Indole-3-Acetic Acid Is Synthesized by the Endophyte Cyanodermella asteris via a Tryptophan-Dependent and -Independent Way and Mediates the Interaction with a Non-Host Plant

The plant hormone indole-3-acetic acid (IAA) is one of the main signals playing a role in the communication between host and endophytes. Endophytes can synthesize IAA de novo to influence the IAA homeostasis in plants. Although much is known about IAA biosynthesis in microorganisms, there is still less known about the pathway by which IAA is synthesized in fungal endophytes. The aim of this study is to examine a possible IAA biosynthesis pathway in Cyanodermella asteris. In vitro cultures of C. asteris were incubated with the IAA precursors tryptophan (Trp) and indole, as well as possible intermediates, and they were additionally treated with IAA biosynthesis inhibitors (2-mercaptobenzimidazole and yucasin DF) to elucidate possible IAA biosynthesis pathways. It was shown that (a) C. asteris synthesized IAA without adding precursors; (b) indole-3-acetonitrile (IAN), indole-3-acetamide (IAM), and indole-3-acetaldehyde (IAD) increased IAA biosynthesis; and (c) C. asteris synthesized IAA also by a Trp-independent pathway. Together with the genome information of C. asteris, the possible IAA biosynthesis pathways found can improve the understanding of IAA biosynthesis in fungal endophytes. The uptake of fungal IAA into Arabidopsis thaliana is necessary for the induction of lateral roots and other fungus-related growth phenotypes, since the application of the influx inhibitor 2-naphthoxyacetic acid (NOA) but not the efflux inhibitor N-1-naphtylphthalamic acid (NPA) were altering these parameters. In addition, the root phenotype of the mutation in an influx carrier, aux1, was partially rescued by C. asteris.


Introduction
The plant hormone indole-3-acetic acid (IAA) plays a role in the communication between host plant and microbes, including plant-associated microorganisms [1][2][3] and endophytes [3,4], but also plant pathogens [5][6][7]. IAA together with other phytohormones is responsible for plant growth and development. It is involved in nearly each plant process such as cell growth, root initiation, tropism, fruit ripening, or senescence [8]. If plant-associated microorganism can regulate the plant IAA levels, they are able to influence the plant metabolism to their benefit, either as pathogens or as beneficial microorganisms. Then, microbial IAA acts as a mediator between plants and microorganisms and plays a key role in the communication between both partners. Microorganisms alters plant auxin levels and can therefore influence the plant development.
Cyanodermella asteris is an endophyte from the inflorescence axis of Aster tataricus [43], which is a perennial plant from the Northern part of Asia. Roots and rhizomes of A. tataricus 2. Results 2.1. The Endophyte C. asteris Induces a Root Phenotype on the Non-Host Plant A. thaliana that Is Reminiscent of High Auxin C. asteris induced a root phenotype with a massive lateral root and a reduced main root growth on the non-host plant A. thaliana that is reminiscent of auxin application (Figure 2a,f). The induction of the auxin response in root tips and lateral root primordia was also detected in a DR5::GUS line of A. thaliana (Figure 2b-e,g-i): Co-cultivated plants showed an increased amount of lateral root primordia that were additionally more intensely stained compared to controls.
Since the co-cultivation studies of C. asteris with A. thaliana showed an increased lateral root growth indicating a change in the plant IAA level, the possibility of an IAA biosynthesis by C. asteris was analyzed.

The Fungal Genome of C. asteris Reveals Several Putative Genes Involved in IAA Biosynthesis
Different candidate genes for IAA biosynthesis enzymes could be identified in the automated annotated genome sequence of C. asteris [58] (Table 1). Following the candidate genes, a TAM pathway would be possible, since all necessary genes were found in the genome of C. asteris. Three genes of the IAOx-IAN pathway were found, but two genes were missing for the IAOx dehydratase/hydrolase (IAOx into IAN) and the nitrile hydratase (IAN into IAM). Only one gene was found in the annotation for the IAM pathway, where the gene for the Trp monooxygenase in the first step (Trp into IAM) was missing. Genes for the conversion of Trp into IPyA were not found in the genome annotation of C. asteris, indicating that this pathway might not be used for IAA biosynthesis. In addition to the Trp-dependent pathways, there were also putative genes for the direct conversion of indole into IAA. However, since this is information from the genome and not transcriptome, it cannot be concluded that the genes detected would be indeed expressed.

Is Reminiscent of High Auxin
C. asteris induced a root phenotype with a massive lateral root and a reduced main root growth on the non-host plant A. thaliana that is reminiscent of auxin application (Figure 2a,f). The induction of the auxin response in root tips and lateral root primordia was also detected in a DR5::GUS line of A. thaliana (Figure 2b-e and 2g-i): Co-cultivated plants showed an increased amount of lateral root primordia that were additionally more intensely stained compared to controls.  root growth on the non-host plant A. thaliana that is reminiscent of auxin application (Fig-ure 2a,f). The induction of the auxin response in root tips and lateral root primordia was also detected in a DR5::GUS line of A. thaliana (Figure 2b-e and 2g-i): Co-cultivated plants showed an increased amount of lateral root primordia that were additionally more intensely stained compared to controls.  Table 1. Candidate genes of C. asteris involved in IAA biosynthesis. Shown are the candidate genes in C. asteris to the corresponding enzyme reaction, but certainly this is no evidence that they are expressed in C. asteris. The putative IAA biosynthesis pathways in C. asteris are highlighted in orange in Figure 1.

C. asteris Secretes the Synthesized IAA into the Liquid Medium
Several candidate gene of C. asteris indicated a de novo biosynthesis of IAA, including different Trp-dependent pathways such as via TAM or IAOx-IAN but also a Trp-independent pathway with indole as the IAA precursor. Therefore, fungal in vitro cultures of C. asteris were analyzed for their IAA production after 10 days of cultivation ( Figure 3). The specific mass spectrum of methylated IAA (Figure 3b), including the internal standard D 5 -IAA, was found in medium and hyphae of C. asteris. IAA was detected in all six experiments with higher IAA amounts in the medium compared to hyphae (Figure 3a, p < 0.05). C. asteris secreted most of the synthesized IAA into the medium (64% to 92%), since only 8% to 36% of the IAA was retained in the hyphae. < 0.05). C. asteris secreted most of the synthesized IAA into the medium (64% to 92%), since only 8% to 36% of the IAA was retained in the hyphae.  Table S1). Significant differences between medium and hyphae are labeled with an asterisk. Sample size n is given in the upper part below the asterisks. (b) Extracted ion chromatogram of IAA (m/z 130) from medium and hyphae samples of C. asteris, including mass spectra (MS) at the specific retention times (red: MS data of IAA; blue: MS data of the internal standard D5-IAA).

IAA Biosynthesis in C. asteris
To elucidate the contribution of different possible pathways, C. asteris was incubated with different precursors and intermediates of IAA pathways ( Figure 1) to evaluate their possible contribution to the biosynthesis. C. asteris was cultivated with 1 mM Trp, 1 mM TAM, 100 µM IAN, 100 µM IAM, 100 µM IPyA, 100 µM IAD, and 1 mM indole to induce the IAA biosynthesis ( Figure 4). All precursors, except IPyA, increased the levels of IAA in C. asteris. IPyA is very unstable compared to other intermediates and was converted non-enzymatically to IAA (data not shown). In general, the direct precursors of IAA (IAN, IAM, and IAD) enhanced the biosynthesis very strongly in medium and hyphae (p < 0.05), whereas an intermediate earlier in the IAA biosynthesis, TAM, increased less strong and only in the hyphae. The early IAA biosynthesis precursors Trp and indole showed a small induction. Data were analyzed using an unpaired t-test with Welch correction (test statistics can be found in Table S1). Significant differences between medium and hyphae are labeled with an asterisk. Sample size n is given in the upper part below the asterisks. (b) Extracted ion chromatogram of IAA (m/z 130) from medium and hyphae samples of C. asteris, including mass spectra (MS) at the specific retention times (red: MS data of IAA; blue: MS data of the internal standard D 5 -IAA).

IAA Biosynthesis in C. asteris
To elucidate the contribution of different possible pathways, C. asteris was incubated with different precursors and intermediates of IAA pathways ( Figure 1) to evaluate their possible contribution to the biosynthesis. C. asteris was cultivated with 1 mM Trp, 1 mM TAM, 100 µM IAN, 100 µM IAM, 100 µM IPyA, 100 µM IAD, and 1 mM indole to induce the IAA biosynthesis ( Figure 4). All precursors, except IPyA, increased the levels of IAA in C. asteris. IPyA is very unstable compared to other intermediates and was converted non-enzymatically to IAA (data not shown). In general, the direct precursors of IAA (IAN, IAM, and IAD) enhanced the biosynthesis very strongly in medium and hyphae (p < 0.05), whereas an intermediate earlier in the IAA biosynthesis, TAM, increased less strong and only in the hyphae. The early IAA biosynthesis precursors Trp and indole showed a small induction.

The IAA Biosynthesis Inhibitor MBI Reveals a Trp-Independent Biosynthesis Pathway in C. asteris
In addition to the Trp-dependent IAA biosynthesis, we were also interested in the possibility of a Trp-independent biosynthetic pathway. Here, two different IAA biosynthesis inhibitors MBI and YDF were used to disrupt the pathways and the two main IAA precursors Trp and indole were added to enhance possible effects ( Figure 5). . IAA levels in medium and hyphae of C. asteris after feeding with different IAA precursors. C. asteris was cultivated in MEAlow for seven days, before IAA precursors (Trp, TAM, IAN, IAM, IPyA, IAD, or indole) were added to the culture. IAA levels were detected after 10 days of cultivation (3 days with precursors). The IAA level in (a) medium and (b) hyphae of C. asteris is correlated to the respective control (without any IAA precursor), which is set to 1. Small letters above the box plots show the statistical significance; test statistics can be found in Table S2. Sample size n is given below the x-axis.

The IAA Biosynthesis Inhibitor MBI Reveals a Trp-Independent Biosynthesis Pathway in C. asteris
In addition to the Trp-dependent IAA biosynthesis, we were also interested in the possibility of a Trp-independent biosynthetic pathway. Here, two different IAA biosynthesis inhibitors MBI and YDF were used to disrupt the pathways and the two main IAA precursors Trp and indole were added to enhance possible effects ( Figure 5).
The IAA biosynthesis inhibitor MBI increased the IAA level 4.2-fold in C. asteris cultures with indole ( Figure 5a, p < 0.001). This was interpreted as the possibility to change to an alternative pathway from indole when the Trp-dependent one was blocked. The controls without any IAA precursor and with Trp did not change their IAA levels in the presence of MBI. Without any biosynthesis inhibitor, similar IAA levels were detected for Trp and indole as substrate in the cultures, and they were increased to the 4.4-to 5.5-fold compared to controls. These differences were only seen in the medium, not in hyphae ( Figure 5c). Only Trp increased the IAA content in the hyphae, no matter whether MBI was present or not.
The addition of YDF, a second IAA biosynthesis inhibitor, did not change the IAA pattern as MBI did (Figure 5b). Controls without any IAA precursor and the cultures incubated with indole showed no significant differences between cultures with and without YDF. Only the cultures where Trp was added increased threefold their IAA levels, when YDF was present. Similar to MBI, the hyphae did not change their IAA levels when YDF was present (Figure 5d). This experiment showed evidence for a Trp-independent pathway to synthesize IAA from the precursor indole in C. asteris. . IAA levels in medium and hyphae of C. asteris after feeding with different IAA precursors. C. asteris was cultivated in MEAlow for seven days, before IAA precursors (Trp, TAM, IAN, IAM, IPyA, IAD, or indole) were added to the culture. IAA levels were detected after 10 days of cultivation (3 days with precursors). The IAA level in (a) medium and (b) hyphae of C. asteris is correlated to the respective control (without any IAA precursor), which is set to 1. Small letters above the box plots show the statistical significance; test statistics can be found in Table S2. Sample size n is given below the x-axis.
The IAA biosynthesis inhibitor MBI increased the IAA level 4.2-fold in C. asteris cultures with indole ( Figure 5a, p < 0.001). This was interpreted as the possibility to change to an alternative pathway from indole when the Trp-dependent one was blocked. The controls without any IAA precursor and with Trp did not change their IAA levels in the presence of MBI. Without any biosynthesis inhibitor, similar IAA levels were detected for Trp and indole as substrate in the cultures, and they were increased to the 4.4-to 5.5-fold compared to controls. These differences were only seen in the medium, not in hyphae ( Figure 5c). Only Trp increased the IAA content in the hyphae, no matter whether MBI was present or not.
The addition of YDF, a second IAA biosynthesis inhibitor, did not change the IAA pattern as MBI did ( Figure 5b). Controls without any IAA precursor and the cultures incubated with indole showed no significant differences between cultures with and without YDF. Only the cultures where Trp was added increased threefold their IAA levels, when YDF was present. Similar to MBI, the hyphae did not change their IAA levels when YDF was present (Figure 5d). This experiment showed evidence for a Trp-independent pathway to synthesize IAA from the precursor indole in C. asteris.

Auxin Synthesis by C. asteris and Uptake into the Host Plant Are Important for the Root Phenotype
The fungus C. asteris was co-cultivated with A. thaliana on medium containing two different IAA transport inhibitors: NOA and NPA ( Figure 6). A. thaliana control plants developed a very strong phenotype in co-cultivation with C. asteris, including reduced main root length, massive lateral root growth, increased root and leaf biomass ( Figure 2 and Figure S1). The addition of 10 µM NOA to the medium changed the phenotype of the A. thaliana plants: shorter main roots ( Figure S1b Table S3 for test statistics).

Auxin Synthesis by C. asteris and Uptake Into the Host Plant Are Important for the Root Phenotype
The fungus C. asteris was co-cultivated with A. thaliana on medium containing two different IAA transport inhibitors: NOA and NPA ( Figure 6). A. thaliana control plants developed a very strong phenotype in co-cultivation with C. asteris, including reduced main root length, massive lateral root growth, increased root and leaf biomass ( Figure 2 and Figure S1). The addition of 10 µM NOA to the medium changed the phenotype of the A. thaliana plants: shorter main roots ( Figure S1b), agravitropic root growth, and a higher root biomass (Figure 6a), resulting from the small, compact phenotype ( Figure 6c) and a delayed plant development compared to the controls (Figure 6b). The co-cultivation with C. asteris changed the root phenotype of the NOA-treated plants only slightly: the root biomass ( Figure 6a) and root length ( Figure S1b) were similar; only the reduction of the root growth decreased slightly (Figure 6c). The plants co-cultivated with C. asteris developed similar to the plants without C. asteris on NOA (Figure 6b). NPA-treated plants had  Table S3 for test statistics).
Increasing NOA and NPA concentrations in the medium ( Figure S2) led to stronger effects: NOA-treated plants developed shorter roots and were delayed in plant development, which was not affected by C. asteris; NPA-treated plants showed still an enhanced lateral root growth and a faster plant development in the presence of C. asteris, but the main root growth was not reduced.
aux1 mutants of A. thaliana (background: Landsberg erecta Ler-0) that lack the auxin resistant 1 (AUX1) influx carrier showed still an increased lateral roots growth (Figure 7c,d) in the presence of C. asteris. The agravitropic root growth, as seen in the aux1 controls, was still present, but it was not as strong as in the aux1 controls. The ecotype Ler-0 showed a similar root phenotype in the presence of C. asteris as the Columbia (Col-0) ecotype (Figures 2 and 7a,b).  (Figure 6a). Co-cultivation with C. asteris led in NPA-treated plants to longer roots ( Figure S1b) and larger rosettes ( Figure S1c) as seen already in control plants.  Table S4.
Increasing NOA and NPA concentrations in the medium ( Figure S2) led to stronger effects: NOA-treated plants developed shorter roots and were delayed in plant development, which was not affected by C. asteris; NPA-treated plants showed still an enhanced lateral root growth and a faster plant development in the presence of C. asteris, but the main root growth was not reduced.
aux1 mutants of A. thaliana (background: Landsberg erecta Ler-0) that lack the auxin resistant 1 (AUX1) influx carrier showed still an increased lateral roots growth ( Figure  7c,d) in the presence of C. asteris. The agravitropic root growth, as seen in the aux1 controls, was still present, but it was not as strong as in the aux1 controls. The ecotype Ler-0 showed a similar root phenotype in the presence of C. asteris as the Columbia (Col-0) ecotype (Figures 2 and 7a,b). Figure 6. Co-cultivation of A. thaliana and C. asteris with IAA transport inhibitors 2-naphthoxyacetic acid (NOA) and N-1-naphtylphthalamic acid (NPA). A. thaliana and C. asteris were co-cultivated on answering these questions, another piece would be added to the IA in microorganisms and might contribute to resolving some of th analysis of genes possibly involved in IAA biosynthesis and feedi asteris using the IAA precursors and intermediates (Trp, TAM, IA dole) and the IAA biosynthesis inhibitors 2-mercaptobenzimidazo sin DF (YDF) [54,55] will show us possible IAA biosynthetic route In addition, the possible contribution of auxin uptake and e the co-cultivation between C. asteris and A. thaliana on plates con naphthoxyacetic acid, NOA) [56] and efflux (N-1-naphtylphthala tors [57]. This could show if the root phenotype might be caused into the medium.   Table S4.

The Endophyte
The mutant tir1, lacking the transport inhibitor response 1 (TIR1) receptor, showed a similar phenotype as already known from control plants with an increase in lateral roots and a reduction of main root growth, but no difference to Col-0 control roots ( Figure S3). Since the TIR receptor family consists of several proteins, it is possible that others are responsible for IAA perception in the tir1 mutant.
Summarizing these experiments, the IAA influx carrier inhibitor NOA reduced the effects caused by C. asteris, whereas NPA did not influence the phenotype of A. thaliana caused by C. asteris. The absence of the AUX1 influx carrier had a slight effect on the root phenotype, namely the agravitropic response was rescued by C. asteris to some extent. The mutant tir1, lacking the transport inhibitor response 1 (TIR1) receptor, showed a similar phenotype as already known from control plants with an increase in lateral roots and a reduction of main root growth, but no difference to Col-0 control roots ( Figure S3). Since the TIR receptor family consists of several proteins, it is possible that others are responsible for IAA perception in the tir1 mutant.
Summarizing these experiments, the IAA influx carrier inhibitor NOA reduced the effects caused by C. asteris, whereas NPA did not influence the phenotype of A. thaliana caused by C. asteris. The absence of the AUX1 influx carrier had a slight effect on the root phenotype, namely the agravitropic response was rescued by C. asteris to some extent.

Discussion
It was shown that the endophytic fungus C. asteris synthesized IAA via different pathways. In addition to Trp-dependent pathways, C. asteris uses also a Trp-independent pathway using indole as precursor to synthesize IAA. C. asteris secreted most of the IAA into the culture medium and retained only small amounts in its hyphae (Figures 3-5). The root phenotype of A. thaliana co-cultivated with C. asteris was reminiscent of high auxin concentrations, and inoculated auxin-responsive promoter reporter lines (DR5-GUS) showed an increased staining in areas where lateral root primordia are formed compared to control roots ( Figure 2). Similarly, it was reported for the interaction of T. virens and A. thaliana that in a DR5::GUS line, an increased activity in the root tip for auxin responses was found [1]. Unfortunately, the pictures did not reveal any changes in other parts of the roots, so it is not possible to compare these with the results found in the interaction of C. asteris and A. thaliana in this study (Figure 2).
Most of the IAA was secreted into the medium, indicating that fungal IAA most likely acts in the host plant and not in the fungal endophyte. In the natural environment inside the plant, secreted fungal IAA would influence the plant IAA homeostasis and thus plant AUX1 growth and development. IAA was synthesized using Trp-dependent and -independent pathways in C. asteris (for the different pathways, see Figure 1). The direct precursor of IAA in the Trp-dependent pathways-IAM, IAN and IAD-led to an increase in the IAA content, and this is corroborated by the existence of the candidate genes involved in the different pathways that were found in the genome of C. asteris (Table 1). Despite the fact that expression of the genes has not been determined so far, the increase of IAA production after incubation with the respective precursors indicate that these gene products could be involved in the auxin production.
The existence of the IPyA pathway could not be clearly verified, since IPyA did not increase the de novo synthesized IAA level in C. asteris. However, since IPyA was rapidly degraded in the fungal medium (data not shown), its possible role as an intermediate cannot completely be ruled out. In addition, candidate genes were not found in the genome,

Discussion
It was shown that the endophytic fungus C. asteris synthesized IAA via different pathways. In addition to Trp-dependent pathways, C. asteris uses also a Trp-independent pathway using indole as precursor to synthesize IAA. C. asteris secreted most of the IAA into the culture medium and retained only small amounts in its hyphae (Figures 3-5). The root phenotype of A. thaliana co-cultivated with C. asteris was reminiscent of high auxin concentrations, and inoculated auxin-responsive promoter reporter lines (DR5-GUS) showed an increased staining in areas where lateral root primordia are formed compared to control roots ( Figure 2). Similarly, it was reported for the interaction of T. virens and A. thaliana that in a DR5::GUS line, an increased activity in the root tip for auxin responses was found [1]. Unfortunately, the pictures did not reveal any changes in other parts of the roots, so it is not possible to compare these with the results found in the interaction of C. asteris and A. thaliana in this study ( Figure 2).
Most of the IAA was secreted into the medium, indicating that fungal IAA most likely acts in the host plant and not in the fungal endophyte. In the natural environment inside the plant, secreted fungal IAA would influence the plant IAA homeostasis and thus plant AUX1 growth and development. IAA was synthesized using Trp-dependent and -independent pathways in C. asteris (for the different pathways, see Figure 1). The direct precursor of IAA in the Trp-dependent pathways-IAM, IAN and IAD-led to an increase in the IAA content, and this is corroborated by the existence of the candidate genes involved in the different pathways that were found in the genome of C. asteris (Table 1). Despite the fact that expression of the genes has not been determined so far, the increase of IAA production after incubation with the respective precursors indicate that these gene products could be involved in the auxin production.
The existence of the IPyA pathway could not be clearly verified, since IPyA did not increase the de novo synthesized IAA level in C. asteris. However, since IPyA was rapidly degraded in the fungal medium (data not shown), its possible role as an intermediate cannot completely be ruled out. In addition, candidate genes were not found in the genome, indicating that IAA might not be synthesized via the IPyA pathway. The IPyA pathway is widely spread in bacteria and fungi: Azospirillum brasilense [59,60], Pseudomonas stutzeri [61], and Ustilago maydis [24] are just a few examples. Pseudomonas species convert Trp directly into IAD without synthesizing IPyA, which is catalyzed by a Trp side chain monooxygenase [39][40][41][42]. Such a pathway cannot be excluded in C. asteris, but a candidate gene for a Trp monooxygenase in the genome was not found.
The IAM pathway would also be possible, but here again, a Trp monooxygenase would be needed. The first IAM pathway was described for Pseudomonas savastanoi [40], but it is found by now also in other plant-associated bacteria [62] and fungi. The fungal pathogens Fusarium proliferatum [27], Colletotrichum fruticola [63], and C. gloeosporioides f.sp. aeschynomene [64] synthesize IAA also over the IAM pathway. The latter fungus uses the IAM and IPyA pathway, preferring IAM as the main biosynthetic way [64]. The IAM pathway is usually used by pathogens to induce an uncontrolled plant cell growth [65]. However, different Streptomyces species indicate more than one IAA biosynthetic pathway, including the IAM pathway as the dominant one [66].
IAM is not only an intermediate in the IAM pathway, but it can also be an intermediate in the IAOx-IAN pathway. Candidate genes were found for the Trp conversion into IAOx in C. asteris, but no gene was found for the IAOx conversion into IAN. IAN can be converted either directly into IAA using nitrilases or via IAM into IAA, but the genome was lacking candidate genes for the nitrile hydratase (conversion of IAN into IAM) in C. asteris. Nitrilases are a large family in bacteria, fungi, plants, and animals [67]. IAN-specific nitrilases were found among others in Pseudomonas species [68,69] and Alcaligenes faecalis JM3 [70,71]. A nitrile hydratase could be isolated from the plant pathogen Agrobacterium tumefaciens and the leguminous Rhizobium spp. [72], which catalyzes the transformation of IAN into IAM.
The most possible IAA biosynthesis pathway in C. asteris might be the TAM pathway: here, candidate genes could be identified for each enzymatic reaction, and an induced biosynthesis with exogenous IAD could be shown (Figure 4). The TAM pathway is known in different bacteria and fungi. Several microorganisms are described to possess a Trp decarboxylase activity, which is responsible for the Trp decarboxylation of Trp to TAM (Taphrina deformans [73], Metarhizium robertsii [74]). Other microorganisms increase their IAA level when they were supplemented with exogenous TAM (Azospirillium [75], Ustilago maydis [76], Leptosphaeria maculans [6]). In addition to the TAM pathway, which might be the main pathway for IAA biosynthesis in C. asteris, there could exist also other pathways for IAA biosynthesis with partial enzymatic coverage, where the substrates for missing steps could be provided by the host plant.
In addition to the different Trp-dependent IAA biosynthesis pathways, there is a Trp-independent pathway. There is evidence that Trp-independent pathways exist in bacteria [60] and fungi [77], but it is not clear whether IAA is synthesized using indole or indole-3-glycerol phosphate (IGP) as a precursor. The IAA biosynthesis inhibitor MBI can inhibit the Trp-dependent biosynthesis of IAA [53,78]. C. asteris synthesizes high levels of IAA in the presence of indole (Figure 5), if the Trp-dependent biosynthesis was inhibited by MBI.
Different plant mutants suggest that IAA can be synthesized from indole and IGP. Orp mutants of maize, which show a defect in the last step of Trp biosynthesis [16,17], synthesized IAA from 15 N 1 -anthranilate (indole precursor), but not from D 5 -Trp. The mutants accumulated up to 50 times more IAA than controls [79]. trp2 mutants of A. thaliana also synthesized IAA from 15 N 1 -anthranilate and not from D 5 -Trp. trp2 and trp3 mutants accumulated very high levels of IAA (38-fold and 19-fold more than controls) [80]. A. thaliana plants with an antisense gene copy of IGP synthase as well as trp2-1 and trp3-1 mutants produced different amounts of IAA: whereas the trp mutants increased their IAA level, IGP antisense mutants decreased it [18]. Whether these results from plants can be transferred to bacteria and fungi has to be shown by further research, but there is a strong evidence for the existence of a Trp-independent IAA biosynthesis starting with indole and/or IGP also in these organisms.
The second IAA biosynthesis inhibitor YDF used in our study is a more stable derivative of yucasin and blocks the conversion of Trp into IPyA by the YUCCA enzymes in plants [54,55]. The application of YDF to C. asteris changed the IAA levels in Trp-treated cultures but not dramatically. The IPyA pathway is only one out of many different IAA biosynthesis pathways, which could compensate the possible reduction of IAA. Furthermore, possible genes involved in the IPyA pathway were not found in the genome, which does not exclude them since it is only an automatic annotation. The cultures supplemented with indole showed an increased IAA level indicating alternative IAA biosynthesis path-ways in C. asteris beside over IPyA. In consequence, the IAA level increased only very slightly (not significant) in these cultures in the presence of YDF ( Figure 5).
In the co-cultivation experiment with A. thaliana and C. asteris on medium containing the IAA transport inhibitors NOA or NPA, the IAA efflux carrier inhibitor NPA did not influence the root phenotype remodeling associated with C. asteris. Only the IAA influx carrier inhibitor NOA inhibited the alterations in the root architecture of A. thaliana in most instances. The roots developed, compared to the non-co-cultivated plants on NOA, only slightly elongated lateral roots without an increase in root biomass. IAA is transported into the cell passively by diffusion and actively by AUX1/LAX transporters [81]. These transporters are blocked by NOA [56], reducing the intracellular IAA level. Diffusion of IAA still takes place and explains why NOA-treated plants slightly increased their root length in the presence of C. asteris. NPA is targeting the IAA efflux carriers ABCB1 and ABCB19 [82][83][84], which blocks the polar and thus basipetal auxin transport [10]. There is disagreement regarding whether NPA directly inhibits the PIN efflux transporters [57]. pin1 loss-of-function mutants of A. thaliana show the same phenotype as NPA-treated plants [85][86][87]. NPA-treated plants showed the typical phenotype, described in the literature [88]: agravitropic root growth with reduced lateral roots. The influx carrier proteins are not affected by NPA [89], so IAA from the fungus can still be transported into the cell and influence the IAA homeostasis of the plant.
A. thaliana plants without any NOA developed a root phenotype (Figures 2 and 6), which is known to be induced by other plant-beneficial microorganisms. Treating plants with IAA transport inhibitor reduces the effect of microorganisms on root architecture [90][91][92][93]. For example, the plant-growth promoting fungus Aspergillus ustus induced lateral root growth and thus root biomass [94]. The bacterium Bacillus megaterium STM196 reduced primary root growth, but it induced lateral root growth and root biomass [51]. Several microorganisms are known to induce these changes in root architecture (reviewed in [95]). The induction of lateral root growth was lost in aux-mutants of A. thaliana when inoculated with different microorganisms [91,96]. The increased root system, induced by plant-beneficial microorganisms, can improve nutrition uptake, which leads to a better tolerance against different environmental stresses [97][98][99].
Further experiments should investigate the potential contribution of the different candidate genes involved in IAA biosynthesis that were identified. Additional feeding experiments with heavy-isotope labeled Trp and identification of the intermediates will also improve our knowledge about which Trp-dependent pathway is preferred by C. asteris.

Plant and Fungal Material
All A. thaliana ecotypes (Col-0 and Ler-0) and mutants (aux1 and tir1) used in this study were obtained from Nottingham Arabidopsis Stock Centre (NASC, Loughborough, UK). C. asteris was derived from our own culture (deposited under the accession number DSM 100826 at DSMZ, Braunschweig, Germany).

Co-Cultivation of A. thaliana with C. asteris
A. thaliana plants of the ecotypes Col-0, Ler-0, and auxin mutants aux1 and tir1 were co-cultivated with C. asteris on Petri dishes, sealed with Parafilm (Bemis, Neenah, WI, USA). A. thaliana seeds were sterilized with 70% ethanol containing 0.1% Triton-X 100 for 1 min, followed by 1.2% sodium hypochlorite containing 0.1% Triton-X 100 and three to four times washing with autoclaved water. Seven sterilized seeds were sown in the upper part of Petri dishes with communication between both partners could shed light on possible mechanisms of ho the synthesis of the antitumor compounds astins might be regulated. Since A. tataric grows too large for laboratory studies, A. thaliana was chosen instead for co-cultivatio experiments with C. asteris under sterile conditions on Petri dishes.
Co-cultivation assays with A. thaliana showed a root phenotype well known to b induced by IAA-producing microorganisms in plant-beneficial interactions [50][51][52] shorter main root, increased lateral root growth, enhanced root biomass, and an aux response detected in auxin-inducible promoter-reporter plants specifically in root tip and lateral root primordia.
A further question was whether C. asteris is capable of synthesizing IAA de novo an whether the IAA can also be secreted into the medium, and if so, by which pathway(s). answering these questions, another piece would be added to the IAA biosynthesis puzz in microorganisms and might contribute to resolving some of the loose ends. Genom analysis of genes possibly involved in IAA biosynthesis and feeding experiments with asteris using the IAA precursors and intermediates (Trp, TAM, IAN, IAM, IAD IPyA, i dole) and the IAA biosynthesis inhibitors 2-mercaptobenzimidazole (MBI) [53] and yuc sin DF (YDF) [54,55] will show us possible IAA biosynthetic routes.
In addition, the possible contribution of auxin uptake and efflux was analyzed fo the co-cultivation between C. asteris and A. thaliana on plates containing IAA influx ( naphthoxyacetic acid, NOA) [56] and efflux (N-1-naphtylphthalamic acid, NPA) inhib tors [57]. This could show if the root phenotype might be caused by fungal IAA secrete into the medium.  MS/MEAlow (1.1 g/L MS incl. vitamins, 5 g/L malt extract, 5 g/L D-glucose, 0.5 g/L peptone, 0.5 mL/L Hutner's trace elements [100]; pH 6.1) and cultivated under long day conditions (16 h with 23 • C with light intensities between 90 and 120 µmol s −1 m −2 ; 8 h with 18 • C in the dark). After one week, C. asteris conidia were applied to the lower part of the Petri dish. Conidia were harvested from a one-week-old C. asteris culture in MEAlow by filtering (gauze, 20 µm mesh size) and centrifugation (5000 rcf, 4 • C, 10 min). The pellet with conidia was dissolved in MEAlow and used for inoculation. Each plant was inoculated with 10 5 conidia in a distance of 4.5 cm to the seeds (seven spots per Petri dish). The phenotype of A. thaliana plants was recorded weekly (root length, growth stage). The plants were harvested after 35 days of cultivation for fresh and dry weight determination. For treatment with auxin transport inhibitors, A. thaliana seeds were sterilized and sown as above. After one week, A. thaliana seedlings were transferred to Petri dishes containing

Incubation of C. asteris with IAA Precursors and/or Biosynthesis Inhibitors
C. asteris was cultivated in liquid MEAlow on a shaking incubator (23 • C, 130 rpm) for 10 days. After centrifugation (5000 rcf, 10 • C, 10 min) to separate medium and hyphae, hyphae were freeze-dried and dry weight was determined. IAA was extracted and analyzed from medium and freeze-dried hyphae.
To determine an IAA biosynthesis pathway in C. asteris, liquid cultures of C. asteris were fed with different IAA precursors (1 mM Trp dissolved in water, sterilized by autocaving, and 1 mM indole dissolved in methanol) and intermediates (1 mM TAM; 100 µM of IAN/ IAM/IAD/IPyA; all dissolved in methanol) after seven days of cultivation. The cultures were harvested (see above) after three additional days for IAA extraction and analysis (see Section 4.4).
In a third experiment, C. asteris was cultivated in liquid MEAlow including the IAA biosynthesis inhibitors MBI (60 µM; dissolved in isopropanol) or YDF (100 µM; dissolved in isopropanol). IAA precursors Trp or indole (each 1 mM) were added after seven days. After an additional three days, the fungal culture was harvested (see above) to extract and analyze the IAA content.

Extraction and Detection of IAA
For IAA extraction, each medium sample was divided into three technical replicates, which was set up with 1 µg internal standard D 5 -IAA (Euriso-top SAS, Saint-Aubin, France). Then, the medium was shaken for 2 h in the dark at 4 to 8 • C to guarantee a uniformly distribution of the internal standard. After this, the pH value was set to three, and the medium was extracted twice with the same volume of ethyl acetate (HPLC grade, Carl Roth, Karlsruhe, Germany) (vortexing for 10 s, centrifugation for 5 min, 4 • C, 5000 rcf). The organic phases were collected and evaporated at a Multivapor P-12 (Büchi Labortechnik GmbH, Essen, Germany). The dried sediments were resuspended twice in 500 µL methanol (HPLC grade, Carl Roth, Karlsruhe, Germany) and again evaporated under gaseous nitrogen. In the next step, the 200 µL of methanol resolved IAA was methylated with the same volume of (trimethylsilyl)diazomethane (1:100 dilution in diethylether) at room temperature for 30 min, before the methanol was evaporated under gaseous nitrogen. The IAA sample was finally resuspended in 50 µL of ethyl acetate and measured by gas chromatography-mass spectrometry (GC-MS).
Freeze-dried hyphae samples were ground and mixed with 80% methanol, including 5% acetic acid. The samples were divided into three technical replicates, and 1 µg internal standard D 5 -IAA was added to each replicate. After shaking for 2 h in the dark at 4 to 8 • C, the samples were centrifuged to remove the pellet (10,000 rcf, 4 • C, 10 min). The methanol of the supernatant was evaporated under gaseous nitrogen until the water phase was left. The pH of the water phase was set to three and extracted twice with the same volume of ethyl acetate. The combined organic phases of ethyl acetate were evaporated under gaseous nitrogen before the methylation of IAA was performed as described above. The sample was treated as described for medium, and the IAA contents were measured by GC-MS.
Samples were separated by GC 3900 on an Agilent HP-5 column (30 m/Ø 0.250 mm/0.25 µm film) with 1 mL/min helium (10 psi). Sample (1 µL) was injected splitless at 250 • C and separated by the following method: 2 min at 70 • C, then a temperature increase about 20 • C/min until 280 • C, and held until 17.5 min. The separated sample was analyzed with an ion trap (Saturn 2100 MS; Varian, Palo Alto, CA, USA) and ionized by electron ionization (70 eV) at 200 • C.
The IAA content of each sample was calculated according to the isotope dilution equation [101] with the corresponding m/z (methylated IAA: m/z 130; methylated D 5 -IAA: m/z 135). The background of the IAA precursors and intermediates at m/z 130 were subtracted to get only the de novo synthesized IAA levels of C. asteris. To compare the different experiments with the different IAA precursors/intermediates, the IAA levels in each experiment were related to the respective control without any supplementation.
Data of the interaction study between A. thaliana and C. asteris on NOA-or NPAcontaining medium were analyzed with different statistical analyses: (a) A Welch-ANOVA (due to an unequal sample size and dissimilar variances) was used for biomarkers with unequal samples size and dissimilar variances, with IAA transport inhibitor included as a six-level between-subjects factor (control, NOA, NPA without and with C. asteris). Post-hoc tests were performed using a Games-Howell test. (b) A Kruskal-Wallis test was used for biomarkers with equal sample size but unsimilar variances or for growth stages, with IAA transport inhibitor included as a three-level (control, NOA, NPA) and C. asteris included as two-level (without or with C. asteris) between-subjects factor. Post-hoc tests were performed using Dunn's test. Effect sizes were calculated [102,103].
For determination of biosynthetic pathways, data of the controls (without any precursor or intermediate) were analyzed using an unpaired t-test on the medians of the three technical replicates, assuming unequal variances (Welch correction).
IAA levels of C. asteris cultures, which were supplemented with different IAA precursors/intermediates and/or IAA biosynthesis inhibitors, were statistically analyzed by performing a Welch-ANOVA on the medians of the three technical replicates, with IAA precursors as an eight-level (control, Trp, TAM, IAN, IAM, IPyA, IAD, indole) or IAA biosynthesis inhibitor as a six level (control, MBI, and YDF without and with C. asteris) between-subjects factor. A post-hoc test was performed (Games-Howell test) and effect sizes were calculated [102,103]. The respective test used is given in the results and/or figure legends.

Conclusions
This study revealed that the endophytic fungus C. asteris produces the plant hormone IAA via different biosynthetic pathways: Trp-dependent and Trp-independent. The most likely Trp-dependent way is the TAM pathway, which does not exclude other pathways such as via IPyA or IAM. The Trp-independent pathway starts with indole as the substrate. By synthesizing and secreting IAA, C. asteris influences the root growth of A. thaliana that results in more lateral but shorter main roots. The phenotype was reduced after the addition of an auxin transport inhibitor for influx but not efflux. Further experiments should lead to the identification of responsible genes in the endophyte for the biosynthesis and secretion of IAA. This study needs to be investigated further in a future study to see how the growth of the roots behave and change through different stages of the plant growth ending to the greenhouse stage.
Supplementary Materials: The following are available online at https://www.mdpi.com/1422-006 7/22/5/2651/s1, Figure S1: Co-cultivation of A. thaliana and C. asteris on IAA transport inhibitor containing medium, Figure S2: Co-cultivation of A. thaliana and C. asteris in the presence of IAA transport inhibitors, Figure S3: Growth of tir1 mutant of A. thaliana with C. asteris after 35 days, Table S1: Test statistics on IAA levels in medium and hyphae of C. asteris, using an unpaired ttest. Table S2: Test statistics on IAA levels in medium and hyphae of C. asteris in presence of IAA precursors, Table S3: Test statistics on IAA levels in medium and hyphae of C. asteris in the presence of MBI or YDF, Table S4: Test statistics on co-cultivation of A. thaliana and C. asteris on 10 µM NOA or 10 µM NPA, Table S5: Test statistics on co-cultivation of A. thaliana and C. asteris on 30 µM NOA or 20 µM NPA. Table S6: Raw data. Data Availability Statement: All data supporting this study are available in this paper and in its supplementary data published online.