Generation of CRISPR-edited homozygous TaPHT1;9 wheat mutant lines
A 23 bp sgRNA sequence (5′-AGCCACATCAGACATGTCCAAGG-3′) was designed targeting the conserved exon regions of three TaPHT1;9 homoeologs (TaPHT1;9-4A/TraesCS4A02G416400.1, TaPHT1;9-4B/TraesCS4B02G317000.1, and TaPHT1;9-4D/TraesCS4D02G313700.1), and the expression cassette was transformed into the immature embryos of wheat cultivar 'Fielder' by using CRISPR/Cas9 (Fig. 1a, b). A total of 87 independent plants (T0) were regenerated. To characterize the mutation types, the sequencing information of T0 plants were amplified with homoeolog specific primers (Table S1) and compared with wild-type (WT) 'Fielder', and these lines were tracked from T0 to T3 generations. Two homozygous mutant lines in T3 with three homologs-edited TaPHT1;9 were confirmed: tapht1;9 − 1 with single base deletion and multi base deletions, tapht1;9 − 2 with single base deletion and single base insertion (Fig. 1c), which were used for this following experiment.
The decreased grain yields and yield components in CRISPR-edited TaPHT1;9 wheat mutant lines under soil condition
The above tapht1;9 − 1 and − 2 wheat mutant lines and WT ('Fielder') were sowed in soils with exogenous low (0 mg kg− 1, P0; 5 mg kg− 1, P5), or exogenous sufficient (20 mg kg− 1, P20) Pi fertilizer supplies. At the filling stage, phenotypes of two wheat mutants in P0 or P5 were visibly weaker than WT, whereas two mutant lines in P20 exhibited similar growth phenotypes with WT (Fig. 2a). At maturity stage, some quantitative parameters (dry weights of roots and stems, grain number per spike, 1000-grain weight and grain yields) of two CRISPR-edited TaPHT1;9 wheat mutant lines sowed in soils with the above three Pi treatments were significantly lower ≤ 14.48% than WT (Fig. 2b-d, f and g), whereas there was no significant change in spike number per plant between the edited mutants and WT in P0, P5 and P20 treatments (Fig. 2e). Grain yields of two wheat mutant lines were obviously decreased by ≤ 13.55%, ≤ 12.41% and ≤ 11.02% than WT in P0, P5 and P20 treatments, respectively.
The inhibited P absorption and PUEs in CRISPR-edited TaPHT1;9 wheat mutant lines under soil condition
Consistent with the above growth parameters and two yield components, P accumulations in roots of two wheat mutant lines were also decreased by ≤ 17.27% and ≤ 14.07% than WT in P5 and P20 treatments, respectively (Fig. 3a), whereas there was insignificant difference in root P accumulations between two wheat mutant lines and WT when they were grown in P0 treatment (Fig. 3a). P accumulation in straws and grains of two wheat mutant lines were also significantly lower (≤ 27.11%) than those in WT under three Pi-fertilizer treatments (Fig. 3b, c). Aboveground (straws and grains) P accumulations of wheat mutants were markedly decreased by ≤ 16.61%, ≤ 15.42% and ≤ 11.69% than WT in P0, P5 and P20 treatments, respectively (Fig. 3b, c). PUEs of only one wheat mutant (tapht1;9 − 2) was significantly lower than WT in P0 treatment, whereas PUEs of both two wheat mutant lines were all decreased by ≤ 4.56% than WT in P5 treatment (Fig. 3d). Under P20 treatment, PUEs of two wheat mutant lines were insignificantly reduced compared to WT (Fig. 3d).
The increased grain yields and yield components in transgenic rive plants expressing TaPHT1;9 under soil condition
Two ectopic transgenic rice lines expressing TaPHT1;9 (OE1 and OE3) and WT ('Fielder') were sowed in soils with three exogenous Pi supplies (no Pi fertilizer, 0 mg kg− 1, P0; low Pi fertilizers, 40 mg kg− 1, P40; or sufficient Pi fertilizers, 80 mg kg− 1, P80). At maturity stage, phenotypes of two transgenic rice lines and WT were photographed. Our results indicated that, different from the above two CRISPR-edited TaPHT1;9 mutant lines, these two transgenic rice lines (OE1 and OE3) sowed in both P0 and P40 treatments were visibly better than WT, while similar phenotypes appeared between two transgenic rice lines and WT in P80 treatment (Fig. 4a). Consistent with the above qualitative phenotypes, some quantitative growth and yield parameters (straws dry weights, grain yields, panicle number per plant, spikelet number per panicle and 1000-grain weight) exhibited the significant increase (≥ 6.08%) in both P0 and P40 treatments, whereas insignificant change between two transgenic rice lines and WT occurred in P80 treatment (Fig. 4b-f).
The increased P absorption and PUEs in transgenic rive plants expressing TaPHT1;9 under soil condition
Under both P0 and P40 treatments, P accumulations in straws and grains of two transgenic rice lines were higher than those in WT (Fig. 5a, b). In general, the aboveground (straws and grains) P accumulation of two rice lines were markedly increased by ≥ 14.64% and ≥ 11.65% than WT under P0 and P40 treatments, respectively. Phosphorus utilization efficiencies (PUEs) of above-mentioned plants were also calculated. The results showed that there was no significant difference in PUEs between transgenic plants and WT when they were grown in P80 soils (Fig. 5c). Under P0 and P40 treatments, however, PUEs of two transgenic rice lines increased by ≥ 3.88% and ≥ 5.67% (Fig. 5c), respectively. With the increase of exogenous Pi fertilizer supplies, above P accumulations of transgenic plants and WT showed an increasing trend, however, PUEs under P80 were lower than P0 and P40 (Fig. 5).