Evolution of legume nodule type

Nitrogen-fixing root nodule symbiosis can sustain the development of the host plants under nitrogen limiting conditions. Such symbiosis only occurs in a clade of angiosperms known as the nitrogen-fixing clade (NFC). It has long been proposed that root nodule symbiosis evolved several times (in parallel) in the NFC. There are two main arguments supporting this parallel hypothesis. First, the occurrence of root nodule symbiosis is scattered across the NFC, hypothesizing a single origin is evolutionarily unparsimonious as this would require massive loss of the trait. Second, legume-type and actinorhizal-type nodules are fundamentally different. Two recent phylogenomic studies compared the genomes of nodulating and related non-nodulating species across the four orders of the NFC and found that genes essential for nodule formation are lost or pseudogenized in the non-nodulating species. As these symbiosis genes are specifically involved in the symbiotic interaction, it means that the presence of pseudogenes and the loss of symbiosis genes strongly suggest that their ancestor that still had functional genes most likely formed the nodule symbiosis. Therefore, the first argument supporting the parallel hypothesis is disproved. Actinorhizal-type nodules have long been regarded as modified lateral roots, originating from root pericycle cells, with newly formed cortical cells being infected by bacteria. Here, we compared the lateral root and nodule development in two plant species forming actinorhizal-type nodules (Parasponia andersonii (Parasponia) and Alnus glutinosa (Alnus)). We showed that in Parasponia and Alnus, lateral roots are mainly originated from pericycle-derived cells, endodermis cells only form the outmost layer of the lateral root cap. During the formation of Parasponia and Alnus nodules, pericycle-derived cells only form the nodule vasculature, instead of a lateral root (as was previously reported for these species). In both cases, parental root cortical cells are mitotically activated and are infected by bacteria. This is similar to the origin of infected tissue in legume nodules. This finding shows that actinorhizal-type nodules are not modified lateral roots, they are more similar to legume-type nodules than previously described. Therefore, the second argument supporting the parallel hypothesis is also disproved. We further showed that a legume-type nodule can be (partially) converted to actinorhizal type by knocking out a gene, namely Medicago truncatula (Medicago) NODULE ROOT1 (MtNOOT1). In Medicago Mtnoot1 mutants, the nodule vasculature is derived from pericycle cells, similar to the ontogeny of actinorhizal-type nodule vasculature. As homeotic mutants usually revert to an ancestral phenotype, we proposed that actinorhizal-type and legume-type nodules share a common evolutionary origin, and actinorhizal type is ancestral. Our study supports the single origin hypothesis that nodulation of actinorhizal type evolved once in the common ancestor of the NFC. It has been suggested that MtNOOT1 maintains nodule identity of the vascular meristem in a cell autonomous manner. However, our study implies that MtNOOT1 can indirectly maintain nodule identity by repressing actinorhizal-type ontogeny of nodule vasculature. It remains unclear whether MtNOOT1 has a direct effect on maintaining nodule identity of the vascular meristem. To investigate this, we generated CRISPR/Cas9 mutants in the MtNOOT1 orthologous gene in Parasponia, named PanNOOT1. We showed that in Parasponia Pannoot1 mutants roots can be formed at the nodule apex, similar to legume noot1 mutants. Both MtNOOT1 and PanNOOT1 are expressed in the nodule vascular meristem, suggesting that Medicago MtNOOT1 and Parasponia PanNOOT1 share a conserved function in maintaining nodule identity of the vascular meristem in a cell autonomous manner. We further showed that MtNOOT1 and PanNOOT1 are required for the intracellular colonization of bacteria in the host cells. The functional analysis of PanNOOT1 could bring insight to the ancestral symbiotic function of NOOT1 in the NFC. Medicago MtNOOT1 has been described as a nodule identity gene. It is also highly expressed in Medicago roots, where its function remains unclear. We showed that knockout of MtNOOT1 leads to accelerated primary root growth due to an enlarged root apical meristem. We further revealed that MtNOOT1 is expressed in the transition zone of roots, suggestive of a promotive role in inducing cell differentiation between two cell groups with different fates (root apical meristem vs. elongation/differentiation zone). This is similar to the function of MtNOOT1 orthologs in Arabidopsis thaliana (Arabidopsis)- AtBOP1 and AtBOP2- setting the boundary of restricted growth between shoot apical meristem and differentiating/elongating leaf primordia. In Arabidopsis, AtBOP1 and AtBOP2 promote cell differentiation in the stem by inducing lignin biosynthesis. Similarly, in Medicago roots MtNOOT1 promotes xylem cell differentiation by inducing a cascade of genes involved in xylem cell differentiation at proper distance from the root tip. A general role played by MtNOOT1 is to promote cell differentiation along the apical-basal axis during root development. These findings indicate that MtNOOT1 is required for a coordinated root development along the apical-basal axis. Shoot architecture of plants is modified by branching. Previous studies have shown that Arabidopsis AtBOP1 and AtBOP2 orthologs in monocot species (Hordeum vulgare and Zea mays) are involved in modulating shoot branching. We showed that in the tropical tree Parasponia PanNOOT1 regulates axillary branch development. Knock-out of PanNOOT1 reduces axillary branch growth, due to delayed axially bud emergence and outgrowth. We further showed that PanNOOT1 functions in boundary specification between the petiole and the stem, and in stipule formation and leaf patterning, in a similar fashion as BOP genes in herbaceous (model) species. This indicates that PanNOOT1 is a BOP gene, playing a conserved role in stipule formation and leaf patterning. This is the first time that the function of a BOP gene is characterized in a tree, revealing a novel role in shoot branching. I discussed the findings obtained in this thesis together with related published studies. I proposed the evolutionary trajectory of legume-type nodules, and the role of legume NOOT1 in this process. NOOT1 is required for the maintenance of nodule identity in actinorhizal-type and legume-type nodules. I further hypothesized the ancestral symbiotic function of NOOT1 in the NFC, and discussed the general function of NOOT1 under various developmental contexts.