Title: Making sense of the complex role of the mitochondria in mediating the plant touch response
Abstract: Mechanical stimulation of plants, thigmorphogenesis, has been studied for many decades with some of the key papers on this subject being published almost 50 years ago (Neel and Harris, 1971; Turgeon and Webb, 1971). Many studies have shown that repeated touching of plants leads to growth retardation and have, moreover, linked this phenomenon to a range of processes including hormone responses, reactive oxygen species and Ca2+ signaling. These observations have led to suggestions that the touch response may confer cross-protection to abiotic and biotic stresses (Chehab et al., 2012). Whilst studies to date implicate auxin, gibberellin and jasmonate in this response the precise mechanisms underlying such cross-protection are as yet unknown. In this issue (Xu et al., 2019) studied 25 mutants targeting mitochondrial function revealing that all affected the touch transcriptome with synergistic and antagonistic effects being observed in subsequently generated double and triple mutants. They additionally demonstrated overlap between the touch response, mitochondrial stress signaling and alternative metabolism and showed that mitochondrial function modifies the touch response both directly at the level of gene expression or indirectly by modifying the levels of phytohormones. The laboratory of Jim Whelan, Professor at the Department of Animal, Plant and Soil Sciences at La Trobe University, in Melbourne is focused on the study of various elements of mitochondrial function including mitochondrial biogenesis. Jim joined the Department 5 years ago from the University of Western Australia in Perth. The work described in this article, which was carried out in collaboration with Michelle Hooi and Vincent Bulone from Adelaide and Inge De Clercq and Frank Van Breusegem from Ghent, resulted from 2 years of intensive work by the first author, post-doc Yue Xu under the close supervision of the early career researcher Dr Yan Wang. However, the actual work involved took far in excess of 2 years as this also encompassed the isolation and generation of the mutants studied, which was carried out over a period of some 10 years. This work was complemented by that of others in the lab including the study of hormone signaling mutants carried out by Yan and Jim, RNAseq data analyses by Oliver Berkowitz, and the cellular response to touch by Reena Narsai. Finally, hormone analysis was carried out in Adelaide and regulatory network analysis in Ghent. As I describe above the morphological response to touch is long characterized. It has additionally been known for almost 30 years now that the transcript abundance of any calcium and calmodulin mediated genes occurs following mechanical stimulus (Braam and Davis, 1990). The current study – and several others preceding it – revealed that indeed the levels of many hundreds of transcripts are altered following such a stimulus and hint at a role for mitochondrial retrograde signaling in the process. Indeed an earlier study from Jim's laboratory revealed that the OUTER MEMBRANE PROTEIN 66 (OM66) displays a rapid increase in transcript abundance upon touch (Van Aken et al., 2016), which was independent from that mediated by the transcription factor ANAC017, involving the classical mitochondrial retrograde signaling pathway including Alternative Oxidase 1a (AOX1a) (Ng et al., 2013). The use of the 25 mutants of mitochondrial function in the study of (Xu et al., 2019), alongside the study of the interaction between hormone and energy signaling, revealed the central role of mitochondria in mediating the touch response via both the up-regulation of genes responsive to (a)biotic stress signaling and a down-regulation of genes responsive to development and circadian rhythms. The work provides several key advances to our state-of-the-art understanding of the touch response. First, the resolution offered by their kinetic analysis clearly demonstrates the complexity of the touch response indicating that it either involves several pathways or that intricate, gene-specific RNA degradation pathways play a role in the response. That said the role played by elements contributing to mitochondrial function is striking. In particular the results from the cyclic dependent kinase E1 (cdke1) mutant, which has previously been demonstrated to interact with the AKIN10 subunit of the SNF1 related kinase complex (Ng et al., 2013), were revealing with only 568 of the 2004 transcripts observed to respond to touch in wild type being observed in the cdke1 mutants. These data thus interlink the touch response with energy and stress response signaling (Wurzinger et al., 2018). Furthermore, the fact that CDKE1 is a constituent of the mediator complex, which is a multi-protein complex functioning as a transcriptional co-activator in all eukaryotes (Dolan and Chapple, 2017), probably explains the route by which a mechanical stimulus effects the broad range of cellular transcriptional programs that mediate the growth retardation apparent on repeated mechanical stimulation. A further detail that this study uncovered was the notable enrichment in calmodulin binding transcription activator (CAMTA) binding sites present in the touch responsive genes leading the authors to postulate the model summarized in Figure 1. Carrying out these experiments was by no means an easy task, as co-ordinating the growth of so many mutants to certify that they were all at the same growth stage and ensuring that touching was performed in a uniform manner and that harvesting was always precisely timed, required painstaking dedication and discipline. A further obstacle was that this had to be carried out in a manner that minimized disturbances which could constitute accidental mechanical stimulation. Similarly, their interpretation of the collected was far from facile due to the sheer volume of data collected, with the 22 RNAseq data sets consisting of approx. 19 000 quantified genes in each, a number that could easily have led to the feeling of not being able to see the forest for the trees. In spite of these challenges (Xu et al., 2019) were able to provide compelling evidence that a significant proportion of the transcriptome response to touch is integrated with cellular energy and stress signaling combining to promote stress resilience over growth. A number of intriguing questions arise from this work such as What is the energy demand of the touch response and if this response was dampened would there be an effect on growth and development? Is the touch response dynamically different with leaf age or between different cell types and To what extent is the touch response maintained in crop and horticultural species? In addition to these forward looking questions a couple of more general overreaching conclusions can be made from the study. First, in shorter-term studies responses to other stimuli may be masked if they overlap with the touch response as these would also be induced in the control. More globally it provides food for thought with respect to the design of growth facilities since air handling units may yield heterogeneous growth conditions whilst automated plant watering and moving systems may additionally trigger mechanical responses. The highlighted paper is thus important on several independent levels; it illuminates previously uncovered details of the touch response, poses several intriguing questions that could be addressed using the same tools and experimental techniques described therein and finally throws up important questions that need to be answered in the era of high-throughput phenotyping in order to prevent the appearance of artefacts or masking of true responses. In doing so, (Xu et al., 2019) open up an important new avenue by which to disentangle the complex underpinnings of the plant response to touch.