Host-dependent specialized metabolism of nitrogen export in actinorhizal nodules induced by diazotrophic Actinomycetota Frankia cluster-2

Abstract

Frankia cluster-2 strains are diazotrophs that engage in root nodule symbiosis with actinorhizal plants of the Cucurbitales and the Rosales. Previous studies have shown that an assimilated nitrogen source, presumably arginine, is exported to the host in nodules of Datisca glomerata (Cucurbitales), while a different metabolite is exported in the nodules of Ceanothus thyrsiflorus (Rosales). To investigate if an assimilated nitrogen form is commonly exported to the host by cluster-2 strains, and which metabolite would be exported in Ceanothus, we analysed gene expression levels, metabolite profiles, and enzyme activities in nodules. We conclude that the export of assimilated nitrogen in symbiosis seems to be a common feature for Frankia cluster-2 strains, but the source of nitrogen is host dependent. The export of assimilated ammonium to the host suggests that 2-oxoglutarate is drawn from the tricarboxylic acid (TCA) cycle at a high rate. This specialized metabolism obviates the need for the reductive branch of the TCA cycle. We found that several genes encoding enzymes of central carbon and nitrogen metabolism were lacking in Frankia cluster-2 genomes: the glyoxylate shunt and succinate semialdehyde dehydrogenase. This led to a linearization of the TCA cycle, and we hypothesized that this could explain the low saprotrophic potential of Frankia cluster-2.

Keywords: Frankia; Actinorhizal symbiosis; GS synthetase; TCA cycle; nitrogenase; root nodules; succinic semialdehyde dehydrogenase.

Fig. 1.

Relative expression levels (ΔCt value) of genes involved in the GS/GOGAT pathway. (A) Illustration of the GS/GOGAT pathway connected to the TCA cycle and the arginine biosynthesis pathway, given in grey. (B) The gene expression data. The Ct value is normalized against the gene infC, encoding the translation initiation factor IF-3 (Alloisio et al., 2010), and the nitrogenase subunit (MoFe protein) gene nifD. An asterisk indicates a significant difference (P<0.5), based on one-way ANOVA followed by Tukey post-hoc analysis, of gene expression measured in four technical replicates of three biological replicates of nodules from Alnus glutinosa induced by Frankia alni ACN14a (red, left), Ceanothus thyrsiflorus induced by Candidatus Frankia californiensis Cv1 (yellow, centre), and Datisca glomerata induced by Candidatus Frankia californiensis Dg2 (black, right). Individual data points of the biological repeats are presented. Abbreviations: gdh, glutamate dehydrogenase; glnA1/glnA2, glutamine synthetase I subunits; glnII, glutamine synthetase II; gltB, glutamate synthase, large subunit; gltD, glutamate synthase, small subunit.

Fig. 2.

Relative expression levels (ΔCt value) of genes involved in the arginine biosynthesis pathway. (A) Biosynthesis pathway and connection to the TCA and GS/GOGAT cycle, illustrated in grey. (B) Gene expression levels. The Ct value is normalized against the gene infC, encoding the translation initiation factor IF-3 (Alloisio et al., 2010), and the nitrogenase subunit (MoFe protein) gene nifD. An asterisk indicates a significant difference (P<0.5), based on Student’s t-test (Cv1 and Dg2), of gene expression of four technical repeats of three biological repeats of nodules from Ceanothus thyrsiflorus induced by Candidatus Frankia californiensis Cv1 (yellow, left), and Datisca glomerata induced by Candidatus Frankia californiensis Dg2 (black, right). Individual data points of the biological repeats are presented. Abbreviations: argB, acetylglutamate kinase; argC, N-acetyl-γ-glutamyl-phosphate reductase; argD, acetylornithine/succinyldiaminopimelate aminotransferase; argE/argJ, bifunctional gene acetylornithine deacetylase; argF, ornithine carbamoyltransferase; argG, argininosuccinate synthase; argH: argininosuccinate lyase.

Fig. 3.

Concentrations of nitrogen metabolites in Ceanothus thyrsiflorus compared with previously published data on Datisca glomerata and Alnus glutinosa (Persson et al., 2016). Data are presented to compare the concentration (log10 nmol g FW^–1) per tissue: uninoculated roots (grey), inoculated roots (brown), and nodules (mustard). Significant differences, based on one-way ANOVA followed by Tukey post-hoc analysis, are indicated with compact letter display.

Fig 4.

GS transferase activity assay in nodules of Alnus glutinosa, Datisca glomerata, and Ceanothus thyrsiflorus. Nodules of A. glutinosa were induced by Frankia alni ACN14a, of D. glomerata induced by Candidatus Frankia californiensis Dg2, and of C. thyrsiflorus by Candidatus F. californiensis Cv1. Activity measurements are based on the amount of enzyme required to produce 1 μmol of γ-glutamyl hydroxamate (nKat). The activity was measured in the negative control in crude protein extract without glutamine (red), and otherwise either using crude protein extracts (yellow), or in protein extract exposed for 10 min to 40, 50, or 60 °C (brown, light grey, and dark grey), to distinguish between Frankia GSI and plant GS/Frankia GSII activity. The absorbance at 530 nm was corrected against the background absorbance of total denatured protein (boiled for 10 min at 95 °C). The assay was conducted on two technical replicates of two biological replicates. The individual data points are given. Different letters indicate significant differences (P<0.05), based on one-way ANOVA followed by Tukey post-hoc analysis.

Fig. 5.

Relative gene expression (ΔCt value) of genes encoding enzymes involved in the TCA cycle. The Ct value is normalized against the gene infC, encoding the translation initiation factor IF-3 (Alloisio et al., 2010), and the nitrogenase subunit (MoFe protein) gene nifD. An asterisk indicates a significant difference (P<0.5), based on Student’s t-test (Cv1 and Dg2), of gene expression of four technical repeats of three biological repeats of nodules from Ceanothus thyrsiflorus induced by Candidatus Frankia californiensis Cv1 (yellow, left), and Datisca glomerata induced by Candidatus F. californiensis Dg2 (black, right). Individual data points of each biological repeat are presented. The pathway with enzymes interacting is illustrated below. Abbreviations used: pepck, phosphoenolpyruvate carboxykinase; gltA/gltA2, citrate synthase; citA/citA4, citrate synthase; acnA, aconitate hydratase A; icd, isocitrate dehydrogenase; gltB, glutamate synthase, large subunit; gltD, glutamate synthase, small subunit; gdh, glutamate dehydrogenase; sucC/sucD, succinate-CoA ligase subunit alpha/beta; sdhA/shdB/sdhC/sdhD, succinate dehydrogenase complex subunit A/B/C/D; fum, fumarate hydratase; mdh: malate dehydrogenase.

Fig. 6.

Protein modelling of 2-oxoglutarate decarboxylase. (A) The solved crystal structure of Mycobacterium smegmatis (red, reference A0R2B1). (B) The model of Candidatus Frankia californiensis Dg2 (yellow, global model quality estimate QMEANDisCo >0.70). (C) The solved crystal structure of 2-oxoglutarate dehydrogenase of Staphylococcus epidermis (rosy brown, reference Q5HPC6). (D) Overlay of three different models, with arrows indicating the major differences between the 2-OG decarboxylase and 2-OG dehydrogenase.

Fig. 7.

Enzyme activity assay for succinic semialdehyde dehydrogenase (SSA-DH) in cluster-1 Frankia alni ACN14a compared with cluster-2 Frankia coriariae BMG5.1. Absorbance was measured at 340 nm, to detect the production of NADPH from NADP⁺; absorbance was blank corrected. Protein extracts were allowed to acclimate for 30 min in the reaction buffer with NADP⁺ but without substrate, after which SSA or solvent only was injected. The absorbance was measured for an additional 50 min. Time is represented in cycles, with increasing darkness (red for ACN14a and yellow for BMG5.1) indicating that more time has passed. Boxplot indicates the absorbance at the beginning, before injection, after injection, and at the end of the experiment, based on two technical replicates of three biological replicates. The compact letters display indicates significant differences based on three-way ANOVA (sample, treatment, cycle), followed by Tukey post-hoc analysis.

Fig. 8.

Summary of the data presented. The host plant provides Frankia with carbon sources which are shuttled into the TCA cycle. Several gene losses were identified, resulting in a linearization of the cycle. As a result, the produced 2-oxoglutarate is used for ammonium assimilation. The export of an assimilated nitrogen source is dependent on the host plant. A heatmap represents all Frankia gene expression analysed, including genes which were not discussed in the manuscript. Gene expression levels are indicated in light (lower) to dark (higher) green.