AU2011213827A1 - Polymeric prodrug with a self-immolative linker - Google Patents

Abstract

A cascade carrier linked prodrug is described which comprises a biologically active moiety and - a masking group having at least one nucleophile and being distinct from the carrier. In one preferred embodiment of the invention, the prodrug has the structure represented by formula (I).

Description

AUSTRALIA Regulation 3.2 Patents Act 1990 Complete Specification Standard Patent APPLICANT: Ascendis Pharma A/S Invention Title: POLYMERIC PRODRUG WITH A SELF-IMMOLATIVE LINKER The following statement is a full description of this invention, including the best method of performing it known to me: Prodrug Linker Field 5 The present invention is directed to polymeric prodrugs having temporary linkages to amino groups of biologically active entities such as peptides, proteins, natural products or synthetic chemical compounds. Backarourd 0Typically Polymers are either used in a non-covalent fashion, with the drug copcaln polymsi ae ia formulated into a solvent-polymer mixture, or by com pound p ysicochem ically form ul en to o e o t ed u ' f n t o a permanent covalent attachment of a polymer reagent to one Of the drug's functional groups. 15dtodptfmltoifrlog Non-covalent drug encapsulation has been applied to depot fonulatios for long acting release profiles. Typically, drug is mixed with polymer material and processed in such fashion, that the drug becomes distributed throughout the bulk polymer material. Such polymer-protein aggregates may be shaped as micropartiCleS which are 20 administered as an injectable suspension or they are o formulated ase t h re administered in a single bolus injection. Drug release occurs when the polymer swells or degradation of the polymer allows for diffusion of the drug to the exterior. Such degradation processesmay be autohydrolyti or enzYie-catalyzed. A example for a marketed drug based onbolus administration of a drug-polymer gel is Lupron Depot. 25 An example for a marketed drug based on suspended microparticles is Nutropin Depot A disadvantage of the non-covalent approach is that in order to preveE~t uncontrolled, burst-type release of the drug, encapsulation has to be highly efficient by creating a 30 sterically highly crowded environment Restraining the diffusion of ar. unbound, water soluble drug molecule requires strong van der Waals contacts, frequently mediated through hydrophobic moieties. Many conformationaly sensitive therapeutics such as proteins or peptides are rendered dysfunctional during the encapsulation process and/or during subsequent storage. In addition, such amino page I containing drug compounds readily undergo side reactions with polymer degradation products (D.H. Lee et al., j. Contr. Rel., 2003, 92, 291-299). Furthermore, dependence of the release mechanism upon biodegradation may cause interpatient variability. 5 Alternatively, drugs may be conjugated to polymers through permanent covalent bonds. This approach is applied to various classes of molecules, from so-called small molecules, through natural products up to larger proteins. Many small molecule medicinal agents, like alkaloids and anti-tumor agents, show 10 low solubility in aqueous fluids. One way to solubilize these small molecule compounds is to conjugate them to hydrophilic polymers. A variety of water-soluble polymers, such as human serum albumin dextran, lectins, poly(ethylene glycol) (PEG), poly(styrene-co-maleic anhydride), poly(N-hydroxypropyhnethacrylamide), poly(divinyl ether-co-maleic anhydride), hyaluronic acid have been described for this 15 purpose (R. Duncan, Nature Rev. Drug Disc., 2003, 2, 347-360). A major challenge in cancer therapy is to selectively target cytotoxic agents to tumor cells. A promising method to accumulate small molecule anticancer agents in tumor tissue and decrease undesirable side effects of these agents is the attachment of the 20 cytotoxin to a macromolecular carrier. The passive targeting of polymeric drug conjugates to tumors is based on the so-called enhanced permeability and retention effect (EPR) as described by Matsumura, Y. and Maeda, H., in Cancer Res., 1986, vol 6, pp 6387-6392. As a result, several polymer-drug conjugates have entered clinical trial as anticancer agents. Covalent modification of biological molecules with poly(ethylene glycol) has been extensively studied since the late 1970s. So-called PEGylated proteins have shown improved therapeutic efficacy by increasing solubility, reducing immunogenicity, and increasing circulation half-live in vivo due to reduced renal clearance and proteolysis 30 by enzymes (see, for example, Caliceti P.,Veronese F.M., Adv. Drug Deliv. Rev. 2003, 55, 1261-1277). However, many medicinal agents such as TNFalfa2, saquinavir or somatostatin are inactive or show decreased biological activity when a polymer is covalently Page 2 conjugated to the drug molecule (T. Peleg-Shulman et al., J. Med. Chem., 2004, 47, 4897-4904). In order to avoid shortcomings imposed by either non-covalent polymer mixtures or 5 permanent covalent attachment, it may be preferable to employ a prodrug approach for chemical conjugation of drug to polymer carrier. In such polymeric prodrugs, the biologically active moieties are typically linked to the polymeric carrier moiety by a temporary bond formed between the carrier moiety and a hydroxy, amino or carboxy group of the drug molecule (such as is shown in Fig. 1). 10 Prodrugs are therapeutic agents that are almost inactive per se but are predictably transformed into active metabolites (see B. Testa, J.M: Mayer in Hydrolysis in Drug and Prodrug Metabolism, Wiley-VCH, 2003, page 4). The carrier prodrug approach may be applied in such a fashion that the medicinal agent is released in vivo from the is polymer in order to regain its biological activity. The reduced biological activity of the prodrug as compared to the released drug is of advantage if a slow or controlled release of the drug is desired. In this case, a relatively large amount of prodrug may be administered without concomitant side effects and the risk of overdosing. Release of the drug occurs over time, thereby reducing the necessity of repeated and frequent 20 administration of the drug. Prodrug activation may occur by enzymatic or non-enzymatic cleavage of the temporary bond between the carrier and the drug molecule, or a sequential: combination of both, i.e. an enzymatic step followed by a non-enzymatic 25 rearrangement. In an enzyme-free in vitro environment such as an aqueous buffer solution, a temporary bond such as an ester or amide may undergo hydrolysis, but the corresponding rate of hydrolysis may be much too slow and not therapeutically useful. In an in vivo environment, esterases or amidases are typically present and may cause significant catalytic acceleration of the kinetics of hydrolysis from twofold up to 30 several orders of magnitude (see, for example, R.B. Greenwald et aL. J.Med.Chem. 1999, 42 (18), 3857-3867). Page 3.

Definitions based on IUPAC (as given under http://www.chem.qmul.ac.uk/iupac/medchem/ (accessed on 8 March 2004) 5 Prodrug A prodrug is any compound that undergoes biotransformation before exhibiting its pharmacological effects. Prodrugs can thus be viewed as drugs containing specialized non-toxic protective groups used in a transient manner to alter or to eliminate undesirable properties in the parent molecule. 10 Carrier-linked prodrug (Carrier prodrug) A carrier-linked prodrug is a prodrug that contains a temporary linkage of a given active substance with a transient carrier group that produces improved physicochemical or pharmacokinetic properties and that can be easily removed in 1s vivo, usually by a hydrolytic cleavage. This is shown graphically in Fig. 1. Cascade prodrug A cascade prodrug is a carrier prodrug for which the cleavage of the carrier group becomes effective only after unmasking an activating group. 20 Polymeric cascade prodrug A polymeric cascade prodrug is a carrier prodrug that contains a temporary linkage of a given active substance with a transient polymeric carrier group for which the cleavage of the carrier becomes effective only after unmasking an activating group. 25 Bioprecursor prodrug A bioprecursor prodrug is a prodrug that does not imply the linkage to a carrier group, but results from a molecular modification of the active principle itself. This modification generates a new compound, able to be transformed metabolically or 30 chemically, the resulting compound being the active principle. Biotransformation Biotransformation is the chemical conversion of substances by living organisms or enzyme preparations. Page 4.

Prodrugs fall in two classes, bioprecursors and carrier-linked prodrugs. Bioprecursors do not contain a carrier group and are activated by the metabolic creation of a functional group. In carrier-linked prodrugs the active substance is linked to a carrier 5 moiety by a temporary linkage. This invention is concerned with polymeric carrier linked or macromolecular prodrugs, where the carrier itself is a macromolecule such as a carrier protein or polysaccharide or polyethylene glycol. Specifically, the invention relates to polymeric carrier-linked prodrugs for which this cleavage between polymer and drug proceeds in two steps according to a cascade mechanism. 10 Cleavage of a carrier prodrug generates a molecular entity (drug) of increased bioactivity and at least one side product, the carrier. This side product may be biologically inert (for instance PEG) or may have targeting properties (for instance antibodies). After cleavage, the bioactive entity will reveal at least one previously 15 conjugated and thereby protected functional group, and the presence of this group typically contributes to the drug's bioactivity. In order to implement a prodrug strategy, at least one certain functional group in the drug molecule is employed for attachment of the carrier polymer. Preferred functional 20 groups are hydroxyl or amino groups. Consequently, both the attachment chemistry and hydrolysis conditions vary greatly between these two functionalities. In a simple one-step mechanism, the prodrug's temporary linkage is characterized by an intrinsic lability or enzyme dependence. The susceptibility of this linkage to 25 hydrolysis in an aqueous environment with our without enzyme catalysis controls the cleavage kinetics between polymeric carrier and drug. Numerous macromolecular prodrugs are described in the literature where the temporary linkage is a labile ester bond. In theses cases, the functional group provided by the bioactive entity is either a hydroxyl group or a carboxylic acid (e.g. Y. Luo, MR Ziebell, GD Prestwich, "A 30 Hyaluronic Acid - Taxol Antitumor Bioconjugate Targeted to Cancer Cells", Biomacromolecules 2000, 1, 208-218, J Cheng et al, Synthesis of Linear, beta Cyclodextrin Based Polymers and Their Camptothecin Conjugates, Bioconjugate Chem. 2003, 14, 1007-1017, R. Bhatt et al, Synthesis and in Vivo Antitumor Activity of Poly(L-glutamic acid) Conjugates of 20(S)-Campthothecin, J. Med. Chem. 2003, Page S.

46, 190-193 ; R.B. Greenwald, A. Pendri, C.D. Conover: H. Zhao, Y.H. Choe, A. Martinez, K. Shum, S. Guan, J. Med- Chem., 1999, 42, 3657-3667; B. Testa, J.M: Mayer in Hydrolysis in Drug and Prodrug Metabolism, Wiley-VCH, 2003,Chapter 8). 5 Especially for therapeutic biomacromolecules but also for certain small molecule drugs, it may be desirable to link the macromolecular carrier to amino groups of the bioactive entity (i.e. N-terminus or lysine amino groups of proteins). This will be the case if masking the drug's bioactivity requires conjugation of a certain amino group of the bioactive entity, for instance an amino group located in an active center or a 10 region or epitope involved in receptor binding. Also, during preparation of the prodrug, amino groups may be more chemoselectively addressed and serve as a better handle for conjugating carrier and drug because of their greater nucleophilicity as compared to hydroxylic or phenolic groups. This is particularly true for proteins which may contain a great variety of different reactive functionalities, where non 15 selective conjugation reactions lead to undesired product mixtures which require extensive characterization or purification and may decrease reaction yield and therapeutic effciency of the product. Amide bonds as well as aliphatic carbamates are much more stable towards hydrolysis 20 than ester bonds, and the rate of clevage would be too slow for therapeutic utility in a carrier-linked prodrug. Therefore it is advantageous to add structural chemical components such as neighbouring groups in order to exert control over the cleavability of the prodrug amide bond. Such additional cleavage-controlling chemical structures that are not provided by the carrier entity nor by the drug are 25 called linker. Prodrug linkers can have a strong effect on the rate of hydrolysis of a given temporary bond. Variation of the chemical nature of these linkers allows to engineer the tinker properties to a great extent. For instance, prodrug linkers may be designed for enzyme-selectivity. Prerequisite for 30 enzymatic dependence is that the tinker structure displays a structural motif that is recognized as a substrate by a corresponding endogenous enzyme (Fig. 2). Enzyme-catalyzed acceleration of prodrug cleavage is a desirable feature for organ or cellular targeting applications. Targeted release of the bioactive entity is effected, if Page 6/ an enzyme, that selectively cleaves the linkage, is specifically present in the organ or cell-type chosen for treatmenL A typical property of an enzyme-dependent temporary linkage is its stability with 5 respect to hydrolysis. The temporary linkage itself will not undergo autohydrolysis at a rate that would release drug to such an extent that a therapeutic effect could be induced in a normal dosing regime. It is only in the presence of the enzyme, that the attack of the enzyme on the linkage causes a significant acceleration of cleavage and concomitant an enhancement of free drug concentration. 10 Several examples have been published for the prodrug activation of amine-containing biologically active moieties by specific enzymes for targeted release. In these cases, cleavage occurs in a one-step process which is catalyzed by the enzyme. G. Cavallaro et al., Bioconjugate Chem. 2001, 12, 143-151 describe the enzymatic release of an 15 antitunoral agent by the protease plasmin. Cytarabin is coupled via the tripeptide sequence D-Val-Leu-Lys to the polymer alpha, beta-poly(N-hydroxyethl)-DL aspartamide (PHEA). Enzymatic release of cytarabin is effected by the protease plasmin which concentration is relatively high in various kinds of tumor mass. 20 Further examples for antitumoral polymeric prodrugs activated by specific enzymes like beta lactamase (R. Satchi-Fainaro et al., Bioconjugate Chem. 2003, 14, 797-804) and cysteine proteases like cathepsin B (R. Duncan et al. J. Contr. Release 2001, 74, 135-146) have been described. Wiwattanapatapee et al. (2003) outline a dendrimer prodrug for colonic delivery of 5-aminosalicylic acid. The drug molecule is 25 conjugated by an azo bond to "generation 3" PAMAM dendrimer. 5-aminosalicylic acid is released in the colon by a bacterial enzyme called azo reductase (W. R. Wiwattanapatapee, L. Lonlim, K. Saramunee, J. Controlled Release, 2003, 88: 1-9). A.J. Garman et al. (A.J. Garman, S.B. Kalindjan, FEBS Lett, 1987, 223 (2), 361-365 30 1987) use PEG5000-maleic anhydride for the reversible modification of amino groups in tissue-type plasminogen activator and urokinase. Regeneration of functional enzyme from PEG-uPA conjugate upon incubation at pH 7.4 buffer by cleavage of the malearnic acid linkeage follows first order kinetics with a half-life of 6.1 h. The prodrug cleavage was not investigated in the presence of enzymes, and it can be Page 7 .expected - as explained above - that proteases present in the in vivo environment will sianificantly contribute to the cleavage of the temporary amide linkage. A further disadvantage of this linkage is the lack of stability of the conjugate at lower pH values. This limiits the applicability of the linker to active agents which are stable at s basic pH values, as purification of the active agent polymer conjugate has to be performed under basic conditions to prevent premature prodrug cleavage. Cascade mechanisms have proven particularly useful in the controlled release of drugs containing amino-group functionalities because linker cleavage characteristics can be 10 optimized with greater flexibility than in simple one-step prodrugs. Cascade cleavage is enabled by linker compounds that are composed of a structural combination of a masking group and an activating group. The masking group is attached to the activating group by means of a first temporary linkage such as an ester 15 or a carbamate. The activating group is attached to an amino-group of the drug molecule through a second temporary linkage, for instance a carbamate. The stability, or susceptibility to hydrolysis of the second temporary linkage is dependent on the presence or absence of the masking group. In the presence of the masking group, the second temporary linkage is highly stable and unlikely to release drug with 20 therapeutically useful kinetics. In the absence of the masking group, this linkage becomes highly labile, causing rapid cleavage and drug release. Cleavage of the first temporary linkage is the rate-limiting step in the cascade mechanism. This first step may induce a molecular rearrangement of the activating 25 group such as a 1,6-elimination. The rearrangement renders the second temporary linkage so much more labile that its cleavage is induced. Ideally, the cleavage rate of the first temporary linkage is identical to the desired release rate for the drug molecule in a given therapeutic scenario. Furthermore, it is desirable that cleavage of the second temporary linkage is instantaneous after its lability has been induced by cleavage of 30 the first temporary bond. A variety of examples exist for cascade carrier prodrugs where the masking group functionality is performed by the carrier polymer itself as shown diagrammatically in Fig. 3. In the systems discussed below, the masking group is not only part of the Page : carrier but has also been engineered for enzyme-dependence (Fig. 4). Only in the presence of a corresponding enzyme is the rate of cleavage of the first temporary linkage sufficiently accelerated for therapeutic use. 5 R.B. Greenwald, A. Pendri, C.D. Conover, H, Zhao, Y.H. Choe, A. Martinez, K. Shun, S. Guan, J. Med. Cherm, 1999, 42, 3657-3667 & PCT Patent Application

WO

A-99/30727 describe a methodology for synthesizing poly(ethylene glycol) prodrugs of amino-containing small molecule compounds based on 1,4- or 1,6-benzyl elimination. In this approach, poly(ethylene glycol) as the polymeric carrier is 10 attached to the benzyl group by means of a first temporary linkage such as an ester, carbonate, carbamate, or axnide bond. The benzyl group serves as the activating group, and the PEG polymer also has the function of the masking group in this cascade cleavage mechanism. The amino group of the drug molecule is linked via a second temporary linkage, containing a carbamate group, to the benzyl moiety. The is release of PEG from the drug molecule is initiated by enzymatic cleavage of the first temporary linkage followed by a rapid 1,4- or 1,6-benzyl elimination, initiating cleavage of the second temporary linkage. The same linker system is also used for releasable poly(ethylene glycol) conjugates of 20 proteins (S. Lee, R.B. Greenwald et al. Bioconj. Chern. 2001, 12 (2), 163-169). Lysozyme is used as model protein because it loses its activity when PEGylation takes place on the epsilon-amino group of lysine residues. Various amounts of PEG linker were conjugated to the protein. Regeneration of native protein from the PEG conjugates occurs by enzymatic cleavage in rat plasma or in non-physiological high 25 pH buffer. Greenwald et al. published in 2000 a poly(ethylene glycol) drug delivery system of amino-containing prodrugs based on trimethyl lock lactonization (R.B. Greenwald et al. J.Med.Chem. 2000, 43(3), 457-487; PCT Patent Application No. WO-A 30 02/089789). In this prodrug system, substituted o-hydroxyphenyl-dimethylpropionic acid is linked to PEG by an ester, carbonate, or carbamate group as a first temporary linkage and to amino groups of drug molecules by means of an amide bond as second temporary linkage. The rate-determining step in drug release is the enzymatic Page 9 cleavage of the first linkage. This step is followed by fast amide cleavage by lactonization, liberating a potentially toxic aromatic lactone side product Similar prodrug systems were described by F.MH. DeGroot et al. (W002083 180 and 5 W004043493A1) and D. Shabat et aL (W004019993A1). W002083 180 discloses a prodrug system with elongated and multiple linkers based on 1,(4+2n) elimination. The masking moieties in these examples were specifically designed for enzymatic cleavage. This approach was extended to dendritic prodrug system where one enzymatic activating event triggered the release of more than one drug molecule 10 (WOO4043493A1). W04019993Al discloses a similar prodrug system based on a self-immolative dendrimer releasing many drug moieties upon a single enzymatic activating event These systems are characterized by the absence of a polymeric carrier. Instead, oligomerization of prodrug linker components provides for a high molecular weight of the prodrug, and prodrug cleavage generates linker residues and is free drug, but no polymeric entity is released. The disadvantage in the abovementioned prodrug systems desribed by Greenwald, DeGroot and Shabat is the release of potentially toxic aromatic small molecule side products like quinone methides after cleavage of the temporary linkage. The 20 potentially toxic entities are released in a 1:1 stoichiometry with the drug and can assume high in vivo concentrations. This risk factor is even greater if self-immolative dendritic structures based on oligomers of the activating group are employed and more aromatic side products than drug molecules are released. 25 More recently, R.B. Greenwald et at. (Greenwald et al. J. Med.Chem. 2004, 47, 726 734) described a PEG prodrug system based on bis-(N-2-hydroxyethyl)glycin amide (bicin aide) linker. In this system two PEG molecules are linked to a bicin molecule coupled to an amino group of the drug molecule. The first two steps in prodrug activation is the enzymatic cleavage of both PEG molecules. Different linkages 30 between PEG and bicin are described resulting in different prodrug activation kinetics. The main disadvantage of this system is the slow hydrolysis rate of bicin amide conjugated to the drug molecule (tl/2 = 3 h in phosphate buffer) which results in the release of a bicin-modified prodrug intermediate that may show different Page 10 pharmacokinetic and pharmacodynamic properties as compared to the parent drug molecule. Cascade prodrugs with masking groups that are part of the carrier polymer are limited 5 in the control of drug release kinetics. As masking group cleavage is the rate-limiting step in the cascade mechanism, its molecular structure governs the kinetics. If the carrier polymer is identical to the masking group, the structural flexibility is restricted to the polymers' features. Alternatively, if the polymer requires structural modification in order to match the requirements for controlled cleavage, synthesis of io corresponding structures may become more difficult. Also, the incorporation of masking group features into a polymer may change its safety profile. Therefore is it preferred to structurally separate the masking group and the carrier. This may be achieved by employing a permanent bond between polymer carrier and 15 activating group. This stable bond does not participate in the cascade cleavage mechanism. If the carrier is not serving as a masking group and the activating group is coupled to the carrier by means of a stable bond, release of potentially toxic side products such as the activating group is avoided.The stable attachment of activating group and polymer also suppresses the release of drug-linker intermediates with 20 undefined pharmacology. Systems have been developed for targeted delivery of therapeutic agents by rendering the masking group enzyme-dependent. Only in the presence of a corresponding enzyme is the rate of cleavage of the first temporary linkage connecting the masking 25 group with the activating group sufficiently accelerated for therapeutic use. Antczak et al. (Bioorg Med Chem 9 (2001) 2843-48) describe a reagent which forms the basis for a macromolecular cascade prodrug system for amine-containing drug molecules. In this approach an antibody serves as carrier, a stable bond connects the 30 antibody to an activating moiety, carrying an enzyrmatically cleavable masking group. Upon enzymatic removal of the ester-linked masking group, a second temporary bond cleaves and releases the drug compound, as shown in Fig. 6. Page il - D. Shabat et al. (Chem. Eur. J. 2004, 10, 2626-2634) describe a polymeric prodrug system based on a mandelic acid activating moiety. In this system the masking group is linked to the activating moiety by a carbamate bond. The activating moiety is conjugated permanently to a polyacrylamide polymer via an aide bond. After s enzymatic activation of the masking group by a catalytic antibody, the masking group is cleaved by cyclization and the drug is released. The activating moiety is still connected to the polyacrylamide polymer after drug release. M.-R. Lee et al. describe (Angew. Chern. 2004,116, 1707-1710) a similar prodrug )0 system based on mandelic acid activating moiety and an enzymatically cleavable ester-linked masking group. In all of these described prodrug-polymer systems the masking group is specifically designed to be substrate to an enzyme, and masking group cleavage will almost is entirely depend upon enzymatic catalysis with the disadvantages of interpatient variability, injection site variability and poor in vitro-in vivo correlation. A major drawback of predominantly enzymatic cleavage is interpatient variability. Enzyme levels may differ significantly between individuals resulting in biological 20 variation of prodrug activation by enzymatic cleavage. Enyzme levels may also vary depending on the site of administration, for instance it is known that in the case of subcutaneous injection, certain areas of the body yield more predictable therapeutic effects than others. To reduce this unpredictable effect, nLon-enzymatic cleavage or intramolecular catalysis is of particular interest (see, for example, B. Testa, LM: 25 Mayer in Hydrolysis in Drug and Prodrug Metabolism, Wiley-VCH, 2003, page 5). Furthermore, it is difficult to establish an in vivo-in vitro correlation of the pharmacokinetic properties for such enzyme-dependent carrier-linked prodrugs. In the absence of a sound in vivo-in intro correlation the optimization a release profile 30 becomes a cumbersome task. Also, the need for enzyme selectivity imposes a severe limitation on the structural features that can be used in the prodrug linker. This restriction greatly hinders the Page 12 development of a sound structure-activitY relationship and consequently the optimization of liner cleavage kinetics. For these reasons, there is a need to provide novel linker and/or carrier technologies 5 for forming polymeric prodrugs of amine containing active agents in order to overcome the limitations of the described polymeric prodrugs. Page 13.

Detailed description of the invention The present invention addresses the disadvantages described above. The invention provides for polymeric cascade prodrugs characterized by a masking group containing 5 a nucleophile and being distinct from the carrier. The nucleophile is in a suitable distance to a first temporary linkage with an aromatic activating group capable of undergoing a 1,(4+ 2 p) elimination reaction (with p = 0, 1, 2, 3, 4,...) after cleavage of the first temporary linkage. The invention is further 10 characterized by the activating group being connected to the amino group of a drug molecule through a second temporary bond which is cleaved as a consequence of the 1,(4+2p) elimination. An additional structural feature is the attachment of a polymeric carrier to the activating group by means of a permanent bond. 15 The masking groups according to the present invention contain at least one nucleophile Nu. This nucleophile, which can be for example a primary, secondary or tertiary amino group can assist the cleavage of the masking group from the activating moiety by intramolecular catalysis or cyclization. Page 34 The invention provides for poly-meric cascade prodrugs and corresponding polymeric cascade prodrug tinker reagents of Formula Ia or Ib. R4 RI 1 Y2 01Y3 [rT Nu-W-Y 4 R3 Ar Formula Ia 10 R4 X-'RI Y R2 Y. 15 Y2 0 Y3 i T Nu-W-Y 4 Ar R3 Formula Ib 20 wherein Y 1 to Ys, RI to R4, T, X, W, Nu and Ar are defined below: Native drug release is effected by a two step mechanism. The first step is the rate determining cleavage of the first temporary linkage between the masking group 25 y Nu-W-Ya and the activating moiety [R4 R4 X X or R2 Y2 0 20 R3 R3 Ar Ar Page is from the polyrneric prodrug in vivo. As described above, cleavage of the masking group may be mediated by an enzymatic or a non-enzymatic step, such as pH-dependent hydrolysis or intramolecular 5 cyclization. In the preferred embodiment of thbe invention, cleavage is effected non enzymatically by intramolecular cyclization or catalysis. The half-life of the cleaveage kinetics in an aqueous buffer of pH 7.4 at 37 o C of the masking group according to the present invention is preferably between 1 hour and 6 months. more preferably between I day and 3 months, and most preferably between 1 day and 2 months. The second and final step in release of the regenerated native drug is the fast, spontaneous and irreversible so-called 1.4- or 1,6 or 1,(4+2p) (in which p = 2,3, 4 or higher) elimination of the R4 ]"IR4 X--'.-R1 n R1 Sz R2

Y

2 or Y2 Ar R3 N R3 20 moiety of the remaining polymeric prodrug of formula la or formula Ib, respectively. This mechanism of native drug release from a- polymeric prodrug triggered by hydrolytic cleavage of the masking group followed by a 1,6-elimination step of the activating group is exemplified by a polymeri c prodrug according to the present 25 invention. Page 16 Polymer R7 RB O Nu-W 04 0 N-Drug H masking group cleava ge by intramolecular catalysis Polymer 10 R7R OH+ + H~ CO I ~ N -[Drug

OH

2 15 Polymer + HN-Drug 20 Defnition of Yi to Y 5 , R1 to R4, T, X, W, Nu and Ar in formula la or Ib T is D or A In the case where the inventive structure is a polymeric cascade prodrug linker reagent, T is A, and A is a leaving group. Non-limiting examples of suitable leaving 25 groups A include but are not limited to chloride, bromide, fluoride, nitrophenoxy, imidazolyl, N-hydroxysuccinimidyl, N-bydroxybenzotriazolyl,

N

hydroxyazobenzotriazolyl, pentafluorphenoxy, N-hydroxysulfosuccinimidyl, or any other leaving group known by those skilled in the art. 30 In the case where the inventive structure is a polymeric cascade prodrug, T is D, and D is a residue of an amine-containing biologically active material including but not limited to small molecule bioactive agents or biopolymers like proteins, polypeptides and oligonucleotides (RNA, DNA), peptide nucleic acids (PNA), Page 17 Note that in this description reference is often made to prodrugs. A true prodrug is found when T is the residue of the amine-containing biologically active material or moiety. If T is a leaving group A then the formula represents a polymeric cascade prodrug linker reagent . For simplicity these will be referred to prodrugs in this 5 description. It will be understood from the context whether a true prodrug or a reagent as a predursor is meant. Suitable organic small molecule bioactive moieties include, without limitation, moieties such as central nervous system-active agents, anti-infective, anti-neoplastic, 10 antibacterial, anti-fungal, analgesic, contraceptive, anti-inflammatory, steroidal, vasodilating, vasoconstricting, and cardiovascular agents with at least one primary or secondary amino group. Non-exclusive examples of such compounds are daunorubicin, doxorubicin, idarubicin, mitoxantron, aminoglutethimide, amantadine, diaphenylsulfon, ethambutol, sulfadiazin, sulfamerazin, sulfamethoxazol, sulfalen, 15 clinafloxacin, moxifloxacin, ciprofloxaxin, enoxacin, norfl oxacin, neomycin B, sprectinomycin, kanamycin A, meropenem, dopamin, dobutamin, lisinopril, serotonin, carbutamid, acivicin, etc. Suitable proteins and polypeptides having at least one free amino group include but 20 are not limited to ACTH, adenosine deaminase, agalsidase, albumin, alfa-1 antitrypsin (AAT), alfa-1 proteinase inhibitor (API), alteplase, anistreplase, ancrod serine protease, antibodies (monoclonal or polyclonal, and fragments or fusions), antithrombin III, antitrypsins, aprotnin, asparaginases, biphalin, bone-morphogenic proteins, calcitonin . (salmon), collagenase, DNase, endorphins, enfuvirtide, 25 enkephalins, erythropoietins, factor VIIa, factor VIII, factor VfIla, factor IX, fibrinolysin, fusion proteins, follicle-stimulating horiaones, granulocyte colony stimulating factor (G-CSF), galactosidase, glucagon, glucagon-like peptides like GLP 1, glucocerebrosidase, granulocyte macrophage colony stizMulating factor (GM-CSF), phospholipase-activating protein (PLAP), gonadotropin chorionic (hCG), 30 hemoglobins, hepatitis B vaccines, hirudin, hyaluronida-ses, idurnonidase, immune globulins, influenza vaccines, interleukins (1 alfa, 1 beta, 2, 3, 4, 6, 10, I1, 12), IL-1 receptor antagonist (rhIL-1ra), insulins, interferons (alfa 2a, alfa 2b, alfa 2c, beta la, beta 1b, gamma la, gamma lb), keratinocyte growth factor (KGF), transforming growth factors, lactase, leuprolide, levothyroxine, luteinizing hormone, lyme vaccine, Page Is natriuretic peptide, pancrelipase, papain, parathyroid hormone, PDGF, pepsin, platelet activating factor acetylhydrolase (PAF-AH), prolactin, protein C, octreotide, secretin, sermorelin, superoxide dismutase (SOD), somatropins (growth hormone), somatostatin, streptokinase, sucrase, tetanus toxin fragment, iilactase, thrombins, 5 thymosin, thyroid stimulating hormone, thyrotropin, tumor necrosis factor (TNF), TNF receptor-IgG Fc, tissue plasminogen activator (tPA), TSH, urate oxidase, urokinase, vaccines, plant proteins such as lectins and ricins. Also included herein is any synthetic polypeptide or any portion of a polypeptide with 10 in vivo bioactivity. Furthermore, proteins prepared by recombinant DNA methodologies including mutant versions of aforementioned proteins, antibody fragments, single chain binding proteins, catalytic antibodies and. fusion proteins are included. is Preferred proteins are antibodies, calcitonin, G-CSF, GM-CSF, erythropoietins, hemoglobins, interleukins, insulins, interferon, SOD, somatropin, TNF, TNF receptor-IgG Fc, and GLP- 1. X is a spacer moiety such as R5-Y6 20 Yi, Y 2 can each be either 0, S, or NR6, independently of each other.

Y

3 , Y 5 can each be either 0 or S, independently of each other.

Y

4 is 0, NR6, or -C(R7)(R8)

Y

6 is 0, S, NR6, succinimide, maleimide, unsaturated carbon-carb on bonds or any 25 beteratom containing a free electron pair, or is not present. R2 and R3 are selected independently from hydrogen, substituted or non-substituted linear, branched or cyclical alkyl or heteroalkyl, aryls, substituted aryls, substituted or non-substituted heteroaryls, cyano, nitro, halogen, carboxy, carbox)yalkyl, alkylcarbonyl, carboxamidoalkyl, etc. 30 The term "heteroalkyl" in the context of the present invention denotes (linear, cyclical or branched) alkyl chains where the alkyl chains contain or are substituted with at any position one or more heteroatoms, selected independently from 0, S, N, P, Si, Cl, F, Br, 1, etc. or groups, selected independently from carboxamide, carboxylic ester, Page 19 phosphonate ester, phosphate ester, double or triple bonds, carbamate, urea, thiourea, thiocarbamate, oxime, cyano, carboxyl, carbonyl, etc. Each R4 substitution on Ar may be the same or different and is selected independently 5 from hydrogen, substituted or non-substituted linear, branched or cyclical alkyl or heteroalkyl, aryl, substituted aryl, substituted or non-substituted heteroaryl, substituted or non-substituted linear, branched, or cyclical alkoxy, substituted or non substituted linear, branched, or cyclical heteroalkyloxy, aryloxy, beteroaryloxy, cyano, halogen, etc. 10 R4 is selected preferably from small substituents such as hydrogen, methyl, ethyl, ethoxy, methoxy, and other Cl to C6 linear, cyclical or branched alkyls and heteroalkyls. 15 n is zero or a positive integer. R7 and R8 are selected independently from hydrogen, substituted or non-substituted linear, branched or cyclical alkyl or heteroalkyl, aryls, substituted aryls, substituted or non-substituted heteroaryls, cyano, halogen, etc. 20 RS is selected from substituted or non-substituted linear, branched or cy-clical alkyl or heteroalkyl, aryls, substituted aryls, substituted or non-substituted beteroaryls, etc. R6 is selected from hydrogen, substituted or non-substituted linear, branched or 25 cyclical alkyl or beteroalkyl, aryls, substituted aryls, substituted or non- substituted heteroaryls, etc. RI is a polymer. Non-limiting examples for suitable polymers are polyalkyloxy-based polymers like 30 poly(pTopylene glycol) or poly(ethylene glycol), dextran, chitosan, hyaLuronic acid and derivatives, alginate, xylan, mannan, carrageenan, agarose, cellulos. e, starch, hydroxyethyl starch (HES) and other carbohydrate-based polmers, poly-(vinyl alcohols), poly(oxazolines), poly(anhydrides), poly(ortho esters), poly(carbonates), poly(urethanes), poly(acrylic acids), poly(acrylamides) such as Page 201 poly(hydroxypropylmethacr3'lamide) (HMPA), poly(acrylates), poly(methaciylates) like poly(hydroxyethylmethacrylate), poly(organophosphazenes), poly(siloxanes), poly(vinylpyrrolidonle), poly(cyanoacrylates), poly(esters) such as poly(lactic acid) or poly(glycolic acids), poly(iminocarbonates), poly(amino acids) such as poly(glutamic 5 acid), collagen, gelatin, copolymers, grafted copolymers, cross-linked polymers, hydrogels, and block copolymers from the above listed polymers. Hydrogels may be defined as three-dimensional, hydrophilic or amphiphilic polymeric networks imbibing large quantities of water. The networks are composed of 10 homopolymers or copolymers, are insoluble due to the presence of covalent chemical or physical (ionic, lydrophobic interactions, entanglements) crosslinks. The crosslinks provide the network structure and physical integrity. Hydrogels exhibit a thermodynamic compatibility with water which allows them to swell in aqueous media.(see.: N.A. Peppas, P. Bures, W. Leobandung, H. Ichikawa, Hydrogels in is pharmaceutical formulations, Eur. J. Pharm. Biopharn. 2000, 50, 27-46). The chains of the network are connected in such a fashion that pores exist and that a substantial fraction of these pores are of dimensions of between I and 1000 nra. By selecting certain polymerization conditions, the hydrogel may be obtained in the form of an amorphous gel or as beaded resin. Such soft beads may have a diameter of between 1 20 and 1000 micrometer. Hydrogels may be synthesized from the polymers and copolymers listed above and physically cross-linked or chemically cross-linked by radical, anionic or cationic polymerization, by chemical reactions like condensation or addition reactions as 25 described in W.E. Hennink and C.F. van Nostrum, Adv. Drug Del. Rev. 2002, 54, 13 36. Further examples include branched and hyperbranched polymers. Examples for such polymers include dendrimers and other dense star polymers. (R. Esfand, D.A, 30 Tomalia, Drug Discov Today, 2001, 6(8), 427-436; P.M. Heegaard, U. Boas, Chem. Soc. Rev. 2004 (33(1), 43-63; S.M. Grayson, J.M. Frechet, Chem. Rev. 2001, 101 (12), 3819-3868). Page 21, R1 can also be a biopolymer like a protein. Non-limiting examples of such polymers include albumin, antibodies, fibrin, casein, and other plasma proteins. Each R1 polymer can carry one or more biologically active substances linked to the 5 polymer by conjugation with a second prodrug linker as described herein or any other linker known to the person sidlled in the arL The polymers may have further substituents and may be functionalized for attachment to the spacer moiety X. Non limiting examples of such functional groups comprise carboxylic acid and activated derivatives, amino, maleimide, thiol, sulfonic acid and derivatives, carbonate and i o derivatives, carbamate and derivatives, hydroxyl , aldehyde, ketone, bydrazine, isocyanate, isothiocyanate, phosphoric acid and derivatives, phosphonic acid and derivatives, haloacetyl, alkyl halides, acryloyl, arylating agents like aryl fluorides, hydroxylamine, disulfides like pyridyl disulfide, vinyl sulfone, vinyl ketone, diazoalkanes, diazoacetyl compounds, epoxide, oxirane, and aziridine. 15 Preferred functional groups for the R1 polymer include but are not limited to thiol, maleimide, amino, carboxylic acid and derivatives, carbonate and derivatives, carbamate and derivatives, aldehyde, and haloacetyl. 20 Especially preferred functional groups include thiol, maleimide, amino, carboxylic acid and derivatives, carbamate and derivatives, and carbonate and derivatives thereof Non-limiting examples for suitable bonds or groups formed between X and RI 25 include disulfide. S-succinimido, amide, amino, carboxylic ester, sulfonamide, carbamate. carbonate, ether, oxime, hydrazone, urea, thiourea , phosphate, phosphonate, etc. Preferred bonds or groups formed between X and RI comprise S-succinimido, amid, 30 carbamate, and urea. Preferably, the R1 polymers are well hydrated, degradable or excretable, nontoxic and non-immunogenic in mammals. Preferred Ri polymers include polyalkoxy-based polymers like polyethylene glycol and polyethylene glycol reagents as those Page 22 described in Nektar Inc. 2003 catalog "Nektar Molecule Engineering - Polyethylene Glycol and Derivatives for Advanced PEGylation" and branched, byperbranched, cross-linked polymers and hydrogels, and proteins like albumin. s W is selected from substituted or non-substituted linear, branched or cyclical alkyl, aryls, substituted aryls, substituted or non-substituted linear, branched or cyclical heteroalkyl, substituted or nonsubstituted heteroaryls, etc. W is selected preferably from non-toxic substituted or non-substituted linear, 10 branched or cyclical alkyls or heteroalkyls. Y7 0 R7 0 Preferred variations of 0are O ,R6 4 2 and 0 R8 R8 15 NR6 O1 R7 0 Especially preferred variations of y 4 y 2 are R8 0 and NR6 0 20 forming polymeric prodrugs of the following-formulas: [R4 L RI I R4 i R1 0Y3 0 Q " Y R3 R3 R6 Ar R7 R8 Ar Ira lab R4\ X.R1 R4 X-R1 30 OR2 Y T 0-

Y

1

-

0 Q2 Y Nu-W-NA R3 Nu-W R3 \ 5 Ar 4 R8 RS R7 Iba Ibb Page 23 In formula Iaa and Iba R6 may also be Nu-W. At least one Nu is present in Nu-W. 5 Nu is a nucleophile that can perform a nucleophilic attack at the carbonyl carbon of 11 and thus catalyse the cleavage of the masking group by intramblecular Y4 WY2 catalysis or cyclization (figure 8). Figure 8 shows an example according to formula la or Ib wherein the cleavage of the masking group is by intramolecular cyclization. In cases where Nu only catalyses the cleavage of the masking group by intramolecular 10 catalysis, no cyclical product of the masking group is formed. Preferred nucleophiles include primary, secondary and tertiary amino groups, thiol, carboxylic acid, hydroxylamine, hydrazine, and nitrogen containing heteroaryl. Especially preferred nucleophiles include primary, secondary and tertiary amino groups. In order to effectively catalyse the cleavage of the masking group, the spacing 15 between the nucleophile Nu and Y 2 is preferably between three and fifteen atoms. More preferably, the spacing betetween Nu and Y 2 is between four and ten atoms. The at least one nucleophile Nu may be attached anywhere to W (e.g. at the terminus or in the middle ofW) or may be part of W. 20 Preferred variations for the masking group Nu-W-Y4 are selected independently from 0 R9\ R80 SN-W-N-J or N-W 25 R10 R6 RI0 R7 R9\ where /N-W forms a primary, secondary or tertiary amine nucleophile Nu. R10 These preferred variations result in polymeric prodrugs according to the following 30 formulas: Page 24- [R4 IX R1 [R4 IX RI 0%0Y R, 0 oY r R9 0 T' R 3T 5 N-W-N R3 /N-W R3 RI R6 Ar R10 R7R8 Ar [R4 X. .R1 [R4 XR1 1' R2 Y 0 R2 Y 10 R9 0C + Y 'rT R9\ 0 Y 3 T N-W-N R3 R3 RIO R6 Ar RIO R7 R8 Ar 15 R9, R10 are selected independently from hydrogen, substituted or non-substituted alkyl or heteroalkyl, substituted or non-substituted aryl or heteroaryl. Especially preferred variations for the masking group Y1 Nu-W-Y 4 20 are selected from R9 N 11 N R9, R8 / I or or N W RIO R12_ m R 6 m = 2 - 10 R10 R7 Preferably, R9, R10, RI I and R12 are selected independently from hydrogen, substituted or non-substituted alkyl and R7 and/or R8 are not hydrogen. 30 R6 may also be R9 R11 N RIO R12 m and is preferably not hydrogen. Pagf 25.

Surprisingly it was found, that the masking group can modify irreversibly the amine containing biologically active moiety when the nucleophile Nu is absent in the masking group. As shown in the example section, during release of the bioactive moiety insulin from a polymeric prodrug with a pentanoyl masking group Which is not 5 part of the present invention (as it does not contain a nucleophile), approximately 30% of the insulin molecule was modified with the masking group by acyl transfer. The mechanism of an example of this modification where D contains an additional free amino group that serve as nucleophile for acyl transfer from the masking group is shown in figure 9. 10 Ar of formula Ia or Ib is a multi-substituted aromatic hydrocarbon or a multi substituted aromatic heterocycle. To be aromatic, the number of pi electrons must satisfy the Hackel rule (4n+2) and the cycle has to be planar. A huge variety of compounds satisfy these criteria and thus are suitable as Ar in formula Ia or Ib. Non is limiting preferred aromatic moieties include: W N N N 20 N ~ N VV -~ N W N NN 25N N N 30 Page 26 wherein W is 0, N, or S, independent from each other. X Y_ R2 Y5 Y3 T Y-3 T 5 Y 2 and R3 in formula Ia or Y2 and R3 in formula Ib have to be arranged on the aromatic ring in such a fashion that a 1,4- or 1.6- or 1,( 4

+

2 p), with p= 2, 3, 4 and higher, elimination can take place (see above). For example, in the case of a 6-membered ring, the substituents have to be arranged ortho or para. I0 Preferred rnoieties for Ar are mono- and dicyclic aromatic hydrocarbons or aromatic heterocycles. Especially preferred moieties are monocyclic five- or six-membered aromatic is hydrocarbons or aromatic heterocycles. General synthesis procedures of the polymeric prodrugs Synthesis of representative examples of polymeric prodrugs according to the present 20 invention is described in the Examples section. Prodrugs of the present invention can be prepared in various different fashions. Figure 10 shows general routes for the synthesis of the polymeric prodrugs of the present invention according to formula Ia. 25 In a first method, intermediate (III) is provided by acylating Y 2 of starting material (II) with the masking group Y For this, X or Nu may have to be protected with a reversible protecting group PGi. Suitable protecting groups are described in TW Greene, P.G.M. Wuts, Protective 30 groups in organic synthesis, 1999, John Wiley & Sons, 3' ed.. From intermediate (II1) two alternative routes can be used to yield (Iaa). In a first route intermediate (Tl) is activated by an activating agent such as 4-nitrophenyl chloroformate or disuccinyl carbonate to yield (IV). The amine containing drug molecule is attached to (IV) to yield (V) by displacing the leaving group of the activated intermediate (IV). After Page 27. ' deprotection of X such as by treating intermediate (V) with reagents like trifluoroacetic acid or DTT (where applicable) deprotected intermediate (V) is then reacted with polymer R1 to yield the polymeric prodrug (Iaa). 5 In a second route the polymer RI is first attached to the intermediate (DI) after deprotection of X (where applicable) to form intermediate (VI). After an activation step intermediate (VII) is formed. (VI) is reacted with the amine containing drug molecule to form the polymeric prodrug (Iac). 1o In a second method, intermediate (VIII) is provided by activating starting material (II) by an activating agent such as 4-nitrophenyl chlorofomate. For this, Y 2 and/or X may have to be protected with a protecting group PG 2 and/or PG 1 .Amine containing drug is reacted with intermediate (VIII) to form (IX. In a first route, Y 2 of (IX) is selectively deprotected and acylated to form intermediate (V) which is further is processed to (Iaa) as described above. In a second route X is selectively deprotected and reacted with polymer RI to form intermediate (X). Y 2 of (X) is then deprotected and acylated to form the polymeric prodrug (lac). In a third method starting material (II) is reacted with polymer R1 to form 20 intermediate (XI). In one route, intermediate (XI) can be acylated to form intermediate (TI) which processed as described above to form polymeric prodrug (Iaa). In a second route, Y 2 is protected by the protecting group P0 2 , activated and reacted with the amine containing drug molecule to form (X). Intermediate (X) is then processed as described above to form the polymeric prodrug (Iac). 25 For all methods described, further functional groups such as Y 3 or nucleophiles present in Nu-W may have to be protected with suitable protecting groups. Polymeric prodrugs according to formula Ib can be prepared by methods described 30 above for prodrugs according to formula Ia using starting material Irb instead of II in figure 10. R4 X R2 2 R3 l1b Page2& It is understood, that linker structures according to the outlined invention and carrying protecting groups or leaving groups as described and used in the synthesis of corresponding polymeric prodrugs are considered within the range of the invention. 5 Application of the polymeric prodrugs in molecular therapy For polymeric cascade prodrugs it is desirable for the cleavage kinetics of the first temporary linkage to proceed under conditions present in the blood circulation of the 10 human body (pH 7.4, 370C). Most importantly, cleavage of the first temporary linkage should be based on hydrolysis and exhibit none or only very limited dependence upon chemical or biochemical or physicochemical entitities present in the human blood circulation such as enzymes, salts or binding proteins. 15 It was now surprisingly found that the rate of cleavage of the first temporary linkage connecting the masking group with the activating group as well as its dependence upon blood components can be controlled by neighbouring group effects mediated by nucleophilic functional groups (such as primary, secondary or tertiary amines) present in the masking group and positioned in a distance to the corresponding temporary 20 bond. If the masking group is structured in this fashion, an intramolecular reaction with contribution of the nucleophile governs the kinetics of the linkage (Fig. 5 and Fig. 8). Key advantage of the polymeric prodrugs of the present invention is their 25 predominantly non-enzymatic cleavage: the half-life of the prodrug in suitably buffered human blood plasma of pH 7.4 (with aqueous buffer concentration < 50 %) is at least 50 % of the half-life of the prodrug in enzyme-free buffer pH 7.4. This feature allows for better predictability and control of release rates after 30 administration to a living organism and reduces interpatient variability. In contrast to the enzymatic dependency of masking group removal as described in abovementioned examples of Antzczak et al., Shabat et al. and Lee et al., a higher Page 29 level of control over release rates can be achieved if the masking group has enzyme independent, self-eliminating properties. The masking groups according to the present invention contain at least one s nucleophile Nu. Structural features of this masking group such as nucleophilicity of the amine group and ring-forming capacity may be systematically optimized in order to precisely adjust the rate of prodrug cleavage. Such intramolecular reactions resulting in unmasking and subsequent rearrangement are highly independent from enzymes due to the fact that intramolecular reactions are generally preferred over 1o intermolecular reactions as shown diagranmatically in Fig. 8. In another embodiment of the invention, independency of prodrug cleavage from enzyme levels is achieved by providing a prodrug containing a sterically demanding carrier group as is shown in Fig. 7. 15 Such encapsulation or sterical protection by the sterically demanding carrier group may be conferred by a branched, hyperbranched, crosslinked or self-assembled structure of the carrier polymer. Such polymers tend to form a densely packed molecular volume, as exemplified for instance in dendrimers, dense star polymers or 20 bead-shaped nano- and mnicroparticles or amorphous gels. If the linkage of the polymer carrier to the drug is located in the interior of the polymer carrier, the linked drug will be efficiently encapsulated and protected from enzymatic attack. In this case, sterical hindrance by the polymer prevents enzymes from accessing and cleaving the temporary linkages. 25 In yet another embodiment, enzyme-independent prodrag cleavage is achieved by combining an intramolecular self-eliminating masking group with an encapsulating hyperbranched or crosslinked or self-assembled carrier. 30 A further advantage of the present invention is the release of an unmodified biologically active moiety. In cases where the biologically active moiety contains further reactive functional groups like arnino groups of lysine residues in proteins, an unwanted side reaction between the masking group and the biologically active moiety can occur. The reactive functional -groups of the biologically active moiety may react Page 30 twith the masking group, forming a stable covalent bond and resulting in the release of a modified biologically active inoiety. This potential side reaction is shown schematically in Figure 9. The occurrence of such side reactions is shown in the examples section using polymeric prodrugs which are not part of the present invention 5 with simple masking groups like a pentanoyl residue without a nucleophile Nu present in the masking group and as described by Antczal et al. or Lee et al. The side reaction in this linker system is suppressed using polymeric prodrugs according to the present invention with intramolecularly activated masking groups that contain nucleophiles Nu (see example section). 10 Enzyme-independent release control enables depot formulations without the need for encapsulation. Until now, many biocompatible ma-terials like hydrogels with large pore sizes could not be used for depot formulations due to their lack of encapsulation properties. From such well-hydrated and mechanically soft biocompatible 15 materials.biologically active moiety would be released too fast for most therapeutic applications. In combination with the prodrug linkers described in this invention, the carrier material may be optimized for its biocompatibility properties as the release is solely governed by the linker cleavage kinetics and does not require chemical or enzymatic degradation of the polymer carrier itself. 20 Release rates are governed by a substantially non-enzymatic chemical reaction which is in turn dependent on the molecular structure of the linker. Systematic or random modifications of the chemical structure, for instance by changing substituents in one or more positions, for instance a masking group in a cascade prodrug, allows for the 25 generation of prodrung linkers with differing release rates. It is therefore possible to create a variety of prodrug linkers and select those fast or slow cleaving prodrug linkers according to the demands posed by a given medicinal or therapeutic application. 30 Another advantageous feature which is part of this invention is the attachment of the polymer carrier through a stable covalent bond to an activating moiety involved in a double or cascade prodrug release mechanism. As part of this invention, the activating moiety remains attached to the polymer carrier after drug release and therefore cannot diffuse into the environment. Permanent linkage of th polymer carrier to the activating Page 31 . .

group greatly reduces the any side-reactivity of the activiating moiety and the probability of unwanted toxic effects. In other polymeric cascade prodrugs known in the art, the activating moiety is released in addition to the drug. Due to the molecular rearrangement mechanisms used in cascade prodrugs, the activating moiety is released 5 in a highly reactive form and may cause direct damage to surrounding biomolecules, or potentially toxic derivatives of the activating moiety may be formed in vivo. Description of the Figures 10 Fig. 1 shows a carrier-linked prodrug. Fig. 2 shows an enzyme-dependent carrier-linked prodrug. Fig. 3 shows a cascade prodrug where the masking group is part of the carrier. Fig. 4 shows an enzyme-dependent cascade prodrug where the masking group is part of the carrier. 15 Fig. 5 shows a self-cleaving cascade prodrug where the masking group is separate from the carrier. Fig. 6 shows an enzyme-dependent cascade prodrug where the masking group is separate from the carrier. Fig. 7 shows a cascade prodrug where the carrier is sterica.lly protecting the masking 20 group. Fig. 8 shows cleavage of the masking group by intramolecular cyclisation. Fig. 9 shows a possible side reaction of polymeric prodrug activation. Fig.10 shows general synthesis methods. Fig. 11 shows mass spectra of prodrug released insulin molecules. Page 32 - Examples Materials Fmoc-amino acids, resins and PyBOP were purchased from Novabiochem and are 5 named according to the catalogue. Fmoc-Ado-OH was obtained from Neosystem. All additional chemicals were purchased from Sigma Aldrich. Recombinant human insulin was from ICN Biomedicals (USA). Maleimide-PEG5k was obtained from Nektar (USA). 5-(and-6)-carboxyfluorescein succinimidyl ester (rnixed isomers) was obtained from Molecular Probes. 10 Solid phase synthesis reaction medium Solid phase synthesis was performed on NovaSyn TG Sieber aide resin with a loading of 0.17 mmol/g or 2-hlorotrityl chloride resin with a loading of 1.4 mmol/g. Syringes equipped with polypropylene fits were used as reaction vessels. Is Standard coupling cycle for fmoc-protected amino acids For fmoc protecting-group removal, the resin was repeatedly threee times, 4 min each) agitated with 2/2/96 (v/v/v) piperidine/DBU/DMF and repeatedly (six times) washed with DMF. 20 Coupling of fmoc-protected amino acids to free amino groups on resin was achieved by agitating the resin with 3 equivalents (eg) of fmoc--amino acid, 3 eq PyBOP and 6 eq DIEA in relation to free amino groups in DMF for 60 min. Finally, the resin was repeatedly (five times) washed with DMF. 25 Standard cleavage protocol for TentaGel Sieber amide resin Upon completed synthesis, the resin was washed with DCM, dried in vacuo and treated repeatedly (five times) with 97/2/1 (v/v).DCMITES/TFA. After evaporation, compounds were purified by preparative RP-HPLC (Waters 600). 30 Standard cleavage protocol for 2-chlorotrityl chloride resin Upon completed synthesis, the resin was washed with DCM, dried in vacuo and treated two times for 30 minutes with 65/35 (v/v) HFIP/DCM. After combining the eluates the volatile components were evaporated. Page 33' Analysis Mass spectrometry (vS) was performed on a Waters ZQ 4000 ESI instrument and spectra were, if necessary, interpreted by Waters software MaxEnt, 5 Size exclusion chroma-tography was performed using an Amersham Bioscience AEKTAbasic system equipped with a Superdex 200 column (Amersham Bioscience). NMR spectra were recorded on a Bruker AC300. Overview - synthesis of Dolymeric prodrues according to formula la with ester-linked 10 masking groups N2 0 N2 PyBopIDIEAJDMF RI HOH H R1 OH + o S'-Mmt N HOS-Mmnt HO 2 OH N or /DIC0 R2 OH 45 ci 20 0 RN S-Mt O0/\ OH NO, RI R2 O O NS-Mmt O)= H 7 R2 OH 25 02N 1. NH,-R, 2. Mmi deprotection 30 0 R1 Polymer atta:hmen R N S R SH N___ ___ OB NH NH R3 R3' Pase 34, .Synthesis of DDIVrneriC DdrSacodntofruaIwthabaat-iked mnaskIn~g roups a) H R UB2 HO R R2 .

R2 1 Rl H NBRi N R7 HOHO O7 OH 0 R213 R2 12 0 -- m Ho S-RA 0 0 Rl Ri R2 0 17 R2 OH 14 0 R6- N, 2. MS6A 15 Z. R5 Ril 0 0 N'S - rl NaBH RI SR7 0___ R-R5 R2 0 8R6-N\ -O H R7 16 RS R 2 j 19 20 Rl R -7 R4 R6-N 0 R5 R2 C)= NH RS 21. 45 ftge 35.

b) D- 0 Rl Rl OH ___W_ 0O 23 R2 22 0 24 0 HN02 Ri 0 R6-4 5 RIH 0 -mmt 2HR0Sff R6-N \5 R2 - R2 R5 R2 26 NaBH 4 O Ri N OH-0 -- Mr - 28 - 29 RB-N RR2 0 R - \ R R 22 7 R 3 3 0 Page- 361.

C-) RI 0r RI Br HSH Tr+ HOi)- Ar-1, H0b 54~~ I DBU Tni s i.~ C)~ l4NO5 RI Ir \Y 56 R24i\ R2-N 0 2. R3 RI R3 NB 0 / t IDSC 10 RI 5L' R2 - 0 Tn\,/ t R3 61 O \__ 'jN R2-N \ O3 INPys,-CI 0 R3R1 / '-- NO 2 RiPY S -O / 15 R1 R2-N 00 O\ / \ r R3 62 0= R2-N - 0 R30 D NO 2 1. HN-R4 1. H,-R2.. NPys deprotacion I.1tWR 3. Polymer attadhmelrt 20 Ri 2. Trt deprolectionSHR / \Polymer attachmernt 0 /PS R2-N 0

R

R3 \$0R 3 163 0 N-R4 H H 25 .30 Page 371 Synthesis of polbmeric prodrugs according to formula Ib with a carbamate-linked masking group a) H 35a i ~0 O H DICHOBT H H / H H H N O - O N 'N 36a 1. BH3 THF or LAIH. 0 OH 2. AcOH, DIC, HOBt 3. HFIP HO

H

2 N O O N~ 37a OH 2, PyBOP, DIE.A HO H OY Mmt-S N N 0 38 OH I M C D 2 H O 2 15b 0 Mmt--S O O) N 0 40a O N O <O \/ NO 2 0 H Mmt-S N O O 00 41 a K 5OO 50 b) Page 38, HN H 35b DIC HOBT H H 10 0 OH 1. BHa THF 36b 2. AcOH, DIC, HOBt t OH N C20 E H 37b OH 2. N,N,N-trimethyethyoendiamine 3. HFIP OH 30 0< H2NN 39 2, PyBOP, DIEA 0 OH 40 Mmnt-S N H 0 N 40b 41b O 50 0 00 42 + R3 R4, S, N N 60 H 43' N 65 Page 39, Synthesis of compound 2 0 HO S-Mmt 5 2 Mint-chloride (1 eq) and mercaptopropionic acid (1.1 eq) were dissolved in TFA and incubated for 30 min. The solvent was removed under reduced pressure. The product was dissolved in pyridine, diluted in water, acidified by acetic acid and extracted with 10 ether. The ether phase was separated and dried over Na-SO4. Solvent was removed under reduced pressure and product 2 was purified by RP-HPLC. Synthesis of compounds 3a and 3b 15 R1 N S-Mmt HO H OH 3a R1= H 20 3b RI = OMe Octopamine hydrochloride (la) (2 eq), DIEA (4 eq), and PyBOP (1 eq) were dissolved in DMF, 2 (leq) was added and the mixture was reacted for 50 min at room temperature. After addition of acetic acid (7 eq) product 3a was purified by RP HPLC. 25 3b was synthesized from normetanephrine hydrochloride (1b) as-described above. 1a: MS [M+Na]*= 536 (MW+Na calculated = 536.2 g/mol) 3b: MS [M+Na] = 566 (MW+Na calculated = 566.2 g/mol) 30 Pane 40.

Synthesis of mercaptothiazolide 4 N 5 4 2-Mercaptothiazoline and triethylamine (1.5 eq) were dissolved in dry THF and pentanoyl chloride (1 eq) was added. The mixture was stirred for 1 h at 50 *C under 10 inert gas atmosphere and was allowed to cool to room temperature. 0.5 N aqueous HC was added and the separated organic phases were dried over Na 2 SO4. After concentration in vacuo the residue was purified by silica gel column chromatography using heptane/ethyl acetate (1/1) as mobile phase. Mercaptothiazolide 4 was collected as a viscous yellow oil. is 4 Rr (beptane/ethyl acetate 1:1)= 0.7 Synthesis of compounds Sa and 5b 20 0 H 0 H Hj 5a Sb 25 General synthesis protocol: 1 g 2-ohlorotrityl chloride resin (loading 1.6 rnmol/g) was incubated for 1 h with 850 mg (2.4 minol) Fmoc-Ile-OH and 840 ±l (4.8 mmol) DIEA in 5 ml 1/1 DCM/DMF. After fmoc removal and washing of the resin with DMF, boc-aminobutyric acid was 30 coupled to 0.5 g resin according to the standard coupling method. Compound 5a was cleaved from the resin with 97/1/2 (v/v) DCM/TA/TES for 45 min. After neutralisation with pyridine, solvents were removed under reduced pressure and 5a was purified by RP-HPLC. Sb was synthesized from boc-aminohexanoic acid as described above. Page 41, Sa MS M+Na3 = 339.2 (MW+Na calculated = 339.4 g/mol) 5b MS [M+Na.]*=367.4 (MW+Na calculated = 367.5 g/mol) 5 Svnthesis of compound 6a 0 0 N - S-Mint OH 10 6a 10 Mercaptothiazolide 4 (1 eq), phenol 3a (4 eq) and DMAP (4 eg) were refluxed in DCM for 2 h under nitrogen atmosphere. After neutralization with acetic acid, the solvent was removed in vacuo and product 6a was purified by RP-HPLC. 15 62 MS [M+Na]*= 620 (MW+Na calculated = 620.3 g/mol) Page 42, Synthesis of comoounds 6b to 6e 0 0 H 0 H S-Mmt /- NO NH 6b 0 0 0 0ON S-Mmt 10 O0H HN Sc 0 0 R SNO S-Mt is N 0 N S-in H N 0 H H - OH 6d RI= OMe General synthesis protocol: Se RI = H Carboxylic acid 5a (leq), phenol 3b (1 eq), DIC (1 eq), and DMAP (2 eq) in DMF 20 were reacted for 1h at room temperature. After addition of acetic acid (4 eq) tbe resulting carboxylic ester 6c was purified by RP-HPLC. 6d was synthesized as described above using 5b and 3b as starting materials. 6b was synthesized as described above using Z-Lys(Boc)-OH and 3b. 6e was synthesized as described above using 5b and 3a. 25 6b MS [M+Na]+= 928 (MW+Na calculated = 928.6 g/mol) 6c MS [M+Na]*= 864 (MW+Na calculated = 864.5 g/mol) 6d MS [M+Na] 892 (MW+Na calculated = 892.6 g/mol) 30 6e MS [M+Na1)*= 862 (MW+Na calculated = 862.6 g/mol) Page 43.

Synthesis of comDounds 7a to 7e 0 / 0 0 N-- S-Mrm )-- s-um 0 0 5 7a 2= NH 7b 0 No2 7- 4,o O' 0/ 0 0 R 1 HN 0 0 15 7d Ri = Me 7e RI=H General synthesis protocol: Alcohol 6a (1 eq), 4-nitrophenyl chloroformate (10 eq), and DIBA (10 eq) were 20 stirred in dry dioxane for 3 h at room temperature under nitrogen atmosphere. After addition of acetic acid (25 eq) the mixtures were diluted with 7/3 (v/v) acetortrile/H20 and the carbonate 7a was purified by RP-HPLC. 7b, 7c, 7d, or 7e were synthesized from 6b, 6c, 6d, or 6e, respectively, as described above. 25 7a MS [M+Na] = 785 (MW+Na calculated = 785.5 g/mol) 7b MS [M+Nar = 1093 (MW+Na calculated = 1093.7 g/mol) 7e MS [M+Na]* = 1029 (MW+Na calculated = 1029.6 g/mol) 7d MS [M+Na]* = 1057 (MW+Na calculated 1057.6 g/mol) 30 7e MS [M+Na]*= 1027 (MW+Na calculated = 1027.6 g/mol) Page 44.

Synthesis of compounds 8a to 8c (N"'linker-insulin) 0

.NH

2 isu n 5 NKC 5e29NK iB1 NH 2 rs ,NH aA1 Ba R4 = H9 H 9a R4 = Suc-PEC-5k UBI 2 1 0 N 0 0 HNN 0 B H N H RH i NH NH aA1 8b R4 = H Insulin 9b R4a uc-EGk c R4 = Su-PEGSk \H 15 General synthesis protocol: Rh-Insulin in 1/1 (v/v) DMSO/DMF was mixed with a solution of 0.9 eq carbonate 7a in DMSO. The resulting solution was adjusted to basic pH with DIEA and stirred for 1.5 h at RT. RP-PLC purification gave Mint-protected intermediate. 20 After lyophilization, the Mimt-protected intermediate was mixed with 95:5 (v/v) TFA/triethylsilane and stirred for 5 min. Volatiles were removed under nitrogen flow and 8a was purified by RP-HPLC and lyophilized. Regioselectivity of insulin modification was verified by DTT reduction and MS analysis. 8b or Sc were synthesized from 7c, or 7d, respectively, as described above. 25 8a MS [M+2H] 2 + 3078.9; [M+3H) = 2053.2 [M-4H] 4 *= 1540.6 (MW calculated = 6158 g/mol) 8b MS [M+2H] 2 = 3152.9; [M+3H] = 2100.6 [M+4H]1 4 += 1575.8 (MW calculated = 6302 g/mol) 30 Sc MS: [M+3H]Ir 2110.7; [M+4HJ* = 1583.7; [M+5H] = 1266.6 (MW calculated= 6330 g/mol) Page 45.

Synthesis of compounds 8d to 8% (NliFluorescein-"Llnker-insul-i)

H

2 N 0 H c2 / 0 FK-N 0,0 Insulin N 5 b-H0 Hr r.629 H Nz fluorescein insulin F \ NH 0 a 8d R4 = H 9d R4 = Suc-PEGSk H2 QB a 10 " N0 0 N 0 0 HN o R I Innsulin 15 - 9Se R4S P = H 15 9e 4 = su-PEG~kSf R1I=oMe, R4 =H 8g R=H,R4'H 9f R1 = OMe, R4 = Suc-PEG5k suC succinimidyl 9g R1 NH R4 =Suc-PEG5k 20 Synthesis of N' 9 -fluorescein insulin: 80 mg (13.8 ±mol) rb-insulin were dissolved in 4 ml 1/1 (v/v) DlvliF/DMSO and 40 p1 DIEA were added. 8 mg (17 pmol) 5-(and-6)-carboxyfluorescein succinimidyl ester 25 were added and the solution was stirred for 30 min at room temperature. 4 ml S/5/1 (v/v/v) acetonitrile/water/acetic acid were added, product NB2-fluorescein insulin was purified by RP-HPLC and lyophilized. The conjugation site was verified by reduction of N8 29 -f1uoreScein insulin with 1,4-dithiotbreitol, protease digestion and MS analysis. 30 MS: [M+2H1] 2 3084.0; [M+3H]" = 2054.6 (MW calculated = 6166 g/mol) Page 46 - N".-fluorescein insulin in 1/1 (v/v) DMIF/MSO was mixed with a solution of 0.9 eq carbonate 7b in DMSO. The resulting solution was adjusted to basic pH with DIEA and stirred for 3 h at RT. RP-HPLC purification gave Mmt-protected intermediate. 5 After lyophilization, the intermediate was dissolved in 95/5 (v/v) TFA/triethylsilane and stirred for 5 min. Volatiles were removed under nitrogen flow and 8d was purified by RP-HPLC and lyophilized. Se, 8f, or Sg were syntesized as described above using 7c, 7d, or 7e, respectively. io d MS: [M+2H] = 3364.1; [M+3H.H] =2242.7; [M+4H]* = 1681.5 (MW calculated = 6724 g/mol) Se MS: [M+3H)"= 2219.2 [M+4H]4*= 1665.9; [M+5H]"= 1332.8 (MW calculated = 6660 g/mol) 8f MS: [M+3H] - 2229.7 [M+4H] 4 = 1673.3; [M+5H]'= 1337.7 (MW calculated is =6689 g/mol) 8g MS: [M+3H]'" = 2218.7 [M+4H] 4 *= 1664.9 (MW calculated = 6659 g/mol) Synthesis of compounds 9a to 92 (mono-peazvlated insulin compounds) 20 70 sI 500 RM 8a in 1/4 (v/v) acetonitrile/water were mixed with 7 sl 10 mM maleimide-PEG5k in 1/4 (v/v) acetonitrile/water and 10 p1 0.5 M sodium phosphate buffer pH 7.0 and incubated for 15 min. Compound 9a was purified by SEC (column: Superdex 200, flow rate: 0.75 ml/min) using 10 mM HEPES buffer (pH 7.4), 150 mM 25 NaCl, 3 mM EDTA, and 0.005% Tween as mobile phase. The collected eluate (approximately 1.5 ml) was directly used as such for release rate determination. 9b, 9c, 9d, 9e, 9f, or 9g were synthesized as described above from 8b, 8c, 8d, Se, 8f, or 8g, respectively. 30 9a through 9g: SEC retention time: 19.5 min Page 47 Synthesis of compounds 11a and _lb RI Br HO / \O R2 1la R1 = R2= Me 1lb Ri = Me, R2= H 1o 3,5-Dimethy1-4-hydroxy-acetophenone (5.0 mmol) (10a) and CuBr2 (1.7 g, 7.5 mmol) were dissolved in 10 rni ethyl acetate and refluxed for 2 h. Solid byproducts were removed by filtration- The filtrate was evaporated and crude product 11a was purified by RP-HPLC. 1lb was synthesized from 4-hydroxy-3-methyl-acetophenone (10b) (0.75 g, 5.0 15 mmol) as described above. 11a: Yield 754 mg (62%) MS [M+H] = 243.1/245.1 (MW+H calculated =244.1 g/mol) 11b: Yield 533 mg (47%) 20 MS [M+H]* = 229.2/231.1 (MW+ H calculated = 230.1 g/mol) Synthesis of compounds 1Za and 12b R1 25 HO/\ N R2 12a RI = R2= Me 12b RI = Me, R2= H 30 500 mg 11a (2.06 mmol) and 576 mg (4.11 mmol) hexametbylenetetramine were dissolved in 20 ml of trichloromethane and refluxed for 30 min. The solvent was removed in vacuo. 4 ml ethanol and 2 nil of concentrated HCl were added and the slurry was heated to 50 *C for 4 h. The mixture was concentrated in vacuo, diluted with acetonitrile/water and 12a was purified by RP-HPLC. Page 48 12b was synthesized from 472 nag (2.06 mmol) 11b as described above. 12a: Yield 547 mg (81%) as TFA-salt MS [M+Na] = 202.2 (MW+Na calculated = 202.2 g/mol) 5 12b: Yield 455 mg (70%) as TFA-salt MS [M+Naj= 188.2 (MW+Na calculated = 188,2 g/mol) Synthesis of compound 13 10 HO /N\ 13 500 mg (1.71 mmol) 12a (TFA. salt) were dissolved in 10 ml 1/1 (v/v) 15 methanol/water, 129 mg (3.41 amnol) NaBH 4 were added and the mixture was stirred for 30 min at RT. 0.5 ml of acetic acid were added and 13 was purified by RP-HPLC. 13: Yield 313 mg (62%) as TFA--salt MS [M+Na]* = 204.2 (MW+Na calculated = 204.2 g/mol) 20 NMR (300 MHz, DMSO-d 6 ) S[ppm] = 8.25 (s, iH, Phenol), 7.84 (bs, 3H, NH3), 6.89 (s, 2H, CH.), 5.85 (d, 1H, Hydroxyl, J=3.7 Hz), 4.62 (m, 1H, CHc.,yj), 2.93 (m, 1H, CHa), 2.80 (m, 1H, CHb). 2. 17 (s, 6H, CH3). Synthesis of compounds 14a to 14d 250 RI S-R4 HO R7 - OH R2 14a RI = R2= Me, R7 = H, R4 = Mmnt 30 14b R1 = R2= H, R7 = Me, R4 = Mmt 14c R = OMe, R2= H, R7 = Me, R4 = Mnt 14dR =H, R2= H, R7 = Me, R4 = Trt Page 49.'- 13 (TFA salt, 159 mg, 0.541 mmol) was coupled to compound 2 as described for compound 3a to yield 14a. 14b, or 14c were synthesized as described above using synephrine (335 mg, 2.00 mmol) or metanepbrine (HCI salt, 281 mg, 1.20 mmol), respectively. 5 Synephrine (335 mg, 2.3 mmol) was coupled to 3-tritylsulfanyl-propionic acid as described above to yield 14d. 14a: Yield 254 mg (87%) MS [M+Na]*= 564.7 (MW+Na calculated = 564.3 g/mol) 10 14b: Yield 760 mg (72%) MS [M+Na]* = 550.2 (MW+Na calculated = 550.3 g/mol) 14c: Yield 530 mg (80%) MS [M+Na]* = 580.4 (MW+Na calculated = 580.4 g/mol) 14d: Yield 567 mg (49%) 15 MS [M+Na)*= 520.5 (MW+Na calculated = 520.7 g/mol) Synthesis of compounds 15c. 15d and 15f H H1 1Sc 15d 15f General synthesis protocol: 25 1 g 2-chlorotrityl chloride resin (loading 1.4 mmol/g) was incubated for 1 h with N,N'-dimethylpropane- 1,3-diamine (for synthesis of 15c), or NN'-diethyl-propane 1,3-diamine (for synthesis of 15d) or NN'-dirnethyl-ethane- 1,2-diamine (4 eq) (for synthesis of 15f) in DCM. After washing of the resin with DMF, amines were acetylated with 1/1/2 (v/v/v) acetic anhydride/pyridine/DMF for 14 h. The resin was 30 washed with THF and dried. LiAlH 4 (I M in THF, 4 eq) was added dropwise to the suspended resin in THF. The resulting suspension was stirred for 3 h at 45*C under nitrogen atmosphere. After cooling, aqueous Rochelle's salt solution was added and the resin was separated and dried. Compounds were cleaved from resin with 2/1 (v/v) Page 5O ..

HFIP/DCM (2 x 30 min). The volatile components were evaporated and the products 15c, 15d or 15f were used in the following steps without further purification. 15c MS [M+H=J*= 131.2 (MW = 130.1 g/mol) 5 15d MS [M+H+ = 159.2 (MW = 158.1 g/mol) 15f MS [M+HJ*= 117.1 (MW = 116 g/mol) Synthesis of compounds 16a to 16f and 16i 0 R1 10 0 N S-R4 O R7 R6-N - OH R5 R2 16a RI = R2 = R5 = Me, R6 = 2-(dimethylanino)ethyl, R7 = H, R4 = Mmt 16b R1 = OMe, R2 = H, PS = Et, R6 = 2-(diethylamino)ethyl, R7 = H, R4 = Mnt is 16c RI = OMe, R2 = H, RS = Me, R6 = 3-(N-ethyl-N-methylamino)propyl, R7 = Me, R4=Mnt 16d R1 = R2= H, R5 = Me, R6 = 3-(N-ethyl-N-methylaiino)propyl, R7 = Me, R4 = Mint 16e RI = OMe, R2= H, RS = Et, R6 = 3-(diethylamino)propyl, R7= Me, R4 = Mmt 20 16f R1 = R2= H, R5 = Et, R6 = 3-(diethylamino)propy1, R7 = Me, R4 = Mmt 16i RI = R2= H, R5 = Et, R6 = 2-(diethylamino)ethyl , R7 = Me, R4 = Trt 14a (120 mg, 0.222 mnmol) was dissolved in 1.5 ml of dry THF. p-Nitrophenyl chloroformate (45 mg, 0.222 nmol) and DIEA (113 ,.1, 0.665 mmol) were added and 25 the mixture was stirred for 30 min at RT. ISa (N,NN'-trimethyl-ethylene-1,2 diamine) (72 p.1, 0.554 mmol) was added and stirring wa-s continued for 30 min. The solvent was removed in vacuo, 100 p1 of AcOH were added and 16a was purified by RP-HPLC. 30 16b was synthesized as described above from 3b (80 mg, 0.15 mmol) and 15b (N,N,N'-triethyl-ethylene- 1.2-diamine) (55 mg, 0.38 mtnol). 16c or 16d were synthesized as describe above from 14c (56 mg, 0.1 mmol) or 14b (53 mg, 0.1 mmol), respectively, and diamine 15c. Page 51 16e or 16f were synthesized as described above from 14c (56 mg, 0.1 mrnol) or 14b, respectively. (53 mg, 0.1 mmol) and diamine 15d. 16i was synthesized as described above from 14d (350 mg, 0.7 mmol) and 1Sb (N,N,N-t~riehyl-ethylene-1,2-diamine) (180 p1, 1 mmol). 16a: Yield 120 mg (69%) as TFA salt MS [M+Na = 692.4 (MW-1+Na calculated = 692.9 g/mol) 16b: Yield 48 ig (40%) as TFA salt MS [M+Na] = 736.3 (MW+Na calculated = 736.4 g/mol) )0 16c: Yield 8 mg (10%) as TFA salt MS [M+Na]* = 736.4 (MW+Na calculated = 736.4 g/mol) 16d: Yield 20 ig (25%) as TFA salt MS [M+Na]*= 706.3 (MW+Na calculated = 706.3 g/mol) 16e: Yield 2 mg (3%) as TFA salt is MS [M+Na]* = 764.6 (MW+Na calculated = 764.4 g/mol) 16f: Yield 6 mg (8%) as TFA salt MS [M+Na] =734.4 (MW+Na calculated = 734.3 g/mol) 16i: Yield 152 mg (28%) as TFA salt MS [M+Na]* = 690.5 (MW+Na calculated = 690.9 g/mol) 20 Synthesis of compound 17 0 N S-Mmt 25 HD H - b40 17 Amine 12b (TFA salt) was coupled to compound 2 as described for compound 3a. 30 17: Yield 608 mg (74%) MS [M+Na]* = 548.3 (MW+Na calculated = 548.7 g/mol) Page 52 Synthesis of compounds 18a and 18b O RI N S-Mmt YO H R6-N 0 5 R5 R 18a R1 = R5 = lde, R2 H, R6 = 3-(dimethylamino)propyl 18b R1 = = Me, R2 H, R6 = 2-(N-ethyl-N-methylamino)ethyl 383 mg (0.729 miniol) of 17 were reacted with p-nitropbenyl-chloroformate and 10 N,N,N'-tr-imethyl-propane-1,3-diamine (15e) or 15f, respectively, to yield 18a or 18b as described for compound 16a. 18a: Yield 287 ng (50%) as TFA salt MS [M+Na] = 690.7 (MW+Na calculated= 690,9 g/mol) 15 18b: Yield 148 ng (26%) as TFA salt MS [M+Na* = 690.9 (MW+Na calculated = 690,9 g/mol) Synthesis of com.nounds 16g and 16h 0 20 RI 0 N S-Mmt )-o H R6-N - OH R5 R2 16g RI = R5 = Me, R2= H, R6 = 3-(dimethylamino)propyl 16h RI = RS = &ve, R2= H, R6 =2-(N-ethyl-N-methylamino)ethyl 25 18a (287 mg. 0.3 67 mmol, TFA salt) was dissolved in 5 ml methanol, NaBR 4 (41 mg, 1.07 mmol) was added and the mixture was stirred for 30 min at RT. 0.5 ml of acetic acid were added and 16g was purified by RP-HPLC. 18b (8 mg. 0.010 namol, TFA salt) was reacted as described above to yield 16h. 30 16g: Yield 201 rag (70%) as TFA-salt MS [M+Na]* = 692.7 (MW+Na calculated = 692.9 g/mol) 16h: Yield 6 mg (77%) as TFA-salt MS [M+Na]* - 692.7 (MW+Na calculated = 692.9 g/mol) Page 53:' Synthesis of compounds _9a to 391 0 R1 o N, 5 -O R7 R4 R6-N - 0 R5 R 2 Oo NO 0 - / NO 2 19a R1 = R2 =RS = Me, P6 2-(dimethylaminO)ethYl, R7 = H, R4 = Mnt 19b R= OMe, R2= H, R5= Et, R6 2-(diethylamino)ethyl, R7 = H, R4 = Mimt 10 19c R1 = OMe, R2 = H, R5 = Me, R6 = 3-(N-ethyl-N-methylamino)propyl, R7 = Me, R4 = Mmt 19d RI = R2 = H, R5 = Me, R6 = 3-(N-ethy1-N-methylamino)propyl, R7 = Me, R4 Mmt 19e R1 = OMe, R2= H, R5= Et, R6= 3-(diethylamino)propy1, R7 = Me, R4 = Mint 15 19f RI = R2= H, R5 = Et, R6 = 3-(diethylamino)propyl, R7 = Me, R4 = Mmt 19g R1 = RS = Me, R2 = H, R6 = 3-(dimethylamino)prpy, R7 = H, R4 Mnmt 19h RI = RS = Me, R2= H, R6 = 2-(N-ethyl-N-methylamino)ethyl, R7 = H, R4= Mmt 19i R1= R2 = H, R.S = Et, R6 = 2-(diethylamino)ethyl, R7 = Me, R4 = Trt 20 Carbonates 19a to 19i were synthesized from 16a to 16i, respectively, as described for compound 7a. 19: Yield 98 mg (72%) as TFA-salt 25 MS [M+Na] = 857.8 (MW+Na calculated = 858.0 g/mol) 19b: Yield 6 mg (11 %) as TFA-salt MS [M+Na]*= 901.8 (MW+Na calculated = 901.5 g/nol) 19c: Yield I mg (15/o) as TFA-salt MS {M+Na]= 901 .4 (MW+Na calculated = 901.5 g/mol) 30 19d: Yield 8 mg (2cP%) as TFA-salt MS [M+Na] = 871-4 (MW+Na calculated = 871.4 g/mol) 19e: Yield 0.3 mig (18%) as TFA-salt MS [M+Na] 4 = 929.4 (MW+Na calculated = 929.5 g/mol) 19f: Yield 4 mg (45 %) as TFA-salt Pae 54, - MS [M+Na] 4 * 899.7 (MW+Na calculated =899.6 g/mol) 19g: Yield 6 mg (6%) as TFA-salt MS [M+Na] =857.8 (MW+Na calculated = 858.0 g/mol) 19h: Yield 0.8 mg (11%) as TFA-salt 5 MS [M+Na] = 857.7 (MW+Na calculated. = 858.0 g/mol) 191: Yield 77 mg (49%) as TFA-salt MS [M+Naj = 856.2 (MW+Na calculated. = 856.0 g/mol) io Synthesis of comDounds 20a to 20f 0 R1 R5 I R5 R27 SH is EB29 NH,2 aA1 \ NH 1 Insulin 20a RI = R2= R5 Me, R6 = 2-(dimethylamino)ethyl, R7 = H 20 20b R1 = OMe, R2 = H., R5 = Et, R6 = 2-(dietylamino)ethyl, R7 = H 20c R1 = OMe, R2= H, R5 = Me, R6 = 3-(N-ethyl-N-methylamino)propyL R7 = Me 20d R1 = R2= H, R5 = Me, R6 = 3-(N-ethlyl-N-methylamino)propyl, R7 = Me 20e RI = 5= Me, R2 = H, R6 = 3-(dimethylamino)propyl, R7 = H 20f RI = S= Me, R2= H, R6= 2-(N-ethyl-N-nehylarnifo)6thyl. R7 H 25 Insulin derivatives 20a, 20b, 20c, 20d, 20a, or 20f were synthesized from 19a, 19b, 19c, 19d, 19g, or 19h respectively, as desc-ribed for compound 8a. ZOa MS [M+3H] = 2077.3 [M+4HJ 4 *= 1 5592 (MW calculated = 6231.3 g/mol) 30 20b-MS [M+3H)" = 2093.0 [M+4H] 4 *= 1569.6 (MW calculated = 6274 g/mol) 20c MS [M+3H]'= 2090.8 [M+4H] 4 *= 1 568.7 (MW calculated = 6274 g/mol) ZOd MS [M+3H] = 2081.3 [M+4H]4= 1561.8 (MW calculated =6244 g/mol) 20e MS [M+3H)"= 2077.1 [M+4H]'=- 1 558.2 (MW calculated = 6231.3 g/mol) ZOf MS [M+3H]= 2076.7 [M+4H] 4 *= 1 559.3 (MW calculated 6231.3 g/mol) Page :- Synthesis of cornoounds 21 a to 21f (mono-DeoVlate I insulin derivatives) o Ri 580 o~ 0 S PEG5K 'B2~ /NH cL 1 Insulin 7 \>NHz io 21a RI= R2 = R5 = Me, R6 = 2-(dimethylamnino)ethyl, R-7 H 7-1b Rl = OMe, R2 = HL R5 = Et, R6 = 2-(diethylamLino)et1y, R7 H Z1c R1I OMe, R2 HR.5 = Me, R6 = {-N-ehy-NI-methylaflifo)propyl, R7 =Me 21d RI = R2 = H, P-5 =Me, R6 = 3-(N-ethiyl-N-mnetyamiflo)propyl, R7 = Me 2le Rl = R5 = Me, R2 = I-, R6 = 3-(dimnethylamino)propyl, R7 = H 15 21f RI = R5= Me. R2 = H, R6 =2-(N-ethy1-N-methYlamlifl0)ethyI, R7 =H Insulin derivatives Zli, Zlb, 21c, Zid, 21e, or 21f we-re synthesized from compound ZNa, 20b, 20c, 2.Od, 20; or 20f. respectively, as described for compound 9a. 20 Zia through 21f: SEC retention time: 19.5 m-in Page 56..

Synthesis of compounds 23a and 23b 0 RI OH HO/ R2 23a R1 = Me, R2= H 23b RI = R2= Me o-Cresol (Z2a) (1 eq), succinic anhydride (1 eq), and AJCl 3 (3 eq) in nitrobenzene i o were heated to 100 0C for I h. The reaction mixture was poured on HClI/ice and extracted with ether. The organic layer was extracted with 1 N NaOH and the aqueous layer was acidified with concentrated HCL. The aqueous layer was extracted with ether and the ether was evaporated. 23a was purified by RP-HPLC. 23b was synthesized from 2,6-dimethylphenol (22b) as described above. 15 23a: Yield 552 mg (31%) MS [M+Na]* = 231.0 (MW+I-Na calculated = 231.2 g/mo 1) NMR (300 MHz, DMSO-d6) S(ppm]= 12.05 (bs, 1H, CO 2 H), 10.23 (s, 1 H, phenol OH), 7.74 (s, 1H, CH.), 7.7 (d, 1H, CH., 3 JH, = 8.4 Hz), 6.86 (d, 1H, CH.,, JH,H = 20 8.4 Hz), 3.13 (t, 2H1, C(O)CH 2 , 3 H.H =6.4 Hz), 2,53 (t, 21H, CH 2 CO2, 3 J3JH.H = 6.4 Hz), 2.16 (s, 3H, CH 3 ) 23b: Yield 166 mg (15%) MS [M+Na]* = 245.4 (MW-+Na calculated = 245.2 g/mo3) 25 Synthesis of compound 24 HN _'Mmt 1.85 g (16.02 mmol) cysteamine hydrochloride were dissolved in 15 ml of TFA and 2.47 g (8.01 mmol) MmtCl were added. After stirring the mixture at RT for 20 min 30 the solvent was evaporated in vacuo. The residue was dissolved in diethyl ether and extracted with saturated aqueous NaHC03, IN H 2 S0 4 arid brine. The solvent was evaporated and 24 was purified by RP-HPLC. Page 57, - : 24: Yield 1.11 g (30%) as TFA salt TLC (AcOEtIEtsN 99/1), R= 0.24 Synthesis of compounds 25a and 25b 0 H HRI N S-m R2 25a RI = Me, R2 H 10 25bRI=R2=Me 23a (1eg), HOBt (1.1 eq) and DIC (1 eq) were dissolved in DMIF and stirred at RT for 30 min. 24 (TFA salt, 1 eq) and DIEA (3 eq) were added and the solution was stirred for 60 min. Acetic acid was added and 25a was purified by R-HPLC. 15 25b was synthesized from 23b as described above. 25a: Yield 552 mg (25%) MS [M+Na]* = 562.7 (MW+Na calculated = 562.7 g/mol) 25b: Yield 15 mg (40%) 20 MS [M+Na]V = 576.6 (MW+Na calculated = 576.6 g/mol) Synthesis of compounds 26a and 26b oH RI N 25 -OS-Mnt 2- R6-N 0 R5 R2 26a R1 = Me, R2 = H, R5 = R6 =3-(dimethylamino)propyl 26b R= R2= R5 = Me, R6= 2-(dimethylamino)ethyl 30 267 mg (0.495 mmol) 25a was reacted with p-nitrophenylchlorofonnate and N-(3 dimethylamino-propyl)-N',N'dimthyl-propane-1,3-dianine (15g) to yield 26a as described for compound 16a. 26b was synthesiszed as described above using 15 mg 25b and N,N,1N'-trinethyl ethane-1,2-diamine (15a). Page 5,- 26a: Yield 282 mg (58%) as double TFA salt MS [M+Na]Y= 775.2 (MW+Na calculated = 776.0 g/mol) 26b: Yield 17 mg (70%) as TFA salt 5 MS [M+Na]* = 704.5 (MW+Na calculated = 704.6 g/mol) Synthesis of compounds 27a and 27b 0 H R1 N 10 / \ S-Mmt R6-N OH R5 P 27a RI = Me, R2= H, R5 = R6 = 3-(dimetbylamino)propyI 27b R1 = R2= RS = Me, R6 = 2-(dimethylamino)ethyl 15 26a (272 mg, 0.277 mmol, double TFA salt) was dissolved in 5 ml metha20ol, NaBH 4 (42 mg, 1.09 mrnol) was added and the mixture was stirred for 30 min at RT. 0.5 ml of acetic acid were added and 27a was purified by RP-HPLC. Alcohol 27b was synthesized likewise from 26b (17 mg, 25 pnol, TFA s alt). 20 27a: Yield 142 mg (52%) as double TFA salt MS [M+Na]+ = 777.9 (MW+Na calculated = 778.0 g/mol) 27b: Yield 6 mg (40%) as TFA salt MS [M+Na]*= 706.5 (MW+Na calculated = 706.6 g/mol) Pag 590 Synthesis of compounds 28a _and28b RI 00 R1 S-Mmt 5 R6-N 0 R5 R2 O NO2 28a RI = Me, R2 = H, RS = R6= 3-(dimethylamino)Propyl 28b R1 = R2 = RS = Me, R6 2-(dimethylamino)ethyl 10 Carbonates 28a or 28b were synthesized from 27a or 27b, respectively, as described. for compound 7a. is 28a: Yield 1 mg (29%) MS [M+Na]*= 942.9 (MW+Na calculated = 943.2 g/mol) 28b: Yield 1.5 mg (19%) MS [M+Na]*= 871.6 (MW+Na calculated = 871.7 g/mol) 20 Synthesis of comoounds 29a and 29b 0 Ri 0 R6SN /O R5 R2 O O SH 25 EB29 NH HA 1 //NH cLAI insulin \\H aB1 2 29a R1 M , R2= H, R5 = R6 = 3-(dimethylamilno)propyl 30 29b R1 = R2= R5 = Me, R6 = 2-(dimethylamino)ethyl Insulin derivatives 29a or 29b were synthesized from 28a or 28b, respectively, as described for compound 8a. Page 60.

29a MS [M+3H)"= 2105.8 [M+4HJ*= 1580.2 (MW calculated= 63164 g/mol) 29b MS [M+3H] = 2081.8 [M+4HJ*= 1562.4 (MW calculated = 6244 g/mol) Synthesis of mono-evylated insulin derivatives 30a and 30b 5 R6-,N>~ / R5 R2 O S- PEG5K 329 NH NH aAl 10 \nsulin J1H2 30a RI = Me, R2 H, R5 = R6 = 3-(dimethylamino)propyl 30b R1 = R2= R5 = Me, R6 = 2-(dimethylamino)ethyl 15 Insulin derivatives 30a or 30b were synthesized from 29a or 29b, respectively, as described for compound 9a. 20 30a and 30b: SEC retention time: 19.5 min Page 61. i Synthesis of a polymeric trodruz accrding to formula la with an ester-linked masicing o-oU and a dendritic carrier (9M) P2 R2R2 P2 IR2 R2 R2 R 50 0 0 0 k HN HH NHNHM HN NH HN NH HN w-I NH HNf 0 ~ 0 00 0 N H HN 0-J H I-I O N NH H-INOC R5 Rl RI RI RI R Ri R 0 0: t 0& 0 D0 0 0o-L 0 0 0 H0 NH HN NH HN NH HN N HNNH HN 20 ?N N 0 H HN H H cB29 0 25 N NH In sulK comp" NH S HN H-N 0 00 KNH 0

NH

2 page 62 0 H 0 O HN H 0 NH 0 -N YHNy-- ,-g o ,- 0 0 0 0 HN H HH 0 0 R2 00 N 1 N HN H H0 0 0 O O OH O--NNNN H 0kH 0 00 NH, 0 HNH 00 0 0 0 0 H { 0 HN j H H H 0 Co 0 -R A 0-o-0O N 0 O ,H N 0~N 0R2 0 0 NP He 0 0 H 8 K 0 00 N 0 H 0 P Page 63 Synthesis of compound 31 0 o 0 o H 0 0 NH O /N O O N N NI 0 O

H

2 NOC N NH 0 H o 0 o H 0 0 H 0 H H 0 0 O 'N,^'O- 0- 0-,N N N O H O O O O H~ O O 0 o 0 ,H 0 0 N H 0 O O0 N O O 31 5 31 was obtained according to the standard solid-phase synthesis protocol. The amino acids Fmoc-Dpr(B3oc)-OH, Fmoc-Dpr(Fmoc)-OH, Fmoc-Dpr(Fmoc)-OHI, Fmoc-Ado OH, and Fmoc-Dpr(Fmoc)-OH were coupled to NovaSyn TG Sieber amide resin. 10 After final fmoc removal the resin was agitated with 5 eq maleimidopropionic acid and 5 eq DIC in relation to amino groups in DMF for 30 min. 31 was cleaved from resin with TFA/TES/water 95/3/2 (v/v/v). After evaporation of solvent, product 31 was purified by RP-HPLC. Page 64.

MS: [M+H]*= 2494.6 (MW calculated = 2495.4 g/mol) Synthesis of compound 32 H H ,N - boc 0 0 H 0 0OH 0 ~ 0 H H HS NH H H HO N'' ON_ 0 oH 0 N

UO-N.

0 .- ~ H 0 H H S NHH 0 H H NN N 8 N boc H 0 N 0 > - N -N, - -N .boc 0 0 H 32 Compound 32 was obtained according to the standard solid-phase synthesis protocol. 5 The amino acids Fmoc-Cys(Mmt)-OH, Fmoc-Dpr(Fmoc)-OH, Fmoc-Dpr(Fmoc)-OH, Fmoc-Ado-OH, and Fmoc-Dpr(Fmoc)-OH were coupled to NovaSynTG Sieber amide resin. kfter final fmoc removal-the resin was agitated with 3 eq Boc-aminoxyacetic acid, 3 eq DIC, and 3 eq HOBt in relation to amino groups in DMF for 30 min. 32 was 10 cleaved from resin with DCM/TFA/TES 97/1/2 (v/v/v). After addition of 0.8 eq pyridine in relation to TFA the solvent was evaporated and product 32 was purified by RP-HPLC. MS: [M+H = 2688.2 g/mol (MW calculated =2688.8 g/mol) 15 Pag 65 Synthesis of compound 33 0

CONH

2 O HN H H o o 0 OH HN 0 N H 00 HH 0 H0 N HN 0 o OHO HN -O H 0 33 5 Compound 33 was obtained according to the standard solid-phase synthesis protocol. The amino acids Fmoc-Dpr(ivDde)-OH, Fmoc-Dpr(Fmoc)-OH, Fmoc-Dpr(Fmoc) OH, Fmoc-Lys(Fmoc)-OR, and Fmoc-Ado-OH were coupled to NovaSyn TG Sieber amide resin. i0 After final fmoc removal the resin was agitated with 3 eq 3,6,9-trioxadecanoic acid, 3 eq PyBOP, and 6 eq DIEA in relation to amino groups in DMF for 60 min. To cleave the ivDde protecting group, the resin was treated three times with 2% hydrazine in DMF. After washing, 3 eq Fmoc-Ser-OH was coupled with 3eq DIC and 3 eq HOBt for 30 min. After final fnoc removal resin was washed and the product is was cleaved from resin with DCM/TFA/TES 88>/1 0/2 (v/v/v). Solvent was evaporated and the residue was oxidized with 10 eq sodium periodate in 3/2 (v/v) 0.1 M sodium phosphate pH 7/acetonitrile for 15 min to yield 33. Product 33 was purified by RP HPLC and lyophilized. Page 66 MS: [M+fi' = 3372.1 g/rnol (3-372.8 g/mol) Synthesis of compound 34 H H ~.-ii Y~0' boc H 0H40 0 o0~ 0 N4 HN GoN.0>~ '0~ ~ 0bH 0 0~o- 0HHS 0 ~HB 0 0 0 N 00 H~ 0~ 0 H 0 HH o 02 00 H1 HH j7W 0R 0 0~ 0 3 1. pH ~ 31 2. HIemdprpoi acidO /DI H H 0~

H

2 0.N 00 C)0 4 H H 00 04 H 0 04 3400 0- ->-o - A.- R " O N. MH 0 0 " C)P0 ge 61H 6 mg (2.4 pmol) of compound 31 were dissolved in 1 ml 2/1 (v/v) acetonitrile/0.1 M sodium phosphate buffer pH 7 and 65 mg (24.2 pnol) of compound 32 were added. s The solution was stirred at room temperature for 2 h and then the product purified by RP-TPLC and lyophilized (yield: 45 mg (78%)). The lyophilized product (45 mg) was dissolved in 0.5 ml DMvF and 10 gl DIEA were added. 5 mg (30pLmol) 3-maleimidopropionic acid and 4.7 pl (30 pimol) DIC in 150 pl 10 DMF were added and the reaction mixture was stirred at room temperature for 20 min, the product purified by RP-ETPLC and lyophilized. The lyophilized product was incubated for 10 min in 95/5 (v/v) TFA/water and then the solvent was removed in a stream of nitrogen. Product 34 was purified by RP 15 HPLC and lyophilized (overall yield for all three steps: 20 mg (47%)). MS: 17700 - 18200 (broad peak) (MV calculated = 17749 g/mol) Synthesis of compound 9h 20 1.5 mg (225 nmol) 8g and 5 mg (280 rumol) 34 were mixed, dissolved in 300 i 2/1 (v/v) 0.1 M sodium phosphate buffer pH 7/acetonitrile and incubated for 15 min at room temperature. The product was purified by RP-HPLC and lyophilized. (yield 4 mg. 160 nrmol, 70%) 25 The lyophilized product was dissolved in 200 p1 0.1 M sodium citrate buffer pH 1.5 and 69 mg (20.5 pmol) 33 in 200 pl 2/1 (v/v) acetonitrile/sodium citrate buffer pH 1.5 were added. The mixture was stirred at room temperature for 24 h and product 9h was purified by size exclusion chromatography (column: Superdex 200, buffer: 10 mM 30 HEPES pH 7.4, 0.005% Tween-20, 3mM EDTA, flow rate: 0.75 ml/min) SEC elution time: 15 min Page 6B.

Synthesis of compound 37a HO O 5 H 2 N, O O I OH 37a 250 mg (0.35 mmol) 2-<hlorotrityl chloride resin (loading 1.4 mmol/g) was incubated io for 1.5 h with 308 mg (4 eq., 1.4 mmol) 4,7,10-trioxatridecane-1,13-diamine in 4 ml DCM to yield 35a. The resin was washed with DCM and dried. 107 mg (0.7 mmol) HOBt, 110 pl (0.

7 mmol) DIC, and 150 mg (0.9 mmol) 5-formyl salicylic acid in 3 ml DMF were added and the resulting suspension was stirred for I h at RT to yield 36a. After washing with DCM and THF, the resin was suspended in 6 ml THF and 3 ml (3 15 mmol) BH 3 THF (1 M in THF, 8.5 eq.) were added dropwise. The reaction mixture was stirred for 18 h at 451C under nitrogen atmosphere. After cooling 4 ml THF, 0.8 ml DIEA and 1.6 ml MeOH were added successively. 210 mg (0.84 mmol) 12 (as a concentrated TIF solution) were added and the suspension was stirred for 1 h. The resin was repeatedly washed (three times each) with THF, DMF, MeOH, and DCM. 20 The dried resin was reacted with 107 rag (0.7 mmol) HOBt, 110 1 (0,7 mmol) DIC, and 55 uL (0.9 mmol) AcOH in 3 ml DMF for 1h. After washing of the resin with DMF and DCM compound 37a was cleaved from resin with 2/1 (v/v) EHFIP/DCM (two times for 30 min). The volatile components were evaporated and the product 37a was used in the following step without furtherpurification. 25 37a: Yield 29 mg (20%) as TFA salt MS [M+Na]* = 421.4 (MW+Na calculated = 421.5 g/mol) 30 Page 69.1 Synthesis of comDound 38 HO H0 Mmt-S NO s 9 O N NO 0 OH 38 24 mg (0.06 mmol) 37a, 31 mg (0.06 mmol) PyBOP, 32 pl (0.18 mmol) DIEA, and i o 23 mg (0.06 mmol) 2 in 0.5 ml DMF were reacted for 50 rnin at room temperature. After addition of 50 ptl acetic acid product 38 was purified by RP-HPLC. 38: Yield 7 mg (15%) MS [M+Na] 4 = 781.3 (MW+Na calculated = 781.6 g/mol) 15 Synthesis of compound 39 OH

H

2 N 390 20 300 mg (0.42 mmol) 2-chlorotrityl chloride resin (loading 1.4 mmol/g) was incubated for 1.5 h with 245 mg (4 eq., 1.7 mmol) 1.8-dianinooctane in 4 ml DCM to yield 35b. The resin was washed with DCM and dried. 107 mg (0.7 rnmol) HOBT, 110 pl (0.7 mmol) DIC, and 150 mg (0.9 mmol) 5-formyl salicylic acid in 3 ml DMF were added and the resulting suspension was stirred for 1 h at RT to yield 36b. After washing with z5 DCM and THF, the resin was suspended in 6 ml THF and 3 m1 (3 mmol) BH 3 THF (1 M in THF) were added dropwise. The reaction mixture was stirred for 18 h at 45*C under nitrogen atmosphere. After cooling 4 ml THF, 0.8 mIl DIEA and 1.6 mi MeOH were added successively. 210 mg (0.84 mmol) 12 (as a concentrated THF solution) were added and the suspension was stirred for I h. The resin was repeatedly washed Page 7M ' (three times each) with THF, DMF, MeOH, and DCM. The dried resin was reacted with 107 mg (0.7 mmol) HOBT, 110 I (0.7 mmol) DIC, and 55 p.L (0.9 mmol) AcOH in 3 ml DMF for ih. After washing of the resin with DMF and DCM cornpound 37b, 78 mg (0.39 mmol) p-nitrophenylchloroformate, and 210 ul (1.2 5 mriol) DIEA in 1/1 (v/v) THF/DCM were reacted for 30 min at RT. The separated resin was suspended in 1/1 (v/v) THF/DCM and 210 .1 (1.

2 mmol) N,NN' trirnethylethylendiamine were added. The resulting suspension was stirred for 25 min at RT. The resin was separated and washed with DCM. Product 39 was cleaved from resin with 2/1 (v/v) HFIP/DCM (two times for 30 min). The volatile components were io evaporated and the product 39 was purified by HPLC. 39: Yield 16 mg (8%) as TFA salt MS [M+Na] = 473.5 (MW+Na calculated = 473.3 g/mol) Synthesis of compound 40a HO 20 w m 40a 25 38 (7 mg, 9 pmol) was dissolved in 200 g1 of dry THF. p-Nitrophenylchlorofornate (2.0 mg. 10pminol) and DIEA (4.4 p1, 25 pmol) were added and the mixture was stirred for 30 min at RT. N,N,N'-Triethylethylenediamine (15b) (18 pl, 0.1 rmmol) was added and stirring was continued for 30 min. The solvent was removed in vacuo, 10 30 pL of AcOH were added and 40a was purified by RP-HPLC. 40a: Yield 1 mg (11%) as TFA salt MS [M+Naj = 951.1 (NW+Na calculated = 951.8 g/mol) Page 7L., - Synthesis of compound 40b OH 0<~ Mmi-S N H O 40b 15 mg (33 mol) 39, 18 nig (33 pmol) PyBOP, 23 p.1 (0.13 mimol) DIEA, and 13 Mn (35 Pmol) 2 in 0.5 ml DMF were reacted for 45 min at room temperature. After addition of 50 [LI acetic acid product 40 was purified by RP-HPLC. 10 40b: Yield 10 mg (37%) as TFA salt MS pM+H] 4 = 811.5 (MW +Na calculated 810.5 g/mol) Page 72.

Synthesis of compound 41a and 41b

NO

2 o CO 0 H 0< 0 10 0 O N 41 a No 2 1.5 O 0 Mmnt-S N 20 4 b 41b Carbonate 41a or 41b vras synthesized from 40a or 40b as described for compound 7a 25 41a: Yield 0.4 mg as TFA salt MS [M+Na] 4 = 1116.8 (MW+Na calculated = 1116.9 g/mol) 41b: Yield 2 mg (16%) as TFA salt MS [M+H)*= 976.8 (1NfW calculated = 975.8 g/mol) 30 Synthesis of compound 42 N N 0 -B29 N / Insulin 42 >H2 10 Insulin derivative 42 was synthesized front 41b as described for compound 8a. 42 MS [M+3H]V= 2124.5 [M+4HJ 4 + 1594.6 (MW calculated 6371 g/mol) 15 Synthesis of compound 43 0 Y H o N'Suc 20 N o / PEG5K FB29 NH 0 Insulin 43 >H2 25 Insulin derivative 43 was synthesized from 42 as described for compound 9a. 43: SEC retention time: 18.0 min 30 Page 74.

Synthesis of rh-insulin loaded PEGA hvdroLel 45 0 PEGA-hydrogel - R2 + H9 N 0 DIC 0 ' N> PEGA.-hydroge N 10 0 44 20b 150 0 PEGA-hydrogel H/ 0 20 \~N ~ o Insulin 45 25 1'aleimide derivatization of polyacr3'lamide based- hydrogel (PEGA):

NI{

2 -PEGA hydrogel beads with 0.4 rnolg loadi-ng and 150-300 im bead size were purchased from Novabiochern. 30 2.5 g mnethan~ol-wet -NH2PEGA-hydrogel (0.4 mraol/g NH 2 -loading) was weighed into a syringe equipped with a polypropylene frit iMaleirnide loading, was adjusted by acylation employing a mixture of activated maleirnidopropioruc acid and acetic acid as described in the following. The hydrogel was vkrasbed 5 times with DMF and reacted with 13.5 mg (0.08 minol) 3-maeimidoprc)pionic acid, 115.2 lii (1.92 rwnol) Page 75.

acetic acid and 313 pl (2 mmol) DIC in 4 ml DMF for 30 main. The maleimide derivatized hydrogel 44 was washed 10 times with DMF and DCM and finally with acetonitrile. 5 30 mg of maleimide derivatized resin 44 (loading 16 pmolg) was reacted with 3 mg of compound 20b (480 nmol, 1.06 eq) in 600 p 1 20/80 (v/v') acetonitrile/50 mM phosphate buffer (pH 7.4) for 10 min to give rh-insulin loacled hydrogel 45. The hydrogel 45 was washed 5 times with 50/50 (v/v) acetonitrile/water and three times with acetonitrile and dried under vacuum. 10 Synthesis of rh-insulin carbohydrate-based bydrosel 46 0 Sepharose-hydrogel- N 15 O s 0 0 0 0O H N HO 0 z:B29NH H 20 k / NH a.A Insulin

NH

2 \ 2 46 25 NHS-activated "Sepharose 4 Fast Flow" hydrogel beads (chemicaly crosslinked agarose, crosslinker epichlorbydria) were purchased from Armersham. 1.5 g ethanol-wet Sepharose hydrogen (150 mg dry hydrogel1) was weighed into a syringe equipped with a polypropylene frit and reacted with I 1M 4,7,10 trioxatridecane-1,13-diamine in DMF for 30 min. After 5 crashing steps with DMF, 30 bydrogel was reacted with 8.5 mg (0.05 mmol) 3-maleimidopropionic acid, 57 ld (0.95 mmol) acetic acid, 151 mg (1 mmol) HOBt and 158 p. (1 mmol) DIC in 4 ml DMF for 30 min to give maleimide derivatized hydrogel. The hydrogel was washed 10 times with DMF and finally with acetonitrile. Page 76: 1.5 mg 8c was dissolved in 25/75 (vA') acetonitrile/50 mM phosphate buffer pH 7.4 and reacted with 10.8 mg maleimide derivatized hydrogel for 10 min. The rh-ins ulin loaded hydrogel 46 was washed five times with 50/50 (v/v) acetonitrile/water arid three times with acetonitrile and dried under vacuum. Synthesis scheme of Fuorescein-Insulin-rHSA (50) 10 H Bf InSUIM S + 4747 15 1 + 4 VSAN S 0 N- H N 20 49 020 C s 0 -,, 0 HSA 4 50, can H -I , F \ 01:

,,-

30 cyMS 0 HSA Page 77, Synthesis of bismaleimide 47 00 0 0 N ~ N H O1O 0 4.7 0 3-Maleimido-propionic acid (92 mg, 0.54 mmol) in 200 p.1 DMF was reacted vith DIC (78 p, 0.50 mmol) at RT for 15 min. 4,7,10-Trioxa-tridecan-1,13-diamin e (43.5 1o pl, 0.20 mmol) was added and the mixture was stirred for 30 min at RT. After addition of 800 gl 1/4 (v/v) acetic acid/water 47 was purified by RP-H1LC. 47: Yield 23 mg (22%) MS [M+Na1] = 545.5 (MW+Na calculated = 545.6 g/mol) 15 Synthesis of rHSA-maleimide (48) 66.5 p1 3 mM rHSA solution in 145 mM NaCI, 32 mM sodium octanoate, 0.0015% Tween-80 was mixed with 66.5 p.1 0.5 M phosphate buffer pH 7.0. 0.41 Mg 20 bismaleinide 47 (0.8 ponol) were added and the mixture was.reacted for 15 min at RT. Compound 48 was purified by SEC (column: Superdex 200, flow rate: 0.75 ml/min) using 10 mM HEPES buffer pH 7.4, 150 mM NaCl, 3 mM EDTA, an d 0.005% Tween as mobile phase. (Yield: 2.6 nil 77.5 p.M 48) 25 SEC retention time: 17.1 min (280 nm) ESI-MS = 66988 (MW calculated = 66984 g/mol) Synthesis of fluorescein-insulin-linker--maleimide (49) 30 40 p1 of 2.4 mM bismaleimide 47 in 1/1 (v/v) acetonitrile/water (96 nmol) wer-e mixed with 40 p1 0.5 M sodium borate buffer pH 5.8. 24 nmol 8f in 16.8 p1l 1/1 (v/v) acetonitrile/water were added and the mixture was incubated for 10 min at RT. 5 pl AcOH were added and 49 was purified by RP-HPLC. Page 78 ESI-MS = 7211 (MW calculated = 7211 g/mol) Synthesis of fluorescein-insulin-inke-rHSA 50 a) from 49 and rHSA 5 b) from 48 and 8f a) 30 pl of 80 p.M 49 in 1/1 (vA) acetonitrile/water (2.4 nmol) were mixed with 70 pl 0.25 M sodium phosphate buffer pH 6.4. 8 p1 3 mM rHSA in 145 mM NaCI 32 mM 10 sodium octanoate, 0.0015% Tween-80, (24 nmol) was added and the mixture was incubated at RT for 20 min. Compound 50 was purified by SEC (column: Superdex 200, flow rate: 0.75 ml/min) using 10 mM HEPES buffer pH 7.4, 150 mM NaCl, 3 mM EDTA, and 0.005% Tween as mobile phase. 15 SEC retention time: 17.3 min (500 nm) ESI-MS = 73676 (MW calculated = 73673 g/mol) b) 20 SEC eluate of rHSA-maleimide 48 (241 p.1, 77.5 M, 18.7 nmol) was mixed with 20 p.

1 0.5 M sodium borate buffer pH 5.8. 14 pl 1.41 mM 8f (19.6 amol) in 1/1 (v/v) acetonitrile/water were added and the mixture was incubated at RT for 10 mii. 1.2 p.1 48.5 mM 3-maleimido propionic acid (58 nmol) in 1/1 (v/v) acetonitrile/water were added and compound 50 was purified by SEC (column: Superdex 200, flow rate: 0.75 25 ml/min) using 10 mM HEPES buffer pH 7.4, 150 mM NaCl, 3 mM EDTA, and 0.005% Tween as mobile phase, SEC retention time: 17.1 min (500 umn) ESI-MS = 73698 (MW calculated = 73673 g/mol) 30 Page 79 Synthesis scheme of rHSA-linker-GLP-1 53a and 53b -o S-amin 0 10 100 40 Suc-Linker-rHSA -N -o N S HR1 Sia R1 = GLP-1 51b RI = FluoreScein-GLP-1 53a RI = GLP-1 53b RI = Fluorescein-GLP-1 rHSA 20 00 H %>Q _- ~-\ Fluorescein-GLP-1 52 25 Svnthesis of 51a GLP(7-36) (sequence: HAEGTFTSDVSSYLEGQAAKEFIAWLVKGR-amide) was synthesized on Rink-amide resin employing fmoc-strategy (Specialty Peptide 30 Laboratories, Heidelberg, Germany). N-terminal fmoc-protecting group was removed and the resin was washed with DCM and dried. 118 mg resin (0.11 mmol/g, 13.2 pmol) was suspended in a solution of 50 mg 19i (53 pmol) in 750 41 dry DMSO and 22.4 g1 DIEA. 2.1 p±l pyridine was added and the mixture was shaken for 48 h at RT. After washing the resin 6 times each with DMF and DCM, cleavage of the peptide Page 80 from resin and removal of protecting groups was achieved with 96/2/2 (v/v/v) TFA/triethylsilane/water. Volatiles were removed under nitrogen flow and 51a was purified by RP-HPLC and lyophilized. s 51a: Yield 4.6 mg (9%) MS: [M+3H13+= 1251.0 (MW calculated = 3750.3 g/mol) Synthesis of 51b Lys28 ivDde side chain protected GLP(7-36) (sequence: io HAEGTFTSDVSSYLEGQAAKEFIAWLVK(ivDd)GR-amnide) was synthesized on Rink-amide resin employing fmoc-strategy (Specialty Peptide Laboratories, Heidelberg, Germany). N-terminal fmoc-protecting group was removed and the resin was washed with DCM and dried. 50 mg resin (0.11 mmol/g, 5.5 ±mol) was suspended in a solution of 25 mg 19i (26 gmol) in 400 41 dry DMSO and 11.2 pL 15 DIEA. 1.1 pl pyridine was added and the mixture was shaken for 48 h at RT. After washing the resin six times with DMF the ivDde protecting group was cleaved by incubating the resin 3 times with 5% hydrazine in DMF for 20 min. Fmoc-8-amino, 3,6-dioxaoctanoic acid was coupled according to the standard coupling cycle. Fmoc protecting group was removed and carboxy-fluorescein was coupled by incubating the 20 resin with 8 mg 5-(and-6)-carboxyfluorescein succinimidyl ester and 2 p.1 DIEA for 60 mn. Resin was washed six times each with DMF and DCM. Cleavage of the peptide from resin and removal of protecting groups was achieved with 96/2/2 (v/v/v) TFAltriethylsilane/water. Volatiles were removed under nitrogen flow. 51b was used for the synthesis of 52 without further purification. 25 MS: [M+3HJ' = 1064.3, [M+2H1 3 = 1418.3 (MW calculated = 4254 g/mol) Synthesis of 52 Raw material 51b was dissolved in 500 .1 1/1 (v/v) acetonitrile / 0.25 M sodium phosphate pH 7 and 8 mg NN'-bis( -maleimidopropionyl)-2-hydroxy-1,3 30 propanediamine were added. The solution was stirred at RT for 15 min and 52 was purified by RP-HPLC and lyophilized. 52: Yield- 5.1 mg MS [M+3H]* = 1162.8, [M+2H)" = 1549.4 (MW calculated = 4645 g/mol) Page 81' Synthesis of compound 53a 0

-

S u-LiUnker-HSA 0 C -N HN \ GLP-1 53a 5 30 pl 1.57 mM 48 (47 nmol) in 10mM HEPES buffer pH 7.4, 150 mM NaCI, 3 mM EDTA, and 0.005% Tween were mixed with 10 il 0.5 M sodium phosphate buffer pH 7.4. A mixture of 2 pl DMSO and 12 pl of 6.06 mM 51a (73 nmol) in water/acetonitrile 9/1 (v/v) was added and the solution was incubated at RT for 30 min. 53a was purified by SEC (column: Superdex 200, flow rate: 0.75 ml/min) using 10 10 mM phosphate buffer pH 7.4, 150 mM NaCl, and 0.005% Tween as mobile phase. SEC retention time: 17.7 min (280 nm) ESI-MS = 70745 (MW calculated = 70734 g/mol) 15 Synthesis of 53b ,Suc-Unker-HSA 00 HN \ Fluorescein-GLP-1 53b 100 pl 3 mM 52 (300 nmol) in 9/1 50 mM sodium phosphate pH 7.0/ acetonitrile were mixed with 100 p.

1 3 mM HSA (300 mnol) and the solution was incubated at RT 20 for 30 min. 53b was purified by SEC (column: Superdex 200, flow rate: 0.75 ml/min) using 10 mM HEPES buffer pH 7.4, 150 mM NaCl, 3 mM EDTA, and 0.005% Tween as mobile phase. SEC retention time: 17.7 min (500 nm) Page 82 Synthesis of comPounds 54a and 54b Br RI HO - 0 54a R1 = Me 5 54b R1 = H AIC1 3 (1.05 eq) was suspended in DCM and 6-bromohexanoic acid chloride (1 eq) was added. After stirring at RT for 20 min o-cresol (1 eq) was added and the mixture was reacted at RT for 25 min. The reaction mixture was poured into ice water and 10 extracted with ethyl acetate. The separated organic phase was dried over Na 2 SO4 and concentrated in vacuo. Product 54a was purified by silica gel column chromatography using heptane/ethyl acetate (4/1) as mobile phase. 54b was synthesized as described above using 6-bromohexanoic acid chloride and phenol. 15 54a: Yield 3.7 g (33%) MS [M+H* = 285.1 and 287.2 (MW+H calculated = 386.2 g/mol) 54b: Yield 620 mg (15%) MS [M+1]*= 271.2 (MW calculated = 271.0 g/mol) 20 Synthesis of compounds 55a and 55b S RI Trt HO / \ - 0 25 55a RI = Me 55b Rl = H Page 83.

DBU (105 g, 701 umol) was added to a solution of bromide 54a (105 mg, 369 jimol) and tritylthiol (204 mg, 738 mrnol) in 50 ml dry DMSO. The reaction mixture was stirred at RT for 40 min and acidified with 1 N H 2

SO

4 . The aqueous layer was extracted with ethyl acetate and evaporated. 55a was purified by RP-HPLC. 5 55b was synthesized according to the same protocol using 54b (180 mg, 0.66 mmol). 55a: Yield 173 mg (97%) MS [M+Na] = 503.6 (MW+Na calculated = 503.7 g/mol) 55b: Yield 160 mg (85%) 10 MS [M+Na]= 489.5 (MW+Na calculated = 489.3 g/mol) Synthesis of como~ounds 56a and 56b S R1 Trt R2 O R3 0 15 56a RI = Me, R2= R3 = 3-(dimethylamino)propyl, 56b RI H, R2= Me, R3 = 3-(N-ethyl-N-methylamino)propyl 56a was prepared from 55a (9 mg, 19 p.mol), p-nitrophenyl-chloroformoate and bis(3 dimethylamino-propyl)amine (21 pl, 94 pimol) as described for compound 16a. 20 56b was synthesized from 55b (160 mg, 0.34 mmol), p-nitrophenyl-chlorofornate and N-ethyl-NN'dimethyl- 1,3-propanediamine (15c) as described for compound 16a. 56a: Yield 12 mg (70%) as TFA salt MS [M+Na] 4 = 716.8 (MW+Na calculated = 717.0 g/mol) 25 56b: Yield 80 mg (32%) as TFA salt MS [M+Na]*= 645.6 (MW+Na calculated = 645.4 g/mol) 30 Pagc 84, Synthesis of compounds 57a and 57b Ri Trt R2 0 \ N -- OH R3 0 5 57a RI = Me, R2 R3 = 3-dimethylamino-propyl, 57b RI = H, R2= Me, R3 = 3-(N-ethyl-N-methylamino)ProPyl 57a and 57b were synthesized from 56a (12 mg, 13 Lmol, double TFA salt) and 56b 10 (80 mg, 110 jimol, TFA salt), respectively, as described for compound 16g. 57a: Yield 9 mg (75%) as TFA salt MS [M+Na]* = 719.0 (MW+Na calculated= 718.7 g/mol) 57b: Yield 60 mg (75%) as TFA salt is MS [M+Na]* =647.4 (MW+Na calculated = 647.4 g/mol) Synthesis of compounds 58a and 58b S RI Trt R2 O R3 0 O= O O / NO 2 20 58a R1 = Me, R2 = R3 = 3-(dimethylamino)propyl, 58b RI = H, R2= Me, R3 = 3-(N-etbyl-N-methylamio)propyl 25 57a (1 eq, 8mg, 9 imol), 4-nitrophenyl chloroformate (3.5 eq, 6 mg, 30 ±mol), DIEA (6 eq, 9 1d, 52 ptmol), and DMAP (1 eq, 1 mg, 9 ptmol) were stirred in 1 ml dry DCM Page 85 at RT for 45 min under nitrogen atmosphere. The volatiles were evaporated and acetic acid was added. The mixture was dissolved in 1/1 (v/v) acetonitrile/water and the carbonate 58a was purified by RP-HPLC. Carbonate 58b was prepared likew-ise from 57b (135 mg, 0.18 mmol). 58a: Yield 7 mg (70%) as TFA salt MS [M+Naf4= 883.8 (MW+Na calculated = 884.1 g/mol) 58b: Yield 110 mg (77%) as TFA salt MS [M+Naf = 812.4 (MW+Na calculated= 812.5 g/mol) 10 Synthesis of compound 59 N 0 SH eB 29 NH H A1 Insulin S1NH2 caB12 is Rh-Insulin (44.5 mg, 7.7 .imol), carbonate 58a (1 eq, 7 mg, 6.4 mmol), DIEA (15 pl, 88 pmol) and DMAP (1.5 mg, 12 Imol) in 0.3 ml DMSO were reacted at RT for 30 min. The reaction mixture was neutralized with acetic acid and diluted with 1/1 (v/v) acetonitrile/water. RP-HPLC purification gave the appropriate Trt-protected intermediate. 20 After lyophilization, the Trt-protected intermediate was mixed with 95/5 (v/v) TFA/triethylsilane and stirred for 5 min. Volatiles were removed under nitrogen flow and 59 was purified by RP-HPLC and lyophilized. Position of insulin modification was verified by DTT reduction and MS analysis. 25 59: MS [M+3H)" = 2095.5 (M+4H] 4 = 1572.2 (MW calculated = 6288 g/mol) Pas: 86 Synthesis of compound 60 N<O ~S--Suc-PEGk Ns eB29 NHO H A INH oLA1 60 Insulin NH2 asi 5 60 was prepared from 59 (0.17 p.mol) as described for compound 9a. 60: SEC retention time: 19.5 min Synthesis of compound 61 10 S Trt O-N 61 57b (70 mg, 90 pLmol), DSC (161 mg, 630 prnol), DIEA (192 Ll, 1.1 mmol), and DMAP (11 Img, 90 .rmol) were stirred in I rn.1 dry acetonitril at RT for 14 h under 15 nitrogen atmosphere. The volatiles were evaporated and acetic acid was added. The mixture was dissolved in 1/1 (v/v) acetonitrile/water and the carbonate 61 was purified by RP-H{PLC. 61: Yield 40 mg (51%) as TFA salt 20 MS [M+Na]* = 788.4 (MW+Na calculated = 788.5 g/mol) Page 87 - Synthesis of compound 62 0 2 N /S \ S O-N 62 S0 61 (12 mg, 13 Vtmol) and NPys-Cl (4 mg, 21 pmol) were stirred in I ml DCM at 1 0*C for 2 h under nitrogen atmosphere. Volatiles were removed under nitrogen flow and 62 was purified by RP-HPLC. 10 62: Yield 7 mg (65%) as TFA salt MS [M+Na]*= 700.9 (MW+Na calculated =701.4 g/mol) 15 Synthesis of coMDound 63 N Sfs-Suc-PEG5k 0 Qzz oz o 63 /NH rhGH 0.9 mg desalted rhGH (ProspecTany, Israel, MW 22250 g/mol, 40 nmol) in 200 jtd 50 20 mM borate buffer (pH 8.0), 8 .i1 of carbonate 62 in acetonitrile (38 mM, 300 nmol), and 40 41 DMSO were reacted at RT for 3 h. The solvent mixture and low molecular Page S8 weighed compounds were replaced by water and subsequently by acetate buffer (25 mM, pH 4.2, 0.005% Tween 20) by ultrafiltration using Centricon 5 filter (cutoff 5 kDa). 8 p.l (80 nmol) 10 mM DTT in 25 mM acetate buffer pH 4-2, 0.005% Tween was added and incubated at RT for 30 min. Low molecular weight compounds were 5 removed by ultrafiltration using Centricon 5 filter and 25 mM acetate buffer pH 4.2, 0.005% Tween as eluate. After concentration to a volume of 100 p1 (Centricon 5) 20 p.1 (100 nmol) 5 mM maleimide-PEG5k in water and 80 ul 0.5 M phosphate buffer pH 7.0 were added. The mixture was incubated at RT for 5 min. Monoconjugate 63 was separated by SEC (column: Superdex 200, flow rate: 0.75 rnl/rain) using 10 mM 10 phosphate buffer pH 7.4, 150 mM NaCl, and 0.005% Tween 20 as mobile phase. The collected eluate (approximately 1.0 ml) was diluted with 0.5 ml buffer containing 0.05% NaN 3 and directly used for release rate determination. 63: SEC retention time: 17.5 min 15 Release of insulin or fluorescein-insulin from coniugates in buffer pH 7.4 Release of (fluorescein)-insulin from (fluorescein)-insulin conjugates 9a to 9h, 21a to 21f, 30a, 30b, 43, 50, and 60, release of fluorescein-GLP-1 from 53b, and release of 20 rbGH from 63 was effected by linker hydrolysis in aqueous buffer pH 7.4. Collected SEC eluates of (fluorescein)-insulin conjugates (see above), fluorescein-GLP-1 conjugate and rhGH-conjugate, respectively, were incubated at 37 *C and samples were taken at time intervals and analyzed by RP-HPLC (iasulin conjugates) or SEC (rhGH conjugate, fluorescein insulin conjugates and fluorescein-GLP- 1 conjugate) 25 and UV detection at 215 or 280 nm or VIS detection at 500 rim. Peaks correlating with the retention time of native insulin, fluorescein-insulin, fluorescein-GLP- 1, and rhGH, respectively, were integrated and plotted against incubation time, and curve fitting software was applied to estimate the corresponding halftime of release. 30 Release of insulin from ydrogel coniuaates 45 and 46 4 mg of 45 or 2 mg 46 were weighed into a test tube and incubated with 1 ml 10 mM HEPES buffer pH 7.4, 150 mM NaCt, 0.005% Tween at 37*C. 45 ji samples were Page B9 taken at different time intervals and quantitatively analyzed for rh-insulin by a RP H{PLC assay. The rh-insulin peaks were integrated and rh-insulin concentration was obtained from a standard curve. A first order release kinetic was fitted to the data points to give the linker half life. 5 MS-analysis of released insulin from compound 9a. 9b and 30a Samples of buffer released insulin (see above) were analyzed by mass spectrometry. Figure 11 shows the mass spectra of released insulin from compound 9a, 9b, and 30a. 10 The mass spectrum of insulin released from compound 9a clearly sbows a major side product indicatedd by arrows), corresponding to the irreversibly pentanoyl-modified insulin. In this case, removal of the pentanoyl masking goup was not by hydrolysis but by acyl transfer to the insulin. The mass spectrum of insulin released from compound 9b and 30a shows no modification. L5 Release of fluorescein-insulin from conj1ate 9d and 9e in 80% human plasma Release of fluorescein-insulin from 9d or 9e was effected by hydrolysis in 80% human plasma in 20 mM HEPES pH 7.4 at 37 *C. Samples were taken at time 20 intervals and analyzed by SEC and VIS detection at 500 nm. Peaks correlating with the retention time of fluorescein-insulin were integrated and plotted against incubation time, and curve-fitting software was applied to estimate the corresponding halftime of release. 25 Table: Polymeric prodrug hydrolysis compound ti buffer pH 7.4 t 1 iahuman plasma 9a 40 h nd 9b 55 h nd 9c 4.5 d nd 9d 7h 4 h 9e 55 h 30 h 9f 90 h nd 9g 37 h nd Page 90, 9h 98 -h nd 21a 64 d nd r1b 8 >d nd 21c 52 d nd 21d 29 d nd 21e 100 d nd 21f 83 h nd 30a 17 d nd 30h >70 d nd 43 4 h nd 45 7 d nd 46 4 d nd 50 57 h nd 53b 19 h nd 60 10 d nd 63 51 d nd nd = not determined The foregoing is considered illustrative of the principles of the invention and since numerous modifications will occur to those skilled in the art, it is not intended. to limit 5 the invention to the exact construction and operation described. All suitable modifications and equivalents fall within the scope of the claims. Page 90 - Abbreviations: Boc t-butyloxycarbonyl DBU 1,3-diazabicyclo[5.4.0)undecene DCM dichloromethane s (iv)Dde 1--(4,4-dinethyl-2,6-dioxo-cyclohexyliden)3-methyl-butyI DIC diisopropylcarbodiimide DIEA diisopropylethylamine DMAP dimethylarnino-pyridine DMF N,N-dimethylformamide 10 DMSO dimethylsulfoxide Dpr diaminopropionic acid DSC disuccinidylcarbonate EDTA ethylenediaminetetraacetic acid Et ethyl 15 eq stoichiometric equivalent fmoc 9-fluorenylmethoxycarbonyl Fmoc-Ado-OH Fmoc-8-arnino-3,6-dioxaoctanoic acid HFIP hexafluoro isopropanol BEPES N-(2-hydroxyethyl) piperazine-N'-(2-ethanesulfonic acid) 20 HOBt N-bydroxybenzotriazole LCMS mass spectrometry-coupled liquid chromatography Mal maleimidopropionyl Me methyl Mmt 4-methoxytrityl 25 MS mass spectrum MW molecular mass Npys 3-nitro-2-pyridinesulfenyl PyBOP benzotriazole-1-yl-oxy-tris-pyrrolidino-phosphonium hexafluorophosphate 30 rHSA recombinant human serum albumin rbGH recombinant human growth hormone RP-HPLC reversed-phase high performance liquid chromatography RT room temperature Page 92 SEC size exclusion chromatography Suc succinimidopropionll TES tiethylsilane TFA trifluoroacetic acid TaF tetahydrofurane, UV ultraviolet vi1S visual z benzylox3ycarboflyl 10 20 25 30 Page 93

Claims (27)

1. 2. The method of claim 1, wherein a first one of the at least one intermediate compound is an intermediate molecule of Formula III [R4]n ,PG 1 y R2 Y 2 03 Nu-W-Y 4 Ar R3 (Ill) 20 Y synthesized by acylating Y2 with 4 and optionally attaching a first protecting group PGI, wherein Nu is a nucleophile; W is selected from substituted or non-substituted linear, branched or cyclical alkyl, 25 aryls, substituted aryls, substituted or non-substituted linear, branched or cyclical heteroalkyl, substituted or nonsubstituted heteroaryls; Y, is selected from the group of moieties consisting of 0, S, or NR6; Y 4 is selected from the group of moieties consisting of -C(R7)(R8)-, 0, or NR6; R7 and R8 are selected independently from hydrogen, substituted or non-substituted 30 linear, branched or cyclical alkyl or heteroalkyl, aryl, substituted aryl, substituted or non-substituted heteroaryl, carboxyalkyl, alkylcarbonyl, carboxamidoalkyl, cyano or halogen. Page 95 R6 is selected selected from hydrogen, substituted or non-substituted linear, branched or cyclical alkyl or heteroalkyl, aryls, substituted aryls, substituted or non-substituted heteroaryls; 5 3. The method of claim 2, wherein a second one of the at least one intermediate compound is an intermediate compound of formula (IV) PG1 [R4], I G y xR2 Y 2 0 A Nu-W-Y 4 Ar R3 Y 5 (IV) formed by activating the compound of Formula III with an activating agent, wherein A is a leaving group and Ys is selected from 0 or S. 10
2. 4. The method of claim 3, wherein A is selected from chloride, bromide, fluoride, nitrophenoxy, imidazolyl, N-hydroxysuccinimidyl, N-hydroxybenzotriazolyl, N hydroxyazobenzotriazolyl, pentafluorphenoxy and N-hydroxysulfosuccinimidyl. 15 5. The method of one of claims 3 or 4, wherein the activating agent is selected from the group of activating agents consisting of 4-nitrophenyl chloroformate or disuccinyl carbonate.
3. 6. The method of one of claims 3 or 4, wherein the amine-containing biologically'active 20 moiety D is attached to the second one of the at least one intermediary compound (Formula IV) by displacement of the leaving group A.
4. 7. The method of claim 6 comprising a step of removal of the reversible first protecting group PG 1 from the second one of the at least one intermediate compound. 25
5. 8. The method of claim 7, wherein the step of removal of the reversible protection group PGI is carried out using a reagent selected from the group of reagents consisting of trifluoroacetic acid or DTT. 30 9. The method of any one of claims 7 or 8, further comprising a step of attaching a polymer RI to X.
6. 10. The method of claim 2, further comprising a step of removing the first polymer protection group PG I and attaching a polymer Rl to X to form a third one of the at 35 least one intermediate compound and having a Fonnula VI: Page 96 [R4], / R1 R2 >-Y 2 0 Y3 Nu-W-Y 4 Ar R3 (VI).
7. 11. The method of claim 10, further comprising a step of activating the third one of the at 5 least one intermediate compound by using an activating agent to form a fourth one of the at least one intermediate compound and having a Formula VII and wherein A is a leaving group: R1 [R4]n I xR 2 y -2 3 YS- A Nu-W-Y 4 Ar R3 Y ( 10 12. The method of claim 11, wherein A is selected from chloride, bromide, fluoride, nitrophenoxy, imidazolyl, N-hydroxysuccinimidyl, N-hydroxybenzotriazolyl, N hydroxyazobenzotriazolyl, pentafluorphenoxy and N-hydroxysulfosuccinimidyl.
8. 13. The method of claim 11 or 12, wherein the activating agent is selected from the group 15 of activating agents consisting of 4-nitrophenyl chloroformate or disuccinyl carbonate.
9. 14. The method of one of claim I 1 to 13, wherein the amine-containing biologically active moiety D is attached to the compound of Formula VII by displacement of the leaving group A. 20
10. 15. The method of claim 1, further comprising a step of reacting the starting molecule of Formula II with a first polymer RI to attach the first polymer RI to X.
11. 16. The method of claim 15, further comprising a step of protecting Y 2 with a second 25 protective group PG 2 .
12. 17. The method of claim 15 or 16, further comprising the step of activating the prodrug with an activating agent and reacting with an amine-containing biologically active moiety D to form a fifth one of the at least one intermediate compound and having 30 Formula (X): Page 97 [R41, ,RI R2 PG 2 7Y 2 s ||D Ar R3 Y X.
13. 18. The method of claim 17, wherein the activating agent is selected from the group of activating agents consisting of 4-nitrophenyl chloroformate or disuccinyl carbonate 5
14. 19. The method of claim 17 of 18, wherein the second protecting group PG 2 is removed from Y 2 and Y2 is acylated with Y Nu-W-Y 4 wherein W is selected from substituted or non-substituted linear, branched or cyclical 1o alkyl, aryls, substituted aryls, substituted or non-substituted linear, branched or cyclical heteroalkyl, substituted or nonsubstituted heteroaryls; Nu is a nucleophile; Yj is selected from the group of moieties consisting of 0, S, or NR6; Y 4 is selected from the group of moieties consisting of -C(R7)(R8)-, 0, NR6; and 15 R7 and R8 are selected from hydrogen, substituted or non-substituted linear, branched or cyclical alkyl or heteroalkyl, aryl, substituted aryl, substituted or non-substituted heteroaryl, carboxyalkyl, alkylcarbonyl, carboxamidoalkyl, cyano or halogen.
15. 20. The method of claim 19, wherein the removal of the second reversible group PG 2 is 20 carried out using a reagent selected from the group of reagents consisting of trifluoroacetic acid or DTT.
16. 21. The method of claim I comprising a step of attaching a removable first protecting group PGI to X and a removable second protecting group PG 2 to the starting 25 compound of Formula 11 and thereafter activation using an activating agent to form a sixth one of the at least one intermediate compound and having Formula (VIll): [R4}., PG 1 X R2 PG2Y 2 Q Y A Ar R3 Y ( wherein A is a leaving group 30 22. The method of claim 21, wherein A is selected from the group of leaving groups selected from the group of leaving groups consisting of chloride, nitrophenoxy, Paqe 98 imidazolyly, N-hydroxysuccinimidyl, N-hydroxybenzotriazolyl, N hydroxyazobenzotriazolyl, pentaflourphenoxy and N-hydroxysulfosuccinimidyl.
17. 23. The method of claim 21 or 22, wherein the activating agent is selected from the group 5 of activating agents consisting of 4-nitrophenyl chloroformate or disuccinyl carbonate.
18. 24. The method of any one of claims 21 to 23, wherein the amine-containing biologically active moiety D is attached to the sixth one of the at least one intermediate compound by displacement of the leaving group A. 10
19. 25. The method of claim 24 comprising a step of removing the first protecting group PGI from X and attaching a polymer RI.
20. 26. The method of claim 25, wherein the second protecting group PG 2 is removed from Y 2 is and Y 2 is acylated with Nu-W-Y 4 wherein W is selected from substituted or non-substituted linear, branched or cyclical alkyl, aryls, substituted aryls, substituted or non-substituted linear, branched or 20 cyclical heteroalkyl, substituted or nonsubstituted heteroaryls; Nu is a nucleophile; Y 1 is selected from the group of moieties consisting of 0, S, or NR6; Y 4 is selected from the group of moieties consisting of -C(R7)(R8)-, 0, NR6; and 25 R7 and R8 are selected from hydrogen, substituted or non-substituted linear, branched or cyclical alkyl or heteroalkyl, aryl, substituted aryl, substituted or non-substituted heteroaryl, carboxyalkyl, alkylcarbonyl, carboxamidoalkyl, cyano or halogen.
21. 27. The method of claim 24, wherein the second protecting group PG 2 is removed from Y 2 30 and Y 2 is acylated with Nu-W -Y4 wherein W is selected from substituted or non-substituted linear, branched or cyclical alkyl, aryls, substituted aryls, substituted or non-substituted linear, branched or cyclical heteroalkyl, substituted or nonsubstituted heteroaryls; 35 Nu is a nucleophile; Yi is selected from the group of moieties consisting of 0, S, or NR6; Y 4 is selected from the group of moieties consisting of -C(R7)(R8)-, 0, NR6; and Pp: 99 R7 and R8 are selected from hydrogen, substituted or non-substituted linear alkyl, branched alkyl or cyclical alkyl, aryl, substituted aryl, substituted or non-substituted heteroalkyl or heteroaryl, carboxyalkyl, alkylcarbonyl, carboxamidoalkyl, cyano or halogen. 5
22. 28. The method of claim 27 comprising a step of removing the first protecting group PG from X and attaching a polymer RI.
23. 29. The method of claim 15 comprising a step acylating Y 2 with Y 10 Nu-W-Ya to form a third one of the at least one intermediate compound; wherein W is selected from substituted or non-substituted linear, branched or cyclical alkyl, aryls, substituted aryls, substituted or non-substituted linear, branched or cyclical heteroalkyl, substituted or nonsubstituted heteroaryls; 15 Nu is a nucleophile; Yj is selected from the group of moieties consisting of 0, S, or NR6; Y 4 is selected from the group of moieties consisting of -C(R7)(R8)-, 0, NR6; and R7 and R8 are selected from hydrogen, substituted or non-substituted linear alkyl, branched alkyl or cyclical alkyl, aryl, substituted aryl, substituted or non-substituted 20 heteroalkyl or heteroaryl, carboxyalkyl, alkylcarbonyl, carboxamidoalkyl, cyano or halogen.
24. 30. The method of claim 29, further comprising a step of activating the third one of the at least one intermediate compound by using an activating agent to form a seventh one of 25 the at least one intermediate compound and having a Formula VII and wherein A is a leaving group [R4), xR1 Y R2 ' Y2 0 YT A Nu-W-Y4 Ar R3 Y 5 s(yII).
25. 31. The method of claim 30, wherein A is selected from the group of leaving groups 30 consisting of chloride, nitrophenoxy, imidazolyl, N-hydroxysuccinimidyl, N hydroxybenzotriazolyl, N-hydroxyazobenzotriazoyl, pentafluorphenoxy and N hydroxysulfosuccinimidyl. Page 100
26. 32. The method of claim 30 or 31, wherein the activating agent is selected from the group of activating agents consisting of 4-ntirophenyl chloroformate or disuccinyl carbonate.
27. 33. The method of one of claims 30 to 32, wherein the amine-containing biologically 5 active moiety D is attached to the compound of Formula VII by displacement of the leaving group A. Paqe 101