Drug discovery

Fighting evolution with chemical synthesis

A synthetic strategy has been developed that provides easy access to structurally diverse analogues of naturally occurring antibiotics, providing a fresh means of attack in the war against drug-resistant bacteria. See Article p.338

Pathogenic bacteria have developed strong resistance to various antibiotics as a result of the rapid evolution of the bacteria in response to widespread antibiotic misuse. This problem is exacerbated by the reduction in antibacterial-drug research by large pharmaceutical companies over the past few decades — we are failing to replenish our armamentarium quickly enough to compensate. An apocalyptic scenario looms wherein a drug-resistant 'superbug' becomes so well established that the most trivial infections would be as life-threatening as they were during the pre-antibiotic era. On page 338 of this issue, Seiple et al.1 offer an ingenious strategy to bolster the antibiotic-discovery process: a highly modular and versatile chemical-synthesis approach that provides access to a rich array of antibiotics possessing different molecular topologies and functionalities.

Naturally occurring compounds derived from bacteria or fungi have long been the primary source of antibiotics. These compounds can be produced on a large scale by fermentation, but few inherently possess adequate bioavailability, in vivo activity or metabolic stability for use as drugs2. Chemical modifications are often necessary to address these issues. The complexity of natural products means that such modifications often involve a long, linear sequence of steps in which chemical groups in the molecules are masked with 'protecting' groups to allow the modifications to be made, and then unmasked. Making variations at certain molecular locations tends to be particularly laborious or impractical, and so synthesizing a diverse collection of antibiotic candidates for drug discovery can be forbiddingly difficult.

Seiple and colleagues show convincingly that chemical synthesis can surmount the challenge of molecular diversification, trumping strategies based on synthetic biology3 or semisynthesis (in which compounds from natural sources are used as starting materials). In their approach, the authors constructed complex macrolides — antibiotics that contain a characteristic large ring of atoms known as a macrocycle — through the assembly of eight modular building blocks. These simple modules could each be varied easily to incorporate different structural motifs at specific locations in the macrolide skeleton.

The researchers' synthetic route included seven key coupling reactions, each of which introduced different building blocks into the macrolide (Fig. 1). Such reactions are said to be convergent because they bring together modules that were prepared in separate, parallel reaction sequences. The convergent coupling reactions serve as branch points in the synthesis that allow a diverse range of structural modifications to be easily made. For example, Seiple et al. introduced reactive groups such as azides, allyl groups and amides at different positions, all of which could be readily converted into various structural motifs. Such manipulation, when combined with variations in the building blocks, enables an exponential expansion of diversity.

Figure 1: A modular approach to macrolide synthesis.

Seiple et al.1 report a synthetic route that allows the rapid preparation of a diverse array of structural analogues of macrolide antibiotics. The authors prepared eight modular building blocks (coloured shapes), the structures of which can be easily modified. The synthetic route includes seven key steps in which the building blocks are joined together. Chemical groups in the modules can also be readily converted into various other groups after the modules have been incorporated into macrolide intermediates (not shown). This approach allowed the authors to prepare more than 300 macrolides, including the clinical candidate solithromycin.

This approach is possible only because the synthetic sequence is extremely well choreographed. Steps such as protecting-group manoeuvres are kept to a minimum; convergent couplings and other complexity-building transformations account for the bulk of the sequence. These key reactions are mechanistically very different from one another (ranging between the addition of an enolate and a transition-metal-catalysed cycloaddition, for those in the know). Yet the authors show that these all work well for modules that carry a variety of groups.

The formation of macrocyclic rings is one of the toughest challenges in macrolide synthesis — many routes to these antibiotics are foiled because attempts to form the ring fail4. Conventionally, macrocycle formation must be individually optimized for each reaction substrate by finding a set of protecting groups that allows the transformation to occur. However, Seiple and colleagues report that, in their work, macrolides containing various ring sizes can be made using almost identical ring-formation conditions.

Similarly, other reactions in the synthetic sequence need minimal adaptation to work for different substrates that contain a variety of chemical groups, and deliver consistent yields of products. Such generality is the cornerstone of the authors' approach, because it allows hundreds of analogues of macrolides, bearing highly dissimilar arrangements of attached groups, to be accessed expeditiously with little perturbation to the overall strategy.

Another prerequisite for rapidly generating structural diversity is an abundant supply of diversifiable intermediates; this in turn requires all the chemical transformations in the syntheses to be conducted on gram scales5. Seiple et al. tackled this challenge using a series of in-house methodologies. The first was a scalable aldol reaction, a type of carbon–carbon bond-forming reaction. This reaction shows exceptional stereoselectivity (it preferentially forms isomers that have specific 3D orientations of atoms) and tolerates the presence of a potentially interfering amino group (NH2) in one of the reactants6. The second was a concise approach7 to preparing the carbohydrate desosamine and its analogues, which form a key part of another building block. These examples highlight the synergy between target-oriented synthesis and methodology developments.

Using such scalable methods allowed the authors to make gram quantities of several macrolides — a large amount for drug-discovery purposes — in remarkably high overall yields. This reaffirms the fact that scalable synthetic methods that can provide ample supplies of structural analogues of molecules are essential for the continued success of drug discovery4. Given the sheer number of possible macrolide analogues, some structural variations will remain challenging to produce synthetically, but further advances in synthetic methods will probably bridge the gaps.

Impressively, Seiple and colleagues prepared more than 300 macrolide antibiotics as candidates for pharmaceutical development. The researchers report that most of these compounds inhibit the growth of drug-resistant pathogens, including some extremely challenging strains. Further studies are necessary for assessing properties (such as toxicity and bioavailability) that will affect the antibiotics' clinical viability, but these compounds represent a gateway to commercial pharmaceuticals. One of the compounds is more effective against Gram-negative bacteria than are many other macrolides. Its activity is still modest, but marks a promising start to tackling an important medical need.

No antibiotic is impervious to the power of evolution — bacteria will always mutate to develop resistance, which means that new antibiotics must be pursued continuously and perpetually. To this end, synthetic platforms such as the present one offer rapid access to unparalleled structural variability for antibiotic candidates, while still allowing exquisite control of each variation. This fine-tunability, along with biological knowledge, can enable chemists to keep up with bacterial evolution. In the midst of the antibiotic-resistance crisis, Seiple and colleagues' work sends out a clear message: chemical synthesis is the way forward.Footnote 1


  1. 1.

    See all news & views


  1. 1

    Seiple, I. B. et al. Nature 533, 338–345 (2016).

  2. 2

    Wright, P. M., Seiple, I. B. & Myers, A. G. Angew. Chem. Int. Edn 53, 8840–8869 (2014).

  3. 3

    Keasling, J. D., Mendoza, A. & Baran, P. S. Nature 492, 188–189 (2012).

  4. 4

    Parenty, A., Moreau, X., Niel, G. & Campagne, J.-M. Chem. Rev. 113, PR1–PR40 (2013).

  5. 5

    Kuttruff, C. A., Eastgate, M. D. & Baran, P. S. Nat. Prod. Rep. 31, 419–432 (2014).

  6. 6

    Seiple, I. B., Mercer, J. A. M., Sussman, R. J., Zhang, Z. & Myers, A. G. Angew. Chem. Int. Edn 53, 4642–4647 (2014).

  7. 7

    Zhang, Z., Fukuzaki, T. & Myers, A. G. Angew. Chem. Int. Edn 55, 523–527 (2016).

Download references

Author information



Corresponding author

Correspondence to Phil S. Baran.

Related audio

Shamini Bundell find out how building antibiotics is a bit like Lego.

Related links

Related links

Related links in Nature Research

Tuberculosis: Drug discovery goes au naturel

Synthesis: A constructive debate

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Yan, M., Baran, P. Fighting evolution with chemical synthesis. Nature 533, 326–327 (2016). https://doi.org/10.1038/533326a

Download citation

Further reading


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.