News & Views | Published:

Cell biology

Thanks for the memory

In response to a transient hormonal cue, a developing egg commits irreversibly to a mature state. Surprisingly, this irreversible switch is composed of intrinsically reversible components.

Few decisions in life are truly irrevocable, even at the cellular level. Yet there are times when it's crucial not to turn back — during the development of multicellular organisms, for instance, when cells must make an irreversible commitment to a new fate in response to a transient stimulus. This process, called differentiation, is well known, but it's not really clear how a cell 'remembers' its commitment long after the signal, frequently a hormone, has disappeared. On page 460 of this issue, Xiong and Ferrell1 show how brief exposure to the hormone progesterone triggers an irreversible switch in cell fate in developing frog eggs. The memory is sustained by positive feedback loops in the underlying control system. When the positive feedback is perturbed, the cell does the unthinkable and 'dedifferentiates', reverting to characteristics of the immature egg.

For more than 30 years, investigation of the maturation of frog eggs has been fertile ground for discoveries about how cells divide and differentiate. In 1971, Masui and Markert2 demonstrated that a bit of cytoplasm taken from a mature frog egg and injected into an immature egg would induce maturation, replacing the need for progesterone. The authors named this mysterious activity MPF, for maturation-promoting factor. MPF was purified in 1988, and was shown to consist of two proteins — an enzyme called Cdc2, which attaches phosphate groups to other proteins, and an unstable co-factor called a cyclin, which is required for Cdc2 activity3,4. A few years later, the activity of a second enzyme, MAPK, was shown to be required for egg maturation5. MAPK also phosphorylates proteins.

Subsequently, intensive biochemical and molecular investigations by many laboratories have produced a detailed picture of the signalling cascades that are elicited in an immature egg by progesterone. The picture has become quite complicated, with more than 30 different molecules involved. In fact, a recent comprehensive review of the literature6 required a poster-sized insert to depict all of the known biochemical responses of eggs to progesterone. In an attempt to sort through this web of molecular information, most scientists organize their thoughts around the two key enzymes — Cdc2 and MAPK — and the signalling events that activate each of them7.

Organizing the network in this way has helped investigators to appreciate the extensive positive feedback that is inherent in the molecular control system underlying egg maturation (see Fig. 1 on page 461). Positive feedback occurs when a molecule that is activated by a signalling pathway activates some earlier step in that pathway, thus strengthening and perpetuating the signal. Cdc2 activates itself in many ways: by inducing the synthesis of cyclin, by phosphorylating and turning off a Cdc2 inhibitor, and by phosphorylating and turning on a Cdc2 activator. Likewise, MAPK triggers the synthesis of Mos, a protein upstream in the MAPK activation cascade. Furthermore, the Cdc2 and MAPK pathways activate each other. Except for the synthesis of a few key proteins such as Mos and cyclin, most of the signalling and feedback events in this system consist of phosphorylation6.

Despite this wealth of molecular information about egg maturation, and the appreciation of positive feedback in the molecular circuitry, until now no one could explain how eggs persist in the mature state — with high Cdc2 and MAPK activity — long after progesterone has been removed. Ferrell and Xiong previously carried out theoretical studies of the problem8, and found that positive feedback such as that seen in the Cdc2 and MAPK cascades could create a biochemical 'memory'. This prediction was certainly not intuitive, as these feedback events operate largely by phosphorylation — known to be rapid and readily reversible. Was it not more likely that a long-term response to a transient signal would be maintained by an irreversible event, such as the destruction of a key regulatory protein, maybe an inhibitor of Cdc2 or MAPK?

It appears not. Proof of Ferrell and Xiong's theoretical prediction required careful experimentation. In particular, the prediction demands that inhibiting the positive feedback will cause a mature egg to 'de-mature', at least with respect to key biochemical events. The authors now have compelling evidence to support this idea1. They find that three different treatments that block the positive feedback between MAPK and Mos result in transient, rather than sustained, activation of both MAPK and Cdc2. So, positive feedback is necessary to maintain the biochemical memory of a mature egg (Fig. 1). Xiong and Ferrell also provide an additional, theoretical explanation for their data, with computational simulations that show how an essentially reversible switch can be rendered irreversible depending on the strength of the positive feedback in the system.

Figure 1: No turning back.
figure1

When a hormone stimulates an egg cell to mature, the switch in cell fate is irreversible even after the hormonal signal is gone. Xiong and Ferrell1 show that this long-term cellular memory is sustained by strong positive feedback in the underlying molecular control system.

Beyond its particular contributions regarding the maturation of eggs, the work of Xiong and Ferrell1 should have much broader impact owing to the approach they present for addressing questions of cellular function. As in work regarding the cellular switch that controls entry into, and exit from, nuclear division9,10, Xiong and Ferrell's experiments were driven by theoretical predictions about the fundamental mechanisms that control such switches. These predictions derive from a systems-level view, but require quantitative, reductionist-style experimentation to test their validity. The ability to think and work on such a broad scale, from system through to molecule, is the most impressive aspect of Xiong and Ferrell's work. The notion of pairing theoretical and computational biology with experimental cell biology should catch on, and the positive feedback inherent in this interdisciplinary brand of science is likely to drive many breakthroughs in our understanding of cellular controls.

References

  1. 1

    Xiong, W. & Ferrell, J. E. Jr Nature 426, 460–465 (2003).

  2. 2

    Masui, Y. & Markert, C. L. J. Exp. Zool. 177, 129–145 (1971).

  3. 3

    Gautier, J., Norbury, C. H., Lohka, M., Nurse, P. & Maller, J. L. Cell 54, 433–439 (1988).

  4. 4

    Gautier, J. et al. Cell 60, 487–494 (1990).

  5. 5

    Kosako, H., Gotoh, Y. & Nisheda, E. EMBO J. 13, 2131–2138 (1994).

  6. 6

    Schmitt, A. & Nebreda, A. R. J. Cell Sci. 115, 2457–2459 (2002).

  7. 7

    Abrieu, A., Doree, M. & Fisher, D. J. Cell Sci. 114, 257–267 (2001).

  8. 8

    Ferrell, J. E. Jr & Xiong, W. Chaos 11, 227–236 (2001).

  9. 9

    Pomerening, J. R., Sontag, E. D. & Ferrell, J. E. Jr Nature Cell Biol. 5, 346–351 (2003).

  10. 10

    Sha, W. et al. Proc. Natl Acad. Sci. USA 100, 975–980 (2003).

Download references

Author information

Rights and permissions

Reprints and Permissions

About this article

Further reading

Comments

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.