Mycorrhizal trifoliate orange has greater root adaptation of morphology and phytohormones in response to drought stress

Plant roots are the first parts of plants to face drought stress (DS), and thus root modification is important for plants to adapt to drought. We hypothesized that the roots of arbuscular mycorrhizal (AM) plants exhibit better adaptation in terms of morphology and phytohormones under DS. Trifoliate orange seedlings inoculated with Diversispora versiformis were subjected to well-watered (WW) and DS conditions for 6 weeks. AM seedlings exhibited better growth performance and significantly greater number of 1st, 2nd, and 3rd order lateral roots, root length, area, average diameter, volume, tips, forks, and crossings than non-AM seedlings under both WW and DS conditions. AM fungal inoculation considerably increased root hair density under both WW and DS and root hair length under DS, while dramatically decreased root hair length under WW but there was no change in root hair diameter. AM plants had greater concentrations of indole-3-acetic acid, methyl jasmonate, nitric oxide, and calmodulin in roots, which were significantly correlated with changes in root morphology. These results support the hypothesis that AM plants show superior adaptation in root morphology under DS that is potentially associated with indole-3-acetic acid, methyl jasmonate, nitric oxide, and calmodulin levels.

Root morphology. Root morphology of five-month-old trifoliate orange seedlings was significantly reduced by the DS treatment, compared with the WW treatment, while it was increased by mycorrhizal inoculation under the same soil water status as compared with non-mycorrhizal treatment (Table 2; Fig. 3). AM seedlings represented significantly (P < 0.05) higher root length, projected area, surface area, average diameter, volume, tips, forks, and crossings than non-AM seedlings, regardless of whether the seedlings were subjected to DS: 13%, 8%, 9%, 5%, 66%, 323%, 91%, and 107% increase under WW, and 10%, 7%, 8%, 7%, 48%, 201%, 83%, and 79% under DS, respectively (Table 2). Figure 4 showed the change in root hairs in response to the DS and mycorrhizal inoculation. DS treatment significantly reduced root hair density, while it notably (P < 0.05) increased root hair length and had no effect on root hair diameter, compared with WW (Fig. 5). Compared with the non-AM seedlings, root hair density in mycorrhizal colonized seedlings was significantly (P < 0.05) decreased by 12% under both WW and DS (Fig. 5). Root hair length was significantly (P < 0.05) 22% lower in AM seedlings than in non-AM seedlings under WW, whereas it was significantly 7% higher in AM seedlings than in non-AM seedlings under DS. Root hair length was not affected by AM fungal colonization, irrespective of the soil water status.   Root endogenous phytohormone levels. Mycorrhizal inoculation and DS did not significantly (P < 0.05) alter the root brassinosteroid (BR) and gibberellins (GAs) concentrations (Fig. 6). The DS treatment induced lower accumulation of root IAA and MeJA than the WW treatment, irrespective of the AM status ( Fig. 6). However, compared with non-mycorrhizal seedlings, root IAA concentration in mycorrhizal seedlings was increased, by 25% under WW and by 19% under DS. Root MeJA concentration was 24% and 11% higher under WW and DS in AM seedlings than in non-AM seedlings, respectively.

Root hair development.
Root calmodulin (CaM) and nitric oxide (NO) concentration. The DS treatment significantly (P < 0.05) decreased the root calmodulin (CaM) and nitric oxide (NO) concentration in both AM and non-AM seedlings, as compared with the WW treatment ( Fig. 7a,b). Compared with the non-AM fungal colonized controls, D. versiformis-colonized seedlings exhibited significantly (P < 0.05) higher root CaM concentration by 6% and 11% under WW and DS (Fig. 7a), and 219% and 117% higher root NO concentration under WW and DS (Fig. 7b), respectively.
Correlation studies. Leaf Ψ and root IAA, MeJA, CaM, and NO, but not BR and GAs, showed a significantly (P < 0.01) positive correlation with the number of 1 st -, 2 nd -, and 3 rd -order lateral roots, root morphological traits, and root hair density. In contrast, they showed a significantly (P < 0.01) negative correlation with root hair length and had no significant correlation with root hair diameter (Table 3).

Discussion
In our study, inoculation with D. versiformis strongly stimulated the formation in 1 st -, 2 nd -, and 3 rd -order lateral roots in trifoliate orange subjected to both WW and DS. This is in agreement with earlier studies reported in Medicago truncatula colonized by Gigaspora margarita 27 . Such mycorrhizal roots became progressively more branched than non-mycorrhizal controls 26 . This was confirmed by an increase in the number of tips, forks, and crossings in the present study. This shows that AM plants had a stronger adaptation in terms of lateral root formation than non-AM plants under DS. Mycorrhizal mycelium is mainly localized in the newly formed lateral roots 39 . Greater number of lateral roots in AM plants would provide better chances of colonization by AM fungi, as well as for absorbing water from the soil to the host plant, as shown in our study through a significant correlation between leaf Ψ and the number of lateral roots. A diffusible factor, 'Myc factor' , from AM fungi, was found to stimulate lateral root formation 27 . Our study showed that root IAA, MeJA, CaM, and NO were involved in lateral root formation. As reported by Felten et al. 40 , mycorrhiza-stimulated lateral root induction is paralleled by an increase in the 1-naphthylphthalamic acid (NPA)-sensitive auxin response at the root apex and in provascular tissues, together with IAA-based auxin signalling.
The present study showed that exposure to DS for six weeks dramatically restricted root morphological development in trifoliate orange seedlings, irrespective of the AM status. This indicates that the volume of soils explored by the whole root systems was dramatically reduced by DS in soils, which is in agreement with earlier studies on rice 3 . However, the D. versiformis that colonized trifoliate orange seedlings caused an improvement in morphological traits (length, area, volume, diameter, tips, forks, and crossings), compared with the non-AM controls under both WW and DS, indicating a larger extensive distribution of roots in AM seedlings under DS, which can contribute to higher leaf Ψ than non-AM plants 41 . The improvement of the root system by mycorrhization was also found in split-root trifoliate orange seedlings colonized by Acaulospora scrobiculata and F. mosseae, grown in a two-chambered split-root system 42 . In general, root morphology, development, and physiology are closely connected with plant growth 41 . A better root morphology in AM plants would promote plant growth, as shown by the positive effect of AM fungi on plant growth performance in trifoliate orange under both WW and DS. As roots have the ability to grow toward the direction of high water availability in the soil 2 , a significantly positive correlation of root morphological traits with leaf Ψ suggests that roots of mycorrhizal plants can explore the soil better for water uptake under DS. As a result, AM plants possess a stronger capacity, owing to a better root morphology, to adapt to soil drought than non-AM plants.
Wu et al. 31 showed that four AM fungal species (Claroideoglomus etunicatum, D. versiformis, F. mosseae, and R. intraradices) induced a significant increase in root hair density in trifoliate orange seedlings grown in sands. The present study further indicated a strong effect of AM fungal inoculation on the root hair density of trifoliate orange seedlings grown in soils under both WW and DS. Li et al. 43 reported that AM fungi mainly enhanced plant drought tolerance by the improvement of P and leaf water status, but root hairs presumably contributed to the shoot P enhancement. The AM-mediated increase in root hair density does not depend on growth substrates and soil water status. On the other hand, AM fungal inoculation affected the root hair length with a significant decrease under WW and with a significant increase under DS, suggesting that the the effect of AM fungus on root    versiformis under well-watered and drought stress conditions. Data (means ± SD, n = 4) followed by different letters above the bars indicate significant differences (P < 0.05) between treatments. Abbreviations: same as for Fig. 1 31 . We also concluded that higher concentrations of IAA, MeJA, NO, and CaM in roots under mycorrhization stimulate AM fungal colonization under WW and also improve root morphology under DS, indicating different functionings of these phytohormones in mycorrhizal plants exposed to water treatments. Earlier studies confirmed a negative effect of GAs on root mycorrhizal colonization 45,46 and a limited role of BR in determining AM development 31 . In this study, both AM fungal inoculation and water treatment produced no changes in the root BR and GAs concentration of trifoliate orange seedlings, suggesting that BR and GAs are not key factors involved in mycorrhizal development.
The root IAA level, in this study, was significantly decreased by DS while it was increased by AM fungal colonization, regardless of the soil water status. The mycorrhiza-induced root IAA increase might be due to small amounts of IAA that could be released by spores of AM fungi 47 . IAA, the major endogenous auxin in plants, is required for root development and root hair formation 48 . Moreover, auxins participate in the induction of cellular rhizogenic competence, root apical meristem differentiation, and development of the root cap and vasculature 49 . A mycorrhiza-induced IAA increment in roots would improve root morphology and lead to root hair modification. Therefore, it is reasonable that root IAA was significantly and positively correlated with root morphological traits and root hair density. Root IAA was negatively correlated with root hair length, whereas root hair length was significantly decreased by AM fungi under WW and increased under DS, suggesting that AM-induced IAA changes are dependent on soil water status and the integrated effects of several phytohormones, besides IAA. In addition, auxin accumulation could be involved in plant responses to DS via the activation of signalling pathways and induction of auxin-responsive genes 50 . The increase in the root IAA by mycorrhization has the potential capacity to enhance drought tolerance of the host plant.
In addition to IAA, NO participates in the induction of root tip elongation and the formation of lateral roots 51 . It is also a critical molecule for root hair formation through the auxin-signalling cascade 52 . In this study, in spite of the negative effect of DS on root NO levels, root NO concentrations were significantly increased in response to AM fungal inoculation under both WW and DS. This is consistent with the results of Calcagno et al. 53 in Medicago truncatula colonized by Gigaspora margarita and Espinosa et al. 54 in olive seedlings inoculated with R. irregularis. It has been established that NO is involved in root mycorrhizal colonization. Furthermore, root NO was significantly and positively correlated with the number of lateral roots, root morphological traits, and root hair density. This implies that mycorrhizal colonization heavily stimulated root NO production, which might play a role in the formation of lateral roots and root hairs. The NO-mediated effect on roots is under the control of auxins 52 . Further works are needed to clarify the mycorrhiza-induced root improvement by NO-mediated auxin signalling 55 . In addition, induction of root NO accumulation under mycorrhization might be a part of the mechanism producing a local response to enhance drought tolerance in plants 51 .
Methyl jasmonate (MeJA) has been identified as a vital cellular regulator to mediate developmental processes (including root hair production), proper arbuscule formation 44 , and defence responses against stresses (e.g., drought) 56 . Our study indicated that the root MeJA concentrations in trifoliate orange seedlings were pronouncedly increased by the AM fungal inoculation, irrespective of the soil water status. This is in agreement with previous studies on barley, soybean, and trifoliate orange plants under mycorrhization 31,57,58 . The root MeJA concentration was significantly and positively correlated with root hair density, but was negatively correlated with root hair length, which is in agreement with our previous work on trifoliate orange 31 . The increase in root MeJA concentration upon mycorrhizal inoculation might be associated with improving drought tolerance of the host plant, as previously mentioned by Anjum et al. 59 .
An earlier study had reported an increase in the intracellular CaM in soybean after being colonized by R. intraradices 60 . The present study also indicated that inoculation with D. versiformis induced the accumulation of root CaM under both WW and DS. Similar results were found in trifoliate orange seedlings colonized by F. mosseae under both WW and DS 16 . It seems that CaM is involved in the process of root mycorrhizal colonization. Lévy et al. 61 further confirmed that Ca 2+ spiking and CaM-dependent protein kinases are necessary for mycorrhizal infection. Calmodulin (CaM), a Ca 2+ receptor, can bind with Ca 2+ as the Ca 2+ /CaM messenger system to activate downstream target proteins, thereby regulating the generation of reactive oxygen species and modulating transcription factors to maintain homeostasis between different cellular processes 62 . In the study of Huang et al. 16 , mycorrhiza-induced CaM mediated the antioxidant enzyme defence system to enhance drought tolerance in plants. Higher root CaM levels in AM plants have the capacity to enhance drought tolerance of the host plant. CaM is a primary decoder of Ca 2+ signals 63 , while the Ca 2+ gradient and Ca 2+ influxes induce root hair formation and growth 64 . Root CaM was significantly and positively correlated with root hair density while negatively correlated with root hair length.
In short, our results supported the preceding hypothesis that AM plants had better root adaptation of morphology in response to drought stress, which is potentially associated with AM-induced changes in IAA, MeJA, NO, and CaM. It is therefore suggested that in citriculture, either stimulating the formation and development of AMs or inoculating native AM fungi into citrus orchards, would be vital for the enhancement of drought tolerance and root morphology in citrus trees.

Methods
Plant culture. Seeds of trifoliate orange were sterilized by 70% of ethanol solution for 10 min, rinsed 4 times with distilled water, and pre-germinated in autoclaved (121 °C, 0.11 Mpa, 2 h) river sand under the condition of 27/20 °C day/night temperature, 740 μ mol/m 2 /s photon flux density, and 80% relative humidity. After 4 weeks, three five-leaf-old seedlings were transplanted into a plastic pot (11.5  characteristics of pH 6.0, 12.1 mg/kg KMnO 4 -N, 15.7 mg/kg Bray-P, and 22.3 mg/kg neutral NH 4 OAc-K. Half of the seedlings received the AM fungal inoculated treatment. A 2200-spore dosage of an AM fungus, Diversispora versiformis (P. Karst.) Oehl, G. A. Silva & Sieverd, was applied into the surroundings of the plant roots. The AM fungus was supplied by the Bank of Glomeromycota in China (BGC). The AM fungus was isolated from the rhizosphere of Astragalus adsurgens in Ejin Horo Banner, Inner Mongolia Autonomous Region, China. The spores of the AM fungus were propagated by white clover in a mixture of sand and soil (1:1, v/v) for 12 weeks. For the non-AM fungal treatment, seedlings were supplied the same amount of autoclaved AM fungal inocula plus 2 mL filtrate (25 μ m filter) of mycorrhizal inoculum to minimize differences in other microbial communities. AM and non-AM seedlings were placed in a green house on the Yangtze University campus, where the photosynthetic photon flux density was 880 μ mol/m 2 /s, day/night temperature 28/21 °C, and relative humidity 85%. During the experiment, a 30 mL Hoagland solution per pot was used to replace 30 mL distilled water weekly.
Water treatments. After AM fungal inoculation, the seedlings were gravimetrically maintained at 75% of maximum water holding capacity (soil WW status) in the growth substrate for 15 weeks. Subsequently, half of AM and non-AM seedlings were changed to 50% of maximum water holding capacity (soil DS status) in the growth substrate for DS for 6 weeks. The other seedlings were still kept in soil WW status for 6 weeks. After 21 weeks of water treatments, seedlings were harvested. Experimental design. The experiment was composed of AM fungal inoculations (with or without D. versiformis) and water treatments (WW and DS) with a completely randomized arrangement, for a total of four treatments. Each treatment was replicated four times, for a total of 16 pots.