Comparative transcriptomic and physiological analyses unravel wheat source root adaptation to phosphorous deficiency

Phosphorus (P) is a crucial macronutrient for plant growth and development. Basic metabolic processes regulate growth; however, the molecular detail of these pathways under low phosphorous (LP) in wheat is still unclear. This study aims to elucidate the varied regulatory pathways responses to LP stress in wheat genotypes. Phenotypic, physiological, and transcriptome analyses were conducted on Fielder (P efficient) and Ardito (P inefficient) wheat genotypes after four days of normal phosphorous (NP) and LP stress. In response to LP, Fielder outperformed Ardito, displaying higher chlorophyll content-SPAD values (13%), plant height (45%), stem diameter (12%), shoot dry weight (42%), and root biomass (75%). Root structure analysis revealed that Fielder had greater total root length (50%), surface area (56%), volume (15%), and diameter (4%) than Ardito under LP. These findings highlight Fielder’s superior performance and adaptation to LP stress. Transcriptome analysis of wheat genotype roots identified 3029 differentially expressed genes (DEGs) in Fielder and 1430 in Ardito, highlighting LP-induced changes. Key DEGs include acid phosphatases (PAPs), phosphate transporters (PHT1 and PHO1), SPX, and transcription factors (MYB, bHLH, and WRKY). KEGG enrichment analysis revealed key pathways like plant hormones signal transduction, biosynthesis of secondary metabolites, and carbohydrate biosynthesis metabolism. This study unveils crucial genes and the intricate regulatory process in wheat’s response to LP stress, offering genetic insights for enhancing plant P utilization efficiency.

9 and 13 PHT1s have been found in Arabidopsis and rice.Functional investigations have been conducted utilizing mutant and transgenic plants of PHT1 genes 6,7 .A universally conserved master regulator known as the transcription factor PHOSPHATE RESPONSE 1 (PHR1) coordinates the transcriptional activation of a significant portion of phosphate starvation-response (PSR) genes 8 .The nuclear proteins SYG/PHO81/XPR1 (SPX) serve as Pi-dependent competitive inhibitors of PHR1 activity in both Arabidopsis and rice 9,10 .Moreover, other Pi-responsive transcription factors like WRKY, MYB, and bHLH families have been identified in several plant species 11 .Furthermore, miR399 decreases the transcript level of PHO2, which encodes a ubiquitin-conjugating E2 enzyme involved in the ubiquitination and degradation of PHT proteins 12 .
Wheat (Triticum aestivum L., BBAADD, 2n = 6x = 42) is the most widely cultivated staple food crop 13 .In recent years, advancements have uncovered several molecular mechanisms governing Pi signaling and homeostasis in wheat.Genes encoding Al-activated malate and citrate transporters (such as TaALMT1 and TaMATE1B) play a crucial role in facilitating the efflux of organic acids (OAs), thereby improving the efficiency of Pi utilization in agricultural settings 14 .Phytases are phosphatases that can initiate the stepwise hydrolysis of phytate and thereby provide phosphate.The presence of phytase activity in the wheat can be attributed to three genes (rTaPAPhy_a1, rTaPAPhy_b1, and rTaPhyIIa2) 15 .In wheat, TaPHT1;2, TaPht1;4, TaPHT1;9-4B, and TaPT2 (all members of the PHT1 family) facilitate Pi uptake under low-Pi conditions [16][17][18][19] , whereas TaPHT2;1 (from the PHT2 family) and TaPHO2-A1 are crucial for regulating Pi translocation 20,21 .Multiple transcription factors (TFs) play a role in wheat's response to phosphate deficiency.For instance, TaPHR-A1, TaNFYA-B1, TaPHR3-A1-A, and TaMYB4-7D have been demonstrated to enhance the expression of several TaPHT1 genes in transgenic wheat plants 19,[22][23][24] .Furthermore, wheat miRNAs such as TaMIR1139 and TaMIR399 target genes across different families, playing pivotal roles in regulating plant tolerance to Pi starvation 20,25 .Metabolomic analysis presents a valuable tool in plant research for gaining deeper insights into plant responses to stress, particularly Pi starvation.This approach sheds light on the metabolic effects of Pi deficiency in wheat, revealing alterations in pathways related to carbohydrates, amino acids, and secondary metabolism 26,27 .Additionally, proteomic analysis revealed significant alterations in the abundance of proteins associated with nitrogen and phosphorus, small molecule, and carboxylic acid metabolic processes in response to Pi deficiency in wheat 28 .Plant hormones, such as auxin, cytokinin (CK), ethylene (ETH), abscisic acid (ABA), gibberellin (GA), jasmonic acid (JA), and salicylic acid (SA), undergo changes that play crucial roles in integrating Pi signaling and regulating root growth [29][30][31] .However, most studies have only investigated the effects of LP stress on one or two phytohormones in roots, with minimal available data on the comprehensive effects of LP stress on the phytohormone metabolome in wheat roots.
The plant's response to low phosphorus (LP) stress is a multifaceted process regulated by the interplay of numerous genes.Despite this complexity, the transcriptional and metabolic mechanisms in wheat under LP stress remain relatively understudied.Advances in sequencing technologies, including metabolome and transcriptome analyses, offer promising avenues to accelerate the identification of molecular mechanisms governing plant responses to LP stress.This study delved into the physiological, transcriptomic, and targeted metabolomic responses of Fielder (P-efficient) and Ardito (P-inefficient) to elucidate key differentially expressed genes and metabolic pathways essential for LP response in wheat.These findings establish a fundamental groundwork and serve as a valuable resource for further exploration of gene regulatory mechanisms and a deeper understanding of the molecular processes involved in LP response and adaptation.

Result Phenotypic changes and phosphorus efficiency under LP stress
After undergoing low phosphorus (LP) treatment, two wheat genotypes experienced significant (p < 0.05) reductions in plant growth.Parameters such as SPAD, plant height, root dry weight, and shoot dry weight were significantly lower following LP treatment compared to normal phosphorus (NP) treatment in both genotypes (Table 1).Root structure analysis revealed decreased total root length, surface area, volume, and diameter in both genotypes under LP stress (Table 2).Notably, Fielder exhibited increased phenotypic indices under low-P stress compared to Ardito.Root images depicted elongated roots (greater root length) in Fielder compared to Ardito under LP (Figure S1).Overall, Fielder demonstrated superior performance to Ardito under LP stress, as evidenced by higher values and lower degradation rates of phenotypic indicators.Regarding root P contents, Fielder and Ardito experienced reductions of 59% and 70% under LP compared to NP, with Fielder showing a 95% increase in roots compared to Ardito.Additionally, shoots exhibited a 59% reduction in Fielder and a 56% reduction in Ardito under LP compared to NP, while shoot P was 79% higher in Fielder than Ardito under LP.Fielder accumulated more total P in both shoots and roots than Ardito, particularly in the roots (Fig. 1).These findings underscore Fielder's enhanced tolerance to Pi starvation and efficient Pi accumulation, designating it as the P-efficient line, while Ardito exhibited lower phosphorus tolerance.S2).The results ensured accurate replication within each group and highlighted distinct differences between genotypes and treatments.To validate RNA-Seq gene expression accuracy, qRT-PCR was conducted on 20 randomly selected DEGs (Table S2) using identical cDNA templates.
The results exhibited a highly significant positive correlation with transcriptome data, affirming the reliability of the RNA-Seq results (Fig. 2).

Identification of differentially expressed genes in response to LP stress
Pairwise comparisons of genotype and treatment identified genes associated with Pi starvation.Fielder exhibited 3029 DEGs (1591 upregulated and 1438 downregulated), while Ardito had 1430 DEGs (1176 upregulated and 254 downregulated) compared to NP treatment.Surprisingly, Fielder had more DEGs than Ardito.P efficient genotype Fielder had 6704 DEGs (3328 upregulated and 3376 downregulated) compared to P inefficient genotype Ardito under LP stress (Fig. 3 and Table S3).

GO annotations and KEGG enrichment analysis
We performed GO annotation and KEGG pathway enrichment analyses of the DEGs identified in wheat under LP stress.In total, 3029 DEGs were grouped in 45 terms in Fielder, 1493 DEGs were grouped in 44 terms in Ardito, and 6704 DEGs were grouped in 47 terms in the comparison Fielder with Ardito were discovered in the GO database.Metabolic process (GO: 0008152), cellular process (GO: 0009987), and response to stimulus (GO: 0050896) were the most abundant GO terms annotated in the biological process (BP) in the three comparison groups (Table S4).Furthermore, membrane part (GO: 0044425), cell part (GO: 0044464), and organelle  (GO: 0043226) were the most abundant GO terms in the cellular component (CC) (Table S4).Finally, binding (GO: 0005488) and catalytic activity (GO: 0003824) were the most enriched GO terms in the molecular function (MF) category (Fig. 4 and Table S4).The next step is to understand the functions of these significantly enriched DEGs.The KEGG pathway enrichment analysis showed that under LP stress conditions, the top 20 enriched KEGG pathways in Fielder, Ardito, and Fielder vs. Ardito are presented in Fig. 5. DEGs from three comparisons were enriched in two identical pathways, 94 genes in Fielder, 82 genes in Ardito, and 169 genes in Fielder vs. Ardito under LP stress were significantly enriched in phenylpropanoid biosynthesis (ko00940) pathway; 13 genes in Fielder, 11 genes in Ardito, and 21 genes in Fielder vs. Ardito were enriched in arginine and proline metabolism (ko00330) pathway.In addition, for Fielder, protein processing in the endoplasmic reticulum (ko04141, 117 genes), diterpenoid biosynthesis (ko00904, 17 genes), and nitrogen metabolism (ko00910, 14 genes) were the three main pathways.For Ardito, tryptophan metabolism (ko00380, 20 genes), Cysteine and methionine metabolism (ko00270, 20 genes), and diterpenoid biosynthesis (ko00904, 16 genes) were the three main pathways with many DEGs.For the comparison of Fielder with Ardito, plant-pathogen interaction (ko04626, 107 genes), glutathione metabolism (ko00480, 57 genes), MAPK signaling pathway-plant (ko04016, 52 genes), and starch and sucrose metabolism (ko00500, 52 genes) were the enriched pathways with the highest number of DEGs (Fig. 5).The differential expression of genes in response to LP stress-activated varied molecular mechanisms in various lines, as revealed by enriched KEGG pathways.

DEGs related to pulper acid phosphatase and phosphate transporter
The transcriptome data revealed the identification of seven differentially expressed genes (DEGs) that encode pulper acid phosphatase (PAP) under LP stress (Fig. 6 and Table S5).Among these, five PAP genes, including one PAP2, three PAP27, and one PAP29, were significantly upregulated in Fielder; two PAP7 and one PAP29 genes were upregulated in Ardito under LP stress.One PAP1, one PAP27, and one PAP29 genes were upregulated, and one PAP1 gene was downregulated in Fielder compared with Ardito under LP stress.Four genes encoding phosphate transporter (PHT), including two PHT1-8 and one PHO1-3, were upregulated, and one PHO1-2 was downregulated in Fielder, but no response in Ardito.In addition, one SPX1 and two SPX6 genes were upregulated, and one SPX3 gene was downregulated in Fielder compared with Ardito under LP stress (Fig. 6 and Table S5).

DEGs related to plant hormones signal transduction
The KEGG enrichment analysis showed that 52 DEGs enriched in the MAPK signaling pathway were identified under LP stress.Thirty-nine DEGs involved in auxin, abscisic acid (ABA), brassinosteroid (BR), ethylene (ETH), gibberellin (GA), jasmonic acid (JA), salicylic acid (SA), and zeatin (ZT) signaling pathways were identified to be associated with plant hormone signal transduction (Fig. 7 and Table S6).In the auxin signaling pathway, one ARF, one GH3, and two IAA genes were upregulated, but one ARF and three SAUR genes were downregulated in Fielder compared with Ardito under LP stress.One ARF, one GH3, and two SAUR genes were upregulated in Ardito, but one SAUR gene was downregulated in Fielder.In the abscisic acid signaling pathway, one PYL and one www.nature.com/scientificreports/PP2C genes were upregulated in Fielder and Fielder vs. Ardito, respectively, while three PP2C and one ABF genes were downregulated.In the brassinosteroid signaling pathway, two C92C genes were downregulated, and one BR gene was downregulated in Fielder compared with Ardito.One ERF1 and one EIN2 genes were upregulated in the ethylene signaling pathway in Ardito and Fielder vs. Ardito.In contrast, two EIN and two ERF1 genes in Fielder, three EIN genes in Ardito, and one ERF1 gene in Fielder vs. Ardito were downregulated.In the gibberellin signaling pathway, two CPS4, one GA3, and one KSL4 genes were upregulated in Fielder and Ardito, while one GA2 gene was downregulated in Fielder.In the jasmonic acid signaling pathway, four JAZ genes were upregulated in Fielder compared with Ardito.In the salicylic acid signaling pathway, one SABP2 gene was upregulated, and one SABP2 gene was downregulated in Fielder compared with Ardito.In the zeatin signaling pathway, four U73C and one CZOG genes in Fielder compared with Ardito and one CKX1 gene in Ardito were upregulated under LP stress.Two U73C genes were significantly downregulated in Fielder and Ardito, and one CKX1, one CZOG, and one U73C3 genes in Fielder compared with Ardito were downregulated under LP stress (Fig. 7 and Table S6).

DEGs related to secondary metabolic synthesis
Twenty-nine critical DEGs involved in secondary metabolic synthesis were significantly expressed in three comparisons, Fielder_LP vs. NP, Ardito_LP vs. NP, and Fielder vs. Ardito under LP stresses (Fig. 8 and Table S7).
In the pathway of phenylpropanoid metabolism, four PAL, one 4CL5, and two 4CL7 genes were upregulated, while one PAL, two 4CL7, and one C73A10 genes were downregulated in Fielder vs. Ardito under LP stress, respectively.One 4CL7 gene in Fielder, two PAL, and two C73A10 genes in Ardito were upregulated under LP stress.Most structural genes implicated in the flavonoid metabolism pathways were significantly upregulated.www.nature.com/scientificreports/For instance, four genes encoding shikimate o-hydroxycinnamoyltransferase (EC 2.3.1.133),four CYP75A genes, three genes encoding chalcone-flavanone isomerase (EC 5.5.16), two CHS, one CAMT, and one F3H genes were upregulated.Moreover, one CHS1 gene was downregulated in Fielder and Fielder vs. Ardito.One FLS and one C75A5 genes were downregulated in Fielder and Ardito.In the pathway of Anthocyanin metabolism, one BZ1 gene was downregulated in Fielder and Fielder vs. Ardito under LP stress (Fig. 8 and Table S7).

DEGs related to carbohydrate metabolism
Twenty-eight crucial DEGs involved in carbohydrate metabolism were significantly expressed in three comparisons under LP stress (Fig. 9 and Table S8).In the starch and sucrose metabolism pathway, three SUS genes were upregulated in Fielder and Ardito, and one SUS4 gene was downregulated in Fielder under LP stress.One UGP2 gene was upregulated in Fielder vs. Ardito under LP stress.One GBE1 gene was significantly downregulated in Fielder under LP stress.Two BAM3 genes were upregulated, and one BAM gene was downregulated.
In the glycolysis pathway, one GALM gene was upregulated in Fielder and Ardito and Fielder vs. Ardito under LP stress, respectively.One HK4 gene in Fielder vs. Ardito was downregulated, and one HK8 gene in Fielder was upregulated.One PFP gene was upregulated in Fielder and Ardito under LP stress.One ALDO3 gene was upregulated, and one ALDO1 gene was downregulated in Fielder vs. Ardito.One GAPDH2 and one PGK genes were downregulated in Fielder vs. Ardito.One ENO1 and one PPDK1 genes were upregulated in Ardito.Genes involved in the citrate cycle (TCA cycle) pathway were significantly expressed in the comparison Fielder with Ardito, but no response in Fielder and Ardito comparison LP with NP, including the upregulation of one ACLY gene and one ACO gene and the downregulation of the MDH2 gene and SDHA gene.

DEGs related to transcription factors (TFs)
GO analysis results showed that numerous DEGs involved in transcription factors (TFs) were annotated in the GO database in the present study.A total of 426 TFs from 29 families were significantly expressed in three comparisons under LP stress.MYB, AP2, and bHLH were the top three leading TF families, containing 82, 50, and 44 differentially expressed genes (Table 3 and Table S9).Three TFs MYB (TraesCS7D02G295400), bHLH (TraesCS5A02G049600), and NAC (TraesCS2A02G101900) families were expressed in three comparisons which showed a downward trend.However, numerous TFs showed different trends in the two cultivars under LP stress.For instance, for the WRKY family, two genes were upregulated, and one gene was downregulated in Fielder; four genes were upregulated, and one gene was downregulated in Ardito; but in Fielder comparison with Ardito, nineteen genes were upregulated, and three genes were downregulated under LP stress.For the HSF family, twenty-three genes were downregulated, and one gene was upregulated in Fielder; six genes were upregulated, and two genes were downregulated in Fielder vs. Ardito, but no significantly responded in Ardito under LP stress.In addition, two genes from the SPB family, two from the RAV family, and one from the Nin-like family  One gene from E2F/DP was upregulated in Ardito under LP stress (Table 3 and Table S9).These transcription factors showed specialty in three comparisons.

Phytohormone profiles in wheat roots
To confirm the transcriptomic findings, we performed targeted metabolomics analysis on 41 plant hormones in the roots, detecting 34 of them, including 7 AUXs (IAA, IAAla, IAAsp, IALeu, IAPhe, IATrp, and IPA), ABA, 6 GAs (GA1, GA3, GA4, GA7, GA8, and GA9), 2 GA synthesis inhibitors (UCZ and PP333), 3 JAs (JA, MeJA, and OPDA), SA, 10 CKs (mT, cZ, cZR, tZ, tZR, DZ, iP, iPR, 6-BA, and DPU), and 4 other plant growth regulators (MH, 1-NAD, 4-CPA, and 5-NG) (Table 4).The results were consistent with the transcriptome data.LP stress significantly increased the total auxin, total JA, and SA contents but decreased the total ABA and total CK contents in Fielder.Additionally, the total GA and SA increased, while the total IAA, CK, JA, and ABA content decreased in Ardito under LP stress.Furthermore, compared to Ardito, the total IAA, ABA, GA, JA, and SA content increased in Fielder under LP stress (Table 4).Regarding auxin, the IAA and IALeu content significantly increased, while the IAPhe content decreased in Fielder.The content of IAA has been increased by more than 2-fold in Fielder under LP stress.In contrast, the IAA, IAAsp, IAAleu, IAPhe, and IATrp content significantly decreased but increased the IAAla content in Ardito.As for GA, the GA1 and GA8 contents significantly increased, but the GA3 content decreased in Fielder.In Ardito, the GA7, GA8, and GA9 contents significantly increased, while the GA1, GA3, and GA4 content decreased.In terms of JA, the JA and OPDA content significantly increased, while MeJA decreased in Fielder.Conversely, the MeJA content decreased in Ardito, and the JA and OPDA content showed no significant influence.The total CK content accounts for about 80% of the total plant hormone content in the wheat root, with the highest being cZR.The CZ and DZ content increased, while the mT, CZR, and tZ content decreased in Fielder.In Ardito, the tZ content increased, while the cZ, cZR, tZR, and 6-BA content decreased.LP stress significantly decreased the MH and 4-CPA content in Fielder.Interestingly, 1-NAD was detected in Ardito under LP stress (Table 4).

PAP and PHT regulated P assimilation and remobilization
To respond to LP stress, plants have evolved lots of strategies known as phosphate starvation responses (PSR) regulated by phosphorus-starvation-induced (PSI) genes (PHT1, PAP, SPX) 32 .In our study, six PAP genes were upregulated.They had a higher expression level in P efficient genotype Fielder under LP stress, which can support the high release of phosphorus from organic phosphorus under LP stress.Plants increase the secretion of purple acid phosphatases (PAPs) into the rhizosphere to scavenge organic phosphorus for plant use under low phosphorus stress 33 .It found that Cm-PAP10.1 and Cm-PAP10.2,genes encoding purple acid phosphatases, were upregulated in melon under LP stress, the activity of PAPs secreted by plant roots under LP stress was positively correlated with the degree of P deficiency 34 .In rice, OsPAP10c overexpression increased acid phosphatase (APase) activity by almost fivefold in both roots and leaves under low phosphorus.Meanwhile, genes encoding phosphorus transporter proteins were screened in transcriptome data, among which four DEGs of high-affinity phosphorus transporter systems (PHT1 and PHO1 families) were screened; three of these were significantly upregulated in P efficient genotype but not response in P inefficient genotype Ardito treated with LP.The high-affinity transporter TaPHT1;9-4B and its transcriptional regulator TaMYB4-7D contributed to efficient Pi acquisition and plant growth under Pi-limiting conditions 35 .In Zygophyllum xanthoxylum, PHO1 and PHT1 genes were upregulated in the roots, suggesting that these phosphorus transporter protein genes may play a vital  role in regulating the distribution, transport, and maintenance of dynamic homeostasis of P in the plant body during P deficiency 36 .Our results that Fielder has a higher total phosphorus content both in the stem and at the root suggests that Fielder may have activated and mobilized phosphorus by secreting more PAPs and enhanced phosphorus absorption by PHTs.SPX-domain-containing proteins (SPXs) play a crucial role in the sensing, signaling, and transport of inorganic phosphate (Pi) in eukaryotes.AtSPX1 is a phosphate-dependent inhibitor of PHR1 in Arabidopsis; OsSPX1, OsSPX4, and OsSPX6 are involved in Pi starvation signaling and acting as a negative regulator of PHR in rice 9,10,37,38 .Transgenic plants overexpressing SPX6 exhibited decreased Pi concentrations and suppression of phosphate starvation-induced (PSI) genes 37 .In our study, four SPX genes were expressed in different trends under LP stress; further research is needed on the function of SPX in response to LP stress.

Plant hormones signal transduction in wheat root
LP stress disrupts plant hormone synthesis and distribution, consequently impacting root growth 5 .Our study reveals significant alterations in the expression of genes involved in plant hormone signal transduction, www.nature.com/scientificreports/underscoring the pivotal role of hormones in wheat root development under low phosphorus stress.Under Pi starvation, heightened auxin signaling at the root tip and lateral root primordia inhibits primary root growth while stimulating lateral root formation 31 .The growth hormone-responsive transcription factor SAUR modulates root morphology and fosters lateral root development 39,40 .Similarly, increasing cytokinin (CK) is crucial in maintaining the root-to-shoot ratio, thereby reducing CK concentrations under low phosphorus conditions 30 .Previous research indicates that the transcription factor MYB62 regulates phosphate starvation response (PSR) by modulating GA metabolism and signaling 41 , suggesting the necessity for MYB TFs and GA signaling regulation in wheat to cope with low P stress.Jasmonate (JA) and salicylic acid (SA) are additional hormones crucial for integrating environmental cues and influencing root growth under LP stress 32,42 .Our study demonstrates the upregulation of genes involved in auxin, CK, GA, JA, SA, and brassinosteroid (BR) signaling in Fielder compared to Ardito under low phosphorus stress, indicating their involvement in regulating wheat root morphology in response to low-P stress.Targeted metabolome results corroborate the transcriptome data, showing increased total IAA, ABA, GA, JA, and SA in Fielder than in Ardito under LP stress (Table 4).Under LP stress, Fielder exhibited a significant increase in total auxin, total JA, and SA contents while experiencing a decrease in total ABA and total CK contents.On the other hand, Ardito showed an increase in total GA and SA contents but a decrease in total IAA, total JA, total CK, and ABA contents under LP stress.Root morphological indicators were significantly higher in Fielder compared to Ardito, indicating Fielder's ability to expedite root development in response to stress through upregulation of gene expression and elevation of hormone contents.Brassinosteroids (BRs) are plant hormones that promote cell elongation and division, crucial for plant growth and development 43 .
Recent studies have implicated brassinosteroid signaling in regulating phosphate starvation-induced malate secretion in plants 44 .Additionally, ABA and ETH signaling were observed to be downregulated in both Fielder and Ardito under low phosphorus stress.While ABA and ETH signals may negatively regulate low phosphorus stress in roots, further investigation is required to ascertain the precise mechanism.

Secondary metabolic synthesis regulated to LP stress
Secondary metabolites, such as phenylpropanoids, flavonoids, and anthocyanins, can scavenge reactive oxygen species, delay microbial degradation of organic acids, and enhance mobilization of rhizosphere phosphorus [45][46][47] .Numerous DEGs were enriched in the phenylpropanoid and flavonoid metabolism pathways, suggesting that P starvation might affect the metabolic product content of secondary metabolic synthesis in wheat roots.In our study, PAL, CYP73A, and 4CL in the phenylpropanoid metabolism pathway were upregulated under LP stress, and E2.3.1.133,PCMT, and CYP75A were significantly upregulated in the comparison Fielder with Ardito, F3H was exclusively upregulated in Fielder and E5.5.1.6exclusively upregulated in Ardito suggesting the increase of genes related to flavonoid metabolism may be enhanced the low P tolerance in Fielder.Similar to our results, flavonoids were highly accumulated in P-efficient cotton genotype Jimain169 roots under low phosphorus due to the upregulation of the genes responsible for flavonoids 48 .In style, a set of genes involved in flavonoid synthesis were found to be upregulated by Pi starvation accompanied by the flavonoid metabolites, phenolic acids, and phenyl amides contents were increased in roots, which might facilitate P solubilization and cooperate with beneficial microorganisms in the rhizosphere, and thus contributing to P acquisition and utilization 49 .The above results indicate that Pi starvation altered the gene expression related to flavonoid biosynthesis.Genes and metabolites that regulate flavonoid metabolism might regulate phosphorus signaling and improve wheat tolerance to low www.nature.com/scientificreports/phosphorus stress.However, the flavonoid content decreased significantly in P-resistant maize roots under low phosphorus stress, which may be attributed to the levels of phosphorus supply 50 .This may be related to the supply of phosphorus levels.In soybeans, there is more anthocyanin production in LP and more isoflavonoid production in NP 51 .Anthocyanin is also one of the most observed flavonoids in the roots and shoots under LP stress.Among various functions of flavonoids in plants, the higher accumulation of anthocyanin in leaves has a vital role in the photo-protection of the leaf.However, the function of anthocyanin in the roots is still to be elucidated 52 .

Changes in carbohydrate biosynthesis under LP stress
Plants produce ATP and CO 2 through photosynthetic product breakdown (e.g., glucose) by respiration to promote root growth and development or maintain root activity for nutrient and water uptake and translocation 53 .
Higher root biomass and physiology are required for higher production and are important traits supporting normal plant growth under stress conditions.P-deficiency was reported to increase carbohydrate translocation via the phloem to roots to favor root growth for better acquisition of Pi from soil 54 .Fielder has a higher root morphology than Ardito under low phosphorus stress but at the cost of an increased response to root-to-shoot ratio.In the previous study, the P efficient genotype had a more remarkable ability to maintain phosphorylated sugars (i.e., glucose-6-P and fructose-6P) by upregulating the genes involved in glycolysis, starch, and sucrose synthesis that are important for glycolysis as well as the biosynthesis of sugars and starch 55,56 .The above results suggest that wheat roots maintain ATP supply by continuously altering the transcription levels of genes encoding key enzymes in the glycolysis/gluconeogenesis pathway.Under low P conditions, SUS, UGP, GALM, HK, PFP, ALDO, and BAM were upregulated in Fielder, suggesting that Fielder better maintains the phosphorylated sugars required for sugar and starch biosynthesis to regulate the metabolic processes required for energy and carbon skeleton production to supply root growth 57 .
Plants secrete organic acids from TCA cycle derivatives (Major citrate and malate) to help solubilize sparingly soluble inorganic phosphate and mobilize Pi from the soil 58 .Citrate exudation from roots increased P uptake from Fe-P in soils 59 .Similarly, in the case of GhmMDH1 from cotton (Gossypium hirsutum L.), overexpressed cotton showed higher malate exudation and stimulation of P uptake from sparingly soluble forms 60 .It found that transgenic rice of maize PEPC showed a more significant increase of oxalate exudation and accumulated P than the wild type 61 .ATP-citrate lyase (ACLY) could catalyze the transformation among acetyl-CoA, oxaloacetate, and citrate, the key node of pyruvate metabolism going into the TCA cycle under low phosphorus.Here, ACLY showed a higher expression in the Fielder roots than Ardito and perhaps played a key role in promoting the TCA cycle.ACO catalyzes the transformation among isocitrate, cis-Aconitate, and citrate, also upregulated in Fielder vs. Ardito, suggesting that wheat roots mainly mobilize P absorption through the secretion of citric acid under LP stress in Fielder.However, the MDH2 was downregulated compared to Fielder with Ardito, perhaps due to energy balance.

Transcription factors responded to LP tolerance in wheat
The transcriptional regulation of PSR involves several transcription factors (TFs), including members of the MYB, WRKY, and bHLH families.These TFs positively or negatively regulate Pi signaling and Pi homeostasis in plants 11 .In our study, the identification of 426 transcription factors suggested complex regulation in wheat root response to LP stress (Table 3).Among those identified transcription factors, 82 DEGs belonging to the MYB transcription factor family were significantly expressed by LP stress.For instance, a MYB gene (TraesC-S7D02G295400) showed a significantly different expression in three comparisons (Table S9-1), indicating that MYB transcription factors are crucial in responding to low phosphorus stress.OsMYB4P-and OsMYB2P-1-overexpression upregulated the expression of the Pi transporter genes, leading to higher Pi accumulation in shoots and roots 62,63 .In addition, MYB genes have also been shown to respond to low phosphorus levels by modulating gibberellin (GA) metabolism and signaling and regulation of the miR399f promoter 41,64 .
Forty-four transcription factor genes belonging to the bHLH family were significantly expressed by LP stress.One bHLH gene (TraesCS5A02G049600) was downregulated in Fielder, Ardito, and Fielder vs. Ardito under low phosphorus stress, suggesting that wheat tends to downregulate bHLH in response to low phosphorus stress.It has also been found in other plants.For instance, AtbHLH32 is a negative regulator of several Pi-starvation responses in Arabidopsis.The bHLH TF OsPTF1 (Pi starvation-induced transcription factor 1) was reported to be involved in the PSR in rice.OsPTF1-overexpression enhanced tolerance to Pi starvation in transgenic rice by modulating root architecture rather than triggering increased expression of phosphate transporters 65 .
Additionally, 30 transcription factors belonging to the WRKY family were identified in this study.Most WRKY genes showed significant upregulation, especially in the comparison between Felder and Ardito, suggesting the WRKY may have a positive regulatory effect on low phosphorus tolerance in Fielder.WRKY45 and WRKY75 positively regulate the expression of PHT1;1.RNA interference (RNAi) lines of both WRKY45 and WRKY75 displayed impaired Pi uptake 66,67 .In rice, transgenic seedlings overexpressing OsWRKY74 display increased Pi uptake, longer roots, increased biomass, and higher iron accumulation levels 68 , WRKY21 and WRKY108 activate the expression of OsPHT1;1 under Pi-sufficient conditions to promote Pi accumulation 69 .The vast and complex regulatory mechanisms indicate that TFs play an important role in responding to low phosphorus stress, and further research is needed on these TFs.

Plant materials and treatment
Two wheat genotypes with different responses to low phosphorus (LP), including Fielder (P efficient) and Ardito (P inefficient), were used in this study.Wheat seeds were surface-sterilized with 0.5% NaClO (v/v) solution for 20 min, rinsed with sterile water, and allowed to germinate in the wet paper towel for 4-5 days at 4 °C before the seedlings were transplanted into continuously aerated 1/5 Hoagland hydroponic solution (pH 4.2) for a 2-day acclimation to low pH in a walk-in growth chamber with a stable temperature of 23 °C, a 14 h photoperiod of 400 μmol m −2 •s −1 illumination level, and relative humidity of 60-80%.For different Pi-level treatments, germinated seedlings were moved into the solutions with two concentrations of NaH 2 PO 4 : 200 μmol/L (normal Pi-level, NP) or 2 μmol/L (low Pi-level, LP) and grow for 4 days.The experimental samples are entirely random and arranged with at least three replicates.

Determination of phenotypic indicators
The plant height of seedlings was measured with a ruler, and Electronic vernier calipers were used to measure stem thickness.The relative chlorophyll content in fully developed leaves, as indicated by SPAD (Soil and Plant Analyzer Development) value, was determined with a portable Konica SPAD-502Plus instrument (Konica Minolta Holdings Inc., Tokyo, Japan).Fresh plant roots from different treatment groups were collected and completely expanded onto the scanner platform (Epson Expression 1000XL, Seiko Epson Corporation, Nagano, Japan).Analyze the scanned root photos using WinRhizo Pro software (Regent Instruments Inc., Quebec, Canada) to obtain the roots' total length, diameter, surface area, and volume.Afterward, dry the stem and root samples at 75 ° C for 4 days and further determine the dry mass of the stem and roots using an electronic balance.

P concentration determination
Oven-dried shoot and root samples were ground into fine powder for P concentration analysis.A 50-microgram sample was weighed out and digested with 9 mL nitric acid and 1 mL H 2 O 2 using a microwave digestion instrument, and phosphorus concentration was measured by ICP-AES (Juguang Co., Beijing, China).Total phosphorus in the shoot (TPL) was calculated based on shoot phosphorus concentration (LPC) and shoot weight.Total phosphorus in roots (TPR) was calculated based on root phosphorus concentration (RPC) and root weight 70 .

RNA extraction, library preparation, sequencing, and read mapping
Follow the instructions of the TRIzol kit (Invitrogen Co., Carlsbad, CA, USA) to extract total RNA from a 1 cm root tip.RNA quality and integrity were measured using a 5300 bioanalyzer (Agilent Co., Santa Clara, CA, USA) and quantified using ND-2000 (Thermo Fisher NanoDrop, Waltham, MA, USA).Only high-quality RNA samples (OD260/280 = 1.8-2.2,OD260/230 ≥ 2.0, > 1 μg, RIN ≥ 6.5, 28S: 18S ≥ 1.0) were used to construct sequencing library and verified by qRT-PCR.Entrust Shanghai Majorbio Biopharmaceutical Biotechnology Co., Ltd.(Shanghai, China) to perform RNA purification, reverse transcription, and library construction according to the manufacturer's instructions (Illumina, San Diego, CA, USA), followed by sequencing analysis.12 RNAseq transcriptome libraries (2 genotypes × 2 treatments × 3 biological replicates) were prepared and sequenced.The FASTP 71 soft was used to remove low-quality readings from the originally paired ends, and HISAT2 72 (Ver.2.2.1) was used to map clean readings onto the wheat genome for comparison.The StringTie 73 (Ver.1.3.6)was then used to assemble the mapped reads in a reference-based approach.All genes were annotated against public databases, including NR, Pfam, COG, Swiss-Prot, KEGG, and GO.

Differential expression and functional enrichment analysis
Express the level of each transcript using Fragments Per Kilobases Per Millionreads (FPKM).RSEM 74 was used to quantify gene abundances.Differential expression analysis was performed using the DESeq2 75 .DEGs with |log 2 FC| ≥ 1 and FDR ≤ 0.05 were considered significantly different expressed genes.Cluster heat diagrams were drawn using Toolkit for Biologists (TBtools Ver.2.030) with default settings 76 .Those data were added to the Comprehensive Gene Expression Database with the accession number PRJNA1033153.In addition, Goatools and KOBAS63 77 were used to analyze these DEGs' GO and KEGG functions.When the Bonferroni corrected, the p-value was ≤ 0.05, and GO terms, and KEGG metabolic pathways were significantly enriched compared to the entire transcriptome background.

Metabolome detection
The frozen wheat root apices were dispatched to Shanghai Majorbio Biopharmaceutical Biotechnology Co., Ltd.(Shanghai, China) for phytohormone extraction and analysis.Precision weighing was conducted on 100 mg root samples, which were then placed into 2 mL grinding tubes.Subsequently, 498 μL of 80% methanol and 2 μL of SA-D4 internal standard solution (2 μg/mL) were added to each tube.The mixture was then ground at low temperature for 3 minutes and subjected to ultrasound extraction at low temperature for 1 hour.To the extracted sample, 25 mg of EN15662 (consisting of magnesium sulfate 4 g, sodium chloride 1 g, sodium citrate 2 hydrate 1 g, disodium hydrogen citrate 1.5 hydrate 0.5 g) was added as a salt packet.The mixture was immediately shaken and vigorously vibrated on an oscillator for 10 minutes.Following this, it was centrifuged at 10 °C for 10 min, and 100 μL of supernatant was extracted.To this, 60 μL of water was added, and the solution was thoroughly swirled before transferring it into a sample vial.Qualitative and quantitative detection of the target substances in the samples was carried out using LC-ESI-MS/MS (UHPLC-Qtrap).The specific parameters included a WatersBEHC18 (2.1 × 100 mm, 1.7 μm) liquid chromatography column with a column temperature of 30 °C and a sample size of 10 μL.The mobile phase consisted of mobile phase A (0.1% formic acid water) and mobile phase B (0.1% formic acid acetonitrile), with a balance time of 3 min and a collection time of 10 min.In Sciex quantitative software OS, default parameters were utilized to automatically identify and integrate each ion fragment and assist in manual inspection.A standard linear regression curve was established with the mass peak area of the analyte as the ordinate and the analyte concentration as the abscissa.For sample concentration calculation, the mass spectrum peak area of the sample analyte was substituted into the linear equation to determine the concentration result.

FielderFigure 1 .
Figure 1.Total P concentrations were shown in the shoot and root under different levels of Pi-supply in two wheat genotypes.DW, dry weight.Values are means ± standard errors (n = 3).Different lowercase letters indicate significant differences (p < 0.05) within each index.

Figure 2 .
Figure 2. Correlation analysis of differentially expressed genes (DEGs) expression between transcriptome data and qRT-PCR results.

Figure 4 .
Figure 4. Top 20 GO annotations terms of DEGs under LP stress in wheat.MF molecular function, CC cellular component, BP biological process.

Figure 5 .
Figure 5. Top 20 enriched KEGG pathways of DEGs under LP stress in wheat.The bubble size indicates the number of DEGs involved in the pathways.NP normal phosphorus, LP low phosphorus.

Table 1 .
Effects of wheat on phenotypic changes under low phosphorus stress.Values are means ± standard errors (n = 3).Different lowercase letters indicate significant differences (p < 0.05) within each index.

Wheat root transcriptome profiling in response to LP stress
Transcriptome sequencing analysis was applied to investigate the molecular responses of wheat roots to Pi starvation.Twelve cDNA libraries were constructed, combining two genotypes, two treatments, and three replicates.These libraries yielded an average of 51.88 M clean reads (ranging from 44.29 M to 61.88 M).Additionally, a total of 113.40 G clean bases were generated, with respective lowest, highest, and average clean base values of 6.47 G, 8.85 G, and 7.56 G. Quality control data exhibited a Phred quality score Q30 ranging from 92 to 93.22%, coupled with an average GC content of 54.32%.Importantly, all samples demonstrated total mapping and uniquely mapping rates of 91.38% and 84.12%, respectively (average values), affirming the high quality and suitability of the sequencing data for subsequent gene expression analysis (TableS1).A comprehensive assembly yielded 118095 expressed genes (91852 known and 26243 new) and 192506 transcripts (113961 known and 78545 new).Subsequently, 91313 genes and 113446 transcripts were annotated by referencing KEGG, Swiss-Prot, Pfam, GO, COG, and NR databases.Principal component analysis (PCA) distinguished the 12 samples into two groups, Fielder and Ardito, effectively capturing differences between NP and LP treatments.PC1 and PC2 explained 18.34% and 39.70% variations, respectively (Fig.

Table 2 .
Effects of wheat on root morphology under low phosphorus stress.Values are means ± standard errors (n = 3).Different lowercase letters indicate significant differences (p < 0.05) within each index.

Fielder_LP vs. Fielder_NP Ardito_LP vs. Ardito_NP Fielder_LP vs. Ardito_LP 0 Figure 3. Differentially expressed genes profile
DEGs involved in purple acid phosphatase and phosphate transporter were identified in the RNA-seq analysis that showed changes in expression under LP stress in wheat.The numbers in the black rectangles show Log 2 (fold-changes) of the comparisons.The comparisons show that each gene in the picture is significantly differentially expressed.PAP purple acid phosphatase, PHT, PHO phosphate transporter, SPX SPX-domaincontaining protein.

Table 3 .
Differentially expressed transcriptional factors (TFs) under LP stress in wheat.

Table 4 .
Concentrations (mean ± SE, n = 3) of 34 hormones detected under LP and NP treatment in two wheat roots.Different letters in the boxes indicate a significant difference at p < 0.05.ND not detected.