Plant biology

Fixation with regulation

A gene has been isolated that controls the number of symbiotic nitrogen-fixing nodules in legumes. Its similarity to a well-characterized regulatory gene in Arabidopsis provides clues about its action.

Leguminous plants produce root nodules, within which symbiotic bacteria capture atmospheric N2 and convert it into nitrogen that can be used by the plant. But this process is energetically expensive and so legumes strictly control the numbers of nodules they form. Papers by Krusell et al.1 and Nishimura et al.2 (pages 422 and 426 of this issue), and by Searle et al.3 in Science, describe the characterization of a regulatory gene that normally limits nodule numbers, and that when mutated increases nodulation. Control of nodule development is of interest in its own right, but this work may also have agricultural applications.

Soybean and pea mutants with enhanced nodulation have been known for about 20 years4, but their complex genomes have hampered attempts to clone the genes responsible. Similar mutants5,6 were recently identified in Lotus japonicus, a legume with a relatively small genome. These mutants have hypernodulation and aberrant roots, hence their designation as har mutants. The numbers of both nodules and lateral roots are increased in these L. japonicus and soybean mutants, indicating that normal legumes possess a common regulatory mechanism that limits the numbers of root and nodule growing points, or meristems.

Grafting experiments showed that HAR1 control of nodule and lateral-root number in L. japonicus depends on the shoots rather than the roots (Fig. 1), a characteristic that had been previously observed with soybean hypernodulation mutants4. So plants with har1-mutant shoots grafted onto wild-type roots had increased root nodulation. In contrast, the reciprocal grafted plants (mutant roots with wild-type shoots) had normal roots and nodules. After positioning HAR1 on a physical map of the L. japonicus genome, two groups1,2 cloned the gene. They went on to identify mutations in the equivalent genes in pea1 and soybean2 hypernodulation plants, which showed a similar shoot control of nodulation3. Independently, following about 15 years of work3, the equivalent gene controlling nodulation in soybean was isolated and was called NARK (nodule autoregulation receptor kinase). It is clear that HAR1 and NARK are the same genes from different species.

Figure 1: Control of nodule number in legumes.

One proposal4 is that a factor produced by nodule and root meristems is detected in shoots, resulting in the production of a second factor that suppresses the development of new nodules. In line with this model, mutation of the HAR1/NARK gene increases the numbers of nodules and lateral-root meristems1,2,3. So HAR1/NARK protein is required for the production of the shoot-derived inhibitor, possibly through its involvement in perception of a nodule- and root-derived signal.

HAR1 and NARK encode a type of receptor protein that is abundant in plants7 and has three components: an extracellular domain of leucine-rich repeats, a membrane-spanning domain, and an intracellular protein kinase domain (Fig. 2). This structure is compatible with the receptor's function being perception of a ligand outside the cell, followed by internal signal transduction through protein phosphorylation by the kinase domain. Mutant genes sequenced from the three species had alterations in the kinase domain, suggesting that phosphorylation by HAR1/NARK protein is essential for signalling. The signal perceived by HAR1/NARK is unknown, as is the kind of inhibitor that suppresses further nodulation.

Figure 2: How HAR1/NARK might function in signalling and root nodulation.

This scheme is based on the model8 for signalling by CLAVATA proteins in Arabidopsis, and on the characteristics of HAR1/NARK described by Krusell et al.1, Nishimura et al.2 and Searle et al.3. HAR1 is most similar to CLAVATA (CLV) 1, a receptor kinase containing leucine-rich repeats which interacts with CLV2, forming a complex that recognizes the signalling peptide CLV3. CLV2 contains leucine-rich repeat and transmembrane domains, but lacks a kinase domain. HAR1/NARK might interact with a CLV2-like protein, yet to be identified, to recognize a signal produced by root and nodule meristems.

Comparison with the model plant Arabidopsis provides insight into how HAR1/NARK might function, even though Arabidopsis does not form nodules. The Arabidopsis genome is predicted to contain 216 leucine-rich-repeat, receptor-like kinases7. They are believed to perceive extracellular signals such as bacterial flagellin, the wounding signal systemin, and the plant hormones brassinolide and phytosulphokine. A real surprise from the new work1,2,3 is that the legume HAR1/NARK is more similar in sequence to Arabidopsis CLAVATA1 than to any other Arabidopsis protein, implying that, like CLAVATA1, it functions as a receptor kinase. Together with CLAVATA2, CLAVATA1 forms a complex that detects the signal peptide CLAVATA3 (Fig. 2). The CLAVATA genes are involved in regulating cell fate in the shoot apical meristems8, clavata1 mutants having an enlarged meristematic zone that leads to fasciation (contiguous parts growing into one).

CLAVATA1 and HAR1/NARK appear to have similar functions, because mutation of the corresponding genes results in increased meristematic activity (Fig. 2), either in shoot apical meristems (clavata1) or in root and nodule meristems (har1/nark). However, they must have different roles, because the roots of clavata1 mutants are unaffected, as are the apical meristems of nark mutants3. Moreover, their patterns of expression are different. CLAVATA1 is exclusively expressed in apical meristems8, whereas HAR/NARK seem to be expressed in most tissues except apical meristems2,3. It will be interesting to see how far the different functions of HAR1/NARK and CLAVATA1 can be attributed to their expression patterns.

Grafting and other experiments with soybean hypernodulation mutants led to a model of how legumes might regulate nodule number4. This postulates that a signal moves from the root to the shoot, and that increased root nodulation is detected in shoots as the root-derived signal increases. As a result, the shoots produce an inhibitor, which is translocated to the root, where it represses further nodule development. Based on the grafting experiments and the nature of HAR1/NARK, it seems that this protein could be involved in the perception of a root-made signal in the shoot (Fig. 1). An unexpected result is that all three groups1,2,3 detected HAR1/NARK messenger RNA in roots, implying that the protein is being made there. But although the gene is transcribed in roots, the grafting experiments show that the root-expressed gene alone cannot function in nodule repression. So some additional factor in the aerial part of the plant may be required for HAR1/NARK to act.

There is a second shoot-controlled hypernodulation gene in pea that, when mutated, can lead to shoot fasciation9. Given the similar fasciation seen with the clavata mutants, there may be some functional overlap of the CLAVATA and HAR1/NARK pathways. It seems likely that the product of this second gene in pea could be part of the same signalling pathway as HAR1/NARK. Identification of the gene will be required to see if its product can interact directly with HAR1/NARK (Fig. 2) and what relationship there is between the HAR1/NARK and CLAVATA pathways. Genes with high sequence similarity to CLAVATA1 have been identified in soybean10, and so HAR1/NARK may have arisen as a duplication of an ancestral CLAVATA1- like gene, with a subsequent divergence of functions.

Finally, the results of Krusell et al.1, Nishimura et al.2 and Searle et al.3 have practical implications. One of the problems with most of the existing hypernodulation mutants is that the abnormal root and nodule development has severe effects on plant growth. However, Searle et al.3 describe one mutation in NARK that has less detrimental effects. A different L. japonicus gene encoding a transcriptional regulator that represses nodulation has also just been described11, but its relationship to the HAR1/NARK receptor pathway is not known. Now that key genes regulating nodulation have been isolated, it may be possible to identify plants carrying subtle mutations — these might allow increased nodulation, with consequent increases in nitrogen fixation, without affecting plant growth too badly. Such plants might show improved growth, and also leave residual nitrogen in the soil to increase the growth of subsequent crops.


  1. 1

    Krusell, L. et al. Nature 420, 422–426 (2002).

    ADS  CAS  Article  Google Scholar 

  2. 2

    Nishimura, R. et al. Nature 420, 426–429 (2002).

    ADS  CAS  Article  Google Scholar 

  3. 3

    Searle, I. R. et al. Science published online 31 October 2002 (doi: 10.1126/science.1077937).

  4. 4

    Caetano-Anolles, G. & Gresshoff, P. M. Annu. Rev. Microbiol. 45, 345–382 (1991).

    CAS  Article  Google Scholar 

  5. 5

    Wopereis, J. et al. Plant J. 23, 97–114 (2000).

    CAS  Article  Google Scholar 

  6. 6

    Kawaguchi, M. et al. Mol. Plant–Microbe Interact. 15, 17–26 (2002).

    CAS  Article  Google Scholar 

  7. 7

    Shiu, S.-H & Bleecker, A. B. Proc. Natl Acad. Sci. USA 11, 10763–10768 (2001).

    Article  Google Scholar 

  8. 8

    Clark, S. E. Nature Rev. Mol. Cell Biol. 2, 276–284 (2001).

    CAS  Article  Google Scholar 

  9. 9

    Sagan, M. & Duc, G. Symbiosis 20, 229–245 (1996).

    Google Scholar 

  10. 10

    Yamamoto, E., Karakaya, H. C. & Knap, H. T. Biochim. Biophys. Acta 1491, 333–340 (2000).

    CAS  Article  Google Scholar 

  11. 11

    Nishimura, R., Ohmori, M., Fujita, H. & Kawaguchi, M. Proc. Natl Acad. Sci. USA 99, 15206–15210 (2002).

    ADS  CAS  Article  Google Scholar 

Download references

Author information



Rights and permissions

Reprints and Permissions

About this article

Cite this article

Downie, J., Parniske, M. Fixation with regulation. Nature 420, 369–370 (2002).

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.


Sign up for the Nature Briefing newsletter for a daily update on COVID-19 science.
Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing