Plant adaptation to changing environmental conditions largely depends on root system developmental plasticity : depending on water and nutrient availability, as well as biotic and abiotic stresses, the root system architecture is modified by dynamically regulating organ growth, number and distribution. In addition to the formation of lateral roots, legume plants are able to form under soil nitrogen limiting conditions another root lateral organ, the nitrogen-fixing nodule (for reviews see Crespi and Frugier, 2008 ; Gonzalez-Rizzo et al., 2009). This legume-specific organogenesis is induced in response to a symbiotic interaction with bacteria collectively referred to as Rhizobia.
We are interested in understanding mechanisms allowing legumes to coordinate root and nodule development depending on changing environmental conditions : nutrient availability, such as nitrate ; abiotic stresses such as salt or drought ; root pathogens or symbionts. This is critical both to improve yield stability of legume crops used in sustainable agriculture, as well as for understanding bottlenecks that restrict symbiotic nitrogen-fixing nodule organogenesis to legume plants, in the perspective of transferring this capacity to other crops.
In this goal, reverse and forward genetic approaches allowed us to identify in the Medicago truncatula model legume different signaling pathways regulating legume root system development depending on environmental cues.
We showed that cytokinin phytohormones regulate antagonistically symbiotic nodule and lateral root development depending on the CRE1 (Cytokinin Response 1) receptor (Gonzalez-Rizzo et al., 2006 ; Frugier et al., 2008 ; Plet et al., 2011 ; Laffont et al., 2015). Even though rhizobia bacteria synthesize and secrete typical bio-active cytokinins, their contribution to nodulation seems nevertheless rather limited (Kisiala et al., 2013). The CRE1 cytokinin signaling pathway regulates nodule organogenesis through the modulation of polar auxin transport (Plet et al., 2011) depending on the accumulation of specific flavonoids (Ng et al., 2015), leading to a local auxin accumulation in dividing cortical cells associated to early nodule formation. More recently, we showed that CRE1 acts redundantly with other CHASE Histidine Kinase (CHK) cytokinin receptors in nodule initiation, and can be replaced by the closest homolog from the Arabidopsis thaliana aposymbiotic plant (Boivin et al., 2016). Cytokinins and CHK receptors have divergent functions in mature nodules, notably in relation to the nitrogen fixation capacity, which can not be replaced by the Arabidopsis CRE1 gene. In addition, cytokinins and the CRE1 signaling pathway also act in the root epidermis where they seem to negatively regulate bacterial Nod factor signaling and potentially rhizobial infections (Jardinaud et al., 2016).
A combination of biochemical and transcriptomic analyses allowed us to identify in M. truncatula roots primary cytokinin response genes depending on the CRE1 signaling pathway (Ariel et al., 2012). Among those, the Nodulation Signaling Pathway 2 (NSP2) transcription factor, previously demonstrated to be critical for nodule organogenesis, was identified. Interestingly, this gene is also posttranscriptionally regulated in response to cytokinins by a specific microRNA isoform (miR171h). This strategy also led us to identify a new basic Helix-Loop-Helix (bHLH) transcription factor (MtbHLH476) as a cytokinin primary response gene positively involved in nodule formation.
These transcriptomic analyses also allowed to identify that the gibberellin (GA) phytohormone metabolism was rapidly regulated by cytokinins. We then showed that GA indeed negatively regulates nodulation, and notably rhizobial infections, depending on DELLA signaling proteins (Fonouni-Farde et al., 2016). Strikingly, DELLA proteins in the epidermis are sufficient to induce a symbiotic infection marker, interact with the critical NF signalling components NSP2 and NF-YA1 and can transactivate ERN1 expression.
Figure 1. Regulatory pathways acting downstream of cytokinins in nodulation
Schematic representation of a M. truncatula symbiotic nodule primordium, including selected components of the Rhizobium Nod factor, cytokinin, and gibberellin signalling pathways.
NFP/LYK, Nod Factor perception / Lys-M Kinase receptor ; DMI3/CCaMK, Does Not make Infections 3 / Calcium - CalModulin protein Kinase ; NSP2, Nod factor Signalling Pathway 2 ; NF-YA1, Nuclear Factor - YA1 ; ERN1, ERF Required for Nodulation 1 ; ENOD11, Early Nodulin 11 ; DELLAs, DELLA proteins ; CRE1, Cytokinin Response 1 ; RRB, Type B Response Regulators ; RR4, Type A RR 4 ; miR171h, microRNA 171h ; bHLH476, basic Helix-Loop-Helix 476 ; CKX1, Cytokinin oxidase 1 ; PINs, Pin-formed proteins
In addition to the role of the CRE1 cytokinin signaling pathway in symbiotic nodulation, we analyzed its involvement in other root environmental responses. Whereas no defect in endomycorhization was observed in cre1 roots, an increased tolerance to an abiotic stress (salt) and to root pathogens (Ralstonia solanacearum and Aphanomyces euteiches) was identified (Moreau et al., 2014 ; Laffont et al., 2015). In addition to direct cross-talks between cytokinin and defense/stress responses, these phenotypes additionally correlate with the increased ability of the cre1 mutant to form lateral roots.
Forward genetic screens allowed us to identify mutants affected both in root and nodule development, referred to as compact root architecture (cra ; Laffont et al., 2010). One of these mutants, cra2 , affects antagonistically nodule and lateral root development and is defective in a CLAVATA1-like LRR-RLK (Huault et al., 2014). Grafting experiments showed that the regulation of symbiotic nodulation was determined by the activity of the receptor in shoots, implying a systemic regulation of root nodulation. More recently, we have showed that the MtCEP1 peptide positive effect on nodulation and negative effect on lateral root formation relies on the MtCRA2 receptor (Mohd-Radzman et al., 2016). We are currently analyzing how this new regulatory pathway integrates with the previously reported systemic Super Numeric Nodules (SUNN)-dependent “Autoregulation of Nodulation” and cytokinin/CRE1 pathways.
Figure 2. A working model for a dynamic fine tuning of nodulation : SUNN and CRA2 Autoregulation of Nodulation Pathways.
CRA2, Compact Root Architecture 2 LRR-RLK ; SUNN, Super Numeric Nodules LRR-RLK ; CLE12, CLE13 : CLAVATA3-like peptides 12 and 13 ; CEP1, C-terminal Encoded Peptides 1 ; AON, Autoregulation Of Nodulation.
We previously characterized adaptive root architecture responses induced by an abiotic stress (salt) in M. truncatula and identified thanks to various transcriptomic approaches different markers and regulatory genes regulated in response to salt stress and linked to root and nodule development (Frugier et al., 2000 ; Gargantini et al., 2006 ; de Lorenzo et al., 2007 ; Chinchilla et al., 2008 ; Gruber et al., 2009). Among those, a connection with cytokinin signaling was highlighted (Merchan et al., 2007 ; Laffont et al., 2015). In the current ABSTRESS EEC project, we now aim to identify in legumes regulatory genes controlling tolerance to “multistress” conditions (a combination of drought stress and of the soil pathogenic fungus Fusarium).
Natural diversity of M. truncatula was also used to select genotypes showing contrasting salt responses to perform transcriptomic approaches (Zahaf et al., 2012). Selected regulatory pathways crucial for M. truncatula root architecture and its adaptation to abiotic stress conditions were functionally analyzed. Among those, we characterized a transcriptional network involving a type I HomeoDomain-Leucine Zipper (HD-ZIP) transcription factor regulating a LOB (Lateral Organ Boundary) transcription factor, which controlled lateral root emergence notably under abiotic stress conditions (Ariel et al., 2010). In addition, we identified a salt-regulated NAC transcription factor that is involved both in root response to an abiotic stress and in symbiotic nodule senescence (de Zélicourt et al., 2012). We are currently expanding molecular analyses of the interactions existing between stress and senescence responses in symbiotic nodules notably in the frame of the STAYPINK ANR project.
Figure 3. MtNAC969 RNAi symbiotic nodules are prematurely senescent
Toluidine blue-stained section of GUS RNAi (A) and MtNAC969 RNAi (B) 21 days post inoculation nodules. MtNAC969 RNAi nodules have an increased amyloplast accumulation and expression of senescence markers.
Zone I (meristem), II (infection/differentiation), and III (nitrogen-fixing).