Plant development

An axis of auxin

Embryos have two distinct ends, which become apparent early on. Quite how this initial polarity is sustained in plant embryos has been unclear. Step forward the agent provocateur of plant development — auxin.

In multicellular organisms, different kinds of cell are specialized for different tasks — reproduction, say, or light perception. Clearly, the proper functioning of these organisms requires that the various cell types are positioned correctly relative to one another. Hence, common to the development of all multicellular organisms is the formation of axes along which the body plan is organized. Such axes are usually manifest very early in development, and indeed it is often impossible to identify an apolar stage1. This is certainly the case in many higher plants, including the experimental organism Arabidopsis thaliana, where a developmental decision that ultimately gives rise to a shoot at one end and a root at the other can be traced back to the first, asymmetric division of the fertilized egg cell2. But how is this initial asymmetry maintained and translated into the polar root–shoot axis? On page 147 of this issue3, Friml and co-workers strikingly demonstrate that such 'elaboration' depends on the movement to and fro of the plant hormone auxin.

As Fig. 1 shows, the initial division of the fertilized Arabidopsis egg cell produces two daughter cells: a small upper cell (the apical cell) and a larger basal cell. The apical cell generates the 'proembryo', which develops through a series of stereotypic divisions to give rise to the upper, central and mid-lower regions of the seedling. Meanwhile, the basal cell produces the suspensor — a stack of cells that attaches the proembryo to maternal tissue. Initially, all of the suspensor cells lie outside the embryo proper, but later the uppermost cell is recruited by the proembryo to form the hypophysis, the founder of the lowest regions of the embryonic root2.

Figure 1: The basics of early Arabidopsis development.

a, The egg cell lies deep in the female parts of the flower, within the ovule. b, Once fertilized, the egg undergoes an asymmetric division, producing a larger basal cell and a smaller apical cell. Friml et al.3 have found that the basal cell exports the hormone auxin (red arrow) and that the apical cell responds to it. c, The basal cell divides horizontally to form the suspensor, while the auxin-receiving apical cell divides vertically to form the 2-cell proembryo. d, Around the stage of the 32-cell globular embryo, the flow of auxin is suddenly reversed3 and now accumulates in the hypophysis, the founder of the basal regions of the embryonic root. At this stage, one can begin to distinguish the regions that give rise to the apical, central and basal parts of the seedling (e).

It has been difficult to pin down how the initial asymmetry is translated into this suspensor–root–shoot axis. One contender for the role is auxin. A truly multi-talented signalling molecule, auxin has a finger in virtually every plant-developmental pie, from the control of branching in both shoot and root to the patterning of the root tip4. All of these processes depend on the cellular responses to local auxin concentrations, and on the generation of patterns of auxin accumulation. The responses to auxin involve changes in gene expression, which are mediated by families of positive and negative gene regulators whose relative abundance is tightly controlled by auxin5. The accumulation of auxin, meanwhile, is directed by two protein families: the AUX influx carriers, which pass auxin into cells, and the PIN efflux carriers, which pass it out6. Directionality of auxin transport is provided by the asymmetric localization of the PIN proteins6.

A role for auxin in elaborating the embryonic axis has been suspected for many years, not least because polarity defects can be induced in embryos by blocking auxin movement7. More recent molecular genetic evidence comes from the analysis of mutations in three Arabidopsis genes, MONOPTEROUS (MP), BODENLOS (BDL) and GNOM (GN), all of which cause defects in axis elaboration. MP encodes a positive and BDL a negative regulator of auxin-inducible genes8,9, whereas GN encodes a protein needed for the proper subcellular targeting of the PIN proteins10,11. What is exciting about the work of Friml et al.3 is that it ties these strands together definitively.

To follow the responses of embryos to auxin, the authors introduced a gene for green fluorescent protein (GFP), attached to a synthetic control region that is activated by auxin. They found that the first signs of a response (that is, the first signs of GFP production and hence auxin activity) occur at a very early stage, in the apical daughter cell of the asymmetrically divided egg. As the proembryo develops, the auxin-response signal persists, remaining absent from the suspensor cells beneath. This pattern continues until the proembryo consists of about 32 cells, when a remarkable thing happens. The axis of auxin response is suddenly reversed, becoming undetectable in the apical regions, with a new maximum in the developing hypophysis beneath. The authors find that these response patterns reflect actual gradients of auxin concentration, with maximal responses occurring at maximal concentrations.

To find out how these gradients arise, Friml et al. investigated the expression patterns of the PIN auxin-efflux proteins. They show that, within the two-cell proembryo, a previously uncharacterized PIN-family member — PIN7 — is expressed in the basal cell at the boundary facing the cell's apical sister. This is consistent with the auxin maximum in the apical cell. Later, the cells of the suspensor continue to express PIN7 at their apical side, while in the proembryo another protein, PIN1, is expressed without apparent polarity10.

But at the 32-cell stage, it's all change. Both PIN1 and PIN7 become localized to the basal membranes of the cells in which they are expressed (proembryo and suspensor cells, respectively) — an event that coincides with the reversal of the auxin gradient and, presumably, with the onset of auxin production in the proembryo. This new direction of auxin flow is apparently reinforced by the expression of two other PIN-family members, so that, although PIN7 is positioned to transport auxin down the suspensor and out of the embryo, the net effect is the accumulation of auxin, and a maximal response to it, in the embryonic root.

Do these auxin fluxes help to maintain the embryonic axis of polarity? The authors' studies of mutant plants show that they do. Embryos with mutations in PIN7 have trouble establishing the initial auxin-response maximum in the apical cell and its daughters, and this coincides with a confused apical/basal identity in the proembryo. The effects of mutations in other PIN proteins are milder and affect later stages of basal embryo development. But matters are not entirely predictable. Knowing the PINs' expression patterns, if you were going to put money on any of these mutants not making it out of embryogenesis it would be on those with PIN7 defects — yet these plants recuperate to produce relatively normal seedlings. Interestingly, the recovered axis is always in the correct orientation, hinting that the polarizing influence of the suspensor is maintained into these later stages. It is not clear how polarity is restored, but it does require auxin efflux, as plants with mutations in all four embryonically expressed PINs fail to recover. These data, together with the requirement for an embryonic response to auxin implicit in the defects caused by MP and BDL mutations, underline the relevance of asymmetric auxin transport and responses in maintaining polarity.

So does auxin do it all? Is the initial polarizing signal from maternal tissue — the signal that directs the first, asymmetric cell division — also auxin? And can auxin itself direct the polar localization of PIN proteins? It is certainly possible that the initial asymmetry is directed by a low level of auxin flow from maternal tissues that passes beneath the radar of current techniques, and that maternal auxin, channelled up the suspensor, could continue to provide axial information later on. Furthermore, auxin has for many years been proposed to act in feedback loops to regulate its own flux6. But it is harder to envisage a mechanism whereby auxin could trigger the dynamic changes in its direction of flow that are observed here.

Are there then other signals, and, if so, what is their relationship to auxin? Is auxin instructive or permissive? If the former, auxin must first instruct apical, and then, almost immediately, basal fates. As this would require that auxin concentration is maintained and interpreted with the utmost rigour, auxin might instead permit development, within an apical or basal context set by additional input. At least we are now in a position to muse on these possibilities, thanks to the work of Friml and colleagues.


  1. 1

    Wolpert, L. Principles of Development (Oxford Univ. Press, 2002).

    Google Scholar 

  2. 2

    Jürgens, G. EMBO J. 20, 3609–3616 (2001).

    Article  Google Scholar 

  3. 3

    Friml, J. et al. Nature 426, 147–153 (2003).

    ADS  CAS  Article  Google Scholar 

  4. 4

    Berleth, T. & Sachs, T. Curr. Opin. Plant Biol. 4, 57–62 (2001).

    CAS  Article  Google Scholar 

  5. 5

    Kepinski, S. & Leyser, O. Plant Cell 14, S81–S95 (2002).

    CAS  PubMed  PubMed Central  Google Scholar 

  6. 6

    Friml, J. Curr. Opin. Plant Biol. 6, 7–12 (2003).

    CAS  Article  Google Scholar 

  7. 7

    Hadfi, K., Speth, V. & Neuhaus, G. Development 125, 879–887 (1998).

    CAS  PubMed  PubMed Central  Google Scholar 

  8. 8

    Hardtke, C. S. & Berleth, T. EMBO J. 17, 1405–1411 (1998).

    CAS  Article  Google Scholar 

  9. 9

    Hamann, T. et al. Genes Dev. 16, 1610–1615 (2002).

    CAS  Article  Google Scholar 

  10. 10

    Steinmann, T. et al. Science 286, 316–318 (1999).

    CAS  Article  Google Scholar 

  11. 11

    Geldner, N. et al. Cell 112, 219–230 (2003).

    CAS  Article  Google Scholar 

Download references

Author information



Corresponding authors

Correspondence to Stefan Kepinski or Ottoline Leyser.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Kepinski, S., Leyser, O. An axis of auxin. Nature 426, 132–135 (2003).

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.


Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing