Tryptophan-independent auxin biosynthesis contributes to early embryogenesis in Arabidopsis Bing Wanga,1, Jinfang Chub,1, Tianying Yua,1, Qian Xua, Xiaohong Sunb, Jia Yuana, Guosheng Xionga, Guodong Wanga, Yonghong Wanga,b, and Jiayang Lia,b,2 a State Key Laboratory of Plant Genomics and bNational Center for Plant Gene Research (Beijing), Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, Beijing 100101, China

Contributed by Jiayang Li, February 27, 2015 (sent for review February 5, 2015; reviewed by Klaus Palme and Hong-Wei Xue)

|

phytohormone IAA biosynthesis Arabidopsis thaliana

| tryptophan | embryogenesis |

T

he phytohormone auxin plays critical roles in almost every aspect of plant development including embryogenesis, architecture formation, and tropic growth. One of the most fascinating topics in plant biology is how auxin can have so many diverse and context-specific functions (1). Dynamic auxin gradients, which are regulated mainly by local auxin synthesis, catabolism, conjugation, and polar auxin transport, are essential for integration of various environmental and endogenous signals (2, 3). Auxin perception and signaling systems are responsible for a read-out of the auxin gradients (1). Local auxin biosynthesis is important for leaf development, shade avoidance, root-specific ethylene sensitivity, vascular patterning, flower patterning, and embryogenesis (4). Indole3-acetic acid (IAA), the naturally occurring principal auxin in plants, is biosynthesized from tryptophan (Trp) through four proposed routes according to their key intermediates, namely indole3-acetaldoxime (IAOx), indole-3-pyruvic acid (IPyA), indole3-acetamide (IAM), and tryptamine (TAM) (5). The TRYPTOPHAN AMINOTRANSFERASE OF ARABIDOPSIS (TAA)/YUCCA (YUC) linear pathway has been considered as a predominant Trp-dependent auxin biosynthetic pathway (4, 6–8). However, labeling studies and analyses of Trp auxotrophic mutants have long predicted the existence of a Trp-independent IAA biosynthetic pathway (9–13). When the Arabidopsis and maize seedlings were fed with isotope-labeled precursor, IAA was more enriched than Trp, and the incorporation of label into IAA from Trp is low, suggesting that IAA can be produced de novo without Trp as an intermediate (11, 12). The Trp biosynthetic mutant trp1, defective in phosphoribosylanthranilate transferase (PAT), and indole-3-glycerol phosphate synthase (IGS) antisense plants are deficient in steps earlier than indole-3-glycerol phosphate (IGP) formation and display decreased levels of both IAA and Trp (10, 14). However, the trp3 and trp2 mutants, defective in tryptophan synthase α (TSA) and tryptophan synthase β (TSB) subunits, respectively, accumulate higher levels of IAA than the wild type despite containing lower Trp levels (10, 11, 15, 16). These data strongly suggested the existence of the Trp-independent IAA biosynthetic pathway that might branch from IGP and/or indole (10).

www.pnas.org/cgi/doi/10.1073/pnas.1503998112

Further studies suggested that a cytosol-localized indole synthase (INS) may be involved in the Trp-independent biosynthesis of indole-containing metabolite (17). However, the molecular basis, biological functions, and spatiotemporal regulation of the Trpindependent IAA biosynthetic pathway have remained a mystery. In higher eukaryotes, embryogenesis initiates the generation of the species-specific body plan. During embryogenesis, a singlecelled zygote develops into a functional multicellular organism with cells adopting specific fates according to their relative positions. In higher plants, essential architecture features, such as body axes and major tissue layers, are established in embryogenesis, and auxin plays a vital role in apical-basal patterning and embryo axis formation (18, 19). Local auxin biosynthesis, polar auxin transport facilitated by PIN-FORMED1/3/4/7 (PIN1/3/4/7), and auxin response coordinately regulate apical–basal pattern formation during embryogenesis (7, 20–22). The TAA and YUC families in Trp-dependent IAA biosynthesis predominantly regulate embryogenesis at or after the globular stage (7, 22). However, the auxin source for embryogenesis before the globular stage remains elusive. In this study, we provide, to our knowledge, the first genetic and biochemical evidence that the cytosol-localized INS is a key component in the long predicted Trp-independent auxin biosynthetic pathway and is critical for apical–basal pattern formation during early embryogenesis in Arabidopsis. Significance The phytohormone indole-3-acetic acid (IAA) plays a vital role in plant growth and development. IAA can be synthesized through the precursor tryptophan (Trp), known as the Trpdependent IAA biosynthetic pathway. However, IAA may also be synthesized through a proposed Trp-independent IAA biosynthetic pathway. Although the Trp-independent IAA biosynthesis was hypothesized 20 years ago, it remains a mystery. In this paper, we provide compelling evidence that the cytosollocalized indole synthase (INS) initiates the Trp-independent IAA biosynthetic pathway and that the spatial and temporal expression of INS plays an important role in the establishment of the apical–basal pattern during early embryogenesis, demonstrating that the Trp-dependent and -independent IAA biosynthetic pathways coordinately regulate embryogenesis of higher plants. Author contributions: J.L. supervised the project; B.W., J.C., T.Y., and J.L. designed research; B.W., J.C., T.Y., Q.X., X.S., J.Y., G.X., and G.W. performed research; B.W., J.C., T.Y., and J.L. analyzed data; and B.W., Y.W., and J.L. wrote the paper. Reviewers: K.P., Albert Ludwigs University of Freiburg; and H.-W.X., Institute of Plant Physiology and Ecology, Shanghai Institutes of Biological Sciences, Chinese Academy of Sciences. The authors declare no conflict of interest. Freely available online through the PNAS open access option. 1

B.W., J.C., and T.Y. contributed equally to this work.

2

To whom correspondence should be addressed. Email: [email protected].

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. 1073/pnas.1503998112/-/DCSupplemental.

PNAS | April 14, 2015 | vol. 112 | no. 15 | 4821–4826

PLANT BIOLOGY

The phytohormone auxin regulates nearly all aspects of plant growth and development. Tremendous achievements have been made in elucidating the tryptophan (Trp)-dependent auxin biosynthetic pathway; however, the genetic evidence, key components, and functions of the Trp-independent pathway remain elusive. Here we report that the Arabidopsis indole synthase mutant is defective in the long-anticipated Trp-independent auxin biosynthetic pathway and that auxin synthesized through this spatially and temporally regulated pathway contributes significantly to the establishment of the apical–basal axis, which profoundly affects the early embryogenesis in Arabidopsis. These discoveries pave an avenue for elucidating the Trp-independent auxin biosynthetic pathway and its functions in regulating plant growth and development.

whereas overexpression of INS could only partially rescue trp3-1 (Fig. 1D). Overexpression of TSA without a cTP (ΔcTP-TSA) could rescue DR5-driven GUS activities in cotyledons of ins-1, whereas TSA or cTP-INS overexpression had only a partial effect (Fig. S3A). Collectively, these data demonstrate that functional differentiation of INS and TSA results from their distinct subcellular localizations. We then investigated the effect of exogenous Trp on phenotypes and IAA levels of wild type, ins-1, and trp3-1. Treatment of trp3-1 with Trp could largely restore the disordered expression pattern of DR5:GUS on root tips and root growth and reduce the IAA accumulation (from 67 to 36% above the wild type) (Fig. S3 B–D). However, Trp could not fully restore the DR5:GUS pattern nor raise the IAA level in ins-1 compared with that in the wild type (Fig. S3 B–D), consistent with the hypothesis that INS participates specifically in the Trp-independent IAA biosynthetic pathway. INS Specifically Participates in the Trp-Independent IAA Biosynthesis.

Fig. 1. INS and TSA display distinct subcellular localizations and different functions in tryptophan biosynthesis and IAA biosynthesis. (A) Subcellular localizations of INS-GFP, TSA-GFP, cTP-INS-GFP, and ΔcTP-TSA-GFP in protoplasts from stable transgenic plants. Ch, chlorophyll. (B) Phenotypes of 12-d-old wild-type (Col-0), ins-1, trp3-1, 35S:TSA/trp3-1, 35S:cTP-INS/trp3-1, and INS:INS/trp3-1 seedlings. (C) Tryptophan (Trp) contents in 7-d-old seedlings. gdw−1, per gram dry weight. The Trp accumulation mutant trp5-1 was used as a control. Data are represented as mean ± SEM; **P < 0.01 by Student’s t test; n = 5. (D) Free IAA contents in 12-d-old seedlings (n = 5). gfw−1, per gram fresh weight. Data are represented as mean ± SEM; *P < 0.05 and **P < 0.01 by Student’s t test; n = 5. (Scale bars, 10 μm in A and 0.5 cm in B.)

To further distinguish the activities of Trp-dependent and -independent IAA biosynthetic pathways, seedlings of Columbia (Col-0), ins-1, and trp3-1 were treated with labeled anthranilic acid in the presence or absence of unlabeled Trp at high temperature and then the labeled IAA contents were measured (Fig. 2A and Fig. S4 A and B). Anthranilic acid is a common precursor of the Trp-dependent and -independent IAA biosynthetic pathways, whereas exogenous Trp should increase the Trp pool size

Results Functional Differentiation of INS and TSA Results from Their Distinct Subcellular Localizations. INS and TSA are close paralogs that

contain a conserved Trp synthase α-subunit domain (Fig. S1A). TSA contains a chloroplast transit peptide (cTP) domain and is localized in chloroplasts, whereas INS lacks the cTP and is localized in cytosol in protoplasts of stable transgenic plants (Fig. 1A). Trp biosynthesis has been proposed to take place in chloroplasts, and TSA and TSB were considered to form a complex and catalyze the formation of Trp (10, 23, 24). Previous work has shown that INS could complement Trp auxotrophy of the Escherichia coli Δtrp A strain, which is defective in Trp biosynthesis due to the deletion of TSA (17). We therefore investigated whether INS participates in IAA biosynthesis through the Trp-independent pathway. Based on the findings that both trp3-1 and trp2-1 mutants, defective in TSA and TSB1, respectively, are Trp auxotrophic (15, 16), we show here that they have lower contents of Trp, whereas the INS null mutant shows little Trp auxotrophic phenotype and has a normal Trp content (Fig. 1 B and C and Fig. S1). Overexpression of INS with a cTP (cTP-INS) could completely rescue the Trp auxotrophic phenotype and restore the Trp level of trp3-1, whereas overexpression of INS could only partially rescue trp3-1 (Fig. 1 B and C and Fig. S2), suggesting that INS could produce indole in cytosol and feed into Trp biosynthesis effectively if relocated to the chloroplast. Furthermore, the trp3-1 and trp2-1 mutants have higher levels of IAA than the wild type, whereas the INS null mutants ins-1 and ins-2 have decreased IAA levels (Fig. 1D and Fig. S1 C–E). INS could complement the deficiency of ins-1 in terms of the IAA content and DR5-driven GUS activity that reflects the read-out of auxin signal transduction (25), suggesting that INS is involved in auxin biosynthesis (Fig. 1D and Fig. S3A). Overexpression of cTP-INS could restore the IAA level of trp3-1, 4822 | www.pnas.org/cgi/doi/10.1073/pnas.1503998112

Fig. 2. INS is a key component in the Trp-independent IAA biosynthetic pathway. (A) Schematic diagrams of 13C6-anthranilic acid (13C6-ANT) feeding experiment. Seedlings were transferred to liquid Murashige and Skoog (MS) media containing 13C6-ANT and cultured for 24 h. In a parallel test, the MS media contained 13C6-ANT and unlabeled Trp, which should increase the Trp pool size and decrease the biosynthesis of 13C6-IAA only through the Trpdependent pathway. (B) 13C6-IAA contents in 12-d-old wild-type (Col-0), ins-1, and trp3-1 seedlings after 13C6-ANT or 13C6-ANT plus Trp treatment. gfw−1, per gram fresh weight; ns, no significant difference. Data are represented as mean ± SEM; **P < 0.01 by Student’s t test; n = 5. (C) Phenotypes of 7-d-old wild-type (Col-0), ins-1, sur2, ins-1 sur2, trp3-1, and trp3-1 sur2 seedlings. (Scale bar, 0.5 cm.) (D) Hypocotyl length of 9-d-old wild-type (Col-0), ins-1, wei8-1, and ins-1 wei8-1 seedlings at 20 °C and 28 °C. Data are represented as mean ± SEM; different letters indicate significant difference by Tukey’s multiple comparison test (P < 0.05); n = 30.

Wang et al.

and decrease the biosynthesis of 13C6-IAA through the Trp-dependent pathway (Fig. 2A). When exogenous Trp was applied, the wild-type and ins-1 seedlings exhibited significant decreases in the 13C6-IAA content, but this was greater in ins-1 (31%) compared with that of the wild type (18%), whereas trp3-1 seedlings displayed no significant decrease (Fig. 2B), indicating that Trp-independent IAA biosynthesis was deficient in ins-1 whereas Trp-dependent IAA biosynthesis was impaired in trp3-1. These phenomena explain the defective root development of trp3-1, such as reduced main root length and impaired gravitropism (Fig. S3C), which resemble the root phenotypes of wei8-1 tar2-1 and wei8-1 tar2-2 tar1-1 seedlings (7). Consistent with this, we found that the high-auxin phenotypes such as the elongated hypocotyl and extended petiole of sur2, in which the Trpdependent IAA biosynthesis is elevated (26), were suppressed by trp3-1 but not by ins-1 (Fig. 2C and Fig. S4C). And the trp3-1 sur2 double mutant displayed similar Trp auxotrophic phenotypes to trp3-1 (Fig. 2C and Fig. S4C). Furthermore, we found that the ins-1 mutant is impaired in the hypocotyl elongation induced by high temperature and has an additive effect with wei8-1 (Fig. 2D), a null mutant of TAA1 in the Trp-dependent IAA biosynthetic pathway (7), suggesting that INS probably acts in a parallel pathway with TAA1 to regulate the hypocotyl length at high temperature. Taken together, these results indicate that INS is involved in the production of IAA through the Trpindependent pathway.

PLANT BIOLOGY

INS Is Required for Early Embryogenesis. In Arabidopsis the cell division pattern of an embryo is nearly stereotypic. In the wild type, a zygote divides asymmetrically, generating a smaller apical cell and a larger basal cell. The apical cell undergoes two rounds of longitudinal, one round of horizontal, and one round of periclinal divisions, giving rise to a 16-cell embryo proper, whereas the basal cell divides transversely to produce the suspensor (18). The uppermost suspensor cell is specialized to the hypophysis (HY), which subsequently produces an upper lens-shaped cell

(LSC) that will form the quiescent center (QC) and a lower daughter cell that will form the columella stem cells (CSC) (Fig. 3A and Fig. S5A). The embryo development of trp3-1 is normal (Fig. 3B); however, the stereotypical pattern of cell division is disturbed in ins-1 from the one-cell stage onward (Fig. 3 C–I). In both ins-1 and ins-2 mutants, ∼6–9% embryos arrested at the 1-, 4-, 8-, or 16-cell stage with a correct cell division pattern (Fig. 3C and Fig. S5 B, M, and O), and more importantly, ∼11–14% embryos displayed abnormal cell division patterns (Fig. 3 D–I and Fig. S5 C–M and O). For example, the apical cell, basal cell, and their progeny underwent divisions with abnormal cell plates at a wrong position or in a wrong orientation (Fig. 3 D–G and Fig. S5 C–F, arrow). From the globular to torpedo stages, the cell group in the basal embryo region, which normally displays characteristic cell arrangement, was consistently disorganized (Fig. 3 G–I and Fig. S5 G–L, bracket). Furthermore, there is no cotyledon primordium bulging out in the apical embryo region (Fig. 3I and Fig. S5L). The embryonic defects could be complemented by overexpression of the wild-type INS gene (Fig. S5 N and O), demonstrating that INS is responsible for the defective embryogenesis of ins-1. Together, it is suggested that INS plays a critical role in early embryogenesis. Furthermore, we can identify neither an ins-1 trp3-1 seedling from more than 500 progeny of the ins-1 trp3-1+/− plants nor an ins-1 trp3-100 seedling from more than 500 progeny of the ins-1 trp3-100+/− plants, indicating that the ins-1 trp3-1 and ins-1 trp3-100 double mutants were completely embryo lethal. Both ins-1 trp3-1 and ins-1 trp3-100 exhibited arrested embryos with normal apical–basal polarity (Fig. 3J and Fig. S6 A–D) and embryos with abnormal cell division planes (Fig. 3 K–P and Fig. S6 E–O). These abnormal divisions resulted in embryos defective in the apical–basal polarity (Fig. 3 K–N and Fig. S6 E–J), embryos with disorganized basal cell regions (Fig. 3 O and P and Fig. S6 K–O), and embryos without cotyledon primordia (Fig. 3P and Fig. S6O). Occasionally, the ins trp3 embryo formed filamentous structure without differentiation of the apical cell (Fig. 3K and

Fig. 3. Effects of INS and TSA on embryogenesis. Embryos of wild-type (Col-0) (A) and trp3-1 (B) plants at 1-, 4-, 8- or 16-cell globular (g), heart (h), torpedo (to), and cotyledon (c) stages. HY, hypophysis; LSC, lens-shaped cell; QC, quiescent center; and CSC, columella stem cells. (C–I) When normal sibling embryos reach the heart stage, defective embryos of ins-1 are arrested before the globular stage with normal body plans (C) or display abnormal cell division patterns (D–I). (J–P) When normal sibling embryos reach the heart stage, defective embryos from the ins-1 trp3-1+/− siliques are arrested with normal body plans (J) or exhibit abnormal cell division patterns (K–P). Arrows indicate aberrant cell divisions in early embryogenesis, and brackets mark aberrant basal cell regions.

Wang et al.

PNAS | April 14, 2015 | vol. 112 | no. 15 | 4823

Fig. S6E). The embryonic defects of ins and ins trp3 resembled the defects of auxin-related mutants, such as pin7, pin1/3/4/7, gnom, monopteros (mp), and bodenlos (bdl) (20, 27–29), suggesting that deficiency in auxin activity is the likely cause of defects during early embryogenesis. Trp-Independent IAA Biosynthesis Contributes to Apical–Basal Axis Establishment. To understand the roles of Trp biosynthesis and

IAA biosynthesis in embryogenesis, we compared the embryo development of Col-0, ins, trp3, ins trp3, trp2, and ins trp2 plants. The Trp auxotrophic mutants trp3 and trp2 underwent normal embryogenesis comparable to the wild type, whereas ins-1 displayed a normal Trp level but a significantly higher percentage of defective embryos (Figs. 1C, 3B, and 4A; Fig. S7 A–E), suggesting that Trp biosynthesis may play a limited role in early embryogenesis. Furthermore, although trp3-100 is a weak allele of TSA (15) (Fig. S1A), the ins-1 trp3-100 double mutant is completely embryo lethal (Fig. 4A and Figs. S6 and S7 A–E), suggesting that ins-1 trp3-1 and ins-1 trp3-100 are embryo lethal probably because of a lack of auxin biosynthesis. Moreover, direct measurement of IAA contents in ovules showed that, compared with the wild type, free IAA levels decreased to 78% in morphologically normal ovules of ins-1, to 31% in morphologically abnormal ovules of ins-1, and to 15% in abnormal ovules of the ins-1 trp3-1+/− plants, whereas IAA levels increased to 115% in trp3-1 ovules (Fig. 4B). Furthermore, we found that the embryonic deficiency of ins-1 was suppressed by trp2-1 and trp2-301, probably due to the increased IAA levels of ins-1 trp2-1 and ins-1 trp2-301 double mutants (Figs. 1D and 4A; Figs. S7 A–E and S8A), suggesting that IAA biosynthesis is important for early embryogenesis.

The Trp auxotrophic phenotypes of ins-1 trp2-1 and ins-1 trp2-301 were similar to those of trp2-1 and trp2-301, respectively (Fig. S8B). These phenomena may be explained by the observation that trp2-1 seedlings over-accumulated indole (11), which might diffuse from chloroplasts to the cytosol and compensate for the deficiency of ins-1. We then compared the spatiotemporal expression patterns of INS and TSA proteins during embryogenesis. INS is expressed uniformly at the 1- and 2-cell stages, predominantly in suspensor cells at the 4- and 8-cell stages, and increases gradually in the proembryo, especially in the periphery and basal domains from the 16-cell stage. By contrast, TSA is expressed from the four-cell stage and exhibits a uniform expression pattern (Fig. 4C). The DR5:GFP reporter has been used to visualize the read-out of auxin signaling during embryogenesis (20, 30). We further compared the DR5 activity represented by GFP signals in Col-0, ins-1, trp3-1, and ins-1 trp3-1 embryos. In the wild-type embryo, the GFP signals were observed in the apical cell and its progeny but very weakly in the suspensor cells before the globular stage (Fig. 4D). At the transition stages, GFP signals accumulated mainly at the root meristem region (Fig. 4E). Compared with the wild type, the DR5:GFP expression patterns were disturbed in arrested or morphologically defective ins-1 embryos (Fig. 4 F and G), but displayed regular patterns in trp3-1 (Fig. 4 H and I) and morphologically normal ins-1 embryos (Fig. 4 J and K). In the ins-1 trp3-1 embryos, the GFP signals were hardly detected in arrested embryos (Fig. 4L), nor in embryos with abnormal cell division patterns (Fig. 4M). Collectively, these results demonstrate that

Fig. 4. Trp-independent IAA biosynthesis contributes significantly to early embryogenesis. (A) Percentage of defective embryos in the wild type (Col-0) and mutants at the globular stage. Data are represented as mean ± SEM; **P < 0.01 by Student’s t test; n = 4, and each biological repeat contains about 250 embryos. (B) Free IAA contents in ovules from the wild-type (Col-0), ins-1, ins-1 trp3-1+/−, and trp3-1 siliques. “ins-1 normal” refers to the ins-1 ovules with normal morphology, “ins-1 abnormal” refers to the ins-1 ovules containing defective embryos, and “ins-1 trp3-1” refers to the abnormal ovules from ins-1 trp3-1+/− siliques that contain defective embryos of ins, ins trp3+/−, and ins trp3-1−/− together. Data are represented as mean ± SEM; *P < 0.05 and **P < 0.01 by Student’s t test; n = 5. (C) Expression patterns of INS and TSA proteins during embryogenesis. 1C, 2C, 4C, 8C, and 16C represent 1-, 2-, 4-, 8-, and 16-cell stages, respectively. (D–M) DR5 activities shown by GFP signals in wild-type embryos (D and E), defective ins-1 embryos (F and G), trp3-1 embryos (H and I), normal ins-1 embryos (J and K), and ins-1 trp3-1 embryos (L and M). (Scale bars, 15 μm in C–M.)

4824 | www.pnas.org/cgi/doi/10.1073/pnas.1503998112

Wang et al.

the Trp-independent IAA biosynthetic pathway provides an important auxin source for the development of early embryogenesis. INS and YUC Coordinately Regulate Embryogenesis. To further evaluate the roles of Trp-dependent and Trp-independent auxin biosynthetic pathways in embryonic development, we carried out genetic analyses using ins-1 and the yuc4 yuc6 double mutant (yuc4/6). Compared with the wild type (Fig. 5A), ins-1 exhibited arrested embryos (Fig. 5 B and C) and embryos defective in apical–basal polarity (Fig. 5 D–F), whereas yuc4/6 embryos displayed mild defects at the basal embryo pole (Fig. 5G). The ins-1 yuc4/6 triple mutant displayed an additive negative effect on embryogenesis and IAA biosynthesis (Fig. 5 H–N; Fig. S7 F and G). It should be pointed out that both ins-1 and yuc4/6 formed more branches than wild type, and the ins-1 yuc4/6 triple mutant displayed an additive promotion on shoot branching (Fig. S9). Together, it is suggested that INS and YUC4/6 cooperatively regulate embryo development and apical dominance through parallel and independent pathways.

Fig. 5. INS and YUCs act in parallel IAA biosynthetic pathways. (A–L) Morphologies of wild-type (Col-0) (A), ins-1 (B–F), yuc4 yuc6 (yuc4/6) (G), and ins-1 yuc4/6 (H–L) embryos at the globular stage. Arrows indicate aberrant cell divisions and brackets mark aberrant basal cell regions. (M) Percentage of defective embryos in Col-0, ins-1, yuc4/6, and ins-1 yuc4/6 at the globular stage. Data are represented as mean ± SEM; different letters indicate significant difference by Tukey’s multiple comparison test (P < 0.05); n = 4, and each biological repeat contains about 250 embryos. (N) Free IAA levels in 12-d-old Col-0, ins-1, yuc4/6, and ins-1 yuc4/6 seedlings. gfw−1, per gram fresh weight. Data are represented as mean ± SEM; different letters indicate significant difference by Tukey’s multiple comparison test (P < 0.05); n = 5.

Wang et al.

Fig. 6. A model of the Trp-dependent and Trp-independent IAA biosynthetic pathways. Solid arrows refer to pathways with identified enzymes and dashed arrows to undefined ones. ANT, anthranilate; AS, anthranilate synthase; CHA, chorismic acid; IAAld, indole-3-acetaldehyde; IAM, indole3-acetamide; IAN, indole-3-acetonitrile; IAOx, indole-3-acetaldoxime; IGP, indole-3-glycerol phosphate; IGs, indole glucosinolates; IGS, indole-3-glycerol phosphate synthase; IPyA, indole-3-pyruvic acid; PAT, phosphoribosylanthranilate transferase; TAA1, tryptophan aminotransferase of Arabidopsis1; TAM, tryptamine; TSA, tryptophan synthase α; TSB, tryptophan synthase β; YUC, YUCCA.

biosynthesis takes place in chloroplasts, where TSA and TSB have been suggested to form a complex and convert IGP to Trp (10, 23, 24). The first step of Trp biosynthesis is catalyzed by anthranilate synthase (AS), which is a rate-limiting enzyme and subject to feedback inhibition by Trp (31). In the Trp-dependent pathway, IAA is biosynthesized through four proposed routes named after their key intermediates from the common precursor tryptophan, whereas, in the Trp-independent pathway, INS may mediate indole production in cytosol and initiate the Trp-independent IAA biosynthesis that plays an important role for apical– basal pattern formation during early embryogenesis. These two pathways coordinately regulate embryogenesis and apical dominance in Arabidopsis. The 13 C 6-anthranilic acid feeding experiment and genetic analysis between trp3-1 and sur2 indicate that the activity of Trpdependent IAA biosynthesis is impaired in trp3-1. However, free IAA contents increase in trp3-1 and trp2-1, probably due to impairment in Trp biosynthesis and release of the Trp feedback inhibition. Furthermore, the 13C6-anthranilic acid feeding experiment and measurements of IAA contents in seedlings and ovules indicate that the Trp-independent IAA biosynthesis is dramatically impaired in ins-1, which produces ∼18–22% defective embryos that are finally lethal during early embryogenesis. The ins-1 seedlings with normal morphology should be generated from healthy embryos. The ins-1 trp3-1 and ins-1 trp3-100 double mutants are completely embryo lethal, most likely due to an impairment in both Trp-independent and -dependent auxin biosynthesis. However, the ins-1 trp2-1 and ins-1 trp2-301 double mutants display normal embryogenesis, similar IAA levels, and PNAS | April 14, 2015 | vol. 112 | no. 15 | 4825

PLANT BIOLOGY

Discussion Although genetic and metabolic studies have clearly established the importance of the Trp-dependent IAA biosynthetic pathway, IAA has also been proposed to be synthesized from IGP and/or indole through a Trp-independent pathway, which remains to be substantiated from genetic, biochemical, and functional studies (9–11). In this study, we demonstrate that the cytosol-localized INS is a key component in the long-speculated Trp-independent IAA biosynthetic pathway and contributes significantly to the establishment of the apical–basal pattern. Based on our findings in this work, we propose an updated model of IAA biosynthetic pathways (Fig. 6). In Arabidopsis Trp

embryo-lethal mutants in the trp2 background may help to identify the key components in the Trp-independent IAA biosynthetic pathway. Further investigation of components and functions of this Trp-independent pathway will likely be an exciting area in the coming years.

comparable morphologies to trp2-1 and trp2-301. These observations may be explained by the high levels of indole in trp2 (11), which may diffuse from chloroplasts to the cytosol and compensate for the deficiency of ins-1. Similarly, in cytosol, INS converts IGP to indole, which may diffuse into chloroplasts and participate in Trp biosynthesis with a low efficiency. This explains why overexpression of INS can partially rescue trp3-1 and why the TSA null mutant can survive under weak light conditions (15). Our work has identified the source of auxin during early embryogenesis. In Arabidopsis, the early axis pattern is predominantly set up during early embryogenesis and provides the positional reference to postembryonic development (18). INS shows spatiotemporal expression patterns during embryogenesis and contributes to the establishment of an auxin gradient during embryogenesis. It is probable that local auxin biosynthesis through the Trp-independent and Trp-dependent pathways coordinately regulates the auxin gradient and orchestrates critical formative steps in the creation of apical–basal pattern during embryogenesis. How indole is converted into IAA is a challenge. It will be interesting to investigate whether indole-3-butyric acid participates in the Trp-independent pathway (32). A systematic analyses of metabolic profiles of wild type, ins-1, and trp3-1 treated with 13C6-anthranilic acid and a large-scale genetic screening for

ACKNOWLEDGMENTS. We thank the Arabidopsis Biological Resource Center for providing ins-1 (SALK_138654), trp3-1 (CS8331), trp3-100 (CS8332), trp2-1 (CS8327), trp5-1 (CS8558), tar2-1 wei8-1 (CS16414), and sur2 (SALK_028573) seeds; the Nottingham Arabidopsis Stock Centre for providing ins-2 (GT_5_101747) seeds; Dr. Yunde Zhao (University of California, San Diego) for providing yuc4/6 seeds; and Dr. Yuxin Hu (Institute of Botany, Chinese Academy of Sciences) for providing trp2-301 seeds. We also thank Dr. Weicai Yang and Dr. Hongju Li (Institute of Genetics and Developmental Biology, Chinese Academy of Sciences) for technical assistance in embryo observation and Dr. Steven M. Smith (The University of Western Australia) for critical reading of the manuscript. This work was supported by the National Natural Science Foundation of China (Grants 91117001, 91117016, and 30830009).

1. Wang R, Estelle M (2014) Diversity and specificity: Auxin perception and signaling through the TIR1/AFB pathway. Curr Opin Plant Biol 21:51–58. 2. Teale WD, Paponov IA, Palme K (2006) Auxin in action: Signalling, transport and the control of plant growth and development. Nat Rev Mol Cell Biol 7(11):847–859. 3. Kieffer M, Neve J, Kepinski S (2010) Defining auxin response contexts in plant development. Curr Opin Plant Biol 13(1):12–20. 4. Zhao Y (2012) Auxin biosynthesis: A simple two-step pathway converts tryptophan to indole-3-acetic acid in plants. Mol Plant 5(2):334–338. 5. Woodward AW, Bartel B (2005) Auxin: Regulation, action, and interaction. Ann Bot (Lond) 95(5):707–735. 6. Tao Y, et al. (2008) Rapid synthesis of auxin via a new tryptophan-dependent pathway is required for shade avoidance in plants. Cell 133(1):164–176. 7. Stepanova AN, et al. (2008) TAA1-mediated auxin biosynthesis is essential for hormone crosstalk and plant development. Cell 133(1):177–191. 8. Zhao Y, et al. (2001) A role for flavin monooxygenase-like enzymes in auxin biosynthesis. Science 291(5502):306–309. 9. Tivendale ND, Ross JJ, Cohen JD (2014) The shifting paradigms of auxin biosynthesis. Trends Plant Sci 19(1):44–51. 10. Ouyang J, Shao X, Li J (2000) Indole-3-glycerol phosphate, a branchpoint of indole3-acetic acid biosynthesis from the tryptophan biosynthetic pathway in Arabidopsis thaliana. Plant J 24(3):327–333. 11. Normanly J, Cohen JD, Fink GR (1993) Arabidopsis thaliana auxotrophs reveal a tryptophan-independent biosynthetic pathway for indole-3-acetic acid. Proc Natl Acad Sci USA 90(21):10355–10359. 12. Wright AD, et al. (1991) Indole-3-acetic acid biosynthesis in the mutant maize orange pericarp, a tryptophan auxotroph. Science 254(5034):998–1000. 13. Baldi BG, Maher BR, Slovin JP, Cohen JD (1991) Stable isotope labeling, in vivo, of D- and L-tryptophan pools in Lemna gibba and the low incorporation of label into indole-3-acetic acid. Plant Physiol 95(4):1203–1208. 14. Last RL, Fink GR (1988) Tryptophan-requiring mutants of the plant Arabidopsis thaliana. Science 240(4850):305–310. 15. Radwanski ER, Barczak AJ, Last RL (1996) Characterization of tryptophan synthase alpha subunit mutants of Arabidopsis thaliana. Mol Gen Genet 253(3):353–361. 16. Last RL, Bissinger PH, Mahoney DJ, Radwanski ER, Fink GR (1991) Tryptophan mutants in Arabidopsis: The consequences of duplicated tryptophan synthase beta genes. Plant Cell 3(4):345–358. 17. Zhang R, Wang B, Ouyang J, Li J, Wang Y (2008) Arabidopsis indole synthase, a homolog of tryptophan synthase alpha, is an enzyme involved in the Trp-independent indole-containing metabolite biosynthesis. J Integr Plant Biol 50(9):1070–1077. 18. Capron A, Chatfield S, Provart N, Berleth T (2009) Embryogenesis: Pattern formation from a single cell. Arabidopsis Book 7:e0126. 19. Lau S, Slane D, Herud O, Kong J, Jürgens G (2012) Early embryogenesis in flowering plants: Setting up the basic body pattern. Annu Rev Plant Biol 63:483–506.

20. Friml J, et al. (2003) Efflux-dependent auxin gradients establish the apical-basal axis of Arabidopsis. Nature 426(6963):147–153. 21. Blilou I, et al. (2005) The PIN auxin efflux facilitator network controls growth and patterning in Arabidopsis roots. Nature 433(7021):39–44. 22. Cheng Y, Dai X, Zhao Y (2007) Auxin synthesized by the YUCCA flavin monooxygenases is essential for embryogenesis and leaf formation in Arabidopsis. Plant Cell 19(8):2430–2439. 23. Miles EW (1991) Structural basis for catalysis by tryptophan synthase. Adv Enzymol Relat Areas Mol Biol 64:93–172. 24. Radwanski ER, Zhao J, Last RL (1995) Arabidopsis thaliana tryptophan synthase alpha: Gene cloning, expression, and subunit interaction. Mol Gen Genet 248(6):657–667. 25. Petersson SV, et al. (2009) An auxin gradient and maximum in the Arabidopsis root apex shown by high-resolution cell-specific analysis of IAA distribution and synthesis. Plant Cell 21(6):1659–1668. 26. Barlier I, et al. (2000) The SUR2 gene of Arabidopsis thaliana encodes the cytochrome P450 CYP83B1, a modulator of auxin homeostasis. Proc Natl Acad Sci USA 97(26): 14819–14824. 27. Hamann T, Benkova E, Bäurle I, Kientz M, Jürgens G (2002) The Arabidopsis BODENLOS gene encodes an auxin response protein inhibiting MONOPTEROS-mediated embryo patterning. Genes Dev 16(13):1610–1615. 28. Hardtke CS, Berleth T (1998) The Arabidopsis gene MONOPTEROS encodes a transcription factor mediating embryo axis formation and vascular development. EMBO J 17(5):1405–1411. 29. Mayer U, Buttner G, Jurgens G (1993) Apical-basal pattern formation in the Arabidopsis embryo: Studies on the role of the gnom gene. Development 117(1):149–162. 30. Szemenyei H, Hannon M, Long JA (2008) TOPLESS mediates auxin-dependent transcriptional repression during Arabidopsis embryogenesis. Science 319(5868): 1384–1386. 31. Li J, Last RL (1996) The Arabidopsis thaliana trp5 mutant has a feedback-resistant anthranilate synthase and elevated soluble tryptophan. Plant Physiol 110(1):51–59. 32. Strader LC, et al. (2011) Multiple facets of Arabidopsis seedling development require indole-3-butyric acid-derived auxin. Plant Cell 23(3):984–999. 33. Mou Z, He Y, Dai Y, Liu X, Li J (2000) Deficiency in fatty acid synthase leads to premature cell death and dramatic alterations in plant morphology. Plant Cell 12(3): 405–418. 34. Gray WM, Ostin A, Sandberg G, Romano CP, Estelle M (1998) High temperature promotes auxin-mediated hypocotyl elongation in Arabidopsis. Proc Natl Acad Sci USA 95(12):7197–7202. 35. Sun X, et al. (2014) An in-advance stable isotope labeling strategy for relative analysis of multiple acidic plant hormones in sub-milligram Arabidopsis thaliana seedling and a single seed. J Chromatogr A 1338:67–76. 36. Fu J, Chu J, Sun X, Wang J, Yan C (2012) Simple, rapid, and simultaneous assay of multiple carboxyl containing phytohormones in wounded tomatoes by UPLC-MS/MS using single SPE purification and isotope dilution. Anal Sci 28(11):1081–1087.

4826 | www.pnas.org/cgi/doi/10.1073/pnas.1503998112

Materials and Methods Plant growth, plant transformation, and high-temperature–induced hypocotyl elongation were performed as previously described (33, 34). Free IAA content was measured in seedlings, aerial parts, and ovules as described previously (35, 36), with minor modifications in ultraperformance liquid chromatography-tandem mass spectrometer (UPLC-MS/MS) conditions. Primers used in this study are listed in Table S1. Details are provided in SI Materials and Methods.

Wang et al.

Supporting Information Wang et al. 10.1073/pnas.1503998112 SI Materials and Methods Plant Materials and Growth Conditions. A. thaliana materials used

in this work include the wild-type ecotype Col-0, Landsberg erecta (Ler), and the following mutants or marker lines: ins-1 (SALK_138654), ins-2 (GT_5_101747), trp3-1 (1), trp3-100 (2), trp2-1 (3), trp2-301 (4), trp5-1 (5), wei8-1 (6), yuc4 yuc6 (yuc4/6) (7), sur2 (8), DR5:GUS (9), and DR5:GFP (10). Arabidopsis seeds were surface-sterilized, vernalized at 4 °C for 2–4 d, and germinated on MS medium containing 1.0% (wt/vol) sucrose and 0.6% agar (type M, Sigma-Aldrich) at 21–23 °C in a 16-h light/8-h dark cycle at an intensity of 60–80 μE·m−2·s−1. For morphological examination, 10-d-old seedlings were transferred to vermiculite saturated with 0.3 × B5 medium and grown under a 16-h light/8-h dark cycle at 21–23 °C as described previously (11). For root observation, seedlings were grown vertically on MS plates containing 2.0% (wt/vol) sucrose and 0.6% phytoagar or on MS plates supplemented with 50 μM Trp for 5 d. The sequences of all of the primers used in this work were listed in Table S1. Plasmid Construction and Plant Transformation. A 7.7-kb DNA fragment containing the 3-kb upstream sequence, entire INS coding region, and 3-kb downstream sequence was amplified by primers INSpro-INS-F and INSpro-INS-R with KpnI and SalI adaptors from the BAC clone T10P11. The amplified DNA fragment was inserted into the KpnI- and SalI-digested pCAMBIA1300 vector (CAMBIA) to generate the complementation construct INS:INS, which was transformed into the ins-1 DR5:GUS and ins-1 plants, respectively, by the Agrobacterium-mediated floral dip method (12). The T3 homozygous progeny of INS:INS/ins-1 DR5:GUS seedlings were used for GUS signal observation, and the T3 homozygous plants of INS:INS/ins-1 transgenic plants were collected for embryogenesis analysis and quantification of free IAA and Trp contents. To construct the 35S:TSA plasmid, a BamHI-KpnI fragment containing the TSA coding sequence (CDS) was amplified by primers TSA-OE-F and TSA-OE-R and then subcloned into the BamHI- and KpnI-digested pBin438 vector. To generate the cTP-INS fusion fragment, the cTP region of TSA was amplified by primers G-cTP-F1 and cTP-INS-R1, and the INS-coding region was amplified by primers cTP-INS-F2 and G-INS-R2. Then the PCR products were mixed and subjected to an amplification with primers G-cTP-F1 and G-INS-R2. To construct the 35S:ΔcTP-TSA and 35S:cTP-INS plasmids, the ΔcTP-TSA1 fragment amplified from Arabidopsis cDNA by primers G-ΔcTP-TSA-F and G-TSA-R and the cTP-INS fusion fragment were cloned into the Gateway entry vector pDONR222 (Invitrogen) and subsequently transferred to the binary vector pB7WG2D (13). To construct the 35S:INS-GFP, 35S:TSA-GFP, 35S:cTP-INSGFP, and 35S:ΔcTP-TSA-GFP plasmids, four BamHI-SalI fragments containing INS CDS, TSA CDS, cTP-INS, and ΔcTP-TSA1 were amplified using primer pairs INS-G-F + INS-G-R, TSAOE-F +TSA-G-R, cTP-INS-F+ cTP-INS-R, and ΔcTP-TSA-F + ΔcTP-TSA-R, respectively, and then subcloned into BamHI- and SalI-digested pBI221-GFP vector. Recombinant plasmids were subsequently digested with BamHI and EcoRI, and the insertion fragments were subcloned into BamHI- and EcoRI-digested pBI121 vector. To construct INS:INS-GFP-GUS and TSA:TSA-GFP-GUS plasmids, the INSpro:INS fragment was amplified from the BAC clone T10P11 using primers G-INSpro-F and G-INS no GA-R, and the TSApro:TSA fragment was amplified from Arabidopsis genomic DNA using primers G-TSApro-F and G-TSA no GA-R. Wang et al. www.pnas.org/cgi/content/short/1503998112

Then INSpro:INS and TSApro:TSA fragments were cloned into the Gateway entry vector pDONR222 (Invitrogen) and subsequently transferred to the binary vector pBGWFS7 (13). Gene Expression Analysis. Total RNA was prepared using a TRIzol kit according to the user manual (Invitrogen) and treated with TUBRO DNase (Ambion) before use. First-strand cDNA was synthesized using the oligo(dT) primer with the SSIII first-strand synthesis system (Invitrogen). Real-time PCR experiments were performed using gene-specific primers in a total volume of 20 μL with 1 μL of diluted cDNA, 0.5 μM gene-specific primers, and 10 μL EvaGreen supermix (BioRad) on a CFX 96 real-time system (BioRad) according to the manufacturer’s instructions. The Arabidopsis Ubiquitin5 (UBQ5) and EF1a genes were used as the internal controls. Hypocotyl Measurements. Seedlings were grown on MS plates at 20 °C and 28 °C for 9 d in continuous light as previously described (14). Hypocotyl length was measured using ImageJ software (rsbweb.nih.gov/ij/). Histological and Microscopy Analyses. Histochemical staining for GUS activities was performed as described (15). Briefly, whole seedlings or various tissues were immersed in the GUS staining solution, which contains 2 mM 5-bromo-4-chloro-3-inolylβ-D-glucuronic acid, 100 mM sodium phosphate (pH 7.2), 10 mM ethylene diamine tetraacetic acid, 0.5 mM potassium ferricyanide, 0.5 mM potassium ferrocyanide, and 0.1% Triton X-100. Plant materials were vacuumed briefly and incubated at 37 °C in dark from 30 min to overnight. The GUS stain solution was removed and replaced with 70% (vol/vol) ethanol to remove chlorophyll from the tissue. Arabidopsis roots were then cleared in the HCG solution (chloroacetaldehyde:water:glycerol = 8:3:1) for several minutes before observation. Finally, the GUS signals were observed under a stereomicroscope or microscope, and individual representative seedlings were photographed with a CCD camera. To observe the subcellular localization of INS-GFP, TSA-GFP, cTP-INS-GFP, and ΔcTP-TSA-GFP, leaves of the stable 35S:INS-GFP, 35S:TSA-GFP, 35S:cTP-INS-GFP, and 35S:ΔcTPTSA-GFP transgenic plants were collected for protoplast preparation as described (16). The GFP signals and chlorophyll autofluorescence were examined under a confocal microscope (FluoView FV1000; Olympus) at an excitation wavelength of 488 and 647 nm, respectively. To observe GFP signals in embryos, ovules were collected in 10% (vol/vol) glycerol solution, and embryos were squeezed out of the ovules by careful extrusion under a cover glass. The GFP signals were examined under a confocal microscope (FluoView FV1000; Olympus) at an excitation wavelength of 488 nm. To study the development of embryos, siliques were dissected with hypodermic needles and cleared in Herr’s solution containing lactic acid:chloral hydrate:phenol:clove oil:xylene (2:2:2:2:1, wt/wt) (17) and observed with a confocal microscope (FluoView 1000; Olympus). Free IAA Measurement in Seedlings and Aerial Parts. Seedlings of 12-d-old Col-0, ins-1, trp3-1, INS:INS/ins-1, 35S:cTP-INS/trp3-1, INS:INS/trp3-1, trp2-1, ins-1 trp2-1, yuc4 yuc6 (yuc4/6), ins-1 yuc4/6, Ler, and ins-2 grown on MS medium were collected and frozen in liquid nitrogen. In addition, aerial parts of 30-d-old Col-0, ins-1, trp2-301, ins-1 trp2-301, trp2-1, and ins-1 trp2-1 plants were used for free IAA measurement. Five or six independent replicates were prepared. For each replicate, ∼200 mg (fresh weight) 1 of 10

materials were homogenized under liquid nitrogen, weighted, and extracted at −20 °C for 24 h with 2 mL of cold methanol containing antioxidant and 2H2-IAA (internal standard, CDN Isotopes). Purification was performed with Oasis MAX solid-phase extract cartridge (150 mg/6 cc; Waters). Measurement of free IAA was carried out using a UPLC-MS/MS system consisting of a UPLC system (ACQUITY UPLC; Waters) and a triple quadruple tandem mass spectrometer (Quattro Premier XE; Waters) as previously described (18). Free IAA Measurement in Ovules. Ovules in the fourth and fifth

siliques of the Arabidopsis main stem were harvested from the funicle carefully with hypodermic needles and hybrid tweezers and collected in 2.0-mL safe-lock tubes with 200 μL of cold methanolcontaining antioxidant. The wild-type ovules, morphologically normal ovules of ins-1 (ins-1 normal), and trp3-1 ovules were light green, whereas the abnormal ovules from the ins-1 siliques (ins-1 abnormal) and from the ins-1 trp3-1+/− siliques (ins-1 trp3-1) were white. Five independent replicates were prepared. Each replicate contained about 80 ovules and 0.2 ng 13C6-IAA (internal standard, Cambridge Isotope Laboratories). Then samples were homogenized by ball mill (MM 400; Retsch GmbH) at a frequency of 28 Hz for 1 min using 3-mm tungsten carbide beads. After extraction at −20 °C for 12 h, the samples were centrifuged at 16,000 × g for 15 min, and the supernatants were collected and dried with nitrogen. The residues were derivatized with bromocholine bromide (BETA, TCI) as previously described (19). After evaporation of the reaction mixture, the residues were reconstituted and subjected to solid-phase extraction with Bond Elut CBA cartridges (1 mL, Agilent) for further enrichment. After evaporation, the residues were dissolved with 300 μL of 10% (vol/vol) acetonitrile containing 0.1% formic acid. The IAA derivative was analyzed using the UPLC-MS/MS system consisting of a UPLC (ACQUITY UPLC) and a triple quadruple tandem mass spectrometer (QTRAP 5500; AB SCIEX) under positive ion mode. Chromatographic separation was performed on an ACQUITY UPLC BEH C18 column (100 × 2.1 mm, 1.7 μm; Waters) with the column temperature set at 25 °C, and the flow rate was 0.4 mL/min. The linear gradient runs from 88 to 66% (vol/vol) A (solvent A, 0.1% formic acid aqueous; solvent B, acetonitrile) in 7 min and from 66–8% (vol/vol) A in the next 1 min and is re-equilibrated with the initial condition for 3 min. The optimized mass spectrometer parameters were set as follows: curtain gas 20 psi, collision gas 6 psi, ion spray voltage 5,450 V, temperature 600 °C. The declustering potential (93 V) and collision energy (42 V) were applied. Multiple reaction monitoring (MRM) mode was used for quantification and the selected MRM transitions were 261.0 > 130.0 for IAA derivative and 267.0 > 136.0 for 13C6-IAA derivative, respectively.

re-equilibrated with the initial condition for 2 min. The optimized mass spectrometer parameters were set as follows: curtain gas at 40 psi, collision gas at 6 psi, ion spray voltage at −4,300 V, and temperature at 550 °C. The declustering potential (−85 V) and collision energy (−13 V) were applied. MRM mode was used for quantification and the selected MRM transitions were 180.1 > 136.1 for 13C6-IAA and 176.1 > 132.0 for 2H2-IAA. Endogenous Trp Measurement. Seedlings of 7-d-old Col-0, ins-1,

trp3-1, trp2-1, 35S:TSA/trp3-1, 35S:cTP-INS/trp3-1, INS:INS/ins-1, and trp5-1 grown on MS medium were collected for Trp measurement. Frozen seedling samples were homogenized under liquid nitrogen. The fine powder of lyophilized tissues (6.0 mg) was transferred into a 4-mL glass vial. Then 1.5 mL of chloroform was added to the dry tissues. The sample was thoroughly vortexed and incubated for 45 min at 37 °C. After cooling down to room temperature, 1.5 mL of HPLC-grade water containing 3 μg 2H5-Trp (internal standard, Sigma-Aldrich) was added to the chloroform. Then the sample was vortexed, incubated for 45 min at 37 °C, and centrifuged at 3,000 × g for 30 min to separate the phases. Then 1.0 mL of aqueous phase was transferred into a 1.5-mL tube and lyophilized to dry. The dry residue was resuspended into 50 μL of pyridine containing 15.0 mg/mL methoxyamine– HCl and incubated for 1 h at 50 °C after a brief sonication. Metabolites were then directly derivatized by mixing 50 μL of N-methyl-N-(trimethylsilyl) trifluoroacetamide (MSTFA) with 1% trimethylchlorosilane (TMCS) (MSTFA + 1% TMCS, Fluka, 69478) for 1 h at 50 °C. The clear supernatant was transferred into a glass insert after centrifugation at 20,000 × g for 10 min, and 1 μL of supernatant was loaded on a gas chromatography-tandem mass spectrometry (GC-MS) system (Trace DSQ II; Thermo Fisher company) equipped with a 30-M DB 5MS column for Trp measurement. The oven temperature program was initially set at 150 °C for 1 min and then ramped to 240 °C at 20 °C/min, to 260 °C at 10 °C/min, and to 300 °C at 20 °C/min and and held for 5 min. The signal was collected using the SIM mode. The intensity of m/z 202 and 207, corresponding to Trp and 2H5-Trp (retention time 19.59 min), respectively, was used for endogenous Trp calculation. Genetic Analysis. Double and triple mutants were generated from

13 C6-IAA Measurement in Seedlings. Five independent replicates were collected for 13C6-IAA quantification. For each replicate, ∼400 mg (fresh weight) whole seedlings were homogenized under liquid nitrogen, weighted, and extracted for 24 h with 3 mL of cold methanol-containing antioxidant and 2.5 ng 2H2-IAA (internal standard, CDN Isotopes). Endogenous 13C6-IAA produced from 13 C6-anthranilic acid was measured similarly to that previously described (18) with some changes under UPLC-MS/MS conditions. The UPLC-MS/MS system consisted of a UPLC (ACQUITY UPLC ) and a tandem mass spectrometer (QTRAP 5500). The chromatographic separation was achieved on an ACQUITY UPLC BEH C18 column (50 × 2.1 mm, 1.7 μm; Waters) with the column temperature at 30 °C and a flow rate of 0.2 mL/min. The linear gradient runs from 90 to 60% (vol/vol) A (solvent A, 0.05% acetic acid aqueous; solvent B, methanol) in 4 min, 60–35% (vol/vol) A in the next 4 min, 35–2% (vol/vol) A in the following 1 min and is

the crosses of relevant homozygous single or double mutants and identified from the F2 progenies. Genotyping of the ins-1, ins-2, wei8-1, sur2, yuc4, yuc6, trp3-1, trp3-100, trp2-1, trp2-301, and trp5-1 mutants was performed by PCR using the primers in Table S1. In detail, the PCR products of the wild type and trp3-1 amplified by primers trp3-1-F ad trp3-1-R were digested with BslI. The PCR products of the wild type and trp3-100 amplified by primers trp3-100-F and trp3-100-R were digested with FokI. The PCR products of the wild type and trp2-1 amplified by primers of trp2-1-F and trp2-1-R were digested with BseRI. The PCR products of the wild-type and trp2-301 amplified by primers of trp2-1-F and trp2-1-R were digested with BglII. The PCR products of the wild-type and trp5-1 amplified by primers of trp5-1-F and trp5-1-R were digested with PstI. The ins-1 mutant was genotyped using primer pairs 654RP + LBb1 and 326RP + 326LP. The ins-2 mutant was genotyped using primer pairs 537G03RP + Ds5-4 and 537G03LP + 537G03RP. The wei8-1 mutant was genotyped using primer pairs 70560R1 + DWLB1 and 70560F8 + 70560R1. The sur2 mutant was genotyped using primer pairs 028573RP + LBb1 and 028573LP + 028573RP. The yuc4 mutant was genotyped using primer pairs YUC4-R + SMP32 and YUC4-F + YUC4-R. The yuc6 mutant was genotyped using primer pairs YUC6-R + LBb1 and YUC6-F + YUC6-R.

1. Last RL, Fink GR (1988) Tryptophan-requiring mutants of the plant Arabidopsis thaliana. Science 240(4850):305–310.

2. Radwanski ER, Barczak AJ, Last RL (1996) Characterization of tryptophan synthase alpha subunit mutants of Arabidopsis thaliana. Mol Gen Genet 253(3):353–361.

Wang et al. www.pnas.org/cgi/content/short/1503998112

2 of 10

3. Last RL, Bissinger PH, Mahoney DJ, Radwanski ER, Fink GR (1991) Tryptophan mutants in Arabidopsis: The consequences of duplicated tryptophan synthase beta genes. Plant Cell 3(4):345–358. 4. Jing Y, et al. (2009) Tryptophan deficiency affects organ growth by retarding cell expansion in Arabidopsis. Plant J 57(3):511–521. 5. Li J, Last RL (1996) The Arabidopsis thaliana trp5 mutant has a feedback-resistant anthranilate synthase and elevated soluble tryptophan. Plant Physiol 110(1):51–59. 6. Stepanova AN, et al. (2008) TAA1-mediated auxin biosynthesis is essential for hormone crosstalk and plant development. Cell 133(1):177–191. 7. Cheng Y, Dai X, Zhao Y (2006) Auxin biosynthesis by the YUCCA flavin monooxygenases controls the formation of floral organs and vascular tissues in Arabidopsis. Genes Dev 20(13):1790–1799. 8. Barlier I, et al. (2000) The SUR2 gene of Arabidopsis thaliana encodes the cytochrome P450 CYP83B1, a modulator of auxin homeostasis. Proc Natl Acad Sci USA 97(26):14819–14824. 9. Ulmasov T, Murfett J, Hagen G, Guilfoyle TJ (1997) Aux/IAA proteins repress expression of reporter genes containing natural and highly active synthetic auxin response elements. Plant Cell 9(11):1963–1971. 10. Benková E, et al. (2003) Local, efflux-dependent auxin gradients as a common module for plant organ formation. Cell 115(5):591–602. 11. Mou Z, He Y, Dai Y, Liu X, Li J (2000) Deficiency in fatty acid synthase leads to premature cell death and dramatic alterations in plant morphology. Plant Cell 12(3):405–418.

12. Clough SJ, Bent AF (1998) Floral dip: A simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. Plant J 16(6):735–743. 13. Karimi M, Inzé D, Depicker A (2002) GATEWAY vectors for Agrobacterium-mediated plant transformation. Trends Plant Sci 7(5):193–195. 14. Gray WM, Ostin A, Sandberg G, Romano CP, Estelle M (1998) High temperature promotes auxin-mediated hypocotyl elongation in Arabidopsis. Proc Natl Acad Sci USA 95(12):7197–7202. 15. Jefferson RA, Kavanagh TA, Bevan MW (1987) GUS fusions: Beta-glucuronidase as a sensitive and versatile gene fusion marker in higher plants. EMBO J 6(13):3901–3907. 16. Sheen J (2001) Signal transduction in maize and Arabidopsis mesophyll protoplasts. Plant Physiol 127(4):1466–1475. 17. Herr JM (1971) A new clearing-squash technique for the study of ovule development in angiosperms. Am J Bot 58(8):785–790. 18. Fu J, Chu J, Sun X, Wang J, Yan C (2012) Simple, rapid, and simultaneous assay of multiple carboxyl containing phytohormones in wounded tomatoes by UPLC-MS/MS using single SPE purification and isotope dilution. Anal Sci 28(11):1081–1087. 19. Sun X, et al. (2014) An in-advance stable isotope labeling strategy for relative analysis of multiple acidic plant hormones in sub-milligram Arabidopsis thaliana seedling and a single seed. J Chromatogr A 1338:67–76.

Fig. S1. Identification of ins and trp3 mutants. (A) Domains of TSA and INS. INS lacks the chloroplast transit peptides (cTP). (B) Phenotypes of wild-type (Col-0), trp3, and trp2 plants. The trp3 and trp2 mutants display typical Trp auxotrophic phenotypes compared with the wild type, for example, growing more slowly, exhibiting light-conditional phenotypes, and dwarf. The trp3-100 mutant is a weak allele of TSA and trp2-301 is a weak allele of TSB1. (Scale bar, 5 cm.) (C) Schematic diagram of the T-DNA insertion sites of ins-1 and ins-2. The ins-1 (SALK_138654) and ins-2 (GT_5_101747) lines carry T-DNA insertions in intron 6 and in exon 4, respectively. (D) Relative expression levels of INS and TSA. INS is knocked out in both ins-1 and ins-2, whereas TSA is knocked down in trp3-1. Data are represented as mean ± SEM; **P < 0.01 by Student’s t test; n = 3. (E) Free IAA contents of 12-d-old wild-type (Ler) and ins-2 seedlings. gfw−1, per gram fresh weight. Data are represented as mean ± SEM; *P < 0.05 by Student’s t test; n = 4.

Wang et al. www.pnas.org/cgi/content/short/1503998112

3 of 10

Fig. S2. Morphological comparison among wild-type (Col-0), ins-1, trp3-1, INS:INS/trp3-1, 35S:cTP-INS/trp3-1, and 35S:TSA/trp3-1 plants. (A) Phenotypes of 40-d-old plants. (Scale bar, 2 cm.) (B–D) Quantitative analyses of plant height (B), plant fresh weight (C), and stem diameter (D) in 40-d-old plants. Data are represented as mean ± SEM; **P < 0.01 by Student’s t test; n = 20.

Fig. S3. Observations of DR5 activities and effects of Trp treatment on wild-type, ins-1, and trp3-1 seedlings. (A) The DR5 activities visualized by GUS signals in 5-d-old wild-type (Col-0), ins-1, INS:INS/ins-1, 35S:ΔcTP-TSA/ins-1, 35S:TSA/ins-1, and 35S:cTP-INS/ins-1 seedlings. (Insets) Enlarged shoot tips. Overexpression of INS and ΔcTP-TSA rescued the DR5 activity of ins-1. (B and C) Expression patterns of DR5:GUS reporters (B) and morphologies (C) of 5-d-old wild-type (Col-0), ins-1, and trp3-1 seedlings grown with or without 50 μM Trp supplement. The trp3-1 seedlings displayed short and curved roots with disordered expression patterns of DR5:GUS. (D) Free IAA contents in 12-d-old Col-0, ins-1, and trp3-1 seedlings grown with or without 50 μM Trp supplement. gfw−1, per gram fresh weight. Data are represented as mean ± SEM; n = 5. (Scale bars, 1 mm in A, 0.1 mm in B, and 0.2 cm in C.)

Wang et al. www.pnas.org/cgi/content/short/1503998112

4 of 10

Fig. S4. Plant materials of 13C6-anthranilic acid feeding experiment and phenotypic analysis of the ins-1 sur2 and trp3-1 sur2 double mutants. (A) Schematic diagrams of 13C6-anthranilic acid (13C6-ANT) feeding experiment. The 11-d-old wild-type (Col-0), ins-1, and trp3-1 seedlings were transferred to liquid MS media containing 64 μM 13C6-ANT and cultured for 24 h at 28 °C. In a parallel test, the MS medium contained 64 μM 13C6-ANT and 100 μM unlabeled Trp, which should increase the Trp pool size and decrease the biosynthesis of 13C6-IAA only through the Trp-dependent pathway. (B) Morphologies of 11-d-old Col-0, ins-1, and trp3-1 seedlings used for 13C6-ANT feeding experiment. The trp3-1 mutant formed small and yellow true leaves. (C) Phenotypes of 30-d-old wild-type (Col-0), ins-1, sur2, ins-1 sur2, trp3-1, and trp3-1 sur2 plants. The phenotypes of sur2 were suppressed by trp3-1, but not by ins-1. (Scale bars, 0.5 cm in B and C.)

Wang et al. www.pnas.org/cgi/content/short/1503998112

5 of 10

Fig. S5. Embryogenesis of ins-2 and INS:INS/ins transgenic plants. (A) Embryos of wild-type (Ler) at the 1-, 4-, 8- or 16-cell, globular (g), and heart (h) stages. HY, hypophysis; LSC, lens-shaped cell; and QC, quiescent center. (B–L) Defective embryos of ins-2 arrested with correct body plans (B) or displayed abnormal cell division patterns (C–L) when normal sibling embryos reached the heart stage. The root meristems (G–L) and cotyledon primordia (K and L) were aberrant. Arrows indicate abnormal cell plates at wrong positions or in wrong orientations, and brackets mark aberrant basal cell regions. (M) Statistical analysis of embryonic defects in Ler and ins-2. (N) Relative expression levels of INS in wild-type (Col-0), ins-1, and INS:INS/ins-1 plants. Data are represented as mean ± SEM; **P < 0.01 by Student’s t test; n = 3. (O) Statistical analysis of embryonic defects in Col-0, ins-1, and INS:INS/ins-1 plants. In M and O, the percentage of defective embryos, arrested embryos with correct body plans, and embryos with abnormal cell division planes (ACD) are analyzed. Data are represented as mean ± SEM; **P < 0.01 by Student’s t test; n = 4, and each biological repeat contains about 200 embryos.

Wang et al. www.pnas.org/cgi/content/short/1503998112

6 of 10

Fig. S6. Embryogenesis of the ins-1 trp3-100 double mutant. (A–O) When normal sibling embryos reached the heart stage, defective embryos from ins-1 trp3100+/− siliques arrested at the 1- (A), 4- (B), 8- (C) or 16-cell (D) stage with regular cell division patterns or displayed abnormal cell division planes (E–O). For example, defective embryos formed filamentous structure without specification of a proembryo (E), possessed an apical cell with transverse division (F and G), underwent incomplete transverse division (H), exhibited disordered polarity at the eight-cell stage (I), formed additional layers of cells in the proembryo (K), and lacked root meristem and cotyledon primordia (K–O). Arrows indicate abnormal cell plates at wrong positions or in wrong orientations, and brackets mark aberrant basal cell regions.

Wang et al. www.pnas.org/cgi/content/short/1503998112

7 of 10

Fig. S7. Statistical analyses of embryonic defects in Col-0, ins-1, trp3, ins-1 trp3+/−, trp2, ins-1 trp2, yuc4/6, and ins-1 yuc4/6 plants. (A and B) Percentages of arrested embryos (A) and ACD embryos (embryos with abnormal cell division planes) (B) in the wild-type (Col-0) and mutants at the globular stage. (C–E) Percentages of arrested embryos (C), ACD embryos (D), and defective embryos (E) in the wild-type (Col-0) and mutants at the heart stage. (A–E) Data are represented as mean ± SEM; **P < 0.01 by Student’s t test; n = 4, and each biological repeat contains about 250 embryos. (F and G) Percentages of arrested embryos (F) and ACD embryos (G) in wild type (Col-0), ins-1, yuc4/6, and ins-1 yuc4/6 at the globular stage. Data are represented as mean ± SEM; different letters indicate significant difference by Tukey’s multiple comparison test (P < 0.05); n = 4, and each biological repeat contains about 250 embryos.

Fig. S8. Comparisons of IAA contents and morphologies among ins-1, trp2, and their double mutants. (A) Free IAA content in 30-d-old Col-0 (n = 5), ins-1 (n = 6), trp2-301 (n = 5), ins-1 trp2-301 (n = 6), trp2-1 (n = 6), and ins-1 trp2-1 (n = 6) plants. gfw−1, per gram fresh weight; ns, no significant difference. Data are represented as mean ± SEM; *P < 0.05 and **P < 0.01 by Student’s t test. (B) Phenotypes of 4-wk-old wild-type (Col-0), ins-1, trp2-301, ins-1 trp2-301, trp2-1, and ins-1 trp2-1 plants. (Scale bar, 1 cm.)

Wang et al. www.pnas.org/cgi/content/short/1503998112

8 of 10

Fig. S9. INS and YUCs cooperatively regulate shoot branching. (A) Phenotypes of 7-wk-old Col-0, ins-1, yuc4/6, and ins-1 yuc4/6 plants. (Scale bar, 4 cm.) (B) Quantitative analysis of branch number per plant. Data are represented as mean ± SEM; **P < 0.01 by Student’s t test; n = 30.

Wang et al. www.pnas.org/cgi/content/short/1503998112

9 of 10

Table S1. Primers used in this study Primer name

Sequence

Primers for plasmid construction INSpro-INS-F 5′-aaggtaccATGTTGTTGTTAATAGTGGACCAATT-3′ INSpro-INS-R 5′-aagtcgacTCCAATAGCTTCGACTTCGACTT-3′ TSA-OE-F 5′-ggatccATGGCGATTGCTTTCAAA-3′ TSA-OE-R 5′-atggtaccTCAAAGAAGAGCAGATTTAAGA-3′ INS-G-F 5′-ggatccATGGATCTTCTCAAGACTCCTT-3′ INS-G-R 5′-gtcgacAGAGACAAGAGCAGACTTCAAAG-3′ TSA-G-R 5′-gtcgacAAGAAGAGCAGATTTAAGAGACTTG-3′ cTP-INS-F 5′-ttcggatccATGGCGATTGCTTTCAAA-3′ cTP-INS-R 5′-tttgtcgacAGAGACAAGAGCAGACTTC-3′ ΔcTP-TSA-F 5′-ttaggatccATGGCTTCTCTCTCCACC-3′ ΔcTP-TSA-R 5′-tccgtcgacAAGAAGAGCAGATTTAAGAGA-3′ G-cTP-F1 5′-GGGGACAAGTTTGTACAAAAAAGCAGGCTTAATGGCGATTGCTTTCAAA-3′ cTP-INS-R1 5′-GAGGAAGGAGTCTTGAGAAGATCCATGGGAGTGAATCTCTTGAAAGAAAGCGATG-3′ cTP-INS-F2 5′-CATCGCTTTCTTTCAAGAGATTCACTCCCATGGATCTTCTCAAGACTCCTTCCTC-3′ G-INS-R2 5′-GGGGACCACTTTGTACAAGAAAGCTGGGTATCAAGAGACAAGAGCAGA-3′ G-ΔcTP-TSA-F 5′-GGGGACAAGTTTGTACAAAAAAGCAGGCTTAATGGCTTCTCT CTCCACC-3′ G-TSA-R 5′-GGGGACCACTTTGTACAAGAAAGCTGGGTATCAAAGAAGAGCAG ATTT-3′ G-INSpro-F 5′-GGGGACAAGTTTGTACAAAAAAGCAGGCTTATACATACGGCTT CGACTC-3′ G-INS no GA-R 5′-GGGGACCACTTTGTACAAGAAAGCTGGGTAAAGAGACAAGAGCAGACT-3′ G-TSApro-F 5′-GGGGACAAGTTTGTACAAAAAAGCAGGCTTATCAGGTTTGTCA AATACC-3′ G-TSA no GA-R 5′-GGGGACCACTTTGTACAAGAAAGCTGGGTAAAAGAAGAGCAGATTTAA-3′ Primers for genotyping 654RP 5′-CCTTGCTCAGGTGATTCAGAC-3′ 326LP 5′-CTGCGACGAGGAGTAGAGAAC-3′ 326RP 5′-TTCGGAACCCGAAAATCTAC-3′ 537G03LP 5′-GACTCGCCCAAGATCTTAACC-3′ 537G03RP 5′-TGATCCATTAGCTGATGGTCC-3′ YUC4-F 5′-TTGTCAAACCGAGGCGTACC-3′ YUC4-R 5′-ACGACCATATGAGGCAGAGC-3′ YUC6-F 5′-TCTGCAACTTCGGTGCTCAG-3′ YUC6-R 5′-TTTAACGAGAACATCTTAAAGCGCTG-3′ 70560F8 5′-CATCAGAGAGACGGTGGTGAAC-3′ 70560R1 5′-GCTTTTAATGAGCTTCATGTTGG-3′ 028573LP 5′-TTTCGTGGTTCTCTCTTTTCG-3′ 028573RP 5′-GTGGTATGGGCCATGACTTAC-3′ trp3-1-F 5′-ATGTGTGTTTGTGTCCCCTGTCA-3′ trp3-1-R 5′-CTCCCCATCCAGCTATCTGTTG-3′ trp3-100-F 5′-GAATACTCTTGGCTGGTAGATTTCA-3′ trp3-100-R 5′-TGTGGAACAACCTGCACCAA-3′ trp2-1-F 5′-CCTGGAGTCGGACCCGAGCA-3′ trp2-1-R 5′-TGTTGGAAACGTACTGACCAGAGAG-3′ trp5-1-F 5′-ACGTTTTGAGCGGCGAACATCTGCA-3′ trp5-1-R 5′-TGCTTTACTTTGGTGAGAATTTCTG-3′ LBb1 5′-GCGTGGACCGCTTGCTGCAACT-3′ Ds5-4 5′-TACGATAACGGTCGGTACGG-3′ SMP32 5′-TACGAATAAGAGCGTCCATTTTAGAGTGA-3′ DWLB1 5′-CATACTCATTGCTGATCCATGTAGATTTCC-3′ Primers for qPCR Q-UBQ5-F 5′-GCTTCATCTCGTCCTCCGTC-3′ Q-UBQ5-R 5′-AGAACAGCGAGCTTAACCTTCTTAT-3′ Q-EF1a-F 5′-TGAGCACGCTCTTCTTGCTTTCA-3′ Q-EF1a-R 5′-GGTGGTGGCATCCATCTTGTTACA-3′ Q-INS-F 5′-ACTGCTCACAACACCCACAACT-3′ Q-INS-R 5′-AACCCCTACTGAACTTACGAGATAGAT-3′ Q-TSA-F 5′-CATTTGCAGACAACATCTTTAGATT-3′ Q-TSA-R 5′-GTGTGGCTTGAGTCACAGTAACTTATT-3′

Wang et al. www.pnas.org/cgi/content/short/1503998112

10 of 10