HPMA copolymers: Origins, early developments, present, and future^☆

2.1. Early history

Our research in the sixties was concentrated on the hydrophilic biomedical polymers based on hydrophilic esters and N-substituted amides of (meth)acrylic acid. The focus was on four tasks [10]: a) selection of a suitable group of polymers and the study of their chemical properties – polymerization kinetics, stability, mechanical properties, permeability, possibility of introduction of functional (reactive) groups; b) study of the relationship between the chemical and physical structure of crosslinked polymers (hydrogels) and their biocompatibility; c) study of the interaction of water-soluble polymers with living organisms – using evaluation criteria from blood plasma expanders; and d) modification of soluble polymers by attachment of biologically active compounds (drugs, hormones, enzymes) as well as the modification of the polymer backbone and side-chains with enzymatically degradable sequences.

The choice of HPMA for development as drug carrier was not random. Based on the detailed studies of the relationship between the structure of hydrophilic polymers and their biocompatibility [11–21], we have chosen N-substituted methacrylamides as our target because the α-carbon substitution and the N-substituted amide bond ensured hydrolytic stability of the side-chains. We synthesized a series of compounds trying to identify a crystalline monomer for easy purification and reproducible synthesis. The first crystalline N-substituted methacrylamide we succeeded to synthesize, HPMA, was chosen for future development [22,23].

2.2. First HPMA copolymer drug and/or protein conjugates

The research on the use of HPMA copolymers as drug carriers commenced in the early 70s. In April 1974 we filed two patent applications [24,25] which covered the synthesis of N-substituted (meth)acrylamides containing oligopeptide sequences and their application as drug (and other biologically active compounds) carriers.

The first HPMA copolymer–drug (N-(4-aminobenzensulfonyl)-N′-butylurea) conjugate (Fig. 1) was presented at the Prague Microsymposiumon Polymers in Medicine in 1977 [26] and published [27]. At the same time studies in the modification of proteins were initiated – HPMA copolymer–insulin [28] and HPMA copolymer–chymotrypsin [29,30] conjugates.

Two major synthetic routes were used for the synthesis of HPMA copolymer–drug conjugates, copolymerization and polymer-analogous attachment [10,21]. Copolymerization of HPMA with a polymerizable derivative of a drug, e.g., N-(4-aminobenzensulfonyl)-N′-butylurea [27] is the example of the first route. Insulin [28], chymotrypsin [29,30] and ampicillin [31] (Fig. 2) were attached to HPMA copolymers by aminolysis [32,33] of reactive polymeric precursors. The development of HPMA copolymers is summarized in Table 1.

2.3. Development of oligopeptide spacers

Macromolecules are internalized by cells via endocytosis and ultimately localize in the (enzyme rich) lysosomal compartment. Consequently, we developed HPMA copolymers containing enzymatically degradable bonds (Fig. 3) [34]. Oligopeptide side-chains were designed as drug attachment/release sites [35] and shown to be degradable in vivo [36].

The relationship between the structure of oligopeptide sequences and the rate of enzymatically catalyzed release of a drug or drug model was studied thoroughly. First, model enzymes, chymotrypsin [30,37–39], trypsin [40], and papain [41] were evaluated, followed by intracellular (lysosomal) enzymes. A study with a mixture of lysosomal enzymes revealed that cysteine (thiol) proteinases were responsible for the cleavage of these polymer–drug conjugates [42,43]. Subsequently, polymers were designed to match the specificity of individual cysteine proteinases: cathepsins B [44], L, H, and artificial mixtures of lysosomal enzymes [45]. The stability of the oligopeptide side-chains in blood plasma and serum was verified [46]. Based on these results it was possible to control the degradability of HPMA copolymers by a particular enzyme. Early in vivo validation of the design was achieved – the drug model (p-nitroaniline) was released in in vivo following intravenous (i.v.) administration in Wistar rats [36].

Later, oligopeptide spacers were combined with self-immolative groups. For bone-specific delivery of prostaglandins a cathepsin K sensitive spacer combined with a 1,6-elimination unit was designed [50]; for oral delivery of 9-aminocamptothecin a combination of a reducible aromatic azo bond with a 1,6-elimination unit was employed [51].

The first degradable polymer carriers based on HPMA were also reported at the Polymers in Medicine Microsymposium in the Prague in 1977 [52] and at conferences in Varna [53] and Tashkent [54]. We used the oxidized insulin B chain (it contains two amino groups at positions 1 and 29) to prepare branched, water-soluble HPMA copolymers by reacting insulin B-chain with HPMA copolymers containing side-chains terminated in p-nitrophenyl esters. The polymers were cleavable (Fig. 4), so we chose the sequence 23–25 (Gly–Phe–Phe) from the insulin B-chain (the bond originating at amino acid 25 is cleavable by chymotrypsin) and synthesized branched, soluble high molecular weight enzymatically degradable copolymers containing the Gly–Phe–Phe segments in crosslinks connecting primary chains [38]. The latter type of polymer carrier was evaluated in vivo in rats and it was shown that the branched polymer carrier is degradable and its molecular weight distribution decreases with time following i.v. administration [36]. These experiments demonstrated the possibility to manipulate the intravascular half-life of polymeric carriers based on HPMA.

2.4. Validation of the targetability of HPMA copolymer–drug conjugates

The choice and design of a targeting system has to be based on a sound biological rationale. The design of the first targetable HPMA copolymer was based on the observation [56] that small changes in the structure of glycoproteins lead to dramatic changes in the fate of the modified glycoprotein in the organism. When a glycoprotein (ceruloplasmin) was administered into rats, a long intravascular half-life was observed. However, when the terminal sialic acid was removed from ceruloplasmin, the asialoglycoprotein (asialoceruloplasmin) formed contains side-chains exposing the penultimate galactose units. The intravascular half-life of the latter was dramatically shortened due to the biorecognition of the molecule by the asialoglycoprotein receptor on the hepatocytes. This receptor recognizes galactose and N-acetylgalactosamine moieties [56]. To determine if one can mimic this process with a synthetic macromolecule, we synthesized HPMA copolymers with N-methacryloylglycylglycine p-nitrophenyl ester and attached galactosamine by aminolysis [57]. These copolymers behaved similarly to the glycoproteins and were biorecognizable in vivo (Fig. 5). Their clearance from the bloodstream was related to the N-acylated galactosamine content (1–11 mol%) of the HPMA copolymer [57–59]. Separation of the rat liver into hepatocytes and non-parenchymal cells indicated that the polymer is largely associated with hepatocytes, and density-gradient subcellular fractionation of the liver confirmed that the HPMA copolymers were internalized by liver cells and transported, with time, into the secondary lysosomes [59,60]. It was very important to find that HPMA copolymers containing side-chains terminated in galactosamine and anticancer drug adriamycin also preferentially accumulated in the liver, i.e., it appeared that non-specific hydrophobic interactions with cell membranes did not interfere with the biorecognition by hepatocytes [61].

In parallel, efforts on the targetability of HPMA copolymer–antibody conjugates started. First HPMA copolymer conjugates with polyclonal and monoclonal anti-Thy-1.2 antibodies and anti-FITC (fluorescein isothiocyanate) antibodies were evaluated. Targetable conjugates containing daunomycin were synthesized and in vitro experiments have shown two orders of magnitude enhanced cytotoxicity of the targeted conjugate (when compared to the nontargeted one) [62]. The targetability and activity of anti-Thy1.2 conjugates with HPMA copolymer–daunomycin conjugates was proven in vivo on a mouse model [63]. Anti-Thy1.2 antibodies were also efficient in targeting HPMA copolymer–photosensitizer (chlorin e₆) conjugates [64].

2.5. Early interdisciplinary collaborations

At the beginning of the eighties, we started collaborations with coworkers from the biological field: John Lloyd and Ruth Duncan from the University of Keele in United Kingdom, and Blanka Říhová from the Institute of Microbiology in Prague. The collaboration with the Keele group was initiated by Helmuth Ringsdorf who gave a lecture at the 1977 Prague symposium (where Kopecek presented first HPMA copolymer–drug conjugates and biodegradable carriers based on HPMA). After the meeting Ringsdorf suggested to Lloyd to contact Kopecek because he thought that the collaboration would be beneficial for both. Kopecek met Lloyd in Dresden in July 1978 and they agreed on the evaluation of HPMA copolymer conjugates. First samples were synthesized (different side-chains terminated in p-nitroanilide as drug model) and evaluated at Keele for their cleavability by lysosomal enzymes [42,65] and their stability in blood plasma and serum [46]. More than 300 different polymer structures containing oligopeptide sequences were synthesized in the Prague laboratory [24,25,35,47], and biological properties of a number of them evaluated at Keele within a 10 year period [66,67]. The collaboration with Vladimír Kostka and coworkers from the Institute of Organic Chemistry and Biochemistry in Prague on the cleavability of peptide sequences in HPMA copolymers by cathepsin B [44, Fig. 4], the most important lysosomal cysteine proteinase, resulted in the identification of GFLG sequence, which is incorporated in all conjugates used in clinical trials. From the two fastest cleaving oligopeptides, GFLG and GFTA (see Fig. 3, example 5), we have chosen the GFLG sequence over the GFTA to avoid T; at that time we were worried about the potential immunogenicity.

In 1978 Kopecek gave a lecture at the Institute of Microbiology in Prague. After the lecture he discussed with Říhová and the collaboration with her group on the immunogenicity/biocompatibility [69–72] and biorecognition (targeting) [62–64] of HPMA conjugates commenced.

These collaborations resulted in the filing of “Polymeric drugs” patent application in 1985 [68]. Kopecek coined the name for the HPMA copolymers evaluated in clinical trials as PK1 and PK2(P for Prague, K for Keele) (Fig. 6).

Rationale

HPMA copolymer–drug conjugates are nanosized (5–20 nm) water-soluble constructs. Their unique structural, physicochemical, and biological properties are advantageous when compared to low molecular weight drugs. The concept of targeted polymer–drug conjugates was developed to address the lack of specificity of low molecular weight drugs for cancer cells. This approach was based on the work of DeDuve, who realized that the endocytic pathway is suitable for lysosomotropic drug delivery [2]. The features needed to design an effective conjugate [10,21,35,73–75] comprise a polymer–drug linker that is stable during transport [46] and able to release the drug in the lysosomal compartment of the target cell at a predetermined rate [34,47,75], adequate physicochemical properties of the conjugate (solubility, conformation in the biological environment) [48,76–78], and the capability to target the diseased cell or tissue by an active (receptor–ligand) [75] or a passive (pathophysiological) mechanism [79]. Since the activity of many drugs depends on their subcellular location, a mechanism for the manipulation of the subcellular drug location would be beneficial.

The advantages of polymer-bound drugs (when compared to low molecular weight drugs) are (reviewed in [10,47,67,73–75,80,81]): a) active uptake by fluid-phase pinocytosis (non-targeted polymer-bound drug) or receptor-mediated endocytosis (targeted polymer-bound drug), b) increased active accumulation of the drug at the tumor site by targeting [82–84], c) increased passive accumulation of the drug at the tumor site by the enhanced permeability and retention (EPR) effect [85], d) long-lasting circulation in the bloodstream [86], e) decreased non-specific toxicity of the conjugated drug [82], f) decreased immunogenicity of the targeting moiety [82], f) immunoprotecting and immunomobilizing activities [87], and g) modulation of the cell signaling and apoptotic pathways [88–94].

3.1. Enhanced permeability and retention effect

The enhanced permeability and retention (EPR) effect is the predominant mechanism by which soluble macromolecular anticancer drugs exhibit their therapeutic effect on solid tumors. The phenomenon is attributed to high vascular density of the tumor, increased permeability of tumor vessels, defective tumor vasculature, and malfunctioning or suppressed lymphatic drainage in the tumor interstitium [79]. Other factors, however, may have an opposite effect. For example, a high interstitial pressure may result in a convective fluid flow from the center of the tumor to the periphery, which might carry macromolecules [95]. Nevertheless, a number of studies showed increased accumulation of macromolecules in tumors as compared to that in normal tissue [85,96]. The degree of accumulation was dependent on molecular weight [85,86], charge [97,98], and their overall hydrophobic –hydrophilic character. The tumor type and microenvironment may influence its transport characteristics (pore cutoff size) [99].

The efficiency of extravasation into solid tumors depends on the concentration gradient between the vasculature and tumor tissue and time. Consequently, high molecular weight (long-circulating) polymer conjugates accumulate efficiently in tumor tissue [85] due to the EPR effect [79,100]. However, if they possess a non-degradable backbone, they may deposit and accumulate in various organs [18]. We have previously synthesized high molecular weight carriers by connecting HPMA chains via lysosomally degradable oligopeptide sequences [34] to form water-soluble branched conjugates [36,38–41,101–103]. Following intravenous (i.v.) administration to rats, the oligopeptide crosslinks were cleaved and the resulting lower molecular weight polymer chains were excreted into the urine [36]. These water-soluble copolymers were synthesized by crosslinking (short of gel point) of HPMA copolymer precursors (containing oligopeptide side-chains terminated in a reactive ester group) with diamines.

Later, we designed a new, reproducible synthetic pathway for long-circulating HPMA copolymers [85,104]. New crosslinking agents were synthesized and high molecular weight copolymers prepared by crosslinking copolymerization. The composition of the monomer mixture, however, has to be such that at the end of the polymerization the system is short of the gel point (water-soluble). This method [104] is also suitable for the synthesis of HPMA copolymers, which contain, in addition to oligopeptide crosslinks, oligopeptide side-chains terminated in doxorubicin (DOX) (or other anticancer drugs).

The influence of the molecular weight of such conjugates on their biological activity was evaluated [85]. Copolymerization of HPMA, a polymerizable derivative of DOX (N-methacryloylglycylphenylalanylleucylglycyl doxorubicin) and a crosslinking agent, N²,N⁵-bis(N-methacryloylglycylphenylalanylleucylglycyl) ornithine resulted in high molecular weight, branched, water-soluble HPMA copolymers containing lysosomally degradable oligopeptide sequences in the crosslinks as well as in side-chains terminated in DOX. Four conjugates with Mw of 22, 160, 895, 1230 kDa were prepared. Biodistribution of the conjugates and their treatment efficacy in nu/nu mice bearing s.c. human ovarian OVCAR-3 carcinoma xenografts were determined (Fig. 7). The half-life of conjugates in the blood was up to 5 times longer and the elimination rate from the tumor was up to 25 times slower as the Mw of conjugates increased from 22 to 1230 kDa. The treatment with HPMA copolymer-bound DOX possessing an Mw higher than 160 kDa inhibited the tumor growth more efficiently than that of 22 kDa or free DOX(p<0.02). The data clearly indicated that the higher the molecular weight of the conjugate the higher the treatment efficacy of human ovarian xenografts in nu/nu mice [85].

3.2. Multidrug resistance

The acquired resistance of malignant tumors to chemotherapeutic agents remains the major cause of cancer therapy failure [105]. A membrane glycoprotein, termed P-glycoprotein, has been shown to be responsible for cross-resistance to a broad range of structurally and functionally distinct cytotoxic agents. This glycoprotein, encoded in humans by the mdr1 gene, functions as an energy-dependent efflux pump and removes cytotoxic agents from the resistant cells. The elucidation of the function of P-glycoprotein [106], other ATP-driven efflux pumps [107], as well as other mechanisms of multidrug resistance [108] has had a major impact on the understanding of multidrug resistance in human tumors. The exclusion of the polymer–drug conjugate from the cytoplasmof the cell, through intracellular trafficking in membrane-limited organelles, renders the efflux pumps ineffective. Furthermore, subcellular trafficking along the endocytic pathway from the plasmamembrane to the perinuclear region changes the gradient of distribution of drugs inside cells [109,110]. The concentration gradient of free drugs is directed from the plasma membrane to the perinuclear region, in contrast to polymer-bound drugs, which have a gradient in exactly the opposite direction. The drug, released from the polymeric carrier in the lysosomal compartment, enters the cytoplasm in the perinuclear region. Consequently, the probability of its interaction with nuclear DNA and/or topoisomerase II is higher than the probability of its recognition by the P-glycoprotein efflux pump [109].

We hypothesized that HPMA copolymer-bound DOX [P(GFLG)- DOX] (P is the HPMA copolymer backbone) would behave differently than free DOX during long term incubation with cancer cells. To verify the hypothesis, we have studied the effect of free DOX and P(GFLG)- DOX on the induction of multidrug resistance and changes in metabolism in human ovarian carcinoma A2780 cells during repeated cyclic (chronic) exposure [111]. Such experiments are of therapeutic relevance. The development of multidrug resistance during adaptation of sensitive human ovarian carcinoma A2780 cells to free DOX and P(GFLG)-DOX was analyzed. Adaptation of sensitive A2780 cells to repeated action of free DOX augmented cellular resistance to DOX and finally led to the over-expression of the MDR1 gene. On the other hand, P(GFLG)-DOX induced neither the multidrug resistance with or without MDR1 gene expression, nor the adaptation of the sensitive A2780 cells to free DOX [111].

The in vivo comparison of the efficiency of DOX and P(GFLG)-DOX in solid tumor mice models of DOX sensitive (A2780) and resistant (A2780/AD) human ovarian carcinoma [89] demonstrated the advantages of polymer conjugation. Free DOX was effective only in sensitive tumors decreasing the tumor size about three times, while P(GFLG)-DOX decreased the tumor size 28 and 18 times in the sensitive and resistant tumors, respectively (Fig. 8). An enhanced accumulation of P(GFLG)- DOX in the tumor was observed, whereas only low concentrations of DOX were detected in other organs (brain, liver, kidney, lung, spleen, and heart) following P(GFLG)-DOX administration [89].

3.3. Combination therapy with polymer-bound drugs

Photodynamic therapy is a newer paradigm in anticancer therapy that involves activation of specific compounds called photosensitizers with specific wavelengths of light to induce cell death. Illumination of these compounds results in the generation of free radicals and singlet oxygen, which cause cell damage and ultimately cell death. A combination of chemotherapy and photodynamic therapy may result in a synergistic response, resulting in a better cure rate than monotherapy. On two cancer models, Neuro 2A neuroblastoma [112] and human ovarian carcinoma heterotransplanted in nude mice [113], we have shown that combination therapy with HPMA copolymer–anticancer drug [DOX and meso chlorin e₆ mono(N-2-aminoethylamide) {Mce₆}] conjugates showed tumor cures that could not be obtained with either chemotherapy or photodynamic therapy alone. Cooperativity of the action of both drugs contributed to the observed effect [114]. Based on biodistribution data [115], we hypothesized that combination therapy of s.c. human ovarian carcinoma OVCAR-3 xenografts in nude mice using multiple doses/irradiation of P(GFLG)–Mce₆ and P(GFLG)–DOX may acquire low effective doses without sacrificing the therapeutic efficacy. Indeed, 10 out of 12 tumors exhibited complete responses in the group of mice receiving multiple photodynamic therapy (PDT) plus multiple chemotherapy [116].

Finally, we have demonstrated the advantages of targeted combination chemotherapy and photodynamic therapy using OV-TL16- targeted HPMA copolymer–DOX and HPMA copolymer–mesochlorin e₆ conjugates. OV-TL16 antibodies are complementary to the OA-3 antigen (CD47) present on the majority of ovarian cancers. The immunoconjugates (Fig. 9) preferentially accumulated in human ovarian carcinoma OVCAR-3 xenografts in nude mice with a concomitant increase in therapeutic efficacy when compared with non-targeted conjugates [83]. The targeted conjugates suppressed tumor growth for the entire length of the experiment (>60 days; unpublished data).

The combination index (CI) analysis was used to quantify the synergism, antagonism, and additive effects of binary combinations of free and HPMA copolymer-bound anticancer drugs, 2,5-bis(5-hydroxymethyl- 2-thienyl)furan (SOS), DOX, and mesochlorin e₆ mono-ethylenediamine (Mce₆) in anticancer effect toward human renal carcinoma A498 cells. The combination of SOS+DOX proved to be synergistic over all cell growth inhibition levels. All other combinations exhibited synergism in a wide range of drug effect levels [117]. Similarly, the targeted (using Fab′ of OV-TL16 antibody) and nontargeted targeted HPMA copolymer–drug conjugates, P(GFLG)–Mce₆ and P(GFLG)–SOS, were evaluated against human ovarian carcinoma OVCAR-3 cells. The observations that most combinations produced synergistic effects will be important for clinical translation [118].

In collaboration with Satchi-Fainaro’s laboratory at the University of Tel Aviv a new therapeutic strategy for bone neoplasms using combined targeted polymer-bound angiogenesis inhibitors was developed [119]. The aminobisphosphonate alendronate (ALN), and the potent anti-angiogenic agent TNP-470 were conjugated with HPMA copolymer. Using reversible addition–fragmentation chain transfer (RAFT) polymerization, we synthesized a HPMA copolymer–ALN-TNP-470 conjugate bearing a cathepsin K-cleavable linker, a protease overexpressed in bone tissues. Free and conjugated ALNTNP- 470 demonstrated their synergistic anti-angiogenic and antitumor activity by inhibiting proliferation, migration and capillary-like tube formation of endothelial and osteosarcoma cells. The bi-specific HPMA copolymer conjugate reduced vascular hyperpermeability and remarkably inhibited human osteosarcoma growth in mice by 96%. These findings indicate that HPMA copolymer–ALN-TNP-470 is the first narrowly dispersed anti-angiogenic conjugate synthesized by RAFT polymerization that targets both the tumor epithelial and endothelial compartments warranting its use on osteosarcomas and bone metastases (Fig. 10) [119].

3.4. Novel targeting strategies

As discussed in 3.1, HPMA copolymer–drug conjugates accumulate passively in solid tumors as a result of the (molecular weight dependent) enhanced permeation and retention (EPR) effect [85]. Active targeting of HPMA copolymer–drug conjugates can be achieved with the incorporation of cancer cell-specific ligands, such as carbohydrates, lectins, antibodies, antibody fragments, and peptides, resulting in enhanced uptake of conjugates by cancer cells through receptor-mediated endocytosis with concomitant improvement of therapeutic efficacy [120,121].

Among different cancer targeting molecules, peptides are of particular interest. Enhanced peptide targeting efficiency can be achieved through multivalent interactions [122] between targets and HPMA copolymer–peptide conjugates containing multiple copies of peptides within a single polymer chain (Fig. 11) [123].

Combinatorial approaches, such as phage display or synthetic peptide libraries, are suitable for the identification of targeting peptides. Overexpression of the CD21 receptor was found on lymphoblastoid cell lines such as Raji cells; consequently, we have used these techniques to identify targeting moieties for lymphomas [124,125]. With phage display, five distinctive peptides (RMWPSSTVNLSAGRR, PNLDFSPTCSFRFGC, GRVPSMFGGHFFFSR, RLAYWCFSGLFLLVC, and PVAAVSFVPYLVKTY) were identified as ligands of CD21 receptor. The dissociation constants of selected peptides were determined to be in the micromolar range [124]. Using a synthetic chemical combinatorial technique, one-bead one-compound (OBOC) method, we identified four heptapeptides (YILIHRN, PTLDPLP, LVLLTRE, and IVFLLVQ) as ligands for the CD21 receptor [125]. The dissociation constants were found to be similar to peptides selected by phage display. Importantly, the peptides retained their biorecognizability towards CD21 receptor after they were conjugated to HPMA copolymers and demonstrated a multivalency effect [125]. Several peptide-targeted HPMA copolymer– drug conjugates displayed anticancer activity [123,126,127]. The combinatorial chemistry approach (OBOC), when combined with a high-stringency screening method, is able to identify peptides with a picomolar affinity [128,129].

3.5. Subcellular fate and targeting

As discussed above, macromolecular therapeutics are internalized by endocytosis with an ultimate location in the lysosomes. Numerous conjugates were synthesized and evaluated based on this biological rationale. Recently, however, research has been focusing on the identification of different routes of cell entry with the aim to deliver drugs into subcellular compartments different from lysosomes [142– 146]. This direction was mainly driven by attempts to deliver genes or oligonucleotides, i.e., compounds, which may degrade in the lysosomes; however, other rationales may be important: a) the activity of many drugs depends on their subcellular location; and b) the mechanism of action of polymer-bound drugs may be different than that of the free drug. Consequently, manipulation of the subcellular fate of macromolecular therapeutics may result in more effective conjugates.

3.6. Mechanism of action and signaling

The hypothesis that free and polymer-bound drugs activate different signaling pathways is based on their internalization mechanisms. Free drug may initiate signaling pathways by interaction with membrane-bound proteins. In contrast, a polymer-bound drug may be hidden in the hydrophilic corona of the random polymer coil and prevented from interaction with membrane proteins during internalization. It may interact with proteins and DNA after being released from the carrier in the lysosomes and translocated into the cytoplasm [87,91]. We shall demonstrate results supporting this hypothesis using two examples: HPMA copolymer conjugates with geldanamycin and DOX.

3.7. Cancer: clinical trials

HPMA copolymer-based macromolecular therapeutics have been developed considerably in the last 20 years – numerous conjugates have entered clinical trials for therapeutic validation in the last decade. These include HPMA copolymer–DOX [163–165], HPMA copolymer–DOX–galactosamine [166], HPMA copolymer–camptothecin [167], HPMA copolymer–paclitaxel [168], and HPMA copolymer–platinates [169]. Results from testing of some of these conjugates are promising; hopefully the FDA approval of a first macromolecular therapeutics will occur soon. In Section 4.1 we summarized our ideas on the design principles of second-generation conjugates with enhanced therapeutic potential.

3.8. HPMA copolymer conjugates in the treatment of non-cancerous diseases

HPMA copolymer–drug conjugates may be used also for the treatment of diseases other than cancer. We designed bone-targeted HPMA copolymer-conjugated with a well-established bone anabolic agent (prostaglandin E₁; PGE₁) for the treatment of osteoporosis and other musculoskeletal diseases [50,170–175]. The biorecognition of the conjugates by the skeleton was mediated by an octapeptide of D-aspartic acid (D-Asp₈) or alendronate [170,172].

This system has the potential to deliver the bone anabolic agent, PGE₁, specifically to the hard tissues after systemic administration. Once bound to bone, the PGE₁ will be preferentially released at the sites of higher turnover rate (greater osteoclasts activity) via cathepsin K (osteoclast specific) catalyzed hydrolysis of a specific peptide spacer and subsequent 1,6-elimination [50,176]. When given in anabolic dosing range, the released PGE₁ will activate corresponding EP receptors on bone cells surface to achieve net bone formation. The main features of the design are HPMA copolymer backbone containing cathepsin K-cleavable oligopeptide side-chains (Gly–Gly–Pro–Nle) terminating in either D-Asp₈ or in p-aminoben-zyloxycarbonyl- 1-prostaglandin E₁, a PGE₁ prodrug (Fig. 17A).

This novel delivery system has several distinct advantages. First of all, it is a double-targeted delivery system, which contains a bone-binding moiety (D-Asp₈) and a cathepsin K (osteoclast specific enzyme) specific releasing mechanism. By directing PGE₁ specifically to the skeleton, the side effects of systemic administration of the drug would be greatly reduced. Secondly, E-series prostaglandins (PGEs) are powerful anabolic agents in bone, and this delivery system will better target these molecules to sites in the skeleton with a high turnover rate, where new bone formation would be more beneficial. Thirdly, the system permits improved control of drug concentration at the target (bone) site after systemic administration [171,173]. Fourthly, the polymeric carrier can be eliminated from hard tissues and, subsequently, cleared from the body via kidney glomerular filtration. It also offers proper protection of the conjugated PGE₁ from metabolism before it reaches bone tissue. Most recently, we observed the preferential deposition of the proposed delivery system to the bone resorption sites in ovariectomized rats; this strongly supports our higher turnover sites/drug-release hypothesis [172]. In vivo experiments on ovariectomized rats have proven the concept. Following a single i.v. administration of the HPMA copolymer–Asp₈–PGE₁ conjugate to aged, ovariectomized rats, bone formation rates were substantially greater than controls when measured 28 days later (Fig. 17B) [173].

HPMA copolymer conjugates have been also successful in the treatment of rheumatoid arthritis [175,177].

3.9. Methods of synthesis of HPMA copolymer conjugates

The topic of conjugate synthesis will be extensively covered throughout this volume. We have demonstrated several approaches through this chapter. Consequently, we shall just mention some of the important points, which were not covered.

4.1. Modification with semitelechelic poly(HPMA)

Semitelechelic (ST) polymers are linear macromolecules having a reactive functional group at one end of the polymer chain [186]. The single functional group provides the opportunity to conjugate or graft the macromolecule to other species or surfaces. Davis and coworkers have shown that attachment of ST poly(ethylene glycol) (PEG) to therapeutical proteins results in an increase of their resistance to proteolysis, reduction of their antigenicity, and prolongation of intravascular half-life [187]. The extent of protein property changes depends on the degree of PEG substitution and PEG molecular weight [188]. Modification of biomedical surfaces with PEG reduces their biorecognizability and prevents protein adsorption [186].

Poly(HPMA) [189] exhibits similar properties as poly(ethylene glycol) [190] when used for modification of enzymes or vesicular carriers. The first report on semitelechelic poly(HPMA) (ST-PHPMA) was published in 1995 [191]. Modification of nanospheres, based on a copolymer of methyl methacrylate, maleic anhydride, and methacrylic acid, with ST-PHPMA resulted in decreased protein adsorption in vitro and increased intravascular half-life, as well as decreased accumulation in the liver, after intravenous administration into rats. The higher the molecular weight of ST-PHPMA, the more pronounced the changes in these properties (Fig. 22). These data seem to indicate the influence of the hydrodynamic thickness of the coating layer on the process of opsonization and capture by Kupffer cells of the liver and macrophages of the spleen [191].

Similarly, carboxyl and amino group modification of chymotrypsin with ST-PHPMA-CONHNH₂ and ST-PHPMA-COOSu (N-hydroxysuccinimide ester) produced conjugates [189] with comparable properties to PEG-modified chymotrypsin [187]. Ulbrich et al. used ST-PHPMA to modify ribonuclease, chymotrypsin [192], and superoxide dismutase [193]. The proteolytic stability of modified protein increased and their immunogenicity decreased [192,193].

4.2. Modification with HPMA copolymers containing multiple reactive side-chains

First protein modified with HPMA copolymers was prepared by the reaction of the copolymer of HPMA with N-methacryloylglycylglycine p-nitrophenyl ester with insulin [28]. The unreacted protein was separated on Sephadex 75; the HPMA copolymer–insulin conjugate exhibited a slower onset and a slight prolongation of hypoglycemic effect in rats when compared to free insulin [28]. Chymotrypsin [29,30] and cobra venom acetylcholinesterase [194] followed.

To obtain a better insight into the steric hindrance of the polymer chains on the formation of enzyme–substrate complex, we have studied the hydrolysis of polymeric substrates (HPMA copolymers with oligopeptide side-chains terminated in p-nitroanilide) catalyzed by HPMA copolymer-bound chymotrypsin. The kinetic analysis showed that the hydrolysis of polymer substrates with polymer-bound chymotrypsin led to a decrease in both k_cat and k_cat/K_M, but the relationship between the individual substrates remained intact. Apparently, the steric effects of two independent polymer chains (one bound to substrate, the other to enzyme) were roughly additive [30].

Different chemistry was used for the modification of cobra venom acetylcholinesterase [194]. The secondary OH groups of poly(HPMA) (Mw 25–30 kDa) were activated with 4-nitrophenyl chloroformate in dimethylformamide followed by attachment of acetylcholinesterase in borate buffer. The poly(HPMA)-modified acetylcholinesterase demonstrated a 70-fold prolongation of enzyme activity in blood after intravenous injection into mice when compared to unmodified enzyme. The thermoinactivation rate of the polyHPMA–acetylcholinesterase conjugate was 74 times smaller than that of native enzyme (Fig. 23) [194].

HPMA copolymer with N-methacryloylglycylphenylanylleucylglycine p-nitrophenylester was used for the modification of ribonuclease [195] and superoxide dismutase [193]. No difference in biological activity of conjugates prepared using ST-PHPMA and HPMA copolymers with reactive side-chains was detected [193].

There is considerable activity in using HPMA copolymers with reactive side-chains to stabilize complexes of DNA with viruses. This research is covered in the chapter by Seymour in this volume and in our review [196].

5.1. HPMA-based hydrogels

First HPMA-based hydrogels were synthesized by crosslinking copolymerization of HPMA and methylene-bis-acrylamide or ethylene- bis-methacrylamide in the early 70s [197]. The kinetic course of copolymerization, interaction parameter polymer–water, the modulus of elasticity, and concentration of elastically effective chains were characterized.

Degradable hydrogels containing oligopeptide crosslinks susceptible to chymotrypsin-catalyzed hydrolysis were synthesized by crosslinking HPMA copolymers containing reactive side-chains (terminated in p-nitrophenoxy groups) with oligopeptide (GGY, GFY, GLF, AGVY, and AGFY)-containing diamines [49]. The degradability of hydrogels was dependent on the length and detailed structure of the oligopeptide sequence and on the network density, and thus the equilibrium degree of swelling (Fig. 24). The higher the degree of swelling, the faster the rate of degradation. The degree of swelling also has an impact on surface vs. bulk degradation of the hydrogel. If the enzyme cannot diffuse into the hydrogel interior, only surface degradation takes place. This was demonstrated by a comparison of the degradation of HPMA copolymer crosslinked with AGVY-containing sequences catalyzed by chymotrypsin and HPMA copolymer-bound chymotrypsin. Only surface degradation of the hydrogel was observed in the latter case due to the larger size of polymer-modified enzyme [49]. HPMA-based hydrogels containing the sequence GFYAA in the crosslinks were cleavable with cathepsin B, a lysosomal thiol proteinase [44]. In further experiments, HPMA-based hydrogels with degradable crosslinks were shown to release FITC-dextran and daunomycin during incubation with a mixture of lysosomal enzymes (Tritosomes) or chymotrypsin [198].

Crosslinking copolymerization of HPMA with N,O-dimethacryloyl-hydroxylamine produced hydrolytically degradable hydrogels [199]. These hydrogels were used as a depo for anticancer drug (DOX); results of combination therapy using DOX release from hydrogels followed by antibody-targeted therapy has been effective in the treatment of terminal stages of Bcl1 leukemia in mice [200].

Interpenetrating network (IPN) hydrogels, composed of a pH-sensitive, aromatic azo group-containing network as one component (A), and a hydrolyzable network as the other (B), were designed to refine their swelling profile in the gastrointestinal tract (A is a copolymer of N,N-dimethylacrylamide, acrylic acid, N-tert-butylacrylamide, and N-methacryloylglycylglycine p-nitrophenyl ester crosslinked with an aromatic azo group-containing diamine; B is a copolymer of HPMA with N,O-dimethacryloylhydroxylamine) [201].

Hennink et al. synthesized ABA triblock copolymers, where block A is thermosensitive poly(HPMA lactate) and block B is PEG. Heating of the copolymer results in the formation of a viscoelastic material, which may be stabilized (crosslinked) by photopolymerization [202]. These materials are suitable for protein delivery [203].

5.2. HPMA-based hybrid hydrogels

Traditional synthetic methods, such as crosslinking copolymerization, crosslinking of polymeric precursors, and polymer–polymer reactions, have produced numerous hydrogel materials with excellent properties [9,204,205]. However, these approaches do not permit a precise control of chain length, sequence, and 3D structure. On the contrary, self-assembly of (macro)molecules into physical hydrogels has the potential to produce precisely defined, hierarchical 3D structures [206–209]. Self-assembly is highly specific and frequently found in biology both in the peptide/protein and DNA worlds.

Hybrid hydrogels are usually referred to as hydrogel systems that possess components from at least two distinct classes of molecules, for example, synthetic polymers and biological macromolecules, interconnected either covalently or non-covalently [210]. Conjugation of peptide domains and synthetic polymers may lead to novel materials with properties superior to those of individual components [206–218].

6.1. Design of biocompatible long-circulating conjugates for passive and active targeting

The extravasation of polymer–drug conjugates into a solid tumor depends on the concentration gradient and time of circulation. The renal threshold limited the molecular weight of the first generation of polymeric carriers to well below 40 kDa; this lowered the retention time of the conjugate in the circulation with concomitant decrease in pharmaceutical efficiency. The design, synthesis, and evaluation of new second-generation anticancer nanomedicines based on high molecular-weight polymeric drug carriers containing degradable bonds in the main chain (polymer backbone) are needed. Recent advances in living radical polymerization [182,183] and in bioconjugate chemistry [220] provide a solid scientific basis for further developments.

6.2. Targetability and multivalency

The targetability of HPMA copolymer conjugates was demonstrated using numerous targeting moieties in vitro, in vivo, and in clinical trials. The synthesis of high molecular weight conjugates will permit binding of several targeting moieties per macromolecule, enhancing the avidity of the conjugate [221]; concomitantly, the efficacy should increase.

6.3. Conformation of multifunctional conjugates

Polymeric drug carriers containing multiple components, such as targeting modules, drug releasing modules, and endosome disruptive modules, have showed the potential to perform multiple functions within a single structure. Ideally, each component within the delivery system should function independently, without affecting the functionality of the other components. However, the physical and biological properties of multifunctional conjugates can be expected to exert some influence on the other components [48,50,78]. For example, poly(HPMA) assumes a random coil conformation in aqueous solutions [23]. Incorporation of hydrophobic side-chains into HPMA copolymers may result in intra- and/or intermolecular association between the hydrophobic moieties. Therefore, awareness of the complexities caused by the introduction of each component (like polymer self-association) is needed to design multi-component drug conjugates. For example, intra- or intermolecular hydrophobic interactions of a targeting peptide [78] or a hydrophobic drug [48,50] can result in the formation of unimolecular micelles with concomitant decrease in the rate of (enzymatically catalyzed) drug release [78]. Combination of physicochemical methods, such as FRET (fluorescence energy transfer) [78], light scattering [76,77], determination (for photosensitizers) of the quantum yield of singlet oxygen formation [76,77], and determination of biological activity can guide the optimal design of particular conjugates. The preference for either inter-polymer association or intra-polymer association is intrinsically determined by polymer concentration and structural parameters such as the nature and the size of the polymer chain, or the type, content, and arrangement of hydrophobes in the polymer. It can also be influenced by extrinsic factors, such as temperature, solvent, ionic strength, and pH [78 and references therein].

6.4. Mechanism of internalization and subcellular trafficking

Studies have shown that multiple endocytic pathways can concurrently take action when a macromolecular drug conjugate is exposed to a single type of cell. Consequently, it is important to study the mechanisms of internalization and subcellular fate when developing a macromolecular therapeutic for a particular disease [180,222]. As shown above, methods to manipulate the subcellular fate have a great potential to enhance the efficiency of macromolecular therapeutics; this line of research should be pursued.

6.5. Further studies on combination therapy

Quantitative in vitro and in vivo evaluation of combination therapy using polymer-bound drugs needs to be undertaken to optimize the design, structure, and dosing of the conjugates. It is important to address the question on the impact of the mechanism of drug action and of pharmacokinetics/pharmacodynamics on the design selection in combination therapy: two drugs per one macromolecule [119] or a combination of two polymer conjugates [112,113,117,118] containing one drug type per macromolecule?

6.6. Dream conjugate

The research summarized above should result in optimized structures of macromolecular therapeutics. An efficient conjugate for the future must have an increased intravascular half-life. Consequently, for the long-circulating conjugate to remain biocompatible and eliminate from the organism, labile bonds, cleavable either enzymatically or pH-sensitive, should be incorporated into the polymer backbone. Two targeting moieties should be attached: one for cellular targeting to direct the conjugate to a subset of cells, the other for subcellular targeting to direct the released drug into a particular organelle. In addition, two biologically active moieties might be attached to the one macromolecule (Fig. 27).

6.7. New design paradigms

Finally, new designs based not on systematic improvements of known concepts, but on very new ideas are needed. For example, inspired by the high level of biorecognition of peptides self-assembling into antiparallel heterodimers (described in Section 5.2.1), we designed drug-free macromolecular therapeutics. Their activity is based on the crosslinking of receptors at the cell surface mediated by coiled-coil formation and concomitant induction of apoptosis [223].