The genome of stress tolerant crop wild relative Paspalum vaginatum leads to increased biomass productivity in the crop Zea mays

A number of crop wild relatives can tolerate extreme stressed to a degree outside the range observed in their domesticated relatives. However, it is unclear whether or how the molecular mechanisms employed by these species can be translated to domesticated crops. Paspalum Paspalum vaginatum is a self-incompatible and multiply stress-tolerant wild relative of maize and sorghum. Here we describe the sequencing and pseudomolecule level assembly of a vegetatively propagated accession of P. vaginatum. Phylogenetic analysis based on 6,151 single-copy syntenic orthologous conserved in 6 related grass species placed paspalum as an outgroup of the maize-sorghum clade demonstrating paspalum as their closest sequenced wild relative. In parallel metabolic experiments, paspalum, but neither maize nor sorghum, exhibited significant increases in trehalose when grown under nutrient-deficit conditions. Inducing trehalose accumulation in maize, imitating the metabolic phenotype of paspalum, resulting in autophagy dependent increases in biomass accumulation.

Responses of maize (Zea mays), sorghum (Sorghum bicolor), and paspalum (Paspalum vaginatum) to nutrient-deficit stress. (A) Representative images of above and below ground organs of maize , sorghum seedlings, and paspalum ramets at 21 days after panting (dap) under optimal (Full) , nitrogen-deficit (-N), or phosphorus-deficit conditions (-P). (B) Change in fresh biomass accumulation under -N or -P conditions in maize, sorghum, and paspalum at 21 dap. (C) Changes in root length relative to Full at 21 daf under -N or -P conditions in maize, sorghum, and paspalum. (B-C) (* = p <0.05; *** = p <0.0005;**** = p <0.00005; t-test). (D-E) Abundance (D) and reduction (E) of N as a proportion of total dry biomass in the shoots of maize, sorghum seedlings, and paspalum ramets at 21 dap. (F-G) Abundance (F) and reduction (G)of P as a proportion of total dry biomass in the shoots of maize, sorghum seedlings, and paspalum ramets at 21 dap. (H) Change in the observed mRNA expression of the conserved and expressed paspalum orthologs of maize genes known to encode starch synthase (GSSS1B) and starch debranching enzymes (ISO3 and ZPU1) in shoots under N-deficient or control conditions. General and paspalum-specific physiological responses to nutrient-deficit stress Only the metabolites with a statistically significance change in abundance (p <0.05; t-test) and an absolute fold change >2 in at least one of the three species evaluated are shown. Cell marked in gray were not significantly different between conditions and/or exhibited an absolute fold change less than 2. 1 3,5-dimethoxy-4-hydroxycinnamic acid; 2 Gamma-aminobutyric acid. Raw data for fold change and t-test results are shown in Supplemental Document 2. (C-D) Change in trehalose abundance in the roots of 3-week-old maize, sorghum seedlings, and paspalum ramets under -N condition (C) and -P condition (D) relative to plants grown under Full condition. Statistically significant changes are indicated in purple (t-test), and non-statistically significant changes are indicated in gray. (E) Number of significantly differentially expressed genes in paspalum (Pv), maize (Zm) and sorghum (Sb) identified in comparisons between roots of 3-week old plants grown under either -N or -P conditions and Full condition. Shading indicates the proportion of differentially expressed genes (DEGs) in each species that are syntenically conserved across species, or present at a unique location in the genome of the individual species evaluated. (F-G) Number of syntenically conserved orthologous triplets exhibiting shared or species-specific differential expression in response to -N and -P conditions (G) in maize (Z. mays), sorghum (S. bicolor), and paspalum (P. vaginatum). (H) Simplified diagram for trehalose metabolic pathway. (I) Expression levels of trehalase-encoding genes in paspalum (Pv), maize (Zm) and sorghum (Sb) in roots of 3-week-old plants grown in nutrient optimal (Full), N-deficient (-N) and P-deficient (-P) conditions. and L-threonine ( Figure 3A). This observation is consistent with the decreases in amino acid metabolism All three species exhibited decreases in 1,3-diaminopropane and allantoin levels under N-deficit conditions ( Figure 3A). Allantoin acts as a pool of relocalizable N that can be catabolized into ammonia 177 for N assimilation and amino acid biogenesis 45 . In addition, the abundance of both succinic acid and maleic 178 acid (MaA) increased in all three species in response to N-deficit stress ( Figure 3B). Maleic acid produced 179 and secreted in response to another abiotic stress (drought) in holm oak (Quercus ilex) 46 . As we examined 180 internal metabolite abundance but did not profile root exudates in the current study, it is not possible to 181 determine whether the internal accumulation of maleic acid resulted in additional secretion in these three 182 grass species. Succinate, the anion of succinic acid, forms part of the tricarboxylic acid (TCA) cycle. The 183 increase in succinic acid levels, combined with the decreased abundance of gamma-aminobutyric acid 184 (GABA), is consistent with these species employing the GABA shunt pathway, which was proposed to act 185 as an additional energy source to support cellular metabolism under stress conditions 47-49 . 186 We observed changes in metabolite abundance in maize and sorghum grown under P-deficient condi-187 tions relative to the nutrient optimal conditions, including L-asparagine, GABA, L-glutamine, L-alanine, 188 capric acid, D-glucose-6-phosphate and glycerol-1-phosphate. However, none of these metabolites ex- The sequencing, assembly, and annotation of the paspalum genome provided the opportunity to quantify 205 differences and commonalities in how maize, sorghum, and paspalum transcriptionally respond to nutrient-206 deficit stress. We collected RNA from the root tissues of three biological replicates of each species and

221
The set of 220 syntenically conserved orthologous gene groups that responded transcriptionally to 222 N-deficit stress in a consistent fashion among maize, sorghum, and paspalum was disproportionately 223 enriched in GO terms related to response to nutrient levels, nitrate assimilation, metal ion transporter 224 activities and divalent inorganic cation transmembrane transporter activity ( Figure 3F; Figure S6A). The 225 set of 37 syntenically conserved orthologous gene groups that responded transcriptionally to P-deficit 226 stress in a consistent fashion among the three grasses was disproportionately enriched in GO terms related 227 to lipid metabolic process, phosphate ion transport, response to nutrient levels and cell communication 228 ( Figure 3G; Figure S6A). Syntenically conserved orthologous gene groups where a transcriptional response 229 to N-deficit stress was unique to paspalum where enriched in genes involved in proton transport, glycoside   In an attempt to phenocopy the reduced plasticity in response to N-deficient treatment originally ob-   Figure S9).

354
Imposing equivalent stress treatment protocols across species presents numerous challenges. One 355 potential concern with the initial finding that paspalum is less phenotypically plastic in response to nutrient-356 deficient treatment than maize is that the slower baseline accumulation of biomass in paspalum may 357 deplete the modest reserves of nitrate and phosphate in soil more slowly than they would be used by maize.

358
Here comparison of paspalum to sorghum may be more informative than comparison of paspalum to maize.

18/44
Codon-level multiple sequence alignments of syntenic orthologous gene groups were generated with ParaAT2.0 120 . Synonymous nucleotide substitution rates (Ks), and non-synonymous nucleotide substi-547 tution rates (Ka) were estimated from these multiple sequence alignments using the 'codeml' package 548 implemented in PAML 133 . The estimation was conducted using the maximum-likelihood method and 549 the parameters runmode=0, Codon-Freq=2, model=1. The known phylogenetic relationships of the six 550 included species were used as a known input tree. Syntenic orthologous groups containing any genes with 551 a Ks greater than 2, a Ka greater than 0.5, and a Ka/Ks ratio greater than 2 were removed.

552
Gas chromatography-mass spectrometry (GC-MS) metabolite profiling 553 Root samples from maize, sorghum, and paspalum seedlings grown as described above were collected 554 in a dark room illuminated solely by a green bulb and ground into a fine powder in liquid nitrogen.

555
Approximately 50 ± 0.5 mg of the ground powder was used for metabolite extraction and derivatization as   Table S2 to ensure that all markers were in the same linkage 584 phase.

585
To assist with scaffold ordering and assessment of the quality of the assembly, 500 bp on either side 586 of mapped SNP markers were excised from the assembly used for GBS read alignment and mapped to   . Gene ontology (GO) analysis of differentially expressed syntenic orthologous genes across the three species and in papspalum alone. (A) Significantly enriched GO terms (false discovery rate (FDR) ≤ 0.05) for 220 and 37 syntenic orthologous genes that were differentially expressed in all of the three species in response to N-deficit and P-deficit conditions, respectively. Bars indicate the log-transformed enrichment factor (number of genes associated with the overrepresented GO terms in the study gene set over the number of genes associated with the GO term in the background gene set) for enriched GO terms. Negative log-transformed multi-test corrected p values are color coded. (B) Significantly enriched GO terms (false discovery rate (FDR) ≤ 0.05) in 825 and 650 syntenic orthologous genes that were differentially expressed only in paspalum in response to N-deficit and P-deficit conditions, respectively. Bars indicate the log-transformed enrichment factor (number of genes associated with the overrepresented GO terms in the study gene set over the number of genes associated with the GO term in the background gene set) for enriched GO terms. Negative log-transformed multi-test corrected p values are color coded. Figure S7. Expression patterns of genes encoding trehalose-6-phosphate synthase in response to nutrient stress across maize (Zea mays), sorghum (Sorghum bicolor), and paspalum (Paspalum vaginatum). (A) Phylogeny of orthologs of Arabidopsis trehalose-6-phosphate synthase 1 (TPS1) genes in the three species. (B) Expression pattern of the trehalose-6-phosphate synthase genes in the three species under nutrient-optimal (Full), nitrogen-deficit (-N), and phosphorus-deficit (-P) conditions. "maize1" and "maize2" indicate the two subgenomes that formed in maize after the recent whole-genome duplication event 12-16 million years ago. (C-I) Expression patterns of other syntenic genes annotated as encoding trehalose-6-phosphate synthase that did not cluster with Arabidopsis homologs in the three species under nutrient-optimal (Full), nitrogen-deficit (-N), and phosphorus-deficit (-P) conditions. Figure S8. Expression patterns of genes encoding trehalose-6-phosphate phosphatase enzymes in response to nutrient stress across maize (Zea mays), sorghum (Sorghum bicolor), and paspalum (Paspalum vaginatum). (A) Phylogeny of orthologs of characterized Arabidopsis trehalose-6-phosphate phosphatase (TPP) genes in the three species.

38/44
(B-C) Expression patterns of the trehalose-6-phosphate phosphatase genes (trpp6 and trpp11) that clustered with their Arabidopsis homologous (TPPA, TPPG, TPPF) under nutrient-optimal (Full), nitrogen-deficit (-N), and phosphorus-deficit (-P) conditions. (D-H) Expression patterns of other syntenic genes annotated as trehalose-6-phosphate phosphatase that did not cluster with Arabidopsis homologs in the three species grown under nutrient-optimal (Full), nitrogen-deficit (-N), and phosphorus-deficit (-P) conditions. For panels B to I, "maize1" and "maize2" indicate the two subgenomes that formed in maize after the recent whole-genome duplication event 12-16 million years ago. Figure S9. ValA treatment alters biomass accumulation and nutrient reallocation to sorghum (Sorghum bicolor) grown under nutrient-deficient conditions. (A) Representative images of sorghum seedlings grown under nutrient optimal and N-deficit conditions with or without validamycin A (ValA) treatment. Images were taken 21 days after planting. For the ValA treatment, a 30 µM solution was added at 6 PM on the day that the plants were watered with the indicated nutrient solutions. (B) Changes in observed trehalose abundance -normalized to an internal reference (ribitol) -in response to validamycin A and/or nutrient conditions in sorghum root tissues. Error bars are standard deviations. Student's t-test (* = p <0.05; ** = p <0.005; *** = p <0.0005). (C) Dry weight of the above-ground tissue of sorghum seedlings grown under nutrient-optimal and nitrogen-deficit conditions harvested at 3 weeks after planting. Plant tissues were freeze-dried for 48 hours after harvesting. (D) Shoot-to-root ratio calculated from the dry weight of above-ground tissues and roots of the same sorghum seedlings. (E) Representative images of paspalum seedlings at 3 weeks after planting grown under nutrient optimal (Full) and nitrogen-deficient (-N) conditions with (ValA) or without (Control) validamycin A treatment. (F) Lack of significant increases in trehalose abundance (normalized to an internal reference [ribitol]) in response to validamycin A treatment (ValA) in 3-week-old paspalum seedlings under either full-nutrient or N-deficient conditions. (G) No significant change observed in above ground dry weight of 3-week-old paspalum seedlings in response to validamycin A treatment (ValA) under full-nutrient or N-deficient conditions. (H) Ratio of shoot-to-root dry weight in 3-week-old paspalum seedlings grown with or without validamycin A under full-nutrient or N-deficient conditions.  , ZmMDH6 (C) and ZmBZIP11 with our without validamycin A treatment under full nutrient and N-deficient conditions. p values were calculated by DESeq2 after correction for false discovery rate lower than 0.05. (F) Trehalose accumulation might lead to a lower T6P level, resulting in the release of inhibition of SNRK1 activity. The active status of SNRK1 would promote autophagy and ZmAKIN11 expression while repressing the expression of MDH and bZIP genes. (G-H) A biological replicate of the immunoblot measuring the abundance of both free ATG8 (upper band) and the ATG8-PE conjugate (lower band) in root samples collected from 3-week-old maize seedlings grown under optimal nutrient (Full) and nitrogen-deficit (-N) conditions with or without ValA treatment (G) and in root samples collected from 1-week-old maize seedlings grown under optimal nutrient conditions with or without ValA treatment (H). Total protein loading control is shown in the lower panel.

43/44
Supplementary Notes attached to this submission as separate files.