Seed collection
Two ecotypes of Chenopodium quinoa were used in this work: Socaire ecotype from Altiplano (San Pedro de Atacama, Atacama Desert, Chile), and Faro ecotype from Central Chile. Socaire seeds were collected in a private field (23º34´S 67º54´W, 3500 m.a.s.l.) with the permission of owner and Socaire community and in accordance to research agreement among Concepcion University and Socaire Community.
Faro seeds were collected in Vicuña City (in an experimental station of the Banco Base Inia Intihuasi (La Pepa, 30º2´S 70º42´W, 616 m.a.s.l.) with all the permission required by the Institution (http://163.247.128.32/gringlobal/search.aspx).
Pedro Leon (Banco Base Inia Intihuasi) performed formal identification of the specimen plants. Plants of C. quinoa are already included in the collection of the Herbarium of the Universidad de Concepción (http://www2.udec.cl/~herbconc/index.htm, curator Alicia Marticorena) and in the Banco Base Inia Intihuasi (http://163.247.128.32/gringlobal/search.aspx, Pedro Leon in charge). It is worth to note that C. quinoa is not endangered species, and experimental research was in according to the practices of SAG (Servicio Agricola y Ganadero) of Chile.
Seeds of both ecotypes were collected in summer (February) in each location. Total percent and available Nitrogen were 0.09% and 46 mg/kg in Socaire soils, and 0.12% and 42 mg/kg in Vicuña soils, respectively. Further characterization of agroecological conditions in both places are described in Garcia et al. and Montes et al. [40, 41]. Collected seeds were stored in dry conditions at room temperature for 5 months, until experiments were performed.
Seed characterization and metabolite profiling
Samples for characterization and metabolite profiling were obtained from three different plants (n₌3). Socaire and Faro quinoa grains were dried in an oven at 80ºC overnight, and 1000-seed weight of each ecotype was measured. N content in seeds was determined by the method of Kjeldahl et al. [42].
Seeds were frozen, ground to powder and lyophilized to generate 50-mg samples for each sample. Metabolite profiling and analysis was performed at the West Coast Metabolomics Center (UC Davis, Davis, CA, USA), according to the methodology described in Botanga et al. [43]. The homogenization, extraction and derivatization of the seed powder was according to Fiehn et al. [44]. Chromatography was performed following the protocol described in Botanga et al. [43]. Metabolites were identified using the Fiehnlib libraries [45]. Data were further processed using the algorithms implemented in the open-source BinBase metabolome database [46].
Raw data were normalized, filtered and analyzed using the MetaboAnalyst 3.0 webserver [47]. Heatmap plots (to visualize metabolites profiles) were performed using the pheatmap R package. Data were scaled by Row Z-Score to capture metabolites with similar behavior. Clustering of metabolites was performed using the default pheatmap Ward method. Metabolites pathway information was obtained from KEGG API (https://www.kegg.jp/kegg/rest/keggapi.html) using a Shell script.
Germination
Seeds from both C. quinoa ecotypes were placed on Petri dishes with filter paper; sterile distilled water was added as needed to ensure adequate moisture for germination. Seeds were incubated for 48 h at 24°C in the dark. The number of germinated seedlings was recorded every 4 h for 48 h. Fifty seeds of each ecotype were used for each experiment, that were performed in triplicate. Germination was considered complete when the radicle emerged from the seed. Germination rate (number of germinated seeds/h) was recorded for seeds of each ecotype during 48 h [48]. The germination percentage, defined as [number of germinated seeds/number of total seeds] x 100 [49] was also recorded every 4 h up to 48 h after seed imbibition.
Determination of sufficient and low nitrate supply
In order to determine the sufficient and limiting NO3- supply concentrations for both Socaire and Faro plant ecotypes, germinated seeds presenting similar lengths of emerged radicle were transplanted into 700 mL pots containing sand:perlite (1:1). Seedlings were grown under growth chamber conditions (21°C, 16 h light/8 h dark photoperiod, 57% relative humidity) for 20 days. Plants were watered with MS 407 nutrient medium, which contained the macro- and micronutrients described by Murashige and Skoog [50], except nitrate, potassium and phosphate. In order to compensate for the lack of those nutrients, nutrient solution was supplemented with 85 mg/L KH2PO4, 950 mg/L KCl, and different concentrations of NO3- as KNO3. Six nitrogen treatments (0.0, 0.5, 2.0, 5.0, 20 and 100 mM NO3-) were applied in order to determine sufficient and limiting nitrate supply conditions for both quinoa genotypes. Ten pots (with 3 plants each) were submitted to each nitrogen treatment.
Seedlings were irrigated with the different solutions at field capacity during the time of the experiment. Plants were also irrigated in order to compensate for water losses during the time of the experiment. At the end of the experiment, samples of whole plants were collected, and growth parameters were measured in order to evaluate sufficient and low nitrogen supply concentrations.
Seedling performance at sufficient and low nitrate supply
In order to study the performance of both Quinoa ecotypes at sufficient and limiting N conditions during establishment, seedling of quinoa of both ecotypes were grown with irrigation at two NO3- concentrations, previously established with the experiment mentioned above (Section 2.4): 20 mM and 0.5 mM NO3-, corresponding to sufficient (HN) and low (LN) NO3- supply, respectively. Thirty pots of 700 mL (3 plants per pot) were used for each N supply condition per ecotype, using the same plant growth conditions described above (Section 2.4). The experiment was run for 24 days in a completely randomized design, that included additional plants grown in order to prevent bordering effect. At the end of the experiment, plants with at least four true leaves were collected.
For the study of the mechanisms involved in N uptake and assimilation, Chlorophyll a fluorescence, biomass, chlorophylls and protein content, NR and GS enzymatic activities and gene expression, and nitrate transporter gene expression were quantified in plants of both ecotypes submitted to both N treatments.
Plant morphometric analyses
Leaf area (n=5 plants of each ecotype submitted to each treatment) was measured through image analysis using the ImageJ software. Dry weights of leaves, shoots and roots (n=5 plants of each ecotype submitted to each treatment) were determined by drying the tissues at 60°C for 48 h till constant weight. Number of secondary roots and total root length (n=5 plants of each ecotype submitted to each treatment) were determined by image analysis using the WinRhizo software [51].
Protein content quantification
Leaves and roots (n=5 plants of each ecotype submitted to each treatment) were ground in liquid nitrogen and homogenized in 400 µL of extraction buffer (50 mM Tris-HCl pH 7.8, 1 mM EDTA, 1 mM DTT, 10 mM MgSO4, 5mM sodium glutamate, 10% ethylene glycol). After centrifugation (10,000 g for 10 min), the supernatant was used to measure total protein content by the Bradford assay [52]. Bovine serum albumin was used as standard.
Chlorophyll quantification and Chlorophyll fluorescence measurements
Chlorophyll a and b contents were determined spectrophotometrically following the method described by Lichtenthaler and Buschmann [53].
Chlorophyll (Chl) fluorescence measurements were performed (n=5 plants of each ecotype submitted to each treatment) using a portable fluorometer (FMS 2, Hansatech Instruments Ltd., Norfolk, UK). Leaves were dark-adapted for 20 min prior to measurements. Measurements were performed at chamber temperature. Actinic light used was of 300 µmol photons m-2 s-1, as described in Bascuñán-Godoy et al. [54]. Fluorescence parameters were calculated as described in Maxwell and Johnson and Kramer et al. [55, 56].
Nitrate reductase and Glutamine synthetase activities of roots
Both nitrate reductase (NR; EC 1.6.6.1) and glutamine synthetase (GS; EC 6.3.1.2) enzymes catalyze the limiting steps in the reduction of NO3- to NH4+ (primary assimilation), and the incorporation of NH4+ into amino acids, respectively. NR activity was measured in roots (n=5 plants of each ecotype submitted to each treatment) according to Kaiser and Lewis [57]. Total protein was extracted as described above, and the reaction was started by addition of 150 µL of reaction buffer (50 mM KH2PO4-KOH buffer, pH 7.5; 10 mM KNO3 and 0.1 mM NADH) to 100 µL of soluble protein extract. Samples were incubated at 30°C for 15 min. The NADH oxidation was measured by spectrophotometrical methods at 340 nm. GS activity was measured in roots (n=5 plants of each ecotype submitted to each treatment) by the formation of γ-glutamyl hydroxamate using the method described by O’Neal and Joy [58]. In order to perform the enzymatic reaction, 400 µL of protein extract were mixed with 150 µL of reaction buffer (100 mM Tris-HCl, pH 7.8; 50 mM sodium glutamate, 5 mM hydroxylamine hydrochloride, 50 mM MgSO4 and 20 mM ATP) and incubated at 30°C for 20 min. The reaction was stopped with detection solution (0.37 M FeCl3, 0.67 M HCl and 20% trichloroacetic acid). Glutamyl hydroxamate was measured spectrophotometrically at 540 nm using γ-glutamyl hydroxamate as a standard.
Phylogenetic analysis of nitrate transporters
Nitrate transporter (LATS and HATS) homolog sequences of C. quinoa and two other species from the Chenopodioideae subfamily (Beta vulgaris and Spinacia oleracea) were obtained from the Phytozome (https://phytozome.jgi.doe.gov/pz/portal.html) and NCBI (https://www.ncbi.nlm.nih.gov) databases, using the BLAST algorithm and the Arabidopsis thaliana homologues as query. The evolutionary history of these genes was inferred using the Maximum-Likelihood method based on the Tamura-Nei model [59]. Node robustness was assessed by the bootstrap method (N = 1000 pseudoreplicates). Phylogenetic analyses were performed with MEGA7 [60]. After a first characterization by phylogenetic analysis, a LATS and a HATS homologue of quinoa were selected for gene expression studies.
RNA extraction, first strand cDNA synthesis and quantitative real-time PCR
Total RNA was isolated from quinoa roots using RNeasy Plant Mini kit (Qiagen). The RNA obtained was treated with RNase-free DNase I (Qiagen) and quantified with a NanoDrop 1000 spectrophotometer (Thermo Scientific, USA). RNA intactness was verified by visual inspection of integrity of 28S and 18S rRNA bands in denaturing formaldehyde/agarose gel electrophoresis. RNA was stored at -80ºC for further use.
RNA (1 μg) was reverse-transcribed into single-stranded cDNA templates using the PrimeScript™ RT reagent Kit (Takara) and oligo-p(dT)15 primer. Reverse transcription was done in equiproportions (i.e., from equal quantity of RNA) within all compared samples from each experiment. The cDNA synthesis reaction mixture was diluted 10-fold in distilled water before using in real-time PCR.
Gene-specific primers for real-time PCR reactions were designed using Primer 3 input software v. 0.4.0 (http://bioinfo.ut.ee/primer3-0.4.0/primer3/input.htm), to have melting temperatures of 58–60°C and generate PCR products of 50-200 bp. The elongation factor (EF1-alpha) was used as endogenous (housekeeping) gene in order to normalize experimental results [26]. Primer sequences and the accession numbers of the corresponding genes are as follows: EF1-alfa (forward, 5´- GTACGCATGGGTGCTTGACAAACTC-3´; reverse, 5´-TCAGCCTGGGAGGTACCAGTAAT-3´; GenBank accession no. XM_021860126); Nitrate reductase [NADH]-like (forward, 5´ -AGGACTGGACCATTGAGGTG-3´, reverse 5´- GCTGCAGAACCCCAATTAAA-3´; Acc. no. XM_021892662.1); Glutamine synthetase cytosolic isozyme 1-1-like (forward 5´- AAAGGATATTTCGAGGACAGGAGG-3´, reverse 5´- CTTGAGAGACAGCTGCAGATT-3´); Acc. no. XM_021911887.1; NRT1/PTR Family (that is, NPF) 6.3-like (a LATS homologue) (forward, 5´-GAGACATGGCTAGCTGAGGA-3´; reverse, 5´-CCTTTTAGGCATGACATTAGCTACT-3´; Acc. no. XM_021912755); High-affinity Nitrate Transporter (that is, HATS) 2.1-like (forward, 5´-ATGTTGCTGAGTACGACGAC-3´; reverse, 5´- GGGACGTTGTGTAGGGGTAG-3´; Acc. no. XM_021864786).
Each PCR reaction contained 10 µL of 2X SYBR Green PCR master mix (Agilent TecHNologies), 50 ng of cDNA, and 0.45 µM (final concentration) of each primer, in a final volume of 20 µL. Real-time PCR reactions were run in an Agilent Mx3000P QPCR System (Agilent TecHNologies). The initial denaturing time was 3 min at 95°C, followed by 35 PCR cycles consisting of 95°C for 30 s and 60°C for 20 s. After the PCR cycles, the purity of the PCR products was checked by analysis of the corresponding melting curve. The comparative 2(-ΔΔCT) method was used to quantify the relative abundance of transcripts [61]. Experiments included three biological replicates, and three technical replicates for each biological replicate were performed.
Statistical analyses
Data were analyzed with the STATISTICA v6.0 software package (Statsoft Inc., Tulsa, OK, USA, www.statsoft.com). One-way ANOVA was used to identify differences among ecotypes during germination.
To analyze metabolite profiles and compare both ecotypes, first an exploratory data analysis was developed using principal components analysis (PCA) by means of scatterplot3d R package (Suppl. Fig 2). Then, from each ecotype the 100 metabolites with the lowest variance between replicates from each ecotype were selected, giving a total of 161 unique metabolites (Suppl. Table S1), and those without assigned molecular identity were removed. Thus, 85 remnant metabolites were analyzed by a heatmap including metabolite clustering analysis.
For the determination of sufficient and low nitrate supply, we used two-way ANOVA (level of significance P ≤0.05). Data from effects of nitrate supply on the different quinoa ecotypes (Socaire and Faro) were analyzed by two-way ANOVA. Fisher test was used to identify means with significant differences (level of significance P ≤0.05).