Chemical biology

How to minimalize antibodies

The success of antibodies as pharmaceuticals has triggered interest in crafting much smaller mimics. A crucial step forward has been taken with the chemical synthesis of small molecules that recruit immune cells to attack cancer cells.

Over the past two decades, a wave of biological medical products, or biologics, dubbed monoclonal antibodies (mAbs) has conquered the pharmaceutical armamentarium. There are more than 30 mAbs marketed for treating cancer, autoimmune diseases and other serious medical conditions, and a similar number are in late-stage clinical trials1. At least five mAbs have each garnered more than US$5 billion in annual revenue. Small molecules that mimic the pharmacological properties of mAbs therefore have the potential to become highly competitive drugs. Writing in the Journal of the American Chemical Society, McEnaney et al.2 provide evidence that this is an achievable goal.

The success of mAbs as pharmaceuticals is remarkable, given their size, composition and heterogeneity (mAbs are populations of similar, but not identical, molecules). They are more than 100 times larger than conventional small-molecule drugs manufactured using chemical synthesis, and therefore require more-expensive and less-precise biological synthesis.

Tripartite Y-shaped antibodies evolved as a cornerstone of the vertebrate immune system. They can selectively and tightly bind to foreign molecules with their two Fab regions (targeting functions) and recruit components of the host immune system with their Fc region (effector functions). Furthermore, the Fc region mediates recycling of the antibody molecule, resulting in its retention in the blood (giving it a prolonged circulatory half-life). In the treatment of cancer, mAbs use their tripartite architecture to bring immune cells readied for the kill in close proximity to cancer cells (Fig. 1a).

Figure 1: Bridging target and effector cells.
figure1

a, Tripartite antibody molecules bridge target and effector cells by simultaneously engaging two structures (antigens) on the target-cell surface with their Fab regions and one effector-cell surface receptor with their Fc region. b, Small molecules that mimic Fab regions can be docked to whole antibodies (not shown) or Fc regions (depicted) to produce molecules that have the pharmacological properties of antibodies. c, McEnaney et al.2 now report a quadripartite synthetic antibody that is 20 times smaller than naturally occurring antibody molecules and readily generated by chemical synthesis. Cartoon versions of the antibody and the small molecules are approximations of the actual molecular structures and not drawn to scale.

These three principal features of antibody molecules bestow pharmacological properties on mAbs that are unmatched by small molecules. But the complexity of mAbs has prevented generic versions of branded drugs from being produced by competitor companies, and slowed the production of similar versions — providing a strong investment incentive for pharmaceutical companies, but potentially driving up health-care costs. By contrast, small molecules that mimic mAbs would have lower manufacturing costs and enable competition from generics. And unlike mAbs, which often trigger immune responses in patients, small molecules are not immunogenic. Moreover, they can penetrate tissues and cells more efficiently, can reach buried sites on target molecules that are inaccessible to mAbs, can be given orally and have a longer shelf life.

The discovery and development of peptides, peptidomimetics and other small molecules that have a specificity and affinity for biological targets comparable to those of mAbs have been key in efforts to replace mAbs by small molecules. In order for them also to have the effector functions and prolonged circulatory half-life of mAbs, these small molecules were designed to dock to antibodies either in vitro, yielding chemically programmed antibodies3, or in vivo, producing antibody-recruiting molecules4. Although the resulting hybrid molecules (Fig. 1b) combine several of the advantages of mAbs and small molecules, their biological component still restricts the scope of their chemical component.

A provocative question has therefore been whether small molecules can copy both the Fab- and Fc-mediated pharmacological properties of mAbs. This is theoretically possible if Fab-mimicking small molecules can be fused to other small molecules that bind to Fc receptors. McEnaney et al. now deliver this missing link. They used chemical synthesis to combine a known Fab-mimicking small molecule that binds to a cell-surface receptor on prostate-cancer cells with a known Fc-mimicking cyclic peptide that selectively binds to an Fc receptor called CD64 on immune cells. The resulting compound mimics two of the three principal natural features of antibodies.

“This compound induces immune cells to engulf and ingest prostate-cancer cells in vitro.”

Bridging two cell-surface receptors is a formidable task for a small molecule. It requires a linker between the Fab- and Fc-mimicking components that has sufficient length, solubility and rigidity. The authors used computer modelling to predict that less than one-third of the naturally occurring distance between Fab and Fc regions in an antibody is required to simultaneously engage receptors on two different cells, and they used this information to design their linker. They also found that two copies of each Fab- and Fc-mimicking component are essential for efficiently mediating targeting and effector functions in vitro. The result is a quadripartite molecule (Fig. 1c) that resembles tripartite antibodies with respect to composition and function, but which is 20 times smaller, homogeneous (all the molecules are the same) and readily generated by chemical synthesis. Encouragingly, McEnaney et al. observed that this compound induces immune cells to engulf and ingest prostate-cancer cells in vitro.

Although in vivo validation studies have still to be performed, it seems that McEnaney et al. have taken a pivotal step towards obtaining antibody-mimicking small molecules that avoid some of the liabilities of biologics. But there is more work to be done. First, the quadripartite molecule is about 7 kilodaltons in size — substantially smaller than antibodies, but still larger than conventional small-molecule drugs (less than 1 kDa), which limits most of the potential advantages discussed earlier. However, the molecular weight can conceivably be cut further by replacing the relatively large Fc-mimicking cyclic peptide with a peptidomimetic or other small molecule. Then again, an intermediate-sized synthetic antibody mimic might be a good thing, because unrestrained access of small molecules to intra- and extracellular nooks and crannies could make their activity and toxicity profiles unpredictable.

Second, the Fc-mimicking component of the synthetic antibody mimic binds to only one kind of Fc receptor (CD64), whereas natural antibodies and mAbs engage other Fc receptors, including CD16 and CD32, on a variety of functionally different effector cells. Moreover, the Fc region of antibodies triggers activation of the complement cascade, which is an additional mechanism of target-cell destruction, and it also mediates prolonged circulatory half-life — effects that generally have been difficult to produce with small molecules. Even so, synthetic antibody mimics that engage only one kind of Fc receptor might allow effector functions to be fine-tuned. The modular, versatile design of McEnaney and co-workers' molecules will also allow their properties to be tailored through chemical synthesis, for example to include a peptide or peptidomimetic that is retained in the blood by binding to circulating albumin protein5.

Third, Fab-mimicking small molecules are still limited in scope when compared with mAbs, which can be generated and evolved to bind to almost any cell-surface receptor selectively and tightly3. Nonetheless, our ability to generate and screen large chemical libraries, which are structurally much more diverse than biological libraries, has afforded access to an increasing number of small molecules that can compete with mAbs in terms of specificity and affinity6.

Meanwhile, proponents of biologics are not sitting idle. Antibody engineers have generated a large variety of antibody molecules that have improved targeting and effector functions. For example, a new class of 'bispecific' antibody7 can recruit and activate T cells, which are particularly potent effector cells that cannot be directly engaged by natural antibodies and mAbs. Although not as miniaturized as synthetic antibody mimics, these bispecific antibodies are three times smaller than mAbs and can be clinically potent, safe and profitable, as demonstrated by the recently marketed anticancer drug blinatumomab7. Intriguingly, however, synthetic antibody mimics might be better copies of these T-cell-engaging biologics than of conventional mAbs, because the biologics bind to just one kind of effector-cell receptor (CD3) and do not need prolonged circulatory half-lives. All things considered, synthetic antibody mimics have the potential to become a new class of pharmaceutical.Footnote 1

Notes

  1. 1.

    See all news & views

References

  1. 1

    Reichert, J. M. mAbs 7, 1–8 (2015).

    CAS  Article  Google Scholar 

  2. 2

    McEnaney, P. J. et al. J. Am. Chem. Soc. 136, 18034–18043 (2014).

    CAS  Article  Google Scholar 

  3. 3

    Rader, C. Trends Biotechnol. 32, 186–197 (2014).

    CAS  Article  Google Scholar 

  4. 4

    McEnaney, P. J., Parker, C. G., Zhang, A. X. & Spiegel, D. A. ACS Chem. Biol. 7, 1139–1151 (2012).

    CAS  Article  Google Scholar 

  5. 5

    Dennis, M. S. et al. J. Biol. Chem. 277, 35035–35043 (2002).

    CAS  Article  Google Scholar 

  6. 6

    Kodadek, T. Curr. Opin. Chem. Biol. 14, 713–720 (2010).

    CAS  Article  Google Scholar 

  7. 7

    Frankel, S. R. & Baeuerle, P. A. Curr. Opin. Chem. Biol. 17, 385–392 (2013).

    CAS  Article  Google Scholar 

Download references

Author information

Affiliations

Authors

Corresponding author

Correspondence to Christoph Rader.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Rader, C. How to minimalize antibodies. Nature 518, 38–39 (2015). https://doi.org/10.1038/518038a

Download citation

Further reading

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.