Skip to content

Another gamma paper

January 27, 2013

In my last post, I referred to a paper about gammaproteobacteria that were not rhizobia.  Here is another one (Huang et al. 2012).  The difference is that these authors try to persuade us that their gammaproteobacterium is a rhizobium.  I am not convinced.

It is not that I don’t believe that gamma-rhizobia can exist.  Indeed, I argued in an earlier post that they probably do.  The discovery of beta-rhizobia in 2001 opened our eyes to the possibility of legume symbionts that were not alphaproteobacteria, and after more than a decade of further research, we now know a fair amount about beta-rhizobia in the genera Burkholderia and Cupriavidus (Gyaneshwar et al. 2011).

Huang et al. start from an unusual perspective on our current knowledge of the taxonomic diversity of rhizobia.  They introduce their paper like this:

Currently 56 species, 11 genera symbiotic nitrogen-fixing bacteria have been identified, including not only traditional rhizobia that belong to genera Rhizobium, Sinorhizobium, Mesorhizobium, Bradyrhizobium, Azorhizobium and Agrobacterium, but also some symbiotic nitrogen-fixing bacteria that have previously been categorized as non-symbiotic: some species in Phyllobacterium, Ochrobacterum, and Methylobacterium can all grow symbiotically with legumes and fix nitrogen [2–6].

In their Discussion section, their list is even more restricted:

Currently there are six genera of bacteria that can nodulate legumes, including Rhizobium, Sinorhizobium, Allorhizobium, Mesorhizobium, Bradyrhizobium, and Azorhizobium [24].

At the end of this post, I have listed the references that Huang et al. used to support these statements. You will notice that references [2-4] are all descriptions of beta-rhizobia.  Reference [24] was published in 2004, so should have covered beta-rhizobia, but it is, unfortunately, only available in China.

It is clear that Huang et al. cannot have been unaware of the existence of beta-rhizobia, so it is a mystery why they chose to air-brush them out of the picture. We must focus, though, on their claim to have discovered the first gamma-rhizobium, which is not dependent on their interpretation of history.

The paper describes a bacterial strain, D5, isolated from nodules of Acacia confusa in Guangxi, south China.  Inoculation with D5 resulted in N-fixing nodules on several Acacia species, but not on soybean.  However, the nodules were formed 90 days after inoculation, which is very late.  PCR based on a culture of D5 yielded a 16S rRNA gene sequence that is indisputably from a species of Pseudomonas, although it has no exact match in GenBank.  PCR also yielded partial sequences of nodA and nifH genes, and these were similar to those previously reported from some Bradyrhizobium strains.

For nodA, the top three hits are strains from Uraria in Australia, Phaseolus in the USA, and Abrus in Senegal.  Not much clue there about host specificity, therefore.  The only common link is that all three were from different studies authored by Tomasz Stępkowski, but I think that just reflects Tomasz’s expertise in exotic bradyrhizobia.  The nodA sequence from D5 has a frameshift caused by two extra bases about 30 bases from the start.  This would make the gene nonfunctional, of course, but it might just be a sequencing error.

The nifH sequence from D5 has a 33-base deletion and a frameshift when compared to other nifH sequences, but the alignable bases are 100% identical to those of Bradyrhizobium sp. SEMIA 695 from Neonotonia (a relative of soybean), and similar to those of many other bradyrhizobia.

How can we interpret these results?  Clearly, the predominant organism in the D5 culture was a Pseudomonas, but there was something in there that had bradyrhizobial nod and nif genes and probably formed the nodules on Acacia.  Pseudomonads grow rapidly and bradyrhizobia very slowly, so it would be easy for a few Bradyrhizobium cells to remain undetected in a Pseudomonas culture.  The plant would then select these few cells, though nodulation might be delayed because the effective inoculum was very dilute.  Unfortunately, Huang et al. did not address Koch’s fourth postulate, which is that the organism that can be isolated at the end of the infection should be the same as the one that was introduced.  It remains possible that the nodules they saw were filled only with bradyrhizobia.  Even if there were pseudomonads in the nodules, though, it would not prove that those pseudomonads caused the formation of the nodule.  The claim for nodulating pseudomonads that Huang et al. make is essentially the same as that of Benhizia et al. (2004)*.  That claim was later retracted by Muresu et al. (2008), who documented extensive colonisation of nodules by non-nodulating gammaproteobacteria.

Most of us who isolate bacteria from wild nodules know that there are frequently “contaminants”, i.e. bacteria that are not capable of inducing nodule formation but grow within nodules that are induced by other bacteria (the actual rhizobia).  Sometimes these free-loaders are the most abundant bacteria in cultures from the nodule, and it can be hard to separate them from the true rhizobia even when taking what appear to be single colonies from plates.  How can we address this problem?  Fortunately, there is a simple strategy that can help.  Most bacteria can mutate spontaneously to a high level of resistance to the antibiotic streptomycin, though the frequency is normally less than one in a million cells.  Plating a culture on streptomycin allows only these rare mutants to grow and form colonies.  Since it is unlikely that two different bacteria will mutate at the same spot on the plate, the colonies that appear are likely to be pure.  I would predict that, if one plated the D5 culture on streptomycin and waited patiently for ten days, one would see small bradyrhizobial colonies in places where they had not been overgrown by the pseudomonad.

I still believe that there are gammaproteobacteria that can nodulate legumes, and we will find them one day.  That day has not yet arrived, though.

* Lionel Moulin mentioned this work in a comment on my earlier post.


Huang, B., Lv, C., Zhao, Y., and Huang, R. A novel strain D5 isolated from Acacia confusa. PLoS ONE 7: e49236.

Gyaneshwar, P., Hirsch, A.M., Moulin, L., Chen, W.-M., Elliott, G.N., Bontemps, C. et al. (2011) Legume-nodulating betaproteobacteria: diversity, host range and future prospects. Molecular Plant-Microbe Interactions.

Benhizia, Y., Benhizia, H., Benguedouar, A., Muresu, R., Giacomini, A., and Squartini, A. (2004) Gamma proteobacteria can nodulate legumes of the genus Hedysarum. Systematic and Applied Microbiology 27: 462-468.

Muresu, R., Polone, E., Sulas, L., Baldan, B., Tondello, A., Delogu, G. et al. (2008) Coexistence of predominantly nonculturable rhizobia with diverse, endophytic bacterial taxa within nodules of wild legumes. FEMS Microbiology Ecology 63: 383-400.

Selected references from the paper by Huang et al.

2.  Moulin L, Munive A, Dreyfus B, Boivin-Masson C (2001) Nodulation of legumes by members of the beta-subclass of proteobacteria. Nature, 411: 948–950.

3.  Chen WM, Laevens S, Lee TM, Coenye T, DeVos P, et al. (2001) Ralstonia taiwanensis sp. nov., isolated from root nodules of Mimosa species and sputum of a cystic fibrosis patient. Int. J. Syst. Evol. Microbiol. 51(5): 1729–1735.

4.  Vandamme P, Goris J, Chen WM, de Vos P, Willems A (2002) Burkholder-iatuberum sp. nov. and Burkholderia phymatum sp. nov., nodulate the roots of tropical legumes. Syst. Applied Microbiology, 25(4): 507–512.

5.  Zurdo-Pineiro JL, Rivas R, Trujillo ME, Vizcaino N, Carrasco JA, et al. (2007) Ochrobactrum  cytisi  sp.  nov.,  isolated  from  nodules  of  Cytisus  scoparius  in Spain. Int J Syst Evol Microbiol, 57(4): 784–788.

6.  Mantelin S, Saux MF, Zakhia F, Béna G, Bonneau S, et al. (2006) Emended description of the genus Phyllobacterium and description of fournovel species associated with plant roots: Phyllobacterium bourgognense sp. nov., Phyllobacterium  ifriqiyense  sp.  nov.,  Phyllobacterium  legum  inum  sp.  nov.  and Phyllobacterium brassicacearum sp. nov. Int J Syst Evo Microbiol, 56(4): 827– 839.

24.  Chen WX, Wang ET, Chen WF (2004) The relationship between the symbiotic promiscuity  of  rhizobia  and  legumes  and  their  geographical  environments. Scientia Agricultura Sinca, 37(1): 81,86.


From → Papers, Taxonomy

  1. I totally agree with your comments. And I can’t believe such story ends in Plos One.

    • Tomasz Stepkowski permalink

      It seems that this rhizobium strain may belong to the genus Bradyrhizobium, but to a very slow-growing lineage. I am rather skeptical that gamma-proteobacteria can form nodules on legumes. I remember that in 1988 or 89 there was a paper from Sharon Long lab describing the lack of nod gene expression in some gamma-proteobacteria spp (E.coli, Pseudomonas, Xanthomonas). Later, Sharon was trying to identify a sigma factor responsible for nod gene expression in Sinorhizobium strain. I think that she has failed in identification of such sigma factor which would control the nodulation process.


      • Thank you, Tomasz! You make a good point about gene expression – I also remember that there were studies that failed to express nod genes in gammaproteobacteria. However, my guess would have been that this was also true for betaproteobacteria, which are more similar in most respects to gammaproteobacteria than they are to alphaproteobacteria. They share many mobile elements, for example, such as IncP plasmids and some genomic islands, and the range of ecological niches occupied by Burkholderia and Pseudomonas is rather similar. I haven’t looked at the literature on sigma factors and gene expression in Burkholderia. Are there differences that would lead you to predict that Burkholderia could nodulate but Pseudomonas could not? Or is it just a case of putting the right promoter sequences in front of the genes?


  2. Tomasz Stepkowski permalink

    Peter, I must admit that I have never heard about such constraints in beta-proteobacteria. Certainly, nod genes may not be expressed in Neisseria. In my opinion, the problem is quite general and concerns the factor(s) limiting the distribution of symbiotic (nodulation) genes in bacteria. It seems to me that it is not lateral transfer, which can donate genes crossing inter-Kingdom barriers. Such limits are rather physiological, and could be related to nod gene expression. Apparently, some bacteria have predispositions enabling expression of nodulation loci – a phenomenon which has not been elucidated even with the advent of OMICS-related approaches. The other symbiotic predisposition is the capability to occupy and live within eukaryotic cell. It is not accidental that most (if not all) alpha-proteobacterial pathogens are intracellular. This shows that despite the substantial progress that has been made there are still big gaps in our understanding of symbiosis even from rhizobium perspective. I think that a right person to comment this issue could be Sharon Long.


  3. Dear Prof. Young,
    If we look at the gel picture of (Figure 2. PCR results of 16S rRNA of strain D5), there are two bands there. So there you may be right about contamination in the culture. I have observed that, generally the PCR product of alphaproteobacteria is approx. 1435 bp and nearly 1500 bp for gammaproteobacteria.

    • Thanks, Praveen – that’s an interesting observation! Presumably the Pseudomonas sequence was the strongest one, though, since that is what the authors reported.


  4. Matt permalink

    Here’s another claim for ‘gamma-rhizobia’: doi:10.1016/j.syapm.2010.04.005.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: