P. Byrant, K. Powell, Y. Cong
The London Bioscience Innovation Centre,
2 Royal College Street, London NW1 0TU
PolyTherics Ltd and The School of Pharmacy
Corresponding author: Dr Yuehua Cong
Recombinant therapeutic proteins are an important class of medicines that are used to treat a wide range of conditions. Many types of proteins have been approved for clinical use including, enzymes, hormones, cytokines, blood factors, antibody fragments and antibodies. The continued development of protein-based medicines is often hampered by problems related to their short half-life in the blood, which must be addressed to transform them into viable therapeutics.
Rapidly cleared proteins must be dosed more frequently. While this poses an inconvenience to the patient and added costs for treatment, a suboptimal dosing frequency also increases the risk for immunogenicity and toxicity. Efficacy is also decreased when proteins clear rapidly because there is a sub-therapeutic concentration between doses. By impacting safety, efficacy, patient compliance and costs, it is clear that optimised pharmacokinetics are a fundamental requirement for the development of protein-based medicines.
Three general approaches have been followed to prolong the circulation half-life of proteins: (i) increase the size of the protein molecule in circulation, (ii) exploit recycling mechanisms and (iii) utilise slow release systems to prolong dosing into the blood stream (e.g. pumps). Of these strategies, the covalent conjugation of poly(ethylene glycol) (PEG) to a protein has been clinically proven to be the most general approach. There are at least 10 PEGylated products currently used in the clinic with a total market value of over $8 billion per annum. No other strategy has emerged that is widely applicable for developing protein-based medicines. While several PEGylated proteins have now been granted marketing approval (see Table 1), other classes of molecule can also be PEGylated. Macugen® is a PEGylated oligonucleotide that is used in the treatment of age related macular degeneration. There are also approximately 20 PEGylated products are undergoing clinical trials.
PEGylation is the chemical attachment of poly(ethylene glycol) (PEG) to a protein, peptide or another candidate molecule. It is a well-established approach for enhancing therapeutic efficacy. PEG is an FDA approved excipient that has been widely used in foods, cosmetics and pharmaceuticals including injectable, topical, rectal and nasal formulations. PEGylation has been clinically proven to be a successful method for enhancing pharmacokinetic, pharmacodynamic, and immunological profile of therapeutic proteins.1,2
In addition to the limitations posed by their rapid clearance, protein-based medicines are complex molecules prone to aggregation, which must be avoided since aggregation increases the risks for immunogenicity. Protein aggregation also results in the loss of bioactivity with the consequence that dose reproducibility can be difficult to achieve. Since protein-based medicines are administered parenterally, formulation options are limited. Often proteins are reconstituted from lyopholised solids to minimise aggregation during transit and storage. A highly trained healthcare professional is often necessary to ensure that reconstitution and administration are conducted safely.
Utilising final dosage forms that can be administered without the need of reconstitution is a much-preferred option. The key limitation however is to ensure the protein is stable and not prone to aggregation in its final dosage form. PEGylation decreases the susceptibility for proteins to aggregate in solution. PEGylated proteins can be formulated more simply in liquid dosage forms resulting in a greater opportunity to utilise more user-friendly dosage forms. The use of pumps, colloids and other depot systems is generally not applicable for proteins due to their propensity to aggregate.
Together the requirements for optimised pharmacokinetics and stable formulations have exerted significant challenges that must be addressed for developing protein-based medicines. PEGylation is a clinically proven strategy to address these two fundamental requirements.
Technologies to PEGylate biopharmaceuticals
The study of the chemical conjugation of PEG molecules to biologics started from the late 1970s. The first generation PEGylation technologies quite often target the ε amino groups of lysine or the N-terminal amino acid group on the polypeptide mainchain. This often causes modification of multiple lysines (i.e. “decorates the surface of the protein”) and yields a difficult to reproduce mixture of PEG positional isomers of the protein that have different physiochemical, biological and pharmaceutical properties.
Linear PEG molecules with a molecular mass less than 12kDa were often applied in the first generation PEGylation methods. Adagen® (approved in 1990), Somavert® (approved in 2002) and Oncaspar® (approved in 2006) have been produced in this way. They all have several (four to six) 5kDa PEG molecules attached to lysine residues. Most recently Krystexxa™ has gained initial US approval for the treatment of chronic gout. Approximately 36 PEG (10kDa) molecules were conjugated to the lysine residues of the tetrameric uricase, which is a porcine derived protein. These non-endogenous proteins products required hyper-PEGylation. Without PEGylation, these types of protein cannot be used effectively as medicines. The multiple PEG chains bound to the protein are intended to shield potentially immunogenic protein epitopes and to block exposure of the uricase core to proteolytic enzymes.
Second generation PEGylation processes use PEG molecules of higher molecular weight with endogenous proteins (e.g. interferon, erythropoietin). These products are mono-PEGylated meaning there is one PEG molecule conjugated to each protein molecules. The PEG is still conjugated to different residues on the protein however. Hence these second generation products are also mixtures with PEG conjugated at different sites on the protein (i.e. PEG positional isomers).
These mixtures must be completely characterised during development. During production, such mixtures must maintain tight specification profiles, which we believe is difficult to achieve and which adds significant costs to the final product. So-called branched PEGs have also been used for these second-generation PEGylated products. These PEG molecules are essentially two separate PEGs bound to an amino acid (e.g. lysine), which is then conjugated to a site on the protein. The branched PEGs offer a practical way to use higher molar masses of PEG.
In 2001 PegIntron® (PEGylated interferon alfa-2b) became available as weekly treatment for hepatitis C. The principle PEGylation site (about 50%) is His34 although lysine and N terminus PEGylation also occur. Pegasys® (PEGylated interferon alfa-2a, approved in 2002) uses a branched PEG comprised of two 20 kDa PEG molecules linked through a lysine that is conjugated to interferon. This results in a 40kDa PEG which is then chemically linked to one of several lysine residues of interferon alfa-2a via an amide bond. Mircera® (approved in 2007 Europe) integrates a 30kDa PEG through an amide bond to the N-terminal amino group or the ε-amino group of lysine. Neulasta®, PEGylated G-CSF was approved in 2002. A 20kDa PEG is predominantly covalently conjugated to the N-terminal methionyl residue.
From these products it is clear that PEGylation is safe and has made them into better medicines than for the non-PEGylated protein. However there remains a need to produce more homogeneous products. Also there is a need to increase the efficiency of PEGylation. All the first and second generation products are derived from PEG having been conjugated predominantly to lysine residues on the protein surface. Such conjugation processes while yielding mixtures are also not very efficient. Third generation PEGylation strategies are therefore seeking to site-specifically PEGylate the protein utilising more efficient processes. Cost is a major issue to overcome in order to provide affordable pharmaceuticals for patients. Site-specific PEGylated medicines are therefore needed for more uniform and potentially efficacious protein-based medicines that are more cost-affordable.
Considerable effort to achieve site specific PEGylation is focused on adding a free cysteine to a protein because the thiol on cysteine increases the specificity for PEGylation. Cimzia® is an example of this and was approved in 2008. Cimzia has a 40kDa branched PEG attached to the cysteine residue of anti tumour necrosis factor antibody fragment (Fab’). Unfortunately Cimzia has native disulfide bonds that can be scrambled with the free cysteine complicating the PEGylation process. Since most therapeutic proteins have disulfide bonds, site-specific PEGylation at the natural disulfide bond of protein has also been reported.3 This avoids the need to add a free cysteine to the protein.
Several other strategies for site-specific PEGylation are in development. Several site specific mono-PEGylated proteins are currently undergoing clinical trials. Site-specific PEGylation that is also cost effective is predicted to be the next stage in the evolution for PEGylated medicines.
In vivo properties of PEGylated biopharmaceuticals
PEG is a large molecule that when conjugated to a protein decreases its rate of renal clearance. PEG also ‘sterically shields’ the protein to protect it from proteolytic cleavage while in circulation. This helps to increase the circulation half-life. This steric shielding of PEG also decreases the interaction of the protein with the immune system, thus decreasing immunogenicity of the therapeutic protein. PEGylation dose not however change the function of the protein, so while the steric shielding of PEG can also decrease its initial interaction with its target, once the interaction between the protein and its target occurs, the biological outcome is the same as for the non-PEGylated protein. Remembering that protein-based medicines tend to be highly potent molecules, means that clinically, PEGylate proteins can be much better medicines than the non-PEGylated proteins. The properties of PEGylated biopharmaceuticals are briefly described in more detail below.
Extended half life
Rapid clearance in vivo as a result of glomerular filtration and proteolytic degradation is a common limitation to the efficacy of many therapeutic proteins. PEGylation increases the circulating half-life of proteins by tackling both of these issues.
When PEG is covalently attached to a protein, the resulting conjugate has been found to have a hydrodynamic volume 10 times that of the native protein.4 This dramatic increase raises the overall size of the drug molecule above the kidney filtration threshold resulting in significant extension of blood circulation time. In addition, PEG molecules can mask the protein surface and consequently reduce both possible immunological responses and proteolytic degradation.1, 5 Increased blood residence time is an important benefit of PEGylated biopharmaceuticals leading to reduced dosing frequency as the therapeutic effect is maintained for longer.
Take interferon as an example: PEGylation of interferon-α (IFN-α) extends the circulating half life in man from less than one hour to greater than two days.6 Treatment of hepatitis C with IFN-α induces adverse effects including flu-like symptoms, depression, anxiety and sleep disturbance.7 The enhanced half life of PEGylated IFN-α allows less frequent dosing and hence unpleasant side effects are reduced. Patient compliance with this therapy is greatly increased compared to that with non-PEGylated IFN-α and PEGylated IFN-α2a (Pegasys) or α2b (PegIntron) are now the preferred therapy for hepatitis C in combination with Ribavarin. It is only with the PEGylated products that it has now been possible to cure patients of hepatitis C.
Metabolism and elimination
The PEG used for protein PEGylation tends to be excreted. The metabolism of PEG itself can involve the oxidation of free alcohol groups, which can be available at the ends of the PEG molecule. There is a molecular weight dependence that has been observed for PEG metabolism. As most biopharmaceuticals have been modified with a total PEG mass of 20 kDa or larger, urinary clearance is the major pathway for elimination. A PEGylated protein cannot be compared to a globular protein when related to kidney excretion. Due to recycling mechanisms for some endogenous proteins excretion can vary due to the differences in receptor binding that are observed when different sites on the protein are PEGylated. Generally speaking, renal clearance decreases with increased PEG mass up to a molecular weight threshold and the metabolism of PEGylated pharmaceuticals is a minor route of clearance.8
Treatments with recombinant therapeutic proteins often lead to antigenic effects. The antibodies formed may reduce the therapeutic efficacy of the biotherapeutic molecule. Improving the pharmacokinetic profile of proteins by PEGylation can help to control the immunogenic response as dosing is reduced. PEGylation may also prevent the formation of antibodies by masking immunogenic sites on proteins. Asparaginase is a non-human protein. PEGylated L-asparaginase is produced by hyperconjugation of 5kDa PEG so that the surface of the protein is saturated with PEG molecules. It has been reported that in patients treated with the PEGylated variant, the levels of antibodies formed are significantly lower.9 The Krystexxa™ case also showed the benefit of utilising PEGs to minimise immunogenicity.
PEGylation is often associated with decreased biological activity in in vitro cell based assays. This is a consequence of steric shielding where the PEG molecule partially blocks interactions between the therapeutic protein and its target receptor.10 Hence the on-rate for receptor binding can be decreased. The off rate though can be little changed from the native protein. The site of PEGylation does influence the biological activity, but does not change protein function. Site-specific PEGylation is therefore also needed to optimise activity. Although steric shielding lowers the binding affinity to the receptor, due to the increased circulating half-life, there are many more chances for the receptor-ligand interactions to take place. This results in an extended in vivo pharmacological effect that has been clinically exploited for the different products now being used.
Preclinical toxicology studies have not revealed any PEG-specific toxic findings. PEG is widely used in society. The acute and chronic administration of PEGs of different molecular weights by a range of routes has not led to any clinical signs of toxicity in humans.8
PEGylation is a clinically proven strategy that is both safe and generally applicable for different structural classes of proteins used for the treatment of a wide range of medical conditions (e.g. infection, oncology, inflammation). Although proteins and peptides can be generally PEGylated, other molecules including small molecule drugs, cofactors, lipids and saccharides can also be PEGylated. For example PEGylated liposome (e.g.Doxil®; doxorubicin HCl liposome) and PEG based hydrogel such as FocalSeal® are examples of PEGlyation from a diverse range of applications.11 It is certain that PEGylation will continue to play a role in the development of new peptide and protein-based medicines.
1. Veronese FM. PEGylated protein drugs: basic science and clinical applications. Birkhaeuser Verlag, Basel, 2009.
2. Roberts MJ et al. Adv Drug Deliver Rev 2002;54:459-476.
3. Shaunak S et al. Nat Chem Bio 2006;2:312-313.
4. Fee CJ Biotechnol Bioeng 2007;98:725-731.
5. Abuchowski A et al. Cancer Biochem Biophys 1984;7:175-186.
6. Fishburn CS J Pharm Sci 2008;97:4167-4183.
7. Fried MW Hepatology 2002;36:S237-44.
8. Webster R et al. Drug Metab Dispos 2007;35:9-15.
9. Avramis VI et al. Blood 2002;99:1986-1994.
10. Kubetzko S et al. Mol Pharmacol 2005;68: 1439-1454.
11. Harris JM et al. Nat Rev Drug Discov 2003; 2:214-220.