Many secondary metabolites have been reported to be involved in microbial interactions. One group of secondary metabolites produced by soil and plant-associated microorganisms, but largely unexplored to date, are the volatile organic compounds (VOCs). VOCs are typically small, odorous compounds (<C15) with low molecular mass (<300 Da), high vapor pressure, low boiling point, and a lipophilic moiety. These properties facilitate evaporation and diffusion aboveground and belowground through gas- and water- filled pores in soil and rhizosphere environments (Vespermann et al., 2007; Insam and Seewald, 2010; Effmert et al., 2012). Microbial volatile organic compounds (mVOCs) belong to different chemical classes including alkenes, alcohols, ketones, benzenoids, pyrazines, sulfides, and terpenes (Schulz and Dickschat, 2007; Lemfack et al., 2014, 2017; Kanchiswamy et al., 2015; Schmidt et al., 2015). A recent meta-analysis by Schenkel et al. (2015) provided a comprehensive overview of VOCs derived from soil-borne microbes.

The production of mVOCs in soil is influenced by various factors including the growth stage of the microbes, nutrient availability, temperature, oxygen availability, pH, and soil moisture content (Wheatley, 2002; Insam and Seewald, 2010). Several recent studies reported that the production of certain mVOCs can be induced or suppressed during inter-specific microbial interactions (Garbeva et al., 2014a; Schulz-Bohm et al., 2015; Tyc et al., 2015; Piechulla et al., 2017). mVOCs were often considered to be by-products of primary metabolism, but recent findings revealed that many mVOCs demonstrate biological activity (Schmidt et al., 2015; Tyc et al., 2017a). Furthermore, in bacteria, the production of certain mVOCs is dependent on the GacS/GacA two-component regulatory system (Cheng et al., 2016; Ossowicki et al., 2017). These findings clearly disagree with the opinion that mVOCs are just waste products.

While soluble metabolites are often responsible for short distance interactions, VOCs are considered to be long-distance messengers (Tyc et al., 2017b; Westhoff et al., 2017). There are many types of microbial interactions occurring belowground such as bacteria–bacteria, fungi–fungi, fungi–bacteria, bacteria–protists, fungi–plant, bacteria–plant, and bacteria–fungi–plant interactions. However, most studies addressing belowground VOCs-mediated interactions are focused mainly on the root-emitted volatiles (recently reviewed by Delory et al., 2016).

The knowledge we have gained from research conducted over the last few years reveals that mVOCs can have both beneficial and harmful effects on other organisms (Effmert et al., 2012; Schmidt et al., 2015). mVOCs can provide organisms with rapid and precise ways to recognize neighboring organisms (both friends and foe) and to launch proper responses.

The aim of this review is to summarize the current knowledge concerning the role of mVOCs in intra- and inter-kingdom interactions, to pinpoint mVOCs (e.g., terpenes) involved in microbe–microbe and microbe–plant interactions as well as to indicate challenges in studying belowground mVOCs-mediated interactions and opportunities for further studies and practical applications.

In recent years, evidence supporting the idea that plants respond strongly to mVOCs has grown. Most of the research carried out so far has investigated the impact of microbial VOCs on the model plant Arabidopsis thaliana. This has revealed that, without physical contact, microorganisms are able to drastically alter plant root system development, plant physiology, hormonal pathways, and biomass production (Ryu et al., 2004; Blom et al., 2011; Wenke et al., 2012; Bailly et al., 2014; Bitas et al., 2015; Ditengou et al., 2015; Li et al., 2016; Piechulla et al., 2017). mVOCs can also function as a direct source of nutrients for plants (Meldau et al., 2013), induce resistance to pathogens in plants (D’Alessandro et al., 2014; Kottb et al., 2015; Song et al., 2015b; Wintermans et al., 2016), affect plant secondary metabolite production (Santoro et al., 2011), directly inhibit plant pathogens (Kai et al., 2009; Garbeva et al., 2014b; De Vrieze et al., 2015; Kottb et al., 2015) and induce soil fungistasis and suppressiveness (Garbeva et al., 2011; Van Agtmaal et al., 2015). Moreover, one single mVOC can show various functions, such as dimethyl disulfide, which improves plant growth by enhancing the availability of reduced sulfur (Meldau et al., 2013). It also protects tobacco and corn plants against Botrytis cinerea and Cochliobolus heterostrophus by directly inhibiting pathogens and inducing systemic resistance in plants (Huang C.-J. et al., 2012). Likewise, a characteristic compound of Trichoderma asperellum, 6-pentyl-pyrone, can increase plant defense reactions and at the same time decrease B. cinerea and Alternaria alternata sporulation (Kottb et al., 2015).

Many independent studies revealed that mVOCs emitted by beneficial soil microorganisms can affect plant growth but only few studies focused on how VOCs produced by soil-borne plant pathogens affect plant growth and development. These studies suggest that mVOCs from plant pathogens may modulate the trade-off between plant growth, development and defense. Bitas et al. (2015) showed that VOCs emitted by pathogenic F. oxysporum promoted the growth of A. thaliana and Nicotiana tabacum and affected auxin transport and signaling. VOCs emitted by the pathogen Alternaria alternaria enhanced growth, early flowering and photosynthesis rates of Arabidopsis, maize and pepper by affecting the levels of plastidic cytocinin (Sanchez-Lopez et al., 2016). A more recent study showed that the soil-borne pathogen Rhizoctonia solani produced an array of mVOCs that promote plant growth, accelerate development, change plant VOCs emission and reduce insect resistance (Cordovez et al., 2017). This must be a successful strategy for the pathogenic fungi since with increased root biomass and stimulation of lateral root formation there is a greater surface area for fungal colonization and infection.

When analyzing mVOCs effects on plant growth, it is important to take into account, that microorganisms can produce high amounts of CO₂ that can promote plant growth (Kai and Piechulla, 2009; Piechulla et al., 2017). Hence, a good experimental setup with appropriate controls are required to avoid artifacts in the results (Piechulla et al., 2017).

Alternatively, plants are able to mediate the belowground plant–microbe interactions via root-emitted VOCs (Wenke et al., 2010). Root-derived VOCs may serve multiple roles such as carbon sources, defense metabolites and chemo-attractants (Van Dam et al., 2016). Rhizobacteria such as Pseudomonas fluorescens and Alcaligenes xylosoxidans have been shown to metabolize α- pinene as their sole carbon source (Kleinheinz et al., 1999). Del Giudice et al. (2008) also reported that bacteria associated with the roots of vetiver grass (Vetiveria zizanioides) use sesquiterpenes as a carbon source. Undoubtedly, plants and soil microorganisms are engaged via VOCs in long-distance interactions (Van Dam et al., 2016). However, so far, limited knowledge exists concerning the role of plant VOCs in attracting beneficial organisms and how plant-associated microorganisms affect the quantity and quality of plant volatile emission. Only recently, using a glass olfactometer system, the attraction of distant soil bacteria by VOCs emitted by plant roots was revealed (Schulz-Bohm et al., 2017). Olfactometer systems have been used successfully to study aboveground plant–herbivores interactions (Ballhorn and Kautz, 2013) or belowground plant–nematode interactions (Rasmann et al., 2005). However, this is the first case to apply an olfactometer to study plant–microbe interactions. Moreover, the same study revealed that upon fungal infection, the blend of root VOCs changed and specific bacteria with antifungal properties were attracted (Schulz-Bohm et al., 2017).

Several reports describe the chemical dialog between microbes, plants, and other organisms by the exchange of soluble compounds (Badri et al., 2009; Lira et al., 2015; Song et al., 2015a; Liu et al., 2016). Most of the studies reporting mVOCs-mediated communication belowground focus on the uni-directional responses and only a few studies reported on bi-directional mVOCs-mediated interactions. For instance, the importance of mVOCs in the dialog between the fungal plant pathogen Verticillium longisporum and its bacterial antagonist Paenibacillus polymyxa was recently revealed in both in vitro and in planta experiments (Rybakova et al., 2017). Both microorganisms responded to one another’s VOCs and this specific mVOCs-mediated interaction resulted in the inhibition of cellular metabolism and growth reduction of the fungal pathogen.

A VOCs-mediated dialog between bacteria and fungi was also reported by Spraker et al. (2014) where VOCs of the fungal plant pathogen Aspergillus flavus reduced the production of the major virulence factor EPS of the bacterial plant pathogen R. solanacearum. In parallel, A. flavus responded to VOCs of R. solanacearum by reducing conidia production and increasing aflatoxin production.

Over the last decades, our understanding of the chemical complexity of mVOCs produced by many different soil microorganisms has grown. It is clear that these small and odorous molecules can modify the behavior and promote or inhibit growth of neighboring organisms (Figure 1).

Most existing studies on mVOCs are focused on describing the uni-directional effect of mVOCs produced by a single organism and the responses of the organisms perceiving them without considering mVOCs-mediated dialog and the bi-directional responses to one another. Furthermore, microbial interactions taking place belowground are far more complex than single one-to-one interactions and involve more organisms, which can significantly affect mVOCs emission. For example, fungal-associated bacteria have been shown to affect the production of VOCs in fungi (Schulz-Bohm et al., 2015; Splivallo et al., 2015) and in addition, they can affect the fungal plant-pathogenicity and repress the expression of fungal virulence genes (Minerdi et al., 2009). Therefore, a holistic approach considering the effect of mVOCs on belowground soil community is needed. For instance, using a metagenomics approach Yuan et al. (2017) revealed that mVOCs could alter the composition of soil bacterial and fungal communities and significantly increased the relative abundance of Proteobacteria, Bacteroidetes, Firmicutes, and Ascomycota. Furthermore, mVOCs influenced genes involved in important soil functions such as N-fixation (nifH), nitrification (amoA), denitrification (nirS) and antibiotic production (NRPS) (Yuan et al., 2017).

From the current scientific literature, it is clear that the most studied belowground mVOCs-mediated interactions are the interactions between bacteria, fungi and plants (Figure 1). There is a lack of knowledge relating to the emission of VOCs by protists, archaea or other rhizosphere organisms, such as nematodes or earthworms, indicating that these groups are currently understudied with regards to this aspect.

Several VOCs are commonly produced and emitted by both plant roots, fungi, bacteria and protists and it is possible that these compounds act as a ‘lingua franca’ for intra- and inter-kingdom communication between these organisms. Let us take as an example only one chemical class, the terpenes. Terpenes are the largest and most diverse class of metabolites known to date. They are best known to humans as plants metabolites. However, recent studies revealed that terpenes can be produced by all kingdoms of life including prokaryotes (Takamatsu et al., 2011; Yamada et al., 2012, 2015; Song et al., 2015b; Chen et al., 2016). Recently, Yamada et al. (2015) described a powerful bioinformatics method based on the use of Hidden Markov Models (HMMs) and Protein Families Database (PFAM) search that has allowed the discovery of terpene synthases of bacterial origin and showed that phylogenetically different bacteria can be a rich source of terpenes. Both the number, the wide distribution, and the structural diversity of terpenes provide enormous potential for mediating significant chemical interactions and communication belowground. Examples of terpene-mediated microbial interactions are presented in Figure 2 and Table 1, indicating the ecological importance of terpenes in interactions between soil micro- and macro-organisms, including plant roots.

Despite the rapid increasing numbers of studies showing the importance of mVOCs in the long-distance belowground chemical interactions, we still do not know exactly how VOCs are recognized and perceived. VOCs receptors or other perception mechanisms have not been identified in any of the described cases. The big challenge is to determine whether VOCs are internalized and transduced by receptor-mediated processes, whether they interact with the cell membrane to initiate signal transduction cascades or whether they are simply taken up by the cell and metabolized (Widhalm et al., 2015; Adebesin et al., 2017; Tissier et al., 2017). For plants, the current view is that due to their lipophilic nature, VOCs such as mono- and sesquiterpenes may interfere with membrane structures, thereby causing depolarization of the membranes and triggering Ca²⁺-signaling in plants (Maffei et al., 2001; Heil and Land, 2014). For further deciphering of mVOC-mediated microbe-microbe interactions, the mVOCs microbial perception mechanism needs to be elucidated. The application of methods for screening of mutant strains may be useful for that purpose, to identify microbial genes and proteins that are required for VOCs perception.

Another big challenge is to determine what concentrations of mVOCs are produced in soil and at what distances these mVOCs are eliciting a biological response in other organisms. There is the possibility that, similar to the roles of antibiotics in nature (Davies et al., 2006; Yim et al., 2006; Romero et al., 2011), mVOCs could have concentration-dependent function either as weapons in intercellular chemical warfare or as signaling compounds when they are present in low concentrations.

Concerning the implementations of mVOCs, our knowledge on the potential use of those compounds in large-scale agriculture and horticulture is still limited. In agriculture systems, mVOCs have to be applied under open-field conditions, which are very different from the in vitro conditions currently used in most studies. There are very few studies assessing the effects of mVOCs application under open conditions and they have been summarized in a recent review from Chung et al. (2016). Since it was discovered that the 2,3-butanediol elicited plant growth and induced systemic resistance (Ryu et al., 2003, 2004), several studies have applied this compound or the producing strains to the soil of open fields to test its effects under agricultural conditions and have revealed promising results (Velivelli et al., 2015). Dimethyl disulfide, frequently emitted by many bacteria, is another compound used in recent years in the novel soil fumigant PALADIN^® that targets nematodes and soil-borne pathogens. However, the research concerning the application of other mVOCs in agriculture is still in its infancy. We now live in a time in which the old methods of using chemicals to protect crops need to be replaced with and, in some cases, complemented by green solutions. The traditional harmful synthetic fungicides currently used could be replaced with the so far under-explored and unique mVOCs for which significant proof of plant growth promoting effects and plant protection ability already exists. In spite of the obvious potential of mVOCs in agriculture, the field suffers from the common ‘translational gap’ because of a lack of studies evaluating other unexpected effects of those bioactive molecules on non-target beneficial soil organisms.

All authors contributed and approved the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.