A shocking protein complex

Heat-shock proteins have been found to form part of a large protein complex, called the epichaperome, that improves the survival of some cancer cells. This complex might offer a new target for cancer treatment. See Letter p.397

Tumour cells are subject to various forms of stress, such as oxygen or nutrient shortages, when tumour blood-vessel formation cannot keep pace with tumour growth — and the cells therefore need to have effective stress-survival strategies. Heat-shock proteins (HSPs) are often active in cells exposed to stressful conditions. On page 397, Rodina et al.1 have investigated the role of HSPs in human cancer, and find that the proteins exist in a large complex in some cancer cells.

Maintenance of the correct 3D structure of a protein is essential for its function, and cells have mechanisms dedicated to protein quality control. Proteins with structural defects are either refolded into the correct conformation or, in the case of severe structural abnormality, targeted for degradation. Protein quality control is regulated by chaperone proteins, which are often involved in facilitating folding. Chaperones include HSPs, which are active in human cells both under normal conditions and in stressful conditions such as inflammation and a state of reduced oxygen known as hypoxia.

Rodina and colleagues investigated HSPs in human cancer cells using a biochemical technique that separates proteins in samples of cellular proteins. The authors found that, in samples of what they call 'type 2' cancer cells, the heat-shock protein HSP90 separated as expected on the basis of its structure. However, in other samples tested, HSP90 had an unexpected separation pattern that could be explained by its presence in a large complex with other HSPs. The authors use the term 'type 1' cancer cells to describe cells containing this large protein complex (Fig. 1).

Figure 1: The epichaperome in cancer cells.

Rodina et al.1 have found that, in certain cancer cells (which the authors call type 2 cells), heat-shock proteins such as HSP90, and other proteins associated with the response to cellular stress such as chaperones, are present as solitary proteins or in small complexes. However, other cancer cells (referred to as type 1 cells) show the formation of a large protein complex called the epichaperome, which contains HSP90 and other proteins, including chaperones. Increasing the levels of MYC protein in type 2 cells induced the formation of the epichaperome. When type 1 cells were treated with the HSP90 inhibitor PU-H71 in vitro, the epichaperome disintegrated and the cells died. The epichaperome might thus represent a target for treating cancer.

The researchers found that, in type 1 cancer cells, HSP90 associated with dozens of other proteins, including scaffolding and adaptor proteins, which are also involved in regulating protein folding. The authors called this structure the epichaperome. By contrast, in type 2 cancer cells and non-cancer cells, HSP90 was associated with only a small set of proteins, and most of the HSPs existed as solitary proteins or were assembled into small complexes.

HSP90 requires the essential nucleotide ATP to function. The authors therefore targeted the epichaperome using the molecule PU-H71, which binds to HSP90's ATP-binding pocket and inhibits the protein's function. The inhibitor bound more tightly when HSP90 was in the epichaperome, and killed more type 1 cancer cells than type 2 or non-cancer cells. This difference was not attributable solely to chemical inactivation of HSP90, because the authors found that selective genetic downregulation of other HSPs resulted in increased death of type 1 cells compared with type 2 cells, suggesting that type 1 cell survival depends on an intact epichaperome.

Using protein-complex analysis techniques and PU-H71 treatment, Rodina and colleagues investigated the epichaperome in different types of cancer cell. The epichaperome was detected in 60–70% of cell lines from breast, pancreatic, lung, leukaemia and other cancers, indicating the potential clinical relevance of this protein complex as a therapeutic target. Strikingly, the epichaperome was not restricted to a particular cancer subset characterized by specific gene or protein expression. Type 1 cells also had high levels of the cancer-associated MYC protein. When MYC was downregulated, the epichaperome disappeared, whereas overexpressing MYC in type 2 cells induced epichaperome formation. Thus, MYC is an important regulator of epichaperome assembly.

Treatments for cancer require a target that is present in most cancer cells but absent in normal cells. Rodina and colleagues' finding that type 1 cancer cells contain HSP90 in the epichaperome, but that normal cells have HSP90 mainly in a non-complexed form, suggests that the epichaperome might be a clinical target.

Despite the encouraging results, the authors found considerable differences in dependence on epichaperome formation, both between tumour types and between tumour cells of the same type (for example, breast cancer), which might allow the emergence of treatment-resistant variants. Moreover, in many cancers, such as breast cancer, the metastatic migration of cancer cells to other locations in the body is the main cause of death, and metastatic cells may differ from the initial primary tumours. It remains to be determined to what extent metastatic cells depend on epichaperome formation, and how such cells might respond to drugs that target the complex.

However, obtaining metastatic tissue from a patient by needle biopsy is a problem because of the procedure's invasive nature, and some lesions (for example, in the lung, bone and brain) are difficult to access. 'Liquid biopsy' of circulating tumour cells in the peripheral blood might be an alternative feasible strategy for assessing the epichaperome in metastatic cells. Analysis of circulating tumour cells or tumour cells that have migrated to the bone marrow in people in the early stages of cancer might also provide information on potential precursors of metastases2,3.

Rodina and colleagues' findings could have broader implications. Many cellular programs consist of signalling pathways that rely on a large number of proteins, and these might form stable complexes in a similar way to how the epichaperome forms. For example, the chaperone machinery in a cellular organelle called the endoplasmic reticulum can form large protein complexes4, and the activity of endoplasmic reticulum chaperones under conditions of cellular stress strongly affects cancer-cell survival5. Moreover, cancer-associated ErbB receptors might assemble into higher-order receptor complexes6. Thus, the discovery of other large, cancer-specific protein complexes might open further avenues of investigation for understanding cancer biology, with potential implications for the design of new strategies for cancer therapy.Footnote 1


  1. 1.

    See all news & views


  1. 1

    Rodina, A. et al. Nature 538, 397–401 (2016).

    ADS  CAS  Article  Google Scholar 

  2. 2

    Alix-Panabières, C. & Pantel, K. Nature Rev. Cancer 14, 623–631 (2014).

    Article  Google Scholar 

  3. 3

    Bartkowiak, K. et al. Cancer Res. 75, 5367–5377 (2015).

    CAS  Article  Google Scholar 

  4. 4

    Meunier, L., Usherwood, Y.-K., Chung, K. T. & Hendershot, L. M. Mol. Biol. Cell 13, 4456–4469 (2002).

    CAS  Article  Google Scholar 

  5. 5

    Lee, A. S. Nature Rev. Cancer 14, 263–276 (2014).

    ADS  CAS  Article  Google Scholar 

  6. 6

    Sweeney, C. & Carraway, K. L. III Oncogene 19, 5568–5573 (2000).

    CAS  Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Klaus Pantel.

Related links

Related links

Related links in Nature Research

Molecular biology: Choose your protein partners

Chemical biology: Many faces of a cancer-supporting protein

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Bartkowiak, K., Pantel, K. A shocking protein complex. Nature 538, 322–323 (2016).

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.