Binder types: classical, recombinant and alternatives

Binders generated through immunisation, monospecific polyclonals and mAbs, have proved themselves to be excellent reagents for many routine life sciences applications as well as diagnostics and will in many cases be regarded as the first-line choice, being the best established and most widely used. The exceptional value of monospecific polyclonals in immunohistochemistry in the ongoing Human Proteome Atlas project (KTH, UU) is an excellent illustration of what can be achieved in affinity proteomics. Nevertheless, the recombinant molecules generated in vitro will extend by far what is possible with classical antibodies, with enormous potential to open up wholly new areas of research (e.g. in structure determination and novel in vitro and in vivo assays), diagnostics and therapy. It is important that a wide range of potential uses is kept in mind during binder development to maximise the impact of this initiative. Moreover, since the number of reagents required for a thorough analysis of the human proteome could well exceed 100,000, the use of recombinant technologies amenable to high-throughput rather than the classical animal-based immunisation methods will become inevitable. Single-chain antibody variants, e.g. scFv (TUBS, ULUND, BBT), which can be obtained from naive libraries, and single domain nanobodies, displaying highest affinity when selected from immune camelid libraries (VIB), together with ‘alternative’ scaffolds - represented by Designed Ankyrin Repeat Proteins (DARPins) (UZH) and Affibody molecules (KTH) - have major advantages in stability, production, fine-tuning of specificity, intracellular activity, rapid access to encoding DNA for further production or manipulation, and indefinite storage, in the extreme case as a computer file. Most importantly, they will allow whole ranges of new investigations and experimentation, ranging from functional interference in the cell to the facilitation of crystallisation.

(Picture provided by Serge Muyldermans, VIB)

While scFv and scFab are antibody derived and therefore share the same recognition properties, the nanobodies and alternative scaffolds introduce new principles. Current understanding indicates that these recombinant antibody variants and alternative scaffolds recognise different surface architectures of the antigen (e.g. flat, concave, convex). Therefore the selection of various recombinant binders (scFv, scFab, nanobodies, DARPins, Affibody molecules) for the same target should not be considered redundant, since this collection of binders will complement each other in epitope recognintion and together enable coverage of the entire surface of target molecules. In contrast, limiting the binder types to e.g. scFv fragments, with their particular epitope preference, would lead to a number of interesting sites on targets being unrecognised and remaining inaccessible to blocking or investigation of protein partners interacting at that interface. In addition, for sandwich ELISA and proximity ligation assays (below) it is absolutely necessary to possess at least two binders that recognise topographically distinct epitopes on the target protein.

(Picture provided by Serge Muyldermans, VIB)

Nanobodies (Partner VIB) are the single V domains originally derived from the unique heavy chain-only antibodies of camelids (camels, llamas), generally readily obtained after a brief immunisation period, followed by cloning of the V-gene repertoire and phage display pannings. These in vivo affinity matured nanobodies are well expressed, highly stable even in the reducing environment of the cytoplasm and have nM to pM affinity. They interact regularly with the concave surfaces of their antigen, such as the active sites of enzymes. Therefore, nanobodies will be particularly useful in two separate areas, namely as enzyme inhibitors and as intracellularly acting binders. Many nanobodies against enzymes - as a result of their small size and prolate shape with the paratope at the narrow end - are inhibitory (or activating) for the catalytic activity of their target. This may be extremely valuable for the kinases and tyrosine phosphatases, and complements the similar inhibitory properties of other scaffolds, already experimentally shown in the case of DARPins. In contrast to most small organic kinase inhibitors that lack specificity, the affinity-matured nanobodies are highly specific for their cognate targets and will have no (or very limited) cross-reactivity to related kinases or tyrosine phosphatases. Moreover, their intracellular activity has also been established. Nanobodies, as well as DARPins (below), fused to GFP as ‘Chromobodies’ and transfected into mammalian host cells trace the antigen location and traffic within living cells when followed by confocal microscopy. The effect of a post-translational modification could be investigated in real time. Since nanobodies are stable molecules, it appears that 80-90% of them could be functional as intrabodies, even in the reducing environment of the cytoplasm, either redirecting the target to another cellular compartment or leading to a knock down of its normal activity. Intracellular nanobodies will therefore be valuable tools for systematic elucidation of function.

A DARPin with four ankyrin repeat motifs (Picture provided by EBI)

DARPins (Designed Ankyrin Repeat Proteins, Partner UZH) with picomolar affinity and high specificity have been selected against a wide range of protein targets, including extracellular proteins and domains, intracellular proteins and domains, and integral membrane proteins, such as conformational epitopes on GPCRs. Generation of pM binders from synthetic libraries does not require animals at any step and may thus be a scalable technology. Three features make DARPins exciting alternatives for proteomics applications. First, their production yield from bacteria far exceeds that of most other classes of proteins tested, allowing a degree of parallelisation not previously possible, e.g. the production and purification of 96 different binders at the mg scale by one person in a single day is now implemented at Partner UZH. Secondly, the absence of disulphide bonds and their correct folding in the cytoplasm of prokaryotes and eukaryotes alike enables their use in intracellular assays, such as functional interference with pathways of the target protein. Thirdly, their favourable folding properties make fusion with many other proteins very convenient; as e.g. GFP fusions to any DARPin can be made, the genes can be transfected into mammalian hosts and will express and fold intracellularly, enabling spatiotemporal fluorescence studies tracing the antigen location and traffic within living cells when followed by confocal microscopy in real time. Importantly, several DARPins have been shown to act in vitro and in vivo as inhibitors of a variety of enzymes including e.g. several kinases, proteases and transporters, using a variety of mechanisms. In ProtAffin, intracellular action by DARPins will be employed against cell signalling targets. Also, because of their rigidity, DARPins crystallise well in complex with the target and often facilitate structure determination of the target in the first place. Because of their modifiable structure, more advanced switches and sensors can be constructed to interfere with the signalling in a cell from the outside.

(Picture provided by Sophia Hober, KTH)

Affibody molecules (Partner KTH) are small (~7 kDa ) alpha-helical molecules based on the protein A scaffold, with a number of attractive properties useful in different settings. These binding molecules have been selected against a large number of different proteins and used in applications such as protein detection and purification in different platforms. Binders with pM affinities and high specificity have been selected and used for molecular imaging and therapeutic applications are currently under investigation. The selected Affibody molecules have proven to be highly soluble and stable and and most of them contain no cysteines. The latter also makes them suitable for intracellular applications, both by expression/production within the cell and for targeted internalisation. Moreover, this also provides the opportunity to introduce a unique cysteine, e.g. for site-specific labelling or directed immobilisation on a solid surface. The solvent-exposed terminus of the Affibody binders allows for independent folding of fused proteins, and hence, multimeric constructs can easily be made by head-to-tail genetic fusions, hence the functional affinity (avidity) can be increased as has been achieved in several examples. Moreover, fusion proteins with dual functions can easily be constructed to facilitate detection (signalling molecules) or to endow the binder with complementary functions.

Thus, it is entirely likely that in future, alternative binders will complement antibody resources, becoming an equal or preferred choice in capture arrays, intracellular analysis and protein isolation. Together, the different types of recombinant binders in AFFINOMICS are key future probes of protein function determination.