News & Views | Published:

Molecular biology

Mature proteins braced by a chaperone

Nature volume 539, pages 361362 (17 November 2016) | Download Citation

Hsp70 chaperone molecules help other proteins to fold, and were thought to bind mainly to unfolded proteins. Single-molecule experiments now suggest that Hsp70s can also stabilize almost fully folded proteins. See Letter p.448

Cell function depends not only on the proper folding and localization of proteins when they are synthesized, but also on maintaining the well-being of fully folded (mature) proteins, and on their disposal when they are irretrievably damaged1. Seventy-kilodalton-sized heat-shock proteins (Hsp70s) are a major class of ubiquitous molecular 'chaperones' that have key roles in maintaining protein homeostasis2,3. Hsp70s interact with short stretches of amino-acid residues in extended conformations, such as those in unfolded proteins4, and thus prevent such proteins from misfolding and aggregating. But on page 448, Mashaghi et al.5 report that Hsp70s are more flexible in their binding ability than was previously appreciated: they also interact with intermediates in folding processes, and with nearly fully folded proteins. This suggests that Hsp70s play a bigger part in protein homeostasis than was thought.

Analysing how protein folding is facilitated by molecular chaperones has been challenging — particularly for those chaperones, such as Hsp70, for which nucleotide binding and hydrolysis drive cycles of interactions with proteins. Such chaperone–protein interactions are transient, and the conformations adopted by the substrate proteins are often highly variable and change over time. This is particularly true for the intermediates formed during protein folding.

Single-molecule analytical techniques can overcome these challenges, and thus are powerful approaches for addressing chaperone function. For example, the force generated by optical tweezers — a laser device that can hold and move microscopic objects — is ideal for pulling apart the folded structures of mature proteins to produce fully or partially unfolded proteins. Subsequent relaxation of the pulling force enables the refolding process to be observed. Chaperones, co-chaperones (proteins that assist chaperones in their functions) and nucleotides can be added to the molecule being studied to probe their effect on protein refolding and stability.

Mashaghi et al. used optical tweezers to analyse how DnaK, an Hsp70 from the bacterium Escherichia coli, interacts with maltose-binding protein. Surprisingly, they found that DnaK binds to and stabilizes nearly fully folded and partially folded proteins, in addition to unfolded proteins. These results suggest that Hsp70s not only assist the folding of unfolded and misfolded proteins, but also help to maintain the conformation of mature proteins — thus maintaining their function. This preventive maintenance for 'healthy' mature proteins has not previously been suggested as a role for Hsp70 chaperones.

Although examples of Hsp70 binding to mature proteins have been reported before, the chaperone was thought to interact with extended loops in proteins4. Such thinking is consistent with structural studies4,6 that revealed Hsp70's peptide-binding cleft to be so constrained that only an extended, unstructured polypeptide segment of a protein can be accommodated in it. The substrate-binding domain (SBD) of Hsp70 is dedicated to binding polypeptide substrates, and comprises about half of the chaperone molecule. A subdomain of the SBD, known as SBDβ, contains the peptide-binding cleft, and another subdomain, SBDα, forms a lid over the bound peptide segment (Fig. 1). SBDα thus stabilizes the chaperone–substrate interaction, but without forming any substantial direct contact with the substrate.

Figure 1: Different roles of Hsp70s in protein folding.
Figure 1

a, Seventy-kilodalton heat-shock proteins (Hsp70s) are molecular 'chaperones' that were thought to mainly assist the folding of unfolded, nascent protein chains. Unfolded proteins can bind to a cleft in a region of Hsp70 known as substrate-binding domain-β (SBDβ), and another region (SBDα) forms a lid that closes over the unfolded protein chain. The nucleotide-binding domain binds to ATP, which can be hydrolysed to form ADP. The lid is predominantly open when ATP is bound and closed when ADP is bound. b, Mashaghi et al.5 report single-molecule experiments with an Hsp70, and find that it interacts with an almost fully folded protein. This suggests that Hsp70s have roles in maintaining the structures of almost fully folded proteins.

This model of substrate binding has provided a satisfying mechanistic explanation for how Hsp70 aids protein folding, especially the early steps in which nascent proteins consist of unfolded chains. Mashaghi et al. now find that the SBDα lid is particularly important in stabilizing almost-mature protein substrates — an intriguing observation that raises the question of how Hsp70s bind to such substrates to increase their stability. A structure of an Hsp70 in complex with a mature substrate would help to provide insight into the mechanics of this.

The substrate-binding activity of Hsp70s depends on which nucleotide is bound7,8,9,10,11: either ATP or the product of ATP hydrolysis, ADP. When ATP is bound, the peptide-binding pocket is predominantly open, allowing easy, rapid access for substrates. When ADP is bound, the peptide-binding pocket is predominantly closed — reducing substrate access but enhancing the stability of the chaperone–substrate interaction after the substrate is bound. The ATP-bound state therefore probably binds large, mostly folded substrates much more readily than the ADP-bound state. Both ATP binding and hydrolysis are crucial to the conformational changes of Hsp70 that allow it to bind and release substrates as they fold.

It is therefore surprising that Mashaghi and colleagues find that Hsp70 is more effective in stabilizing the (nearly) mature conformation of maltose-binding protein when ADP is the predominant nucleotide present, rather than ATP. However, the authors note that the ratio of ADP to ATP in cells is high during heat stress, and therefore propose that the observed stabilization is beneficial to the cell under such conditions — it might provide a way of maintaining proteins in their functional, folded conformations for little expenditure of energy. This explanation makes sense, but more work is needed to understand the physiological relevance of the results.

Mashaghi and co-workers' study provides a wealth of information, but the limitations of the analysis should be kept in mind. The authors tested a single Hsp70: DnaK, a well-studied model of these chaperones. But even though Hsp70s are highly evolutionarily conserved, there are substantial differences in both their amino-acid sequences and their specific cellular functions. Does the observed stabilization of folded protein substrates also occur in the Hsp70s of eukaryotic organisms (which include plants, animals and fungi)?

Moreover, Hsp70 activity in protein folding depends on the cooperation of two types of co-chaperone: J-proteins and nucleotide-exchange factors (NEFs)12,13. These co-chaperones, especially NEFs, have much more diverse functions and amino-acid sequences than Hsp70s, so how do they fit into this expanded view of Hsp70 function? And most of Mashaghi and colleagues' single-molecule assays involved just one kind of protein substrate. How generalizable are the observed effects for other substrates?

Finally, it must be remembered that the unfolding force generated by optical tweezers is artificial. It is not clear how representative these findings are of the situation in cells. More experiments, from single-molecule studies to in vivo investigations, will therefore be needed to better understand the diverse roles of Hsp70s. But in the meantime, Mashaghi et al. have provided much food for thought.



  1. 1.

    , & Science 353, aac4354 (2016).

  2. 2.

    & Cell. Mol. Life Sci. 62, 670–684 (2005).

  3. 3.

    , & J. Mol. Evol. 38, 1–17 (1994).

  4. 4.

    , , & J. Mol. Biol. 427, 1575–1588 (2015).

  5. 5.

    et al. Nature 539, 448–451 (2016).

  6. 6.

    et al. Science 272, 1606–1614 (1996).

  7. 7.

    , & Science 245, 385–390 (1989).

  8. 8.

    , , & Science 263, 971–973 (1994).

  9. 9.

    et al. Nature Struct. Mol. Biol. 20, 900–907 (2013).

  10. 10.

    , , & Mol. Cell 48, 863–874 (2012).

  11. 11.

    , & Cell 151, 1296–1307 (2012).

  12. 12.

    & Nature Rev. Mol. Cell Biol. 11, 579–592 (2010).

  13. 13.

    & Front. Mol. Biosci. 2, 10 (2015).

Download references

Author information


  1. Qinglian Liu is in the Department of Physiology and Biophysics, Virginia Commonwealth University, Richmond, Virginia 23298-0551, USA.

    • Qinglian Liu
    •  & Elizabeth A. Craig
  2. Elizabeth A. Craig is in the Department of Biochemistry, University of Wisconsin–Madison, Madison, Wisconsin 53706-1544, USA.


  1. Search for Qinglian Liu in:

  2. Search for Elizabeth A. Craig in:

Corresponding authors

Correspondence to Qinglian Liu or Elizabeth A. Craig.

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