Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.


Drug discovery goes au naturel

A Correction to this article was published on 16 April 2014

In a study that showcases the potential of semisynthetic drug design, structural modification of an existing antibiotic with little activity against Mycobacterium tuberculosis has generated a new class of effective antitubercular lead.

Discovering useful drugs to treat tuberculosis is not an exercise for the faint of heart. Screening for agents that inhibit the growth of the causative mycobacterium produces very few hits, and optimizing those hits into drug-like 'lead' molecules is exceptionally difficult, owing to the bacterium's impermeable cell envelope. Even armed with a promising lead, animal models and early-stage clinical trials have little predictive value, and the long duration of therapy currently required to treat tuberculosis makes safety hurdles much higher than for most bacterial infections. So, an attractive strategy, and one that lies at the heart of a report by Lee et al.1 in Nature Medicine, is to repurpose an existing antibiotic class that is already known to be safe and effective in other infections, but that has poor antitubercular activity.

The authors focused their efforts on spectinomycin, a natural product isolated from Streptomyces bacteria that has a long history of safe use for the treatment of gonorrhoea in patients who are allergic (or resistant) to penicillin. Despite targeting the bacterial ribosome — a complex of proteins and RNA molecules that is essential for cellular protein translation and that is highly evolutionarily conserved — spectinomycin shows little activity against Mycobacterium tuberculosis. Other translation inhibitors, such as streptomycin, kanamycin and amikacin, are widely used in antituberculosis chemotherapy, but their binding site on the ribosome is distinct from that of spectinomycin.

Armed with the crystal structure of spectinomycin bound to the bacterial ribosome2 and a mutant strain of M. tuberculosis defective in a cellular efflux pump that showed increased susceptibility to spectinomycin, Lee and colleagues produced derivatives of the antibiotic, called spectinamides, that showed potent antituberculosis activity both in vitro and in mice. The co-crystal structure of the drug bound to the bacterial ribosome allowed the authors to deduce which sites in the drug were crucial for binding and which were not; they then focused on the latter, chemically modifying the natural product to improve cellular activity by reducing drug efflux. They assessed the effect of the modifications on efflux by comparing the activity of the compounds in wild-type and efflux-mutant bacteria, and used separate in vitro assays to monitor the modified products' ribosome-binding ability.

Drug developers have long had a love–hate relationship with natural products. On the one hand, the sheer complexity and three-dimensionality of natural products provides more opportunity for highly specific, tight binding to potential targets than do the small, flat synthetic molecules that are typical of the large libraries amassed by most pharmaceutical companies (Fig. 1). Many natural products have evolved as weapons deployed by bacteria to kill other bacteria, and taking advantage of that evolutionary war has obvious attractions. On the other hand, natural compounds rarely have the ideal pharmacological properties needed for direct use, and therefore require chemical transformation — a process called semisynthesis — to make them suitable as medicines. Such semisyntheses can be extraordinarily complex and require enormous investment to understand the chemical properties of the parent molecule, thereby allowing a systematic exploration of where alterations can be made.

Figure 1: Natural complexity.

a, For several decades, pharmaceutical companies have focused their efforts in tuberculosis drug discovery on entirely synthetic drugs, such as isoniazid and ethambutol, which are cheap and easy to produce. b, The structural complexity of natural products (for example, rifampicin), however, provides advantages over simple synthetic structures, such as more-effective binding to therapeutic targets. Lee and colleagues' study1 of spectinomycin demonstrates how our increasing knowledge of drug metabolism and protein structures makes semisynthetic alteration of natural compounds an efficient and effective alternative to complete chemical synthesis.

Current practice among scientists working in tuberculosis drug discovery is heavily weighted towards screening and optimizing small synthetic lead molecules. One such molecule, bedaquiline, was, in 2012, the first drug in 40 years to receive US Food and Drug Administration approval for use in the treatment of tuberculosis. However, over these past 40 years, about three-quarters of all approved antibacterial drugs have been the result of semisynthetic efforts from natural-product starting points3.

In fact, the biggest advance in tuberculosis chemotherapy so far was unquestionably the addition of rifampicin to multidrug cocktails in the 1970s. Before the introduction of rifampicin, 18–24 months of therapy with a combination of two or three different agents was required to achieve a sterile cure of tuberculosis. But, following landmark clinical trials in Africa by the British Medical Research Council, the standard therapy for tuberculosis infections became combinations of four drugs, including rifampicin, for a mere six months4. The starting point for rifampicin development was a complex mixture of metabolites from the bacterium Nocardia mediterranei, isolated by the Italian pharmaceutical company Lepetit in 1957. It took 8 years for this company (partnering with Ciba-Geigy, based in Switzerland) to understand the various rifampicin-related metabolites and their chemistry sufficiently to allow the development of an orally available analogue, which became the basis of these new short-course chemotherapy regimens5.

The molecule described by Lee et al. is not the 'next rifampicin', but there are two exciting features of the authors' paper that should encourage more effort in the semisynthetic modification of natural products for tuberculosis drug discovery. First, the authors took full advantage of recent advances in crystallography to guide their strategy for modifying the natural product. This allowed them to focus on a limited set of chemical modifications rather than the inefficient empirical approaches used in the past (for example, several hundred rifampicin analogues had to be explored before it even became clear which positions could be modified without completely losing biological activity). Second, they used our expanding understanding of the role of drug efflux in tuberculosis6 to assess the efficacy of the modified compounds. The proof-of-concept achieved in this study will nudge the pendulum of interest back towards natural products as viable starting points for tuberculosis drug developers.


  1. 1

    Lee, R. E. et al. Nature Med. (2014).

  2. 2

    Borovinskaya, M. A., Shoji, S., Holton, J. M., Fredrick, K. & Cate, J. H. ACS Chem. Biol. 2, 545–552 (2007).

    CAS  Article  Google Scholar 

  3. 3

    Newman, D. J. & Cragg, G. M. J. Nat. Prod. 75, 311–335 (2012).

    CAS  Article  Google Scholar 

  4. 4

    Fox, W., Ellard, G. A. & Mitchison, D. A. Int. J. Tuberc. Lung Dis. 3, S231–S279 (1999).

    CAS  PubMed  Google Scholar 

  5. 5

    Sensi, P. Rev. Infect. Dis. 5, S402–S406 (1983).

    CAS  Article  Google Scholar 

  6. 6

    Balganesh, M. et al. Antimicrob. Agents Chemother. 56, 2643–2651 (2012).

    CAS  Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Clifton E. Barry.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Barry, C. Drug discovery goes au naturel. Nature 506, 436–437 (2014).

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.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing