The role of PROTON1 in controlling the plasticity of ERD in response to nitrogen
The significant association for plasticity of ERD in response to N availability was detected based on the SNPs of 58 fully sequenced accessions, but was not found when using the 250K SNPs due to the absence of the identified SNPs in this dataset. To provide additional support for this association, we sequenced the missing region in the 250K SNP data that contains the biallelic SNP for the 90 accessions that were included in the initial phenotyping panel (Table S5). We then looked at the associations of the sequenced SNP to the plasticity in ERD in response to nitrogen. From all analyzed accessions, 42 carried the minor alleles (A and T, for positions 6904858 and 6904935, respectively), and 106 accessions carried the major alleles (C and G, respectively). On average, the accessions with minor alleles showed 25 % lower FC in ERD in comparison to the accessions with major alleles (0.55 and 0.41, respectively) (Figure 3d). The difference between the ERD FC of the two haplogroups was significant (p = 0.0016, Mann Whitney U test), giving further support to the association of PROTON1 to the ERD plasticity in response to N.
Further investigations of growth using the proton1 T-DNA mutant showed that its smaller FC in ERD was due to a significantly larger early rosette under limiting N, and to a smaller rosette diameter under optimal N in comparison to the WT (Figure 3c). This observation suggests that the reduced plasticity, thus more stable growth, between limiting and optimal N comes at a cost of decreasing the maximum attainable size. Therefore, our findings point that PROTON1 mediates the trade-off between stable and maximal growth under different N conditions.
Gene expression analysis of rosette leaves of plants grown under limiting and optimal N revealed that proton1 is a strong knockdown mutant, showing more than 1000 times reduced expression ofPROTON1 in comparison to WT (Figure 4a). Furthermore, the expression of PROTON1 was increased in WT under limiting N (Figure 4b). Additionally, we observed that the proton1 flowered earlier than the WT under limiting N, but that its FT was not different from that of the WT under optimal N (Figure 4c and d). To investigate if PROTON1 controlled plasticity of ERD in response to N availability is due to plasticity of N uptake, transport or assimilation, we evaluated expression of 18 genes involved in these processes in WT andproton1 mutant grown under either limiting or optimal N. From the 18 analyzed genes, seven showed a significant expression difference between WT and the mutant line in both optimal and limiting conditions including the ammonium transporters AMT1;1 , AMT1;4 ,AMT2;1 ; NIA2 , involved in N assimilation; NLP7transcription factor; and NRT1;5 and NRT3;1 , which are nitrate transporters (Figure 4e and 4f). Furthermore, the levels of N assimilation NIA1 gene and NLP1 transcription factor under limiting N, and the ammonium transporter AMT1;2 and NLP6transcription factor were also significantly different betweenproton1 and the WT under optimal N (Figure 4f). All of the transcripts that showed significantly perturbed expression had higher expression in the mutant line in comparison to the WT, indicating that PROTON1 is involved in processes that lead to repression of expression of genes involved in the aforementioned N-related processes. In addition, it also suggests that the increased plasticity is associated with increased expression of genes involved in N responses. The mechanism by which PROTON1 modulates the expression of these genes remains unknown.
To further confirm the role of PROTON1 in controlling the plasticity of ERD to N availability, we searched the published genome sequences of Arabidopsis accessions for polymorphisms in PROTON 1(At1g19880 ) gene. We identified that Xan-5 has a missense substitution that results in a disruption in the start codon ofPROTON1 . The presence of this SNP in Xan-5 was further confirmed by sequencing. When we grew Xan-5, Col-0 and proton1 under optimal and limiting N, we observed reduction in the plasticity of ERD in Xan-5 as well, thus further supporting the proposed role ofPROTON1 (Figure S3).