Discovery of a second‐site nia2 mutation in the background of multiple Arabidopsis PIF‐related mutants containing the pif3‐3 allele

Pleasenote:Wileyisnotresponsible forthecontentorfunctionality of any Supporting Information supplied by the authors. Any queries (other than missing material) should be directed to the New Phytologist Central Ofﬁce.

Nitrogen (N) is one of the most needed mineral nutrients for plants due to its involvement in the biosynthesis of proteins, nucleic acids, and other essential cellular components such as chlorophyll and phytohormones (Crawford, 1995).Plants can extract N from the soil in a variety of forms, but nitrate (NO 3 À ) represents the predominant source of N in most agricultural soils (Crawford & Forde, 2002).Nitrate is taken up by the root via NO 3 À transporters belonging to the NITRATE TRANSPORTER1/ PEPTIDE TRANSPORTER FAMILY (NRT1/NPF) and NRT2 families (Wang et al., 2012;Krapp et al., 2014).Subsequently, most NO 3 À is translocated to shoot tissue to be reduced by nitrate reductase (NR) to nitrite (NO 2 À ), which is converted into ammonium (NH 4 + ) before being incorporated into the amino acid pool via the action of the glutamine synthase/glutamine oxoglutarate aminotransferase or glutamate synthase (GS/ GOGAT) pathway (Campbell, 1999).As the enzyme catalysing the rate-limiting step of NO 3 À assimilation, and a major enzymatic source for the biosynthesis of the nitric oxide (NO) signalling molecule (Chamizo-Ampudia et al., 2017;Khan et al., 2023), NR is under extensive regulation at both transcriptional and posttranscriptional levels in response to various endogenous and environmental factors (e.g. Park et al., 2011;Lambeck et al., 2012;Konishi & Yanagisawa, 2013;Marchive et al., 2013;Creighton et al., 2017;Jamieson et al., 2022).Among them, light, in part through activating sugar assimilation, tightly regulates NR activity because photosynthesis provides the expensive reducing energy required for NO 3 À reduction and needs to be coordinated with NO 3 À metabolism to maintain the carbon (C)/N metabolic balance (Hoff et al., 1994;Baslam et al., 2020).However, the molecular mechanism of how light signalling directly modulates NO 3 À metabolism remains largely elusive.PHYTOCHROME-INTERACTING FACTORs (PIFs) are basic helix-loop-helix (bHLH) transcription factors that negatively regulate light responses, and are degraded upon the activation of the phytochrome (phy) photoreceptors (Bae & Choi, 2008;Leivar & Quail, 2011;Xu et al., 2015).Moreover, accumulating evidence has demonstrated that PIFs are also targeted by other environmental (e.g.temperature) and developmental (e.g.phytohormones) signals to modulate plant responses, making them central hubs that integrate external stimuli to downstream biological activities (Balcerowicz, 2020;Sanchez et al., 2020).Therefore, we investigated whether PIFs additionally play a role in regulating NO 3 À metabolism in the model plant Arabidopsis thaliana.
In this study, we used plant chlorate resistance to assess NO 3 À metabolic capacity (see Supporting Information Notes S1 for Methods and materials).Chlorate is a NO 3 À chemical analogue that can be taken up by NO 3 À transporters and reduced by NR to chlorite, which is toxic to plants and causes leaf chlorosis (Wilkinson & Crawford, 1991;Tsay et al., 1993;Wang & Crawford, 1996).Under our growth conditions, Arabidopsis nrt1.1 (lacking the dualaffinity NO 3 À transporter NRT1.1) and chl3-5 (lacking the major isoform of NR, NIA2, which accounts for 80-90% of total NR activity) mutants were resistant to chlorate toxicity (Fig. S1; Wilkinson & Crawford, 1991;Tsay et al., 1993), whereas the nrt2.1 nrt2.2 (lacking two high-affinity NO 3 À transporters NRT2.1 and NRT2.2) and nia1-4 (lacking the minor NR isoform, NIA1) were not (Fig. S1).Next, given that PIFs have both shared and distinct regulatory roles (Jeong & Choi, 2013), we tested the chlorate response of the sextuple mutant pqp6p7 (pif1-1 pif3-3 pif4-2 pif5-3 pif6-2 pif7-1), which lacks six of the eight PIFs, thus reducing the risk of potential PIF functional redundancy masking any mutant phenotype.Compared with wild-type (WT), pqp6p7 plants displayed significantly less leaf chlorosis caused by chlorate toxicity (Fig. 1a,b), suggesting that NO 3 À metabolism was compromised in this mutant.Since the pqp6p7 mutant was generated by crossing the quadruple pifq mutant (pif1-1 pif3-3 pif4-2 pif5-3) with pif6-2 and pif7-1, we attempted to narrow down the causal mutation by treating these mutants with chlorate.Although a moderate level of chlorate resistance was observed for pif7-1, only pifq mutant plants exhibited a comparable level of chlorate resistance to pqp6p7 (Fig. 1c).Finally, we subjected the individual pif mutants that makeup pifq to chlorate treatment and found that pif3-3 exhibited chlorate resistance comparable with that of pifq and pqp6p7 (Fig. 1d), suggesting that the resistance originated from there.
Since chlorate resistance can be attributed to a root-dependent deficiency in NO 3 À /chlorate uptake, or a shoot-dependent deficiency in NO 3 À /chlorate reduction, or both, we speculated that grafting might provide useful insight into the mechanisms of modified chlorate responses exhibited by mutants.As positive controls, we demonstrated that the graft chimaeras with WT shoot on nrt1.1 root (WT/nrt1.1)and chl3-5 shoot on WT root (chl3-5/ WT) were resistant to chlorate, whereas the reciprocal graft chimaeras (nrt1.1/WTand WT/chl3-5) were as sensitive to chlorate toxicity as the WT/WT controls (Fig. S2a,b).Subsequently, grafting was performed between WT and pif3-3 mutant plants.Among the resultant chimaeras, only those with a pif3-3 mutant shoot (pif3-3/pif3-3 and pif3-3/WT) were resistant to chlorate (Fig. 2a), suggesting a clear shoot-dependency of the chlorate resistance originating from the pif3-3 mutant.Accordingly, compared with WT and the other pif mutants, mutants containing the pif3-3 allele exhibited a drastically reduced NR activity (Fig. 2b).Despite the lack of a detectable difference in the transcript levels of the two NR genes, NIA1 and NIA2 (Fig. S2c,  d), the reduction in NR activity was reflected by a lower NR protein abundance in pif3-3 and pifq than that of WT (Fig. 2c).Cell-free degradation assay showed that the rate of NR protein degradation was accelerated in pif3-3 compared with WT, indicating a reduction in NR protein stability (Fig. 2d).
The pif3-3 allele was generated by fast neutron mutagenesis, which induced a 2.5-kb deletion in the promoter region of PIF3 that resulted in no detectable transcript of this gene (Fig. S3a; Monte et al., 2004; see Table S1 for primers used in this study).Two other pif3 mutant alleles, pif3-1 and pif3-2, contain T-DNA insertions in the fourth intron of PIF3 and produce truncated transcripts encoding a potential protein lacking a functional bHLH domain (Fig. S3a; Monte et al., 2004).Intriguingly, unlike pif3-3, the pif3-1 and pif3-2 mutants exhibited similar chlorate responses and NR activities to WT (Figs 3a, S3b).To rule out the possibility that the truncated transcripts produced by pif3-1 and pif3-2 were functional in maintaining a WT level of NR activity, we generated our own pif3 mutants using clustered regularly interspaced palindromic repeats (CRISPR)/CRISPR-associated protein 9 (Cas9) that introduced frame shifts early in the gene (Fig. S3c), and confirmed that their NR activity was not affected by the mutations (Fig. 3b).Finally, we showed that the overexpression of PIF3 in the pif3-3 mutant background led to an elongated hypocotyl phenotype characteristic of PIF over-accumulation, but did not revert the repression of NR activity (Figs 3c, S3d,e).Collectively, we conclude that the observed low NR activity in pif3-3 is not due to the mutation in PIF3 per se, but rather to another mutation in the background, presumably caused by fast neutron bombardment.In order to identify the actual mutation in the background of pif3-3 that is responsible for its chlorate resistance and reduced NR activity, whole-genome sequencing analysis was performed for the pif3-3 single mutant.We identified multiple single nucleotide variants (SNVs) and INDELs (insertions and deletions) present in the pif3-3 mutant (Datasets S1, S2).Among them, a SNV in the NIA2

Forum
New Phytologist gene of pif3-3 led to a single amino acid substitution (Y851N) in the NADH-binding domain of the NIA2 protein (Fig. 3d).To confirm that this mutant nia2 allele (which we name nia2-2) compromised NR activity, we performed allelism tests by crossing pif3-3 (with nia2-2 in the background) with known nia2 loss-of-function mutants, chl3-5 (deletion of the entire NIA2 gene) and nia2-1 (T-DNA insertion nia2 mutant), respectively.The F1 progeny from these crosses exhibited a similar level of NR activity as the parental pif3-3 nia2-2 and chl3-5 or nia2-1 mutants (Figs 3e, S3f).On the contrary, F1 plants from crossing pif3-3 nia2-2 with WT had an intermediate level of NR activity (Figs 3e, S3f), which is consistent with the fact that the NIA2 gene is haploinsufficient (Wilkinson & Crawford, 1991).The F1 PIF3/pif3-3 NIA2/nia2-2 heterozygous plants were allowed to self, and from the F2 population, we isolated individuals with only one of these two mutations, that is pif3-3 NIA2 and PIF3 nia2-2.NR activity assay revealed that the nia2-2 mutation is solely responsible for the decreased NR activity in the original pif3-3 mutant and without it in the background, independent pif3-3 NIA2 lines isolated from the F2 population of a WT (PIF3 NIA2) 9 pif3-3 nia2-2 cross do not have a detectably different NR activity compared with WT (PIF3 NIA2, Fig. 3f).
The Y851 residue locates within the C terminus domain of the NIA2 protein that binds to NADH, which provides the electron essential for NO 3 À reduction (Campbell & Kinghorn, 1990).Furthermore, Y851 was found to be highly conserved in NIA orthologs across the plant kingdom from green algae to angiosperms, except for one of the two NIA homologues in Physcomitrium patens, whose corresponding site was occupied by a chemically similar phenylalanine (F) residue (Fig. S3g).Therefore, it is reasonable to speculate that Y851 at this position confers an important function, and substituting it with an asparagine (N) that is both structurally and chemically distinct from Y probably destabilises the protein (Fig. 2c,d).However, the protein abundance of NR in nia2-2 was still higher than that in chl3-5 or nia2-1, even though the NR activity was indistinguishable between the three nia2 mutants (Fig. 3g,h).Therefore, the Y851N substitution might additionally disrupt normal NR enzymatic activity.
Taken together, we report here the discovery of a hidden nia2-2 mutation in multiple mutants containing the pif3-3 allele and show that nia2-2 confers reduced NR activity, thus compromising NO 3 À metabolism.This finding indirectly suggests that light-mediated regulation of NR activity is unlikely to involve most of the PIF family proteins, although the mild chlorate resistance phenotype conferred by pif7-1 is worth exploring in future studies.The characterisation of the nia2-2 mutation also highlights the functional importance of Y851 in maintaining NR protein stability and enzymatic activity and provides a valuable mutant resource for studying NR functions.Lastly, we present a pipeline that can be referred to when studying a regulatory component of NO 3 À metabolism.Specifically, we highlight the use of grafting as a novel and effective approach for spatially separating and distinguishing the effect on NO 3 À uptake in the root and NO 3 À assimilation in the shoot.The pif3-3 mutant and higher-order mutants containing the pif3-3 allele (e.g.pifq and pqp6p7) have been widely used to study PIF functions (e.g.Jiang et al., 2020;Bernula et al., 2021;Yoo et al., 2021;Piskurewicz et al., 2023;Sng et al., 2023).The identification of the hidden nia2-2 mutation revealed here might render it worthwhile re-examining the alternative interpretation that some of the reported mutant phenotypes were due to the deficiency in N metabolism and/or NO production, especially for any studies investigating the crosstalk between the PIF and NO signalling pathways (Lozano-Juste & Le on, 2011;Bai et al., 2014).We were initially surprised that the nia2-2 allele somehow evaded being eliminated through multiple rounds of crossing and persisted in the quadruple and sextuple pif mutants (especially considering that the original pif3-3 mutant was outcrossed twice with Col-0 following its isolation, Monte et al., 2004).There is clearly not an absolute linkage between the PIF3 and NIA2 genes because we have successfully separated the two mutant alleles in this study.However, the fact that they reside on the same chromosome (PIF3: AT1G09530; NIA2: AT1G37130; Fig. S3h) suggests that they might be partially linked, and therefore, the mutant alleles may not assort completely independently during meiosis.
Forward genetics is a powerful approach for understanding gene function underlying a particular mutant phenotype induced by random mutagenesis in an unbiased manner that requires no prior knowledge about the gene of interest (Peters et al., 2003).However, random mutagenesis poses the risk of generating second-site mutations that are responsible for the observed phenotype but are overlooked, which led to a plethora of recent publications unmasking these mutations in previously reported mutants (e.g.Bennett et al., 2006;Westphal et al., 2008;Enders et al., 2015;Gao et al., 2015;Kriegel et al., 2015;Wu et al., 2015;Yoshida et al., 2018;Vlad & Langdale, 2022;Yu et al., 2023).Therefore, we wish to remind the science community to not discount the possibility of a mutant material harbouring secondary mutations in the background, which seems to be particularly common for those generated from nonspecific mutagenesis.If possible, the use of other independent mutant alleles should always be included in their studies.Alternatively, complementation tests (using complete complementation gene constructs) can be performed to validate gene functions.
Shoot-dependent deficiency in nitrate reductase (NR) activity underlies pif3-3conferred chlorate resistance.(a) Chlorate responses of graft chimaeras made by exchanging shoots and roots of wild-type (WT) and pif3-3 mutant plants.The graft chimaeras are labelled as shoot genotype/root genotype.All graft chimaeras were constructed at seedling stage before being transplanted to soil.Four-week-old plants were treated with dH 2 O (control) or 1.5 mM chlorate every 5 d.When necessary, bolt stems were removed for ease of viewing.Bars, 2 cm.(b) Mean relative shoot NR activity of WT and PHYTOCHROME-INTERACTING FACTOR (PIF)-related mutants.Red dots indicate individual values (n = 4), error bars indicate SD, and different letters (a, b) indicate significant differences (one-way analysis of variance (ANOVA) with Tukey's test).(c) Abundance of immuno-detected NR in WT, pif3-3, and pifq plant extracts, quantified against Actin control (arbitrarily set at 1.00 for WT).Ponceau S staining serves as loading control.(d) Destruction rates of endogenous NR in WT and pif3-3 plant extracts, with immunodetectable NR quantified against Actin control (arbitrarily set at 1.00 for time point 0).ND, band not detected.Ó 2023 The Authors New Phytologist Ó 2023 New Phytologist Foundation New Phytologist (2024) 241: 17-23 www.newphytologist.com