed at a ratio of 3: 7₄pH5.5, 150mM NaCl, 0.5 mM DTT) and PBS (Gibco product # 20012-. 1/400 volumes of 60mM octylmaleimide in DMSO and 1/400 volumes of DMSO were added to form a final concentration of 150 μ M N-octylmaleimide and 0.5% DMSO. Finally, 1/1000 volumes of 0.5M DTT were added. Approximate composition of 3 mg/ml N-octylmaleimide sHh solution: PBS (ca. pH 7.2) 150. mu.M sHh-N150. mu.M bound to N-octylmaleimide DTT 500. mu. MDTT 0.5% DMSON-octylmaleimide carrier solution approximate composition: PBS (ca. pH7.2) 150. mu.M DTT 500. mu.M DTT bound to N-octylmaleimide 0.5% DMSO

Reference 1. Perrimon, N. (1995) cell 80, 517-5202. Johnson, R.L. and Tabin, C. (1995) cell 81, 313-3163. Riddle, R.D. et al (1993) cell 75, 1401-14164. Niswender, L. et al (1994) Nature 371, 609-6125. Laufter, E. et al (1994) cell 79, 993-10036. Roberts, D.J. et al (1995) Development 121, 3163-31747. Chiang, C. et al (1996) Nature 382, 609-4138. Bellusci, S. et al (1997) Development 124, 53-639. Marigo, V. et al (1996) Nature 384, 176-Stone, D.M. et al (384), Nature 13411. J. 1996) cell 16210, Alcez. 1996 (16210. 1996) cell 272, 1996-16210. J. 1996) Nature 16211. D.D.J. 10. et al (1996) cell 16210. 1996) Nature 16263-31747. D.7. Wa, 2003-201314. Therond, P.P. et al (1996) Proc.Natl.Acad.Sci.USA 93, 4224-4228515. Lee, J.J. et al (1994) science 266, 1528-153616. Bumcrot, D.A. et al (1995), mol.cell biol.15, 2294-230317. Porter, J.A. et al (1995) Nature 374, 363-36618. Porter, J.A. et al (1996) science 274, 255-25819. Porter, J.A. et al (1995) cell 86, 21-3420. Marigo, V.et al (1995) Genomics 28, 44-5121. Sanicola, M.et al (1997) Proc.Natl.Acad.Sci.USA 94, 38-4322. Specification, B.P.et al (1992) Kintmu.6221. Sanicola, M.1997) Proc.Natl.Acad.Sci.94, 38-6322. Spec.P.P.et al (1996) Natl.Acad.Sci.Sci.Sci.Sci.USA 93, D.D.D.D.D.D.D.D.D.D.D.D.D.D.D.D.D.D.D.D.D.D.D.D.D.D.D.D.D.D.D.D.D.D, n. et al (1997) FEBS letters.404, 319-32326. Ericson, J.et al (1996) cells 87, 661-67327. Wen, D.et al (1996) biochemistry 30, 9700-970928. Bizzozezezezero.O.A. (1996) meth.enzymol.250, 361-38229. Wedegaremer, P.B.et al (1995) J.biol.chem.270, 503-50630. Grosenbach, D.W.et al (1997) J.biol.chem.272, 1956-196431. Pepinsky, R.B.et al (1991) J.biol.m.266, 18244-1822. Tanaka Hall, T.M.et al (1995) Nature 378, 212-33. Mohler and Vanilla (1992) Dev.m.m.266, 18244-1822. Tanaka Hall, T.M.378, T.M.M.M.et al (1995) cells 3339, 1995-166. cell EP.37, 1995-3339. 16, 1995) cell EP 9535, 1995-21635, 1995-EP 9535, 1995, 2, 1995, EP 21635, 1995, EP, EP.7, EP, EP.7, EP, EP.7, EP, 1414-143039. Ericson et al (1995) cells 81, 747-75640. Zoeller et al (1984) Proc. Natl.Acad.Sci.USA, 81, 5662-6641. Kauffman and Sharp, (1982) mol.cell.biol., 2, 1304-131942. Leung et al (1989) technical 1, 11-1543. Mayers et al (1989) science 229, 24244. Harang, SA., (1983) Tetrahedron 39, 345. Itakura et al (1984) Ann.Rev.biochem.53, 32346. Cunningham and Wells (1989) science 244, 1081-108547. Adelman et al, (1983) DNA 2, 18348. Wells et al (1985) genes 34, Acringson et al (1995) cells 81, 747-75640. Zoellman and Sharp et al (1983) natural Ash.1983. Natl.1983. Hassel.1983) genes 34, 1963. 1985, 1963. Wolk.W.3172. 1985. Natl.1983. Hassel et al (1983) genes 34, 1983) and Natl.1983. Hassel et al (1983) genes 34, 1963. Hassel. 1983. Hassel et al (1983) genes 34, 1963. W.3172. Hassel et al (1983) genes) and Natl.1983. Hassel et al (1983) genes 34, Natl.1983. Haslag. 1983, Natl.1983, 1983, Natl,

4475-447854. Szoka et al (1980) Ann. Rev.Biophys.Bioeng.9, 467-50855. Ohsawa et al, (1984) chem.pharm.Bull.32, 244256. Batzri et al, (1973) Biochim.Biophys.acta 298, 1015-101957. Szoka et al, (1978) Proc.Natl.Acad.Sci.USA 75, 4191-419958. Pick, (1981) Arch.biochem.Biophys.212, 186-19459. Kasahara et al, (1977) J.J.251, 7384-739060. Racker et al, (1979) Arch.Biophys.198, 470-47761. Padhaobloc et al, (Natchem.19988. USA, USA 11488. Biophys.1141. Biophys.

1146462. Chou, T.C. and Lipmann, F. (1952) J.biol.chem.196, 8963. Kaminski et al, (1993) New Engl.J.Med.329, 45964. Tomac et al (1995) Nature 373, 335-33965. Gash et al (1996) Nature 380, 252-25566. Hoffer et al (1994) Neuroscience Lett.182.107-11167. Duchen, L.W. and Strich, S.J. (1968) J.Neuro 1. Neurosurg.Psychiatry 31,

535-54268. Kennel et al (1996) Neurobiology of Disease 3, 137-14769. Ripps et al (1995) Proc.Natl.Acad.Sci.USA, 92: 689-69371. Nicholson, L.et al (1995) Neuroscience 66, 507-52172. Beal, M.F. et al (1993) J.neuroscience 13, 4181-419273. Davies, S.et al (1997) cell 90, 537-54874. Hebr-Katz, R. (1993) int.Rev.unal.9, 237-28575. Borg et al (1990) Brain Res.518, 295-29876. Apfel et al (Dufel. 1) Ann.neuros, 29, 87-9077. Nor, C.J. et al (1989) science 244, 182-18878. Thorson, J.S. et al (1989) Methods mol.biol.77, 43-909. Das, A.K. 1997 (11080), Berthon.19880, Biorm. J.1989) Biol.19880, Cherok.11080, cell 11075. 19880, Biorm. 1989, Biorm. 19880, 1989, Biol.19820. 19880, Biorm. 19820, Biol.31. 19880, Biol.31. 19814. J.31. 19880, Biol.31. 19832. 19880, Biol.31. 19832. J.31, Bio, R.J. et al (1993) Lipid Modification of proteins (Lipid Modification of proteins)

Proteins), m.j.schlesinger, ed 83. Krutsch, H.C. & Inman, J.K. (1993) anal.biochem.209, 109-11684. Heitz, j.r. et al (1968) arch.biochem.biophysis.12, 627-63685. stephanini, s. et al (1972) arch.biochem.biophysis.151.28-3486. Kawaguchi, A.J. (1981) Biochem. (Tokyo)89, 337-33987. House, H.O. (1972), modern synthesis reaction (modern synthesis Reactions) W.A.

Benjamin edition

All of the above documents and publications are incorporated herein by reference. Equivalents of

While we have described a number of embodiments of this invention, it will be apparent to those of ordinary skill in the art that the basic embodiments described above may be varied to form other embodiments that utilize the compositions and methods of this invention. It is therefore to be understood that the scope of the present invention includes all alternative embodiments and variations defined in the foregoing description and for the appended claims, and that the invention is not limited to the specific embodiments described in the examples.

Claims (86)

1. An isolated protein comprising an N-terminal amino acid and a C-terminal amino acid, wherein the protein is selected from the group consisting of:

(a) a protein with an N-terminal cystine, having at least one hydrophobic group suspended from the cystine;

(b) a protein with an N-terminal amino acid that is not a cystine appended with at least one hydrophobic group; and

(c) a protein having at least one hydrophobic group substituted for the N-terminal amino acid.

2. The protein of claim 1, wherein the hydrophobic group is a peptide comprising at least one hydrophobic amino acid.

3. The protein of claim 1, wherein the hydrophobic group is a lipid.

4. The protein of claim 1, wherein the protein further comprises a hydrophobic group substituted for or appended to the C-terminal amino acid.

5. The protein of claim 1, wherein the protein is an extracellular signaling protein.

6. The protein of claim 1, wherein the N-terminal amino acid is a functional derivative of cystine.

7. The protein of claim 1, wherein the protein is modified at both the N-terminal amino acid and the C-terminal amino acid.

8. The protein of claim 4 or 7, wherein the protein has a hydrophobic group substituted or suspended from at least one amino acid intermediate the N-terminal and C-terminal amino acids.

9. The protein of claim 1, wherein the protein has a hydrophobic group substituted or pendant from at least one amino acid intermediate the N-terminal and C-terminal amino acids.

10. The protein of claim 3, wherein the lipid moiety is a fatty acid selected from saturated and unsaturated fatty acids having 2-24 carbon atoms.

11. The protein of claim 1, wherein the protein is a hedgehog protein derived from a vertebrate.

12. The protein of claim 11, wherein the hedgehog is derived from a human or a mouse.

13. The protein of claim 11, wherein the vertebrate hedgehog is selected from the group consisting of: sonic, Indian, and Desert hedgehog.

14. The protein of claim 1, further comprising a vesicle in contact with a hydrophobic moiety.

15. The protein of claim 14, wherein the vesicle is selected from the group consisting of: cell membranes, micelles, and liposomes.

16. The protein of claim 11, wherein the protein has an amino acid sequence according to SEQ id no: 1-4.

17. The protein of claim 13, wherein the hedgehog protein has from 1 to 10 amino acids deleted from its C-terminus compared to a wild-type hedgehog protein.

18. The protein of claim 16, wherein at least 60% of the amino acids in the protein are identical to Sonic, Indian, and Desert hedgehog.

19. An isolated protein of the form A-Cys- [ Sp ] -B- [ Sp ] -X, wherein

A is a hydrophobic group, and A is a hydrophobic group,

cys is cystine or a functional derivative thereof,

[ Sp ] is an optional spacer peptide sequence,

b is a protein comprising a plurality of amino acids, and optionally other spacer peptide sequences; and

x is an optional additional hydrophobic group attached to protein B.

20. The isolated protein of claim 19, wherein the isolated protein is a hedgehog protein.

21. The isolated protein of claim 20, wherein if X is present, it is cholesterol.

22. The isolated protein of claim 19, wherein protein B is modified at least one other amino acid with at least one hydrophobic group.

23. The isolated protein of claim 19, wherein the a-Cys linkage is through the amino group of cystine.

24. The isolated protein of claim 19, further comprising a vesicle in contact therewith.

25. The isolated protein of claim 24, wherein the vesicle with which it is contacted is selected from the group consisting of: cell membranes, micelles, and liposomes.

26. A vesicle having a plurality of molecules attached thereto, wherein at least two of the plurality of molecules have the form of claim 19.

27. The vesicle of claim 26, wherein the vesicle is selected from the group consisting of: cell membranes, liposomes, and micelles.

28. An isolated protein having a C-terminal amino acid and an N-terminal thioproline group, wherein the group is formed by the reaction of an aldehyde with the N-terminal cystine of the protein.

29. An isolated protein having a C-terminal amino acid and an N-terminal amide group, wherein the group is formed by the reaction of a fatty acid thioester with an N-terminal cystine of the protein.

30. An isolated protein having a C-terminal amino acid and an N-terminal maleimide group, wherein the N-terminal maleimide group is formed by reaction of a maleimide group with an N-terminal cystine of the protein.

31. The isolated protein of claim 28, 29 or 30, wherein the C-terminal amino acid of the protein is modified with a hydrophobic group.

32. The isolated protein of claim 31, wherein the protein is a hedgehog protein.

33. The isolated protein of claim 32, wherein the C-terminal hydrophobic group is cholesterol.

34. A method of producing a multivalent protein complex, comprising the steps of: the hydrophobic group is attached to the N-terminal cystine, or functional equivalent of the N-terminal cystine, of the protein in the presence of the vesicle.

35. The method of claim 34, wherein the linking step comprises linking a lipid group selected from saturated and unsaturated fatty acids having 2-24 carbon atoms.

36. The method of claim 34, wherein the protein is a hedgehog protein.

37. The method of claim 36, wherein the hedgehog is selected from the group consisting of: sonic, Indian, and Desert hedgehog.

38. The method of claim 36, wherein the hedgehog has a sequence according to seq id NOS: 1-4.

39. The method of claim 34, wherein the linking step comprises linking to a vesicle selected from the group consisting of a cell membrane, a liposome, and a micelle.

40. A method of altering the physico-chemical properties of a protein comprising introducing at least one hydrophobic group on the N-terminal cystine or functional equivalent of the N-terminal cystine of said protein.

41. The method of claim 40, further comprising contacting a hydrophobic group with the vesicle.

42. The method of claim 40, wherein the hydrophobic group is a lipid group selected from saturated and unsaturated fatty acids having 2-24 carbon atoms or a hydrophobic protein.

43. The method of claim 40, wherein the protein is a hedgehog protein.

44. The method of claim 43, wherein the hedgehog protein is selected from the group consisting of: sonic, Indian, and Desert hedgehog.

45. The method of claim 43, wherein the hedgehog has a sequence according to SEQ ID NOS: 1-4, any-number.

46. The method of claim 41, wherein the contacting step comprises contacting with a vesicle selected from the group consisting of a cell membrane, a liposome, and a micelle.

47. A protein complex prepared according to the method of claim 34.

48. A modified protein prepared according to the method of claim 40.

49. The complex of claim 47, wherein said protein is selected from the group consisting of: gelsolin, interferon, interleukins, tumor necrosis factor, single cell colony stimulating factor, granulocyte macrophage colony stimulating factor, erythropoietin, platelet derived growth factor, growth hormone, and insulin.

50. A method of modifying a biologically active protein comprising an N-terminal cystine, comprising reacting the N-terminal cystine with a fatty acid sulfate to form an amide, wherein said modification increases the biological activity of said protein.

51. The method of claim 50, wherein the protein is a hedgehog protein.

52. The method of claim 51, wherein the hedgehog protein is selected from the group consisting of: sonic, Indian, Desert hedgehog, and functional variants thereof.

53. A method of modifying a biologically active protein comprising an N-terminal cystine, comprising reacting the N-terminal cystine with a maleimide group, wherein said modification increases the biological activity of said protein.

54. The method of claim 53, wherein the protein is a hedgehog protein.

55. The method of claim 54, wherein the hedgehog protein is selected from the group consisting of: sonic, Indian, Desert hedgehog, and functional variants thereof.

56. A method of modifying a biologically active protein comprising suspending a hydrophobic peptide on the protein.

57. The method of claim 56, wherein the hydrophobic peptide is suspended from amino acids of a protein selected from the group consisting of: n-terminal amino acids, C-terminal amino acids, amino acids between the N-terminal amino acid and the C-terminal amino acid, and mixtures of the foregoing amino acids.

58. The method of claim 69, wherein the protein is a hedgehog protein.

59. The method of claim 71, wherein the hedgehog protein is selected from the group consisting of: sonic, Indian, and Desert hedgehog.

60. A therapeutic use of the protein of any one of claims 1 or 20, comprising administering the protein to a host.

61. A method of treating a neurological disease in a patient comprising administering to said patient a protein according to any one of claims 1 or 20.

62. The protein of claim 1, wherein the protein is an extracellular signaling protein.

63. The method of claim 57, wherein the suspending step comprises replacing at least the N-terminal amino acid of the protein with at least one hydrophobic amino acid.

64. The method of claim 63, wherein said at least one hydrophobic amino acid is a plurality of isoleucine residues.

65. The method of claim 63, further comprising chemically modifying at least one isoleucine residue.

66. An isolated protein having a C-terminal amino acid and an N-terminal acetamide group formed by the reaction of a substituted acetamide with an N-terminal cystine of the protein.

67. An isolated protein having a C-terminal amino acid and an N-terminal thiomorpholine group formed by reaction of a halo ketone group with an N-terminal cystine of the protein.

68. A method of modifying a biologically active protein comprising an N-terminal cystine, comprising reacting the N-terminal cystine with a substituted acetamide group, wherein the modification increases the biological activity of the protein.

69. The method of claim 68, wherein the protein is a hedgehog protein.

70. The method of claim 69, wherein the hedgehog protein is selected from the group consisting of: sonic, Indian, Desert hedgehog, and functional variants thereof.

71. A method of modifying a biologically active protein comprising an N-terminal cystine, comprising reacting the N-terminal cystine with a halo ketone group, wherein the modification increases the biological activity of the protein.

72. The method of claim 71, wherein the protein is a hedgehog protein.

73. The method of claim 72, wherein the hedgehog protein is selected from the group consisting of: sonic, Indian, Desert hedgehog, and functional variants thereof.

74. A hedgehog polypeptide modified with one or more lipophilic groups, with the proviso that if the hedgehog polypeptide is the mature N-terminal proteolytic fragment of a hedgehog protein, the lipophilic group is not a sterol at the C-terminal residue.

75. A hedgehog polypeptide modified at an internal amino acid residue with one or more lipophilic groups.

76. A hedgehog polypeptide modified with one or more lipophilic aromatic hydrocarbons.

77. The hedgehog polypeptide of any one of claims 74-76, which is a purified protein preparation.

78. The hedgehog polypeptide of any one of claims 74-76, which is a pharmaceutical agent.

79. The hedgehog polypeptide of claim 74 or 75, wherein the lipophilic group is selected from the group consisting of: fatty acids, lipids, esters, alcohols, cage structures, and aromatic hydrocarbons.

80. The hedgehog polypeptide of claim 76 or 79, wherein the aromatic hydrocarbon is selected from the group consisting of: benzene, northern, phenanthrene, anthracene, naphthalene, pyrene, chrysene and tetracene.

81. The hedgehog polypeptide of claim 80, wherein the aromatic hydrocarbon is pyrene.

82. The hedgehog polypeptide of claim 74 or 75, wherein the lipophilic group is selected from the group consisting of: isoprenoids, terpenes, and polycyclohydrocarbons.

83. The hedgehog polypeptide of claim 82, wherein the lipophilic group is selected from the group consisting of: adamantane, fullerene, vitamin, polyethylene glycol, oligoethylene glycol, (C1-C18) alkylphosphoric acid diester, -OCH2-CH (OH) -O- (C12-C18) -alkyl

84. The hedgehog polypeptide of claim 83, wherein the lipophilic group is selected from the group consisting of: 1-or 2-adamantylacetyl, 3-methyladamantan-1-ylacetyl, 3-methyl-3-bromo-1-adamantylacetyl, 1-decallacetyl, camphoracetyl, bornaneacetyl, noradamantylacetyl, norbornaneacetyl, bicyclo [ 2.2.2. ] -oct-5-eneacetyl, 1-methoxybicyclo [ 2.2.2 ] -oct-5-ene-2-carbonyl, cis-5-norbornene-endo-2, 3-dicarbonyl, 5-norbornene-2-ylacetyl, (1R) - (-) -myrtenone acetylyl, 2-norbornaneacetyl, anti-3-oxo-tricyclo [ 2.2.1.0 <2, 6> ] -heptane-7-carbonyl, decanoyl, dodecanoyl, dodecenoyl, tetradecadienoyl, decynoyl or dodecenoyl.

85. The hedgehog polypeptide of any one of claims 74-76, wherein the lipophilic group enhances a biological activity of the polypeptide relative to an unmodified hedgehog polypeptide.

86. A method of altering the growth state of a cell that responds to hedgehog signaling, comprising contacting the cell with the lipophilically modified hedgehog polypeptide of any one of claims 74-76.