News & Views | Published:

Structural biology

Enzyme assembly line pictured

Nature volume 510, pages 482483 (26 June 2014) | Download Citation

Many enzymes form 'assembly lines' containing a series of catalytic modules. Visualization of how the structure of a module shifts during catalysis provides a clearer idea of how such enzymes work. See Article p.512 & Letter p.560

Living cells contain many molecular machines that integrate and orchestrate essential processes. One of the most familiar is the ribosome, which polymerizes amino acids to form proteins. In the early 1990s, a radically different type of polymerizing assembly line was discovered1,2: polyketide synthase (PKS) multi-enzymes, which make polyketide compounds, including many important antibiotics. Remarkably, these complexes use a different module of catalytic sites for each cycle of a chemical-chain extension. Efforts to understand these sophisticated catalysts have today received a great boost from two studies in this issue from the same research group: Dutta et al.3 (page 512) report the three-dimensional structure of a PKS protein that houses an intact catalytic module, and Whicher and colleagues4 (page 560) investigate how individual domains move during the extension cycle.

Biochemical evidence5 suggests that modular PKS multi-enzymes are homodimeric (composed of two identical subunits), and that the subunits are aligned head-to-head, tail-to-tail and are tightly wound around each other. A functional module thus contains two copies of each type of domain — just as in animal fatty-acid synthase, a complex that uses a very similar set of enzymes to those in a PKS module. In 2008, the crystal structure of animal fatty-acid synthase was obtained6, revealing an X-shaped molecule in which several of the domains are highly mobile. An independent study7 using single-particle electron cryo-microscopy (cryo-EM) confirmed that animal fatty-acid synthase is exceptionally flexible and that sets of domains change position in response to substrate binding, hinting that PKS modules might behave like this too. Many excellent studies have provided static X-ray crystal structures of individual domain fragments of PKS modules, but Dutta et al. and Whicher et al. are the first to visualize an intact homodimeric module in sufficient detail to reveal its architecture and the dynamic changes that occur during the chain-extension cycle.

The researchers studied PikAIII, a PKS protein that forms module 5 of the bacterial assembly line that makes the antibiotic pikromycin. The extension cycle in PikAIII starts when the protein accepts a partially built polyketide chain (containing an acyl chemical group) from the carrier domain of the previous module, and recruits a new building block (another acyl group) to its own carrier domain (Fig. 1). The tethered acyl groups are then bonded together in a reaction catalysed by a ketosynthase domain, forming an extended chain that ends up on PikAIII's carrier protein. An additional processing domain catalyses a reduction reaction of the extended chain, which is finally handed on to the next module.

Figure 1: Close-up of a modular assembly-line enzyme.
Figure 1

PikAIII is the fifth module of six in the multi-enzyme that constructs the antibiotic pikromycin. It contains domains that perform substrate recruitment (red), substrate carriage (orange) and catalysis (blue). Dutta et al.3 report that PikAIII is a homodimer that defines an internal reaction chamber (for clarity, only one subunit is shown in full colour; the other is shown in outline). Whicher et al.4 suggest how the domains move as PikAIII catalyses a chain-extension cycle. a, An acyl group (COR1) is transferred from a carrier domain in module 4 to module 5 through a 'side entrance'. R1 represents a chain containing five molecular building blocks. b, The carrier domain of module 5 recruits another acyl building block. Me, methyl group. c, d, It then moves between the catalytic sites to enable a bond-forming reaction between the two acyl groups and the reduction of a carbonyl group (C=O) to a hydroxyl group (OH), before flipping out of the reaction chamber to present the extended chain to module 6.

Dutta and co-workers used cryo-EM to obtain three-dimensional reconstructions of PikAIII at a resolution of 7 to 10 ångströms. This is not high enough to resolve individual atoms, but by fitting known crystal structures of individual domains into place, the authors were able to visualize the module in its entirety. They found that the identical subunits were indeed aligned head-to-head, tail-to-tail and clasped around each other, as predicted7. The subunits enclose a single inner chamber into which all the active sites of the domains face. This arrangement is different from that found for animal fatty-acid synthase, and from that recently proposed for PKSs on the basis of low-resolution X-ray scattering analysis8, but it helps to explain how PKSs control the sequence of events. The authors also observed that the carrier proteins are mobile and adopt alternative positions near one or other of the catalytic domains.

Next, Dutta and colleagues made a construct in which PikAIII was covalently tethered to the carrier protein of the previous module, with the carrier bearing a synthetic polyketide acyl group. Cryo-EM analysis of this revealed that two different channels provide access to the catalytic site of the bond-forming ketosynthase domain. The externally delivered acyl group gains access through a 'side entrance', whereas the PikAIII carrier protein offers the new building block for bond formation through an opening from the internal chamber (Fig. 1).

Whicher et al. used cryo-EM to study PikAIII samples in which a synthetic acyl group was directly attached either to the ketosynthase domain or to the PikAIII carrier protein, to mimic intermediates in the extension cycle. In each case, significant shifts were noted in the position of the carrier protein and the other domains, compared to Dutta and colleagues' structure of non-acylated PikAIII. It therefore seems that the acyl chain is specifically recognized by the catalytic domains, which helps to explain why naturally occurring PKSs normally yield a single product. A similar conclusion has been reached for fatty-acid synthase in recent work9 that used chemical probes to trap carrier domains as they dock to catalytic domains. Deeper understanding of this aspect of PKS selectivity should help synthetic biologists to redesign these assembly-line synthases to prepare potentially useful analogues of naturally occurring antibiotics.

A key issue that cannot be fully addressed by structural studies at this level of resolution is whether some mechanism actively directs the carrier protein and its cargo to the next 'correct' reaction partner, or whether they simply find their way by diffusion within the chamber. Also unclear is whether the loaded carrier protein can reach all catalytic active sites on either subunit, or is confined to one subset. Likewise, what prevents the extended chain from being prematurely passed to the next module before all the programmed operations in module 5 have taken place? Addressing these questions will require further genetic, biochemical and biophysical work. Nonetheless, by showing that an intact module is much more than the sum of its parts, this elegant study has given fresh impetus to our search for the answers.


  1. 1.

    , , , & Nature 348, 176–178 (1990).

  2. 2.

    , , , & Science 252, 675–679 (1991).

  3. 3.

    et al. Nature 510, 512–517 (2014).

  4. 4.

    et al. Nature 510, 560–564 (2014).

  5. 5.

    et al. Nature Struct. Mol. Biol. 3, 188–192 (1996).

  6. 6.

    , & Science 321, 1315–1322 (2008).

  7. 7.

    , & Nature Struct. Mol. Biol. 16, 190–197 (2009).

  8. 8.

    , , & J. Mol. Biol. 426, 2229–2245 (2014).

  9. 9.

    et al. Nature 505, 427–431 (2014).

Download references

Author information


  1. Peter F. Leadlay is in the Department of Biochemistry, University of Cambridge, Cambridge CB2 1GA, UK.

    • Peter F. Leadlay


  1. Search for Peter F. Leadlay in:

Corresponding author

Correspondence to Peter F. Leadlay.

About this article

Publication history



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.

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