News & Views | Published:


The silence of the genes

The discovery that a transcriptional repressor is turned on in prostate tumours as they become metastatic, leading to the silencing of many genes, suggests a new mechanism for tumour progression.

Tumour cells don't usually start out bad. They generally acquire their malignant properties over years or decades, going from benign to invasive to metastatic to lethal, in a process known as tumour progression1. This process occurs through a series of mutations that result in the accumulation of aggressive characteristics by the tumour cells2. There is also evidence that large numbers of cancer-related genes can simply be switched on or off over relatively short periods of time. This suggests the existence of control molecules that can alter the expression of large groups of genes, thereby influencing tumour progression. On page 624 of this issue, Varambally and colleagues3 describe such a molecule, which may silence a whole constellation of genes during the progression of human prostate cancer.

Previously, this same group of researchers used DNA microarrays, also known as gene chips, to compare genes expressed in normal prostate glands with those in early and advanced prostate cancers4. Curiously, their results showed that far more genes were repressed than activated (or 'transcribed') as the prostate tumours became metastatic — that is, acquired the ability to spread to distant organs.

Varambally et al.3 now show that a major gene that is activated in advanced prostate cancer encodes the EZH2 protein, a known repressor of gene transcription5. They report that when the EZH2 gene is activated in prostate-cancer cells, a substantial number of other genes are shut down. If some of the products of these genes suppress tumour development, their repression by EZH2 could accelerate a tumour's progress towards metastasis (Fig. 1).

Figure 1: Tumour progression and transcriptional repressor proteins.

a, Tumour cells progress from benign, to angiogenic (meaning that they can recruit their own blood supply), to invasive (they can move through surrounding tissue), to metastatic (they can spread to other organs). At the genetic level, this progression is characterized by the net loss of functional metastasis-suppressor genes (S), together with the net gain of active tumour- or metastasis-promoting (P) genes1,2. b, Varambally et al.3 have found that the EZH2 protein is activated as prostate tumours progress towards metastasis, and that this protein represses the transcription of whole groups of other genes, many of which may inhibit tumour progression.

Varambally et al. first found that the level of messenger RNA encoding EZH2 significantly increased (indicating increased transcription of the EZH2 gene) in malignant versus benign human prostate cancers. This finding is supported by a recent report6 of a 12-fold increase in EZH2 mRNA levels in prostate-tumour metastases compared with tumours still confined to the prostate. Realizing that they had uncovered a tumour-related gene repressor, Varambally et al. then increased EZH2 levels in prostate-cancer cells in culture, and examined the consequences.

They found that the expression levels of 163 genes fell, whereas no genes appeared to be activated — consistent with a functional role for EZH2 as a transcriptional repressor in tumour cells. The authors also conducted a complementary experiment, using small interfering RNA to repress the EZH2 gene in cultured prostate-cancer cells. This resulted in a dramatic decrease in the rate of cell proliferation, showing a direct effect of EZH2 on a specific tumour-promoting function. These results imply that the upregulation of EZH2 in prostate-cancer cells leads to transcriptional repression and increased cell proliferation. Moreover, the mechanism may not be limited to prostate cancer, as EZH2 expression is also increased in certain lymphomas7.

Do the genes that are repressed by EZH2 have the potential to inhibit tumour progression? In some cases, the answer is yes. For example, a number of the EZH2-repressed genes encode putative tumour suppressors, such as the transcription factor Znfn1a1, or proteins that inactivate cell signalling, such as Rho GTPase-activating protein 1. On the other hand, some of the changes are counter-intuitive, involving genes whose functions one might expect to see upregulated as cancer progresses. These genes encode, for instance, MMP-7, an enzyme associated with the ability of tumour cells to invade surrounding tissues; and v-erb-b2/HER2, a growth-factor receptor associated with increased cell proliferation. It is important to remember that, although molecules such as EZH2 may lead to gene repression in cancer cells, gene activation also occurs during tumour progression2, and these processes are not mutually exclusive. Over time, it is the balance between enhancers and suppressors of metastasis that will determine the rate of tumour progression (see Fig. 1).

Microarray technology has revolutionized our ability to detect the temporal molecular changes that occur as cancers progress towards metastasis. Combining this technology with advances in bioinformatics provides growing sophistication for analysing such molecular patterns. What is commendable about Varambally and colleagues' approach3 is that, in addition to conducting complex pattern analysis, they have gone on to investigate the potential functions of individual genes. In the study described here, they selected one promising candidate for more intensive investigation using the tools of contemporary molecular and cellular biology. Their results provide new insight into the role of one particular protein, EZH2, in tumour biology, and in so doing reveal a potential new mechanism underlying tumour progression.

Could prostate-cancer patients benefit from this finding? Quite possibly. Although blood levels of the prostate-specific antigen (PSA) protein allow physicians to diagnose prostate cancer, this test is less useful for prognosis. What is needed is a way of distinguishing prostate cancers that are, or will become, metastatic from those that will remain localized to the prostate. The best prognostic indicators would be molecules whose levels fall or rise during the transition to metastasis — as is the case with EZH2. Significantly, Varambally and colleagues' data3 also indicate that the presence of EZH2 at the time of diagnosis correlates with future tumour recurrence. So this protein, along with other molecules such as thymosin beta 15 (ref. 8) and PTEN9, may offer physicians new predictive tools with which to guide decisions about how and when to treat prostate-cancer patients.


  1. 1

    Nowell, P. C. Science 194, 23–28 (1976).

  2. 2

    Hanahan, D. & Weinberg, R. A. Cell 100, 57–70 (2000).

  3. 3

    Varambally, S. et al. Nature 419, 624–629 (2002).

  4. 4

    Dhanasekaran, S. M. et al. Nature 412, 822–826 (2001).

  5. 5

    Laible, G. et al. EMBO J. 16, 3219–3232 (1997).

  6. 6

    LaTulippe, E. et al. Cancer Res. 62, 4499–4506 (2002).

  7. 7

    van Kemenade, F. J. et al. Blood 97, 3896–3901 (2001).

  8. 8

    Chakravatri, A. et al. Urology 55, 635–638 (2000).

  9. 9

    McMenamin, M. E. et al. Cancer Res. 59, 4291–4296 (1999).

Download references

Author information

Correspondence to Bruce R. Zetter or Jacqueline Banyard.

Rights and permissions

Reprints and Permissions

About this article

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.