News & Views | Published:

Protein folding

Protection from the outside

Nature volume 471, pages 4243 (03 March 2011) | Download Citation

Protein folding is a high-stakes process, with cell dysfunction and death being the unforgiving penalties for failure. Work in bacteria hints that organisms manage this process beyond the boundaries of the cytoplasm — and even the cell.

Protein misfolding can instigate disease one way or another1: it can cause both loss of function by leading to an insufficient amount of functional proteins, and gain of toxic function through the aggregation of misfolded proteins. Suppressing misfolding and aggregation is the job of the proteostasis network2,3, particularly the various classes of chaperones — evolutionarily conserved proteins that help other proteins to fold productively. Folding protection must operate in many environments, both inside and outside the cell. Writing in Nature Structural and Molecular Biology, Quan et al.4 identify in bacteria a new structural class of chaperone called Spy that, unusually, functions outside the typical cellular remit for chaperone activity.

For their analysis, Quan and colleagues created two 'sandwich fusion proteins' by inserting L53A I54A Im7 — an unstable version of the protein Im7, which is often used in protein-folding studies5 — into two other proteins: β-lactamase and DsbA. When folded, β-lactamase and DsbA confer resistance to the antibiotic penicillin and to cadmium ions (Cd2+), respectively. However, the insertion of a foreign protein into their sequences makes their folding dependent on the folding of the inserted protein. Thus, in the sandwich fusion proteins, L53A I54A Im7 folding leads to two independent selectable markers: penicillin resistance and Cd2+ resistance.

The authors4 induced expression of their fusion proteins in the periplasm of the bacterium Escherichia coli; the periplasm is the space between the inner and outer membranes in Gram-negative bacteria. In most cases, they observed no resistance to either penicillin or Cd2+, presumably because the inability of L53A I54A Im7 to fold prevented β-lactamase and DsbA from folding. A number of strains, however, did gain both penicillin and Cd2+ resistance.

The resistant strains also produced a massive amount of Spy, suggesting that this little-known periplasmic protein had a hitherto unrecognized chaperone activity. The researchers corroborate this result in vitro, showing that Spy can inhibit both aggregation and promote folding, even at sub-stoichiometric concentrations.

Quan et al. also show that Spy activity is independent of the cellular energy molecule ATP. This is not surprising, given that the protein functions outside the cytoplasm. However, operation of Spy at sub-stoichiometric concentrations is surprising, because chaperones that work in this way generally use ATP6. According to conventional wisdom, it is difficult — if not impossible — to imagine a mechanism for how a chaperone actively remodels the protein-folding energy landscape without an energy input. It is equally difficult to reconcile Spy's effects on protein folding and aggregation with a simple holdase mechanism, in which a chaperone passively binds to unfolded proteins.

There could be several explanations. To protect nascent peptides emerging through the inner membrane, Spy could work during protein translation, binding transiently to nascent proteins to stabilize them. Spy could be an efficient protective osmolyte, and thus thermodynamically stabilize proteins' native states by promoting the formation of hydrogen-bonded secondary structures7, which would be consistent with its high levels in the periplasm. Or Spy could be a steric foldase — a type of chaperone that stabilizes the folded state of proteins by binding to them8. Clearly, Spy's mechanism of action merits further investigation.

The discovery of Spy adds to the current repertoire of chaperones functioning in the periplasmic space of Gram-negative bacteria9 and raises questions about the existence of extra-cytoplasmic, or outer, proteostasis networks (the outPN) in complex eukaryotes (plants and animals). Whereas the bacterial inner membrane rigorously protects the cytoplasm and the intracellular proteostasis networks (inPN), the outer membrane is permeable to small molecules (those with a molecular mass of less than roughly 600). It functions as a filter to retain periplasmic proteins close to the surface of E. coli, thus preventing their dilution in the environment. It is perhaps only a modest stretch to compare the bacterial periplasmic space to the interstitial spaces in vertebrates (Box 1).

Box 1: Chaperone networks

In addition to the intracellular proteostasis network (inPN) in its cytoplasm, Escherichia coli produces many chaperones8 — including Spy, identified by Quan et al.4 — that protect protein folding in the periplasm in an ATP-independent manner (a).

Mammals have a number of distinct interstitial spaces filled with bodily fluids that could also operate independently of ATP to protect the major organ systems (b). However, unlike the periplasmic space of E. coli, which is open to the environment, the interstitial systems are closed. Interstitial fluids ultimately communicate with the environment through the kidney filtration system, or through uptake and metabolism by the liver.

Plasma (red) provides components of the extracellular chaperone network (outPN) to the peritoneal (abdomen), pericardial (heart), pleural (lungs), synovial (joints) and amniotic fluids (for simplicity, all grouped in pink). Each might form an interstitial system protecting a separate organ system, and all have a rich protein content, reflecting their passive coupling to plasma. Both the lymphatic system (green), which houses a key arm of the immune system, and the central nervous system's cerebrospinal fluid (CFS; blue) seem to be separate from the plasma outPN-related fluids. CFS is largely devoid of protein, but is possibly protected by the blood–brain barrier through the plasma outPN. E.T.P. & W.E.B.

Unfortunately, our knowledge of the composition and function of the outPN in complex eukaryotes is limited. Although small amounts of the classic chaperones Hsp70 and Hsp90 can be found outside the cell under stress conditions10, their roles remain controversial, and the lack of extracellular ATP makes them ill-suited to a chaperoning role outside the cell. In addition, abundant plasma proteins such as albumin and globulins can bind to other proteins, but their potential role as outPN components remains to be carefully explored. Nonetheless, there is evidence for potential outPN players that chaperone defective proteins — including α1-acid glycoprotein11, α-1-antitrypsin12,13, asialoglycoproteins14, plasma gelsolin15, clusterins16, α2-macroglobulins17 and transthyretin, which is thought to be protective against Alzheimer's disease18.

Is there an equivalent of stress-related Spy induction in humans? At least one possibility is the proteins whose levels increase during the acute-phase response to inflammation17 (such as α1-acid glycoprotein and haptoglobulin) and that have protein-folding protective functions. Even the innate and adaptive immune responses could be seen as highly evolved outPN systems (Box 1).

Undoubtedly, the intracellular proteostasis network is conserved and universal2,3. But the observations4,9 that the seemingly lowly E. coli can protect itself from a periplasmic folding problem by the production of Spy and other non-ATP-dependent chaperones could shift our view of the role of the interstitial space towards it being a home for a comparable extracellular proteostasis network in vertebrates2,3. Indeed, the outPN in vertebrates could report on and manage extracellular protein-folding stress, working in parallel with inflammatory and immune responses (Box 1). After all, like E. coli, vertebrates experience stressful situations every day.


  1. 1.

    , , , & Annu. Rev. Biochem. 78, 959–991 (2009).

  2. 2.

    , , & Science 319, 916–919 (2008).

  3. 3.

    & Curr. Opin. Cell Biol. doi:10.1016/ (2010).

  4. 4.

    et al. Nature Struct. Mol. Biol. (2011).

  5. 5.

    , & Nature Struct. Biol. 9, 209–216 (2002).

  6. 6.

    Curr. Opin. Struct. Biol. 18, 35–42 (2008).

  7. 7.

    , & Proc. Natl Acad. Sci. USA 103, 13997–14002 (2006).

  8. 8.

    , , & Mol. Microbiol. 64, 917–922 (2007).

  9. 9.

    , & Adv. Protein Chem. Struct. Biol. 78, 51–97 (2009).

  10. 10.

    , , & Biochim. Biophys. Acta 1778, 1653–1664 (2008).

  11. 11.

    Bioorg. Med. Chem. Lett. 20, 1205–1209 (2010).

  12. 12.

    Biochem. Biophys. Res. Commun. 393, 242–247 (2010).

  13. 13.

    , & Proc. Am. Thorac. Soc. 7, 415–422 (2010).

  14. 14.

    Physiol. Rev. 75, 591–609 (1995).

  15. 15.

    et al. Proc. Natl Acad. Sci. USA 106, 11125–11130 (2009).

  16. 16.

    , , , & Adv. Cancer Res. 104, 89–114 (2009).

  17. 17.

    , & Comp. Med. 59, 517–526 (2009).

  18. 18.

    & Cell. Mol. Life Sci. 66, 3095–3101 (2009).

Download references

Author information


  1. Evan T. Powers is in the Department of Chemistry, The Scripps Research Institute, La Jolla, California 92037, USA.

    • Evan T. Powers
  2. William E. Balch is in the Department of Cell Biology, The Skaggs Institute for Chemical Biology, Department of Chemical Physiology and Institute for Childhood and Neglected Diseases, The Scripps Research Institute, La Jolla, California 92037, USA.

    • William E. Balch


  1. Search for Evan T. Powers in:

  2. Search for William E. Balch in:

Corresponding authors

Correspondence to Evan T. Powers or William E. Balch.

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