*2.7. Dynamic Variations of Expression Levels of Flavonoid Biosynthetic Genes*

The metabolic pathways of three common flavonoids in *A. roxburghii* are shown in Figure 8. qRT-PCR was used to assess the relative expression levels of nine key enzyme genes in the conserved flavonoid biosynthesis pathway in every treatment group. The results showed that the expression of the *4CL* gene in the treatment group was significantly upregulated during the whole growth process, compared with the control group (Figure 9). The expression of the *PAL* and *CHS* genes in the M was significantly upregulated at the same growth stage (2 months, 3 months, and 4 months), while it was almost the same at other growth stages. The AR2 showed an inhibitory effect on the expression of the *C4H* and *F3 H* genes in the 1 month old and 2 month old *A. roxburghii*, while it showed an inhibitory effect on the expression of the *CHI* gene in the 1 month old, 5 month old and 6 month old *A. roxburghii*. The *F3H*/*FLS* gene had an upregulated expression in the M grown for 1 month and 2 months and the *GT* gene also had an upregulated expression in the M grown for 1 month, 2 months, 3 months, 5 months, and 6 months. Finally, the expression of the *RT* gene was upregulated in the M grown for 1 month and 6 months. In conclusion, AR2 could induce the expression of flavonoid synthesis related genes in varying degrees and at different growth stages.

**Figure 8.** The flavonoid biosynthetic pathway in *A. roxburghii*. Bold words indicate the key enzymes in flavonoid biosynthesis. Compounds in the box show flavones, flavonols and flavonol-glycosides studied in this study.

**Figure 9.** Expression levels of flavonoid biosynthetic genes in the M and NM growth 0 month to 6 months. M and NM represent the mycorrhizal *A. roxburghii* and non-mycorrhizal *A. roxburghii*, respectively. DNxx\_Cx\_g1 represents the ID of gene. The x-axis indicates the relative expression level of the genes. Each value is the mean of three replicates, and error bars indicate standard deviations. Statistical analysis of the data was performed by independent samples *t*-test using the SPSS 22.0 software (IBM, Chicago, IL, USA). \* and \*\* above the columns are significantly different at *p* ≤ 0.05 and *p* ≤ 0.01, respectively.

#### **3. Discussion**

Symbiotic association of mycorrhizal fungi with plants has been shown to affect flavonoid content. The application of co-cultivation of plants with orchid mycorrhizal (OM) fungi and arbuscular mycorrhizal (AM) fungi are progressing gradually. For AM fungi, they can penetrate and colonize the root of the host to form intracellular haustoria-like structures known as arbuscules, which are the principal sites of metabolic exchange between the two organisms [39,40]. Flavonoid content in *Medicago truncatula* was increased by AM inoculation [41]. Xie et al. [42] reported that AM colonization in the soybeans attributed to the increase of certain flavonoids in the root exudates. For OM fungi, they can form pleloton inside the root cells. And the association of OM with the orchid is the focus of our laboratory. Some studies indicated that flavonoid accumulated significantly in the mycorrhizal orchidaceae [16,43]. However, the dynamic changes of flavonoid content in mycorrhizal host at different developmental stages have rarely been studied. In our study, AR2 was belonging to a member of the orchid mycorrhizal fungi, and the co-cultivation between *A. roxburghii* and AR2 was performed. The dynamic changes of several flavonoids showed that the flavonoid had its special accumulation content at a defined growth time and that AR2 had different effects on different flavonoids at different growth stages. Also, AR2 induced the narcissin, rutin and quercetin-7-o-glucoside accumulations in mycorrhizal plantlets across the growth stage. For narcissin and isorhamnetin-3-o-beta-d-glucoside, their content in mycorrhizal *A. roxburghii* growth at three months reached the highest and was more than 420 ng/g and 120 ng/g, respectively. This was the first report regarding the changes of flavonoid content induced by AR2 in *A. roxburghii* at different growth stages.

To investigate the effects of mycorrhizal fungi on metabolites in its host, transcriptome and metabolome analyses were performed. Zhao et al. [29] reported that the secondary biosynthesis and hormone balance in the *Cymbidium hybridum* were induced by mycorrhizal fungus through transcriptome analysis. Schliemann et al. [44] reported that the biosynthesis of some constitutive isoflavonoids and plastidial metabolism could be activated by mycorrhizal fungus *Glomus intraradices* through metabolome analyses. In our study, metabolome analysis revealed that all 709 metabolites and 135 DAMs were putatively annotated among the NM and M. Among them, 148 flavonoid metabolites and 9 flavonoid DAMs were investigated. Furthermore, transcriptome analysis revealed that 4341 DEGs were identified between the two groups, of which 2915 DEGs were up-regulated and 1426 DEGs were down-regulated; KEGG pathways of the more DEGs were involved in the biosynthesis of secondary metabolites including flavonoid. These results implied that AR2 might change internal metabolism in *A. roxburghii*, especially for flavonoids, which would provide a basis for further study on the molecular mechanisms of AR2 promoting the flavonoid accumulation in *A. roxburghii*.

PAL, a key enzyme in the first step of the phenylpropanoid biosynthetic pathway, could be activated by fungal elicitors. Our study also revealed that the *PAL* gene had a significant upregulation, especially in the 2, 3, and 4 month mycorrhizal herbs, compared with the uninoculated ones. This result is in agreement with the results of Zhou et al. [45] and Xu et al. [46]. It is worth mentioning, the expression of the *4CL* gene in the plantlets inoculated AR2 during the whole growth process was significantly upregulated, with the highest expression being 13.3 fold. This is also consistent with Wang et al.'s report [20]. These data imply that AR2 might activate the downstream pathways of phenylpropanoids including flavonoids.

In addition, CHS, is the key enzyme in the flavonoids synthesis pathway [47]. Harrison and Dixon [48] reported that the expression level of the gene *CHS* in the roots of *Medicago truncatula* was enhanced by mycorrhizal fungus *Glomus versiforme*. Xie et al. [49] reported that mycorrhizal symbiosis induced the expression of the *CHS* gene of *Glycyrrhiza uralensis*, and the liquiritin accumulation and the expression of *CHS* gene showed a positive correlation. In our study, the expression level of the *CHS* gene was also upregulated in the mycorrhizal herbs growth for 2, 3, and 4 months; Meanwhile, the corresponding flavonoids (narcissin, rutin, isorhamnetin and quercetin-7-O-glucoside) accumulated in different degrees. Our data added new evidence to support mycorrhizal symbiosis induced the expression of the *CHS* gene and promoted the flavonoids accumulation. Additionally, our study showed that the *GT* gene expression was significantly upregulated in the 1–4 month mycorrhizal herbs, while the *RT* gene was induced in the 1–6 month ones. The corresponding flavonol-glycoside (narcissin, rutin, isorhamnetin-3-O-beta-d-glucoside, quercetin-7-O-glucoside and kaempferol-3-O-glucoside) showed basically the same induction trend in mycorrhizal *A. roxburghii*. These data again indicated that AR2 might activate the metabolic pathway of flavonoids.

In summary, this study provides much information about the changes that occur in the main active ingredient flavonoids and its related genes during different growth stages in M and NM. AR2 has different induction effects on flavonoid content and gene expression in *A. roxburghii* at different growth stages. These will provide a theoretical basis for reasonable harvest time of *A. roxburghii* and a new insight into improving the quality of the *A. roxburghii*.
