Arbuscular mycorrhizal fungi and production of secondary metabolites in medicinal plants

Medicinal plants are an important source of therapeutic compounds used in the treatment of many diseases since ancient times. Interestingly, they form associations with numerous microorganisms developing as endophytes or symbionts in different parts of the plants. Within the soil, arbuscular mycorrhizal fungi (AMF) are the most prevalent symbiotic microorganisms forming associations with more than 70% of vascular plants. In the last decade, a number of studies have reported the positive effects of AMF on improving the production and accumulation of important active compounds in medicinal plants. In this work, we reviewed the literature on the effects of AMF on the production of secondary metabolites in medicinal plants. The major findings are as follows: AMF impact the production of secondary metabolites either directly by increasing plant biomass or indirectly by stimulating secondary metabolite biosynthetic pathways. The magnitude of the impact differs depending on the plant genotype, the AMF strain, and the environmental context (e.g., light, time of harvesting). Different methods of cultivation are used for the production of secondary metabolites by medicinal plants (e.g., greenhouse, aeroponics, hydroponics, in vitro and hairy root cultures) which also are compatible with AMF. In conclusion, the inoculation of medicinal plants with AMF is a real avenue for increasing the quantity and quality of secondary metabolites of pharmacological, medical, and cosmetic interest.


Introduction
Medicinal plants have been a valuable source of therapeutic agents to treat various ailments and diseases such as diarrhea, fever, colds, and malaria since ancient times (Dambisya and Tindimwebwa 2003;Ghiaee et al. 2014;Mathens and Bellanger 2010;Titanji et al. 2008). Nowadays, they also represent a source for the development of new drugs to cure important diseases such as cancer (Newman and Cragg 2007;Beik et al. 2020). Their therapeutic value often is attributed to the presence and richness of active compounds belonging to the secondary metabolism, such as alkaloids, flavonoids, terpenoids, and phenolics (Hussein and El-Anssary 2018). Today, up to 80% of people in developing countries are totally dependent on herbal drugs for their primary healthcare (Ekor 2014), and over 25% of prescribed medicines in developed countries have been derived from plants collected in the wild (Hamilton 2004).
Numerous methods, such as isolation from plants and other natural sources, synthetic chemistry, combinatorial chemistry, and molecular modeling, have been used for drug discovery (Ley and Baxendale 2002;Geysen et al. 2003;Lombardino and Lowe 2004). However, natural products, and particularly medicinal plants, remain an important source of new drugs, new drug leads, and new chemical entities (Newman et al. 2000(Newman et al. , 2003Butler 2004) because of their cultural acceptability, high compatibility, and adaptability with the human body compared to synthetic chemicals (Garg et al. 2021). According to the International Union for Conservation of Nature and the World Wildlife Fund (Chen et al. 2016), an estimate of as many as 80,000 flowering plant species are used for medicinal purposes. For several thousands of plants worldwide, the activity or composition in bioactive compounds remains poorly documented, requiring further in-depth analysis to fully exploit their medicinal potential (Ali 2019).
In nature, plants are associated with an overwhelming number of beneficial microorganisms (e.g., endophytic or symbiotic bacteria and fungi) that play a significant role in plant health, development, and productivity, and in the modulation of metabolite synthesis (Berendsen et al. 2012;Panke-Buisse et al. 2015;Mendes et al. 2011;Castrillo et al. 2017;de Vries et al. 2020;Brader et al. 2014;Compant et al. 2021). Among these are the arbuscular mycorrhizal fungi (AMF), a ubiquitous group of soil microorganisms, forming symbiosis with more than 70% of vascular plants (Brundrett and Tedersoo 2018). Arbuscular mycorrhizas are characterized by the formation of finely branched structures called arbuscules within root cortical cells of host plants (Coleman et al. 2004), which are the site of bidirectional transport, i.e., minerals from the fungal cell to the plant cell and carbon compounds in the opposite direction.
The establishment of the AMF symbiosis requires recognition between the two partners. Lipochitooligosaccharides, the so-called Myc factors, are perceived by the plant in response to signaling molecules (i.e., strigolactones) released by the roots (Akiyama and Hayashi 2006). After reciprocal recognition, AMF hyphae form a hyphopodium on the root epidermis and colonize the root cortex. At the same time, fungal hyphae spread into the surrounding soil as an extensive extraradical mycelium, representing 9 to 55% of the total soil microbial biomass (Olsson et al. 1999). This dense extraradical mycelium considerably enhances the access of roots to water and mineral nutrients (e.g., P, N, K, Ca, S, Zn, Cu), often increasing plant biomass (Smith and Read 2008;Bowles et al. 2016) and quality of crops (Baum et al. 2015;Bona et al. 2016;Noceto et al. 2021). Moreover, this extraradical mycelium modifies the soil structure , which improves soil quality and fertility (Zou et al. 2016;Thirkell et al. 2017). AMF also are well known to improve plant resistance or tolerance to stress conditions, such as drought, salinity, nutrient deprivation, extreme temperatures, heavy metals, pests, and diseases (Ahanger et al. 2014;Salam et al. 2017;Porcel et al. 2012;Cicatelli et al. 2014). In addition to these benefits, they also quantitatively and qualitatively could affect the production of secondary metabolites produced by their hosts (Ahanger et al. 2014;Salam et al. 2017;Porcel et al. 2012;Cicatelli et al. 2014;Kaur and Suseela 2020). Taber and Trappe (1982) were the first to document the presence of AMF in a medicinal plant (in their study conducted on ginger growing in the Fiji Islands and Hawaii). Since then, most medicinal plants were found capable of associating with mycorrhizal fungi (Chen et al. 2014). Recently, single or combinations of AMF have been inoculated to various medicinal plants to investigate their impact on plant biomass as well as on phytochemical constituents in seeds, fruits, leaves, shoots, and roots (e.g., Rydlová et al. 2016;Kapoor et al. 2004;Selvaraj et al. 2009;Dave et al. 2011;Zubek et al. 2012). The majority of studies revealed that AMF were able to enhance plant biomass as well as to promote the accumulation of several active compounds. For example, Lazzara et al. (2017) reported an increased above-and belowground biomass in Hypericum perforatum associated with a mixture of nine different AMF species. Interestingly, the concentrations of pseudohypericin and hypericin, two anthraquinone derivatives that exhibit important photodynamic, antiviral, antiretroviral, antibacterial, antipsoriatic, antidepressant, and antitumoral biological activities (Zubek et al. 2012;Bombardelli and Morazzoni 1995;Gadzovska et al. 2005;Guedes and Eriksson 2005), were increased by 166.8 and 279.2% in the AMF-colonized plants as compared to non-mycorrhizal controls (Lazzara et al. 2017). However, these results should not obviate other studies in which no effects on biomass were reported. For instance, Nell et al. (2010) found that AMF colonization decreased the biomass of rhizomes and roots of Valeriana officinalis, while significantly increasing the levels of sesquiterpenic acids. Another study by Engel et al. (2016) reported an increased content of rosmarinic acid and lithospermic acid A isomer (two phenolic compounds) in Melissa officinalis, while both compounds were diminished in Majorana hortensis, in the presence of three mixtures of AMF. More recently, Duc et al. (2021) showed that a mixture of different AMF species improved the salt stress tolerance of Eclipta prostrata, inducing major changes in polyphenol profile.
In this publication, we provide a thorough review of the literature on AMF mediation of secondary metabolites production in medicinal plants. We also review the different methods that are used to increase/stabilize the production of secondary metabolites. Indeed, the quantity and quality of secondary metabolites obtained from plants grown in natural habitats are critically influenced by various abiotic and biotic stresses (e.g., drought, extreme temperatures, and pathogen attack). This results in high variability of bioactive substances and influences the metabolic pathways responsible for the accumulation of the related natural compounds (Dayani and Sabzalian 2017;Giurgiu et al. 2017;Ramakrishna and Ravishankar 2011). Therefore, we additionally review the most widely used methods of cultivation (i.e., greenhouse, hydroponics, aeroponics, in vitro and hairy root cultures (HRCs)) of medicinal plants, and we investigate their possible application to AMF to further increase the quantity and quality of secondary metabolites produced.

Effect of AMF on growth and secondary metabolite production of medicinal plants
Since the pioneer work of Wang (1989, 1991), reporting the positive effect of AMF inoculation of Datura stramonium and Schizonepeta tenuifolia on the production of active compounds, numerous studies have been conducted.
The literature focusing on AMF in medicinal plants involves 81 plant species belonging to 28 families (Table 1). These medicinal plants present different characteristics to be studied with AMF: important medicinal herbs to treat certain disease such as Artemisia annua producing artemisin to treat malaria in developing countries (Domokos et al. 2018); important condiment plants such as Allium sativum in India (Borde et al. 2009); rare plant species difficult to culture such as Arnica montana (Jurkiewicz et al. 2011); aromatic plants to produce essential oil widely used in the pharmaceutical, cosmetic, and food industries such as most plant species from the Apiaceae (Anethum graveolens and Coriandrum sativum (Rydlová et al. 2016)) and the Lamiaceae (Origanum vulgare (Karagiannidis et al. 2012)); and health foods such as Dioscorea spp. yam (Lu et al. 2015). For the majority of these plants, studies were focused on the effects of AMF on biomass increase and production of bioactive compounds simultaneously (45 studies) or only focused on bioactive compound production (30 studies). For a few other studies, the attention was focused on AMF community composition (four studies) or on the effects of AMF on plant growth under different conditions (15 studies) (Table 1). Among the AMF species tested, Funneliformis mosseae 1 is the most investigated one (25 studies), followed by Rhizophagus intraradices (16), Claroideoglomus etunicatum (14), Rhizophagus fasciculatus (14), Rhizophagus irregularis and Rhizophagus clarus (six studies each), and Gigaspora margarita (five studies) (Table 1). Only a few medicinal plant species were inoculated with AMF present in the soil native to those plants, while the vast majority were inoculated with commercial inoculants.
Another beneficial aspect of AMF is their capability to improve plant nutrient uptake (Bowles et al. 2016), influencing directly or indirectly the concentration of secondary metabolites (Yamawaki et al. 2013). For instance, F. mosseae improved shoot and root biomass, root system architecture, and flavonoid accumulation in Glycyrrhiza uralensis growing under P-deficient nutrient conditions .
A number of studies also have reported enhanced survival and increased growth of micropropagated medicinal plants at the transfer stage from in vitro to ex vivo conditions (Rai 2001). For instance, with F. mosseae, the survival rate of micropropagated Spilanthes acmella and Glycyrrhiza glabra plantlets was 100%, and plant growth and development were improved under glasshouse and greenhouse conditions (Yadav et al. 2012(Yadav et al. , 2013 while in the absence of AMF, the survival rate was only 60-70%. Similarly, height and fresh weight of shoots, roots, and seeds of Scutelleria integrifolia seedlings inoculated with C. etunicatum were significantly increased in pots following micropropagation (Joshee et al. 2007).
These studies clearly evidenced the potential of using AMF inoculants for improving the yield of raw materials (e.g., roots, shoots) of medicinal plants, thus potentially increasing the quantity of active compounds.
Different groups of secondary metabolites whose production was enhanced by AMF inoculation are detailed below.

Alkaloids
Alkaloids are nitrogen-containing organic compounds produced by plants constitutively or in response to pests, diseases, or other external stimuli ). They are found in different organs of important medicinal plants (Table 1) and are characterized by a diverse array of pharmacological properties including analgesia, local anesthesia, cardiac stimulation, respiratory stimulation and relaxation, vasoconstriction, muscle relaxation, antineoplastic, and hypertensive and hypotensive properties (Hussein and El-Anssary 2018).
Since Wei and Wang (1989) first observed that AMF symbiosis can increase the total content of hyoscyamine and scopolamine in Datura stramonium, numerous studies have reported a positive role of AMF in the accumulation of alkaloids. For example, a positive correlation was found between AMF colonization (a mixture of R. intraradices The column "AMF" shows the current names, not the one at the time of publication b There are no studies or available data found online Table 1 (continued) and G. margarita) of Castanospermum australe tree and the castanospermine content (which was reported to inhibit the HIV virus) of leaves (Abu-Zeyad et al. 1999). The contents of some commonly used "heat-clearing" herb compounds, such as berberine, jatrorrhizine, and palmatine, were increased in seedlings of Phellodendron amurense inoculated with AMF (Fan et al. 2006). Other active compounds were increased in the presence of AMF: trigonelline in roots and leaves of Prosopis laevigata colonized by Gigaspora rosea under in vitro conditions; colchicine in tubers of Gloriosa superba colonized by F. mosseae growing under glasshouse conditions, and scopolamine in leaves of Eclipta prostrata colonized by a mixture of C. etunicatum, Claroideoglomus claroideum, F. mosseae, F. geosporum, R. irregularis, and Rhizoglomus microaggregatum growing in climate chamber conditions (Rojas- Andrade et al. 2003;Pandey et al. 2014;Vo et al. 2019).

Terpenoids
The largest and most diverse group of secondary metabolites are terpenoids, which are primary constituents of essential oils (Cox-Georgian et al. 2019). Essential oils are volatile lipophilic mixtures of secondary metabolites, consisting mostly of monoterpenes, sesquiterpenes, and phenylpropanoids, which often are used as flavors and fragrances, as antimicrobials and antioxidants, and as medicines (Deans and Waterman 1993).
Several studies have reported an AMF impact on the production of essential oils by medicinal and aromatic plants (Table 1). For instance, the production of these compounds was increased in Corianderum sativum, Trachyspermum ammi, Atractylodes lancea, Inula ensifolia, Artemisia umbelliformis, Plectranthus amboinicus, Satureja macrostema, Salvia officinalis, Origanum vulgare and Origanum onites, Thymus daenensis, Thymus vulgaris, and Foeniculum vulgare colonized by AMF (Rydlová et al. 2016;Kapoor et al. 2002;Liang et al. 2018;Zubek et al. 2010;Binet et al. 2011;Merlin et al. 2020;Carreón-Abud et al. 2015;Sete da Cruz et al. 2019;Karagiannidis et al. 2012;Arpanahi et al. 2020;Machiani et al. 2021;Kapoor et al. 2004). The content of artemisinin, an important sesquiterpene lactone compound found in Artemisia annua and well known for its effects on malaria and more recently on cancer (Krishna et al. 2008), was increased in leaves of plants colonized by F. mosseae or a combination of Glomus macrocarpum and R. fasciculatus or Diversispora epigaea and R. irregularis grown in pots or under field conditions (Huang et al. 2011;Chaudhary et al. 2008;Domokos et al. 2018). The forskolin content, a diterpene extensively used to treat heart diseases, glaucoma, asthma, and certain types of cancers (Kavitha et al. 2010), was significantly increased in roots of Coleus forskohlii inoculated with Glomus bagyarajii growing under greenhouse conditions (Sailo and Bagyaraj 2005). Similarly, Singh et al. (2013) reported an increased content of forskolin in tubers of Coleus forskohlii associated with R. fasciculatus growing under organic field conditions.
Researchers also have studied the impact of AMF symbiosis on medicinal plants derived from tissue cultures. An example is the increased content of the essential oil carvacrol, a phenolic monoterpenoid with antimicrobial, antioxidant, and anticancer activities (Sharifi-Rad et al. 2018) in micropropagated Origanum vulgare subsp. hirtum after association with the AMF Septoglomus viscosum (Morone Fortunato and Avato 2008).

Phenolics
Phenolics represent a wide group of compounds, sharing one or more phenol groups (Hussein and El-Anssary 2018), among which are flavonoids, curcuminoids, coumarins, tannins, stilbenes, lignans, phenolic acids, and quinones (Cosme et al. 2020). Arbuscular mycorrhizal fungi have been shown to increase the content of phenols in medicinal plants (Table 1). For instance, the production of formononetin (an antimicrobial, antioxidant, antilipidemic, antidiabetic, antitumor, and neuroprotective compound) (Vishnuvathan et al. 2016), was increased in Medicago sativa grown in the presence of R. intraradices (Volpin et al. 1994). The production of curcumin (an anti-inflammatory, antioxidant, anticancer, antiseptic, antiplasmodial, astringent, digestive, diuretic compound) was increased by circa 26% in Curcuma longa colonized by AMF species belonging to the genera Glomus/Rhizophagus, Gigaspora, and Acaulospora sp., under greenhouse conditions (Dutta and Neog 2016). The concentration of total tannins, used to treat tonsillitis, pharyngitis, hemorrhoids, and skin eruptions (Britannica 2021), was increased by 40% in the fruits of Libidibia ferrea inoculated with Acaulospora longula under field conditions (Santos et al. 2020). Additionally, the concentrations of cichoric acid in Echinacea purpurea colonized by R. intraradices (Araim et al. 2009) and p-hydroxybenzoic acid and rutin in Viola tricolor colonized by R. irregularis (Zubek et al. 2015), and the total content of flavonoids in Libidibia ferrea colonized by Gigaspora albida and gallic acid in Valeriana jatamansi colonized by a consortium of three different isolates of R. intraradices spp. (Silvia et al. 2014;Jugran et al. 2015) were increased by the AMF symbiosis.

Saponins
Saponins are characterized by a polycyclic aglycone moiety with either a steroid (steroidal saponins) or triterpenoid (triterpenoidal saponins) attached to a carbohydrate unit (a monosaccharide or oligosaccharide chain) (Hussein and El-Anssary 2018). Among these compounds, a few have demonstrated pharmacological properties, such as antitumor, sedative, expectorant, analgesic, and anti-inflammatory (Hussein and El-Anssary 2018). Arbuscular mycorrhizal fungi were reported to enhance the production of saponins in medicinal plants (Table 1). For instance, the content of glycyrrhizic acid, a triterpenoid saponin used to alleviate bronchitis, gastritis, and jaundice (Pastorino et al. 2018), was increased by 0.38-1.07-fold and by 1.34-1.43-fold after 4 and 30 months, respectively, in Glycyrrhiza glabra (liquorice) plants colonized by F. mosseae and D. epigaea alone or in combination, grown in sand under greenhouse conditions . Similarly, Johny et al. (2021) reported an increase of glycyrrhizic acid concentration in Glycyrrhiza glabra inoculated with C. etunicatum under greenhouse conditions.

Other chemical compounds
Hypericin and pseudohypericin are naphthodianthrones (anthraquinone derivatives) mainly extracted from Hypericum species (Ayan and Cirak 2008). They have many pharmaceutical properties, such as sedatives, antiseptics, and antispasmodics (Baytop 1999). Zubek et al. (2012) reported an increased content of hypericin and pseudohypericin in Hypericum perforatum colonized by R. intraradices alone or by a mixture of F. constrictum, F. geosporum, F. mosseae, and R. intraradices, under greenhouse conditions.
Withaferin-A, a steroidal lactone, traditionally used in ayurvedic medicine (Mirjalili et al. 2009), has a wide range of pharmacological activities including cardioprotective, antiinflammatory, immuno-modulatory, anti-angiogenesis, antimetastasis, and anti-carcinogenic properties. Johny et al. (2021) reported that association between the medicinal plant Withania somnifera and R. irregularis increased the concentration of withaferin-A as compared to non-inoculated plants under greenhouse conditions. It should be noted, however, that AMF showed a neutral or decreased effect on the production of certain secondary metabolites. For example, Nell et al. (2010) found that F. mosseae has no effect on the total concentrations of phenolic and rosmarinic acid in the roots of Salvia officinalis; and Geneva et al. (2010) showed that R. intraradices decreased total phenol and flavonoid contents in the leaves of Salvia officinalis. Similarly, Zubek et al. (2010) reported significant differences in the effectiveness of different AMF species tested in Inula ensifolia. An increased production of thymol derivatives was found in plant roots inoculated with Rhizophagus clarus, while a decreased production of these metabolites was reported in roots inoculated with R. intraradices under greenhouse conditions (Zubek et al. 2010). Moreover, changes in secondary metabolite composition have been observed in medicinal plants inoculated with AMF. For instance, Geneva et al. (2010) observed a modified composition of essential oils and promotion of the relative quantities of bornylacetate, 1,8-cineole, α-thujones, and β-thujones in Salvia officinalis associated with R. intraradices. Similarly, Artemisia umbelliformis inoculated with an alpine microbial community containing Planticonsortium tenue (formerly Glomus tenue), R. intraradices, G. claroideum/etunicatum, and a new Acaulospora species showed a significant increase of E-ocimene concomitant with a decrease of E-2-decenal and (E, E)-2-4-decadienal (Binet et al. 2011). Therefore, the selection of the most effective AMF strains for improving the accumulation of desirable active compounds needs to be taken into account.

Effect of AMF on biomass and production of secondary metabolites in medicinal plants under biotic and abiotic stress conditions
Drought, salinity, heavy metals, pests, and diseases can impact plant growth, reducing their biomass (Hashem et al. 2014;Alwhibi et al. 2017) and consequently affecting the production of secondary metabolites. Arbuscular mycorrhizal fungi can increase the tolerance/resistance of plants against those abiotic and biotic stresses, potentially influencing secondary metabolites production (Hashem et al. 2018).
Several studies have shown that AMF symbiosis can improve the growth and secondary metabolite production of medicinal plants under water deficit conditions. For example, a recent study by Machiani et al. (2021) showed that inoculation with F. mosseae significantly improved biomass and essential oil content (mainly thymol, p-cymene and γ-terpinene) of Thymus vulgaris plants grown in a 2-year field experiment in intercropping with soybean under water deficit conditions. Similarly, Mirzaie et al. (2020) reported that inoculation with F. mosseae significantly increased geranial and β-pinene (both belong to oxygenated monoterpenes essential oils) yields of Cymbopogon citratus grown in a greenhouse pot experiment under moderate water stress conditions (50% field capacity).
Salt stress stimulates the accumulation of phenolic compounds in plants as a general defense mechanism to stress (Parvaiz and Satyawati 2008). Intriguingly, this abiotic stress is a principal elicitor influencing synthesis of compounds in many herbs (e.g., cinnamic, gallic, and rosmarinic acids in Thymus vulgaris; glycyrrhizin in Glycyrrhiza glabra; quinic, gallic, and protocatechuic acids in Polygonum equisetiforme) (Bistgani et al. 2019;Behdad et al. 2020;Boughalleb et al. 2020). A recent study by Amanifar and Toghranegar (2020) reported that moderate salt stress stimulated higher production of valerenic acid in Valeriana officinalis than a situation without salt stress. Interestingly, this increase was significant when the plants were colonized by F. mosseae. Duc et al. (2021) found that a mixture of six AMF species (C. etunicatum, C. claroideum, F. mosseae, F. geosporum, Rhizoglomus microaggregatum, and R. intraradices) increased the tolerance of Eclipta prostrata under moderate salt stress in a pot experiment under controlled conditions, inducing major changes in its polyphenol profile.
Minerals, such as cadmium (Cd) and zinc (Zn), also were reported to impact secondary metabolite production in medicinal plants colonized by AMF. For instance, Hashem et al. (2016) observed that an AMF mixture comprising C. etunicatum, F. mosseae, and R. intraradices enhanced the chlorophyll and protein content and considerably reduced lipid peroxidation in Cassia italica plants under Cd stress in a pot experiment. Moreover, AMF inoculation caused a further increase in proline and phenol content ensuring improved plant growth under stress conditions. Arbuscular mycorrhizal fungi symbiosis improved the disease tolerance of medicinal plants through the mediation of secondary metabolites. For instance, Jaiti et al. (2007) reported that a complex of native AMF species increased the tolerance of Phoenix dactylifera (a plant characterized by high nutritional and therapeutic value of its fruits (Al-Alawi et al. 2017)) against bayoud disease (the most damaging vascular disease of date palm caused by Fusarium oxysporum f. sp. albedinis) by increasing the enzymatic activities of peroxidases and polyphenoloxidases, which are associated with an increase of phenolic compounds in the cell wall.

Mechanisms by which AMF symbiosis promotes secondary metabolism in medicinal plants
It is often considered that the increased concentrations of various secondary metabolite groups (e.g., flavonoids, phenolics) in AMF-colonized plants are a result of the elicitation of several defense response pathways as reviewed by Zeng et al. (2013). For instance, terpenoids in the carotenoid pathway, flavonoids, phenolic compounds, and some alkaloids (such as hyoscyamine and scopolamine) in the phenylpropanoid pathway are often increased in AMF-colonized plants (Kaur and Suseela 2020). These pathways play different roles in the plant-AMF symbiosis, such as signaling, stress tolerance, nutrient uptake, and resistance against biotic and abiotic stresses. However, it is still not totally clear how AMF trigger changes in the concentrations of phytochemicals in plant tissues (Toussaint et al. 2007).
Many studies have focused on the mechanisms by which AMF modulate the production of terpenoids, phenolic compounds, and alkaloids in plants. Terpenoids are synthesized from isoprene units in the methyleritrophosphate (MEP) and the mevalonic acid (MVA) pathways (Zhi et al. 2007).
Phenolic compounds (e.g., phenols, flavonoids, protanthocyanidins, tannins) are synthesized in the shikimic acid pathway where phenylpropanoids are formed and in the malonic acid pathway (Oksana et al. 2012). Most of the alkaloids are synthesized from various biological precursors (most amino acids) such as tyrosine and tryptophane in the shikimic acid pathway (Facchini 2001) (Fig. 1).
Several common nutritional and non-nutritional factors have been proposed to explain the increased production of secondary metabolites in AMF-colonized plants Sharma et al. 2017;Dos Santos et al. 2021) (Fig. 2).
Regarding nutritional factors, the increase was first attributed to the enhanced uptake of nutrients by AMF-colonized plants (Lima et al. 2015;Oliveira et al. 2015;Riter et al. 2014). For example, the role of phosphorus in the synthesis of terpenoids precursors via the MVA (acetyl-CoA, ATP, and NADPH) as well as the MEP (glyceraldehyde phosphate and pyruvate) pathways is widely recognized . Phosphorus enhances terpenoid biosynthesis by increasing the concentration of pyrophosphate compounds, such as isopentenyl pyrophosphate (IPP) and dimethylallyl pyrophosphate (DMAPP) (Kapoor et al. 2002(Kapoor et al. , 2004Zubek et al. 2010), which contain high-energy phosphate bonds. However, Khaosaad et al. (2006) found that the concentration of essential oils significantly increased in two Origanum sp. genotypes colonized by F. mosseae, while the levels of essential oils in plants treated with P did not change. This suggests that the increased concentration of essential oils in AMF-colonized Origanum sp. plants may directly depend on the association with the fungus. In another study by Zubek et al. (2012), AMF colonization improved hypericin and pseudohypericin concentrations in Hypericum perforatum, probably because of an improved plant P and/or N nutrition in presence of the fungi. The increased growth through improved nutrients and water uptake of AMF-colonized plants also explains the enhanced production of these compounds in plants. It is well known that the AMF symbiosis increases shoot biomass, shoot length, and number of nodes in Ocimum basilicum (Gupta et al. 2002;Khaosaad et al. 2008;Rasouli-Sadaghiani et al. 2010;Copetta et al. 2006). Elevated leaf biomass results in increased photosynthetic capacity (Dave et al. 2011;Zubek et al. 2010), thus increasing the production of total photosynthates (e.g., ATP, carbon substrate, glyceraldehyde-3-phosphate, pyruvate, phosphoenolpyruvate, or erythrose-4-phosphate) required for terpenoids, phenolics, and alkaloid biosynthesis (Cao et al. 2008;Hofmeyer et al. 2010;Niinemets et al. 2002).
Regarding non-nutritional factors, alterations in the levels of phytohormones in AMF-colonized plants may reflect their enhanced production (Mandal et al. 2013(Mandal et al. , 2015aZubek et al. 2012). Indeed, it has been shown that the AMF symbiosis changes the concentrations of phytohormones, such as jasmonic acid (JA), gibberellic acid (GA 3 ), and cytokinins (Allen et al. 1980(Allen et al. , 1982Hause et al. 2002;Shaul-Keinan et al. 2002) in plants. Moreover, it has been reported that phytohormones play a role in the secondary metabolism of plants Maes and Goossens 2010;Maes et al. 2008). For instance, JA has been reported to coordinate transcriptional activation of sesquiterpenoid biosynthetic gene expression in Artemisia annua (Maes et al. 2011). Furthermore, the phytohormonal alterations of GA 3 , BAP (6-benzylaminopurine), and JA have been reported to promote the formation of glandular trichomes (Maes et al. 2011) which is positively correlated with an enhanced concentration of terpenoids in plant leaves. Glandular trichomes are the epidermal secretory structures in which terpenoids are synthesized and stored in plants (Covello et al. 2007). The enhanced concentration of terpenoids (essential oils) and increased glandular trichome density has been observed in a number of plants (e.g., Mentha x piperita, Phaseolus lunatus, and Lavendula angustifolia) (Ringer et al. 2005;Bartram et al. 2006;Behnam et al. 2006). Thus, an increase in trichome density upon mycorrhization often has been linked with an enhanced concentration of terpenoids (Copetta et al. 2006;Kapoor et al. 2007;Morone-Fortunato and Avato 2008). The modification of these secondary metabolite concentrations in AMF-plants also may be due to signaling mechanisms between host plants and the fungi (Larose et al. 2002;Rojas-Andrade et al. 2003;Xie et al. 2018). For example, Zhang et al. (2013) have reported that F. mosseae associated with Trifolium repens promoted changes in the concentration of signaling molecules, such as nitric oxide, salicylic acid (SA), and hydrogen peroxide, which influence the activation of key enzymes in phenolics biosynthesis (e.g., l-phenylalanine ammonia lyase (PAL), and chalcone synthase (CHS)). Moreover, AMF may increase the expression of genes encoding enzymes leading to the biosynthesis of these compounds in mycorrhizal plants (Andrade et al. 2013;Battini et al. 2016;Mandal et al. 2015a, b;Xie et al. 2018). For example, induction of terpene synthase (TPS) family genes TPS31, TPS32, and TPS33 has been observed in AMF-colonized tomato plants and probably can explain the changes in their terpenoid profile (Zouari et al. 2014). Mandal et al. (2015a) reported the increase of artemisinin in leaves of Artemisia annua inoculated with R. intraradices. This result was correlated with a higher expression of key biosynthesis genes (such as an allene oxidase synthase gene encoding one of the key enzymes for JA production) via enhanced JA levels. In addition, AMF may enhance the biosynthesis of these compounds either by increasing the production of precursors through the induction of metabolic biosynthetic pathways (Lohse et al. 2005;Zimare et al. 2013;Dos Santos et al. 2021) and/or by induction of key synthase enzymes (Mandal et al. 2013;Shrivastava et al. 2015; Sharma et al. Fig. 1 Main pathways of secondary plant metabolism resulting in the production of alkaloids, phenolics, saponins, and terpenes (in gray, green, pink, and brown shaded portions, respectively) mentioned in this review. Examples of upregulated compounds or classes of compounds in medicinal plants associated with AMF are highlighted with green type. This figure is modified from Dos Santos et al. (2021) 2017; Dos Santos et al. 2021). For example, mycorrhizal colonization (R. intraradices) has been found to elevate the transcript levels of two of the pivotal enzymes of the MEP pathway, 1-deoxy-D-xylulose 5-phosphate synthase (DXS), and 1-deoxy-D-xylulose 5-phosphate reductoisomerase (DXR) in wheat roots (Walter et al. 2000). DXS is an enzyme that catalyzes the initial step of the MEP pathway, where many isoprenoids are biosynthesized, and DXR is an enzyme that is immediately downstream from DXS in the MEP pathway (Walter et al. 2000). In another study by Walter et al. (2002), DXS2 transcript levels were strongly stimulated in Medicago truncatula roots upon colonization by AMF (a mixture of F. mosseae and R. intraradices), and were correlated with the accumulation of carotenoids and apocarotenoids. Finally, alterations in these secondary metabolites' production also can result from plant defense responses to AMF colonization (Mechri et al. 2015;Zubek et al. 2012Zubek et al. , 2015Torres et al. 2015).

Fig. 2
Non-nutritional and nutritional factors influencing the production of secondary metabolites (i.e., terpenoids, phenolics, and flavonoids) in AMF-colonized plants. Non-nutritional factors (leftside in orange): AMF colonization results in the activation of plant defense mechanisms with the production of phenolics and flavonoids. Change in phytohormone levels, such as jasmonic acid (JA), gibberellic acid (GA 3 ), and 6-benzylaminopurine (BAP), increases the number and size of glandular trichomes and leads to transcriptional activation of sesquiterpenoid biosynthetic gene expression. AMF induce the production of signaling molecules, such as nitric oxide, salicylic acid (SA), and hydrogen peroxide, which influence the activation of key enzymes such as l-phenylalanine ammonia lyase (PAL) and chalcone synthase (CHS), for the biosynthesis of phenolic compounds. Nutri-tional factors (rightside in blue): AMF colonization increases plant nutrients and water uptake leading to increased plant growth and leaf biomass. This results in enhanced plant photosynthetic capacity and increased production of photosynthates which are precursors of different secondary metabolites. Increased leaf biomass leads to an increased density of glandular trichomes in which terpenoids are synthesized and stored. This figure is adapted with permission from Springer Nature Customer Service Centre GmbHS: Springer Nature, Phytochemistry Reviews. Insight into the mechanisms of enhanced production of valuable terpenoids by arbuscular mycorrhiza . We thank Evangelia Tsiokanou (National and Kapodistrian University of Athens, Greece) for graciously providing the picture of the plant used in this figure The level of endogenous jasmonate was shown to increase after wounding or pathogen attack. There is no direct study on the effects of AMF on A/S production in these Boragenaceous medicinal plants. However, these findings suggest that AMF could be a potential factor enhancing A/S production in mycorrhizal Boraginaceae plants through the regulation of jasmonate.

Cultivation techniques for secondary metabolite production in mycorrhiza-associated medicinal plants
Plant secondary metabolites are often extracted from individuals grown in nature. For instance, around 95% of the medicinal plants used in the Indian herbal industry today are collected from the wild (Lakshman 2016). However, the quantity and quality of secondary metabolites from plants grown in nature are erratic, often influenced by abiotic and biotic factors, such as extreme temperatures, drought, alkalinity, salinity, and plant pathogens, impacting the metabolic pathways responsible for the accumulation of bioactive substances (Dayani and Sabzalian 2017;Giurgiu et al. 2017;Ramakrishna and Ravishankar 2011). Furthermore, overharvesting of medicinal plant species in nature could place them at a high risk of extinction (Roberson 2008). Finally, growing medicinal plants under field conditions may be time consuming, especially for woody plants (e.g., Taxus brevifolia and Lithospermum erythrorhizon) and slow-growing perennial plants (e.g., Panax ginseng), which can take several years to reach the desired metabolites production (Malik et al. 2011;Chandran et al. 2020;Yazaki 2017;Murthy et al. 2014). Therefore, there is a need for alternative production systems.
Production of medicinal herbs in controlled environments provides opportunities for improving the quality, purity, consistency, bioactivity, and biomass production of the raw material (Hayden 2006). In order to secure the commercial production of secondary metabolites, several cultivation techniques have been developed, potentially compatible with AMF application.

Greenhouse cultivation
Greenhouses are widely used for crop production allyear round. Environmental parameters (e.g., temperature, humidity) are controlled, providing optimal growth conditions to the target crop or plant, favoring development, and thus safeguarding the yield and consistent production of high-quality bioactive compounds (Panwar et al. 2003).
Many medicinal plant species, such as Echinacea angustifolia, Echinacea purpurea, Ocimum basilicum, Withania somnifera, and Psoralea croylifolia, have been grown under greenhouse conditions (Zheng et al. 2006;Panwar et al. 2003). Similarly, many, such as Artemisia annua, Curcuma longa, Coleus forskohlii, Glycyrrhiza glabra, and Gloriosa superba, have been associated with AMF under greenhouse conditions with high production of bioactive compounds reported (Huang et al. 2011;Dutta and Neog 2016;Sailo and Bagyaraj 2005;Liu et al. 2007;Pandey et al. 2014). Therefore, growing medicinal plants in association with AMF under greenhouse conditions could represent a suitable method for improving the quality and production of bioactive compounds at large scale.

Aeroponics
In the aeroponics cultivation system, the roots of plants are hung inside a sealed container in darkness and exposed to a water nutrient-rich spray through atomizers (Lakhiar et al. 2018) (Fig. 3a). This technique has been developed for the cultivation of many different plants, such as horticultural crops (e.g., Lactuca sativa, Cucumis sativus, and Solanum lycopersicum) (Movahedi and Rostami 2020), medicinal herbs (e.g., Anemopsis californica, Crocus sativus, and Valeriana officinalis) (Hayden 2006;Souret and Weathers 2000;Tabatabaei 2008), and medicinal crops (e.g., Arctium lappa and Zingiber officinale) used to extract secondary metabolites from their roots (Hayden et al. 2004a, b). It has been reported that Ocimum basilicum grown under aeroponic conditions had a higher yield, comparable phenolic and flavonoid contents, and antioxidant properties compared to plants grown in a solid substrate (Chandra et al. 2014). Similarly, Cichorium intybus, Withania coagulans, and Echinacea sp. grown in an aeroponic system had higher yields compared to the same plants grown in soil (Movahedi and Rostami 2020). This system was efficient for the production of bioactive molecules from roots of medicinal crops, such as chlorogenic acid in A. lappa and β-sitosterol in Cannabis sativa (Hayden 2006;Ferrini et al. 2021). For various medicinal plants, root apices constitute the main sites where active substances are produced and stored (Watson et al. 2015). However, these active substances are almost impossible to harvest through conventional farming methods. By using the Plant Milking Technology (Plant milking®) (https:// www. plant advan ced. com/ home) for Morus alba (an emblematic tree of traditional Chinese medicine, rich in alkaloids and flavonoids), Chajra et al. (2020) obtained an extract enriched in prenylated flavonoids that was 18-fold higher than commercial root extracts (Fig. 3a, b).
Interestingly, aeroponic cultivation systems also have been developed and used for the production of AMF inoculum in which roots (and AMF) were bathed in a nutrient solution mist (Zobel et al. 1976;Hung and Sylvia 1988). For the production of AMF, plants are precolonized prior to their introduction into the system, through preculturing plant seedlings and AMF propagules (both preferably surface-sterilized) in pots containing a substrate (e.g., mixture of sand and perlite). Then the precolonized plants are transferred to the aeroponic container where the roots (and AMF) develop. The container is usually protected from light to prevent the development of algae (Jarstfer and Sylvia 1995). The mist can be applied by various techniques that differ mainly in the size of the fine droplets produced (e.g., atomizing disk, pressurized spray through a microirrigated nozzle, an ultrasonically generated fog of nutrient solution with droplets of 3-10-μm diameter, and ultrasonic nebulizer technology resulting into microdroplets of 1 μm in diameter) (IJdo et al. 2011;Jarstfer and Sylvia 1995;Mohammad et al. 2000). Mohammad et al. (2000) reported a high number of viable AMF propagules obtained in aeroponic culture, and such inoculum was used in a field experiment (Mohammad et al. 2004). Thus, aeroponic systems could potentially be used for growing medicinal plants associated with their AMF partners in order to obtain substantial biomass and production of secondary metabolites.
Different hydroponic culture systems also exist for the mass production of AMF. They mainly differ in the mode of aeration and application of the nutrient solution (IJdo et al. 2011). For instance, in the static hydroponic culture system, the nutrient solution is not flowing and needs to be aerated via an aeration pump to prevent roots of mycorrhizal plants from suffering oxygen deprivation (IJdo et al. 2011). Via this system, Dugassa et al. (1995) obtained large quantities of mycorrhizal Linum usitatissimum plant roots as well as extramatrical mycelium and chlamydospores free of residues from solid substrate components. In another nutrient film technique (NFT) hydroponic system, a thin nutrient solution (i.e., film) flows into inclined channels (also called gulls) where the plant roots and AMF develop (IJdo et al. 2011). This technique has been used to culture AMF since the 1980s with the production of many sporocarps by F. mosseae (Elmes and Mosse 1984). Later, IJdo et al. (2011) developed an innovative low-cost in vitro plant-based bioreactor system for the mass production of AMF. In this system, Medicago truncatula roots and AMF (Glomus sp.) were grown in a sterilized tube connected at both extremities to a reservoir containing sterilized liquid culture medium. This nutrient solution circulates across the mycorrhizal root system, feeding the plant/fungus associates, while the plant shoot develops in open-air conditions inside a controlled growth chamber (Fig. 3e). The hydroponic system also has been developed for studying the effect of the AMF symbiosis (e.g., P uptake) on maize plants (Garcés-Ruiz et al. 2017). Therefore, hydroponic or semi-hydroponic systems could potentially be combined with medicinal plants and AMF in order to obtain increased production of secondary metabolites. In a recent study, Cartabia et al. (2021) showed how the R. irregularis modified the primary and secondary metabolism and the root exudates of the medicinal plant Anchusa officinalis growing under a semi-hydroponic cultivation system (Fig. 3c). Moreover, permeabilization treatments can be conducted in these cultivation systems, in order to extract the compounds exuded by roots in a non-destructive process that "milks" the same plants several times a year. For example, in the study by Gontier et al. (2002), Datura innoxia plants were cultivated in hydroponic conditions (no AMF were involved) and the plant roots subsequently permeabilized with Tween 20. As a result, a high concentration of tropane alkaloids (TA) (e.g., hyoscyamine and scopolamine) was detected in the nutrient solution. Interestingly, all the plants were able to survive after being rinsed and replaced in the hydroponic system. This approach allows the permeabilization of the plant multiple times without loss of viability (Gontier et al. 2002). Moreover, different permeabilization treatments (e.g., doses and duration of Tween 20, addition of TA precursors) can be chosen to release additional bioactive compounds in the nutrient solution (i.e., TA precursors (phenylalanine and ornithine) leading to 10-80 mg/l TA in the nutrient solution) (Gontier et al. 2002). This study, however, did not include association with AMF.

In vitro production systems
Micropropagation or in vitro propagation is the clonal propagation of plants by tissues, cells, or organs. It involves the aseptic culture of explants of tissues or organs in closed vessels using defined culture media in a controlled environment (Debnath and Arigundam 2020).

Whole plant in vitro culture
In vitro cultivation of whole plants is widely used for mass propagation, conservation of germplasm, production of bioactive compounds, and genetic improvement of a large number of medicinal plant species (Nalawade and Tsay 2004 (Nalawade and Tsay 2004).
The association of AMF with whole plants in vitro has been described in several studies (e.g., Dupré de Boulois et al. 2006;Voets et al. 2005;Lalaymia and Declerck 2020). For instance, using the mycorrhizal donor plant in vitro cultivation system, Voets et al. (2009) obtained fast and homogenous mycorrhization of Medicago truncatula seedlings by placing the plantlets in an actively growing mycelial network arising from a mycorrhizal donor plant (Fig. 4a). In another system, called the half-closed arbuscular mycorrhizal plant in vitro culture system, roots of micropropagated potato plantlets were associated with actively growing AMF propagules, while the shoots developed in open-air conditions (Voets et al. 2005) (Fig. 4c). Applying both systems, several thousand spores of R. intraradices were produced on an extensive extraradical mycelium and abundant root colonization has been obtained. Hence, they could be extended to medicinal plants to enhance secondary metabolite production (Fig. 4). (c) a 90-mm half-closed arbuscular mycorrhizal plant in vitro culture system allowing the growth of the roots of Lithospermum erythrorhizon in close association with R. irregularis MUCL 41833; (d) close-up view of the reddish roots due to shikonin production; (e) a 90-mm root organ culture in vitro system allowing the growth of Ri T-DNA transformed A. tinctoria hairy root (Rat et al. 2021) in assocation with R. irregularis MUCL 41833 in the RC; (f) close-up view of the red AMF spores produced in the RC (arrows). We thank Alicia Varela Alonso (Institut für Pflanzenkultur, Germany) for graciously providing the pictures c and d and Angélique Rat (Ghent University, Belgium) for providing the Alkanna tinctoria hairy roots used in this figure. The system (a) starts with a donor plant (Crotalaria spectabilis) introduced into the RC of a bi-compartmented system (a small Petri dish indicated by a dashed circle (RC) (90 mm diameter)) placed in a large Petri dish (HC) (145 mm diameter). A hole is made in both Petri dishes allowing the shoot to extend outside the system. Approximately 500 spores from an AMF in vitro culture are placed in contact with the roots. The roots and AMF are kept in the dark during the whole growth period, while shoots remain under light. Once the donor plant is well colonized, the extraradical mycelium starts to cross the partition wall separating the RC from the HC, developing profusely in the HC. At that time, one or several receiver micropropagated plants (Alkanna tinctoria) are placed in the HC with their roots in contact with the extraradical mycelium. The plants are planted inside the HC under a lid. Briefly, the base of a cylinder (150 mm high, 100 mm diameter) matches a hole made in the lid of the 145-mm Petri dish. The cylinder top is glued to a 100-mm Petri dish lid. The culture dishes containing the A. tinctoria plants are sealed and covered, up to the base of the cylinder, by black plastic bags. The systems are incubated in a growth chamber to allow plant and AMF growth (detailed procedures of this system can be found in Lalaymia and Declerck (2020)). For system (c), homogenously chopped agar containing AMF propagules from an AMF in vitro culture is inoculated to the newly growing roots of a micropropagated seedling of Lithospermum erythrorhizon. After a few days, the new hyphae growing from the spores colonize the roots of L. erythrorhizon. In system (e), fine root structures of Ri T-DNA transformed Alkanna tinctoria hairy roots are cut and placed in the RC part of a bi-compartmental Petri dish. Chopped agar containing AMF propagules is spread on the young parts of the hairy roots. After a few days, new hyphae growing from spores colonize the A. tinctoria hairy root, producing new spores and extensive mycelium after several months. All these three techniques should be conducted under a laminar flow hood with sterilized laboratory materials

Hairy root cultures (HRCs)
Hairy root cultures are cultures raised after the infection of explants/cultures by the gram-negative soil bacterium Agrobacterium rhizogenes 2 (Tepfer and Casse-Delbart 1987). This bacterium leads to the neoplastic growth of roots which are characterized by high growth rates in hormone-free media and genetic stability (Pistelli et al. 2010). Ri T-DNA transformed hairy roots grow faster than adventitious roots, or even conventional cultures in which plants are growing in soil or substrate (Paek et al. 2009;Yu et al. 2005). For instance, Panax ginseng, a valuable medicinal plant originating from Asia, has been used as a healing drug and health tonic since ancient times (Tang and Eisenbrand 1992) due to its production of triterpene saponins, collectively called ginsenosides (Dewick 1997;Huang 1999). However, it generally takes 5 to 7 years in the field to attain maturity and to reach the harvesting stage for extraction of bioactive compounds (Murthy et al. 2014). To solve this problem, many different techniques have been explored, such as culture of callus tissues, suspended cells, adventitious roots, and HRCs (Furuya et al. 1973(Furuya et al. , 1983Paek et al. 2009;Shi et al. 2021). It has been reported that HRCs grow more rapidly and produce a higher level of ginsenosides than the suspended cells and adventitious root cultures (Inomata et al. 1993;Yoshikawa and Furuya 1987). Moreover, HRCs produce the same phytochemical patterns as the wild-type root organs and accumulate higher levels of certain valuable compounds compared with adventitious roots and native-grown plant roots Hao et al. 2020;Miao et al. 2017). For instance, the total tanshinone content reached 15.4 mg/g dry weight (DW) in transgenic Salvia miltiorrhiza hairy roots, while only 1.7-9.7 mg/g DW tanshinone was produced in roots of field-grown plants Hao et al. 2020). High stability and productivity, high biomass production, and efficient biosynthetic capacity make HRCs valuable biotechnological tools for the production of plant secondary metabolites (Pistelli et al. 2010;Gutierrez-Valdes et al. 2020). Some examples of metabolites produced using HRCs are tropane alkaloids, such as scopolamine and hyoscyamine (Jouhikainen et al. 1999;Häkkinen et al. 2016;Guo et al. 2018;Khezerluo et al. 2018), catharanthine (Hanafy et al. 2016), ginsenosides (Woo et al. 2004;Ha et al. 2016), solanoside (Putalun et al. 2004), and anthraquinones (Perassolo et al. 2017). The studies mentioned here, however, did not involve inoculation with AMF.
In order to upscale production and commercialize secondary metabolites, various conventional bioreactors, broadly classified as liquid phase, gas phase, or hybrid reactors, have been employed for the mass production of HRCs, which permit the growth of interconnected tissues normally unevenly distributed throughout the vessel. For instance, increased production of terpenoid indole alkaloid and artemisinin was obtained with HRCs of Rauwolfia serpentina and Artemisia annua grown in different reactors (no AMF involved) (Mehrotra et al. 2015;Patra and Srivastava 2014). Interestingly, the in vitro large-scale production of AMF is mostly based on HRCs (Declerck et al. 1996;Declerck 2006) which involve the association of AMF propagules with transformed hairy roots on synthetic mineral media. Arbuscular mycorrhizal fungi species grown in HRCs produce viable and contaminant-free spores (Fortin et al. 2005) (Fig. 4e). From these studies, we can suggest HRCs of medicinal plants associated with AMF in bioreactors for the commercial production of secondary metabolites.

Conclusions
Arbuscular mycorrhizal fungi may confer several benefits to medicinal plants, such as growth promotion and improved tolerance to stress conditions. Interestingly, AMF also may enhance the accumulation of active substances in those plants. This makes mycorrhizal technology a potential and sustainable tool for improving the growth and secondary metabolite production of medicinal plants. Factors such as light, temperature, humidity, soil fertility, and cultivation techniques also could influence secondary metabolite production by medicinal plants (Szakiel and Pączkowski 2011a, b). It is thus essential to consider these parameters in finetuning the conditions for optimal production of plants and associated secondary metabolites. In order to guarantee the quality of the bioactive substances produced by mycorrhizal medicinal plants, different substrate or substrate-free systems were described: aeroponic and hydroponic or semihydroponic systems, micropropagated medicinal plants in half-closed arbuscular mycorrhizal-plant or mycorrhizal donor-plant in vitro culture systems, and HRCs. These systems may provide adequate environmental conditions to the plants, resulting in improved crop yield and production of bioactive compounds (Nazari Deljou et al. 2014;Dayani and Sabzalian 2017).
Whatever the system considered, high yields of secondary metabolites are dependent on the AMF strain, the plant, and the environmental growth conditions. Arbuscular mycorrhizal fungi are not host specific, but their affinity to a particular host can be preferential (Cesaro et al. 2008). Early studies have documented that different AMF species may induce differences in secondary metabolite production in the same host or genotype (Zeng et al. 2013). For example, Glomus caledonium increased rosmarinic acid and caffeic acid production in Ocimum basilicum, whereas F. mosseae enhanced only caffeic acid production (Toussaint et al. 2007). In a recent study, Frew (2020) showed that inoculation with four commercial AMF species (C. etunicatum, Funneliformis coronatum, F. mosseae, and R. irregularis) had stronger effects on cereal crop plant allometric partitioning, foliar nutrient, and phenolic concentrations than inoculation with a single commercial AMF species (R. irregularis). Interestingly, the results also showed that the effects of inoculating with these four commercial AMF species were not different from the effects of applying a native AMF inoculant extracted from field soil, suggesting that commercial AMF assemblages may provide little to no additional benefit compared with a resident AMF community (Frew 2020). Thus, a thorough study of AMF species native to medicinal plants should be conducted before choosing the most effective AMF species (native or not; single or combinations of different AMF strains) to inoculate target plant species.
The use of commercial inoculants is an option which should be considered with caution. Indeed, a number of studies have shown that the absence of a regulatory context for the industry of AMF inoculants may have contributed to inoculants of questionable quality. For instance, Salomon et al. (2022) reported that over 80% of tested commercial inoculants contained no viable propagules when screened in sterilized soil. Moreover, when preparing AMF inoculum, adequate phytosanitary controls must be implemented to avoid proliferation of unwanted microbes which may subsequently contaminate plant production. The use of root organs in vitro may provide a solution by avoiding the presence of such contaminants. However, genetically modified plants may represent a drawback for field application and thus for commercial mass production, and the number of species grown in this system remains limited (Ijdo et al. 2011). Finally, whatever system is used, environmental parameters should be considered seriously. For instance, light intensity is known to strongly impact the development of AMF. Konvalinková and Jansa (2016) reported that a sudden decrease of light availability to an AMF-colonized plant resulted in a rapid decrease of P transfer from the AMF to the plant, and when arbuscular mycorrhizal plants were exposed to long-periods of shading (weeks to months), positive mycorrhizal growth responses often declined. Ballhorn et al. (2016) also reported that under light limited conditions, vegetative and reproductive traits were inhibited in AMF inoculated Phaseolus lunatus plants relative to noncolonized plants. Thus, in controlled conditions, light intensity and quality (e.g., blue/red ratio) should be modulated to improve and guarantee the symbiosis between AMF and plants of interest (Konvalinková and Jansa 2016). Adequate timing and harvestable plant parts also are crucial factors to increase the production of secondary metabolites. The best time to harvest (quality peak season/time of day) should be determined according to the quality and quantity of biologically active constituents rather than the total vegetative yield of targeted medicinal plant parts. Taking into consideration all these environmental factors would help optimize plant-AMF associations, increasing biomass and secondary metabolite production.
Funding This project has received funding from the European Union's Horizon 2020 Research and Innovation Programme under the Marie Skłodowska-Curie grant agreement No 721635.

Conflict of interest
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.