Control of arbuscular mycorrhiza development by nutrient signals

Inorganic phosphate (Pi), the main form of phosphorus used by plants, is one of the most important limiting factors for plant growth. In the soil soluble Pi that is readily available for uptake, occurs at very low concentrations (Schachtman et al., 1998). One adaptation of plants to low Pi availability is the symbiosis with arbuscular mycorrhiza fungi (AMF) of the phylum Glomeromycota. The fungi efficiently take up phosphate and other mineral nutrients and deliver them to the host, in exchange for carbohydrates. Thereby, arbuscularmycorrhiza compatible plants have two Pi uptake pathways, which are defined by different sets of phosphate transporters: a direct uptake pathway through the epidermis and root hairs, and a symbiotic uptake pathway for the Pi provided by the fungus (Smith and Smith, 2011). For successful symbiosis the fungus colonizes the root. This involves initial recognition via diffusible molecules, hyphal docking to the root surface by a hyphopodium, re-differentiation of plant cells and their subsequent penetration by fungal hyphae and formation of highly branched fungal arbuscules in the root cortex, which release mineral nutrients to the host (Gutjahr and Parniske, 2013). Plants control the degree of arbuscular mycorrhiza (AM) colonization depending on their nutritional status and it has been repeatedly reported that under high Pi supply, AM development is repressed (e.g., Menge et al., 1978; Braunberger et al., 1991; Balzergue et al., 2010; Breuillin et al., 2010). This suppressive effect of high Pi on root colonization by AMF is partially overruled by nitrogen (N) starvation, and to a lesser extent by potassium, calcium or iron starvation (Nouri et al., 2014), suggesting that plants control the symbiosis in function of their nutrient requirements according to Liebig’s law of the minimum. The molecular mechanisms underlying the control of AM development by nutrient conditions are largely unknown. Conceptually, two scenarios are possible: AM development might be actively suppressed at high Pi conditions (Figures 1A,C). Alternatively or in addition, root cells might be conditioned by Pi starvation to actively promote AM formation (Figures 1B,D). At sufficient Pi supply this promotion might be simply absent. Here we examine the available literature for evidence for one or the other scenario. Although several nutrients influence AM development (Nouri et al., 2014) most research has focused on the role of Pi. Hyphopodium numbers on maize roots were inversely correlated with the Pi status of the shoot (Braunberger et al., 1991), indicating that AM suppression by high Pi occurs systemically. Indeed, split root experiments in pea and Petunia, showed an inhibition of AM colonization in the entire root system, even if only one half of the root system was fertilized with a high Pi concentration and the other half maintained a low Pi content (Balzergue et al., 2010; Breuillin et al., 2010). This calls for a long distance signal traveling from the shoot to the root to regulate AM colonization that might either suppress AM at high Pi or promote its development at low Pi. Candidates for long-distance signaling molecules could be members of the miR399 family since they play a well-established role in systemic Pi-starvation signaling (Lin et al., 2008; Pant et al., 2008; Gu et al., 2011). Interestingly, upon AM colonization, the expression of some miR399 family members was increased in Medicago and tomato leaves (Branscheid et al., 2010; Gu et al., 2014). Consistently, transcript levels of the miR399 target, PHO2 an ubiquitin E2 conjugase, that mediates the degradation of proteins required for phosphate starvation responses (Liu et al., 2012; Park et al., 2014), remained low (Branscheid et al., 2010). It was postulated that increased expression of miR399 family members might serve to keep phosphate starvation responses high to allow continuous colonization, in spite of increased shoot phosphate content resulting from functional symbiosis (Branscheid et al., 2010). However, miR399 over-expression failed to support colonization under high Pi supply (Branscheid et al., 2010), indicating that other regulatory mechanisms link AM development to the nutrient status of the plant. Plant endosymbiosis (AM and root nodule symbiosis) development requires a common set of genes called common SYM genes. Their protein products belong to a signal transduction cascade that is triggered by perception of fungal signals (Myc factors) through receptor-likekinases (Gough and Cullimore, 2011). Myc factor perception induces nuclear Ca2+-spiking that is decoded by a nuclear localized calcium-calmodulin kinase (CCaMK) and leads to transcriptional activation of symbiosis-related genes by