Nutrient regulation of lipochitooligosaccharide recognition in plants via NSP1 and NSP2

Many plants associate with arbuscular mycorrhizal fungi for nutrient acquisition, while legumes also associate with nitrogen-fixing rhizobial bacteria. Both associations rely on symbiosis signaling and here we show that cereals can perceive lipochitooligosaccharides (LCOs) for activation of symbiosis signaling, surprisingly including Nod factors produced by nitrogen-fixing bacteria. However, legumes show stringent perception of specifically decorated LCOs, that is absent in cereals. LCO perception in plants is activated by nutrient starvation, through transcriptional regulation of Nodulation Signaling Pathway (NSP)1 and NSP2. These transcription factors induce expression of an LCO receptor and act through the control of strigolactone biosynthesis and the karrikin-like receptor DWARF14-LIKE. We conclude that LCO production and perception is coordinately regulated by nutrient starvation to promote engagement with mycorrhizal fungi. Our work has implications for the use of both mycorrhizal and rhizobial associations for sustainable productivity in cereals.

MtHistone as a loading control. The assay was repeated three times with similar results. c Fungal colonization of wild type (A17) and nsp2-2 mutants under P-deficient conditions, measured at 3 weeks post inoculation. d Related to Fig. 5c, a repeat of R. irregularis colonization in the critical mutants. n=5 biologically independent samples. Total: total colonization; A: arbuscules. e Nodulation of M. truncatula in strigolactone biosynthesis and strigolactone/karrikin signaling mutants under N-limited conditions. Number of nodules formed per plant was measured at 2 weeks post inoculation. n=10 biologically independent plants. For statistical analysis a one-sided Wilcoxon test was performed. **: p<0.01. Fig. 3f, showing root-length colonization of barley wild type, nsp1a and nsp2 mutants at 7 weeks post inoculation. n=5 biologically independent samples. b Related to Fig. 5d, Rootlength colonization at 9 weeks post inoculation in barley wild type and d14l mutant roots under P deficient conditions. n=8 biologically independent samples. Total: total colonization; A: arbuscules. p-values for colonization levels were determined by a one-sided Wilcoxon test. **: p<0.01; *: 0.01<p<0.05. Heatmaps showing the M. truncatula (a) and barley (b) genes upregulated by either N and/or P starvation, repressed in nsp1 or nsp2 mutants in at least b one nutrient condition and activated by overexpression of miRR-MtNSP2 in M. truncatula (a) and MtNSP2 in barley (b). +P-N, -P+N and -P-N represent the expression of these starvation-induced genes in wild-type plants by comparing -N or/and -P conditions to +P+N conditions. nsp show gene expression in nsp mutants compared to wild-type plants grown in the same nutrient conditions (-N or/and -P), while NSPox show NSP overexpression roots compared to wild type under nutrient replete conditions. Genes are clustered by expression pattern. A 1.5-fold cutoff was used for log2 fold changes with all the heatmaps.

Supplementary Figure 8. NSP1 and NSP2-regulated genes involve the apocarotenoid biosynthetic pathway. a A subset of the biosynthetic pathway of apocarotenoids in plants.
The enzymes in each step are highlighted in bold and their full names are listed in the inset. The genes regulated by nutrient starvation in an NSP-dependent manner (Fig. 5a, b) are highlighted on the pathway, with blue stars representing M. truncatula genes and red circles representing barley genes. b Root expression pMtD27::GUS in M. truncatula wild type and nsp1-1 mutants. The plants were grown on modFP plates for 4 weeks after hairy-root transformation. Bars = 200µm. The assay was repeated three times with similar results. c Expression levels of strigolactone biosynthetic genes in M. truncatula roots with overexpression of GFP, NSP1, NSP2 and a combination of NSP1 and NSP2, using PT4 as a negative control. The plants were grown on modFP plates for 4 weeks after hairy-root transformation of the respective construct. Bars represent means of 4 biological replicates. ± s.e.m. Different letters indicate different statistical groups (ANOVA, post hoc Tukey, P < 0.05).

Supplementary Figure 9. Nutrient regulation of barley RLK genes via 5DS and KARs. a
Relative expression of barley RLK genes in wild-type roots grown under different nutrient conditions. The expression values were obtained by conducting TMM normalization from transcriptome data 91 . n=2-3 biologically independent samples. ±s.e.m. b qPCR showing regulation of symbiosis genes by 5DS and KARs. Wild type plants were grown under repressive P-conditions and pretreated for 2 days on solid media containing 1 µM 5deoxystrigol (5DS) or a mixture of 1 µM karrikin 1 and karrikin 2 (KARs). D53 acts as a marker gene for strigolactone treatment 112 , while DLK2 responding to both strigolactone and karrikin treatment 113 . n=4 biologically independent samples, ±s.e.m.; **indicates p<0.01, *indicates 0.01<p<0.05, measured using a Student's t-test (one-tailed, two-sample equal variance). c Root-length colonization of barley wild type and rlk2-2 mutant grown under low P conditions. The colonization levels were measured at 5 weeks post inoculation. n=3 biologically independent samples. d. A repeat of Fig. 6e, showing root-length colonization of barley rlk2, rlk10 and rlk2/rlk10 double mutants grown under P-limited conditions, measured at 7 weeks post inoculation. The RLK10 mutation in the rlk2/rlk10 double mutant is equivalent to rlk10-1. n=14-15 biologically independent samples. Total: total colonization; A: arbuscules. p-values for colonization levels were determined by a one-sided Wilcoxon test. **: p<0.01.

MtNSP1-dependent genes of those nutrient upregulated in at least one starvation condition (%)
11.5

MtNSP2-dependent genes of those nutrient upregulated in at least one starvation condition (%)
17