Auxin and Vesicle Traffic

Questionable Evidence for Auxin-Carrying Secretory Vesicles

Dear Editor,

The plant hormone auxin (indole-3-acetic acid [IAA]) controls almost every aspect of plant development and growth. Auxin is transported within the plant from cell to cell through auxin transporters, including the PIN-FORMED (PIN) efflux transporters (Geisler et al., 2017; Naramoto, 2017) and ABCB transporters (Borghi et al., 2015; Geisler et al., 2017). These transporters can operate independently of one another or in concert (Geisler et al., 2017). They are transmembrane proteins that are delivered to the plasma membrane (PM) by vesicle transport. Their orientation in the vesicle membrane during intracellular transport could potentially lead to the vesicles accumulating auxin and consequently directly cause auxin to be released into the extracellular space upon vesicle fusion with the PM.

This hypothetical scenario has inspired the idea that auxin transport resembles the release of neurotransmitters from synaptic vesicles during neurotransmission in animals and led to the premise of “plant synapses.” Consequently, the term “Plant Neurobiology” arose about a decade ago (Brenner et al., 2006) to promulgate these concepts and the notion that plants are not so different from animal cells in terms of signal transduction. Although the usage has been heavily criticized by many plant scientists (Alpi et al., 2007; Rehm and Gradmann, 2010), the existence of plant synapses is still promoted by some (Baluška et al., 2009; Baluška and Mancuso, 2013).

There is no information about the transport of ABC transporters to the PM or their intracellular activity. With the PINs, the situation is different. There are two classes, short and long ones. The short PINs, for example, PIN5, and related PIN-LIKES reside on the endoplasmic reticulum membrane, but their specific transport activities have not been clarified (Barbez and Kleine-Vehn, 2013). They might have a negative effect on nuclear auxin response by pumping auxin into the endoplasmic reticulum lumen (where it might get conjugated), thereby effectively reducing the pool of cytosolic and nuclear auxin that mediates auxin-dependent gene expression responses. Thus, their role, if any, in intercellular transport of auxin remains hypothetical and will not be discussed here. The long PINs, however, mediate auxin efflux from the cell and are well known to cycle between endosome(s) and the PM (Adamowski and Friml, 2015).

There are two populations of trans-Golgi network-derived exocytic vesicles transporting PINs to the PM: one for newly synthesized PINs, and the other for recycled PINs, including the polarly localized PIN1 (Richter et al., 2014). PINs had long been considered active auxin transporters that are not subject to any posttranslational control, and most models predicted auxin transport and distribution solely based on the presence and absence of PINs in frequently polar domains of the PM (Wisniewska et al., 2006). It has, however, recently become clear that PIN activity requires phosphorylation by protein kinases such as D6 PROTEIN KINASE (D6PK) and PINOID (PID; Zourelidou et al., 2014). These protein kinases reside at the PM, and in the case of D6PK, often coincide with PINs in polar domains, and cycle to and from the PM (Barbosa et al., 2014; Barbosa et al., 2016). However, PINs and D6PK have very different recycling kinetics, and this is just one of many observations indicating that PINs and their regulatory kinases are transported independently to the PM (Barbosa et al., 2014). In the case of PID, differential transport mechanisms have been proposed at least between PID and PIN2 (Kleine-Vehn et al., 2009).

With antibodies specific for PIN1-activating phosphosites, it has been shown that PIN1 is only phosphorylated and, thus, active at the PM, and that PIN1 phosphorylation is efficiently antagonized by dephosphorylation when the kinases with similar properties to D6PK are removed from the PM using trafficking inhibitors (Weller et al., 2017). Moreover, internalized PIN1 does not stain positively for the activating phosphorylation events (Weller et al., 2017). In conclusion, these observations suggest that PIN proteins should be inactive during vesicular transport, and therefore PIN transport vesicles would be unable to load vesicles with auxin. Nevertheless, the mechanism of PIN-mediated vectorial auxin efflux is still not fully understood, and it could be that other known or as yet unknown transporters are active in vesicles that load them with auxin. Therefore, the unequivocal demonstration of IAA molecules sequestered inside bona fide intracellular vesicles, whether secretory or cycling from the PM, would make an important contribution to resolving this important question. This is exactly what Baluška and coworkers have now claimed to have achieved through immunogold electron microscopy (EM; Mettbach et al., 2017), a conclusion that we do not share.

In two papers (Nishimura et al., 2011; Mettbach et al., 2017), Baluška and coworkers have attempted to localize IAA by immunogold EM. These studies used high-pressure freezing/freeze substitution (HPF/FS) for specimen preparation, a technique now commonly used to preserve ultrastructure and to retain antigenicity. However, the results in the two papers are contradictory. In the Nishimura et al. (2011) paper, immunogold particles are visible virtually everywhere (cytoplasm, plastids, mitochondria, Golgi), but not in the cell wall or vacuole lumen. In the Mettbach et al. (2017) paper, a single micrograph is presented as figure 3A that shows clustered IAA immunogold particles. The authors claim that these clusters are surrounded by a membrane and represent IAA-containing secretory vesicles. However, a limiting membrane cannot be seen. In fact, and in contrast to the Nishimura et al. (2011) paper, membranes in general were poorly resolved in the Mettbach et al. (2017) paper. A control micrograph (figure 3B in Mettbach et al., 2017) purportedly lacking gold particles is also given; however, the usual scale bars are missing and the magnification appears to be much lower, thus preventing gold particle recognition. Clusters of up to 80 gold particles are reported, albeit not shown. In the absence of controls, these clusters may well represent artificial associations of gold colloids without a protein layer, a common problem when older or poor-quality antibody preparations are used.

In fact, we are surprised that a specific labeling was at all possible with the staining procedure described by Mettbach et al. (2017), since antigens are normally no longer detectable when sections are first blocked, then postfixed with 2% glutaraldehyde before exposing them to primary IAA antibodies. We would have expected both positive and negative controls to have been performed, including infiltrating auxin into the tissue, presaturating auxin antibodies with the antigen, and labeling with antisera against marker proteins such as tubulin. However, no such controls were reported.

Mettbach et al. (2017) also performed a statistical analysis of the presence of the IAA gold clusters in various cell types in the roots of Arabidopsis seedlings, associating the clusters with vesicles, but the value of such data is unclear considering that surrounding membranes could not be visualized. Also reported was a reduction in IAA immunogold clusters following brefeldin A (BFA) treatments, which blocks secretory and endocytic recycling traffic in most cell types. However, in the absence of representative EM images, it is difficult to assess the credibility of such claims.

The unequivocal visualization of a boundary membrane is of paramount importance for claims that a transport vesicle contains high concentrations of immunodetectable IAA molecules. Numerous examples exist in the literature for well-recognizable transport vesicles both in cryosections (e.g. Pimpl et al., 2000) and in sections from HPF/FS-processed samples (e.g. Donohoe et al., 2013). But even if this prerequisite is achieved, there remains the basic question as to whether a small molecule like IAA can be effectively retained in place after HPF/FS and methacrylate embedding. Unequivocal controls are needed here as already noted. In this respect, we should point out that a chemical cross-linker (ethyldimethylaminopropyl carbodiimide for IAA; carbodiimide EDC for jasmonic acid) is absolutely essential for reliable immunofluorescence detection of chemically fixed phytohormones (Hause et al., 2013; Pasternak et al., 2015).

Clearly, it would be of great interest if vesicular traffic of IAA could be demonstrated, as such evidence would be a true “first” for plant hormone signaling. From the foregoing, however, the arguments in favor of such vesicular traffic are little more than clutching at straws. The evidence currently available indicates that PINs only become transport-active after insertion into the PM. What then would be their role in auxin transport if vesicles can sequester and concentrate IAA molecules, which are subsequently released into the extracellular space by a mere SNARE-dependent fusion of the vesicles with the PM? Detection of IAA in transport vesicles via immunofluorescence has been attempted and points to the presence of IAA in BFA bodies (Schlicht et al., 2006), but individual transport vesicles were not resolved with this technique. And as we have described above, the immunogold EM data published in support of IAA-enriched transport vesicles are insufficient and flawed. These facts seriously detract from the notion that transport vesicles in plants operate analogously to neurotransmitter-containing vesicles in the synapses of nerve cells by delivering IAA molecules to the cell surface. If it is to enjoy acceptance at large, Plant Neurobiology must be less esoteric and needs to be supported by rigorously performed peer-reviewed experimentation. Speculations alone are insufficient proof of concept.

Rebuttal: Substantial Evidence for Auxin Secretory Vesicles

Note: The Rebuttal to this Letter, provided by Baluška et al., has been removed from this article and will instead be published in the April 2018 issue and linked to this article. The content of the Baluška Rebuttal has not been changed from the original text.