Mycobiota of Mexican Maize Landraces with Auxin-Producing Yeasts That Improve Plant Growth and Root Development

Compared to agrochemicals, bioinoculants based on plant microbiomes are a sustainable option for increasing crop yields and soil fertility. From the Mexican maize landrace “Raza cónico” (red and blue varieties), we identified yeasts and evaluated in vitro their ability to promote plant growth. Auxin production was detected from yeast isolates and confirmed using Arabidopsis thaliana plants. Inoculation tests were performed on maize, and morphological parameters were measured. Eighty-seven yeast strains were obtained (50 from blue corn and 37 from red corn). These were associated with three families of Ascomycota (Dothideaceae, Debaryomycetaceae, and Metschnikowiaceae) and five families of Basidiomycota (Sporidiobolaceae, Filobasidiaceae, Piskurozymaceae, Tremellaceae, and Rhynchogastremataceae), and, in turn, distributed in 10 genera (Clavispora, Rhodotorula, Papiliotrema, Candida, Suhomyces, Soliccocozyma, Saitozyma Holtermaniella, Naganishia, and Aeurobasidium). We identified strains that solubilized phosphate and produced siderophores, proteases, pectinases, and cellulases but did not produce amylases. Solicoccozyma sp. RY31, C. lusitaniae Y11, R. glutinis Y23, and Naganishia sp. Y52 produced auxins from L-Trp (11.9–52 µg/mL) and root exudates (1.3–22.5 µg/mL). Furthermore, they stimulated the root development of A. thaliana. Inoculation of auxin-producing yeasts caused a 1.5-fold increase in maize plant height, fresh weight, and root length compared to uninoculated controls. Overall, maize landraces harbor plant growth-promoting yeasts and have the potential for use as agricultural biofertilizers.


Introduction
Since the second industrial revolution, poor management, excessive use, and mistimed application of chemical fertilizers and pesticides in agriculture have had negative effects on public health and the environment. High concentrations of agrochemicals cause acidification, physical deterioration, a reduction in soil biodiversity, and an increase in greenhouse gas emissions [1]. In water, the main effects of fertilizers are eutrophication, leaching, and the contamination of aquifers [2].
Corn (Zea mays) has become the leading cereal grain by global production volume, with around 197 million hectares of land cultivated annually [3]. In Mexico, maize's center of origin and domestication, some landraces have adapted to different environmental conditions and are locally grown [4]. Maize landraces are reservoirs of genetic diversity and have been used as donor material for global breeding activities. The production and

Results and Discussion
The intensive use of synthetic fertilizers and pesticides to improve crop production has caused the deterioration of natural resources, and has caused damage to human health. Developing sustainable agricultural practices that take advantage of crop-associated microbiota to promote plant growth and soil fertility is therefore crucial. There is a need to isolate local yeast strains from native crops and evaluate their ability to stimulate plant growth and be used in biofertilizers.

Yeast Diversity Associated with Maize Plants
A total of 87 yeast strains were obtained from the Mexican maize landrace "Raza cónico," of which 50 were isolated from the blue variety and 37 from the red variety (Supplementary Table S1). In blue corn, 29 strains came from the phyllosphere, and the remaining strains were found as leaf endophytes. In the red maize, there was a greater distribution of yeasts: nine strains were isolated from kernels, 12 from the phyllosphere, 15 from the root endosphere, and one from the leaf endosphere. Phylogenetic analysis placed the isolates within the two phyla Ascomycota and Basidiomycota, eight families and 10 genera (Clavispora, Rhodotorula, Papiliotrema, Candida, Suhomyces, Soliccocozyma, Saitozyma, Holtermaniella, Naganishia, and Aeurobasidium) ( Figure 1). strains with the most outstanding characteristics were used as bioinoculants on plants in pot trials, and morphometric parameters were evaluated.

Results and Discussion
The intensive use of synthetic fertilizers and pesticides to improve crop prod has caused the deterioration of natural resources, and has caused damage to h health. Developing sustainable agricultural practices that take advantage of crop-a ated microbiota to promote plant growth and soil fertility is therefore crucial. The need to isolate local yeast strains from native crops and evaluate their ability to stim plant growth and be used in biofertilizers.

Yeast Diversity Associated with Maize Plants
A total of 87 yeast strains were obtained from the Mexican maize landrace cónico," of which 50 were isolated from the blue variety and 37 from the red variety plementary Table S1). In blue corn, 29 strains came from the phyllosphere, and t maining strains were found as leaf endophytes. In the red maize, there was a great tribution of yeasts: nine strains were isolated from kernels, 12 from the phyllosphe from the root endosphere, and one from the leaf endosphere. Phylogenetic analysis p the isolates within the two phyla Ascomycota and Basidiomycota, eight families a genera (Clavispora, Rhodotorula, Papiliotrema, Candida, Suhomyces, Soliccocozyma, Sait Holtermaniella, Naganishia, and Aeurobasidium) ( Figure 1).  Most of the isolates belonged to the phylum Ascomycota [blue maize (52%) and red maize (57%)]. In contrast, in a study of maize plants in Thailand, 217 yeast isolates were obtained, and those from the phylum Basidiomycota were more abundant (81.6%) [22]. In contrast, yeasts associated with corn in Brazil belonged entirely to the phylum Ascomycota [23]. These findings showed that yeast diversity appears to be largely influenced by maize variety, tissue type, and the soil in which the plants are grown. The different maize varieties had contrasting anthocyanin content and profiles of metabolites exuded by roots which probably affected the mycobiota structure in the rhizosphere and endosphere [24].
Distribution of yeast genera in this work is shown in Figure 2. Strains of Aeurobasidium, Naganishia, and Saitozyma were obtained only from blue corn, and Suhomyces, Holtermanniella, Kurtzmaniella, Papiliotrema, and Solicoccozyma were exclusive to red maize plants. In common, we found strains of Clavispora, Papiliotrema, Rhodotorula, and Candida.
Distribution of yeast genera in this work is shown in Figure 2. Strains of Aeurobasidium, Naganishia, and Saitozyma were obtained only from blue corn, and Suhomyces, Holtermanniella, Kurtzmaniella, Papiliotrema, and Solicoccozyma were exclusive to red maize plants. In common, we found strains of Clavispora, Papiliotrema, Rhodotorula, and Candida.
Here, Clavispora lusitaniae was the most predominant species (Figure 2). The monophyletic genus Clavispora (Anamorph: Candida) belongs to the Metschnikowiaceae family, which includes isolates from plants, fruits, and soil [25,26]. Other members of this family are well-known for producing exometabolites (e.g., 3-amino-5-methylhexanoic acid) that provide protection against post-harvest pathogens. Rhodotorula is an excellent endophyte; it reaches up to 90% root colonization after 8 days of inoculation of plants [27], which explains why it was the second most abundant in both maize varieties. Papiliotrema and Naganishia (previously classified as Cryptococcus) have also been described as members of the microbiota in soil, roots, and phyllosphere [28]. These yeasts can colonize diverse environments due to their production of capsules and melanin that help them adapt to soils with low nutrients, desiccation, and UV rays [29]. Some species of the genera Candida and Suhomyces have been described as symbionts of sugarcane and maize plants [22,30], while Aerobasidium, which is a notable biocontrol agent, is more commonly found in fruits and vegetables [31]. On the other hand, we found a low prevalence of the genera Solicoccozyma, Saitozyma, and Holtermmaniella, which are well known for their ability to release auxins and lipids [32][33][34].  Here, Clavispora lusitaniae was the most predominant species (Figure 2). The monophyletic genus Clavispora (Anamorph: Candida) belongs to the Metschnikowiaceae family, which includes isolates from plants, fruits, and soil [25,26]. Other members of this family are well-known for producing exometabolites (e.g., 3-amino-5-methylhexanoic acid) that provide protection against post-harvest pathogens. Rhodotorula is an excellent endophyte; it reaches up to 90% root colonization after 8 days of inoculation of plants [27], which explains why it was the second most abundant in both maize varieties. Papiliotrema and Naganishia (previously classified as Cryptococcus) have also been described as members of the microbiota in soil, roots, and phyllosphere [28]. These yeasts can colonize diverse environments due to their production of capsules and melanin that help them adapt to soils with low nutrients, desiccation, and UV rays [29]. Some species of the genera Candida and Suhomyces have been described as symbionts of sugarcane and maize plants [22,30], while Aerobasidium, which is a notable biocontrol agent, is more commonly found in fruits and vegetables [31]. On the other hand, we found a low prevalence of the genera Solicoccozyma, Saitozyma, and Holtermmaniella, which are well known for their ability to release auxins and lipids [32][33][34].

Plant Growth-Promoting Characteristics
Using a set of biochemical tests and enzyme assays, we explored the potential of yeast strains to promote plant growth.
In soil, most phosphorus is found in minerals (such as apatite and hydroxyapatite) that are not assimilated by plants. We found that 14 yeasts could solubilize tricalcium phosphate (Table S1). Among these, the best phosphate solubilizers were: Rhodotorula mucilaginosa RY1 and RY3, Clavispora lusitaniae RY2, Suhomyces prunicola RY4, and Kurtzmaniella quercitrusa RY11. The latter three were formally described as Candida and display high fermentation rates. Under phosphate starvation, corn roots exude more sugars [35]. These sugars can be fermented by yeasts producing organic acids (e.g., citric acid) that dissolve inorganic phosphate, making it bioavailable in the rhizosphere [36]. Phosphate-solubilizing yeasts have increased the biomass, shoot height, and cellular inorganic phosphate content of plants growing under phosphate limitation [37]. Improved maize plant growth and phosphate nutrition have been observed in response to inoculation with a phosphate-solubilizing Candida railenensis, as well as synergism in phosphate uptake by arbuscular mycorrhizal fungi [38]. Therefore, these strains would be useful for agriculture in phosphorus-deficient soils.
Siderophores are low-weight molecules that can bind to ferric ion and transport it into the cell. There, it is reduced to its ferrous form, which is bioavailable. In this study, Papiliotrema flavecens Y27 and Rhodotorula glutinis Y41 produced siderophores under iron-limiting conditions, reaching an activity of 87 ± 4% and 57 ± 3% SU, respectively (Supplementary Table S1). In contrast, Papiliotrema strains isolated from wild cherry tomato rhizosphere did not show siderophores production, but Rhodotorula strains did [39]. Biosynthesis of rhodotorulic acid, a dihydroxamate siderophore, by Rhodotorula has been reported with antagonistic activity against B. cinerea [40]. This suggests that some plant yeasts use siderophores to solubilize iron in soils or as a competitive strategy against other microorganisms in the phyllosphere [41]. Siderophores can also bind other divalent or trivalent cations, such as Co, Cu, Zn, As, Cr, Pb, and Cd [42]. Siderophore-producing yeasts allow plant growth under metal stress [43]. These yeasts have been successfully applied as bioinoculants to assist phytoremediation processes in metal-contaminated soils [44].
Enzyme activities were detected in 13 isolates, of which cellulase (44%), pectinase (27%), and protease (27%) production were the most frequent (Supplementary Table S1). Some isolates were positive for more than one enzyme. Candida sp. RY26, Papiliotrema flavescens RY32, A. pullulans Y64, P. flavecens Y65 and R. glutinis Y41, Y43, Y44, and Y58 showed cellulolytic capacity. Recently, cellulases from Rhodotorula have been used in the food processing and biofuel industries [45]. On the other hand, pectinolytic activity was more evident in Solicoccozyma sp. RY30, A. pullulans Y64, and Naganishia sp. Y19, Y52 and Y53. These isolates were found as root and leaf endophytes. Yeasts may use cellulases and pectinases as colonization mechanisms in plant tissues through the partial degradation of the cell wall [46]. The production of these enzymes and others such as β-1,3-glucanase and chitinase may also serve as biocontrol mechanisms by destroying the cell wall of phytopathogenic fungi such as Aspergillus flavus, B. cinerea, Penicillium expansum, Sphaerotheca fuliginea, and Podosphaera xanthii [17,47]. Aureobasidium pullulans Y64 and R. glutinis Y41, Y43, Y57, and Y58 were positive for proteases. Proteases enhance an endophytic lifestyle by increasing nitrogen uptake from the apoplast during plant colonization [48]. None of the isolates produced amylases. Yeasts that did not show enzymatic activity could use alternative colonization mechanisms, including entry through injuries or cracks. Alternatively, they could colonize through vectors, such as insects or nematodes [28].

Auxin Production
Auxins control several functions in plants, such as vascular tissue differentiation, root initiation, cell division, and elongation. Hence it is crucial that bioinoculants include strains that produce this phytohormone. Auxin production was detected in YPD cultures of 4 out of 87 strains: Solicoccozyma sp. RY31, C. lusitaniae Y11, R. glutinis Y23, and Naganishia sp. Y52. Indolic compound concentration increased 3 three times when tryptophan was present ( Table 1). The auxin biosynthetic pathways in yeasts have not yet been fully determined; however, it is known that their main products are indole-3-acetic acid (IAA) and indole-3pyruvic acid (IPYA) from L-Tryptophan (Trp) [49,50]. This amino acid is transformed into pyruvic acid by aminotransferases and then carboxylated to produce indole-3-acetaldehyde (IAAld) by indole-3-pyruvate decarboxylase. Finally, IAAld is oxidized to IAA. Another scenario is that IPYA is transformed to IAA by indole-3-pyruvate monooxygenase in a single step in yeast [51,52]. Analyses by high-performance liquid chromatography (HPLC) or Gas Chromatography-mass spectrometry (GC-MS) need to be performed to identify the type of auxin released by each strain. The auxin concentrations observed here (from 11.9 ± 1.9 to 52 ± 4.2 µg/mL), along with those reported by Nassar et al. [50], show that yeasts isolated from corn have a higher rate of production than those associated with non-crop plants such as Flaveria angustifolia, Sphaeralcea angustifolia, and Prosopis sp. [29].
19.0 ± 7.7 a 20 ± 7.1 a ‡ Auxins were quantified in yeast cultures using Salkowski's reagent and a standard curve with indole-3-acetic acid. The mean ± standard deviation of three independent experiments is shown. Different lower-case letters between strains indicate statistically significant differences (p-value < 0.05) according to ANOVA with Tukey post hoc.
The strains investigated in this study were able to produce auxins from root exudates, but this was at a lower concentration than those observed in cultures with L-Trp (Table 1). Maize exudates may have L-Trp or its precursors (anthranilate or chorismate) that would support the auxin biosynthesis pathway in yeast [24]. Interestingly, the concentration of auxins produced by the yeasts was higher in the exudates of the maize variety from which they were isolated. Strains likely have a higher affinity to uptake nutrients released by their host. It has previously been reported that Williopsis saturnus, an endophyte yeast, synthesized IAA by obtaining precursors from maize roots [50]. IAA also plays a role as a metabolic regulator that allows fungal cells to enter a quiescent state [53], and in complex communities, this hormone restricts the growth of neighboring fungi [54]. Therefore, auxin production by yeasts could increase their competitiveness and ability to colonize the rhizosphere and plant tissues.

Arabidopsis Thaliana-Yeast Interaction Assays
We extended the study to confirm the plant growth-promoting effect of auxin-producing yeasts using A. thaliana plants ( Figure 3A). We observed that Solicoccozyma sp. RY31, C. lusitaniae Y11, and Naganishia sp. Y52 caused a reduction in primary root elongation as there was an increase in the number of lateral roots compared to the non-inoculated control ( Figure 3B,C). We observed the formation of root hairs on seedlings inoculated with yeast (3A). It is likely that the yeasts induced the cell cycle of the root stele, and consequently, more lateral roots were formed [37]. This modification of the root structure has been stimulated by other yeasts (e.g., Hannaella coprosmaensis, Ustilago esculenta, Sporidiobolus ruineniae, Pseudozyma aphidis, and Dothideomycetes sp.) in Nicotiana benthamiana and the effect have been linked to auxin production in a concentration-dependent manner [55,56]. In our assay, the growth-promoting activity was most prominent during the interaction with the root endophyte Solicoccozyma sp. RY31, which consistently produced auxins in root exudates (22.5 ± 9.7 µg/mL). Although R. glutinis Y23 was a high auxin producer in  Figure 3A). A similar phenomenon was found by Fernandez-San Millan et al. [57], who reported that some strains of Lachancea thermotolerans, Meyerozyma guilliermondii, and Torulaspora delbruekii were harmful to N. benthamiana seedlings. The reasons for these findings are unclear; however, it is possible that yeasts compete with plants for nutrients or that some fungal-derived metabolites at high concentrations may be phytotoxic under in vitro conditions. Sporidiobolus ruineniae, Pseudozyma aphidis, and Dothideomycetes sp.) in Nicotiana benthamiana and the effect have been linked to auxin production in a concentration-dependent manner [55,56]. In our assay, the growth-promoting activity was most prominent during the interaction with the root endophyte Solicoccozyma sp. RY31, which consistently produced auxins in root exudates (22.5 ± 9.7 µg/mL). Although R. glutinis Y23 was a high auxin producer in culture with L-Trp (52 ± 4.2 µg/mL), negative effects on A. thaliana plants were observed by inhibiting their development and causing leaf necrosis and yellowing ( Figure 3A). A similar phenomenon was found by Fernandez-San Millan et al. [57], who reported that some strains of Lachancea thermotolerans, Meyerozyma guilliermondii, and Torulaspora delbruekii were harmful to N. benthamiana seedlings. The reasons for these findings are unclear; however, it is possible that yeasts compete with plants for nutrients or that some fungal-derived metabolites at high concentrations may be phytotoxic under in vitro conditions. To understand whether the positive effects of yeast were mediated by auxin production, we used the A. thaliana DR5:uidA line, which is an efficient IAA biosensor. It has TGTCTC motifs commonly found in the promoters of auxin-inducible genes [58]. The To understand whether the positive effects of yeast were mediated by auxin production, we used the A. thaliana DR5:uidA line, which is an efficient IAA biosensor. It has TGTCTC motifs commonly found in the promoters of auxin-inducible genes [58]. The roots of R. glutinis Y23 were not included because plants showed deficiencies during their development, as explained above. Histochemical analysis showed β-glucuronidase activity in the lateral roots of A. thaliana in co-culture with Solicoccozyma sp. RY31, C. lusitaniae Y11, and Naganishia sp. Y52 (Figure 4). This supports that the auxins produced by yeasts were responsible for the changes in the root structure. roots of R. glutinis Y23 were not included because plants showed deficiencies during their development, as explained above. Histochemical analysis showed β-glucuronidase activity in the lateral roots of A. thaliana in co-culture with Solicoccozyma sp. RY31, C. lusitaniae Y11, and Naganishia sp. Y52 (Figure 4). This supports that the auxins produced by yeasts were responsible for the changes in the root structure.

Evaluation of Auxin-Producing Yeasts as Bioinoculants
Microbes, bacteria, or fungi with in vitro growth-promoting capabilities need to be evaluated on both model plants and crops to fully understand the potential of a strain for use in agriculture. Experiments in pots with vermiculite were carried out to measure the effect of individual auxin-producing yeasts on the corn varieties from which they were grown. After 30 days of treatment with the bioinoculant, a drastic increase in plant development was observed.
In red maize plants, the strains tested had similar values of increased plant height, and their fresh weight was up to 1.5 times greater than the non-yeast controls. However, at the root length level, Naganishia sp. Y52 did not induce significant changes, and C. lusitaniae Y11 showed the best promotion capacity ( Figure 5A). We isolated the Y11 strain from the phyllosphere as well as other strains that have been remarkable for their production of 2-phenylethyl alcohol and 2-phenylethyl acetate that can promote the growth of Agave, A. thaliana and N. attenuate [26].

Evaluation of Auxin-Producing Yeasts as Bioinoculants
Microbes, bacteria, or fungi with in vitro growth-promoting capabilities need to be evaluated on both model plants and crops to fully understand the potential of a strain for use in agriculture. Experiments in pots with vermiculite were carried out to measure the effect of individual auxin-producing yeasts on the corn varieties from which they were grown. After 30 days of treatment with the bioinoculant, a drastic increase in plant development was observed.
In red maize plants, the strains tested had similar values of increased plant height, and their fresh weight was up to 1.5 times greater than the non-yeast controls. However, at the root length level, Naganishia sp. Y52 did not induce significant changes, and C. lusitaniae Y11 showed the best promotion capacity ( Figure 5A). We isolated the Y11 strain from the phyllosphere as well as other strains that have been remarkable for their production of 2-phenylethyl alcohol and 2-phenylethyl acetate that can promote the growth of Agave, A. thaliana and N. attenuate [26].
We observed that all strains of the blue corn plants showed improved plant height and root length, the most significant of which were N. liquefaciens Y52 and R. glutinis Y23 ( Figure 5B). In the field, this would be beneficial to increase the yield of Mexican landraces by developing roots with larger surface areas and improved efficiency in soil nutrients and water uptake. R. glutinis Y23 was the only strain that significantly increased fresh plant weight in blue corn. This finding was unexpected, as this yeast was harmful to A. thaliana. The genus Rhodotorula is well known for its ability to degrade coumaric acid, ferulic acid, vanillic acid, and 4-hydroxybenzoic acid [59], all of which are secreted by maize roots [60]. This may have given R. glutinis Y23 an advantage in feeding on exudates in the rhizosphere while also promoting maize growth. We observed that all strains of the blue corn plants showed improved plant h and root length, the most significant of which were N. liquefaciens Y52 and R. glutinis ( Figure 5B). In the field, this would be beneficial to increase the yield of Mexican land by developing roots with larger surface areas and improved efficiency in soil nutr and water uptake. R. glutinis Y23 was the only strain that significantly increased plant weight in blue corn. This finding was unexpected, as this yeast was harmful thaliana. The genus Rhodotorula is well known for its ability to degrade coumaric acid ulic acid, vanillic acid, and 4-hydroxybenzoic acid [59], all of which are secreted by m roots [60]. This may have given R. glutinis Y23 an advantage in feeding on exudates i rhizosphere while also promoting maize growth.
In our work, yeasts were beneficial for both maize landraces, although there slight variations among strains. This may be due to the ability of each species to colo plant tissues, produce phytostimulants, or uptake nutrients produced by the roots inoculation of W. saturnus in maize cv. Merit resulted in increased weight and leng both roots and shoots [50]. Furthermore, S. cerevisiae Sc-6 and Lachancea thermotoleran 69 improved the height, chlorophyll content, and fresh weight of maize cv. Julliet, a posed to Debaryomyces hansenii Dh-17, which caused harmful effects [57].
Biofertilizers are often Synthetic Microbial Communities (SynCom) made strains with complementary plant promotion capabilities that ensure beneficial effec diverse crops [61]. SynCom has significant and positive impacts on maize develop [62]. In order to progress this research, it would be interesting to explore the comb effect of our yeast strains to improve corn crop yields in the field. In our work, yeasts were beneficial for both maize landraces, although there were slight variations among strains. This may be due to the ability of each species to colonize plant tissues, produce phytostimulants, or uptake nutrients produced by the roots. The inoculation of W. saturnus in maize cv. Merit resulted in increased weight and length of both roots and shoots [50]. Furthermore, S. cerevisiae Sc-6 and Lachancea thermotolerans Lt-69 improved the height, chlorophyll content, and fresh weight of maize cv. Julliet, as opposed to Debaryomyces hansenii Dh-17, which caused harmful effects [57].
Biofertilizers are often Synthetic Microbial Communities (SynCom) made from strains with complementary plant promotion capabilities that ensure beneficial effects on diverse crops [61]. SynCom has significant and positive impacts on maize development [62]. In order to progress this research, it would be interesting to explore the combined effect of our yeast strains to improve corn crop yields in the field.

Maize Accessions
During the 2017-2018 crop cycle, plants of the Mexican maize landrace "Raza Cónico" (Figure 6) were collected from an agricultural field in the locality of "La Paila, Tezoncualpa" in the state of Hidalgo, Mexico (20 • 31 41.9 N, 98 • 43 17.7 W). This landrace has high emergence vigor and cold tolerance with a predominant distribution in temperate areas between 1800 and 3000 m above sea level [63]. Leaf, root, and grain samples were taken from plants (R4 stage) with no chlorosis, necrosis, or parasite infestation and stored in polystyrene bags at 4 • C until they were analyzed. Twelve plants of each variety were sampled.
Cónico" (Figure 6) were collected from an agricultural field in the locality of "La Paila, Tezoncualpa" in the state of Hidalgo, Mexico (20°31′41.9″ N, 98°43′17.7″ W). This landrace has high emergence vigor and cold tolerance with a predominant distribution in temperate areas between 1800 and 3000 m above sea level [63]. Leaf, root, and grain samples were taken from plants (R4 stage) with no chlorosis, necrosis, or parasite infestation and stored in polystyrene bags at 4 °C until they were analyzed. Twelve plants of each variety were sampled. Figure 6. Varieties of the Mexican maize landrace "Raza cónico" were used in this study. Blue (above) and red (below) conical cobs with kernels semi-crystalline and semi-dentate.

Yeast Strain Isolation
For phyllosphere yeast communities, 10 g of leaves from each plant were shaken into flasks with 90 mL of sterile distilled water on a rotary shaker (Inch Incubated Shaker, BEING Scientific Inc., Ontario, CA, USA, BIS-3, 230 V) for 30 min; the solution was collected, and serial dilutions were made to 10 −3 and spread on a YPDA medium (g/L; 10 yeast extract, 20 peptones, 20 D-glucose, 20 agar, 10% tartaric acid (w/v), pH 4.0). Endophytes were isolated according to the suggestion of [29,64], although slight modifications were made to these. Ten grams of leaves, roots or kernels were surface-sterilized by immersion in 70% ethanol for 30 s and then in 4% sodium hypochlorite solution for 6 min. Plant tissues were then crushed in a sterile blender (Classic Blender Model, Oster, Boca Raton, FL, USA) with 90 mL of sterile distilled water for 5 min, and serial dilutions up to 10 −3 were spread on a YPDA medium. All plates were incubated at 28 °C for 5 days. Yeast colonies were selected based on morphological differences and purified on a YPDA medium. The purified strains were preserved in microtubes with 30% glycerol at −80 °C.

Molecular Identification and Phylogenetic Analysis
Genomic DNA was isolated from fungal cultures with 5 days of growth on YPDA, using the Quick-DNA Fungal/Bacterial Miniprep Kit (Zymo Research, Irvine, CA, USA), Figure 6. Varieties of the Mexican maize landrace "Raza cónico" were used in this study. Blue (above) and red (below) conical cobs with kernels semi-crystalline and semi-dentate.

Yeast Strain Isolation
For phyllosphere yeast communities, 10 g of leaves from each plant were shaken into flasks with 90 mL of sterile distilled water on a rotary shaker (Inch Incubated Shaker, BEING Scientific Inc., Ontario, CA, USA, BIS-3, 230 V) for 30 min; the solution was collected, and serial dilutions were made to 10 −3 and spread on a YPDA medium (g/L; 10 yeast extract, 20 peptones, 20 D-glucose, 20 agar, 10% tartaric acid (w/v), pH 4.0). Endophytes were isolated according to the suggestion of [29,64], although slight modifications were made to these. Ten grams of leaves, roots or kernels were surface-sterilized by immersion in 70% ethanol for 30 s and then in 4% sodium hypochlorite solution for 6 min. Plant tissues were then crushed in a sterile blender (Classic Blender Model, Oster, Boca Raton, FL, USA) with 90 mL of sterile distilled water for 5 min, and serial dilutions up to 10 −3 were spread on a YPDA medium. All plates were incubated at 28 • C for 5 days. Yeast colonies were selected based on morphological differences and purified on a YPDA medium. The purified strains were preserved in microtubes with 30% glycerol at −80 • C.

Molecular Identification and Phylogenetic Analysis
Genomic DNA was isolated from fungal cultures with 5 days of growth on YPDA, using the Quick-DNA Fungal/Bacterial Miniprep Kit (Zymo Research, Irvine, CA, USA), following the manufacturer's instructions. The region from the Internal transcribed spacer 1 (ITS1) to the D1/D2 domain of large-subunit (LSU) ribosomal DNA (rDNA) was amplified by PCR with the primers ITS1 and NL4 [65,66] in a thermocycler Veriti (ThermoFisher, Waltham, MA, USA). The amplicons were purified and sequenced by Macrogen in Seoul, South Korea. The raw sequences were edited with BIOEDIT 7.2.5 [67] and then clustered into ITS haplotypes using the DAMBE7 package [68]. Taxonomically related sequences were obtained by BLAST in NCBI. All sequences were aligned using the ClustalX program, and the best nucleotide substitution model was determined using JModelTest2 [69]. A phylogenetic analysis was performed in MEGAX with the maximum likelihood method and 1000 bootstrap replicates [70]. All sequences obtained in this study were deposited in the GenBank under accession numbers MN299221-MN299307. were inoculated with 5 µL of yeast solution (1 × 10 8 cells/mL). Plates were incubated for 7 days at 28 • C, and the colonies with clear halos surrounding them were considered positive [71]. Three experiments were performed for each isolated strain.

Siderophore Quantification
Siderophore production was induced by inoculating 1 × 10 8 yeast cells/mL in 5 mL of Iron deferrated Grimm-Allen liquid media [72] and incubating this at 28 • C for 15 days at 120 rpm. The cultures were centrifuged, and 100 µL of the supernatant was mixed with 100 µL of chrome azurol S (CAS)-Fe solution and 2 µL of shuttle solution (0.2 M 5-sulfosalicylic acid) [73]. The reaction mixture was incubated in the dark at room temperature for 10 min and then read at 630 nm on a VersaMax™ Tunable Microplate Reader. The results were reported as percentage siderophore units (% SU) according to the formula suggested by Chowdappa et al. [74]. Quantification was carried out in triplicate.

Production of Plant Cell Wall-Degrading Enzymes
Amylase and pectinase activities were performed in a minimal medium using 2% potato starch (Sigma Aldrich, Burlington, MA, USA) and 1% pectin from the citrus peel (Sigma Aldrich, Burlington, MA, USA) as a carbon source, respectively [75]. Cellulase production was evaluated in Congo red medium with 2% methylcellulose [76] and protease production in skim milk medium [77]. All plates were inoculated with 5 µL of yeast solution (1 × 10 8 cell/mL) and incubated at 28 • C for 5 days. In order to detect amylases, the colony was flooded in an iodine solution (1% w/v). The presence of a clear halo around the colony was considered positive. Pectinase activity was visualized by a clear halo around the colony after it was flooded with a CTAB solution (cetyltrimethylammonium bromide, 5% w/v). Cellulase and protease activities were demonstrated by a clear halo around the colony. The enzymatic index (EI) was determined by subtracting the colony's diameter from the halo's diameter [64]. Assays were performed in triplicate for each strain.

Auxin Production in Culture Media
The auxin production was tested in 5 mL of YPD broth (g/L; 10 yeast extract, 20 peptone, 20 dextrose, pH 6.0) with or without 0.1% L-tryptophan (L-Trp) inoculated with 1 × 10 8 yeast cells/mL. The cultures were incubated for 15 days at 28 • C at 120 rpm. Following this, the supernatants were obtained by centrifugation and filtered through a 0.2 µm filter (Millipore, Burlington, MA, USA). One hundred microliters of the filtered supernatant were mixed with 100 µL of Salkowski reagent (2% of 0.5M FeCl 3 in 35% HClO 4 ) and incubated in the dark at 28 • C for 30 min. The intensity of the pink color formed was read on a VersaMax™ Tunable Microplate Reader at wavelength 530 nm. Following the suggestion of Herrera-Quiterio et al. [77], a commercial IAA (Sigma Aldrich, Burlington, MA, USA) standard curve was made to estimate the auxin concentration produced by each strain. Three experiments for each isolated strain were performed.

Induction of Auxin Production by Root Exudates
In order to obtain maize root exudates, 10 grains of both varieties were disinfected and germinated according to the process adopted by Rosenblueth and Martínez-Romero [78]. Two-day-old seedlings were transferred to a stainless-steel pedestal inside glass tubes ( [79]. Plants were cultivated in a light/dark photoperiod (16 h/8 h) for 5 days at 28 • C. Root exudates were collected using a miraclot filter and subsequently dried by rotary evaporation in an Eppendorf Concentrator 5301 (Eppendorf, Hamburg, Germany) at 60 • C for 1 h and stored at −80 • C. Five mL of YPD broth was supplemented with the root exudates (final concentration approximately 1 mg/mL) and inoculated with 1 × 10 8 yeast cells/mL. Cultures were incubated for 15 days at 28 • C at 120 rpm, and IAA production was determined from the supernatants as described above. Three experiments for each of the four auxin-producing yeasts were performed. A negative control containing YPD with non-inoculated exudates showed no reaction with the Salkowski reagent.

Arabidopsis Thaliana-Yeast Interaction Assays
Seeds from the Arabidopsis thaliana ecotype Columbia-0 (Col-0) and the DR5:uidA line were disinfected, according to Rosenblueth and Martinez-Romero [78]. They were then transferred to plates with 0.2× Murashige and Skoog (MS) medium with 0.8% agar and 1% sucrose. Plates were maintained at 21 • C, 16/8 h light/dark photoperiod until germination. Four days later, seedlings were inoculated with 10 6 cells of the auxin-producing yeasts 4 cm from the root tip. After one week, the length of the primary root and lateral roots emerged, as expected by Matus-Acuña et al. [80]. Thirty plants were evaluated for each yeast and non-inoculated control.

Evaluation of Auxin-Producing Yeasts as a Bioinoculant
Seeds from each maize variety were disinfected and germinated, as outlined in previous sections. Three-day-old seedlings were transferred to pots with 1 kg of vermiculite autoclaved (3 times for 30 min at 121 • C, 15 psi) and then individually inoculated with 30 mL of a solution of auxin-producing yeasts (1 × 10 8 yeast cells/mL). Plants were kept in a growth room at 28 • C with a 16/8-h light/dark photoperiod. They were watered every 3 days with 30 mL of Fahraeus solution [82], which was supplemented with 0.08 g/L NH 4 NO 3 . After 30 days, plants were harvested, and morphometric parameters such as plant height, root length, and fresh weight were measured. The experiment included 5 replicates of the two maize varieties for each of the auxin-producing yeasts alongside two treatments without inoculation as controls. A total of 50 plants were analyzed.

Statistical Analysis
All statistical analyses and graphics were performed in GraphPad Prism 9. Enzymatic activities, plant growth-promoting characteristics, and auxin production were evaluated in three independent experiments, and results were compared between fungal genera using an ANOVA test with Tukey's post hoc. For bioassays with A. thaliana and maize plants, Dunnett's test was conducted to compare the yeast treatments and the control without inoculation. A value of p < 0.05 was considered to be statistically significant.

Conclusions
This study showed yeast diversity in different maize tissues of two Mexican landraces, with a predominant isolation of Clavispora lusitaniae. A high frequency (32%, 12 isolates) of Rhodotorula glutinis was observed in red maize and Papiliotrema flavescens in blue maize (20%, 10 isolates). Solicoccozyma sp. RY31, C. lusitaniae Y11, R. glutinis Y23, and Naganishia sp. Y52 were efficient auxin producers, and when inoculated in A. thaliana, they induced