Introduction

According to the Food and Agriculture Organization of the United Nations, climate change generates considerable uncertainty about future water availability in many regions. Increased water scarcity under climate change will present a major challenge for climate adaptation, while sea-level rise will affect the salinity of surface and groundwater in coastal areas. Stress caused by climate change has led to increased agricultural losses and threatened global food security1. Drought is considered the most damaging environmental stress, which directly affects the entire growth period of plant seeds from germination to final fruiting2. Drought stress can lead to increased plant osmotic regulators, the inhibition of photosynthesis, and the change of plant endogenous hormone content3,4,5. Drought stress also induces reactive oxygen species, such as superoxide radicals, hydrogen peroxide, and hydroxyl radicals, leading to oxidative stress6. Crop loss due to soil salinisation is another increasing threat to agriculture worldwide, which is more severe in agricultural land in coastal and arid regions7. Irrigation with saline water, low precipitation, and high evapotranspiration are key factors of the rapid salinisation of agricultural land8. These abiotic stresses of drought and salinity have brought unprecedented challenges to the development of crop farming. Compared to the heavy use of chemical fertilisers, the use of plant growth-promoting bacteria to improve plant growth under drought and salinity environments is more sustainable and gaining more attention9,10.

Soil microbial communities are critical to plant health and their resistance to both biotic and abiotic stressors, such as pathogens, drought, salinity, and heavy metal pollution11. A few studies have demonstrated that many beneficial soil bacteria harbour plant growth-promoting activities, e.g. by helping plants with disease suppression12, nutrient acquisition13, phosphorus uptake14 and nitrogen fixations15. Beneficial root microbiota also regulate biosynthetic pathways in the plant itself, leading to differential alterations in the plant metabolome in response to stresses16. Streptomyces are Gram-positive filamentous bacteria, widely distributed in soil and marine environments. While they have long been considered the richest source of bioactive secondary metabolites17, Streptomyces have recently drawn attention as a class of plant growth-promoting bacteria that help plants respond to adversity stress18. The growing evidence showed that Streptomyces can promote plant growth or tolerance to stressors in direct or indirect ways, by secreting plant growth regulator auxin (indole-3-acetic acid, IAA) and siderophores, inducing systemic resistance in plants, and regulating the rhizosphere microbiome via producing antibacterial compounds or signalling molecules19,20,21. Notably, the commercial product Actinovate® and Mycostop® are two Streptomyces-based formulations that have been widely used to suppress a wide range of diseases in a variety of crop groups as a biological fungicide/bactericide for the long term. Recently, the enrichment of Streptomyces has also been shown to play a subsequent role in the drought/salt tolerance of plants22. Despite the widespread claims of efficacy of inoculation of plant growth-promoting Streptomyces, the molecular basis of the growth-promoting effects and the key role of secondary/specialised metabolites in this process are largely unknown.

Here, we report that Streptomyces iranensis HM 35 has profound beneficial effects on helping barley alleviate osmotic, drought and salinity stress. The active components were identified as bioactive spiroketal polyketides pteridic acids H (1) and its isomer F (2) through large-scale fermentation and bioactivity-directed purification followed by NMR, MS, and X-ray crystallography. The abiotic stress mitigating effects of pteridic acids H and F have been confirmed on the model plant Arabidopsis thaliana, where it effectively reversed both drought and salinity stress as phytohormone-like small biomolecules at concentrations as low as 0.5 ng mL−1 (1.3 nM). RNA sequencing results suggested that pteridic acids may assist plants in stress resistance via activating photosynthesis and regulating multiple stress response genes. Moreover, the Biosynthetic Gene Cluster (BGC) of pteridic acids (pta) was identified and analysed in silico, and functionally confirmed by in vivo CRISPR-based genome editing. We have furthermore conducted a survey of 81 potential producers of pteridic acids, which are widely distributed around the world. Phylogenetic and comparative genomic analysis of pta-containing streptomycetes suggested that these strains have evolutionary convergence in disseminating pta BGC through main vertical transmission and occasional horizontal gene transfer. In summary, we reveal a strategy of Streptomyces to secret plant growth regulators that help plants cope with abiotic stresses, which is a promising alternative solution for plant development and crop yields under the current climate change-induced environmental stresses.

Results

Abiotic stress-mitigating activities exhibited by S. iranensis

The promotion of barley growth induced by S. iranensis was tested under multiple abiotic stresses including osmotic, salinity and drought. The osmotic stress experiment was simulated using soils supplemented with 20% (w/v) PEG-6000 by transiently reducing the water potential of the plant. We found that S. iranensis played a significant role in alleviating osmotic stress in barley seedlings. The treated seedlings showed a significant increase in height, fresh weight, and dry weight compared to the control group without any extra treatment (Fig. 1a). The culture broth of S. iranensis also showed considerable activity for the growth of barley in alleviating salinity stress mediated by 100 mM NaCl (Fig. 1b), while S. iranensis was not significantly enriched in the soil around the roots of barley seedlings (Supplementary Fig. 1). Additionally, based on the analysis of barley seedling phenotypes, the treatment with S. iranensis resulted in a significant improvement in plant growth recovery from drought stress (Fig. 1c). Surprisingly, S. iranensis promoted the growth of barley seedlings even in non-stress growth condition and thus indicate that S. iranensis may have potential for use as biostimulant (Supplementary Fig. 2).

Fig. 1: The effects of S. iranensis on barley seedlings under abiotic stress and the bioactive components produced by S. iranensis.
figure 1

a The box plots depict the plant height, fresh weight and dry weight of barley seedlings growing under osmotic stress mediated by 20% (w/v) PEG-6000 (mean ± SD, n = 18). b the box plots depict the plant height, fresh weight and dry weight of barley seedlings growing under salinity stress mediated by 100 mM NaCl (mean ± SD, n = 18). In a and b, statistical significance was assessed by one-way ANOVA with post hoc Dunnett’s multiple comparisons test. Asterisks indicate the level of statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001. Abbreviation: Mock: control; Si: treatment of S. iranensis culture broth; D: treatment of S. iranensis/ΔptaA culture broth; M: treatment of blank medium (ISP2). c The box plots depict the plant height, fresh weight and dry weight of barley seedlings growing under drought stress (mean ± SD, n = 18). Different colours of box plots indicate different growing conditions: green, 21 days water; brown, 7 days treatment after 7 days water followed by 7 days drought; red, 7 days water followed by 14 days drought. Statistical significance was assessed by one-way ANOVA with Tukey test. Asterisks indicate the level of statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001. Abbreviation: Water: well water for 21 days; CK, 7 days treatment of water after 7 days water + 7 days drought; Si: 7 days treatment of S. iranensis culture broth after 7 days water + 7 days drought; D: 7 days treatment of S. iranensis/ΔptaA culture broth after 7 days water + 7 days drought; M: treatment of blank medium (ISP2); Drought: 14 days drought after 7 days water. Statistical significance was assessed by one-way ANOVA with Tukey test. d The metabolite profile of the native S. iranensis growing in liquid ISP2 medium, the known secondary metabolites were identified by HR-LC-MS and highlighted; e the bioactive components pteridic acid H (1) and pteridic acid F (2) isolated from S. iranensis. All box plots with centre lines showing the medians, boxes indicating the interquartile range, and whiskers indicating a range of minimum to maximum data beyond the box. Source data are provided as a Source Data file.

Full size image

Genomic and metabolomic profiles of S. iranensis

To reveal the potential bioactive components, we first annotated BGCs responsible for the biosynthesis of secondary metabolites in S. iranensis using antiSMASH 6.023. Genome sequence analysis of S. iranensis revealed the presence of 47 putative secondary metabolites BGCs with a variety of biosynthetic categories (Supplementary Table 1). The clusters 3, 6, 7, 8, 23, 31, 35 and 40 were annotated to have greater than 80% similarity with BGCs responsible for the biosynthesis of coelichelin, azalomycin, nigericin, elaiophylin, desferrioxamin B, ectoine, rapamycin and hygrocin, respectively. Their corresponding products were also detected and identified through High-resolution Liquid Chromatography-tandem Mass Spectrometry (HR-LC-MS/MS) as well as Global Natural Products Social (GNPS) molecular networking (Fig. 1d, Supplementary Fig. 3)24. However, a large number of metabolites from S. iranensis are still unknown. Since none of the previously identified compounds have been associated with mitigating abiotic stress in plants, we were prompted to expand the fermentation process and identify the potential bioactive compounds.

Characterisation of the bioactive compound pteridic acid

To uncover the bioactive components, fermentation of S. iranensis was scaled up to 175 litres and the culture broth was subjected to separation through open-column chromatography on Amberchrom CG161Me resin, silica gel, and Sephadex LH-20. Bioactivity-guided fractionation led to the isolation of bioactive compound 1 (15.0 mg) together with its isomer compound 2 (4.0 mg) (Fig. 1e).

The bioactive component compound 1 was isolated as a white solid. Its formula of C21H34O6 was deduced by m/z 383.2439 [M + H]+ (calculated for 383.2428, Δ 2.84 ppm). The 1H NMR spectrum exhibited signals for four olefinic protons (δ 7.33, 6.26, 6.13, 5.90) corresponding to two conjugated double bonds, four oxygen-bearing methines (δ 3.85, 3.69, 3.59, 3.43), five methyls, and other aliphatic protons. The 13C NMR spectrum indicated the presence of one carbonyl group (δ 168.7) and one oxygen-bearing quaternary carbon (δ 103.2, C-11). The COSY spectrum established two partial substructures, which could be connected via a spiral function by analysing HSQC and HMBC correlations (Supplementary Figs. 4–11). Compound 1 was crystallised in methanol solution, and the structure was determined via X-ray crystallography (Supplementary Fig. 12). Therefore, compound 1 was identified as a new Streptomyces-derived natural product named pteridic acid H.

Compound 2, a white solid, is an isomer of 1 deduced by MS with the same molecular formula of C21H34O6. The 1H NMR spectrum exhibited signals for four olefinic protons (δ 7.16, 6.25, 6.07, 5.97) corresponding to two conjugated double bonds, four oxygen-bearing methines (δ 3.88, 3.66, 3.56, 3.32), five methyls, and other aliphatic protons. The 13C NMR spectrum indicated the presence of one carbonyl group (δ 170.2) and one oxygen-bearing quaternary carbon (δ 103.2, C-11). HSQC and HMBC correlations confirmed a spiroketal skeleton. NOESY spectrum confirmed its relative configurations, where correlations between H-21 and H-13 and H-15, H-7 and H-12a, and H6, H-10, and Me-18 were observed. The key NOESY correlations between H-7 and H-12a revealed a different spiroketal structure than 1 (Supplementary Fig. 11, 13-20). This can be reflected by the relative up-field NMR data for C-12 (δ 33.8 vs δ 37.4 in 1). Compound 2 was identified as pteridic acid F, previously isolated from Streptomyces pseudoverticillus YN17707 and a marine-derived Streptomyces sp. SCSGAA 002725,26.

Abiotic stress mitigation of pteridic acids in planta

Initially, we tested the effects of different concentrations of pteridic acids H and F on Arabidopsis growth in the absence of abiotic stress. A concentration of 0.5 ng ml−1 of both pteridic acids H and F was found to significantly promote the growth of Arabidopsis seedlings (Supplementary Fig. 21). Under drought stress, pteridic acid H at a concentration of 0.5 ng mL−1 increased the root length and fresh weight of Arabidopsis seedlings by 54.5% and 89%, respectively, and its activity was significantly better than IAA and ABA at the same molar concentration (Fig. 2a, c). The treatment of pteridic acid F also showed great activity in alleviating drought stress, and the root length and fresh weight were increased by 30.5% and 56.7%, respectively (Fig. 2c). Pteridic acids H and F also showed significant activity in alleviating NaCl-mediated salinity stress (Fig. 2b, d). Compared to the non-treated groups, the treatment of 0.5 ng mL−1 pteridic acids H and F increased root length of Arabidopsis seedlings by 74.0% and 61.8%, as well as fresh weight by 126.2% and 110.9%, respectively (Fig. 2d).

Fig. 2: The effect of pteridic acids H and F on Arabidopsis seedlings under abiotic stress.
figure 2

a Phenotype of Arabidopsis seedlings growing under drought stress mediated by 15% (w/v) PEG-6000 using other treatments (bars = 1 cm); b Phenotype of Arabidopsis seedlings growing under salinity stress mediated by 80 mM NaCl using other treatments (bars = 1 cm); c The box plots depict the primary root length and fresh weight of Arabidopsis seedlings growing on non-stress condition (mean ± SD, n = 16); d The box plots depict the primary root length and fresh weight of Arabidopsis seedlings growing on drought stress condition (mean ± SD, n = 16). In c and d, statistical significance was assessed by one-way ANOVA with Tukey test. Asterisks indicate the level of statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001; e differences of lateral root growth of Arabidopsis seedlings growing in other conditions. Mock: control; PH: treatment of 0.5 ng mL−1 pteridic acid H; PF: treatment of 0.5 ng mL−1 pteridic acid F; IAA: treatment of 1.3 nM indole-3-acetic acid; ABA: treatment of 1.3 nM abscisic acid. All box plots with centre lines showing the medians, boxes indicating the interquartile range, and whiskers indicating a range of minimum to maximum data beyond the box. Source data are provided as a Source Data file.

Full size image

To get a first understanding how pteridic acids help Arabidopsis mitigate salinity stress, we used messenger RNA sequencing (mRNA-seq) to profile the transcripts of Arabidopsis seedlings treated with pteridic acid F or H under NaCl-mediated salt stress (Fig. 3a). Compared to the control (CK, treatment with an equal amount of water under NaCl-mediated salt stress), we observed significant differences in the gene expression patterns upon treatment with pteridic acid (PH or PF) by Pearson correlation analysis (Supplementary Fig. 22). The Differentially Expressed Genes (DEGs) analysis revealed 3575 DEGs (1405 upregulated and 2170 downregulated), and 3727 DEGs (1555 upregulated and 2172 downregulated) in the pteridic acid H treatment versus control (PH vs. CK) and pteridic acid F treatment versus control (PF vs. CK), respectively (Supplementary Data 1). Meanwhile, 1226 up-regulated genes and 1860 down-regulated genes were shared between PH and PF treatments. Several abiotic stress-related genes were significantly upregulated after pteridic acids treatments (Fig. 3b), including SLAC1 HOMOLOGUE 1 (SLAH1, AT1G62280)27, PIN-FORMED 6 (PIN6, AT1G77110)28, FERRIC REDUCTION OXIDASE 6 (FRO6, AT5G49730)29 and TOO MANY MOUTHS (TMM, AT1G80080)30. Compared with the control, ALPHA CARBONIC ANHYDRASE 8 (ATACA8, AT5G56330)31, and IAA-LEUCINE RESISTANT 2 (ILR2, AT3G18485)32 were uniquely upregulated in the PH treatment samples, while ATP-BINDING CASSETTE G17 (ABCG17, AT3G55100)33, DIACYLGLYCEROL KINASE 4 (DGK4, AT5G57690)34 and MAP KINASE KINASE 7 (MKK7, AT1G18350)35 were upregulated in the PF treatment samples.

Fig. 3: Effect of exogenous pteridic acids on transcription in Arabidopsis seedlings under salt stress.
figure 3

a The phenotype of Arabidopsis seedling samples in different treatments under salinity stress. CK: control; PF: pteridic acid F treatment; PH: pteridic acid H treatment. b The volcano plots of DEGs identified by mRNA-seq in Arabidopsis seedlings treated by pteridic acid H and F under salinity stress. c heat-maps of DEGs enriched in photosynthesis, response to stimulus, auxin-activated signalling pathway, abiotic stress defense and TFs.

Full size image

Next, we performed the Gene Ontology (GO) enrichment analysis of DEGs focusing primarily on the biological processes (Supplementary Data 2). Among all upregulated genes, several highly enriched GO terms such as “Photosynthesis (GO:0015979)”, “Plastid organization (GO:0009657)”, “Chloroplast organization (GO:0009658)” and “Response to light stimulus (GO:0009416)” suggested pteridic acids could enhance photosynthesis under abiotic stress. For these DEGs with log2FoldChange ≥ 2, we observed upregulation of 12 genes belonging to the “auxin-activated signalling pathway (GO:0009734)” (Fig. 3c), suggesting that pteridic acids also trigger auxin-activated signalling transduction. Intriguingly, we observed the downregulations of genes related to “response to stimulus (GO:0050896)”, “response to chemical (GO:0042221)” and “response to stress (GO:0006950)”, which was speculated to be an interference between different signalling pathway or a negative feedback mechanism of plants to maintain homoeostasis (Fig. 3c)36. Moreover, similar conclusions were reached by the most enriched Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway, pointing to the “Photosynthesis (ath00195)” and “Plant hormone signal transduction (ath04075)” pathways in the Arabidopsis seedlings. In addition, the pathway of “Motor proteins (ath04814)”, “Ribosome (ath03010)”, “Glucosinolate biosynthesis (ath00966)”, “Porphyrin metabolism (ath00860)” and “Flavonoid biosynthesis (ath00941)” were also activated, which had been previously reported that these are closely associated with plant abiotic stress resistance37,38. On the contrary, pteridic acids downregulated several genes enriched in “Phenylpropanoid biosynthesis (ath00940)”, “Cyanoamino acid metabolism (ath00460)”, “Glutathione metabolism (ath00480)”, “MAPK signalling pathway-plant (ath04016)” and “Starch and sucrose metabolism (ath00500)” pathways.

Transcription Factors (TFs) play an important function in coping with abiotic stress tolerance. Several TFs have been proved to participate in plant salt stress responses, such as AP2/ERF, bZIP and MYB39,40,41. Within these DEGs, we identified various TFs, comprising AP2/ERF (42 unigenes), MYB (35 unigenes), WRKY (27 unigenes), bHLH (31 unigenes) and bZIP families (6 unigenes) (Supplementary Data 3). A handful of plant stress resistance-related TFs were observed, such as MYB25, MYB111, WRKY46, PRE1 and SCRM242,43,44,45,46. TFs associated with resistance to abiotic stress, such as MYB60 and ERF120, exhibited discernible upregulation in the PH treatment samples compared to the control47,48. Conversely, MYB29, MYB113 and PRE6 showed unique upregulation in the PF vs. CK group49,50.

A previous study suggested that pteridic acids A and B might have a plant growth-promoting effect like IAA and could stimulate the formation of adventitious roots in kidney beans25. However, we observed that pteridic acids H and F displayed different IAA-induced phenotypes. Pteridic acids did not exhibit the function to significantly promote lateral root growth of Arabidopsis seedlings like IAA (Fig. 2e)51. They were also not capable of promoting the formation of adventitious roots in kidney beans, as shown in Supplementary Fig. 23. Except for drought and salinity stress, we also tested the CuSO4-mediated heavy metal stress alleviation activity of pteridic acids on mung beans. The results showed that the 1 ng mL−1 pteridic acid H was as effective as ABA in helping mung beans to relieve heavy metal stress (Supplementary Fig. 24). In conclusion, pteridic acids H and F are widely applicable potent plant growth regulators produced by Streptomyces to assist plants in coping with different abiotic stress.

Biosynthesis of pteridic acids

The retro-biosynthesis analysis indicated that pteridic acids could derive from a modular type I polyketide synthase. A putative pta BGC was identified in the whole genome sequence of S. iranensis, which shows 87% antiSMASH similarity to the BGC of elaiophylin (BGC0000053 in MiBiG database)52. The pta BGC spans approximately 56 kb and encodes 20 individual biosynthetic genes responsible for the biosynthesis of core polyketide backbones, precursor, glycosylated substituents, transporters and regulators (Supplementary Table 2). The five consecutive Type I polyketide synthase (PKS) encoding genes within the pta BGC consist of one loading module and seven extender modules, which are sequentially extended to form a linear polyketide chain by ketosynthase (KS) domain, acyltransferase (AT) domain, acyl carrier protein (ACP), with additional ketoreductase (KR), dehydratase (DH), and enoyl reductase (ER) domains. The substrate specificity predictions for individual AT domains fit well with the structure of pteridic acids (Supplementary Table 3). The last “Asn” residue is absent in the conserved Lys-Ser-Tyr-Asn tetrad of the KR domain in module 3 (PtaB), which is predicted to be inactive (Supplementary Fig. 25). This is consistent with the nonreduced carbonyl group on the α-carbon in module 3. The first DH domain in module 1 is inactive since it does not have a conserved active motif LxxHxxGxxxxP (Supplementary Fig. 26). Following the thioesterase-mediated release of the polyketide chain, the 6,6-spiroketal core structure is likely formed by spontaneous spiroketalisation of the carbonyl group on C11 and the two hydroxyl groups on C17 and C25. Following a loss of H2O, two differentially oriented spirocyclic rings were formed to yield pteridic acids F and H (Fig. 4a). Remarkably, pteridic acid H showed molecular instability under extreme conditions. In the water solution with high temperature (65°C) or acidity (pH = 3), pteridic acid H is transformed into pteridic acid F (Supplementary Fig. 27). A similar spontaneous transformation from (S) to (R) chirality at the centre of the spiroketal ring was also observed in 6,6-spiroketal avermectin53.

Fig. 4: Biosynthesis mechanism of pteridic acids and CRISPR base editing application in S. iranensis.
figure 4

a the proposed biosynthetic pathway of pteridic acids H (1) and F (2); b the proposed biosynthesis mechanism of elaiophylin (3), and its macrodiolide formation is catalysed by thioesterase (TE) domain. c Sanger sequencing and HR-LC-MS output of CRISPR base editing application of STOP codon introduction targeting the ptaA of S. iranensis. The 20-nt protospacer sequence is highlighted in light green, whereas the 3-nt PAM sequence is shown in yellow. The codons and corresponding amino acids are indicated, and the black double-headed arrow represents the position of the editing window; Extracted Ion Chromatography (EIC) for 1 and 2 (m/z 383.2428 [M + H]+) and 3 (m/z 1047.5863 [M+Na]+) in the wild type S. iranensis (Control) and the mutant S. iranensis/ΔptaA; d Sanger sequencing and HR-LC-MS output of CRISPR base editing application of site-directed mutagenesis targeting the TE domain of pta BGC. EIC for 1 and 2 (m/z 383.2428 [M + H]+) and 3 (m/z 1047.5863 [M+Na]+) in the wild type S. iranensis (Control) and the mutant S. iranensis/M2089I + E2090K + D2091N.

Full size image

CRISPR base editing in S. iranensis

To validate the in silico prediction, we utilised the efficient base editing tool CRISPR-cBEST to experimentally confirm the pta BGC54. As a non-model Streptomyces strain, S. iranensis is hard to genetically manipulate through intergeneric conjugation55. Therefore, the conjugation process was systematically optimised in this study (Supplementary Fig. 28). The core polyketide synthase ptaA was targeted and inactivated by converting a TGG (Trp) codon at position 916 into the stop codon TAA using CRISPR base editing. The editing event was confirmed by PCR amplification and Sanger sequencing of the editing site (Fig. 4c). As expected, the production of both pteridic acids and elaiophylin was abolished in S. iranensis/ΔptaA (Fig. 4c). Plant experiments showed that the treatment of S. iranensis/ΔptaA fermentation suspension led to the abolishment of the abiotic stress mitigating effects (Fig. 1a–c). To further confirm the pta gene cluster, a bacterial artificial chromosome (BAC) library of S. iranensis was constructed. BAC-based cross-complementation of ptaA in S. iranensis/ΔptaA restored the production of pteridic acids and elaiophylin (Supplementary Fig. 29).

Interestingly, based on isotope-labelled precursor feeding and partial cosmid sequencing-based bioinformatics prediction, this BGC has long been inferred to be responsible for the biosynthesis of the antibacterial elaiophylin56,57. In 2015, Zhou et al. reported that the thioesterase in the last module catalysed the formation of symmetrical macrodiolide using two units of linear elaiophylin monomeric seco acid (Fig. 4b)58. To confirm whether the biosynthesis of pteridic acids is also thioesterase-dependent, site-specific mutations of residues Met-Glu-Asp to Ile-Lys-Asn were introduced into the active sites of the TE domain in vivo (Supplementary Fig. 30)59. HR-LC-MS analysis showed that the mutant strain (M2089I + E2090K + D2091N) no longer produced pteridic acids and elaiophylin (Fig. 4d). Hence, we provide additional evidence via in vivo inactivation and site-directed mutagenesis. Co-production of the plant growth-regulating pteridic acids and the antimicrobial elaiophylin through a shared BGC is intriguing and points to possible joint efforts in helping plants cope with both biotic and abiotic stress.

Geographical distribution of pteridic acid producers

We surveyed available gene cluster family (GCF) data for all bacteria in the BiG-FAM database60. We found that the pta BGC (GCF_02696) is strictly restricted to the Streptomyces genus. In addition, a total of 55 BGCs with high similarity to the pta BGC were detected by BiG-SCAPE61, among a total of 9386 type I polyketides BGCs in 1965 Streptomyces from the NCBI assembly database. Through literature supplementation and data dereplication of other reported producers without sequence information, at least 81 Streptomyces are known to produce pteridic acids/elaiophylin or have specific pta BGC up to date (Supplementary Table 4). Based on the known sampling information, the pta-containing Streptomyces display a variety of geographic distribution and biological origins (Fig. 5a). We selected two available Streptomyces strains (Streptomyces violaceusniger Tu 4113 and Streptomyces rapamycinicus NRRL 5491) to test the potential plant growth-promoting activity of these potential pteridic acid producers. The HR-LC-MS analysis of both culture broths revealed that they shared similar metabolite profiles, and both produced pteridic acids H and F (Supplementary Fig. 31). Treatment with both culture broths on barley seedlings also exhibited significant plant growth-promoting activities under osmotic, salinity and drought stress (Supplementary Fig. 32). This evidence suggests that this class of Streptomyces and its specific secondary metabolite pteridic acids have unique ecological significance involved in plant abiotic stress resistance.

Fig. 5: Geographical distribution and phylogenetic analysis of pta-containing streptomycetes possessing pta BGC.
figure 5

a A total of 61 streptomycetes were displayed on the map and distinguished by different colours. For detailed strain information, see (Supplementary Table 4). This figure was partly generated using Servier Medical Art (http://smart.servier.com/), licensed under a Creative Common Attribution 3.0 Generic License; b the phylogenetic tree of 16S rRNA nucleotides sequences of pta-containing streptomycetes and other streptomycetes; c the heatmap depicts similarity differences of core biosynthetic genes of pta BGC between S. iranensis and other 14 pta-containing streptomycetes.

Full size image

Phylogeny and evolution of pta BGC

To explore the evolutionary clues of pteridic acid producers, 16S rRNA genes were initially used to assess the relatedness of the collected 34 potential producers of pteridic acids with other streptomycetes that do not contain pta BGC (Fig. 5b). The results revealed that, except for Streptomyces albus DSM 41398 and Streptomyces sp. GMR22, and other pta-containing streptomycetes cluster together and are distinct with divergent lineages. To further confirm this hypothesis, two high-resolution Streptomyces housekeeping genes, tryptophan synthase subunit beta (trpB) and RNA polymerase subunit beta (rpoB) were employed to analyse the phylogeny relationship among these strains (Supplementary Fig. 33)62. Consequently, only S. albus DSM 41398 was classified in a distinct phylogenetic lineage among the Streptomyces strains containing pta BGC. The strict congruence among the clades of the housekeeping genes indicated dominant vertical transmission and potential horizontal gene transfer of the pta BGC in Streptomyces.

A total of 15 pta-containing streptomycetes with complete genome sequence information were selected to conduct the comparative genomics investigation. The genetic diversity in these streptomycetes was initially revealed using genome sequence similarity analysis. Except for S. albus DSM 41398 and Streptomyces sp. NA02950, we observed a high degree of similarity in the aligned region, as indicated by both the average nucleotide identity (ANI) and the alignment percentage (AP) among these strains (Supplementary Fig. 34). Genome synteny analysis revealed that partial genome rearrangements happened among strains even with high sequence similarities (Supplementary Fig. 35). Notably, the pta BGC in S. albus DSM 41398 (pta-alb) is located at the end of the chromosome, a high variable region in Streptomyces, suggesting its existence by accepting heterologous biosynthetic gene fragments. The nucleotide sequence alignment of pta BGC results showed that pta-alb is relatively complete, and the similarity of core genes is proportional to evolutionary relatedness (Fig. 5c). The metabolite profile of S. albus DSM 41398 also confirmed the integrity of the pta-alb by detecting the production of elaiophylin and pteridic acid H (Supplementary Fig. 36). To further assess the biosynthesis diversity in remaining genetically related strains, we performed the similarity analysis of these BGCs (Supplementary Figs. 37 and 38). The connections between their secondary metabolites, BGCs, revealed that these vertically inherited Streptomyces strains also harbour striking similarities. Combining phylogenetic and comparative genomics analysis, we expect that S. albus DSM 41398 is evolutionarily the most distinct member from other pta-containing streptomycetes and obtained the pta BGC via horizontal gene transfer. However, most pta-containing streptomycetes have vertically inherited pta and other BGCs from their ancient ancestors that may be ecologically important and rarely studied.

Discussion

Drought and salinisation of soil are increasing globally, driving a reduction in crop yields that threatens food security. Plant growth-promoting bacteria is a class of beneficial microorganisms that positively interact with the plant to confer environmental stresses63. Although some Streptomyces species have been reported to have plant growth-promoting activity, the molecules mediating such positive effects are poorly understood. Deciphering the molecular mechanism is key to understanding the complex plant microbiota interaction. In this study, we present an example of S. iranensis secreting a family of secondary metabolites, pteridic acids, to assist plants to cope with abiotic stresses like osmotic, salinity and drought. Pteridic acids H and F were chemically isolated, structurally characterised and functional validated as plant-beneficial molecules.

Plants respond to harsh environments by changing their physiological processes for better survival64. Salt and drought stress signal transduction consists of ionic and osmotic homoeostasis signalling pathways, detoxification (i.e., damage control and repair) response pathways, and pathways for growth regulation65,66. Based on mRNA-seq analysis, we identified a total of 3086 DEGs, some of which are associated with diverse abiotic stress defenses in plants. For example, we observed that the upregulation of SLAH1 is crucial for alleviating the toxicity of salt by root-to-shoot Cl− transport in Arabidopsis27. PIN6 was previously identified as the key salt tolerance-related gene in the roots of the mangrove Avicennia officinalis28. AtACA8 is a plasma membrane localised Ca2+ pump and plays a role in sucrose signalling, ion homeostasis and root development during early seedling germination31. MKK7 positively regulates plant salt tolerance and promotes primary root growth in Arabidopsis seedlings35. The differentiation and development of Guard cells (GCs), which are in the epidermis of leaves and stems that regulate stomatal development, are regulated by TMM30. Under abiotic stress conditions, the upregulation of ABCG17 expression facilitated the translocation of ABA from the shoot to the root, consequently stimulating lateral root growth33. Additionally, our study revealed an enhancement in the lateral root growth of Arabidopsis seedlings upon exposure to pteridic acid F, with the expression of the associated ABCG17 gene observed exclusively in the PF vs. CK group.

Salt stress negatively affects photosynthesis in plants, prompting them to regulate the photosynthetic process, either through intrinsic mechanisms or in response to external stimuli, to enhance their salt tolerance67. Based on GO and KEGG pathway enrichment analysis, photosynthesis and its related physiological events of Arabidopsis seedlings treated with pteridic acids were immensely upregulated. Interestingly, previous studies also showed that small molecules, such as exogenously applied melatonin, can assist plants in resisting salt stress by improving photosynthesis in different plants68,69. Some studies have confirmed that IAA is involved in response to salt stress in plants and a link between IAA signalling and salt stress has been established70,71. Herein, we speculated that pteridic acids might also promote plant growth under salt stress via activating auxin signal transduction to promote plant growth under salt stress. B-ARR, as a positive regulator that regulated downstream activity in the cytokinin signalling pathway, was also uniquely upregulated in PF vs. CK group.

Among the various TFs, MYBs participate in various biological processes in plants such as growth, reproduction, secondary metabolism and stress responses72. For example, upregulated MYB25 has been shown to reduce sensitivities toward osmotic and salt stress in Arabidopsis and upregulated MYB111 is a positive regulator of salt stress in Arabidopsis by binding directly to the cis-acting element in the promoter region of genes encoding flavonoid synthesis enzymes42,43. MYB60, which was upregulated in the PH vs. CK group, was previously demonstrated to play a dual role in abiotic stress responses in Arabidopsis through its involvement in stomatal regulation and root growth for increased water uptake47. MYB29, which was uniquely upregulated in the PF vs. CK group, has been demonstrated to be an important factor in promoting Arabidopsis lateral root growth under salinity stress49. MYB113 has also been reported to promote anthocyanin biosynthesis in Arabidopsis and pear for defense against abiotic and biotic stresses50. Previous studies have shown that overexpression of WRKY46 enhanced root development during salt stress in Arabidopsis through modulation of ABA signalling44. We found that WRKY46 was upregulated in pteridic acid-treated samples, which may also serve as a positive regulator in the ABA signalling pathway to confer abiotic stress resistance to plants, although we didn’t observe other ABA-related DEGs. The two upregulated bHLH family TFs PRE1 and SCRM2 have also been experimentally proved to promote plant growth and resist abiotic stress45,46.

Horizontal gene transfer is an integral driver of BGC evolution, revealing the independent processes of species phylogeny and BGCs distribution73. However, vertical inheritance also influences BGCs evolutionary dynamics, evident from BGCs conservation among closely related strains74. We found that the pta BGC in Streptomyces are widely dispersed geographically and mainly inherited through vertical gene transmission. Some of these strains have also been described to have remarkable biocontrol capabilities. For example, Streptomyces sp. AgN23 activates Arabidopsis defence responses to fungal pathogen infection by secreting plant elicitors75, Streptomyces rhizosphaericus 0250 and Streptomyces sp. 5–10 displayed significant biocontrol potential to fusarium wilt of bitter gourd76,77. The family of pta-containing Streptomyces was also previously described as a specific phylogenetic lineage with the highest BGC abundance and largest genome size across diverse streptomycetes78. Although the biosynthetic diversity of these Streptomyces strains is likely due to horizontal transfer events that occurred relatively recently in their evolutionary history instead of genetic diversification through a vertical transfer of BGCs. The multiple Type I PKSs presenting among these strains are highly conserved based on genetic similarity network analysis. There is currently some evidence supporting potential complex cross-BGC regulation in this class of Streptomyces strains. Jiang et al. demonstrated that a TetR family transcriptional regulator, GdmRIII, controls the biosynthesis of geldanamycin and elaiophylin meanwhile, in Streptomyces autolyticus CGMCC 051679. Recently, He et al. found that the rapamycin BGC-situated LAL family regulator RapH co-ordinately regulated the biosynthesis of both rapamycin and elaiophylin in S. rapamycinicus NRRL 549180. Although these reports correspond to cross-regulation between evolutionarily conserved BGCs, more details of these communications need to be investigated.

In conclusion, pteridic acids are secondary metabolites produced by streptomycetes enhancing plant resistance to abiotic stress. Transcriptomics profile revealed a higher expression of a diverse set of genes, e.g., in photosynthesis and abiotic stress response genes after pteridic acids treatment. This is a useful illustration of the bacterial metabolite-mediated alteration of plants in response to environmental stress. It will open avenues for utilising Streptomyces to rewild plant microbiomes and improve plant abiotic stress resistance to tackle climate change81.

Methods

Strains, plasmids, and cultivation

All strains and plasmids used in this study are listed in (Supplementary Table 5). All Streptomyces strains were obtained from the German Collection of Microorganisms and Cell Cultures GmbH (DSMZ, Germany). All Escherichia coli strains were grown in liquid/solid LB medium (5.0 g L–1 yeast extract, 10.0 g L–1 peptone, 10.0 g L–1 NaCl) at 37 °C. All Streptomyces strains were grown on SFM medium (20.0 g L–1 mannitol, 20.0 g L–1 soya flour, 20.0 g L–1 agar), and the SFM medium with the addition of 120 mM calcium chloride solution was used for the step of conjugation at 28 °C. The ISP2 medium (4.0 g L–1 yeast extract, 10.0 g L–1 malt extract, 4.0 g L–1 dextrose, and 1.0 L distilled water) was used for liquid fermentation of all Streptomyces strains used in plant assay and metabolomics analysis. Appropriate antibiotics were supplemented with the following working concentrations: apramycin (50 µg mL–1), chloramphenicol (25 µg mL–1), and kanamycin (50 µg mL–1). All chemicals utilised in this study were from Sigma-Aldrich, USA.

Metabolomics analyses

High performance liquid chromatography was carried out on the Agilent Infinity 1290 UHPLC system (Agilent Technologies, USA). The 250 × 2.1 mm i.d., 2.7 μm, Poroshell 120 Phenyl Hexyl column (Agilent Technologies, USA) was used for separation. The 2-μL samples were eluted at a flow rate of 0.35 mL min−1 using a linear gradient from 10% acetonitrile in Milli-Q water buffered with 20 mM formic acid increasing to 100% in 15 min. Each starting condition was held for 3 min before the next run. Mass spectrometry detection was performed on an Agilent 6545 QTOF MS equipped with Agilent Dual Jet Stream electrospray ion source (ESI) with a drying gas temperature of 160 °C, a gas flow of 13 L min−1, sheath gas temperature of 300 °C, and flow of 16 L min−1. The capillary voltage was set to 4000 V and the nozzle voltage to 500 V in positive mode. MS spectra were recorded as centroid data at an m/z of 100–1700, and auto MS/HRMS fragmentation was performed at three collision energies (10, 20, and 40 eV) on the three most intense precursor peaks per programme. Data were analysed with MassHunter software (Agilent Technologies, USA) and compared with known compounds and crude extract spectral libraries stored in the GNPS platform45. The precursor and fragment ion mass tolerance were set as 0.1 Da and 0.02 Da, respectively. In addition, the minpairs cos was set as 0.65, and the minimum matched fragment ions were set as 6.0. The metabolites profile of wild-type S. iranensis was visualised by MS-Dial 4.9.282.

Large-scale fermentation and isolation

S. iranensis was cultivated in medium 2 (3.0 g L–1 CaCl2·2H2O, 1.0 g L–1 citric acid/Fe III, 0.2 g L–1 MnSO4·H2O, 0.1 g L–1 ZnCl2, 0.025 g L–1 CuSO4·5H2O, 0.02 g L–1 Na2B4O7·10H2O, 0.01 g L–1 Na2MoO4·2H2O, and 20.0 g L–1 oatmeal in 1.0 L distilled water), at 175 L filling volume in a 300 L fermentation vessel (Sartorius, Germany). The fermentation was carried out for 6 days with aeration of 25-50 L min–1, stirring at 200 rpm with a temperature of 28 °C and at a pH range of 5.4-6.4. The fermentation broth was separated, filtered, and loaded onto an Amberchrom CG161Me resin LC column (200 × 20 cm, 6 L). Elution with a linear gradient of H2O-MeOH (from 30% to 100% v/v, flow rate 0.5 L min−1, in 58 min) afforded seven fractions (A–G). Fraction G was firstly fractionated by silica gel chromatography with a CH2Cl2/CH3OH gradient to yield 16 fractions, F01-F16. F07 was separated by a Sephadex LH-20 (MeOH) column and twelve sub-fractions F07a-l were obtained. From F07e, 1 (15.0 mg) and 2 (4.0 mg) were obtained by repeated HPLC RP-C18 (CH3CN/H2O as gradient).

Pteridic acid H (1): white solid; \({[\alpha ]}_{D}^{20}\) 81 (0.32 mg mL−1, CH3OH), 1H NMR (800 MHz, MeOD): 7.33 (dd, 15.4 Hz, 11.1 Hz, 1H), 6.26 (dd, 15.2 Hz, 10.8 Hz, 1H), 6.13 (dd, 15.3 Hz, 8.8 Hz, 1H), 5.90 (d, 15.4 Hz, 1H), 3.85 (dd, 10.2 Hz, 2.2 Hz, 1H), 3.69 (overlapping, 1H), 3.59 (m, 1H), 3.43 (m, 1H), 2.48 (m, 1H), 2.29 (dd, 14.9 Hz, 6.1 Hz, 1H), 2.01 (m, 1H), 1.64 (dd, 14.9 Hz, 1.9 Hz, 1H), 1.52 (m, 1H), 1.21 (d, 6.1 Hz, 3H), 1.49 (m), 1.01 (d, 6.8 Hz, 3H), 0.96 (d, 6.8 Hz, 3H), 0.93 (t, 7.3 Hz, 3H), 0.91 (d, 7.0 Hz, 3H); 13C NMR (200 MHz, MeOD): 168.7, 151.0, 146.7, 129,6, 120.6, 103.2, 75.5, 72.8, 70.3, 70.1, 51.3, 42.2, 40.4, 37.5, 37.4, 24.8, 20.9, 16.0, 12.1, 10.1, 5.0; UV/vis (CH3CN/H2O) λmax 262 nm; IR (ATR) vmax 2967, 2934, 2879, 1712, 1642, 1600, 1458, 1410, 1383, 1300, 1266, 1223, 1187, 1142, 1109, 1058, 1002, 973 cm−1; (+)-HR-ESI-MS (m/z) [M + H]+ calcd for C21H35O6, 383.2428; found, 383.2439. 1H NMR and 13C NMR see Supplementary Tables 6 and 7.

Pteridic acid F (2): white solid; \({[\alpha ]}_{D}^{20}\) −18 (10 mg mL−1, CH3OH), 1H NMR (800 MHz, MeOD): 5.97 (d, 15.1 Hz, 1H), 7.16 (dd, 15.1 Hz, 10.9 Hz, 1H), 6.25 (dd, 15.1 Hz, 10.9 Hz, 1H), 6.07 (dd, 15.1 Hz, 8.6 Hz, 1H), 3.88 (m, 1H), 3.66 (td, 10.8 Hz, 4.3 Hz, 1H), 3.56 (dd, 11.5 Hz, 4.7 Hz, 1H), 3.32 (m, 1H), 2.49 (m, 1H), 2.19 (dd, 13.1 Hz, 4.3 Hz, 1H), 2.02 (m, 1H), 1.69 (m, 1H), 1.32 (dd, 13.2 Hz, 11.2 Hz, 1H), 1.14 (d, 6.2 Hz, 3H), 1.02 (m, 1H), 1.02 (d, 6.8 Hz, 3H), 0.95 (d, 6.9 Hz, 3H), 0.98 (d, 6.8 Hz, 3H), 1.60 (m, 1H), 1.44 (m, 1H), 0.82 (t, 7.6 Hz, 3H); 13C NMR (200 MHz, MeOD): 170.2, 148.1, 148.1, 129,6, 122.9, 103.2, 78.0, 74.7, 66.3, 66.3, 52.2, 42.1, 40.5, 38.0, 33.8, 20.4, 15.9, 12.7, 5.3; UV/vis (CH3CN/H2O) λmax 264 nm; IR (ATR) vmax 2968, 2931, 2877, 1692, 1643, 1618, 1458, 1410, 1380, 1299, 1270, 1188, 1138, 1106, 1059, 1002, 973, 850 cm−1; (+)-HR-ESI-MS (m/z): [M + H]+ calcd for C21H35O6, 383.2428; found, 383.2433. 1H NMR and 13C NMR see Supplementary Tables 6 and 7.

Structure identification

NMR spectra were recorded on an 800 MHz Bruker Avance III spectrometer equipped with a TCI CryoProbe using standard pulse sequences. NMR data were processed using MestReNova 11.0. UHPLC-HRMS was performed on an Agilent Infinity 1290 UHPLC system equipped with a diode array detector. UV-Vis spectra were recorded from 190 to 640 nm. Specific rotations were acquired using Perkin-Elmer 241 polarimeter. IR data were acquired on Bruker Alpha FTIR spectrometer using OPUS version 7.2. TLC analysis was performed on silica gel plates (Sil G/UV254, 0.20 mm, Macherey-Nagel). The Biotage Isolera One Flash Chromatography system was used for flash chromatography and performed on silica gel 60 (Merck, 0.04–0.063 mm, 230–400 mesh ASTM). Sephadex LH-20 was from Pharmacia.

Crystal structure determination

X-ray data collection of 1 was performed on an Agilent Supernova Diffractometer using CuKα radiation. Data were processed and scaled using the CrysAlisPro software (Agilent Technologies, USA). The structure was solved using SHELXS and refined using SHELXL. Hydrogen atoms were included in ideal positions using riding coordinates. The absolute configuration was determined based on the Flack parameter. Crystal Data for 1: C21H34O6, M = 382.50, monoclinic, a = 8.4619(1) Å, b = 15.6161(2) Å, c = 8.4994(1) Å, α = 90.00°, β = 107.768(1)°, γ = 90.00°, V = 1069.55(2)Å3, T = 120.(2) K, space group P21, Z = 2, μ(Cu Kα) = 0.698 mm–1, 17514 reflections collected, 4275 independent reflections (Rint = 0.0226, Rsigma = 0.0155). The final R1 values were 0.0249 (I > 2σ(I)). The final wR2 values were 0.0648 (I > 2σ(I)). The final R1 values were 0.0252 (all data). The final wR2 values were 0.0651 (all data). The goodness of fit on F2 was 1.057. Flack parameter = 0.13(10).

Genetic manipulation

All primers used were synthesised by IDT (Integrated DNA Technologies, USA) and listed in (Supplementary Table 8). Plasmids and genomic DNA purification, PCR and cloning were conducted according to standard procedures using manufacturer protocols. PCR was performed using OneTaq Quick-Load 2X Master Mix with Standard Buffer (New England Biolabs, USA). DNA assembly was done by using NEBuilder HiFi DNA Assembly Master Mix (New England Biolabs, USA). DNA digestion was performed with FastDigest restriction enzymes (Thermo Fisher Scientific, USA). NucleoSpin Gel and PCR Clean-up Kits (Macherey-Nagel, Germany) were used for DNA clean-up from PCR products and agarose gel extracts. One Shot Mach1 T1 Phage-Resistant Chemically Competent E. coli (Thermo Fisher Scientific, USA) was used for cloning. NucleoSpin Plasmid EasyPure Kit (Macherey-Nagel, Germany) was used for plasmid preparation. Sanger sequencing was carried out using a Mix2Seq Kit (Eurofins Scientific, Luxembourg). All DNA manipulation experiments were conducted according to standard procedures using manufacturer protocols.

Gene inactivation and site-directed mutagenesis

To use the pCRISPR-cBEST for base editing applications, an oligo was designed as Del-ptaA by the online tool CRISPy-web, and the pCRISPR-cBEST plasmid was linearised by NcoI. Mixing the linearised pCRISPR-cBEST plasmid and Del-ptaA with the NEBuilder HiFi DNA Assembly Master Mix (New England Biolabs, USA). The linearised pCRISPR-cBEST plasmid was then bridged by Del-ptaA, ending with the desired pCRISPR-cBEST/ΔptaA. Chemically competent E. coli were transformed with the recombinant plasmid and confirmed via PCR amplification (programme: 94 °C for 30 s, followed by 30 cycles consisting of 94 °C for 15 s; 54 °C for 15 s; 68 °C for 40 s, and 68 °C for 2 min) and Sanger sequencing. The experimental procedure for site-directed mutagenesis for the TE domain is the same as described above. The E. coli-Streptomyces conjugation experiment was conducted according to the modified protocol in this study, and the mutant Streptomyces strains were also confirmed by PCR and Sanger sequencing (Eurofins, France).

Construction of BAC and genetical complementation

The BAC library of S. iranensis was constructed using pESAC13-A from Bio S&T (Montreal, Canada). Based on a high-throughput screening method (unpublished), we selected two BACs 1J23 and 6M10 that cross-cover the ptaA gene (Supplementary Fig. 28). The selected BAC clones were further confirmed using four sets of primers, including ID-1J23-right-F/R, ID-1J23-left-F/R, ID-6M10-right-F/R, and ID-6M10-left-F/R (Supplementary Table 8). Subsequently, we introduced these two BAC clones 1J23 and 6M10 into S. iranensis/ΔptaA (remove the pCRISPR-cBEST/ΔptaA to obtain antibiotics resistance free strain) separately by conjugation. Exconjugants of mutants were further validated by apramycin resistance screening and PCR.

Enrichment evaluation of S. iranensis in rhizosphere soil

The S. iranensis (with the apramycin resistance gene) spore suspension was well mixed with fully sterilised soil and was transferred to a 250 mL flask. The sterilised germinated barley seed was placed in the centre position of soil in flask and grown at 24 ± 2 °C, 8 h dark/16 h light in the growth chamber for 7 days. Samples were collected from soil within 0.1 cm and 3 cm distance from the barley root, with a specification of 0.1 g soil per sample. Then, these samples were transferred to sterilised 1.5 mL Eppendorf tubes and mixed with 500 μL sterilised H2O. 200 μL of each sample was spread evenly over the solid MS medium with the addition of 50 μg mL−1 apramycin and grown at 28°C for 7 days. The number of Streptomyces colonies grown on each plate were counted and statistically analysed.

Arabidopsis growth assays

A. thaliana ecotype Columbia (Col-0) was used to test the effects of pteridic acids treatment under drought stress mediated by PEG-6000 (Duchefa Biochemie BV) and salinity stress mediated by NaCl (Duchefa Biochemie BV). The modified Murashige & Skoog medium (2.2 g L–1 Murashige & Skoog medium including B5 vitamins, 5.0 g L–1 of sucrose, 250 mg L–1 MES monohydrate, 7.0 g L–1 agar, and adjusted pH to 5.7 with KOH) was used in this study. PEG-6000 (15% w/v) was dissolved in water and filtered through 0.2-micron Sartorius Minisart™ Plus Syringe Filters (Fisher Scientific). 50 mL of the filtered solution was overlaid onto the surface of solidified Murashige & Skoog medium. The plates were left for 24 h to diffuse the PEG into the Murashige & Skoog medium. NaCl was added to the medium to a final concentration of 80 mM for the salinity stress alleviation test. The pure compound pteridic acids H and F (0.50 ng mL−1), IAA (0.23 ng mL−1), and ABA (0.34 ng mL−1) were mixed with different media to a final concentration of approximately 1.3 nM and poured into the plates. Seeds were surface sterilised by washing with 70% ethanol for 2 min, then in sterilisation solution (10% bleach) for 1 min by inverting the tubes, and finally washed five times with sterilised water. The seeds were stratified for 2 days at 4 °C in the dark. Sterilised seeds were placed in Petri dishes (approx. 100 seeds per Petri dish) on Murashige & Skoog medium and grown for 3–4 days in the vertical position in a culture chamber at 22 °C under standard long-day conditions (16/8 h light/dark photoperiod). After three days of growth, seedlings with similar root lengths (7-10 mm) were transferred to square plates containing Murashige & Skoog medium (control) or Murashige & Skoog medium supplemented with 15% (w/v) PEG-6000 or 80 mM NaCl. 16 seedlings were used per replicate for each treatment. The initial position of the plant root tip was marked with a marker. The plants were grown in the vertical position under standard long day conditions (22 °C, 16/8 h light/dark) for 8 days, and then each plate was scanned using Image Scanner. Primary and lateral root lengths and the total plant weight were then scored. The primary and lateral root length measurements were performed by analysing pictures with the Image J software. Fresh weight measurements were estimated using a precision balance. Whole 8-day-old plants grown in the medium were removed using forceps, dried in tissue paper, and then weighed using a precision balance.

Barley and mung bean growth assays

The pteridic acid streptomycetes producers were tested for their effects on barley cultivars Guld grown in soil. S. iranensis HM 35, S. rapamycinicus NRRL 5491, and S. violaceusniger Tu 4113 were cultivated in ISP2 medium for 7 days at 28 °C. 500 µL culture broth (ca. 3 × 109 CFU/mL) was added to 100 g sand soil and well mixed. Treated soils were infiltrated with Milli-Q water, 20% PEG-6000 solutions to simulate osmotic stress, and 100 mM NaCl solutions to simulate salinity stress in soil environments. To simulate drought stress during barley growth, the plant was initially watered for 7 days, then subjected to 7 days of drought stress, and finally allowed to recover with various treatments for another 7 days. Barley seeds were rinsed in distilled water and sterilised with 1% sodium hypochlorite for 15 min. They were then washed with distilled water and germinated in distilled water at 24 °C for 2 days. Barley seeds were planted in each plastic pot (5 cm × 5 cm × 6 cm, six seedlings per pot) supplemented with different treated soil and grown at 24 ± 2 °C, 8 h dark/16 h light in the growth chamber for 7 days. Plant height (cm) was measured as the aerial part of the plant, and the fresh shoot weight (g) and fresh root weight (g) of each seedling were measured separately. Then, the seedlings were dried in the hot-air oven at 70 °C for 6 h to obtain the dry shoot and dry root weights (g). For heavy metal stress experiment, mung beans were pre-germinated and placed on top of the modified Murashige & Skoog medium agar, supplemented with 10 mM CuSO4, with 1.0 ng mL−1 pure substances. All mung beans were grown in the dark at 24 ± 2 °C for 4 days.

Kidney beans growth assays

The seeds of kidney beans (appr. 2 cm in length, from organic farming) were firstly sterilised successively with ethanol (70% v/v) and sodium hypochlorite (5% v/v), each for 2 min and then rinsed with sterile Milli-Q water (three times). The sterilised seeds were cultivated on modified Murashige & Skoog agar plates for 3–4 days. After germination, the seedlings (with 1.5–2-cm-long roots) were soaked in 10 mL aliquots of testing compounds (pteridic acids H and F, 1.0 ng mL−1, dissolved in sterile Milli-Q water) in ultra-clear polypropylene containers (ø 34 mm, vol. 20 mL) with polyethylene caps. The control group was treated with 10 mL sterile Milli-Q water. For each treatment, three repetitions (containers) were used, and each repetition included four seedlings. After 24 h, the seeds were transferred into a cut square petri dish, put on the top layer of sandy soil, and then incubated vertically in a growth chamber (24/22 °C, day/night cycle of 16/8 h, 50%, 60%, 70%, 100% of circulated wind velocity for 12 h, 2 h, 2 h, 8 h) for 7 days. Solutions of pteridic acids H and F (2 mL, 1.0 ng mL−1 for both) were added separately into corresponding containers, with sterile Milli-Q water as control, and an extra 8 mL of Milli-Q water was added to each petri dish every other day.

RNA extraction and mRNA-seq

Arabidopsis seeds were sterilised, and seven to ten seeds were cultured in each well of a 6-well plate in 2 mL medium containing modified Murashige & Skoog and 30 mM sucrose with 16 h light, at 24 °C in a controlled environment room. After 7 days, seedlings were washed in 20 mL modified Murashige & Skoog liquid medium and then moved to 20 mL of fresh liquid medium containing 80 mM NaCl and cultured in 50 ml E-flasks with 1 ng mL−1 of each pteridic acids H or F (or equal amount of water as control). Whole-seedling plant samples were collected after 72 h when significant phenotypic variation occurred between groups. Total RNA was extracted using RNeasy Plant Mini Kit (Qiagen, Germany). mRNA was purified from total RNA using poly-T oligo-attached magnetic beads. After fragmentation, the first strand cDNA was synthesised using random hexamer primers, followed by the second strand cDNA synthesis using dTTP for the non-directional library construction. The reads were generated using an Illumina NovaSeq 6000 (Novogene, UK) with a paired-end 150 bp configuration. DESeq2 was used to estimate DEGs between different treatments with the threshold of FDR-adjusted p values ≤ 0.05 and |log2(FoldChange)|≥1 if there is no additional statement83. The online software g:Profiler (http://biit.cs.ut.ee/gprofiler/gost) was used for GO enrichment and clusterProfiler R package was used to test the statistical enrichment of DEGs in KEGG pathways84,85.

Bioinformatics analyses

The identification and annotation of all BGCs of Streptomyces secondary metabolites were carried out with antiSMASH 6.026. The threshold for similar BGCs was selected as greater than 45% sequence similarity. The BiG-FAM database and BiG-SCAPE software were used to identify the distribution of pta gene clusters in different bacteria and generate the pta-gene cluster family similarity network60,61. A cut-off of 0.3 was used as a raw index similarity metric for the BiG-SCAPE analysis. The alignment of 15 pta-containing Streptomyces genome sequences was performed using the whole-genome alignment plugin of the CLC Genomics Workbench version 22.0.2 (Qiagen). The minimum initial seed length was 15 bp, and the minimum alignment block length was 100 bp. The networks of Big-SCAPE and GNPS analysis were visualised using Cytoscape 3.9. The phylogenetic analysis was conducted by the online multiple sequence alignment tool MAFFT and visualised by iTOL v586,87.

Statistical analysis

Statistical significance was assessed by one-way ANOVA with post hoc Dunnett’s multiple comparisons test, one-way ANOVA with Tukey test or t test (see each figure legend). All analyses were performed using GraphPad Prism version 9. p values < 0.05 were considered significant. Asterisks indicate the level of statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. For all relevant figures, source data and exact p values are provided in the Source Data file.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.