Volatile profile of Echinacea purpurea plants after in vitro endophyte infection

Abstract The differences in volatile profile of Echinacea purpurea plants not-inoculated (EpC) and inoculated with their endophytes from roots (EpR) and stem/leaves (EpS/L) were analysed and compared by GC-FID/GC–MS in an in vitro model. Non-terpenes and sesquiterpene hydrocarbons were the most abundant classes with an opposite behaviour of EpS/L showing a decreased emission of sesquiterpenes and an increase of non-terpene derivatives. The main compounds obtained from EpS/L were (Z)-8-dodecen-1-ol and 1-pentadecene, while germacrene D and β-caryophyllene were the key compounds in EpC and EpR. For the first time, this work indicates that bacterial endophytes modify the aroma profiles of infected and non-infected in vitro plants of the important medicinal plant E. purpurea. Therefore, our model of infection could permit to select endophytic strains to use as biotechnological tool in the production of medicinal plants enriched in volatile bioactive compounds. Graphical Abstract


Introduction
Symbiotic plant-bacteria interactions represent a mutualistic interplay between a host plant and not-harmful microorganisms (named endophytes) able to influence plant physiology both directly producing bioactive compounds and modulating plant secondary metabolism and vice versa (Wani et al. 2015). Bacteria emit volatile organic compounds (VOCs) influencing host plant growth and defence response (Bailly and Weisskopf 2012). On the other hand, plants can emit VOCs enhancing or limiting the endophytic ability to colonize (Farre-Armengol et al. 2016).
The microbiota of a medicinal plant can modulate the plant metabolome influencing the therapeutic properties of the plant itself (Todd et al. 2015). For example, we used the medicinal plant Echinacea purpurea (L.) Moench (Asteraceae) as a model plant to demonstrate the effect of plant microbiota on the regulation of alkamide accumulation (Maggini et al. 2017).
Since E. purpurea releases VOCs such as a-phellandrene, b-myrcene, caryophyllene, caryophyllene epoxide, germacrene D and borneol (Mazza and Cottrell 1999), present work aims to determine whether there are variations on VOC emission after infection of E. purpurea with its endophytes.

In planta bacterial growth
Bacterial pools isolated from stem/leaves (S/L) and roots (R) of E. purpurea plants (Table S1) were separately inoculated in three axenic in vitro 2 months old E. purpurea plants (EpS/L and EpR, respectively); three plants (EpC), used as control, were inoculated with sterilized saline solution. The infection experiment was performed in triplicate. After twenty days, R strains growth around roots was remarkable as opposed to S/L strains (Table S2). Our previous data showed that different endophytic communities characterized R and S/L compartments of the plant (Chiellini et al. 2014): this diversity could account for the differences in VOC emission of the relative inoculated plants.

VOC emission
Thirty-seven compounds were identified, ranging from 31 in EpC to 19 in EpS/L ones (Table 1) and the chemical classes resulted differently represented between control and infected plants (Table S3). Mazza and Cottrell (1999) found a different blend of the main compounds reporting the predominance of terpenoids (46.0%) and non-terpenes (53.0%) in the headspace of aerial parts and roots of three Echinacea species. On the other hand, the paper by Mirjalili et al. (2006) reported an essential oil composition of E. purpurea cultivated in Iran similar to that observed in this study with sesquiterpene hydrocarbons (SH) around 71% and germacrene D as principal component (57%).
Our data showed SH as the main volatiles but with a decreased quantity in EpS/L (53.15%) respect to EpC (66.00%) and EpR (73.70%). EpR and EpS/L different aroma profiles could be due to the different endophytic pools used to inoculate the plants since E. purpurea compartments contained different endophytes characterized by specific phenotypes (Maggini et al. 2018). Also, the only oxygenated sesquiterpenes (OS) and apocaroteinod (AC) present in EpC (Germacrene D-4-ol, Nerolidol acetate and (E)-B-ionone) resulted lost in all inoculated plants. In planta SH (e.g. germacrene D in EpS/ L) and OS reduction or disappearance could be due to an endophytic influence on plant metabolic pathways (Maggini et al. 2017) to block the antimicrobial effect of these compounds (Dahham et al. 2015). On the other hand, bacteria can modulate VOC chemical composition in planta producing volatiles (Farre-Armengol et al. 2016) to protect against pathogens, attract mutualistic microbes and act both as plant growth promoters and inhibitors (Park et al. 2015). Notably, plants producing antibacterial VOCs may favour tissue colonization by resistant microbial communities able to detoxify and use plant volatiles as nutrient sources (Baldwin 2010). In this concern, our data evidenced that non-terpene (NT) derivatives were slightly lower in EpR (21.80%) than in EpC (29.25%) but increased in EpS/L (44.20%) suggesting that these compounds could be useful for plant-bacteria interaction as reported for 2-Methyl-n-1tridecene (Park et al. 2015).
Regarding single VOC, only 17 compounds out of 37 were present in all the plant samples (Table 1). Data submitted to principal component analysis (PCA) showed that the vector accounting for EpS/L resulted differentially oriented than that of EpC (Mann-Whitney Utest (MWU): P value ¼ 0.02) and EpR (MWU: P value ¼ 0.01). The major responsible constituents for this position were (Z)-8-dodecen-1-ol and 1-pentadecene for EpS/L whilst germacrene D and b-caryophyllene characterized EpC and EpR (Figure 1). Mucciarelli et al. (2007) evidenced an increase of terpene emission from peppermint leaves and roots after colonization by endophytic fungus. Moreover, the application of epiphytic methylotrophic bacteria was shown to enhance the concentration of 2,5dimethyl-4-hydroxy-2H-furanon (DMHF), in strawberry plants (Verginer et al. 2010).

Conclusions
For the first time, our data indicated that bacterial endophytes could modify the aroma profile of the important medicinal plant E. purpurea currently used for immune stimulation and anti-infection therapy. In particular, our model of infection could permit to select strains with different colonization specificity in order to increase the production of different aromatic compounds. Thus, this strategy strictly supports the application of endophytic bacteria as biotechnological tool (Abrahão et al. 2013) in the production of E. purpurea plants enriched in bioactive compounds and, consequently, with improved therapeutic effects.  Table 1.

Disclosure statement
No potential conflict of interest is reported by the authors.

Funding
This work was supported by Tuscany Region (Italy; Resolution n. 1224/2016 "Medicine Complementari") and by Fondazione Cassa di Risparmio di Firenze (project #2016.0936). The funding sources had no involvement in any part of the study.