The plant hormone auxin has long been known to play a significant role in plant growth, even featuring in Charles Darwin’s book ‘The Power and Movement of Plants’.
However until the mid-2000s the identity of any receptor for the hormone remained a mystery. Until that time, the site of auxin reception was somewhat of an enigma with the main candidate being a protein named ‘Auxin Binding Protein (ABP)’ that, in many biochemical studies, was shown to bind auxin at physiological concentrations (for review of ABP1 work1). This protein is present in Arabidopsis but investigations into its in vivo function were somewhat stalled by a 2001 publication that demonstrated that an abp1 T-DNA insertion mutant was embryo lethal2. This did not overly surprise the research community given the fundamental and wide-ranging role of auxin in plant development.
Meanwhile researchers at York University in the UK and at Indiana University in the US were working on a novel hypothesis that proposed the auxin receptor was linked to the degradation of AuxIAA proteins, which are negative regulators of the auxin response. Ultimately after years of painstaking research the labs of Ottoline Leyser and Mark Estelle demonstrated that the F-box protein TIR1, when in complex with an AuxIAA protein, was also able to bind auxin3-4. This finding nicely pieced together an elegant pathway of control that involved tightly regulated protein degradation controlling both positive and negative feedback regulation of the hormone signal.
At this time the work on ABP1 had mostly retreated into the long weeds as research in the community revolved around the TIR1-AuxIAA receptor paradigm. However toward the end of the 2000s ABP1 had a renaissance as molecular techniques such as RNAi and cellular immunisation allowed identification of transgenic plants with reduced expression of ABP1. Plants with a deficiency in ABP1 exhibited robust auxin-deficient phenotypes including cell-cycle arrest in tobacco BY2 cells5 and changes in cell-wall composition during cell expansion6. Initially these effects were thought to occur independent of TIR1 but more recently ABP1 was shown to lie genetically upstream of the TIR1-receptor pathway7. In keeping with its initially proposed role in the rapid auxin response, ABP1 was found, by use of a weak abp1-5 allele, to have a defect in auxin-induced transcriptionally-independent clathrin-mediated endocytosis of PIN proteins8 as well as playing a critical role in the activation of ROP proteins that control epidermal cell interdigitation9. Therefore reducing the expression of ABP1 had demonstrated, as would be predicted, that the protein plays a critical and wide-ranging role in cell division and expansion.
Therefore at this time the functional relationship between ABP1 and TIR1 appeared to satisfy researchers…… that was until a paper was published at the start of 2015 that greatly altered perceptions within this research community10.
The labs of Yunde Zhao and Mark Estelle at UC-San Diego initially used a CRISPR-based strategy to generate a null abp1 mutant allele, with the full expectation that the resulting plant would be unable to survive. Surprisingly they discovered that this abp1-c1 allele exhibited a completely wildtype phenotype. This prompted the isolation of a new T-DNA insertion mutant that was an abp1 null mutant. These plants also showed wildtype phenotypes across a range of assays that were thought to be ABP1-dependent. Ultimately the authors concluded that ‘ABP1 is not a key component in auxin signaling or Arabidopsis development’.
The discrepancy between these findings and the previous 15 years of data is striking and inevitably leads to questions about the previous work. In a recent calm editorial comment in the Journal of Integrative Plant Biology, Professor Chun-Ming Liu called for a careful reexamination of the previous data and for experimental materials to be exchanged in order to get to the bottom of this growing controversy11.
Even more recently the lab of Lucia Strader at Washington University in St Louis published a short paper in Plant Cell that questioned the abp1-5 mutant12. As stated above this allele has been used in a number of studies where its many interesting phenotypes include long hypocotyl growth in red light9. The Strader lab were interested in this response but ran into difficulties when characterising the genetic basis of this phenotype. Ultimately they decided to sequence the abp1-5 genome and surprisingly found that it contained significant portions from the Arabidopsis ecotype Wassilewskija, which is known to display mild resistance to auxin. Perhaps most worrying was that they discovered abp1-5 contained a null mutation in phyB, which is likely the causative effect of the long hypocotyl phenotype. The authors conclude by warning that their findings ‘provide a cautionary tale illustrating the need to use multiple alleles or complementation lines when attributing roles to a gene product’.
So what has happened in these experiments? In the Gao et al paper10 the authors took the simple approach of testing abp1 mutants for phenotypes and didn’t find anything different from the wildtype. It would be difficult to imagine where any errors would have occurred in these experiments especially given the nature of the genetic lesion in these new abp1 mutants. Gao et al suggest that the previous work based on ABP1 knock-down lines might have in fact been the consequence of off-target transgenic effects. Given the varied role that auxin plays in Arabidopsis development this is not an impossible conclusion to draw as numerous genes are involved in some aspects of this hormone response.
Most concerning is the initial characteristic of the original null abp1 mutant lines that was found to be embryo-lethal. This work is the cornerstone of the subsequent work that aimed to create plants with reduced ABP1 expression. The characterised abp1 mutant lines in Chen et al and Gao et al have T-DNA insertions close together at the 5’ end of the coding sequence so it is unclear why there are such differing results. It was unusual that in the first study only a single abp1 mutant allele was characterised, especially when the authors claimed embryo-lethality2. It would have been helpful for Gao et al (2015) to reexamine the original abp1 mutant line assessed in Chen et al (2001) but perhaps that might have appeared rather too confrontational.
At it stands the future direction of research into ABP1 is in flux with the onus now on those researchers with a vested interest, of which there are plenty, to try to understand the role of this protein.
In many ways this story is an excellent example of how science should work, where claims are independently tested to ensure that earlier experiments have been conducted or interpreted correctly. However this can be difficult as there is usually little value in retroactive testing of published claims. The plant science community awaits the resolution of the mystery of ABP1 whilst commiserating with those who have been negatively affected by this new development in this story.