Therapeutic proteins are a growth area offering promise of countering diseases such as Alzheimer’s and Parkinson’s, but protein drug formulation is challenging due to their structural complexity and instability. Susan Birks highlights some recent advances in this field.
The therapeutic proteins market is forecast to grow at a compound annual growth rate of 6.2% between 2010 and 2017, to reach US$141.5bn in 2017, according to market research company GBI Research.1 There are currently hundreds of such proteins under development and many of the major pharmaceutical companies, such as Roche, Amgen and Johnson & Johnson, are expanding into this market. However, despite the progress made in this field over the past decade, overcoming some of the complexities surrounding protein therapies is still a challenge and a major area for research.
Scientists have long sought methods for designing antibodies to combat specific ailments. But the complexity of designing antibodies that attach only to a target molecule of interest has so far prevented them from realising this goal. Some researchers claim to have made advances recently, however, which could help to bring new, more effective therapies to market.
At Rensselaer Polytechnic Institute in Troy, New York, for example, researchers have developed a new peptide-based strategy for the generation of monoclonal antibodies. They say the surprisingly simple process could be used to design antibodies aimed at neutralising the harmful protein particles that lead to Alzheimer’s disease. They also believe the process could be used both as a tool to understand complex disease pathology and for developing new antibody-based drugs in the future.
Antibodies comprise a large Y-shaped protein topped with small peptide loops. The loops bind to harmful invaders in the body, such as a viruses or bacteria. Once an antibody is bound to its target, the immune system sends cells to destroy the invader.
When attempting to design an antibody, the arrangement and sequence of the antibody loops is of utmost importance. Only a very specific combination of antibody loops will bind to and neutralise each target. And with billions of different possible loop arrangements and sequences, it is seemingly impossible to predict which antibody loops will bind to a specific target molecule.
The researchers at Rensselaer, led by assistant Professor of Chemical and Biological Engineering Peter Tessier, used the same molecular interactions that cause Alzheimer’s proteins to stick together and form the toxic particles that are a hallmark of the disease. ‘We are actually exploiting the same protein interactions that cause the disease in the brain to mediate binding of antibodies to toxic Alzheimer’s protein particles,’ he said.
Alzheimer’s disease is due to specific proteins sticking together to form protein particles. These particles then damage the normal, healthy functions of the brain. The formation of similar toxic protein particles is central to diseases such as Parkinson’s and BSE.
Crucially, the new Alzheimer’s antibodies developed by Tessier and his colleagues attach only to the harmful clumped proteins and not to the harmless monomers or single peptides that are not associated with disease.
The research findings are described in a paper in the journal Proceedings of the National Academy of Sciences (PNAS)2 and Tessier and his colleagues see the potential for their technique being used to target and better understand similar types of protein particles in disorders such as Parkinson’s disease.
‘By binding to specific portions of the toxic protein, we could test hypotheses about how to prevent or reverse cellular toxicity linked to Alzheimer’s disease,’ Tessier says.
no generic approach
Monoclonal antibody therapy is very successful and has major advantages compared with small molecule therapy. However, due to technical constraints, it is very difficult to generate monoclonal antibodies for certain important drug targets, and so far no generic approach has been documented. According to the company Pepscan Therapeutics, this is particularly valid for G-protein coupled receptors (GPCRs) – the transmembrane receptors that sense molecules outside the cell and activate cellular responses, for which no FDA-approved antibody exists.
Speaking at the plenary session of the IBC’s International Antibody Conference in San Diego in December 2011, Pepscan Therapeutics’ chief development officer Dr Klaus Schwamborn presented the company’s protein mimicry platform for the discovery of therapeutic antibodies. Schwamborn introduced Chemical LInkage of Peptides onto Scaffolds (CLIPS) technology that makes use of synthetic scaffolds to increase the activity and stability of a peptide.3
Schwamborn showed data demonstrating that Pepscan’s synthetic immunogens produced using CLIPS induce potent antibodies against GPCRs and are therefore able to mimic the native receptor. Not only do the resulting antibodies bind strongly to the native receptor, they also show functionality with neutralising activity in different cell based assays.
The company believes the platform provides a systematic and efficient way of making synthetic immunogens that induce antibodies against the native target.
The majority of small peptides (20–30 amino acids) derived from intact proteins lack a well-defined structure in solution. CLIPS can be used to solve this problem by affixing the loose ends of the peptide. In this way, the scaffolded peptide may be able to adopt the same spatial structure as the corresponding sequence in the intact protein.
The CLIPS technology fixes the linear peptides into cyclic structures (‘single-loop’ format) and brings together different parts of a protein binding site (‘double-loop’, ‘triple-loop’ etc. format). It involves the cyclisation of peptides containing two or three thiol functionalities (in most cases cysteines) and a small organic molecule having the corresponding number of reactive benzyl bromide groups.
The reaction can be performed on native cysteines in the peptide sequence, but also on artificially introduced (homo)cysteines at any desired position in the peptide. Hence the structure and dimensions of the CLIPS peptides can be varied at will.
Pepscan has developed a toolbox of more than 70 different types of CLIPS templates, varying mainly in polarity, solubility and thiol-thiol spanning distance. Over the past few years, the company has validated and optimised its technology, enabling it to design and synthesise ‘conformationally stabilised’ peptides with a well-defined 3D-structure, resembling the native functional protein surface.
With these immunogens, monoclonal antibodies can be generated using common antibody technologies, like phage display libraries or hybridoma generation procedures.
How Pepscan’s Chemical LInkage of Peptides onto Scaffolds (CLIPS) technology works. A) Schematic representation of a CLIP reaction; B) Molecular structure of the CLIPS linkage; C) Different CLIPS-based topologies (‘bicycles’ and ‘tricycles’) for mimicry of discontinuous epitopes; D) New ‘double-loop’ CLIPS technology for functional reconstruction of discontinuous protein binding sites
Picture courtesy of Pepscan
Dr Schwamborn says: ‘Our approach has proven to be successful and the fact that we have the possibility to tailor antibodies to certain GPCR domains enables us to engineer antibodies that fulfill certain biological require-ments. The immunogen design concept for mimicking extracellular domains of discontinuous epitopes might be also applicable to other target classes, such as ion channels.’
The technology is highly versatile and very easy to apply. The cyclisation reaction lasts no longer than 30 minutes, runs at room temperature and does not require any sort of catalysis. Moreover, it can be applied under fully aqueous conditions and neutral pH of 7.5–8.0 and is therefore compatible with highly sensitive biological systems, such as bacterial phages as used in PDL-screening. Finally, the reaction can be run at extremely dilute conditions (10–100μm), which promote high yields of cyclic products and avoid polymerisation.
Several other chemical methods are currently available for structural fixation of peptides into stable secondary structures. However, the company says all of these are restricted to single loop peptides, are relatively complex or not compatible with all amino acids.
improving the fit
Designing proteins for specific functions often relies on grafting functional groups onto existing protein scaffolds but backbone remodelling, which might allow more complex grafting, has so far been limited because it is a computational challenge, says a collaborative group of researchers working together on both sides of the Atlantic.
Several US-based research institutes, including the Fred Hutchinson Cancer Research Center, National Institute of Allergy and Infectious Diseases, University of Washington, Seattle and the Scripps Research Institute, along with the Computational Biology department at the Instituto Gulbenkian de Ciência, Oeiras, Portugal, have published a paper that outlines the use of a hybrid computational-experimental method for grafting the backbone and side chains of functional motifs (groups) onto scaffolds.4
The researchers were able to integrate computational design with experimental selection for grafting the backbone and side chains of a two-segment HIV gp120 epitope, targeted by the cross-neutralising antibody b12, onto an unrelated scaffold protein. The final scaffolds bound b12 with high specificity and with affinity similar to that of gp120, and crystallographic analysis of a scaffold bound to b12 revealed high structural mimicry of the gp120-b12 complex structure.
The researchers believe this method can be generalised to design other functional proteins through backbone grafting.
One of the other challenges of using proteins is that fact that the native proteins themselves are often unstable in physiological conditions, reducing bioavailability and therefore necessitating a greater dose. Current methods used to increase stability, can often have a detrimental effect on bioactivity.
no compromise
Now researchers in the US have discovered a new way to stabilise and protect protein molecules without affecting the protein’s biological activity. They believe this chemistry opens up a new avenue for the development of protein therapeutics by avoiding the need to compromise between stability and affinity.
Currently, the most effective way to stabilise proteins is to attach PEG to the protein core. The hydrophilic nature of PEG attracts water molecules to form a cushion around the protein and give it extra bulk. This increases the effective size of the molecule, making it more difficult to be filtered from the blood by the kidneys. It also physically prevents attack by enzymes or the immune systems through steric hindrance.
However, the downside is that the biological activity of the molecule can be significantly decreased because PEG also has hydrophobic characteristics. Proteins bind their targets through hydrophobic interactions, so the presence of hydrophobic portions on the attached PEG molecules can interfere with this process.
Now, Andrew Keefe and Shaoyi Jiang, of the University of Washington in Seattle, have shown that a different type of protecting molecule, a zwitterionic polymer, can stabilise a protein without impinging on its activity.5
The researchers attached molecules of polycarboxybetaine (pCB) to the enzyme alpha-chymotrypsin. pCB is a zwitterion – it has both positive and negative charges but is overall neutral. When pCB is attached to the protein, the particular pattern and distribution of charge cause a specific ionic environment around the protein. This in turn – for reasons that are not entirely understood – affects the distribution of water molecules around the protein molecule.
‘As with PEG, the presence of pCB provides protection through steric hindrance,’ says Keefe. In addition, the ionic interactions are such that water molecules are pulled further away from the hydrophobic regions of the protein. One effect is to stabilise the hydrophobic core of the protein molecule, making it less fragile. Another effect is to increase the affinity of the hydrophobic binding sites of the protein to the target.
The team found that both PEG and pCB increased the enzyme’s stability in the face of heat and urea. However, PEG reduced the enzyme’s affinity for its substrate, while pCB actually increased it. ‘We think that using these types of conjugates might have advantages, and we are now looking to investigate their use with therapeutic proteins,’ says Keefe.
The biological activity of most recombinant proteins emanates specifically from their three- dimensional structure, which needs to remain unaltered throughout the shelf-life of the product. However, cleavage or aggregation incidents may not only reduce efficacy but also produce adverse immunologic effects.
This means that the current industry standard for the formulation development of biopharmaceutical drugs is a time consuming, trial-and-error driven process that requires accelerated stability testing to identify the best out of many formulation alternatives.
Occasionally, the large number of formulation options is tested by means of high throughput screening, but this has the potential drawback of being a highly artificial testing environment.
optimising formulations
Now ProJect Pharmaceutics, a company based in Martinsried, Munich, Germany, that designs optimised pharmaceutical formulations and delivery systems, says the Predictive Formulation Analytics that it is using can optimise protein formulations and reduce the need for extensive stability testing.
According to the company, Predictive Formulation Analytics involves applying state-of-the-art analytical methods to characterise the physicochemical state of proteins and then analysing their response to certain excipients, enabling it to quickly and reliably identify promising formulation candidates.
By using Predictive Formulation Analytics ProJect Pharmaceutics can reduce the testing regimes needed to ensure new proteins are stable
www.project-pharmaceutics.com
The company says that the stability of proteins in solution is mainly determined by intramolecular and intermolecular interactions. Intramolecular stability is characterised by protein thermodynamics, which are measured by means of nano differential scanning calorimetry (nanoDSC). Intermolecular stability is represented by the attractive or repulsive interaction between protein molecules, which is quantified through composition gradient static light scattering.
A systematic algorithm based on design of experiments (DoE) is used to determine the most favourable composition for the native structure of a given protein with regard to its intra- and intermolecular physicochemical properties. Beneficial effects resulting from pH value, ionic strength, ion types and stabilising agents can be identified and quantified without stability testing.
The company says: ‘Data obtained from our research provide the basis for a drug product composition that is tailored to the protein, its packaging system and its application.’ A final stability test programme, carried out on samples filled into the final packaging system under genuine pharmaceutical manufacturing conditions, then serves to confirm the suitability and stability of the composition and provides trustworthy data for initiating clinical trials.
References
1. Therapeutic Proteins Market to 2017 – High Demand for Monoclonal Antibodies will Drive the Market. Market report GBIHC080MR, published Sep-2011 by GBI Research, www.gbiresearch.com
2. Peter Tessier. ‘Structure-based design of conformation – and sequence-specific antibodies against amyloid b,’ 5.12. 2011, Early Edition journal Proceedings of the National Academy of Sciences (PNAS)
3. Klaus Schwamborn, ‘CLIPS meets GPCRs’ IBC’s International Antibody Conference, San Diego, December 6, 2011
4. Mihai L. Azoitei et al, Computation-Guided Backbone Grafting of a Discontinuous Motif onto a Protein Scaffold, Science 21 October 2011: Vol. 334 No. 6054 pp373–376 DOI: 10.1126/science.1209368
5. Andrew J. Keefe and Shaoyi Jiang, Nature Chemistry 4, 59–63 (2012) doi:10.1038/nchem.1213. Published online 11 December 2011