Tree Physiology 34, 1399–1410 doi:10.1093/treephys/tpu091

Research paper

Oak powdery mildew (Erysiphe alphitoides)-induced volatile emissions scale with the degree of infection in Quercus robur

Lucian Copolovici1,2,5, Fred Väärtnõu1, Miguel Portillo Estrada1,3 and Ülo Niinemets1,4 Downloaded from 1Institute of Agricultural and Environmental Sciences, Estonian University of Life Sciences, 1 Kreutzwaldi, 51014 Tartu, Estonia; 2Institute of Research, Development, Innovation in Technical and Natural Sciences, ‘Aurel Vlaicu’ University, 2 Elena Dragoi, 310330 Arad, Romania; 3Centre of Excellence PLECO (Plant and Vegetation Ecology), Department of Biology, University of Antwerp, Universiteitsplein 1, B-2610, Wilrijk, Belgium; 4Estonian Academy of Sciences, 6 Kohtu, 10130 Tallinn, Estonia; 5Corresponding author (lucian. [email protected])

Received July 8, 2014; accepted October 2, 2014; published online November 25, 2014; handling Editor Jörg-Peter Schnitzler http://treephys.oxfordjournals.org/

Oak powdery mildew (Erysiphe alphitoides) is a major foliar pathogen of Quercus robur often infecting entire tree stands. In this study, foliage photosynthetic characteristics and constitutive and induced volatile emissions were studied in Q. robur leaves, in order to determine whether the changes in foliage physiological traits are quantitatively associated with the degree of leaf infec- tion, and whether infection changes the light responses of physiological traits. Infection by E. alphitoides reduced net assimilation rate by 3.5-fold and isoprene emission rate by 2.4-fold, and increased stomatal conductance by 1.6-fold in leaves with the largest degree of infection of ∼60%. These alterations in physiological activity were quantitatively associated with the fraction of leaf area infected. In addition, light saturation of net assimilation and isoprene emission was reached at lower light intensity in infected leaves, and infection also reduced the initial quantum yield of isoprene emission. Infection-induced emissions of lipoxygenase at Tartu University on March 3, 2015 pathway volatiles and monoterpenes were light-dependent and scaled positively with the degree of infection. Overall, this study indicates that the reduction of foliage photosynthetic activity and constitutive emissions and the onset of stress volatile emissions scale with the degree of infection, but also that the infection modifies the light responses of foliage physiological activities.

Keywords: green leaf volatiles, induced emissions, isoprene emission, light dependency, monoterpene emission, photosynthesis, quantitative responses, volatile organic compounds.

Introduction Pedunculate oak (Quercus robur) is the most susceptible spe- cies to infections of powdery mildew caused by Erysiphe alphitoi- Plants in the field are exposed to a plethora of biotic stresses des (formerly Microsphaera alphitoides). In fact, powdery mildew that strongly curb their photosynthesis, growth and survival. caused by E. alphitoides is one of the major foliar diseases Among key biotic stresses, fungal infections constitute a major of Q. robur in Europe (Desprez-Loustau et al. 2011) that can stress factor, especially under humid conditions that support ­significantly reduce tree growth and trigger tree decline Mougou-( the spread of fungal pathogens. There is a large difference in Hamdane et al. 2010). Usually, young plants are more susceptible virulence of different fungal pathogens, and many of them are to infections, but severe disease outbreaks can also occur in older highly host-specific Poland ( et al. 2009, Schulze-Lefert and plants in years with optimum weather conditions for E. alphitoides­ Panstruga 2011). Thus, infection by a compatible pathogen (Mougou-Hamdane et al. 2010). Although powdery mildew can result in major disease outbreaks in their host species infection constitutes an important stress in Q. robur, physiologi- (Gururani et al. 2012), thereby selectively reducing the com- cal responses to powdery mildew infection have been studied petitive potential of the host (Alexander and Holt 1998). only in a few cases. It has been demonstrated that E. alphitoides

© The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected] 1400 Copolovici et al. infection decreases net assimilation rates and stomatal conduc- metabolism (Rasulov et al. 2009, Monson et al. 2012), any tance (Hajji et al. 2009), and reduces the rate of isoprene emis- possible modification in the quantum efficiency of photosyn- sion (Brüggemann and Schnitzler 2001). However, quantitative thesis is also expected to result in alterations in the quantum relationships between these physiological modifications and efficiency of isoprene emission. Furthermore, induced mono- the degree of infection have not been studied. As the shape of terpene emissions are typically light-dependent (Niinemets stress severity vs plant response can differ for various stresses 2010b) and can compete for the same chloroplastic precursor (Beauchamp et al. 2005, Niinemets 2010a, 2010b), predicting pool, potentially modifying the light dependence of isoprene stress responses requires understanding of quantitative relation- emissions. Such potential modifications in the shape of consti- ships between stress severity and plant physiological response. tutive isoprenoid emission due to pathogen infection are not Apart from stress effects on photosynthesis and constitu- considered in contemporary isoprene emission models. tive isoprenoid emissions, biotic and abiotic stresses result in a We investigated the influences of E. alphitoides infection on multitude of additional stress responses operating at molecu- foliage photosynthetic characteristics, constitutive emissions lar, cellular, leaf and whole plant levels. Among the early stress of isoprene, induced emissions of LOX volatiles and monoter- responses activated upon contact with pathogenic elicitors are penes in Q. robur, hypothesizing that (i) there are quantitative Ca2+ signals, phosphorylation/dephosphorylation of proteins relationships between the degree of infection and reductions Downloaded from and production of signaling molecules such as salicylic acid, in foliage photosynthetic activity and isoprene emission, and jasmonic acid, ethylene and reactive oxygen species (ROS) increases in stress volatile emission, and that (ii) infection ultimately leading to gene expression and defense responses results in modifications of the shapes of the light responses of (Berger et al. 2007, Pieterse and Dicke 2007, Koornneef and photosynthesis and isoprene emission.

Pieterse 2008). Activation of lipoxygenases and resulting http://treephys.oxfordjournals.org/ emission of volatile products of lipoxygenase pathway (LOX Materials and methods products) consisting of various C6 aldehydes and alcohols, also called green leaf volatiles, also belong to the early stress Study site and plant material responses that can be involved in subsequent alterations in The study was conducted in the vicinity of Tartu, Estonia gene expression patterns (Farag and Paré 2002, Matsui 2006, (58°23′N, 27°05′E, elevation ∼40 m above sea level) in the mid- Jansen et al. 2009). Stress-elicited LOX product emission is dle of September 2011 and 2013. Both years were exception- a ubiquitous stress response and a sensitive indicator of the ally warm and humid with average air temperatures of 18.3 °C presence of abiotic and/or biotic stress (Niinemets 2010a, (year 2011) and 18.2 °C (2013) for the summer months June– at Tartu University on March 3, 2015 Harrison et al. 2013, Possell and Loreto 2013). Stress-induced August and of 12.7 °C (2011) and 11.5 °C (2013) for September emissions of LOX products are typically followed by emissions (data of Laboratory of Environmental Physics, Institute of Physics, of volatile isoprenoids such as mono- and sesquiterpenes and University of Tartu, http://meteo.physic.ut.ee). These averages are benzenoids such as methyl salicylate, whereas the induced 2.5–2.6 °C higher for June–August and 2.4 °C (2011) higher for emission blend is stress dependent, reflecting selective activa- September than the corresponding long-term averages for 1971– tion of different sets of genes (Vuorinen et al. 2007, Hartikainen 2000 (the Estonian Environment Agency, http://www.emhi.ee). et al. 2009, Jansen et al. 2009, Loreto and Schnitzler 2010, Analogously, in 2011, the average values of relative air humid- Staudt et al. 2010, Semiz et al. 2012, Copolovici et al. 2014). ity of 81.8% for June–August and 93% for September were There is evidence that stress-induced emissions of LOX prod- higher than the corresponding long-term averages of 75% for ucts and volatile isoprenoids scale with the severity of different June–August and 83% for September (the Estonian Environment abiotic stresses such as heat, ozone (Beauchamp et al. 2005) Agency). Due to the favorable weather conditions for fungal and mechanical damage (Brilli et al. 2011), but also with biotic development, almost all oak trees were infected with E. alphitoi- stresses such as the degree of herbivory infestation (Copolovici des and exhibited visible signs of damage. et al. 2014). The presence of such quantitative relationships South-exposed leaves showing varying degrees of visual provides an important means to scale up the emissions of bio- infection symptoms were collected from three trees for the genic volatiles from leaves to vegetation and assesses the role measurements. The twigs (∼25 cm long) with multiple leaves of stress-elicited volatiles in air quality and aerosol formation were excised under water, maintained with the cut ends processes in biosphere–atmosphere system (Grote et al. 2013). immersed in water and immediately transported to the labora- In powdery mildew-infected Beta vulgaris leaves, the quan- tory for the measurements. The cut ends were maintained in tum yield of photosynthesis was moderately reduced (Gordon water and a representative leaf was clamped in the cuvette. and Duniway 1982). However, no effects on chlorophyll ­content and PSII quantum yield were observed in powdery Foliage gas-exchange measurements mildew-infected Vitis vinifera (Moriondo et al. 2005) leaves. As To measure gas-exchange and volatile organic compound isoprene emissions are strongly associated with photosynthetic (VOC) emission rates from Q. robur leaves with varying

Tree Physiology Volume 34, 2014 Volatile emissions scale with the degree of infection 1401 degrees of powdery mildew infection, we used a custom- were calculated from measurements of masses 81 (the major made gas-exchange system described in detail in Copolovici monoterpene fragment) and 137 (for details see Copolovici and Niinemets (2010). Briefly, the system used has a 1.2-l et al. 2005 and Filella et al. 2007). The sum of volatile octa- temperature-controlled double-walled glass chamber with decanoid pathway products (LOX products also called green glass and stainless steel bottom. The flow rate through the leaf volatiles), consisting of a variety of C6 aldehydes, alcohols system was 1.4 l min−1, and synthetic air was used by mixing and derivatives produced by C13 hydroperoxide lyases, was −1 purified 2 N (80%), O2 (20%) and CO2 (385 μmol mol ) by found as the sum of the individual mass signals at m/z 83, m/z mass flow controllers (Bronkhorst, Ruurlo, The Netherlands). 85, m/z 99 and m/z 101 (for more details, see Copolovici and The chamber air was vigorously mixed with a fan installed in Niinemets 2010). the system, resulting in fully turbulent conditions such that the To assess the composition of emitted terpenes, volatiles half-time of the chamber air was ∼30 s assuming first-order were also sampled for gas chromatography–mass spectrom- decay kinetics (Niinemets et al. 2011). The air was humidified etry (GC–MS) analyses onto stainless steel tubes with three- with a custom-made humidifier to a relative humidity of 60%. bed filling consisting of different Carbotrap fractions optimized Light intensity was set at 1000 μmol m−2 s−1, and leaf temper- to quantitatively adsorb all volatiles in the C5–C15 range. Four −1 ature was set at 25 °C. CO2 and H2O concentrations at the liters of air was sampled at a constant-flow rate of 0.2 l min Downloaded from chamber in- and outlets were measured with an infra-red dual- using an air sample pump (1003-SKC, SKC, Inc., Houston, TX, channel gas analyzer operated in differential mode (CIRAS II, USA) (Niinemets et al. 2011). Adsorbent cartridges were ana- PP-systems, Amesbury, MA, USA). After enclosure of the leaf lyzed for lipoxygenase (LOX) pathway products, mono-, homo- in the chamber, the leaf was stabilized under the measure- and sesquiterpene emissions, with a combined Shimadzu

ment conditions until steady-state gas-exchange rates were TD20 automated cartridge desorber and Shimadzu 2010 plus http://treephys.oxfordjournals.org/ observed (in ∼30 min after enclosure), and the steady-state GC-MS instrument (Shimadzu Corporation, Kyoto, Japan) using values of foliage gas-exchange and volatile emission rates (see a method as detailed in Toome et al. (2010) and Kännaste the next section) were recorded. Altogether 25 leaves with et al. (2014). The emission rates of isoprene, monoterpenes different degrees of infection were measured (13 in 2011 and and volatile LOX products per unit projected leaf area were cal- 12 in 2013). culated according to the equations of Niinemets et al. (2011). The light-response curves of leaf gas-exchange and VOC emission were measured with a portable gas-exchange/fluo- Quantification of the degree of infection rescence system (GFS-3000, Heinz Walz GmbH, Effeltrich, As E. alphitoides primarily grows on the leaf surface (Mougou at Tartu University on March 3, 2015 Germany). This system has a clip-on type leaf cuvette with et al. 2008), the coverage of foliage by white fungal hyphae 8 cm2 window area. The LED illumination (10% blue, 90% red) can be used to assess the level of infection. To estimate the is provided by the leaf chamber fluorimeter, and the chamber fraction of leaf area infected (percentage of leaf area covered temperature is regulated by Peltier elements. The measure- by the hyphae of E. alphitoides), both sides of the leaf were ments of the light-response curve were carried out in the fol- photographed (for example of a control leaf and a heavily lowing sequence (light intensities in μmol m−2 s−1): infected leaf, see Figure 1) at the end of each experiment and the infected areas were determined by UTHSCSA ImageTool 11000→→21 0051 00→→ 0008 00 2.0 (Dental Diagnostic Science, The University of Texas Health →→5400 00→→2 00 0 Science Center, San Antonio, TX, USA). The degree of infec- tion was separately estimated for the lower and upper leaf sur- Light-response curve measurements were carried out in tripli- faces, and also an average degree of infection was calculated cate for both control and seriously infected leaves (∼60% leaf for the entire leaf. area infected). Equations of von Caemmerer and Farquhar (1981) were Data analyses used to calculate the rates of net assimilation (A), transpira- The light (Q) response curves of emission rates (I) were fitted tion (E) and stomatal conductance to water vapor (gs) per unit by a hyperbolic equation in the form of: projected leaf area enclosed in the chamber. αQ I = , (1) Measurement of the emission rates of volatiles 1+ ()α22QI/ 2 max,Q Emission of VOCs was measured in parallel with physiological parameters by diverting a part of the outgoing air to a proton- where α is the quantum yield and Imax,Q is the emission capac- transfer reaction mass spectrometer (high sensitivity version of ity (Sun et al. 2012). We note that in VOC emission studies, PTR-QMS, Ionicon, Innsbruck, Austria). Isoprene was detected light responses are often fitted to normalized data (so-called as a protonated parent ion at m/z of 69, while monoterpenes Guenther et al. model) (Guenther et al. 1993, Grote et al.

Tree Physiology Online at http://www.treephys.oxfordjournals.org 1402 Copolovici et al.

Results Variation in the degree of oak powdery mildew (E. alphitoides) infection

The infection by E. alphitoides was more prevalent on the upper leaf surface (average ± SE = 49.5 ± 4.1% across all leaves) than on the lower leaf surface (28.8 ± 4.1%, means are signifi- cantly different at P < 0.001). The degrees of infection of the lower and upper surface were positively, but relatively weakly, correlated (r2 = 0.32, P < 0.03). The correlations of the degree of infection with foliage physiological characteristics and emis- sion rates were stronger with the degree of upper leaf surface infection than with the degree of lower leaf surface infection or the average degree of infection of both surfaces (data not shown). Thus, in the following section, we only report the cor- Downloaded from relations with the degree of infection of the upper leaf surface.

Photosynthesis characteristics of leaves infected with E. alphitoides

Infection by E. alphitoides reduced leaf net CO2 assimilation rates, http://treephys.oxfordjournals.org/ and even a moderate infection of 10–20% leaf area resulted in significantly reduced net assimilation rates of ∼4 μmol m−2 s−1 (Figure 2a). With further increases in infection, the net assimi- lation rate decreased, reaching values of 1–2 μmol m−2 s−1 in leaves with ∼60% infection. Differently from the net assimilation rate, stomatal conduc- tance to water vapor increased with increasing percentage of leaf infection by >60%, from 91 ± 11 mmol m−2 s−1 (aver- at Tartu University on March 3, 2015 age ± SE) in control plants to ∼150 mmol m−2 s−1 in leaves with >50% infection (Figure 2b). As a result of the decreases in net Figure. 1 Representative photographs of control (a, b) and oak pow- dery mildew (E. alphitoides)-infected (c, d) leaves of Q. robur. The assimilation rate and increases in stomatal conductance, the images for both the upper (a, c) and lower sides (b, d) of the leaves intercellular CO2 concentration increased with increasing the are demonstrated. degree of leaf infection (Figure 2c).

2013). However, in the original Guenther et al. model, the yield Modification of light response of photosynthesis parameter is implicitly dependent on the emission capacity by mildew infection (Monson et al. 2012, Grote et al. 2013), and we therefore favor Dark respiration rate was similar, 1.6 μmol m−2 s−1 for both con- Eq. (1), which includes the true quantum yield. trol and mildew-infected leaves, but mildew infection resulted The statistical dependencies between the degree of infection in moderate reductions in the light-limited photosynthesis and the emission of volatiles were explored by linear regres- between light intensities 200 and 600 μmol m−2 s−1, and major sions. Separate regressions with the percentage of infection changes in net assimilation rates at high light, ∼2.3-fold from visible in the upper and lower leaf surfaces and the average ∼9 μmol m−2 s−1 in control to 3.9 ± 0.8 μmol m−2 s−1 (aver- for upper and lower surface damage were calculated to com- age ± SE) in infected leaves (Figure 3a). Furthermore, net pare the descriptive power of different damage indicators. The assimilation rates in infected leaves saturated at a lower quan- parameters of Eq. (1) were compared among different com- tum flux density of ∼800 μmol m−2 s−1, while the saturation pounds and for infected and non-infected plants using ANOVA. light intensity was reached at ∼1500 μmol m−2 s−1 in control ANOVA was also used to analyze the differences in the content plants (Figure 3a). of various volatiles among control and heavily infected (∼60% Stomatal conductance increased with increasing light inten- infection) plants. A paired sample t-test was used to assess the sity similar to net assimilation rates (Figure 3b). As with net statistical differences in the degree of infection of the upper assimilation rates, stomatal conductance reached a light- and lower surfaces. All statistical tests were considered signifi- saturated value at a lower light intensity in infected leaves cant at P < 0.05. (cf. Figure 3a and b). Thus, while the values of stomatal

Tree Physiology Volume 34, 2014 Volatile emissions scale with the degree of infection 1403 Downloaded from http://treephys.oxfordjournals.org/

Figure. 3 Comparison of representative light responses of net CO2 assimilation rate (a) and stomatal conductance to water vapor (b) among Q. robur control (non-infected) and E. alphitoides-infected leaves (percentage of leaf area infected ∼60%, see Table 2). Data are

averages ± SE (n = 3 for both control and infected leaves). at Tartu University on March 3, 2015 (50–60% leaf area infected), and a strong negative correlation between the isoprene emission rate and the percentage of leaf damage was observed (Figure 4a). Emission of a variety of compounds was induced by infec- tion, including volatiles of the lipoxygenase pathway ((Z)-3- hexenol, (E)-2-hexenal, (Z)-3-hexenyl acetate and 1-hexanol),

Figure 2. Leaf net CO2 assimilation rate (a), stomatal conductance to ubiquitous monoterpenes (α-pinene, camphene, 3-carene, water vapor (b) and intercellular CO2 concentration (c) in relation to limonene and β-phellandrene), typical stress monoterpenes the percentage of leaf area (upper surface) infected with E. alphitoides in Q. robur. The data correspond to different leaves and the relation- such as (E)-β-ocimene, linalool and monoterpene derivative ships were fitted by linear regressions. Values for visually completely geranyl acetone and benzenoid methyl salicylate (Table 1). healthy non-infected leaves are shown by open symbols. Data are The emissions of total monoterpenes (Figure 4b) and the sum shown separately for 2011 and 2013 measurement campaigns. of lipoxygenase pathway products (LOX products, Figure 4c) increased from close to zero in control leaves to values as high conductance in infected leaves exceeded those in control as ∼0.4 nmol m−2 s−1 in strongly infected leaves (Figure 4b leaves over most of the light range, similar stomatal conduc- and c). Isoprene emission was negatively correlated with both tance of 140 ± 8 mmol m−2 s−1 (average ± SE) was reached at monoterpene and LOX product emission rates for all leaves the highest light intensity of 1500 μmol m−2 s−1 (Figure 3b). with different degrees of infection pooled (Figure 5).

Effects of E. alphitoides infection on isoprene and Light responses of volatile emissions in control stress-elicited volatile emissions and infected plants Infection by E. alphitoides led to a reduction of isoprene emis- Infection by E. alphitoides resulted in both reduced light- sion from 10.6 ± 0.6 nmol m−2 s−1 (average ± SE) in healthy saturated isoprene emission rate and initial quantum yield leaves of Q. robur to 4–5 nmol m−2 s−1 in heavily infected leaves (Figure 6a, Table 2). In addition, isoprene emission in infected

Tree Physiology Online at http://www.treephys.oxfordjournals.org 1404 Copolovici et al.

Table. 1 Volatile emission rates (nmol m−2 s−1) of Q. robur control leaves and oak powdery mildew (E. alphitoides)-infected leaves. Control leaves had essentially no visible symptoms of infection (infec- tion level < 5%), while in the case of infected leaves, leaves with the infection level >60% (59.8–65.9%, on average ± SE of 61.6 ± 0.9%) were included in the analysis (see Figure 4). The means were com- pared by ANOVA and the significance is denoted as P < 0.001 (***), P < 0.01 (**) or P < 0.05 (*). nd, not determined. Compound Control Mildew-infected

Green leaf volatiles (lipoxygenase pathway volatiles, LOX) (Z)-3-Hexenol 0.0139 ± 0.0022 0.0926 ± 0.0043*** (E)-2-Hexenal 0.021 ± 0.006 0.149 ± 0.007*** (Z)-3-Hexenyl 0.0121 ± 0.0 011 0.098 ± 0.007*** acetate 1-Hexanol 0.0119 ± 0.0018 0.082 ± 0.006*** Isoprenoids and derivatives

Isoprene 10.6 ± 0.6 5.36 ± 0.16** Downloaded from α-Pinene 0.0184 ± 0.0038 0.150 ± 0.007*** Camphene 0.0070 ± 0.0018 0.056 ± 0.005*** β-Pinene 0.0034 ± 0.0005 0.0177 ± 0.0018*** 3-Carene 0.00363 ± 0.00041 0.0520 ± 0.0020*** Limonene 0.0080 ± 0.0009 0.0678 ± 0.0036*** http://treephys.oxfordjournals.org/ (E)-β-Ocimene 2.5 × 10−5 ± 0.8 × 10−5 0.0004 ± 0.00005*** β-Phellandrene 0.0036 ± 0.0008 0.0206 ± 0.0009*** Linalool nd 0.0164 ± 0.0027*** Geranyl acetone 0.0072 ± 0.0013 0.0112 ± 0.0006* Aromatics Benzaldehyde 0.079 ± 0.012 0.122 ± 0.016 Methyl salicylate nd 0.437 ± 0.011*** at Tartu University on March 3, 2015

Figure. 4 Correlations of isoprene (a, ΦIso), monoterpene (b, ΦMon) and lipoxygenase pathway volatiles (LOX, ΦLox, various C6 alcohols, alde- hydes and derivatives, also called green leaf volatiles), (c) emission rates with the percentage of E. alphitoides infection (upper surface) in Q. robur leaves. The sum of all monoterpenes and LOX volatiles is reported in Table 1. Data presentation and fitting as in Figure 2. The regressions with the degree of infec tion (DI) are: ΦIso = 8.91 − 0.0743DI; ΦMon = 0.0732 + 0.00532DI; ΦLox = 0.00213 + 0.00624DI. Figure 5. Correlations between isoprene emission rate and monoter- pene and LOX volatile emission rates among the Q. robur leaves with different degrees of E. alphitoides infection. Data were fitted by lin- leaves also saturated at a lower quantum flux density of∼ 600– ear regressions (solid line, relationship with monoterpene emission; 2 1 800 μmol m− s− than in the non-infected control leaves where dashed line, relationship with LOX volatile emission). the emission rate was still increasing at the highest light level of 1500 μmol m−2 s−1 used (Figure 6a). Both monoterpene (Figure 6b, Table 2) and LOX product (Figure 6c, Table 2) was close to linearly correlated with light-dependent changes emission rates also increased with increasing light intensity. in stomatal conductance (inset in Figure 6c). The initial quan- The rate of emission of LOX products increased particularly tum yields for monoterpene and LOX product emissions in strongly with light without apparent light saturation (Figure 6c). infected leaves were lower than those for isoprene emission in The increase of LOX emission rate through the light response control and infected leaves (Table 2).

Tree Physiology Volume 34, 2014 Volatile emissions scale with the degree of infection 1405

Table. 2 Comparison of light-response curve parameters, the initial

quantum yield (α) and the emission capacity (Imax,Q, Eq. (1)) for dif- ferent compound classes and among the control and E. alphitoides- infected leaves of Q. robur. Infected leaves had the infection level >60% (59.8–65.9%, on average ± SE of 61.6 ± 0.9%). Light-response curve parameters for monoterpenes and LOX emissions are only provided for the infected leaves. The emission rates of these compound classes in control leaves were close to the detection limit (Table 1). Means of the parameter values were compared by ANOVA and different letters denote statistically significant differences at P < 0.05.

−1 −2 −1 Compound α (μmol mol ) Imax,Q (nmol m s )

Isoprene (control) 23 ± 6A 12.8 ± 0.6A Isoprene (infected) 16.2 ± 3.5B 6.14 ± 0.06B Monoterpenes (infected) 0.586 ± 0.023C 0.55 ± 0.07C LOX volatiles (infected) 0.532 ± 0.016C 0.73 ± 0.06C Downloaded from photosynthesis was reduced in Vaccinium ashei and Vaccinium corymbosum affected by leaf spot fungus Septoria albopunc- tata (Roloff et al. 2004), sugarcane infected with orange rust Puccinia kuehnii (Zhao et al. 2011) and Prunus cerasus

affected by cherry leaf spot fungus Blumeriella jaapii (Gruber http://treephys.oxfordjournals.org/ et al. 2012), in Quercus fusiformis infected by oak wilt Ceratocystis fagacearum (Anderson et al. 2000), and in Salix burjatica × Salix dasyclados infected by rust fungus Melampsora epitea (Toome et al. 2010). In contrast, in some cases, compensatory photosynthesis has been demonstrated. For example, infection by the rust fungus Puccinia hordei increased photosynthesis in Hordeum distichum compared with control plants (Scholes and Farrar 1986). Such compensatory responses have some- at Tartu University on March 3, 2015 times also been observed in herbivory-damaged leaves where increases in the photosynthetic rate of remaining leaf areas can compensate for consumed leaf area. Our study further demonstrated that the degree of reduc- tion in assimilation rate was quantitatively associated with the degree of leaf infection (Figure 2). Such a response can result from a uniform reduction in foliage assimilation rate across the leaf surface in a coordinated manner analogously to senesc- ing leaves undergoing programmed cell death (Munné-Bosch Figure. 6 Light-response curves of isoprene (a), monoterpene (b) and LOX volatile (c) emissions in control and E. alphitoides-infected leaves 2008, Niinemets et al. 2012). Such a response can also be of Q. robur. Data were fitted by Eq. (1) (Table 2 for the parameters). The explained by reductions in net assimilation rate in infected inset in (c) demonstrates the correlation between LOX volatile emis- leaf regions, especially in the case of necrotrophic patho- sion rate and stomatal conductance (gs) through the light-response curve (Figure 3b for the light response of stomatal conductance). gens, while the assimilation rate is maintained almost con- stant in non-infected leaf regions. In fact, reduction in the net Discussion assimilation rate by Vaccinium infected by leaf spot fungus S. albopunctata was correlated with the fraction of necrotic area Alteration of photosynthetic characteristics of Q. robur (Roloff et al. 2004). However, net assimilation rates reached by E. alphitoides infection close to zero already at the necrotic area of ∼50% (Roloff et al. Our results for Q. robur (Figure 2) are in agreement with previous 2004), suggesting that severe infection also resulted in overall observations demonstrating strong reduction in net assimilation reduction in net assimilation rate. In our study, net assimilation rate by various powdery mildew species (Gordon and Duniway rate reached very low values of 1–2 μmol m−2 s−1 in severely 1982, Moriondo et al. 2005, Hajji et al. 2009). Furthermore, infected leaves (50–60% leaf area infected), suggesting that reduction of photosynthesis is a general response to infection both mechanisms were likely responsible for reduction in net by different pathogens across a variety of species. For example, assimilation rates in E. alphitoides-infected leaves.

Tree Physiology Online at http://www.treephys.oxfordjournals.org 1406 Copolovici et al.

Multiple physiological mechanisms can be responsible for the Emission of isoprene and terpenoids from leaves reduction of photosynthesis in fungal-infected leaves. Among infected with E. alphitoides these mechanisms, changes characteristic to programmed cell Reduction of isoprene emissions by powdery mildew infection death including inhibition of photosynthetic electron transport (Figure 4a) is in agreement with observations of Brüggemann and reductions in the amount or activity of photosynthetic rate- and Schnitzler (2001). Analogously, isoprene emissions were limiting enzymes and light-harvesting pigments have been dem- reduced by oak wilt C. fagacearum infected Q. fusiformis onstrated for different fungal infections (Gordon and Duniway (Anderson et al. 2000), and by rust fungus M. epitea-infected 1982, Tang et al. 1996, Mandal et al. 2009, Zhao et al. 2011). Salix (Toome et al. 2010). Our study further demonstrates that Pathogenic infections can also induce reductions in stoma- the reduction of isoprene emission in powdery mildew-infected tal conductance by rendering stomata in necrotic areas non- leaves is quantitatively associated with the degree of propaga- functional and by affecting stomatal functioning in surrounding tion of the infection (Figure 4a). leaf tissues (Lopes and Berger 2001, Pinkard and Mohammed However, compared with the net assimilation rate, the reduc- 2006). In fact, decreases in stomatal conductance have often tion in isoprene emission was less under most severe infection been observed in fungal-infected plants, especially when infec- (cf. Figures 2a and 4a), indicating that the fraction of carbon tion led to necrotic lesions (Roloff et al. 2004, Moriondo et al. going into isoprene emission increased. On the other hand, iso- Downloaded from 2005, Mandal et al. 2009, Toome et al. 2010). However, in our prene emission is more tightly linked to the rate of photosyn- study, stomatal conductance actually increased with increase thetic electron transport than to light-saturated net assimilation in the degree of infection (Figure 2b) and this was associated (Niinemets et al. 1999, Rasulov et al. 2009, 2011, Harrison with increased intercellular CO2 concentration (Figure 2c). This et al. 2013). Thus, lower sensitivity of light-limited photosynthe- suggests that the reduction in net assimilation rate in E. alphitoi- sis to infection can provide an explanation for smaller decreases http://treephys.oxfordjournals.org/ des infected Q. robur leaves was associated with reductions in isoprene emissions than in photosynthetic capacity. in the amount and/or activity of rate-limiting proteins such as Nevertheless, isoprene emission rates also depend on inter- ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco). cellular CO2 concentration according to a curve with a maximum Impairment of stomatal conductance and photosynthesis and at ∼100 μmol mol−1 (Loreto and Sharkey 1993, Rasulov et al. resulting enhanced water loss in infected plants have also 2009, Sun et al. 2012). This reduction has been associated been observed in other studies (Robert et al. 2005, Major with CO2 effects on the pool size of the substrate for isoprene et al. 2009). In addition, despite the parallel reductions in net synthesis, dimethylallyl diphosphate (DMADP) (Rasulov et al. assimilation rate and stomatal conductance in downy mildew 2009, Possell and Hewitt 2011). Thus, increases in the inter- at Tartu University on March 3, 2015 Peronospora plantaginis-infected Plantago ovata leaves, intercel- cellular CO2 concentration of E. alphitoides leaves might have lular CO2 concentration actually increased in infected leaves, also suppressed isoprene emissions by reduction of DMADP also indicating that the demand for CO2 decreased relatively pool size. In fact, reduction in the initial quantum yield of iso- more than the supply of CO2 (Mandal et al. 2009). On the other prene emission (Figure 6a, Table 2) also suggests that the hand, in V. vinifera leaves infected by powdery mildew Uncinula DMADP pool size was reduced in the infected leaves. We note necator, intercellular CO2 concentration decreased in infected that the relationships among isoprene emission, photosynthetic leaves (Moriondo et al. 2005). rate and CO2 concentration differ between the infected and Overall, this evidence collectively indicates that the photo- non-infected leaves. In particular, in non-infected leaves, the synthetic responses to infection are strongly host and patho- reduction of isoprene emissions at higher CO2 is characteris- gen specific. In the case of biotrophic pathogen E. alphitoides tically associated with increased photosynthetic rate (Loreto infecting Q. robur leaves, clearly a partial loss of stomatal con- and Sharkey 1990, Morfopoulos et al. 2014), but this was not trol increased intercellular CO2 concentration and reduced the the case for infected leaves where both photosynthesis and control of assimilation rates by CO2 concentration. However, the isoprene emission rates were reduced. light saturation of net assimilation rate (when cross-over from Apart from the CO2 effects, the DMADP pool size depends light limitation to Rubisco limitation occurred) was observed at on the competition between isoprene emission and other a lower light intensity in the infected leaves compared with the DMADP consuming chloroplastic pathways such as synthe- control leaves (Figure 3). This suggests that Rubisco activity sis of larger isoprenoids including monoterpenes, diterpenes decreased relatively more than the activity of photosynthetic and carotenoids (Dudareva et al. 2013, Rasulov et al. 2014). electron transport in the infected leaves. Analogously, a stron- Because geranyl diphosphate synthase, the initial enzyme in ger effect of infection on high-light photosynthesis than on the synthesis of longer chain-length isoprenoids, has much light-limited photosynthesis has been reported in several plant higher affinity toward DMADP than the isoprene synthase species including rice (Oryza sativa) (Bastiaans and Roumen (Rajabi Memari et al. 2013, Rasulov et al. 2014), competition 1993), soybean (Glycine max) (Kumudini et al. 2008) and by alternative reactions can significantly curb isoprene emis- sugar beet (B. vulgaris) (Gordon and Duniway 1982). sion. Thus, part of the inhibition of isoprene emission can be

Tree Physiology Volume 34, 2014 Volatile emissions scale with the degree of infection 1407 associated with elicitation of monoterpene emissions from the sites of emission. In the case of mildew infection, sustained infected leaves. In fact, close to proportional negative corre- LOX emissions likely reflect propagation of cellular damage to lation between isoprene emission and monoterpene emission leaf areas not yet infected (Toome et al. 2010) as well as con- was observed through all leaves with varying degrees of infec- tinuous repair and damage of infected leaf regions. We suggest tion (Figure 5), supporting the hypothesis of the competition of that detailed anatomical studies are needed to gain insight into different biochemical processes for DMADP. the nature of powdery mildew-induced LOX emissions at the ultrastructural level. Powdery mildew-induced volatile emissions Surprisingly, LOX product emission rates were light depen- Studies with different fungal pathogens have demonstrated dent (Figure 6c). Such light dependence has not yet been that infection leads to induction of emissions of several classes demonstrated, and we suggest that several factors could be of stress volatiles. For example, infection of willow (Salix) responsible for the light dependence of these emissions. First, leaves by the rust fungus M. epitea resulted in emissions of the increase of emissions of green leaf volatiles with light volatiles of the lipoxygenase (LOX) pathway, and mono- and intensity can result from light-dependent formation of ROS due sesquiterpenes (Toome et al. 2010). Analogously, silver birch to photooxidative stress, and accordingly greater lipoxygen- (Betula pendula) infected with the leaf spot fungus Marssonina ase activity and higher level of lipid peroxidation. In fact, an Downloaded from betulae emitted an increased amount of (Z)-ocimene and (E)- increase in lipoxygenase activity with increasing light level has β-ocimene (Vuorinen et al. 2007). Tomato (Solanum lycop- been observed in Phalaenopsis (Ali et al. 2005). Second, as ersicum) plants inoculated with necrotrophic fungus Botrytis water-soluble compounds, several LOX products can be tem- cinerea emitted a different number and amount of volatiles after porarily controlled by the stomata (Niinemets and Reichstein

inoculation compared with control plants (Jansen et al. 2009). 2003, Harley 2013), and thus, light-dependent changes in http://treephys.oxfordjournals.org/ Although E. alphitoides itself could emit different VOCs, as stomatal conductance could be partly responsible for the light an obligate pathogen, its VOC emissions cannot be measured dependence of LOX product emission. This suggestion is cor- in the absence of plants in vivo, and thus, we were unable roborated by a strong positive correlation between stomatal to determine the profile of possibly emitted fungal volatiles. conductance and LOX product emission (Figure 6c). According to Peñuelas et al. (2014), fungal emissions are dom- Activation of defense pathways following infection is typi- inated by alcohols, benzenoids, aldehydes and ketones, and cally associated with the release of volatile isoprenoids, in there are only marginal emissions of terpenes. However, there particular, with light-dependent emissions of monoterpenes is a large variability in the emission profiles among different (Vuorinen et al. 2007, Toome et al. 2010) as was also dem- at Tartu University on March 3, 2015 fungal species (Müller et al. 2013). onstrated in powdery mildew-infected Q. robur (Table 1, In our study, the main compounds emitted were the LOX Figures 4b and 6b). As discussed above, these emissions products, monoterpenes and methyl salicylate that are charac- likely compete with constitutive isoprene synthesis for chlo- teristic of fungal infections. As a result of an attack by patho- roplastic DMADP. However, while the DMADP pool size likely gens, specific elicitor molecules are generated by chemical or decreased with increasing degree of infection as suggested physical damage to plant cell walls and cellular membranes. by the reductions in the initial quantum yield for isoprene Usually the response of plants to the stress includes formation emission (Table 2), monoterpene emission rate increased of different reactive compounds which can be detected and with increasing degree of infection (Figure 4b). This indicates recognized as a signal and initiate the activation of secondary that the activity of monoterpene synthases scaled positively elicitors and transcription of defense genes (Panka et al. 2013). with the degree of leaf infection. On the other hand, the light Rapid emissions of LOX products are triggered by the release dependence of monoterpene emission (Figure 6b) suggests of free fatty acids from cell membranes, and their peroxidation that the light-dependent changes in the monoterpene precur- by lipoxygenase enzymes (Liavonchanka and Feussner 2006). sor pool size did affect monoterpene emission. This might As inoculation experiments with fungal pathogens demonstrate reflect the competition by isoprene synthase that at any light (Jansen et al. 2009, Toome et al. 2010), LOX products are level consumed more DMADP than monoterpene synthases. emitted within hours after inoculation. This is somewhat slower compared with mechanical damage or herbivory damage (see Conclusions the Introduction section), and likely reflects differences in time when the inoculation signal is perceived. In the case of her- This study provides important evidence of strong effects of bivory damage, it has been demonstrated that LOX product pathogenic infections on constitutive and induced volatile emis- emission rate scales quantitatively with the degree of leaf dam- sions in Q. robur. As the study demonstrates, powdery mildew age (Copolovici et al. 2005, 2014, Niinemets et al. 2011). In infection in oak is associated with infection-dependent reduc- the case of herbivory, LOX emissions have been associated tions in photosynthesis, and isoprene emission and increases with continuous increase of damaged leaf parts that serve as in stomatal conductance and induced emissions. These results

Tree Physiology Online at http://www.treephys.oxfordjournals.org 1408 Copolovici et al. provide encouraging evidence that the severity of pathogen Copolovici L, Filella I, Llusià J, Niinemets Ü, Peñuelas J (2005) The stress can be assessed by monitoring the emissions of stress- capacity for thermal protection of photosynthetic electron trans- port varies for different monoterpenes in Quercus ilex. Plant Physiol elicited volatiles. Such quantitative relationships form a valuable 139:485 –496. basis to incorporate pathogen stress in to isoprene emission Copolovici L, Kännaste A, Remmel T, Niinemets Ü (2014) Volatile models and in to models of induced volatile emissions. organic compound emissions from Alnus glutinosa under interacting drought and herbivory stresses. Environ Exp Bot 100:55–63. Desprez-Loustau ML, Feau N, Mougou-Hamdane A, Dutech C Acknowledgments (2011) Interspecific and intraspecific diversity in oak powdery mil- dews in Europe: coevolution history and adaptation to their hosts. The authors thank Kadri Põldmaa (University of Tartu) for the Mycoscience 52:165–173. identification of the powdery mildew species. Dudareva N, Klempien A, Muhlemann JK, Kaplan I (2013) Biosynthesis, function and metabolic engineering of plant volatile organic com- pounds. New Phytol 198:16–32. Conflict of interest Farag MA, Paré PW (2002) C6-green leaf volatiles trigger local and systemic VOC emissions in tomato. Phytochemistry 61:545–554. None declared. Filella I, Wilkinson MJ, Llusià J, Hewitt CN, Peñuelas J (2007) Volatile organic compounds emissions in Norway spruce (Picea abies) in

response to temperature changes. Physiol Plant 130:58–66. Downloaded from Funding Gordon TR, Duniway JM (1982) Effects of powdery mildew infection on the efficiency of CO2 fixation and light utilization by sugar beet Funding for this study has been provided by the Estonian leaves. Plant Physiol 69:139–142. Ministry of Science and Education (institutional grant IUT- Grote R, Monson RK, Niinemets Ü (2013) Leaf-level models of con- stitutive and stress-driven volatile organic compound emissions. In: 8-3), the European Commission through the European Niinemets Ü, Monson RK (eds) Biology, controls and models of tree http://treephys.oxfordjournals.org/ Regional Fund (the Center of Excellence in Environmental volatile organic compound emissions. Springer, Berlin, pp 315–355. Adaptation), the European Research Council (advanced grant Gruber BR, Kruger EL, McManus PS (2012) Effects of cherry leaf 322,603, SIP-VOL+) and the Romanian National Authority for spot on photosynthesis in tart cherry ‘Montmorency’ foliage. Phytopathology 102:656–661. Scientific Research, CNCS—UEFISCDI (project number PN-II- Guenther AB, Zimmerman PR, Harley PC, Monson RK, Fall R (1993) RU-TE-2011-3-0022). Isoprene and monoterpene emission rate variability: model evalua- tions and sensitivity analyses. J Geophys Res 98:12609–12617. Gururani MA, Venkatesh J, Upadhyaya CP, Nookaraju A, Pandey SK, References Park SW (2012) Plant disease resistance genes: current status and

future directions. Physiol Mol Plant Pathol 78:51–65. at Tartu University on March 3, 2015 Alexander HM, Holt RD (1998) The interaction between plant com- Hajji M, Dreyer E, Marcais B (2009) Impact of Erysiphe alphitoides on petition and disease. Perspect Plant Ecol Evol Syst 1–2:206–220. transpiration and photosynthesis in Quercus robur leaves. Eur J Plant Ali MB, Hahn EJ, Paek KY (2005) Effects of light intensities on anti- Pathol 125:63–72. oxidant enzymes and malondialdehyde content during short-term Harley PC (2013) The roles of stomatal conductance and compound acclimatization on micropropagated Phalaenopsis plantlet. Environ volatility in controlling the emission of volatile organic compounds Exp Bot 54:109–120. from leaves. In: Niinemets Ü, Monson RK (eds) Biology, controls Anderson LJ, Harley PC, Monson RK, Jackson RB (2000) Reduction of and models of tree volatile organic compound emissions. Springer, isoprene emissions from live oak (Quercus fusiformis) with oak wilt. Berlin, pp 181–208. Tree Physiol 20:1199–1203. Harrison SP, Morfopoulos C, Dani KGS et al. (2013) Volatile isoprenoid Bastiaans L, Roumen EC (1993) Effect on leaf photosynthetic rate by emissions from plastid to planet. New Phytol 197:49–57. leaf blast for rice cultivars with different types and levels of resis- Hartikainen K, Nerg A-M, Kivimäenpää M, Kontunen-Sopplea S, tance. Euphytica 66:81–87. Mäenpää M, Oksanen E, Rousi M, Holopainen T (2009) Emissions Beauchamp J, Wisthaler A, Hansel A, Kleist E, Miebach M, Niinemets Ü, of volatile organic compounds and leaf structural characteristics of Schurr U, Wildt J (2005) Ozone induced emissions of biogenic VOC European aspen (Populus tremula) grown under elevated ozone and from tobacco: relations between ozone uptake and emission of LOX temperature. Tree Physiol 29:1163–1173. products. Plant Cell Environ 28:1334–1343. Jansen RMC, Miebach M, Kleist E, van Henten EJ, Wildt J (2009) Berger S, Sinha AK, Roitsch T (2007) Plant physiology meets phytopa- Release of lipoxygenase products and monoterpenes by tomato thology: plant primary metabolism and plant–pathogen interactions. plants as an indicator of Botrytis cinerea-induced stress. Plant Biol J Exp Bot 58:4019–4026. 11:859 – 86 8 . Brilli F, Ruuskanen TM, Schnitzhofer R, Müller M, Breitenlechner M, Kännaste A, Copolovici L, Niinemets Ü (2014) Gas chromatography Bittner V, Wohlfahrt G, Loreto F, Hansel A (2011) Detection of plant mass-spectrometry method for determination of biogenic volatile volatiles after leaf wounding and darkening by proton transfer reaction organic compounds emitted by plants. In: Rodríguez-Concepción M “time-of-flight” mass spectrometry (PTR-TOF). PloS ONE 6:e20419. (ed) Plant isoprenoids: methods and protocols. Humana Press, New Brüggemann N, Schnitzler JP (2001) Influence of powdery mildew York, pp 161–169. (Microsphaera alphitoides) on isoprene biosynthesis and emission Koornneef A, Pieterse CMJ (2008) Cross talk in defense signaling. of pedunculate oak (Quercus robur L.) leaves. J Appl Bot 75:91–96. Plant Physiol 146:839–844. Copolovici L, Niinemets Ü (2010) Flooding induced emissions of vola- Kumudini S, Prior E, Omielan J, Tollenaar M (2008) Impact of tile signalling compounds in three tree species with differing water- Phakopsora pachyrhizi infection on soybean leaf photosynthesis and logging tolerance. Plant Cell Environ 33:1582–1594. radiation absorption. Crop Sci 48:2343–2350.

Tree Physiology Volume 34, 2014 Volatile emissions scale with the degree of infection 1409

Liavonchanka A, Feussner N (2006) Lipoxygenases: occurrence, Panka D, Piesik D, Jeske M, Baturo-Ciesniewska A (2013) Production of functions and catalysis. J Plant Physiol 163:348–357. phenolics and the emission of volatile organic compounds by peren- Lopes DB, Berger RD (2001) The effects of rust and anthracnose nial ryegrass (Lolium perenne L.)/Neotyphodium lolii association as a on the photosynthetic competence of diseased bean leaves. response to infection by Fusarium poae. J Plant Physiol 170:1010–1019. Phytopathology 91:212–220. Peñuelas J, Asensio D, Tholl D, Wenke K, Rosenkranz M, Piechulla B, Loreto F, Schnitzler J-P (2010) Abiotic stresses and induced BVOCs. Schnitzler JP (2014) Biogenic volatile emissions from soil. Plant Cell Trends Plant Sci 15:154–166. Environ 37:1866–1891. Loreto F, Sharkey TD (1990) A gas-exchange study of photosynthesis Pieterse CMJ, Dicke M (2007) Plant interactions with microbes and and isoprene emission in Quercus rubra L. Planta 182:523–531. insects: from molecular mechanisms to ecology. Trends Plant Sci Loreto F, Sharkey TD (1993) On the relationship between isoprene 12:564–569. emission and photosynthetic metabolites under different environ- Pinkard EA, Mohammed CL (2006) Photosynthesis of Eucalyptus mental conditions. Planta 189:420–424. globulus­ with Mycosphaerella leaf disease. New Phytol 170:119–127. Major IT, Nicole MC, Duplessis S, Seguin A (2009) Photosynthetic and Poland JA, Balint-Kurti PJ, Wisser RJ, Pratt RC, Nelson RJ (2009) respiratory changes in leaves of poplar elicited by rust infection. Shades of gray: the world of quantitative disease resistance. Trends Photosynth Res 104:41–48. Plant Sci 14:21–29. Mandal K, Saravanan R, Maiti S, Kothari IL (2009) Effect of downy Possell M, Hewitt CN (2011) Isoprene emissions from plants are

mildew disease on photosynthesis and chlorophyll fluorescence in mediated by atmospheric CO2 concentrations. Glob Change Biol Plantago ovata Forsk. J Plant Dis Protect 116:164–168. 17:1595–1610.

Matsui K (2006) Green leaf volatiles: hydroperoxide lyase pathway of Possell M, Loreto F (2013) The role of volatile organic compounds Downloaded from oxylipin metabolism. Curr Opin Plant Biol 9:274–280. in plant resistance to abiotic stresses: responses and mechanisms. Monson RK, Grote R, Niinemets Ü, Schnitzler J-P (2012) Tansley In: Niinemets Ü, Monson RK (eds) Biology, controls and models review. Modeling the isoprene emission rate from leaves. New of tree volatile organic compound emissions. Springer, Berlin, Phytol 195:541–559. pp 209–235. Morfopoulos C, Sperlich D, Peñuelas J et al. (2014) A model of plant Rajabi Memari H, Pazouki L, Niinemets Ü (2013) The biochemistry isoprene emission based on available reducing power captures and molecular biology of volatile messengers in trees. In: Niinemets http://treephys.oxfordjournals.org/

responses to atmospheric CO2. New Phytol 203:125–139. Ü, Monson RK (eds) Biology, controls and models of tree volatile Moriondo M, Orlandini S, Giuntoli A, Bindi M (2005) The effect of organic compound emissions. Springer, Berlin, pp 47–93. downy and powdery mildew on grapevine (Vitis vinifera L.) leaf gas Rasulov B, Hüve K, Välbe M, Laisk A, Niinemets Ü (2009) Evidence that exchange. J Phytopathol 153:350–357. light, carbon dioxide and oxygen dependencies of leaf isoprene emission Mougou A, Dutech C, Desprez-Loustau ML (2008) New insights into are driven by energy status in hybrid aspen. Plant Physiol 151:448–460. the identity and origin of the causal agent of oak powdery mildew in Rasulov B, Hüve K, Laisk A, Niinemets Ü (2011) Induction of a longer- Europe. For Pathol 38:275–287. term component of isoprene release in darkened aspen leaves: ori- Mougou-Hamdane A, Giresse X, Dutech C, Desprez-Loustau M-L gin and regulation under different environmental conditions. Plant (2010) Spatial distribution of lineages of oak powdery mildew Physiol 156:816–831.

fungi in France, using quick molecular detection methods. Ann Rasulov B, Bichele I, Laisk A, Niinemets Ü (2014) Competition between at Tartu University on March 3, 2015 For Sci 67:212. isoprene emission and pigment synthesis during leaf development Müller A., Faubert P, Hagen M, zu Castell W, Polle A, Schnitzler J- in aspen. Plant Cell Environ 37:724–741. P, Rosenkranz M (2013) Volatile profiles of fungi—chemotyping of Robert C, Bancal MO, Ney B, Lannou C (2005) Wheat leaf photosyn- species and ecological functions. Fungal Genet Biol 54:25–33. thesis loss due to leaf rust, with respect to lesion development and Munné-Bosch S (2008) Do perennials really senesce? Trends Plant leaf nitrogen status. New Phytol 165:227–241. Sci 13:216–220. Roloff I, Scherm H, van Iersel MW (2004) Photosynthesis of blueberry Niinemets Ü (2010a) Mild versus severe stress and BVOCs: thresh- leaves as affected by Septoria leaf spot and abiotic leaf damage. olds, priming and consequences. Trends Plant Sci 15:145–153. Plant Dis 88:397–401. Niinemets Ü (2010b) Responses of forest trees to single and multiple Scholes JD, Farrar JF (1986) Increased rates of photosynthesis in environmental stresses from seedlings to mature plants: past stress localized regions of a barley leaf infected with brown rust. New history, stress interactions, tolerance and acclimation. For Ecol Manag Phytol 104:601–612. 260:1623–1639. Schulze-Lefert P, Panstruga R (2011) A molecular evolutionary con- Niinemets Ü, Reichstein M (2003) Controls on the emission of plant cept connecting nonhost resistance, pathogen host range, and volatiles through stomata: sensitivity or insensitivity of the emis- pathogen speciation. Trends Plant Sci 16:117–125. sion rates to stomatal closure explained. J Geophys Res Atmos Semiz G, Blande J, Heijari J, Işik K, Niinemets Ü, Holopainen JK (2012) 108:4208. Manipulation of VOC emissions with methyl jasmonate and carra- Niinemets Ü, Tenhunen JD, Harley PC, Steinbrecher R (1999) A model geenan in evergreen conifer Pinus sylvestris and evergreen broad- of isoprene emission based on energetic requirements for isoprene leaf Quercus ilex. Plant Biol 14:57–65. synthesis and leaf photosynthetic properties for Liquidambar and Staudt M, Jackson B, El-Aouni H, Buatois B, Lacroze J-P, Poëssel J-L, Quercus. Plant Cell Environ 22:1319–1336. Sauge M-H (2010) Volatile organic compound emissions induced Niinemets Ü, Kuhn U, Harley PC et al. (2011) Estimations of isopren- by the aphid Myzus persicae differ among resistant and susceptible oid emission capacity from enclosure studies: measurements, data peach cultivars and a wild relative. Tree Physiol 10:1320–1334. processing, quality and standardized measurement protocols. Sun Z, Niinemets Ü, Hüve K, Noe SM, Rasulov B, Copolovici L, Vislap Biogeosciences 8:2209–2246. V (2012) Enhanced isoprene emission capacity and altered light

Niinemets Ü, García-Plazaola JI, Tosens T (2012) Photosynthesis during responsiveness in aspen grown under elevated atmospheric CO2 leaf development and ageing. In: Flexas J, Loreto F, Medrano H (eds) concentration. Glob Change Biol 18:3423–3440. Terrestrial photosynthesis in a changing environment. A molecular, Tang X, Rolfe SA, Scholes JD (1996) The effect of Albugo candida physiological and ecological approach. Cambridge University Press, (white blister rust) on the photosynthetic and carbohydrate metabo- Cambridge, pp 353–372. lism of leaves of Arabidopsis thaliana. Plant Cell Environ 19:967–975.

Tree Physiology Online at http://www.treephys.oxfordjournals.org 1410 Copolovici et al.

Toome M, Randjärv P, Copolovici L, Niinemets Ü, Heinsoo K, Luik A, Vuorinen T, Nerg A-M, Syrjala L, Peltonen P, Holopainen JK (2007) Epirrita Noe SM (2010) Leaf rust induced volatile organic compounds sig- autumnata induced VOC emission of silver birch differ from emission nalling in willow during the infection. Planta 232:235–243. induced by leaf fungal pathogen. Arthropod Plant Interact 1:159–165. von Caemmerer S, Farquhar GD (1981) Some relationships between Zhao D, Glynn NC, Glaz B, Comstock JC, Sood S (2011) Orange rust the biochemistry of photosynthesis and the gas exchange of leaves. effects on leaf photosynthesis and related characters of sugarcane. Plant a 153:376 –387. Plant Dis 95:640–647. Downloaded from http://treephys.oxfordjournals.org/ at Tartu University on March 3, 2015

Tree Physiology Volume 34, 2014 Trees (2015) 29:1805–1816 DOI 10.1007/s00468-015-1262-8

ORIGINAL ARTICLE

Bias in leaf dry mass estimation after oven-drying isoprenoid- storing leaves

1,2 1,3 1,4 Miguel Portillo-Estrada • Lucian Copolovici • U¨ lo Niinemets

Received: 23 February 2015 / Revised: 1 June 2015 / Accepted: 17 July 2015 / Published online: 26 July 2015 Ó Springer-Verlag Berlin Heidelberg 2015

Abstract underestimating the content of volatiles and overestimating Key message Air-drying, cooling and freezing leaf that of non-volatile constituents. Three leaf preservation samples rich in isoprenoid content evaporates iso- (flash freezing in liquid nitrogen, freezing at -82 and at prenoids and consequently causes an underestimation -20 °C, and cooling at 4 °C) and two leaf desiccation of the leaf dry matter content. (freeze-drying at -54 °C and 1.2 mbar, and air-drying at Abstract Foliage isoprenoid content or essential oil con- 30, 40 60, 70, 80, 100, 110, 125, 150, 200, and 300 °C) tent is a key plant characteristic important in plant science, procedures were applied to replicate leaves, and the leaf but also for food, cosmetic and pharmaceutical industry. dry matter content as well as the leaf isoprenoid content Commonly, the amount of foliage chemicals is normalized was measured afterwards. The results of this experiment with respect to leaf dry mass (content per dry mass). suggested that freeze-drying method (FD) performed the However, the protocols for foliage drying have received best leaf isoprenoid preservation followed by FF method, little attention. In particular, volatile and semi-volatile whilst air-drying methods performed a dramatic isoprenoid isoprenoids and other compounds with low boiling point loss in the leaf samples, as well as preserving the leaves at may partly volatilize during drying, reducing both iso- freezing and cooling temperatures. Leaf isoprenoid content prenoid and leaf dry matter contents, thereby potentially and leaf dry to fresh mass ratio were correlated and there were evidences of leaf dry matter content (LDMC) mis- calculation in isoprenoid-storing species when using tra- Communicated by H. Rennenberg. ditional air-drying methods and preservation of samples, & Miguel Portillo-Estrada possibly related to the isoprenoid loss as well as other [email protected] volatile compounds. In addition, losses of high C content volatiles and respiration of sugars and carbohydrate Lucian Copolovici [email protected] pyrolysis with low carbon content may differently affect the bulk leaf C content. U¨ lo Niinemets [email protected] Keywords Abies balsamea Á Betula pendula Á Leaf dry 1 Department of Plant Physiology, Institute of Agricultural and matter content Á Pinus mugo Á Quercus ilex Á Volatile Environmental Sciences, Estonian University of Life organic compounds Sciences, Kreutzwaldi 1, 51014 Tartu, Estonia 2 Department of Biology, Centre of Excellence PLECO (Plant and Vegetation Ecology), University of Antwerp, Universiteitsplein 1, B-2610 Wilrijk, Belgium Introduction 3 Institute of Technical and Natural Sciences Research- Development of Aurel Vlaicu University, Elena Dragoi 2, In plant science, several key leaf traits, including specific 310330 Arad, Romania leaf area and amounts of foliage chemical elements such as 4 Estonian Academy of Sciences, Kohtu 6, 10130 Tallinn, nitrogen, phosphorus and carbon as well as foliage chem- Estonia ical compounds such as isoprenoids, flavonoids, etc., are 123 1806 Trees (2015) 29:1805–1816 commonly expressed per leaf dry mass unit. Furthermore, 88 %), leaf carbon content can even be affected more the rate of plant physiological processes such as photo- strongly (Niinemets 1997; Poorter 1994). synthetic capacity is often expressed per unit dry mass to A wide range of methods are available for the measure- characterize the physiological activity of unit mass (Wright ment of leaf dry matter content (LDMC); also used previous et al. 2004). ‘Dry mass’ as an expression basis is favored leaf isoprenoid content determination. They can be summa- over fresh mass, because foliage fresh mass can change rized in: (1) placing leaves under a low relative humidity during the day and among the days due to changes in leaf desiccating atmosphere (i.e., desiccation with silica gel) water status in transpiring leaves (Correia et al. 2001; (Whittaker et al. 2001), (2) increasing the wind speed (air- Davey and Ellis-Evans 1996; Flexas and Medrano 2002; drying with a fan) (Maisuthisakul et al. 2007), (3) increasing Hinckley and Bruckerhoff 1975; Medina and Francisco the air temperature to create a low relative humidity envi- 1994; Tuba et al. 1996; Vos and Groenwold 1988). At the ronment as well as increasing leaf temperature and evaporate extreme, in desiccation-tolerant species, water content can the leaf water rapidly (incubation and sun-drying) (Akah et al. vary more than an order of magnitude (Davey and Ellis- 2003), (4) heating selectively the water molecules in the leaf Evans 1996; Tuba et al. 1996). Thus, dry mass is com- until evaporation (with a microwave) (Jeni et al. 2010), (5) monly considered as a more reliable expression basis than rapid cryogenic treatment of leaves which avoids water fresh mass (Garnier et al. 2001; Niinemets et al. 2011; crystallization and allows water sublimation (flash freezing in Zeiller et al. 2007). On the other hand, dry to water satu- liquid nitrogen), and (6) freezing the leaves and reducing the rated fresh mass ratio itself can be a relevant trait (Castrillo surrounding pressure to sublimate the frozen water (vacuum et al. 2005; Niinemets et al. 2007; Vile et al. 2005) in leaf freeze-drying, also called lyophilization or cryodesiccation) economics and leaf anatomy, again emphasizing the (Ormen˜oetal.2007, 2011). importance of reliable estimation of ‘‘dry’’ mass. But The most widely used method for LDMC determination estimating leaf dry mass carries itself potential problems. unites the air-drying (1) and heating desiccation (2) proce- Even in ‘‘dry’’ plant material, there is a residual amount dures. The time and temperatures used vary largely: com- of water in equilibrium with humidity in the air (Zeiller monly 60–80 °C during 48 h. The widespread protocol for et al. 2007). This adsorbed water is gradually released with measuring LDMC by Garnier et al. (2001)uses60°Cduring increasing temperature or reducing water vapor concen- at least 2 days. In the last decades, methods like freeze-dry- tration in ambient air, but increasing temperatures can also ing, flash freezing and microwave drying have won popu- result first in evaporation of compounds with low boiling larity. Freeze-drying at higher or lower pressure is used in point (such as volatile and semi-volatile isoprenoids). With food industry to preserve flavor quality, and performs better further increases in temperature, pyrolysis of certain leaf essential oils retention during the drying process than oven- constituents may occur (Font et al. 2009). On the other drying (Dı´az-Maroto et al. 2003). On the other hand, micro- hand, drying at physiological temperatures (e.g., 30 °C) wave drying has been presented as a good alternative to can be long-lasting and result in mass loss due to respira- freeze-drying for the determination of organic compounds tion. Some plant tissues, such as foliage of CAM plants, (Popp et al. 1996), also because microwave drying is less cannot be dried under physiological temperatures without a expensive and faster technique than freeze-drying. loss of bulk of dry matter. Volatilization of the studied This experiment is focused on the effect of different compounds during drying can led to major losses of the drying methods on the leaf isoprenoid content and the compound, and accordingly to significant underestimation consequences on the determination of leaf isoprenoid of their content, even when expressed per unit leaf area. content per dry matter content. The main hypothesis is that Foliage in species with specialized isoprenoid storage conifer leaves (which contain a large isoprenoid content), structures such as in conifers or species from the families when highly manipulated, stored or oven-dried, will dra- Lamiaceae (e.g., Mentha), Myrtaceae (e.g., Eucalyptus) matically loose volatile isoprenoids before the chemical and Rutaceae (e.g., Citrus), can contain large amounts of determination. Furthermore, when leaves are dried at high isoprenoids: between 1 and 10 % of total foliage dry mass temperature, this will significantly mislead the calculation (Pen˜uelas and Llusia` 2004). During leaf drying, the effects of dry matter content as a consequence of isoprenoid loss. of compound volatilization will be higher with the increasing of heating temperature and the volatile com- pounds content in the leaf. On the other hand, mass loss or Materials and methods variation in the content of residual water will affect the concentration of other non-volatile compounds, and esti- Plant material mation of mass-based content of chemical elements. Fur- thermore, if the mass lost has different carbon content than Foliage from two evergreen conifers, Abies balsamea (L.) the bulk of leaf matter (e.g., monoterpene carbon content is Mill. and Pinus mugo Turra, an evergreen (Quercus ilex L.) 123 Trees (2015) 29:1805–1816 1807 and a deciduous (Betula pendula Roth) broad-leaved spe- converted to grayscale bitmaps corresponding to each RGB cies was studied. All the four species are known to be channel (red/green/blue), and the gray-scale image of the significant isoprenoid emitters, especially the conifers and blue channel was selected and transformed to a black and Q. ilex are strong constitutive emitters (Keenan et al. 2009; white image. This transformation minimized pale areas and Kesselmeier and Staudt 1999). The constitutive emissions shadows as the green color is essentially black in the blue of B. pendula are weaker, but this species can be triggered channel (Niinemets et al. 2004). The projected leaf area for high isoprenoid emissions by stress events (Hakola was estimated using Inkscape v.0.91 software (www. et al. 2001). Leaves of A. balsamea (Re´gimbal and Collin inkscape.org). For needles, the projected leaf area was 1994) and P. mugo (Tsitsimpikou et al. 2001) store iso- estimated by multiplying the length and width of the prenoids in resin ducts, while Q. ilex (Alessio et al. 2008) needles. and B. pendula (Haapanala et al. 2009) leaves lack spe- cialized isoprenoid storage structures and can store these Treatment protocols compounds non-specifically in only small amounts (Loreto et al. 1998; Niinemets and Reichstein 2002). All foliage for given species was pooled, and random sam- The mature coniferous (height 10–12 m for Abies and ples of ca. 3 g were taken for different treatments as 3–4 m for Pinus) and B. pendula trees (height 10–15 m, explained in Fig. 1. Five different treatment types were age ca. 30 years) were located in the campus of the Esto- applied, including various fresh sample processing methods nian University of Life Sciences (58.39°N, 26.70°E, ele- and drying procedures. Different treatment temperatures vation 41 m), while the 7-years-old Q. ilex trees had been were applied for some treatment types, resulting in alto- grown from seed [seed source Biotopo di Marocche, Drena, gether 15 different treatments (Fig. 1). Fresh sample (as is Trentino, Italy (45.97°N, 10.94°E)] in 10 L pots and reg- common for analyses of foliage isoprenoid contents) treat- ularly cut back not to exceed the height of 1 m. The plants ment procedures included flash freezing in liquid N (FF), were grown in a growth chamber at 25 °C day/20 °C night 2 freezing at -82 and -20 °C (F-82, F-20) for 48 h, and temperature, 60 % of relative air humidity, and light cooling at 4 °C (C) for 48 h. Foliage desiccation procedures intensity of 350 lmol m-2 s-1 provided by Philips HPI-T were freeze-drying (FD) and air-drying in forced-convection Plus 400 W metal halide lamps for 12-h photoperiod. drying oven at 30, 40, 60, 70, 80, 100, 110, 150, 200 and Fully mature current-year foliage was harvested in late 300 °C until a constant mass (designated as AD30, AD40, summer. Branches selected were harvested between 8:00 AD60, etc.). AD30 and AD40 lasted usually 72 h, AD60 to and 9:00 morning hours. The branches were re-cut under AD110 about 48 h, and AD200 and AD300 about 8 h. water, the cut end was maintained in water and the bran- FF, F, and C treatments are not procedures for drying the ches were immediately transported to the laboratory. leaf samples, but they simulate common practices to pre- Leaves were rapidly removed from the branch by tweezers. serve leaf properties or stop metabolic reactions when In deciduous species, petioles were excised and only leaf samples cannot be analyzed directly (e.g., during field lamina was used, while in P. mugo the non-photosynthetic work). part of the needle enclosed in the needle sheath was excised and only green needle parts were used. A total of ca. 1500 • In the case of flash freezing in liquid nitrogen (FF), leaf leaves (ca. 950 needles of A. balsamea needles, ca. 380 samples are submerged into liquid nitrogen and leaf needles of P. mugo needles, 70 leaves of B. pendula leaves, water is frozen within seconds, essentially immediately and 100 leaves of Q. ilex) were harvested. stopping any biological activity. Depending on plant tissues, extracellular ice crystal growth may draw water Morphological traits from the cytosol during freezing (Buchner and Neuner 2011; Wisniewski et al. 2003). This may potentially Leaf fresh mass was measured to the nearest 0.0001 g (ALJ lead to loss of some leaf water by sublimation. 220-4NM, Kern & Sohm GmbH, Balingen, Germany). Nevertheless, intra- and extracellular ice formation Leaf thickness for all species and width and length for are expected to occur rapidly in non-frost acclimated needle-leaved species was measured with a digital caliper samples, and thus, the distribution of water between (Mitutoyo CD-15DC, Mitutoyo Ltd, Hampshire, UK) to the different compartments is expected to be maintained nearest 0.01 mm. Five measurements in different parts of under flash freezing. the leaves were taken and averaged to obtain one mean • In the case of freezing (F), leaf temperature is reduced value. In broad-leaved species, major veins were avoided more slowly, and accordingly, the redistribution of during the thickness measurements. water from cytosol to extracellular compartments is The leaves of broad-leaved species were scanned at a more likely to happen than during FF treatment. On the resolution of 300 dpi. The 24-bit (RGB) color images were other hand, deep supercooling and vitrification (glassy 123 1808 Trees (2015) 29:1805–1816

FFF flash freezing in liquid N2 and homogenization FD Freeze-drying in a mortar (-54 °C, 1.2 mbar, 5 days) terpene extraction F Freezing -82 with hexane °C (48 h) -20 (4 °C, 24 h)

C Cooling (4 °C, 48 h) GC-MS analysis AD Air-drying 30, 40, 60, 70, 80, 100, fresh sample 110, 150, 200, 300 °C dry sample

Fig. 1 Scheme of five types of treatments applied to fresh leaves of nitrogen, FD freeze drying under vacuum, F freezing at given four tree species: coniferous Abies balsamea and Pinus mugo, temperature, C cooling and AD air-drying at given temperature, and evergreen broad-leaved Quercus ilex, and deciduous broad-leaved the study tested their effect on estimation of leaf isoprenoid content Betula pendula. The treatments were FF flash freezing in liquid and dry mass (drying treatments)

cellular solutions) reducing ice formation can also was cooled at a desiccator with fresh silica gel and dry occur (Buchner and Neuner 2011; Wisniewski et al. mass estimated. 2003). • Cooling (C) slows down the rate of metabolic processes and results in no ice formation, but certain loss of leaf Leaf isoprenoid content estimation mass due to respiration and loss of water due to transpiration inevitably occurs during cooling. In the After the treatments, foliage isoprenoids were extracted. A case of F and C treatments, the samples were kept in random foliage sample of 70–100 mg (average ± SE of air-tightly sealed polyethylene bags to minimize water 80 ± 14 mg of dry mass across all samples) was taken, put

evaporation. into ceramic mortar together with liquid N2, and homog- • Freeze-drying (FD) is a leaf drying process based on enized with a ceramic pestle into fine powder (Fig. 1). sublimation of ice from frozen samples in low air After trying several ways of sample homogenization and pressure conditions to speed up sublimation. As grinding to fine powder, this was the method which per- metabolic activity is rapidly stopped, and also the key formed a better and faster result. Samples from all treat- plant volatiles have low saturated vapor pressure at ments were homogenized as described. typical freeze-drying temperatures, it is supposed to be The homogenized plant material powder was mixed the most effective drying method preserving both with 1 mL GC-grade hexane (Sigma Aldrich, St. Louis, foliage dry mass and volatiles. Freeze drying at MO, USA) and left at 4 °C for 24 h. The supernatant was -54 °C was conducted under 1.2 mbar air pressure filtered using a PTFE syringe filter (0.45 lm pore size, for 5 days using a freeze dryer ALPHA 1-2 LD (Martin VWR International, Radnor, PA, USA) and immediately Christ Gefriertrocknungsanlagen GmbH, Osterode am injected into a Shimadzu gas chromatograph mass-spec- Harz, Germany). trometer (GC–MS) QP2010 Plus instrument equipped with • Air-drying (AD) is a routine method for estimation of an auto-injector AOC-20i (Shimadzu Corporation, Kyoto, leaf dry mass in plant science (Garnier et al. 2001; Japan) and with a ZB5-MS (0.25 mm i.d. 9 30 m, Zeiller et al. 2007). Typically, dry mass is estimated at 0.25 lm film Zebron, Phenomenex, Torrance, CA, USA) temperatures between 50 and 80 °C, but in some cases column. The MS was operated in electron-impact mode as high as 105 °C is used (Purahong et al. 2014). We (EI) at 70 eV, and the mass scan range was m/z 33–400. applied drying treatments at 30, 40, 60, 70, 80, 100, The GC oven program, transfer line and ion-source tem- 110, 150 and 200 °C in a forced convection ventilated perature were the same as in Noe et al. (2012). drying oven (100–800, Memmert GmbH?Co.KG, Isoprenoid compounds were quantified for A. balsamea Germany). For drying treatment at 300 °C, a muffle and P. mugo by the comparison of their mass spectra with furnace was used (LE 4/11/R6, Nabertherm GmbH, the data from the NIST library (National Institute of Germany). In each case, the drying lasted until a Standards and Technology) and from Adams (2001). For constant mass at given temperature was achieved (e.g., calibration, commercial authentic standards were used typically 72 h at 30 °C and 48 h at 60 °C). The foliage (Sigma-Aldrich, St. Louis, MO, USA) following the

123 Trees (2015) 29:1805–1816 1809 procedures as described in detail in Copolovici et al. (2009) MD MD DF ¼ ¼ : ð5Þ and Niinemets et al. (2013). Isoprenoids detected and MD þ MW MF quantified included monoterpenes, oxygenated monoter- For AD treatments, the ratio of leaf masses estimated penes, sesquiterpenes and oxygenated sesquiterpenes; a after drying at T and T , R is: total of 53 compounds in A. balsamea and 51 compounds 1 2 T1,T2 in P. mugo. Leaf isoprenoid concentration was expressed as MD;T1 þ MRW;T1 RT1;T2 ¼ : ð6Þ mg or lg compound per leaf dry mass as follows: after the MD;T2 þ MRW;T2 isoprenoid extraction, the dry mass of the isoprenoid-free Depending both on the differences in the residual water powder (D ) was estimated by drying in its original vial M,IF content as well as on possible differences in dry mass loss under a forced convection hood for several days at room upon drying. temperature. The mass of isoprenoids, I, was added to D to calculate the total dry mass, D . The volatile M,IF M Theoretical estimation of isoprenoid loss isoprenoid content per dry mass, I , was calculated as: M after treatments I I IM ¼ ¼ : ð1Þ DM DM;IF þ I In the case of fresh sample extracts (i.e., FF, F and C treatments), a sub-sample of foliage was dried at 60 °Cto Theoretical estimation of mass loss estimate leaf dry to fresh mass ratio. As with bulk leaf dry during treatments matter, volatile isoprenoid content per dry mass (IM) depends on residual leaf water content, overall loss of dry The drying treatments (AD treatments) can affect both the mass during the treatment, and particularly strongly on the content of residual water that remains at given temperature T possible loss of isoprenoid content due to evaporation. Thus, IM at given temperature T is given as: (MRW,T), but also the amount of leaf dry matter (MD,T) due to differences in mass loss as the result of respiration and loss I À IL;T of volatiles. Leaf respiration rate increases exponentially IM;T ¼ ; ð7Þ M þ M with temperature up to a temperature of ca. 50 °C and D;T RW;T decreases with further increases in temperature due to ces- where I is the isoprenoid amount in the leaf sample (mg) sation of leaf physiological activity [e.g., Hu¨ve et al. (2011)], and IL,T the possible isoprenoid loss at given temperature but respiration rate also declines with leaf desiccation, (mg): which is expected to occur faster at higher temperature. In I ¼ I À I ðM þ M Þ: ð8Þ contrast, evaporation of volatiles is a purely physical process L;T M;T D;T RW;T and is expected to increase monotonically with increasing In the case of isoprenoid content expressed per unit leaf temperature. Furthermore, at higher temperatures, beyond dry mass (IM, i.e., estimated for the freeze drying 100 °C, pyrolysis of several leaf constituents such as sucrose treatment): and starch may occur, thereby further reducing leaf dry I mass. Thus, leaf dry to fresh mass ratio corresponding to the I ¼ : ð9Þ M M drying temperature T (DF,T) is given as: D

MD;T þ MRW;T Substituting I in Eq. (8) using Eq. (9), DF;T ¼ ; ð2Þ MD þ MW IL;T ¼ IMMD À IM;TðMD;T þ MRW;TÞ: ð10Þ where MW is the mass of water and MD is the water-free Combining Eqs. (10) and (4) leaf mass (dry mass). Moreover, leaf fresh mass, M ,is F I I M I D M : 11 defined as: L;T ¼ M D À M;T F;T F ð Þ

MF ¼ MD þ MW: ð3Þ To express the isoprenoid loss at given temperature per gram of sample (mg g-1), dividing by M and combining Combining Eqs. (2) and (3), F with Eq. (5), DF;TMF ¼ MD;T þ MRW;T: ð4Þ I ¼ I D À I D : ð12Þ Since the FD procedure was supposed to keep better the L;T M F M;T F;T isoprenoid content and therefore give a better estimate of

MD than the AD procedures, the leaf traits measured by Carbon analysis freeze drying the sample were considered the reference for comparison to other analytical procedures. Leaf control dry Total C content of freeze-dried and air-dried leaf samples to fresh mass ratio DF (FD treatment) is defined as: were determined by dry combustion method with a Vario 123 1810 Trees (2015) 29:1805–1816

MAX CNS elemental analyzer (Elementar Analysensys- where the normality test failed, we prioritized the use of teme GmbH, Hanau, Germany). As with the bulk dry mass, parametric tests for further analyses. We accepted statisti- C contents can change due to residual water content as well cal differences between two groups if P \ 0.05 after run- as due to loss of dry mass. In addition, losses of high C ning a two-tailed Student’s t test. For pairwise multiple content volatiles and respiration of sugars and carbohydrate comparisons between several groups, we used a One-way pyrolysis with low carbon content may differently affect ANOVA test with a post hoc analysis by Holm–Sidak test the bulk leaf C content. Carbon content at a drying tem- (P \ 0.05) (e.g., differences in leaf traits between the four perature (g g-1) is defined as: species studied in Table 1, differences in the leaf terpene ÂÃ content between six drying and preserving treatments in MD CM;T ¼ C À CLðMD À MD;TÞ ; ð13Þ Fig. 2a, etc.). MD;T In the case of AD treatments, we did not analyze them where the term MD/MD,T corresponds to the ratio between by testing statistical differences between groups, but we freeze-dried leaf dry mass and the measured dry mass at used a non-linear regression (two-parameter exponential given air-drying temperature, C is the carbon content of decay equation) or a Pearson’s linear regression analysis to -1 freeze-dried sample (g g ), the term MD - MD,T corre- find a correlation based on the drying temperature. sponds to potential biomass loss during drying and CL (g g-1) is the fraction of carbon in the biomass lost, around 0.88 g g-1 in purely monoterpene-based biomass. Results

Statistical analysis Leaf water content

Two types of statistical tests were used to identify signif- Freeze-drying leaf samples as well as air-drying at tem- icant differences between groups of independent replicates. peratures above 60 °C fully removed the leaf water con- We firstly applied a Shapiro–Wilk normality test with a tent. On the other hand, the sample preservation treatments P [ 0.05 significance threshold to the data, which was FF, F-82, F-20 and C, did not significantly change the passed in the majority of cases. Even having some cases initial leaf water content (P = 0.98, P = 0.98, P = 0.97,

Table 1 Leaf morphological and chemical traits (mean ± standard error) of freeze-dried leaves at -54 °C and 1.2 mbar for 5 days and fresh leaves (A) of four tree species Species Abies balsamea Pinus mugo Betula pendula Quercus ilex

Morphological traits Fresh mass (mg)A 23.44 ± 0.43a 58.3 ± 2.0b 217 ± 5c 446 ± 20d Dry mass (mg) 15.67 ± 0.33a 21.0 ± 0.6b 86.3 ± 2.2c 238 ± 13d Water content (%) 46.3 ± 0.9a 50.9 ± 1.6a 60.10 ± 0.31b 49.0 ± 0.5a Dry to fresh mass ratio (%) 53.7 ± 0.9a 49.2 ± 1.6b 39.91 ± 0.29c 51.90 ± 0.49ab Thickness (lm)A 511 ± 7a 752 ± 8b 176.4 ± 1.5c 195.5 ± 2.9d Width (lm)A 1863 ± 16a 1396 ± 15b –– Length (mm)A 30.23 ± 0.23a 63.6 ± 1.5b –– Projected leaf area (mm2)A 56.3 ± 0.6a 88.9 ± 2.3b 1291 ± 29c 2470 ± 100d Leaf dry mass per area (g m-2) 260 ± 6a 313 ± 13b 66.9 ± 0.8c 94.0 ± 2.3d Chemical composition Carbon content (%) 47.347 ± 0.038a 47.34 ± 0.06a 45.8 ± 1.4a 47.18 ± 0.25a Nitrogen content (%) 1.090 ± 0.012a 1.190 ± 0.046a 2.647 ± 0.046b 2.177 ± 0.083c Monoterpene content (mg g-1) 17.9 ± 0.7a 8.920 ± 0.045b \0.01c \0.01c Oxygenated monoterpene content (mg g-1) 1.49 ± 0.05a 0.040 ± 0.021b \0.01c \0.01c Sesquiterpene content (mg g-1) 3.73 ± 0.20a 15.66 ± 0.06b \0.01c \0.01c Oxygenated sesquiterpene content (mg g-1) 0.228 ± 0.038a 4.87 ± 0.07b \0.01c \0.01c Total isoprenoid content (mg g-1) 23.38 ± 0.66a 29.49 ± 0.22b \0.01c \0.01c Superscript letters denote significant differences between species after a Holm–Sidak pairwise comparison test (P \ 0.05). The differences in the isoprenoid content between the two isoprenoid-storing species were tested by a two-tailed Student’s t test (P \ 0.05)

123 Trees (2015) 29:1805–1816 1811

P = 0.83, respectively; two-tailed Student’s t test, all respectively, in the pairwise comparisons with AD70. The species pooled). In all species, AD30-40 reached a ‘‘con- leaf dry to fresh mass ratio did not change significantly stant mass’’ state during the treatment, but the water con- from the FD treatment to the air-dried samples at 60–80 °C tent was not likely fully removed from the leaf samples as in the non-storing species: P = 0.136 in B. pendula and explained theoretically (Eq. 2). After 48 h of drying at P = 0.256 in Q. ilex, reflected also in Fig. 4e, g, respec- 30 °C A. balsamea and P. mugo leaves contained tively, in the pairwise comparisons with AD70. 45.3 ± 3.1 and 48.8 ± 2.4 % of residual water content, Interestingly, air-drying at 40 °C resulted in a decrease and 21.3 ± 1.8 and 3.4 ± 1.1 % after drying at 40 °C, of the dry to fresh mass ratio compared to drying at 30 °C respectively, compared to further drying the leaves at in the non-storing species (P \ 0.05; Holm–Sidak pairwise 60 °C (zero water content). comparisons) (Fig. 4f, h). Leaf dry to fresh mass ratio in relation to drying tem- Leaf isoprenoid content perature (AD treatments) showed a highly significant linear relationship in the isoprenoid-storing species (Fig. 4b, d), Freeze-dried samples preserved more isoprenoid content contrarily than for the non-storing species (Fig. 4f, h). For per leaf dry mass in both isoprenoid-storing species than A. balsamea and P. mugo, the increase of drying temper- any other drying method or fresh sample preservation pro- ature underestimated the dry mass per initial fresh mass. cedure (Fig. 2a, c) (Holm-Sidak pairwise test; P \ 0.05). The average (±SE) maximum isoprenoid contents mea- sured were 23.38 ± 0.66 and 29.49 ± 0.22 mg g-1 in A. Discussion balsamea and P. mugo, respectively. Compared to the FD treatment, the leaf isoprenoid Analysis of leaf isoprenoid content content decreased dramatically in A. balsamea leaves after the FF (36 %), F-20 (37 %), F-82 (73 %), C (33 %) and The analysis of leaf isoprenoid content used needs of a AD70 (82 %) treatments (Fig. 2a). A similar isoprenoid liquid extraction with hexane. To optimize the isoprenoid loss effect occurred for P. mugo leaves in relation to FD extraction to the liquid phase, the leaf sample must be treatment: FF (36 %), F-20 (80 %), F-82 (84 %), C (79 %) ground to a fine powder to increase the sample contact and AD70 (92 %) (Fig. 2c). For both species, the leaf surface area with the extraction liquid. When the sample isoprenoid loss in relation to the increase of drying tem- size is relatively small, a grinder machine might not be perature (AD treatments) was fitted to a two-parameter successful in powdering the sample; therefore, we manu- exponential decay function (Fig. 2b, d). ally ground the samples with a ceramic mortar and a pestle. Air-drying at 30 and 40 °C preserved more isoprenoids In fresh samples, fresh tissues will stick together, fibers than the more usually used temperatures 60 to 80 °Cin would remain flexible and difficult to grind. Therefore, we both isoprenoid-storing species (P \ 0.0001 for both concluded that helping the manual grinding procedure with analyses; Student’s t test). liquid N2 would help to dry and brittle the tissues. In fact, Leaf isoprenoid content of B. pendula and Q. ilex leaves the frozen samples cracked easily into smaller pieces, and was very low, in the range of 1–10 lgg-1. The isoprenoid facilitated the processing into fine powder. But as this compounds were hardly differentiable from the baseline in technique produces shrinkage of tissues and sublimation of the chromatograms, and thus, the leaf isoprenoid content some leaf water content, there is some potential isoprenoid was considered close to zero. loss during the grinding. Nevertheless, we believe that the importance of reaching a fine powder format for a liquid Dry to fresh mass ratio extraction overrides the disadvantage of potential iso- prenoid loss during the process. The log-transformed leaf isoprenoid content data was During the calculation of the leaf isoprenoid concen- correlated positively to the dry to fresh mass ratio data tration, considering the isoprenoid content, I (Eq. 1), as (Fig. 3): the leaves which preserved more leaf isoprenoid part of the dry matter content made the calculation valid for content after the treatment applied accordingly preserved isoprenoid and non-isoprenoid-storing species, where I is more dry mass per initial fresh mass. essentially zero. Furthermore, the calculation of isoprenoid Considering the dry to fresh mass ratio of FD samples as concentration per dry mass was not dependent on the a reference for minimally disturbed samples during drying potential residual water content of leaves undergoing the process, air-drying temperatures of 60 to 80 °C applied to preservation or drying methods AD30 and AD40. This was isoprenoid-storing species significantly underestimated the because whether the leaf sample powder contained moist or value by 10 % in A. balsamea and by 4.5 % in P. mugo not, the isoprenoid content was always referred to the dry

(P \ 0.001; two-tailed t test), reflected also in Fig. 4a, c, mass, DM. 123 1812 Trees (2015) 29:1805–1816

25 2 (a) (b) y = 7.00x - 2.58 (a) a 2 r = 0.70 ; P < 0.001 20

1 )]

15 b -0.022x -1 b b y = 25.420 × e R2 = 0.90 P < 0.001 10 bc ) 0 -1

5 d 2 y = 19.83x - 8.43 (b) 0 r 2 = 0.60 ; P < 0.001 30 a (c) (d) (mg g terpene content [leaf 10 1 log

Leaf terpene content (mgcontent g Leaf terpene 20 b

-0.046x y = 106.206 × e 0 R2 = 0.93 P < 0.001 10 0.35 0.40 0.45 0.50 0.55 0.60 Leaf dry to fresh mass ratio c c cd Fig. 3 Relationship between (log10-transformed) leaf isoprenoid d -1 0 content (mg g ) remaining after different sample treatments and -1

C leaf dry to fresh mass ratio (g g ). All data pooled of independent FF FD 0 100 200 300 leaf samples undergoing different leaf storing and drying procedures F-20 F-82 AD70 (see ‘‘Materials and methods’’ section). The data was fitted by a linear Treatment Temperature (°C) regression

Fig. 2 Leaf total isoprenoid content per dry mass (mg g-1) (aver- age ± SE of three to five independent replicates) after applying The reason could be that freeze drying under vacuum does different sample processing methods to two tree species leaves: Abies not cause shrinkage of the leaf sample, contrarily to FF. balsamea (a, b) and Pinus mugo (c, d). Treatments legends in (a) and We theorize that preserving leaf samples at -82 and (c): FD freeze drying at -54 °C and 1.2 mbar for 5 days, FF flash freezing in liquid nitrogen, F-82 freezing at -82 °C for 48 h, F-20 -20 °C forms ice crystals inside the leaf tissues. These freezing at -20 °C for 48 h, C cooling at 4 °C for 48 h, AD70 air- might have broken cell structures and tissues and conse- drying at 70 °C for 48 h. In b, d x-axis corresponds to air-drying quently release part of the isoprenoids stored. In fact, after temperature until constant weight was reached: normally 72 h at the freezing treatment, when the air-tight plastic bags 30–40 °C, 48 h at 60–110 °C, and\12 h at higher temperatures. The data was fitted by a two-parameter exponential decay function. Letters containing the leaves were opened, the volatile compounds in a, c denote statistical differences between groups after Holm–Sidak smell was evident; a sign that these had been released out pairwise test at P \ 0.05 significance level of the leaf storing organs during the freezing. Storing the leaf samples at 4 °C (C treatment) and air- Leaf isoprenoid content preservation drying near to ambient temperatures (AD30 and AD40) did not stop completely metabolic processes and leaves con- FF, F and C treatments are used to store leaf samples for tinued losing the isoprenoids stored as a natural process of further isoprenoid content determination. We hypothesized leaf decay (Fig. 2). For AD30 and AD40 leaves, because of that these preservation methods would be effective and the favorable temperature for metabolic processes and therefore result in similar isoprenoid content values than long-lasting drying period, there was presumably a loss of the reference method FD. However, the leaf isoprenoid carbon through the respiration of sugars. content remaining after FF, F and C methods was signifi- Leaf drying at 60–80 °C, performed a very low leaf cantly lower compared to FD treated samples (Fig. 2a, c): isoprenoid content preservation compared to FD treatment. The FF method is used to prevent enzymatic reactions We conclude that the best method to store the samples and vaporization losses of isoprenoids (Ormen˜o et al. for further analysis is to freeze-dry these under vacuum, or

2011), although from our experience, FD method per- dipping these in liquid N2 in the absence of a vacuum formed a better leaf isoprenoid conservation (Fig. 2a, c). freeze-drier. 123 Trees (2015) 29:1805–1816 1813

0.55 (a) (b) Leaf dry matter content and leaf isoprenoid content a a a Our coniferous leaves contained 2–3 % of total dry mass in 0.50 ab b form of volatile isoprenoids, which were mostly evaporated b from the leaf after usual leaf drying procedures (i.e., air- 2 0.45 r = 0.976 drying at 60–80 °C). This fact can lead to a systematic P < 0.0001 -3 error in the leaf dry matter determination for leaves con- y = -1.096x ·10 + 0.564 0.40 taining a substantial storage of volatile compounds, as 0.50 theorized previously. Consequently, the underestimation of (c) (d) a the leaf dry matter content further overestimates the results 0.48

) of other leaf traits and chemical analyses when divided by

-1 b b the calculated dry mass. 0.46 In Fig. 4b, d, the dry to fresh mass ratio dropped more c r 2 = 0.987 0.44 c c than 3 %, as it would be expected from isoprenoid content P < 0.0001 evaporation. We hypothesize that as well as stored iso- y = 4.987x ·10-4 + 0.504 0.42 prenoids, other compounds with high volatility such as fats, 0.44 oils, waxes, phenols, alcohols, etc. could also be evapo- (e) (f) rated, decreasing the remaining leaf dry matter after the 0.42 b drying treatment. Nevertheless, we did not detect a ab noticeable decrease in the dry to fresh mass ratio in relation 0.40 ab ab to these non-isoprenoid potential volatiles in B. pendula Leaf dry to fresh mass to (g g ratio Leaf dry ab a (Fig. 4f) and Q. ilex (Fig. 4h). For this reason, more 0.38 r 2 = 0.060; P = 0.524 research is needed to investigate the semi-volatile com- y = 4.769x ·10-5 + 0.391 0.36 pound content in these species and, in general, the potential sources of mass loss when preserving or drying leaf 0.54 (g) (h) samples. Shrinkage of broad leaves of B. pendula and Q. ilex b 0.52 during FF treatment was stronger than for needle-like ab ab ab ab leaves of A. balsamea and P. mugo. Therefore, we assume 0.50 that the significant drop in dry to fresh mass ratio after flash a r 2 = 0.071; P = 0.488 freezing in liquid N2 (Fig. 4e, g) was caused by evapora- y = 5.949x ·10-5 + 0.508 tion of volatile compounds after the rapid shrinkage and 0.48 breakage of tissues. Analogously, we also believe that in

C 0 50 100 150 FF FD broad leaves, the transpiration was higher than in needles at F-82 F-20 AD70 low drying temperatures (30 and 40 °C): the low dry to Treatment Temperature (°C) fresh mass ratio observed after air-drying at 40 °C (Fig. 4f, Fig. 4 Leaf dry to fresh mass ratio (g g-1) (average ± SE of five to h) is probably related to the respiration of sugars during the eight independent replicates) after applying different sample process- long-lasting drying procedure. In addition, the dark respi- ing methods to four tree species leaves: Abies balsamea (a, b), Pinus ration is maximum at around 40 °C (Sharkey and Ber- mugo (c, d), Quercus ilex (e, f), and Betula pendula (g, h). Treatments nacchi 2012). Nevertheless, the variation of dry to fresh legends in the left panels: FD freeze drying at -54 °C and 1.2 mbar for 5 days, FF flash freezing in liquid nitrogen, F-82 freezing at mass ratio in non-storing species across the treatments FF, -82 °C for 48 h, F-20 freezing at -20 °C for 48 h, C cooling at 4 °C FD, F, C and AD30-AD110 was smaller than for the ter- for 48 h, AD70 air-drying at 70 °C for 48 h. In the right panels, x-axis penoid-storing leaves (Fig. 4). corresponds to air-drying temperature until constant weight was There are several evidences to confirm the hypothesis reached: normally 72 h at 30–40 °C, 48 h at 60–110 °C, and\12 h at higher temperatures. The data was fitted by a linear regression. Letters exposed on biased dry mass estimation when using pre- in (a), (c), (e) and (g) denote statistical differences between groups serving or drying methods that release the leaf isoprenoid after Holm–Sidak pairwise test at P \ 0.05 significance level content in isoprenoid-storing leaves. The application of preserving and drying treatments (specially AD) resulted in a dramatic decrease of isoprenoid content (Fig. 2) as well The leaf isoprenoid content in the broad-leaved species as dry mass content (Fig. 4)inA. balsamea and P. mugo, was very low in any case, as found by Alessio et al. (2008) stronger than for the non-storing species. Moreover, we in Q. ilex, and by Haapanala et al. (2009)inB. pendula. found a positive correlation between the isoprenoid content

123 1814 Trees (2015) 29:1805–1816 and the dry to fresh mass ratio (Fig. 3) which suggests that other stable conditions for isoprenoid storage disap- both parameters are related. These relationships are to be pear. The isoprenoid volatilization or transformation taken in consideration by researchers using highly iso- can be done very quickly in some cases, and especially prenoid-storing species because there is a potential bias of if the homogenization is slow and difficult to perform. the dry mass estimation and further variables related to Our experience is that dried samples are easier, faster that. and handy to homogenize. Flash freezing during that process helps it. Recommendations on leaf sampling, handling, and isoprenoid extraction Conclusion Leaf isoprenoid content determination remains a delicate issue. From the moment of leaf sample collection, the The results of this experiment suggested that freeze-drying manipulation of the samples starts releasing volatile com- method (FD) performed the best leaf isoprenoid conser- pounds. Meticulous and gently care should be given to the vation because of its gentleness to the leaf samples during samples to measure the leaf isoprenoid content, as well as the drying process. Flash freezing preserved isoprenoids in other traits including VOCs, like VOCs emission. The a lower level. Air-drying methods performed a dramatic following recommendations optimize the estimation of leaf isoprenoid loss in the leaf samples, as well as storing the isoprenoid content: leaves at freezing and cooling temperatures. 1. Sampling: we suggest that leaves should not be Leaf isoprenoid content and leaf dry to fresh mass ratio sampled directly from the tree, but a branch should were correlated, moreover generating a bias in the dry mass be cut off and transported to the lab. In this way, the estimation after leaf preserving or drying methods due to leaves are kept as better as possible in their original the loss of isoprenoid content. environment, preserving the branch-leaf water contin- Researchers must take in account the potential dry mass uum. This will reduce the leaf water loss or evapo- loss from isoprenoid storage when choosing the leaf drying transpiration, and analogously the potential volatile method. Thus, to gain precision in highly isoprenoid-stor- compound loss. Leaf samples should be handled with ing species, freeze-drying under vacuum can be the best tweezers to prevent warming up the samples with option to estimate the leaf dry mass. empty hands. The isoprenoid extraction must be done ¨ as soon as possible after the sampling. Author contribution statement UN and MPE: research hypothe- ses; MPE and LC: conducted the data collection; All authors con- 2. Transport to the lab: if the plant sampled is not in the tributed to interpretation of results, writing, and edits. lab and requires a short transport (minutes), this can be done by placing the branch in a jar in contact with Acknowledgments This work was supported by the International- water at ambient temperature, or inside a plastic or ization Program DoRa for doctoral studies through the European vinyl bag at cooling conditions if longer time. If the Social Fund, and the Estonian Ministry of Science and Education (institutional grant IUT8-3), the European Commission through the leaves must wait longer for analysis, leaves can be European Regional Fund: the Estonian Center of Excellence in freeze-dried preferably or nitrogen cooled and kept in Environmental Adaptation (project F11100PKTF), and the Environ- paper bags with silica gel. In any case, the researcher mental Conservation and Environmental Technology R&D Pro- must avoid to mechanically perturb the samples. gramme: BioAtmos (project 8-2/T13006PKTF). We thank Dana Copolovici, Milvi Purgas and U¨ lle Pu¨ttsepp for helping with their lab 3. Storing: storing fresh leaf samples before isoprenoid facilities. extraction should be avoided. 4. Drying the leaves: freeze-drying or submerging in Compliance with ethical standards liquid N fresh leaf samples is the best option. Samples 2 Conflict of interest The authors declare that they have no conflict can be kept for several days into paper bags with silica of interest. gel before the isoprenoid extraction. In addition, FD was the method which performed better conservation of the leaf isoprenoid content, followed by FF. These References procedures will also prepare the samples for the homogenization step. Akah PA, Ezike AC, Nwafor SV, Okoh CO, Enwerem NM (2003) 5. Leaf sample homogenization before extraction. This is Evaluation of the anti-asthmatic property of Asystasia gangetica leaf extracts. J Ethnopharmacol 89:25–36. doi:10.1016/S0378- a key step in the leaf isoprenoid content determination. 8741(03)00227-7 It is when the leaf material and volatile molecules get Alessio GA, Pen˜uelas J, De Lillis M, Llusia` J (2008) Implications of exposed to oxidation, cell structures are broken, and foliar terpene content and hydration on leaf flammability of

123 Trees (2015) 29:1805–1816 1815

Quercus ilex and Pinus halepensis. Plant Biol 10:123–128. Maisuthisakul P, Gordon MH, Pongsawatmanit R, Suttajit M (2007) doi:10.1111/j.1438-8677.2007.00011.x Enhancing the oxidative stability of rice crackers by addition of Buchner O, Neuner G (2011) Winter frost resistance of Pinus cembra the ethanolic extract of phytochemicals from Cratoxylum measured in situ at the alpine timberline as affected by formosum Dyer. Asia Pac J Clin Nutr 16:37–42 temperature conditions. Tree Physiol 31:1217–1227. doi:10. Medina E, Francisco M (1994) Photosynthesis and water relations of 1093/treephys/tpr103 savanna tree species differing in leaf phenology. Tree Physiol Castrillo M, Vizcaino D, Moreno E, Latorraca Z (2005) Specific leaf 14:1367–1381 mass, fresh: dry weight ratio, sugar and protein contents in Niinemets U¨ (1997) Energy requirement for foliage construction species of Lamiaceae from different light environments. Rev depends on tree size in young Picea abies trees. Trees Struct Biol Trop 53:23–28 Funct 11:420–431 Copolovici L, Ka¨nnaste A, Niinemets U¨ (2009) Gas chromatography- Niinemets U¨ , Reichstein M (2002) A model analysis of the effects of mass spectrometry method for determination of monoterpene nonspecific monoterpenoid storage in leaf tissues on emission and sesquiterpene emissions from stressed plants. Stud U Babes- kinetics and composition in Mediterranean sclerophyllous Bol Che 54:329–339 Quercus species. Glob Biogeochem Cycle 16:1110. doi:10. Correia MJ, Coelho D, David MM (2001) Response to seasonal 1029/2002GB001927 drought in three cultivars of Ceratonia siliqua: leaf growth and Niinemets U¨ , Cescatti A, Christian R (2004) Constraints on light water relations. Tree Physiol 21:645–653 interception efficiency due to shoot architecture in broad-leaved Davey MC, Ellis-Evans JC (1996) The influence of water content on Nothofagus species. Tree Physiol 24:617–630 the light climate within Antarctic mosses characterized using an Niinemets U¨ , Portsmuth A, Tena D, Tobias M, Matesanz S, optical microprobe. J Bryol 19:235–242 Valladares F (2007) Do we underestimate the importance of Dı´az-Maroto MC, Pe´rez-Coello MS, Vin˜as MAG, Cabezudo MD leaf size in plant economics? Disproportional scaling of support (2003) Influence of drying on the flavor quality of spearmint costs within the spectrum of leaf physiognomy. Ann Bot (Mentha spicata L.) J Agr. Food Chem 51:1265–1269. doi:10. 100:283–303. doi:10.1093/Aob/Mcm107 1021/Jf0208505l Niinemets U¨ et al (2011) Estimations of isoprenoid emission capacity Flexas J, Medrano H (2002) Drought-inhibition of photosynthesis in from enclosure studies: measurements, data processing, quality C-3 plants: stomatal and non-stomatal limitations revisited. Ann and standardized measurement protocols. Biogeosciences Bot 89:183–189. doi:10.1093/Aob/Mcf027 8:2209–2246 Font R, Conesa JA, Molto J, Mun˜oz M (2009) Kinetics of pyrolysis Niinemets U¨ ,Ka¨nnaste A, Copolovici L (2013) Quantitative patterns and combustion of pine needles and cones. J Anal Appl Pyrol between plant volatile emissions induced by biotic stresses and 85:276–286. doi:10.1016/j.jaap.2008.11.015 the degree of damage Front. Plant Sci. doi:10.3389/Fpls.2013. Garnier E, Shipley B, Roumet C, Laurent G (2001) A standardized 00262 protocol for the determination of specific leaf area and leaf dry Noe SM, Hu¨ve K, Niinemets U¨ , Copolovici L (2012) Seasonal matter content. Funct Ecol 15:688–695. doi:10.1046/j.0269- variation in vertical volatile compounds air concentrations 8463.2001.00563.x within a remote hemiboreal mixed forest Atmos. Chem Phys Haapanala S, Ekberg A, Hakola H, Tarvainen V, Rinne J, Hellen H, 12:3909–3926. doi:10.5194/acp-12-3909-2012 Arneth A (2009) Mountain birch—potentially large source of Ormen˜o E, Ferna´ndez C, Mevy JP (2007) Plant coexistence alters sesquiterpenes into high latitude atmosphere. Biogeosciences terpene emission and content of Mediterranean species. Phyto- 6:2709–2718 chemistry 68:840–852. doi:10.1016/j.phytochem.2006.11.033 Hakola H, Laurila T, Lindfors V, Hellen H, Gaman A, Rinne J (2001) Ormen˜o E, Goldstein A, Niinemets U¨ (2011) Extracting and trapping Variation of the VOC emission rates of birch species during the biogenic volatile organic compounds stored in plant species growing season Boreal. Environ Res 6:237–249 Trac-Trend. Anal Chem 30:978–989. doi:10.1016/j.trac.2011.04. Hinckley TM, Bruckerhoff DN (1975) The effects of drought on 006 water relations and stem shrinkage of Quercus alba. Can J Bot Pen˜uelas J, Llusia` J (2004) Plant VOC emissions: making use of the 53:62–72. doi:10.1139/b75-009 unavoidable. Trends Ecol Evol 19:402–404. doi:10.1016/j.tree. Hu¨ve K, Bichele I, Rasulov B, Niinemets U¨ (2011) When it is too hot 2004.06.002 for photosynthesis: heat-induced instability of photosynthesis in Poorter H (1994) Construction costs and payback time of biomass: a relation to respiratory burst, cell permeability changes and H2O2 whole plant perspective. In: Roy J, Garnier E (eds) A whole formation. Plant Cell Environ 34:113–126. doi:10.1111/j.1365- plant perspective on carbon-nitrogen interactions. SPB Aca- 3040.2010.02229.x demic Publishing, The Hague, pp 111–127 Jeni K, Yapa M, Rattanadecho P (2010) Design and analysis of the Popp M, Lied W, Meyer AJ, Richter A, Schiller P, Schwitte H (1996) commercialized drier processing using a combined unsymmet- Sample preservation for determination of organic compounds: rical double-feed microwave and vacuum system (case study: tea microwave versus freeze-drying. J Exp Bot 47:1469–1473. leaves). Chem Eng Process 49:389–395. doi:10.1016/j.cep.2010. doi:10.1093/jxb/47.10.1469 03.003 Purahong W et al (2014) Influence of different forest system Keenan T, Niinemets U¨ , Sabate S, Gracia C, Pen˜uelas J (2009) management practices on leaf litter decomposition rates, nutrient Process based inventory of isoprenoid emissions from European dynamics and the activity of ligninolytic enzymes: a case study forests: model comparisons, current knowledge and uncertainties from Central European Forests. Plos One. doi:10.1371/journal. Atmos. Chem Phys 9:4053–4076. doi:10.5194/acp-9-4053-2009 pone.0093700 Kesselmeier J, Staudt M (1999) Biogenic volatile organic compounds Re´gimbal J-M, Collin G (1994) Essential oil analysis of balsam fir (VOC): an overview on emission, physiology and ecology. Abies balsamea (L.) Mill. J Essent Oil Res 6:229–238. doi:10. J Atmos Chem 33:23–88. doi:10.1023/A:1006127516791 1080/10412905.1994.9698369 Loreto F, Ciccioli P, Brancaleoni E, Cecinato A (1998) Measurement Sharkey TD, Bernacchi CJ (2012) Photosynthetic responses to high of isoprenoid content in leaves of Mediterranean Quercus spp. temperature. In: Flexas J, Loreto F, Medrano H (eds) Terrestrial by a novel and sensitive method and estimation of the isoprenoid photosynthesis in a changing environment: a molecular, phys- partition between liquid and gas phase inside the leaves. Plant iological and ecological approach. Cambridge University Press, Sci 136:25–30. doi:10.1016/S0168-9452(98)00092-2 New York, pp 290–298

123 1816 Trees (2015) 29:1805–1816

Tsitsimpikou C, Petrakis PV, Ortiz A, Harvala C, Roussis V (2001) response to drying in the desiccation-tolerant species Sporobolus Volatile needle terpenoids of six Pinus species. J Essent Oil Res stapfianus and Xerophyta viscosa. J Exp Bot 52:961–969. doi:10. 13:174–178 1093/jexbot/52.358.961 Tuba Z, Csintalan Z, Proctor MCF (1996) Photosynthetic responses of Wisniewski M, Bassett C, Gusta LV (2003) An overview of cold a moss, Tortula ruralis, ssp ruralis, and the lichens Cladonia hardiness in woody plants: seeing the forest through the trees. convoluta and C. furcata to water deficit and short periods of Hortscience 38:952–959 desiccation, and their ecophysiological significance: a baseline Wright IJ et al (2004) The worldwide leaf economics spectrum. study at present-day CO2 concentration. New Phytol Nature 428:821–827. doi:10.1038/Nature02403 133:353–361. doi:10.1111/j.1469-8137.1996.tb01902.x Zeiller E, Benetka E, Koller M, Schorn R (2007) Dry mass Vile D et al (2005) Specific leaf area and dry matter content estimate determination: what role does it play in combined measurement thickness in laminar leaves. Ann Bot 96:1129–1136. doi:10. uncertainty? A case study using IAEA 392 and IAEA 413 algae 1093/Aob/Mci264 reference materials Accredit. Qual Assur 12:295–302. doi:10. Vos J, Groenwold J (1988) Water relations of potato leaves. 1. 1007/s00769-006-0248-z Diurnal changes, gradients in the canopy, and effects of leaf- insertion number, cultivar and drought. Ann Bot 62:363–371 Whittaker A, Bochicchio A, Vazzana C, Lindsey G, Farrant J (2001) Changes in leaf hexokinase activity and metabolite levels in

123 Chemosphere 138 (2015) 751–757

Contents lists available at ScienceDirect

Chemosphere

journal homepage: www.elsevier.com/locate/chemosphere

Temperature dependencies of Henry’s law constants for different plant sesquiterpenes ⇑ Lucian Copolovici a,b, , Ülo Niinemets b,c a Institute of Technical and Natural Sciences Research-Development of ‘‘Aurel Vlaicu” University, Elena Dragoi 2, Arad 310330, Romania b Institute of Agricultural and Environmental Sciences, Estonian University of Life Sciences, Kreutzwaldi 1, Tartu 51014, Estonia c Estonian Academy of Sciences, Kohtu 6, 10130 Tallinn, Estonia highlights

We studied the sesquiterpenes temperature dependences of Henry’s law constants. At 25 °C Henry’s law constants varied 1.4-fold among different sesquiterpenes. We demonstrate moderately variation in Henry’s law constants temperature responses. article info abstract

Article history: Sesquiterpenes are plant-produced hydrocarbons with important ecological functions in plant-to-plant Received 5 May 2015 and plant-to-insect communication, but due to their high reactivity they can also play a significant role Received in revised form 17 July 2015 in atmospheric chemistry. So far, there is little information of gas/liquid phase partition coefficients Accepted 24 July 2015 (Henry’s law constants) and their temperature dependencies for sesquiterpenes, but this information is needed for quantitative simulation of the release of sesquiterpenes from plants and modeling

atmospheric reactions in different phases. In this study, we estimated Henry’s law constants (Hpc) and Keywords: their temperature responses for 12 key plant sesquiterpenes with varying structure (aliphatic, mono-, Flower and leaf volatiles bi- and tricyclic sesquiterpenes). At 25 °C, Henry’s law constants varied 1.4-fold among different Sesquiterpenes Henry’s law constant sesquiterpenes, and the values were within the range previously observed for monocyclic monoterpenes. ° Partition coefficients Hpc of sesquiterpenes exhibited a high rate of increase, on average ca. 1.5-fold with a 10 C increase in Temperature dependence temperature (Q10). The values of Q10 varied 1.2-fold among different sesquiterpenes. Overall, these data

demonstrate moderately high variation in Hpc values and Hpc temperature responses among different sesquiterpenes. We argue that these variations can importantly alter the emission kinetics of sesquiter- penes from plants. Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction volatiles, volatile and semi-volatile terpenoids including sesquiter- penes play a major role in tropospheric chemistry due to their high All plant species emit different volatile organic compounds reactivity with OH radicals and ozone-forming potentials (VOC) either constitutively, or during certain ontogenetic stages (Fehsenfeld et al., 1992). Sesquiterpenes and especially their oxida- or upon exposure to different stresses (Schnitzler et al., 2010; tion products with relatively low vapor pressure of less than 5 Pa at Peñaflor et al., 2011; Fineschi et al., 2013; Niinemets et al., 2013; 25 °C are considered semi-volatiles (Hoskovec et al., 2005; Kosina Niinemets and Monson, 2013). Plants synthesize more than et al., 2013) and have a high capacity to form and condense on solid 100,000 chemical products and at least 1700 of these are known particles, thus, playing an important role in secondary aerosol for- to be volatile (Kesselmeier and Staudt, 1999). Around 90% of global mation and in formation of cloud condensation nuclei (Huff Hartz terrestrial non-methane VOC emissions are of vegetation origin et al., 2005; Helmig et al., 2006). (Guenther et al., 1995, 2012; Arneth et al., 2008). Among biogenic Plants under different types of stresses emit a great variety of sesquiterpenes (Beauchamp et al., 2005; Copolovici et al., 2011, 2012). The composition of the emissions depend on the stress type ⇑ Corresponding author at: Institute of Agricultural and Environmental Sciences, and on the stress severity (Niinemets et al., 2013). The induced Estonian University of Life Sciences, Kreutzwaldi 1, Tartu 51014, Estonia. emissions of caryophyllene, humulene, valencene, etc. are usually E-mail address: [email protected] (L. Copolovici). http://dx.doi.org/10.1016/j.chemosphere.2015.07.075 0045-6535/Ó 2015 Elsevier Ltd. All rights reserved. 752 L. Copolovici, Ü. Niinemets / Chemosphere 138 (2015) 751–757

(but not only) due to abiotic stresses such as heat, frost, drought solubility of taxol (polyoxygenated diterpene) have been deter- and flooding (Copolovici et al., 2012). Biotic stresses (herbivory, mined by Fichan et al. (1999) and Kapoor and Mahindroo (1997) fungal attacks, etc) result in emissions of farnesene, and nerolidol but no data are available for sesquiterpenes (C15) with the excep- derivative homoterpenes (Joo et al., 2011). However, it is not easy tion of farnesene (Schuhfried et al., 2015). Furthermore, there is a to generalize because the induced sesquiterpene emissions are general lack of temperature dependencies of higher molecular species-specific and the blend of emission could be different for mass volatiles. different stresses. Apart from stress-elicited sesquiterpene release, In the current study, we determined the temperature depen- sesquiterpenes are constitutively emitted by a huge variety of dencies of Henry’s law constant (air–water equilibrium partition 3 1 flowers. As floral scent is the key long-distance attractant to polli- constant; Hpc,Pam mol ) for 12 structurally different sesquiter- nators that are often plant species specific, the composition of the penes (b-caryophyllene, a-cedrene, a-farnesene, c-gurjonene, scent is a characteristic trait of every pant species (Pichersky et al., a-humulene, isosativene, a-longipinene, nerolidol, a-neoclovene, 1994; Dudareva et al., 1998), and plays a major role in plant repro- b-neoclovene, c-neoclovene, and valencene) and the monoterpene ductive success (Pichersky and Gershenzon, 2002). For example, derivative bornyl acetate (C12) that is frequently emitted by many flowers of Asarum species emit a-cedrene, a-humulene, plant species as well. b-caryophyllene, (E)-nerolidol and many other oxygenated and 1 1 non-oxygenated sesquiterpenes at a high level of 1000 ng g h 2. Methods (Azuma et al., 2010; Farré-Armengol et al., 2014), while the volatile profile of Alstroemeria ‘‘Sweet Laura” flowers is dominated by Terpenoid standards were obtained from Sigma–Aldrich (E)-b-caryophyllene, humulene and an ocimene-like compound (Munich, Germany) and were all of high purity (P98%). For the (Aros et al. (2012). Arabidopsis thaliana flowers release a complex determination of Henry’s law constant we used EPICS (equilibrium mixture of volatile monoterpenes and sesquiterpenes with partitioning in closed systems) method of Gossett (1987) as (E)-b-caryophyllene as the dominant component (Chen et al., applied previously (Copolovici and Niinemets, 2005, 2007). The 2003) whereas the formation of all sesquiterpenes found in the EPICS method is based on analysis of compound partitioning in dif- Arabidopsis floral volatile blend is due to only two terpene syn- ferent liquid volumes of two binary solute–solvent equilibrium thases specifically expressed in flowers (Tholl et al., 2005). Due systems. The liquid- or gas-phase concentrations in the two sys- to high emission rates during the flowering period, atmospheric tems are measured, and the ratio of these values is used to com- sesquiterpene concentrations strongly increase during the bloom- pute the dimensionless Henry’s law constant (Hcc) using the ing period (Baghi et al., 2012). This evidence collectively indicates combined mass balance equations of these two systems. Shortly, that both stress-elicited emissions and constitutive flower emis- two identical glass vials (total volume, VT, of 64.5 mL) were used. sions constitute an important source of atmospheric BVOC. 4mL(VL1) distilled water was added to the first vial and 50 mL As different sesquiterpenes have widely different atmospheric (VL2) to the second vial. Thereafter, 0.2 lL of given sesquiterpene rate constants for reactions with ozone and OH radicals (Arneth was added to both vials using a 1 lL micro-syringe (SGE and Niinemets, 2010; Holopainen and Blande, 2013; Holopainen International, Ltd., Victoria, Australia). After the injection, the vial et al., 2013), predicting atmospheric reactions of sesquiterpenes was immediately sealed with a Teflon-coated silicone septum, vig- requires understanding of dynamics of fractional composition of orously shaken and ultrasound until the organic phases has been emitted sesquiterpenes. Moreover, short atmospheric lifetimes completely dissolved. The vials were equilibrated in a thermostatic (two minutes for b-caryophyllene and a-humulene) (Richters water bath at desired temperature for 3 h. According to our work in et al., 2015) due to reactions with ozone implicate the sesquiterpe- the previous study (Copolovici and Niinemets, 2005) the phase nes in secondary organic aerosol formation (van Eijck et al., 2013). equilibrium was reached in ca. 2 h after the injection. The head- From the perspective of plant-insect interactions in reactive atmo- space was further sampled for GC analyses using a multibed steel spheres, this information is also important for predicting the cartridges filled with different Carbotrap fractions as described changes in composition of sesquiterpene blend with distance from previously (Niinemets et al., 2010; Kännaste et al., 2014). The car- the emission source (Holopainen et al., 2013; Blande et al., 2014). tridges were analyzed with a Shimadzu 2010 Plus Gas The emission models can predict the fractional composition of Chromatograph Mass Spectrometer with a TD20 thermodesorption complex compound mixtures once the physico-chemical charac- system. The method and chromatographic programs have been teristics of specific compounds are available (Niinemets and described in detail in (Copolovici et al., 2009; Toome et al., 2010). Reichstein, 2002; Niinemets et al., 2002). In particular, due to the Given the presence of phase-equilibria in both vials, the Henry’s controls of compound-physico-chemical characteristics on gas- law constant was calculated according to Gossett (1987). Under phase transfer resistance between the leaves and the atmosphere, equilibrium conditions, the same amount of sesquiterpene injected the kinetics of emission can importantly depend on the Henry’s into both vials 1 and 2 is partitioned among the gas and liquid law constant (air–water equilibrium partition constant; Hpc, phases as: Pa m3 mol1)(Niinemets and Reichstein, 2002). Furthermore, tem- þ ¼ þ ð Þ perature of leaves and flowers strongly varies through the day, and CL1V L1 CG1V G1 CL2V L2 CG2V G2 1 thus, there is a huge diurnal variation in the emission of sesquiter- penes due to both temperature-dependent changes in the synthe- where CL1 is the equilibrium sesquiterpene liquid-phase concentra- sis rate and due to temperature-dependent changes in compound tion in vial 1 and CL2 the liquid-phase concentration in vial 2, CG1 is physico-chemical characteristics (Niinemets and Reichstein, 2003). the equilibrium sesquiterpene gas-phase concentration in vial 1 and The temperature dependencies of equilibrium coefficients are CG2 the concentration in vial 2, VG1 is the equilibrium gas-phase already available for more than 5000 different compounds (headspace) volume in vial 1 and VG2 the gas-phase volume in vial (Staudinger and Roberts, 1996, 2001; Sander, 2015), but few data 2. We assume that liquid-phase volumes remain constant during are available for relevant plant volatiles. Recently, Henry’s law con- equilibration in both vials. From Eq. (2), the dimensionless CG1 CG2 Henry’s law constant (Hcc ¼ ¼ ) is given as: stants for 10 monoterpenes (C10) (Copolovici and Niinemets, 2005, CL1 CL2 2007) and benzenoid methyl salicylate (C8) and, methyl jasmonate CG1 V L2 V L1 (C13) (Karl et al., 2008) have been determined, but there are very CG2 Hcc ¼ ð2Þ CG1 few data on higher molecular mass plant volatiles compounds. ðV T V L1ÞðV T V L2Þ CG2 As a few exceptions, the solubility of C13 compound b-ionone and L. Copolovici, Ü. Niinemets / Chemosphere 138 (2015) 751–757 753

D where VT = VL1 + VG1 = VL2 + VG2 is the total vial volume. The HVol ln Hpc ¼ þ C; ð5Þ compound concentration in the headspace, CG, is proportional to RT the signal peak area (A for the first vial and A for the second vial) 1 2 where DH is the enthalpy of volatilization and C is a compound- from the gas chromatogram: vol specific constant. Thus, DHvol was derived for each compound as C A the slope of lnH vs. 1/T relationship. In addition, Q10 values G1 ¼ 1 ð3Þ pc CG2 A2 (van’t Hoff equation) were also calculated.

Combining Eqs. (2) and (3), Hcc is ultimately given as: 3. Results A1 V L2 V L1 A2 Hcc ¼ ð4Þ A1 For all studied sesquiterpenes, Henry’s law constants were of ðV T V L1ÞðV T V L2Þ A2 the same order of magnitude varying from 2548 to The data were finally converted to units Pa m3 mol1, i.e. pre- 3505 Pa m3 mol1 at 25 °C(Table 1) with an average ± SE (standard sented as Hpc values (gas-phase vapor pressure to liquid-phase error of the mean) value for all sesquiterpenes of 3 1 concentration ratio). Hpc = HccRT, where R is the gas constant 3135 ± 84 Pa m mol . The Hpc value for the monoterpene deriva- 3 1 1 (Pa m mol K ) and T the absolute temperature (K). The deter- tive bornyl acetate was within the range of Hpc values observed for minations of Hpc were conducted at temperatures 25, 32, 38, 45, sesquiterpenes (Table 1). 50 °C. At each temperature, average Henry’s law constant was cal- Simplified temperature response (Eq. (5)) fitted well the culated for three to four independent determinations. temperature relation of Hpc over the entire temperature range of 2 Over a relatively small temperature range, Hpc temperature 25–50 °C for all compounds studied (r > 0.98, on average response can be fitted as (Dewulf et al., 1999; Görgényi et al., 2002): 0.989 ± 0.005; Figs. 1 and 2). The enthalpy of volatilization

Table 1

Henry’s law constants (Hpc)at25°C, corresponding enthalpies of volatilization and Q10 values for 12 studied sesquiterpene species (C15) and for monoterpene derivative bornyl acetate (C12).

3 1 1 Compound Chemical structure Hpc (Pa m mol ) DHvol (kJ mol ) Q10 b-Caryophyllene 2731 ± 21a 37.08 ± 0.26 1.58

a-Cedrene 3510 ± 70b 40.5 ± 0.8 1.64

a-Farnesene 2960 ± 50ac 35.7 ± 0.5 1.54

a-Humulene 3410 ± 41b 38.7 ± 1.1 1.60

c-Gurjunene 3052 ± 13c 30.8 ± 0.5 1.45

(continued on next page) 754 L. Copolovici, Ü. Niinemets / Chemosphere 138 (2015) 751–757

Table 1 (continued)

3 1 1 Compound Chemical structure Hpc (Pa m mol ) DHvol (kJ mol ) Q10 Isosativene 2548 ± 22a 29.7 ± 0.6 1.42

a-Longipinene 2960 ± 50a 34.01 ± 0.43 1.52

a-Neoclovene 3410 ± 41b 25.7 ± 0.8 1.36

b-Neoclovene 3415 ± 6b 38.5 ± 0.9 1.58

c-Neoclovene 3421 ± 16b 26.3 ± 0.8 1.36

Nerolidol HO 3120 ± 80d 35.8 ± 1.2 1.53

Valencene 3090 ± 140d 37.2 ± 0.9 1.56

Bornyl acetate 2650 ± 14a 14.54 ± 0.38 1.19

OOCH3

Data are averages ± SE (standard error of the mean) of three independent determinations. The atomic formulas of all sesquiterpenes are C15H24, except for nerolidol C15H26O.

The atomic formula for bornyl acetate is C12H20O2. L. Copolovici, Ü. Niinemets / Chemosphere 138 (2015) 751–757 755

1 9.5 (DHvol) varied ca. 1.6-fold from 25.7 to 40.5 kJ mol among β-caryophyllene sesquiterpenes with an average (±SE) of 34.2 ± 1.4 kJ mol1, corre-

) α-humulene sponding to an average Q10 value of 1.512 ± 0.027 (Table 1). Both -1 α-farnesene 9.0 DHvol and Q10 were lower for oxygenated monoterpene derivative mol

3 bornyl acetate (Table 1). The enthalpy of volatilization varied significantly among the 8.5 neoclovene isomers which differ in the location of the double bond , Pa m in the molecule. The enthalpy of volatilization was 25.7 kJ mol1 Pc 1 H for a-neoclovene and 26.3 kJ mol for c-neoclovene that have 8.0

ln ( the double bond located within the hydrogenated indene scaffold- D 1 b (a) ing, while Hvol was 38.5 kJ mol for -neoclovene that has the double bond in the methylene group attached to the hydrogenated 7.5 indole scaffolding (Table 1). The enthalpy of volatilization was not 9.5 significantly different (p = 0.94) between two aliphatic sesquiter- -cedrene α pene a-farnesene and nerolidol, which is surprising given that ) valencene -1 nerolidol has an OH group in its structure (Table 1). 9.0 nerolidol

mol bornyl acetate 3 4. Discussion 8.5 °

, Pa m 4.1. Henry’s law constants of sesquiterpenes at 25 C Pc H 8.0 Due to greater molecular size of sesquiterpenes compared with ln ( ln monoterpenes, the saturated vapor pressure (Pv) and, aqueous sol- (b) ubility (S) are expected to be much smaller for sesquiterpenes than 7.5 for monoterpenes. In fact, saturated vapor pressure of non- 0.0031 0.0032 0.0033 0.0034 oxygenated sesquiterpenes is more than two orders of magnitude 1/Temperature (K-1) less than that in non-oxygenated monoterpenes (Hoskovec et al., 2005; Kosina et al., 2013). Compared with monoterpenes, the val- 3 1 Fig. 1. Temperature dependencies of Henry’s law constants (Hpc,Pam mol )of ues of Hpc for sesquiterpenes are somewhat lower than those different sesquiterpenes and bornyl acetate. Data were fitted by Eq. (5) to derive the observed for bicyclic monoterpenes a-pinene and b-pinene and enthalpy of vaporization (Table 1). All determinations were replicated three times, monocyclic monoterpenes a-and b-phellandrene, but they are of and the error bars provide ± SE. the same order of magnitude than Hpc values in other studied monocyclic monoterpenes (Copolovici and Niinemets, 2005), and in the monoterpene derivative bornyl acetate studied here

(Table 1). Given that Hpc = Pv/S, this evidence suggests that lower 9.5 P of sesquiterpenes is associated with analogous reductions in S α-longipinene v compared with monoterpenes.

) γ-gurjonene -1 In the recent work Schuhfried et al. (2015) determined the H 9.0 isosativene pc for farnesene using another method based on on-line PTR-MS mea- mol 3 surements in non-equilibrium conditions. There are no data on 8.5 aqueous solubility of sesquiterpenes, but Fichan et al. (1999) demonstrated that the C13 oxygenated compound b-ionone had , Pa m

Pc a much lower aqueous solubility and vapor pressure at 25 °C than H 8.0 the C10 oxygenated monoterpenes fenchol, linalool, borneol, and ln ( a-pinene and limonene oxides. (a) 7.5 4.2. Temperature responses of Henry’s law constants for 9.5 sesquiterpenes α-neoclovene β-neoclovene We determined the temperature dependencies of Hpc for impor-

) γ-neoclovene

-1 tant plant volatile sesquiterpenes which are elicited in response to 9.0 different abiotic and biotic stresses. These estimations are, to our mol

3 knowledge, the first determinations of temperature characteristics of equilibrium coefficients for plant volatile sesquiterpenes. The average (±SE) enthalpy of volatilization determined for 204

, Pa m 8.5

Pc different compounds by Staudinger and Roberts (2001) was 1 H 39.1 ± 2.3 kJ mol , while our average value for 12 sesquiterpenes 1 ln ( of 34.2 ± 1.2 kJ mol is in the same range. On the other hand, (b) previous estimations of monoterpene volatilization enthalpies 8.0 1 0.0031 0.0032 0.0033 0.0034 have yielded an average value of 36.5 ± 0.4 kJ mol for non- oxygenated monoterpenes and 17.8 ± 0.4 kJ mol1 for oxygenated -1 1/Temperature (K ) monoterpenes (Copolovici and Niinemets, 2005). Thus, the volatility of sesquiterpenes is expected to increase more with Fig. 2. Temperature dependencies of Henry’s law constants (H ,Pam3 mol1)of pc temperature than the volatility of oxygenated monoterpenes. In key plant sesquiterpenes. Data were fitted by Eq. (5) to derive the enthalpy of vaporization (Table 1). All determinations were replicated three times, and the error contrast, the temperature responses of Hpc for non-oxygenated bars provide ± SE. monoterpenes and sesquiterpenes are expected to be similar. 756 L. Copolovici, Ü. Niinemets / Chemosphere 138 (2015) 751–757

Although the values of the volatilization enthalpy were similar Aros, D., Gonzalez, V., Allemann, R.K., Mueller, C.T., Rosati, C., Rogers, H.J., 2012. among most compounds, still a significant variation in tempera- Volatile emissions of scented Alstroemeria genotypes are dominated by terpenes, and a myrcene synthase gene is highly expressed in scented ture responses, ca. 1.2-fold (and 1.6-fold when also considering Alstroemeria flowers. J. Exp. Bot. 63, 2739–2752. bornyl acetate) was observed across the compounds (Table 1). Azuma, H., Nagasawa, J.-I., Setoguchi, H., 2010. Floral scent emissions from Asarum yaeyamense and related species. Biochem. Syst. Ecol. 38, 548–553. This variability is currently difficult to explain as similar Hpc values Baghi, R., Helmig, D., Guenther, A., Duhl, T., Daly, R., 2012. Contribution of flowering were observed for aliphatic compounds. Even for isomers differing trees to urban atmospheric biogenic volatile organic compound emissions. in the location of the double bonds such as neoclovene isomers Biogeosciences 9, 3777–3785. (Fig. 2), important variability was observed (Table 1). We suggest Beauchamp, J., Wisthaler, A., Hansel, A., Kleist, E., Miebach, M., Niinemets, Ü., Schurr, U., Wildt, J., 2005. Ozone induced emissions of biogenic VOC from tobacco: that future studies with more sesquiterpenes and with experimen- relations between ozone uptake and emission of LOX products. Plant, Cell tal characterization of additional compound physico-chemical Environ. 28, 1334–1343. characteristics such as saturated vapor pressure, aqueous solubility Blande, J.D., Holopainen, J.K., Niinemets, Ü., 2014. Plant volatiles in polluted atmospheres: stress responses and signal degradation. Plant, Cell Environ. 37, and molar volume are needed to gain an insight into variation in 1892–1904. Hpc at a given temperature and into Hpc temperature responses Chen, F., Tholl, D., D’Auria, J.C., Farooq, A., Pichersky, E., Gershenzon, J., 2003. among different sesquiterpenes. Biosynthesis and emission of terpenoid volatiles from Arabidopsis flowers. The overall increase in the rate of enzymatic reactions and inte- Plant Cell 15, 481–494. Copolovici, L., Kaennaste, A., Pazouki, L., Niinemets, U., 2012. Emissions of green leaf grated biological processes for a 10 °C rise in temperature (Q10, volatiles and terpenoids from Solarium lycopersicum are quantitatively related Table 1) is in the range of 1.5–3 (Nobel, 1991). The value of Q10 to the severity of cold and heat shock treatments. J. Plant Physiol. 169, 664–672. for physical processes such as diffusion is typically low, on the Copolovici, L., Kännaste, A., Niinemets, U., 2009. Gas chromatography-mass spectrometry method for determination of monoterpene and sesquiterpene order of 1.05–1.1 (Gates, 1980; Niinemets and Reichstein, 2003). emissions from stressed plants. Studia Univ. Babes-Bolyai Chemia 54, 329–339. However, the volatility increase across the sesquiterpenes studied, Copolovici, L., Kännaste, A., Remmel, T., Vislap, V., Niinemets, Ü., 2011. Volatile ca. 1.5-fold for a 10 °C rise in temperature, is very large for a non- emissions from Alnus glutinosa induced by herbivory are quantitatively related to the extent of damage. J. Chem. Ecol. 37, 18–28. biological process. From a biological perspective, there are two Copolovici, L., Niinemets, U., 2007. Salting-in and salting-out effects of ionic and important implications of this large Q10 value for sesquiterpenes. neutral osmotica on limonene and linalool Henry’s law constants and octanol/ First, it is well-known that leaves release sesquiterpenes in the water partition coefficients. Chemosphere 69, 621–629. Copolovici, L.O., Niinemets, Ü., 2005. Temperature dependencies of Henry’s law stress conditions (Niinemets, 2010). Due to their semivolatile char- constants and octanol/water partition coefficients for key plant volatile acter, their emission level is typically quite low. Due to their high monoterpenoids. Chemosphere 61, 1390–1400. enthalpy of volatilization, emission rates can be strongly enhanced Dewulf, J., Van Langenhove, H., Graré, S., 1999. Sediment/water and octanol/ater equilibrium partitioning of volatile organic compounds: temperature with rising temperatures. This needs to be considered when dependence in the 2–25 C range. Water Res. 33, 2424–2436. predicting plant volatile emission responses when stress occurs Dudareva, N., Cseke, L., Blanc, V.M., Pichersky, E., 1996. Evolution of floral scent in at different temperatures. The second aspect is the involvement Clarkia: novel patterns of s-linalool synthase gene expression in the C. breweri of sesquiterpenes in pollinator attraction (e.g., Tholl et al., 2005; flower. Plant Cell 8, 1137–1148. Dudareva, N., Raguso, R.A., Wang, J.H., Ross, E.J., Pichersky, E., 1998. Floral scent Nieuwenhuizen et al., 2009). Increases in temperature, e.g. even production in Clarkia breweri – III. Enzymatic synthesis and emission of long-term increases such as due to global warming, can benzenoid esters. Plant Physiol. 116, 599–604. importantly enhance the flower attraction for pollinators due to Farré-Armengol, G., Filella, I., Llusià, J., Niinemets, Ü., Peñuelas, J., 2014. Changes in floral bouquets from compound-specific responses to increasing temperatures. facilitated sesquiterpene release (Valterova et al., 2007; Glob. Change Biol. 20, 3660–3669. Farré-Armengol et al., 2014). However, flower volatiles are often Fehsenfeld, F., Calvert, J., Fall, R., Goldan, P., Guenther, A.B., Hewitt, C.N., Lamb, B., characterized by a high fraction of oxygenated monoterpenes and Liu, S., Trainer, M., Westberg, H., Zimmerman, P., 1992. Emissions of volatile organic compounds from vegetation and the implications for atmospheric phenolic compounds known to be perceived by pollinator insects chemistry. Glob. Biogeochem. Cycle 6, 389–430. (Dudareva et al., 1996, 1998; Nagegowda et al., 2008). Because Fichan, I., Larroche, C., Gros, J.B., 1999. Water solubility, vapor pressure, and activity these compounds have lower Q10 values, flower detection by pol- coefficients of terpenes and terpenoids. J. Chem. Eng. Data 44, 56–62. Fineschi, S., Loreto, F., Staudt, M., Peñuelas, J., 2013. Diversification of volatile linators attracted to these oxygenated volatiles might be hindered isoprenoid emissions from trees: evolutionary and ecological perspectives. In: due to the increased odor background of non-oxygenated mono- Niinemets, Ü., Monson, R.K. (Eds.), Biology, Controls and Models of Tree Volatile and sesquiterpenes. We also emphasize that although the role of Organic Compound Emissions. Springer, Berlin, pp. 1–20. Gates, D., 1980. Biophysical Ecology. Springer-Verlag, New York. sesquiterpenes is generally neglected in air quality estimations Görgényi, M., Dewulf, J., Van Langenhove, H., 2002. Temperature dependence of due to low rates of emissions in cool conditions, these emissions Henry’s law constant in an extended temperature range. Chemosphere 48, 757– are expected to become gradually more important in warmer 762. atmospheres, especially given their high potential to produce Gossett, J.M., 1987. Measurement of Henrys law constants for C1 and C2 chlorinated hydrocarbons. Environ. Sci. Technol. 21, 202–208. ozone and secondary organic aerosols (Huff Hartz et al., 2005; Guenther, A., Hewitt, C.N., Erickson, D., Fall, R., Geron, C., Graedel, T., Harley, P., Helmig et al., 2006). Klinger, L., Lerdau, M., McKay, W.A., Pierce, T., Scholes, B., Steinbrecher, R., Tallamraju, R., Taylor, J., Zimmerman, P., 1995. A global model of natural volatile compound emissions. J. Geophys. Res. 100, 8873–8892. Acknowledgements Guenther, A.B., Jiang, X., Heald, C.L., Sakulyanontvittaya, T., Duhl, T., Emmons, L.K., Wang, X., 2012. The model of emissions of gases and aerosols from nature Funding for this study has been provided by the Estonian version 2.1 (MEGAN2.1): an extended and updated framework for modeling biogenic emissions. Geosci. Model. Dev. 5, 1471–1492. Ministry of Science and Education (institutional grant IUT-8-3), Helmig, D., Ortega, J., Guenther, A., Herrick, J.D., Geron, C., 2006. Sesquiterpene the European Commission through the European Regional Fund emissions from loblolly pine and their potential contribution to biogenic (the Center of Excellence in Environmental Adaptation), the aerosol formation in the Southeastern US. Atmos. Environ. 40, 4150–4157. Holopainen, J.K., Blande, J.D., 2013. Where do herbivore-induced plant volatiles go? European Research Council (advanced grant 322603, SIP-VOL+), Front. Plant Sci. 4, 185. and the European Commission and the Romanian Government Holopainen, J.K., Nerg, A.-M., Blande, J.D., 2013. Multitrophic signalling in polluted (project POSCCE 621/2014). atmospheres. In: Niinemets, Ü., Monson, R.K. (Eds.), Biology, Controls and Models of Tree Volatile Organic Compound Emissions. Springer, Berlin, pp. 285– 314. References Hoskovec, M., Grygarová, D., Cvacˇka, J., Streinz, L., Zima, J., Verevkin, S.P., Koutek, B., 2005. Determining the vapour pressures of plant volatiles from gas chromatographic retention data. J. Chromatogr. A 1083, 161–172. Arneth, A., Monson, R.K., Schurgers, G., Niinemets, Ü., Palmer, P.I., 2008. Why are Huff Hartz, K.E., Rosenørn, T., Ferchak, S.R., Raymond, T.M., Bilde, M., Donahue, N.M., estimates of global isoprene emissions so similar (and why is this not so for Pandis, S.N., 2005. Cloud condensation nuclei activation of monoterpene and monoterpenes)? Atmos. Chem. Phys. 8, 4605–4620. sesquiterpene secondary organic aerosol. J. Geophys. Res. Atm. 110, D14208. Arneth, A., Niinemets, Ü., 2010. Induced BVOCs: how to bug our models? Trends http://dx.doi.org/10.11029/12004JD005754. Plant Sci. 15, 118–125. L. Copolovici, Ü. Niinemets / Chemosphere 138 (2015) 751–757 757

Joo, E., Dewulf, J., Amelynck, C., Schoon, N., Pokorska, O., Simpraga, M., Steppe, K., Niinemets, Ü., Reichstein, M., Staudt, M., Seufert, G., Tenhunen, J.D., 2002. Stomatal Aubinet, M., Van Langenhove, H., 2011. Constitutive versus heat and biotic constraints may affect emission of oxygenated monoterpenoids from the foliage stress induced BVOC emissions in Pseudotsuga menziesii. Atmos. Environ. 45, of Pinus pinea. Plant Physiol. 130, 1371–1385. 3655–3662. Nobel, P.S., 1991. Physicochemical and Environmental Plant Physiology. Academic Kännaste, A., Copolovici, L., Niinemets, Ü., 2014. Gas chromatography mass- Press, Inc., San Diego – New York – Boston – London – Sydney – Tokyo – spectrometry method for determination of biogenic volatile organic Toronto. compounds emitted by plants. In: Rodríguez-Concepción, M. (Ed.), Plant Peñaflor, M.F.G.V., Erb, M., Miranda, L.A., Werneburg, A.G., Bento, J.M.S., 2011. Isoprenoids: Methods and Protocols, 1153. Humana Press, New York, pp. Herbivore-induced plant volatiles can serve as host location cues for a 161–169. generalist and a specialist egg parasitoid. J. Chem. Ecol. 37, 1304–1313. Kapoor, V.K., Mahindroo, N., 1997. Recent advances in structure modifications of Pichersky, E., Gershenzon, J., 2002. The formation and function of plant volatiles: Taxol (Paclitaxel). Ind. J. Chem. Sect. B 36, 639–652. perfumes for pollinator attraction and defense. Curr. Opin. Plant Biol. 5, 237– Karl, T., Guenther, A., Turnipseed, A., Patton, E.G., Jardine, K., 2008. Chemical sensing 243. of plant stress at the ecosystem scale. Biogeosciences 5, 1287–1294. Pichersky, E., Raguso, R.A., Lewinsohn, E., Croteau, R., 1994. Floral scent production Kesselmeier, J., Staudt, M., 1999. Biogenic volatile organic compounds (VOC): an in Clarkia (Onagraceae).1. Localization and developmental modulation of overview on emission, physiology and ecology. J. Atmos. Chem. 33, 23–88. monoterpene emission and linalool synthase activity. Plant Physiol. 106, Kosina, J., Dewulf, J., Viden, I., Pokorska, O., Van Langenhove, H., 2013. Dynamic 1533–1540. capillary diffusion system for monoterpene and sesquiterpene calibration: Richters, S., Herrmann, H., Berndt, T., 2015. Gas-phase rate coefficients of the quantitative measurement and determination of physical properties. Int. J. reaction of ozone with four sesquiterpenes at 295 ± 2 K. Phys. Chem. Chem. Environ. Anal. Chem. 93, 637–649. Phys. 17, 11658–11669. Nagegowda, D.A., Gutensohn, M., Wilkerson, C.G., Dudareva, N., 2008. Two nearly Sander, R., 2015. Compilation of Henry’s law constants (version 4.0) for water as identical terpene synthases catalyze the formation of nerolidol and linalool in solvent. Atmos. Chem. Phys. 15, 4399–4981. snapdragon flowers. Plant J. 55, 224–239. Schnitzler, J.-P., Louis, S., Behnke, K., Loivamäki, M., 2010. Poplar volatiles – Nieuwenhuizen, N.J., Wang, M.Y., Matich, A.J., Green, S.A., Chen, X., Yauk, Y.-K., biosynthesis, regulation and (eco)physiology of isoprene and stress-induced Beuning, L.L., Nagegowda, D.A., Dudareva, N., Atkinson, R.G., 2009. Two terpene isoprenoids. Plant Biol. 12, 302–316. synthases are responsible for the major sesquiterpenes emitted from the Schuhfried, E., Aprea, E., Märk, T.D., Biasioli, F., 2015. Refined measurements of flowers of kiwifruit (Actinidia deliciosa). J. Exp. Bot. 60, 3203–3219. Henry’s Law constant of terpenes with inert gas stripping coupled with PTR-MS. Niinemets, Ü., 2010. Mild versus severe stress and BVOCs: thresholds, priming and Water Air Soil Pollut. 226, 120. consequences. Trends Plant Sci. 15, 145–153. Staudinger, J., Roberts, P.V., 1996. A critical review of Henry’s law constants for Niinemets, Ü., Copolovici, L., Hüve, K., 2010. High within-canopy variation in environmental applications. Crit. Rev. Environ. Sci. Technol. 26, 205–297. isoprene emission potentials in temperate trees: implications for predicting Staudinger, J., Roberts, P.V., 2001. A critical compilation of Henry’s law constant canopy-scale isoprene fluxes. J. Geophys. Res.-Biogeosci. 115, G04029. temperature dependence relations for organic compounds in dilute aqueous Niinemets, Ü., Kännaste, A., Copolovici, L., 2013. Quantitative patterns between solutions. Chemosphere 44, 561–576. plant volatile emissions induced by biotic stresses and the degree of damage. Tholl, D., Chen, F., Petri, J., Gershenzon, J., Pichersky, E., 2005. Two sesquiterpene Front. Plant Sci. 4, 262. synthases are responsible for the complex mixture of sesquiterpenes emitted Niinemets, Ü., Monson, R.K. (Eds.), 2013. Biology, Controls and Models of Tree from Arabidopsis flowers. Plant J. 42, 757–771. Volatile Organic Compound Emissions. Springer, Dordrecht – Heidelberg – New Toome, M., Randjärv, P., Copolovici, L., Niinemets, Ü., Heinsoo, K., Luik, A., Noe, S.M., York – London. 2010. Leaf rust induced volatile organic compounds signalling in willow during Niinemets, Ü., Reichstein, M., 2002. A model analysis of the effects of nonspecific the infection. Planta 232, 235–243. monoterpenoid storage in leaf tissues on emission kinetics and composition in Valterova, I., Kunze, J., Gumbert, A., Luxova, A., Liblikas, I., Kalinova, B., Borg-Karlson, Mediterranean sclerophyllous Quercus species. Glob. Biogeochem. Cycle 16, A.-K., 2007. Male bumble bee pheromonal components in the scent of deceit 1110. http://dx.doi.org/10.1029/2002GB001927. pollinated orchids; unrecognized pollinator cues? Arthropod-Plant Interact. 1, Niinemets, Ü., Reichstein, M., 2003. Controls on the emission of plant volatiles 137–145. through stomata: sensitivity or insensitivity of the emission rates to stomatal van Eijck, A., Opatz, T., Taraborrelli, D., Sander, R., Hoffmann, T., 2013. New tracer closure explained. J. Geophys. Res.-Atmos. 108, 4208. http://dx.doi.org/10.1029/ compounds for secondary organic aerosol formation from beta-caryophyllene 2002JD002620. oxidation. Atmos. Environ. 80, 122–130. Plant, Cell and Environment (2016) 39,2027–2042 doi: 10.1111/pce.12775

Original Article How specialized volatiles respond to chronic and short-term physiological and shock heat stress in Brassica nigra

Kaia Kask1, Astrid Kännaste1, Eero Talts1,LucianCopolovici1,2 & Ülo Niinemets1,3

1Institute of Agricultural and Environmental Sciences, Estonian University of Life Sciences, Tartu 51014, Estonia, 2Institute of Technical and Natural Sciences Research-Development of “Aurel Vlaicu” University, Arad 310330, Romania and 3Estonian Academy of Sciences, Tallinn 10130, Estonia

ABSTRACT moderate heat stress that does not result in visible lesions can result in significant reductions in leaf photosynthetic activities Brassicales release volatile glucosinolate breakdown products (Sharkey 2005; Zhang & Sharkey 2009; Zhang et al. 2009) upon tissue mechanical damage, but it is unclear how the re- and modifications in volatile emission profiles (Loreto et al. lease of glucosinolate volatiles responds to abiotic stresses such 1998; Kleist et al. 2012; Possell & Loreto 2013). In fact, release as heat stress. We used three different heat treatments, simulat- of several constitutive and induced volatiles can be extremely ing different dynamic temperature conditions in the field to temperature sensitive and only moderate increases in tempera- gain insight into stress-dependent changes in volatile blends ture, even in the range of 30–38 °C can result in major changes and photosynthetic characteristics in the annual herb Brassica in the emissions (Hartikainen et al. 2009; Kleist et al. 2012; Hu nigra (L.) Koch. Heat stress was applied by either heating et al. 2013; Farré-Armengol et al. 2014). leaves through temperature response curve measurements Apart from ubiquitous stress responses elicited in a wide from 20 to 40 °C (mild stress), exposing plants for 4 h to temper- range of species in response to practically any severe stress, atures 25–44 °C (long-term stress) or shock-heating leaves to several plant taxonomic groups have specialized volatile de- 45–50 °C. Photosynthetic reduction through temperature re- fence pathways (Karban 2011). Glucosinolates constitute the sponse curves was associated with decreased stomatal conduc- unique secondary metabolites in the order Brassicales (Fahey tance, while the reduction due to long-term stress and collapse et al. 2001; Redovniković et al. 2008; Ishida et al. 2014), and so of photosynthetic activity after heat shock stress were associ- far the occurrence of more than 130 natural glucosinolates ated with non-stomatal processes. Mild stress decreased consti- has been documented (Agneta et al. 2014). Depending on their tutive monoterpene emissions, while long-term stress and molecular structure, they can be divided among aliphatic-, shock stress resulted in emissions of the lipoxygenase pathway aromatic- or indole glucosinolates (Hopkins et al. 1997; Ishida and glucosinolate volatiles. Glucosinolate volatile release was et al. 2014). Members of each group are biosynthesized from more strongly elicited by long-term stress and lipoxygenase different precursors via slightly different pathways. Yet, in gen- product released by heat shock. These results demonstrate that eral, the biosynthesis starts with the chain elongation of an glucosinolate volatiles constitute a major part of emission amino acid, continues with the creation of glucosinolate basic blend in heat-stressed B. nigra plants, especially upon chronic structure and ends up with the transformation of the core struc- stress that leads to induction responses. ture into the final glucosinolate molecule (Ishida et al. 2014). Glucosinolates are hydrolyzed to toxic volatile products by Key-words: Brassicales; glucosinolate breakdown products; myrosinases that are released from specialized cells upon me- heat shock; high temperature; lipoxygenase pathway; terpe- chanical wounding, for example, upon insect herbivory (Barth noid emission; volatile organic compounds. & Jander 2006; Wittstock & Burow 2010; Najar-Rodriguez et al. 2015). These breakdown products can be isothiocyanates, INTRODUCTION thiocyanates, nitriles, epithionitriles and oxazolidines (Bones fi Among abiotic stresses, heat stress is one of the most deleteri- & Rossiter 2006; Kos et al. 2012), disul des and thiols (Olivier ous factors resulting in major cellular damage once the heat et al. 1999; Agrawal & Kurashige 2003; Crespo et al. 2012) fi stress threshold has been exceeded (Bidart-Bouzat & Imeh- shown to signi cantly reduce herbivory by omnivorous insects Nathaniel 2008). Such deleterious heat effects are manifested (Hopkins et al. 2009), but also performance of specialist herbi- in ubiquitous stress responses such as collapse of leaf photosyn- vores (Bruinsma et al. 2007; de Vos et al. 2008). thetic activities and formation of reactive oxygen species in leaf Apart from mechanical damage due to herbivory, multiple tissues (Vacca et al. 2004; Hüve et al. 2011) and elicitation of re- stress factors including heat stress can lead to cellular damage lease of lipoxygenase (LOX) pathway volatiles (Maccarrone with potential release of myrosinases, but there is surprisingly et al. 1992; Copolovici et al. 2012). Nevertheless, even mild to little information about the relationship of abiotic stressors and volatile glucosinolate degradation products in brassicoid Correspondence: K. Kask. E-mail: [email protected] species (Wittstock & Burow 2010). Provided myrosinases are

©2016JohnWiley&SonsLtd 2027 2028 K. Kask et al. indeed released as the permeability of cellular membranes in- chemicals (Turk & Tawaha 2003). However, as a rapidly grow- creases upon developing heat stress, release of glucosinolate ing plant, B. nigra has become an aggressive weed in temperate breakdown products is likely, and the release of these special- Europe where it colonizes old fields (Gomaa et al. 2012). Be- ized volatiles might importantly contribute to the total heat- cause of extensive spread, more complex genome and greater triggered volatile blend next to ubiquitous emissions of LOX stress tolerance, it has become next to A. thaliana, an additional products. Furthermore, there is recent evidence that exoge- brassicoid model system in studies on plant biology, ecology nously applied glucosinolate volatiles, isothiocyanates, en- and plant-insect interactions (Dicke & van Loon 2000; hanced the development of heat-tolerance of Arabidopsis Fatouros et al. 2012). thaliana (Hara et al. 2013). Thus, heat-dependent induction of glucosinolate volatile emissions might contribute to develop- ment of induced abiotic stress resistance, but which tempera- MATERIALS AND METHODS ture conditions lead to the release of glucosinolate volatiles Plant material and how these potential emissions are related to ubiquitous stress responses is not known. Plants of B. nigra were grown from the seeds provided by the During their lifetime, plants can be exposed to a wide variety Department of Entomology, University of Wageningen, the of heat episodes differing in duration and temperature during Netherlands. This standardized seed-lot corresponds to a the stress, including short-term to mid-term excursions of leaf wild-grown Dutch B. nigra population that has been used in temperature to high values upon light flecks and upon clearing multiple studies on plant-insect interactions (Bruinsma et al. up the sky when shaded leaves are suddenly exposed to strong 2008; Khaling et al. 2015; Pashalidou et al. 2015). The seeds fi beam irradiance (Singsaas & Sharkey 1998; Sharkey 2005; were sown in 0.8 L plastic pots lled with a mixture of commer- Behnke et al. 2007; Way et al. 2011), as well as during heat cial garden soil with slow-release nutrients (Biolan Oy, Eura, waves that are predicted to become more common in the future Finland) and quartz sand. Day length was 12 h, and light inten- μ À2 À1 (Ameye et al. 2012). Given this variety, it is relevant to consider sity at plant level of 400 mol m s was provided by metal that the heat stress threshold is determined by the heat dose halide lamps (HPI-T Plus 400 W, Philips, Eindhowen, The (heat sum) that is dependent on both the actual temperature Netherlands). Day/night temperatures were maintained at and the duration of the heat episode (Bilger et al. 1984; 24/20 °C and relative humidity of 60%. The plants were fi Niinemets 2010a) as well as on possible increases of heat stress watered every other day to soil eld capacity. Five-to six- resistance due to acclimation and priming responses occurring weeks-old non-bolted plants with at least three fully developed through the heat wave (Niinemets 2010b). Thus, modifications leaves were used in the experiments. Temperature response in the volatile blend triggered by heat stress can depend on the curves were measured, and two different heat stress treatments type of heat stress that ultimately determines whether the stress were conducted in three to seven replications with different threshold for physiological damage is exceeded and whether plants (Fig. 1 for entire experimental protocol). New plants acclimation or priming responses can occur. were used for individual temperatures within heat stress treat- The goal of the present study was to investigate how foliage ments, and each plant was stressed and analysed only once. photosynthetic characteristics and emissions of constitutive and Hence, emissions from 57 plants were analysed with GC-MS, ubiquitous and specialized stress-elicited volatiles respond to and 65 plants were used in gas-exchange measurements (8 vol- heat-stress of various types in black mustard (B. nigra (L.) atiles samples were lost due to malfunctioning of GC-MS car- Koch, Brassicaceae). To our knowledge, there is no informa- tridge autosampler, but nevertheless, the sample size was tion about the relationships between the volatile glucosinolate never below three for individual treatments). degradation products and volatiles of other biosynthetic path- ways through abiotic stress treatments. Hence, next to the vol- Long-term heat stress treatment atile glucosinolates, we also studied the heat responses of emissions of lipoxygenase, terpenoid and shikimate pathway For long-term stress application, the potted plants were placed products. We used three different heat treatments simulating in a Percival growth chamber (model E-36HO, Percival Scien- dynamic temperature conditions in field environments that tific, Inc., Perry, IA, USA) under controlled conditions of light À À can occur during short-term heat episodes and longer-term intensity at plant level of 400 μmol m 2 s 1 provided for À heat waves to gain an insight how the share of different vola- 16 h day 1 (6:00–22:00 h), 60% of humidity and day/night tem- tiles changes in dependence on reversible and irreversible perature of 25/21 °C. Before the start of the heat stress treat- stress conditions of different duration. We hypothesized that ment, the plants were acclimated for 24 h under these growth severe heat stress leads to elicitation of glucosinolate volatiles chamber conditions. After the acclimation, the heat stress was and that the emissions are quantitatively more significant upon applied in two heat waves, one in the evening between 20:00– long-term stress due to elicitation of induction responses. 22:00 h and the second in the following morning between B. nigra is a fast-growing 1- to 2-m-tall annual herb native to 06:00–8:00 h, providing a total treatment period of 4 h, but in- the southern Mediterranean region of Europe, growing over a tervened with a night-time non-stressed period at 21 °C to al- broad temperature range and therefore classified as a stress tol- low for a recovery and induction of volatile stress responses. erant species (Duke 1983). It is occasionally cultivated for its Although the increase of temperature to preset conditions took seeds (Rajamurugan et al. 2012) as well as for leaves, extracts ~0.5 h, the chamber cool down to 21 °C after turning off the of which have allelopathic effects due to its secondary plant lights in the evening took ~1 h. Thus, the total stress period © 2016 John Wiley & Sons Ltd, Plant, Cell and Environment, 39,2027–2042 Responses of Brassica nigra to heat stress 2029

the custom-made gas-exchange system at 25 °C as described later (Fig. 1). The heat shock response was studied at tempera- tures of 25 (control), 45, 48 and 50 °C, and individual plants were used for each treatment.

Gas-exchange measurement system Foliage gas-exchange rates were measured with a custom- made open gas-exchange system described in detail in Copolovici & Niinemets (2010). The system has a temperature-controlled 1.2 L chamber made of double-walled glass and stainless steel bottom ring specially designed for vol- atile compound measurements. The chamber temperature is controlled by circulating thermostated water between the chamber walls (Copolovici & Niinemets 2010). An infra-red dual-channel gas analyzer operated in differential mode (CIRAS II, PP-Systems, Amesbury, MA, USA) was used to

measure CO2 and H2O concentrations at the chamber inlets and outlets (Copolovici & Niinemets 2010). The ambient air was drawn from outside, passed through a 10 L buffer volume Figure 1. A schematic representation of the study experimental and an HCl-activated copper tubing to scrub ozone and was hu- design. Brassica nigra plants were subjected to three different heat midified to ~60% humidity using a custom-made humidifier. treatments: moderate exposure through stepwise raising temperatures After passing the ozone scrubber, ozone concentrations were À1 through temperature response curve measurements, long-term heat less than 1 nmol mol (Sun et al. 2012). The chamber CO2 con- À stress and shock stress both achieved by stepwise increase in centration was 380–400 μmol mol 1 in these experiments. temperature, but differing in duration and treatment temperature. Plants were placed in a gas exchange system to collect volatiles and measure foliage net assimilation and transpiration rates. Collected Gas-exchange measurements and volatile volatiles were analysed with a GC-MS system followed by data analyses. In the case of temperature response curve measurements, the sampling in heat-stressed leaves treatment and physiological measurements occurred simultaneously at For plants subjected to long-term heat stress, at least three top the treatment temperature, while in the case of the two other experimental protocols, the physiological measurements were leaves, and for heat shock treated leaves all treated leaves were – 2 performed after the heat stress treatment at 25 °C. inserted into the leaf chamber (approximately 80 100 cm leaf area), and standard conditions of light intensity of À À 800 μmol m 2 s 1 and chamber temperature of 25 °C (leaf tem- was somewhat longer than 4 h. The temperatures applied in the perature was within ± 1 °C of chamber temperature) were growth chamber in individual stress experiments were 25 °C established. The measurements of net assimilation and transpi- (control), 30, 35, 40 and 44 °C. After completion of the stress ration rates were taken immediately after the gas flows had experiment at the given temperature, individual plants were stabilized in the system, typically in 10–20 min after plant enclosed in the custom-made gas-exchange system, and foliage enclosure. photosynthesis, transpiration and volatile organic compound After the gas flows had stabilized, volatiles were collected onto emission measurements were immediately carried out at 25 ° multi-bed stainless steel cartridges filled with three different C as described later (Fig. 1). carbon-based adsorbents for optimal adsorption of all volatiles between C3–C17 (Kännaste et al. 2014) (for details of cartridges). A portable 210-1003MTX air sampling pump (SKC Inc., Hous- Heat shock stress ton, TX, USA) was used for sampling the chamber outlet air fl À1 Shock stress treatment started at 9:00 in the morning, approxi- with a constant ow rate of 200 mL min . The sampling time mately 1 h after the light regime was automatically turned on in was20min,andthus,4Lofairwassampled.Blanksamplesfrom the plant room. Heat shock treatments followed the protocol of the empty cuvette were taken before the plant measurements. Copolovici et al. (2012). A temperature-controlled glass vessel equipped with a magnetic stirrer (Heidolph MR Hei-Standard Measurements of temperature response curves of with an EXT Hei-Con temperature sensor, Heidolph, net assimilation and volatile release Schwabach, Germany) was used. In the glass vessel, distilled water was heated to the desired temperature, and the plant’s Temperature response curve measurements started at 9:00 in the uppermost part with three fully developed leaves was inserted morning, 1 h after automatic turn-on of the light in the plant in the water with given temperature for 5 min. After the treat- room. Three upper leaves were enclosed in the leaf chamber ment, leaves were left to air-dry for approximately 5 min and and stabilized at the reference conditions of light intensity at leaf À2 À1 then measured for gas-exchange and volatile emissions using level of 800 μmol m s , chamber temperature of 20 °C, CO2 ©2016JohnWiley&SonsLtd, Plant, Cell and Environment, 39,2027–2042 2030 K. Kask et al.

À concentration of 380–400 μmol mol 1 and air humidity of 60% 2009), shikimate pathway volatiles (different benzenoids such as until stomata opened and gas-exchange rates stabilized, typically methyl salicylate) (Wahid et al. 2007; Betz et al. 2009) and gluco- in 20–30 min since leaf enclosure. After reaching the steady- sinolate breakdown compounds (various sulphur- and nitrogen- state, net assimilation and transpiration rates were recorded containing compounds, often containing the CN functional group) and volatile organic compounds were collected for 20 min as de- (Sønderby et al. 2010; Ishida et al. 2014). It is primarily these scribed earlier (Fig. 1). The chamber temperature was raised to latter compounds that give the plants from Brassicaceae the char- the next higher temperature, the plant was conditioned again at acteristic ‘cabbage smell’ (Buttery et al. 1976). No sesquiterpenes this temperature for 20–30 min, and gas-exchange rates were re- were observed in the emission blends in these experiments. corded and volatiles collected. Foliage gas-exchange rates were measured at temperatures 20, 25, 30, 35 and 40 °C, while volatile organic compounds were collected at 20, 25, 30, and 40 °C. Data analyses

Net assimilation rate (A) and stomatal conductance (gs)per GC-MS analyses leaf area and intercellular CO2 concentration (Ci)werecalcu- lated according to von Caemmerer & Farquhar (1981) and The cartridges with adsorbed volatiles were analysed with a the volatile emission rates according to Niinemets et al. (2011). combined Shimadzu TD20 automated cartridge desorber con- For normalization of data and residuals, logarithmic data nected to a Shimadzu 2010 Plus GC–MS system (Shimadzu transformation was used, and average values of gas-exchange Corporation, Kyoto, Japan). Adsorbent cartridges were back and volatile emission rates at different temperatures were com- flushed with high purity He (99,9999% AGA, Linde Group, pared with one-way ANOVA followed by Tukey’s post-hoc Tallinn, Estonia) during thermal desorption with the following test. In addition, linear- and non-linear regression analyses À1 TD20 parameters: He purge flow rate of 40 mL min , primary were conducted to explore the relationships among gas- desorption temperature of 250 °C, primary desorption time of exchange and volatile emission rates and among the emission 6 min, the second stage trap temperature during primary de- rates of different volatile compound classes. (StatSoft Inc., sorption of À20 °C, the second stage trap desorption tempera- Tulsa, OK, USA) was used in these analyses. Heat stress effects ture of 280 °C, hold time of 6 min. The compounds were on volatile bouquets and changes in the volatile bouquets separated on a Zebron ZB-624 fused silica capillary column through the temperature response curve were evaluated by (0.32 mm i.d., 60 m length, 1.8 μm film thickness, Phenomenex, principal component analysis (PCA) (Wold et al. 1987). Load- Torrance, CA, USA) using He with a flow rate of ing and score plots were derived after mean-centering and À1 1.48 mL min as the carrier gas. The following GC oven pro- logarithmic data transformation. Redundancy data analysis À1 gramme was employed: 40 °C for 1 min, 9 °C min to 120 °C, was also used to test for the differences in bouquets among À1 À1 2°Cmin to 190 °C, 20 °C min to 250 °C for 5 min. The stress treatments. Monte-Carlo permutation tests were used Shimadzu QP2010 Plus mass spectrometer was operated in to evaluate the statistical significance of the model. Multivari- the electron impact mode. The transfer line temperature was ate data analyses were performed with CANOCO 5.0 software 240 °C and ion-source temperature 150 °C. The GC-MS system (ter Braak and Smilauer, Biometris Plant Research Interna- was calibrated as explained in Kännaste et al. (2014) and the tional, the Netherlands). All statistical tests were considered compound quantification follows Copolovici et al. (2009). The statistically significant at P < 0.05. compounds were identified by comparing the mass-spectra of volatiles with the spectra of authentic standards of the highest purity purchased from Sigma-Aldrich (St. Louis, MO, USA, RESULTS GC purity, most of the standards) and Fluka (Buchs, Switzer- Effects of heat stress on foliage photosynthetic fi land, GC purity, 1-hexanol and methyl salicylate). For quanti - characteristics cation of emissions, the GC-MS system was calibrated with standard compounds in hexane solution. Six concentrations of Temperature response curve measurements indicated that leaf each compound (range 0.1–1 μL/mL) were prepared, and net assimilation rate (A)ofBrassica nigra had a broad temper- 1 μL of each sample was injected into the adsorbent cartridge. ature optimum between 20 and 30 °C (Fig. 2a). Net assimilation

The cartridge was back flushed with a stream of N2 at rate declined over temperatures 35–40 °C, reaching ~4-fold À 200 mL min 1 to simulate conditions during sampling of vola- lower values at the highest measurement temperature than at tiles. Ultimately, the calibration factor for each compound 20 °C (Fig. 2a). Stomatal conductance to water vapor (gs)de- was estimated as the slope of the GC chromatogram peak area creased with increasing temperature through the entire tem- À À versus compound mass concentration. perature range from on average (±SE) 99 ± 13 mmol m 2 s 1 À À We grouped the volatile compounds released according to at 20 °C to 13 ± 4 mmol m 2 s 1 at 40 °C (Fig. 2b). Thus, the their formation pathways (Table 1) as fatty acid derived com- intercellular CO2 concentration (Ci) decreased with increasing À pounds (lipoxygenase volatiles, also called green leaf volatiles) temperature to low values of 60–100 μmol mol 1 at the highest (Matsui 2006), geranyl diphosphate (GDP) derived volatiles measurement temperature (data not shown). (GDP-pathway, various monoterpenoids synthesized from Exposure of plants to long-term heat stress (4 h exposure to GDP) (Maffei 2010), geranylgeranyl diphosphate (GGDP) de- given temperature, measurements of photosynthetic character- rived volatiles (homoterpenes such as DMNT and some caroten- istics at 25 °C) was associated with minor modifications in A oid breakdown products such as geranyl acetone) (Arimura et al. and gs over the treatment temperature range of 25–35 °C, but © 2016 John Wiley & Sons Ltd, Plant, Cell and Environment, 39,2027–2042 Responses of Brassica nigra to heat stress 2031

À À Table 1. Average ±SE emission rates (pmol m 2 s 1) of different volatiles released from leaves of Brassica nigra in response to three different heat treatments grouped according to the compound formation pathways

No Compound Temperature response curve Long-term stress 20 °C (3) 25 °C (3) 30 °C (3) 40 °C (3) 25 °C (4) 30 °C (5)

Fatty acid derived compounds 1 (E)-3-Hexen-1-ol 2 (E, E)-2, 4-Hexadienal 3 (Z)-3-Hexen-1-ol 0.7# 10 2.9 ± 1.4 a* 4 (Z)-3-Hexenyl acetate 6 5 (Z)-3-Hexenyl formate 6 1-Hexanol 0.17 1.5 ± 0.6 a 1.2 ± 0.7 a 7 1-Pentanol 3.1 8 1-Penten-3-ol 9 1-Penten-3-one 10 2-Ethylfuran 11 2-Pentanone 2.7 2.2 ± 0.7 a 12 Heptanal 3.8 ± 2.0 a 0.8 1.15 ± 0.14 a 0.71 ± 0.11 a 10.0 ± 4.2 ab 3.8 ± 1.9 ab 13 Hexanal 5.8 ± 4.2 a 6.4 ± 0.9 a 2.8 ± 0.9 a 0.96 ± 0.34 a 17 ± 5 ab 3.5 ± 1.1 a 14 Nonanal 22 ± 14 a 4.9 ± 3.7 a 6.8 ± 2.9 a 3.6 ± 3.0 a 54 ± 22 ab 20 ± 10 ab 15 Octanal 7.6 ± 4.7 a 0.9 ± 0.7 a 3.5 ± 1.4 a 1.5 ± 1.0 a 22 ± 9 ab 87.6 ± 3.9 ab GDP-pathway 16 3-Carene 6.0 ± 0.6 a 2.55 ± 0.20 a 2.509 ± 0.032 a 0.103 ± 0.024 b 6.9 ± 2.8 a 2.4 ± 0.9 a 17 Camphene 0.66 ± 0.32 a 0.27 ± 0.08 a 18 Limonene 0.63 ± 0.25 a 0.42 ± 0.10 a 0.45 ± 0.09 a 1.6 ± 0.5 a 0.26 ± 0.16 a 19 α-Pinene 10.9 ± 2.1 a 3.75 ± 0.27 a 2.604 ± 0.023 a 0.272 ± 0.021 b 17 ± 8 a 1.8 ± 1.0 a 20 β-Pinene 0.54 ± 0.17 a 0.302 ± 0.021 a Shikimate pathway 21 Methyl salicylate 0.031 Glucosinolate breakdown products 22 Tetramethylthiourea 0.47 ± 0.20 a 0.42 23 2-Propenenitrile 12 9 14 ± 10 a 8.9 ± 1.5 ab 24 Allyl isothiocyanate 20 2.6 25 Cyclohexyl isocyanate 2.1 3.4 26 Cyclohexyl isothiocyanate 1.3 ± 0.6 a 1.2 ± 0.6 a 0.91 4.7 ± 2.1 a 0.7 27 Methanethiol 0.28 0.6 28 Methyl isothiocyanate 29 Tetramethylurea 2.7 ± 0.9 a 1.5 ± 1.2 a GGDP-pathway 30 6-Methyl-5-hepten-2-one 11 ± 8 a 7.29 ± 0.7 a 2.57 0.36 ± 0.20 a 7.5 ± 3.9 a 2.2 ± 2.0 a 31 Geranyl acetone 20 ± 6 a 23.3 ± 3.5 a 3.0 ± 2.6 a 6.7 ± 2.6 a 1.35 ± 0.27 ab 2.43 ± 1.02 ab

Different stress treatments as outlined in Fig. 1. Five primary compound groups were distinguished on the basis of compound synthesis pathways: fatty acid derived volatiles (products of lipoxygenase pathway, also called green leaf volatiles), geranyl diphosphate (GDP) derived volatiles (GDP-path- way, various monoterpenoids), geranylgeranyl diphosphate (GGDP) derived volatiles (homoterpenes such as DMNT and some carotenoid break- down products including geranyl acetone), and shikimate pathway volatiles (different benzenoids such as methyl salicylate) and glucosinolate breakdown compounds (various sulphur- and nitrogen-containing volatiles). The compound number corresponds to the number in the PCA factor loading plot (Fig. 4). Number of replicates (individual plants) is shown in parenthesis below each stress temperature. #for emissions of compounds, which were above the detection limit in only one of the replicate experiments, no SE values could be calculated. *different letters indicate statistical significance at P < 0.05. further increases in temperature were associated with both rates observed at treatment temperatures 48 and 50 °C reduced A (4.5-fold reduction at 44 °C compared with the value (Fig. 2a). Intercellular CO2 concentration was greater in heat at 25 °C) and gs (2.8-fold reduction), from 69 to shock treated than in control leaves (P < 0.005). À2 À1 24 mmol m s (Fig. 2a, b, respectively). Intercellular CO2 concentration was similar through temperatures 25–40 °C À Temperature response curves of volatile emission (192 ± 16 μmol mol 1), but there was a significant increase in À1 Ci at 44 °C (334 ± 30 μmol mol , P < 0.01 for the difference Total emission of fatty acid derived compounds was low and among the means). Heat shock treatment (exposure for weakly affected by temperatures through the temperature 5 min to given temperature, measurements of photosynthetic response curves (Fig. 3a). Among C6-volatiles, only 1-hexanol characteristics at 25 °C) was associated with major reductions and (Z)-3-hexen-1-ol were above the detection limit at 25 and in both A and gs, with barely positive rates of net assimilation 40 °C, and the rest of the emissions were due to aliphatic alde- observed after 45 °C treatment, and negative net assimilation hydes (Table 1). Total emission of monoterpenoids (GDP- ©2016JohnWiley&SonsLtd, Plant, Cell and Environment, 39,2027–2042 2032 K. Kask et al.

À À Table 1. Average ±SE emission rates (pmol m 2 s 1) of different volatiles released from leaves of Brassica nigra in response to three different heat treatments grouped according to the compound formation pathways

No Long-term stress Shock stress

35 °C (6) 40 °C (6) 44 °C (7) 25 °C (3) 45 °C (4) 48 °C (3) 50 °C (7) Fatty acid derived compounds 1 3900 ± 1600 26 45 ± 17 3 9.4 ± 2.2 a 20 ± 13 a 300 ± 170 a 18 ± 18 a 1.1 ± 0.8 a 14 130 ± 50 a 414±5a 17±9a 5 36 6 0.587 ± 0.042 a 7.6 ± 3.6 a 6.3 ± 3.2 a 2.7 0.087 ± 0.027 a 64 ± 30 b 7 14 ± 10 a 23 ± 10 a 17 ± 5 8 21 73 ± 19 180 ± 90 966 43 ± 8 10 2.613 ± 0.012 a 11 ± 7a 5.0 133 ± 47 11 16 ± 8 b 12 3.9 ± 1.7 a 49 ± 13 b 14 ± 6 ab 1.8 ± 0.7 a 3.6 ± 1.0 a 15.9 ± 1.3 a 6.0 ± 3.1 a 13 6.0 ± 1.9 a 106 ± 41 b 44 ± 19 ab 5.2 ± 2.0 a 6.8 ± 2.0 a 31.4 ± 2.6 a 22 ± 9 a 14 7.8 ± 4.6 a 53 ± 16 b 55 ± 22 b 11.7 ± 3.0 a 11.4 ± 2.2 a 54 ± 33 a 21 ± 5 a 15 4 ± 1.9 a 43 ± 13 b 27 ± 11 b 3.9 ± 2.0 a 4.1 ± 0.6 a 22 ± 6 b 7.8 ± 20 a GDP-pathway 16 5.1 ± 2.3 a 5.5 ± 1.6 a 5.4 ± 1.5 a 0.61 ± 0.44 a 6.9 ± 2.9 a 9.5 ± 4.1 a 13.9 ± 4.7 a 17 0.66 ± 0.12 a 0.23 ± 0.11 a 0.19 9 0.29 ± 0.19 a 1.17 ± 0.20 a 2.17 ± 1.02 a 18 1.4 ± 0.7 a 2.9 ± 1.1 a 4.7 ± 4.4 a 6 ± 6 a 1.7 ± 0.9 a 3.3 ± 2.2 a 6.23 ± 1.03 a 19 6.1 ± 3.0 a 4.6 ± 1.3 a 4.6 ± 2.3 a 3.0 ± 1.0 a 8 ± 6 a 27 ± 6 ab 38 ± 10 a 20 1.06 ± 0.22 a 0.47 ± 0.20 a 0.31 ± 0.23 a 0.85 ± 0.14 a 1.69 ± 0.10 a Shikimate pathway 21 0.134 ± 0.042 a 0.118 ± 0.031 a Glucosinolate breakdown products 22 0.6 14 ± 7 a 110 ± 80 a 96 ± 38 a 178.9 ± 0.7 ab 400 ± 90 b 23 14.7 ± 4.2 ab 30 ± 8 ab 81 ± 25 b 19 24 12.3 ± 4.8 a 1300 ± 700 b 2.5 250 ± 15 25 1.6 14.8 ± 0.8 a 38 ± 22 a 22.4 ± 4.3 a 17 ± 5 a 10.5 ± 3.1 a 26 0.35 28 110 ± 80 a 0.86 ± 0.30 a 10 ± 9 a 2.7 ± 0.5 a 27 3.3 ± 1.2 a 13 ± 8 a 28 3.9 21 ± 11 15.1 ± 2.2 29 3.1 30 ± 24 a 63 ± 26 a 10 ± 6 a 14 ± 6 a GGDP-pathway 30 3.2 ± 1.7 a 3.6 ± 1.3 a 3.1 ± 1.3 a 1.60 ± 0.27 a 1.4 ± 0.5 a 3.4 ± 1.5 a 5.1 ± 1.9 a 31 0.94 ± 0.70 a 5.7 ± 1.8 b 2.01 ± 0.5 ab 7.5 ± 0.9 a 4.7 ± 1.9 a 4.9 ± 3.4 a 8.6 ± 3.2 a

Different stress treatments as outlined in Fig. 1. Five primary compound groups were distinguished on the basis of compound synthesis pathways: fatty acid derived volatiles (products of lipoxygenase pathway, also called green leaf volatiles), geranyl diphosphate (GDP) derived volatiles (GDP-path- way, various monoterpenoids), geranylgeranyl diphosphate (GGDP) derived volatiles (homoterpenes such as DMNT and some carotenoid break- down products including geranyl acetone), and shikimate pathway volatiles (different benzenoids such as methyl salicylate) and glucosinolate breakdown compounds (various sulphur- and nitrogen-containing volatiles). The compound number corresponds to the number in the PCA factor loading plot (Fig. 4). Number of replicates (individual plants) is shown in parenthesis below each stress temperature. #for emissions of compounds, which were above the detection limit in only one of the replicate experiments, no SE values could be calculated. *different letters indicate statistical significance at P < 0.05. pathway compounds) decreased considerably from 20 to 40 °C GDP-pathway volatiles, GGDP-pathway volatiles decreased À À (Fig. 3b). Among GDP-pathway compounds, α-pinene followed from 0.0309 ± 0.0027 nmol m 2 s 1 at 20–25 °C to 0.0043 À À by 3-carene was the dominating volatiles at all temperatures ± 0.0012 nmol m 2 s 1 at 30–40 °C (Fig. 3d). (Table 1). Similarly to LOX-compounds, total emission of gluco- sinolate breakdown products was low and not affected by tem- perature (Fig. 3c). Allyl isothiocyanate and 2-propenenitrile Effects of long-term heat stress on volatile were rare volatiles at only at 20 °C, while the emissions of emissions cyclohexyl isothiocyanate were not affected by temperature (Table 1). Total emission of GGDP-pathway volatiles (caroten- Long-term heat stress had no significant effect on LOX- oid breakdown products) was overall low, and the emissions compounds over the treatment temperatures 25–35 °C, were dominated by geranyl acetone (Table 1). Similarly to but the emissions were strongly enhanced upon exposure © 2016 John Wiley & Sons Ltd, Plant, Cell and Environment, 39,2027–2042 Responses of Brassica nigra to heat stress 2033

Figure 2. Effects of three different heat treatments on mean (±SE) net assimilation rate (A) and stomatal conductance to water vapor (B) in leaves of Brassica nigra. The treatments included raising temperatures through temperature response curve measurements (mild stress, total exposure for 1 h to given moderately high temperature), long-term stress (chronic stress, 4 h treatment with given temperature), and short-term heat shock stress (severe stress, exposure for 5 min to potentially lethal temperature). In the case of long-term and heat shock stress treatments, the measurements were conducted at 25 °C after heat exposure. Three to seven replicate plants were used for each treatment temperature (in each case, the number of biological replicates is shown above the temperature values). For statistical analyses, data were log-transformed and compared with one-way ANOVA followed by Tukey’s post-hoc test. Different letters indicate statistically significant differences at P < 0.05. to 40 °C and 44 °C (Fig. 3a). Moreover, the emission com- position significantly changed as at 40 °C treatment, the À2 À1 plants began to release 2-ethylfuran and 1-penten-3-ol Figure 3. Rates of emission (nmol m s , mean ± SE) of fatty acid and at 44 °C treatment, the emission of these volatiles derived compounds (A), GDP-pathway compounds (B), glucosinolate breakdown products (C) and GGDP-pathway compounds (D) from increased even more (Table 1). At 44 °C treatment, the the foliage of Brassica nigra in three different temperature stress LOX bouquet was dominated by (Z)-3-hexen-1-ol, treatments – temperature response curve (black bars), long-term (grey 1-penten-3-ol and 1-penten-3-one and aliphatic aldehydes bars), and shock stress (white bars). Number of biological replicates is hexanal, nonanal and octanal (Table 1). Heat stress effects shown above the temperature values. Definition of compound groups on the release of glucosinolate breakdown products and emissions of individual volatiles within each group is demonstrated followed the same pattern as that observed for LOX- in Table 1. Stress application, statistical analysis and data presentation compounds (Fig. 3c). At treatment temperatures of 25– as in Figs 1 and 2. 35 °C, total emission of glucosinolate breakdown products À À remained at a low level of 0.0087–0.0193 nmol m 2 s 1 followed by tetramethylthiourea, cyclohexyl isothiocya- (Fig. 3c) but increased somewhat already at 40 °C treat- nate and 2-propenenitrile (Table 1). Emissions of GDP- ment, and a major emission burst of 1.10 pathway and GGDP-pathway compounds were not af- À À ±0.43nmolm 2 s 1 was observed at the highest applied fected by treatment temperature (Figs. 3b, d), but emis- temperature treatment (Fig. 3c). At this temperature sions of the benzenoid and methyl salicylate were treatment, allyl isothiocyanate was the dominating volatile detected after higher temperature treatments (Table 1). ©2016JohnWiley&SonsLtd, Plant, Cell and Environment, 39,2027–2042 2034 K. Kask et al.

Influences of heat shock stress on volatile release curve measurements constituting a mild-short stress, long-term heat stress and heat shock stress) differed significantly from Short-term exposure of leaves to severe heat stress increased each other (Fig. 4. Monte-Carlo permutation test, P < 0.05). total emissions of all volatiles at heat shock temperatures A certain plant-to-plant variability was observed in the release higher than 48 °C and for glucosinolate breakdown products al- of LOX volatiles, (E)-3-hexen-1-ol, (Z)-3-hexenyl formate and ready at 45 °C treatment (Fig. 3). A particularly strong en- 1-penten-3-one (Table 1), upon heat stress. The plants emitting hancement was observed for LOX volatiles that increased À À these volatiles, and experiencing a severe stress under the im- from 0.036 ± 0.019 nmol m 2 s 1 at 25 °C treatment to 1.4 À À posed conditions were grouped in the upper right corner of 2 1 E ± 1.0 nmol m s at 50 °C treatment (Fig. 3a). ( )-3-hexen- PCA score plot (Fig. 4b). In the case of other stressed plants, 1-ol was released at temperature treatments of 45 and 48 °C C5-volatiles such as 1-pentanol, 1-penten-3-ol and 2-ethylfuran as well, but at 50 °C treatment, the plants started to release ad- and some glucosinolate degradation products such as methyl E,E - ditional C5-volatiles and C6-volatiles such as ( ) 2,4- isothiocyanate constituted a stress signal of heat-stressed B. Z hexadienal, ( )-3-hexenyl formate, 1-pentanol, 1-penten-3-ol nigra. In the case of emissions higher than approximately and 1-penten-3-one (Table 1). The total emission of GDP- À2 À1 À À 10 pmol m s (Z)-3-hexen-1-ol, tetramethylthiourea, pathway volatiles rose from 0.013 ± 0.008 nmol m 2 s 1 at 25 ° À À isocyanides, methyl isothiocyanate and tetramethylurea be- C to 0.048 ± 0.016 nmol m 2 s 1 at 50 °C treatment, mainly came stress signals (Table 1 and Fig. 4a, b). In the case of allyl due to enhanced emissions of α-pinene and 3-carene (Table 1, isothiocyanate, the limit of emission for classification the plant À À Fig. 3b). Glucosinolate breakdown products were not detected as stressed in the PCA plot was about 200 pmol m 2 s 1 at 25 °C treatment (Fig. 3c), but their emission increased from À À (Table 1 and Fig. 4a, b). 0.113 ± 0.035 nmol m 2 s 1 at 45 °C treatment to 0.37 À2 À1 In the case of emissions observed at different temperatures ±0.09nmolm s at 50 °C treatment (Fig. 3c). Methyl iso- through the temperature response curve, the blend of emis- thiocyanate was detected only at 50 °C treatment and sions at 25 °C did not differ from that at 20 °C (Fig. 4b). Anal- tetramethylthiourea together with allyl isothiocyanate were ogously, the emission blends at higher temperatures (30 and the dominating volatiles at 50 °C treatment (Table 1). Total 40 °C) did not differ from those at 20 and 25 °C (Fig. 4b). In emission of GGDP-pathway volatiles remained similarly low À2 À1 contrast, in long-term stress, emissions after treatments at 40 as in the long-term stress (0.006 to 0.012 nmol m s ), and it and 44 °C differed significantly from the control treatment fi was not signi cantly different among the heat shock treatments (Fig. 4b). Analogously, heat shock of 45–50 °C resulted in ma- (Fig. 3d). jor changes in the emission blend compared with the controls (Fig. 4b). Changes in emission blends among different As emissions of LOX products and glucosinolate breakdown temperature treatments and among different products were low through the entire temperature response – temperatures within treatments curve, 20 40 °C (Fig. 3a, c), all temperature response curve data were distributed close to the control plants for long-term – and Principal component analysis demonstrated that the emission heat shock stresses in the upper corner of the left side of PCA blends in three different treatments (temperature response score plot (Fig. 4b). High emission of glucosinolate breakdown

Figure 4. Loading-plot (A) and score-plot (B) of PCA analysis based on the emissions of volatiles released (Table 1 for the emission rates) from non- stressed and heat-stressed Brassica nigra plants. In the loading plot, the numbers represent different volatiles (Table 1 for the coding of individual compounds), while in the score plot, each symbol represents an individual non-stressed (empty symbols) or heat-stressed plant (filled symbols). Red symbols correspond to temperature response curve measurements, black symbols to long-term heat stress and green symbols to heat shock stress (Fig. 1 for the details of heat shock treatments and Fig. 3 for the heat stress effects on key volatile groups). In the loading plot, the impact of chemical compounds on PCA increases with the distance from the origin of the co-ordinate system.

© 2016 John Wiley & Sons Ltd, Plant, Cell and Environment, 39,2027–2042 Responses of Brassica nigra to heat stress 2035 volatiles (Fig. 4a), elicitation of methyl salicylate emissions and changes in the composition of GDP-pathway (e.g. induction of camphene emissions) were characteristic to plants exposed to long-term stress at 40 and 44 °C (Fig. 4b). Finally, heat shock treatments differed from long-term heat stress by greater elici- tation of LOX-compounds and GDP-pathway volatiles and lower induction of glucosinolate breakdown products (Figs 3 &4).

Correlations among emissions of different volatile compound classes and among emissions and photosynthetic characteristics Through the temperature response curves, the emissions of GDP-pathway compounds (Fig. 3b) were positively correlated with A (r = 0.71) and gs (r =0.83, P < 0.05 for both), but low- level emissions of LOX-compounds (Fig. 3a) and glucosinolate breakdown products (Fig. 3c) were not correlated with foliage photosynthetic characteristics. In long-term stress treatment, emissions of glucosinolate breakdown products and LOX-compounds were positively correlated through treatment temperatures of 25 to 40 °C, but the correlation was lost in leaves at 44 °C treatment where the increase in the emission of glucosinolate breakdown prod- Figure 6. Emission of GDP-pathway compounds (various ucts was much stronger than that in LOX-compounds (Fig. 5). monoterpenoids, Table 1) in relation to the emission of fatty acid Emission of GDP-pathway compounds was also positively cor- derived compounds in Brassica nigra in long-term (A) and shock (B) related with LOX-compound emission, but the slopes differed stress treatments (Fig. 3 for average emission rates). The emission rates – were measured after heat treatments at 25 °C (heat stress treatments as for treatment temperature range 25 35 °C and 40 and 44 °C, in Fig. 1). Number of biological replicates is shown after the treatment fl re ecting the stronger increase of LOX-compounds over this temperature values. Each symbol corresponds to an individual temperature range (Fig. 6a). Foliage photosynthetic character- replicate experiment. Data fitting as in Fig. 5. istics were negatively correlated with emissions of glucosino- late breakdown products (Fig. 7a, b) and LOX-compounds (Fig. 7c, d), whereas the correlations were strongly non-linear (Figs 7 & 8). In the heat shock treatments, glucosinolate breakdown prod- ucts and LOX emission were not correlated (P > 0.8), although the slope was shallower than that observed for long-term heat treatments due to greater elicitation of emissions of LOX- compounds in heat shock treatments (cf. Figs 5 & 3). The emissions of GDP-pathway volatiles and LOX-compounds were positively correlated over the temperature range of 25 to 48 °C (Fig. 6b), but the treatment at 50 °C was characterized by much stronger elicitation of LOX-compounds (Fig. 3). Analysis of relationships among photosynthetic characteristics and emissions of glucosinolate and LOX-compounds, indicated that photosynthetic activity was lost earlier than stress volatile emissions were elicited (cf. Figs 2 & 3).

DISCUSSION Figure 5. Emission of glucosinolate breakdown products in relation How different types of heat stress affect leaf to emission of fatty acid derived compounds in Brassica nigra in the long-term stress treatment. The plants were exposed for 4 h to given photosynthetic characteristics in B.nigra temperature, and volatile release was measured after the treatment at High temperature stress alters a plethora of plant physiological 25 °C. Number of biological replicates is shown after the temperature values. Each symbol corresponds to an individual plant (Table 1 for functions ranging from cellular to organ and whole plant pro- stress effects on average emissions). Data over 25–40 °C were fitted by cesses (Ludwig-Müller et al. 2000; Sung et al. 2003; Loreto a linear regression. et al. 2006; Velikova et al. 2009; Usano-Alemany et al. 2014). ©2016JohnWiley&SonsLtd, Plant, Cell and Environment, 39,2027–2042 2036 K. Kask et al.

Figure 7. Correlations of emissions of glucosinolate breakdown products (A, B) and fatty acid derived compounds with net assimilation rate (A, C) and stomatal conductance to water vapor (B, D) in Brassica nigra in the long-term stress treatment (Fig. 2 for average values of photosynthetic characteristics and Fig. 3 for average emission rates). Heat stress treatment consisted of a 4 h exposure of plants to given temperature, after which the volatile release was measured at 25 °C. The insets demonstrate emissions with the y-scale reversed and log-transformed. Individual symbols stand for replicate experiments. Additionally number of biological replicates is shown after the temperature values. Data were fitted by non-linear regressions.

In the current study high temperature resistance of B. nigra was agreement with previous studies indicating heat dose depen- studied by three sets of experiments with differing severity of dent reductions in foliage photosynthetic capacity after a cer- heat stress, including measurements of temperature responses tain threshold heat dose has been exceeded (Kadir et al. 2007; where temperature was raised up to 40 °C (mild stress), long- Hüve et al. 2011). Such decreases in photosynthetic capacity term moderate heat stress where plant temperature was raised might reflect inactivation of foliage photosynthetic electron up to 44 °C for 4 h and heat shock stress where leaves were ex- transport processes due to increased leakiness of membranes posed to sublethal to lethal temperatures of 45–50 °C for 5 min. and enhanced non-photochemical quenching (Havaux 1993; Given that Brassicaceae have a specialized defense system con- Lu & Zhang 2000; Zhang & Sharkey 2009; Zhang et al. stituting of high constitutive levels of glucosinolates and release 2009), but they might also result from irreversible cellular dam- of glucosinolate breakdown products, the key aim of the study age (Hüve et al. 2011). As the result of sustained inhibition of was to gain insight into the relationships among ubiquitous photosynthetic activity or cellular damage, foliage photosyn- stress responses and brassicoid-specific stress responses thetic activity does not recover upon return to lower tempera- through the different heat stress treatments. tures as was also observed in our study after long-term heat Among the ubiquitous stress responses, foliage net assimila- stress at 44 °C and heat shock treatments between 45–50 °C tion rate (A) and stomatal conductance (gs) decreased in all (Figs 2 & 8a). heat stress treatments (Fig. 2), but the mechanism of photosyn- thetic decline differed among the different types of heat treat- ment (Figs 2 & 8a). In temperature response curve Different heat stresses have varying effects on measurements, the temperature-dependent reduction in A re- lipoxygenase pathway volatiles sulted from reduced intercellular CO2 concentration (Ci)due to a reduction in gs (Fig. 2). Closure of stomata is often ob- The release of LOX volatiles in low amounts from flowers, served at higher temperatures (Cui et al. 2006; Hüve et al. leaves or fruits is a widespread phenomenon (Bengtsson et al. 2006; Hüve et al. 2011; Copolovici et al. 2012), and this response 2001; Ceuppens et al. 2015). In our study, characteristic C6 reduces water loss in conditions of higher vapor pressure deficit LOX volatiles such as 1-hexanol and (Z)-3-hexen-1-ol were typical to high temperature (Shinohara & Leskovar 2014). emitted in small quantities, close to the analytical detection However, after long-term heat stress at 40 and 44 °C and heat limit, at low temperatures (Table 1). In addition, aliphatic satu- shock stress at 45–50 °C, Ci actually increased, indicating that rated aldehydes hexanal, heptanal, octanal and nonanal were heat stress resulted in stronger reductions in leaf photosyn- consistently emitted at low level through all three sets of exper- thetic capacity than in stomatal conductance. This result is in iments (Table 1). Although in the literature, the LOX-pathway © 2016 John Wiley & Sons Ltd, Plant, Cell and Environment, 39,2027–2042 Responses of Brassica nigra to heat stress 2037

elicitation of emissions of LOX-compounds constitutes a classic indicator of cellular damage (Matsui 2006; Jansen et al. 2009; Matsui et al. 2012). In our study, long-term heat stress at 40 and 44 °C and heat shock treatment at 48 and 50 °C resulted in a major increase in lipoxygenase pathway volatiles, while the lipoxygenase volatile emissions remained low through the temperature response curve measurements (Figs 3a & 8, Table 1). At these high temperatures in both heat stress treat- ments, the plants began to release next to C6-volatiles also var- ious C5-volatiles such as 1-pentanol, 1-penten-3-ol, 1-penten-3- one, and C7-volatile (Z)-3-hexenyl formate (Table 1) that are also formed through the LOX-pathway (Shen et al. 2014) and are emitted upon several other stresses (de Gouw et al. 1999; Brilli et al. 2012). Previous studies indicate that upon short-term heat pulses as those applied in the heat shock treatments, LOX product emis- sions are elicited between temperatures 46–49 °C according to a highly non-linear switch-type response (Loreto et al. 2006; Copolovici et al. 2012) as was also observed in our study (Fig. 3a). This temperature range corresponds to major increases in plasmalemma membrane permeability and time- dependent reductions in foliage photosynthetic activity upon return to lower temperature (Hüve et al. 2011). However, pho- tosynthesis rate of B. nigra strongly decreased upon heat shock at 45 °C as well (Fig. 2), but no significant elicitation of LOX- compounds was observed (Fig. 3a). This discrepancy suggests that the reduction in photosynthetic activity at this temperature likely reflected enhanced engagement of non-photochemical processes or impaired photochemistry without direct membrane-level damage. Similar to our study (Fig. 3a), long-term exposure, from sev- eral hours to days, to moderately high temperatures of 35–45 ° C resulted in elicitation of LOX product emissions that was ac- companied by reduced A (Fig. 7c) in several tree species (Kleist et al. 2012). This evidence together with our observations fur- Figure 8. A comparative scheme of the main effects of three different ther underscores that heat stress impact on cellular processes heat stress applications (physiological temperature response, long-term is dose-dependent, and even moderately high sustained heat and shock heat stress) on photosynthetic characteristics (A) and on waves can result in major cellular lesions progressively leading total emission of main classes of emitted volatiles (B). to the cessation of photosynthetic activity. is primarily associated with the emission of C6 aldehydes and Constitutive terpenoid release upon heat stress their derivatives (Wildt et al. 2003), longer chain length alde- hydes are often found in plant emissions, including emissions Several plant species emit GDP-pathway compounds, mainly from Brassica rapa var. rapa (Taveira et al. 2009), tomato (So- monoterpenes, constitutively. Constitutive monoterpene syn- lanum lycopersicum)(Wanget al. 2001) and hybrid poplar thesis occurs in plastids where the terminal enzymes and mono- (Populus simonii x Populus pyramidalis)(Huet al. 2011), and terpene synthases are located (Tholl 2006; Chen et al. 2011; there is evidence that activation of LOX-pathway is responsi- Rajabi Memari et al. 2013). Constitutive monoterpene emis- ble for the emissions of all these aliphatic aldehydes (Hu et al. sions either come from specialized storage tissues or from im- 2009; Hu et al. 2011). mediate de novo synthesis (Grote et al. 2013). In the latter In addition to the low-level emissions of LOX volatiles in case, the emissions are strongly related to foliage photosyn- non-stressed conditions, a major burst of LOX volatiles upon thetic characteristics, and thus, reduction in foliage photosyn- severe stress constitutes a key ubiquitous stress response thetic rate upon heat stress typically also leads to reduction in (Matsui 2006; Copolovici et al. 2012). Multiple LOXs are con- constitutive monoterpene emissions (Loreto et al. 1998; stitutively active in leaves, and thus, the release of volatile Peñuelas & Llusià 2002; Kleist et al. 2012). In B. nigra, α-pinene LOX-compounds occurs rapidly as soon as the substrate, poly- and 3-carene followed by limonene were the main monoter- unsaturated fatty acids, becomes available because of mem- penes emitted under moderate temperatures in heat stress brane lesions (Feussner & Wasternack 2002; Liavonchanka & treatments and through the temperature response curve mea- Feussner 2006; Andreou & Feussner 2009). Accordingly, surements (Table 1). Through the temperature response curve ©2016JohnWiley&SonsLtd, Plant, Cell and Environment, 39,2027–2042 2038 K. Kask et al. measurements, the rates of total monoterpene emission and conditions and protein cofactors (Ahuja et al. 2009; Borgen net assimilation were positively correlated, and the emissions et al. 2012). In B. nigra, we observed eight different glucosino- decreased with increasing temperature parallel to photosynthe- late breakdown products (Table 1). As with the emissions of sis (Figs 2a & 3b), suggesting that these emissions resulted from LOX volatiles, emissions of glucosinolate breakdown products de novo synthesis. was enhanced at 40–44 °C in long-term stress and at 45–50 °C in However, sustained heat stress can result in the induction of heat shock treatments (Figs 3c & 8b). Moreover, in long-term monoterpene synthesis (Staudt & Bertin 1998), although not stress treatments, the emissions of LOX volatiles and glucosin- always (Kleist et al. 2012). These induced monoterpene emis- olate breakdown products were correlated over temperatures sions typically consist of different monoterpenes than consti- of 25 to 40 °C (Fig. 5). Similar elicitation of LOX volatiles tutive emissions, reflecting expression of new terpene and volatile glucosinolate products suggests that both reflect synthases (Staudt & Bertin 1998; Niinemets et al. 2010a,b; the propagation of lesions with increasing the severity of heat Copolovici & Niinemets 2016). In B. nigra, the total emission stress. rate of monoterpenes was not affected by long-term heat However, long-term and heat shock stresses importantly dif- treatment, but the emission rates of GDP volatiles correlated fered in the quantitative relationship between LOX volatiles with the emissions of LOX-compounds (Fig. 6a). These corre- and glucosinolate breakdown products (Figs 3a, c & 8b). In lations differed for different treatment temperature ranges, particular, long-term heat stress led to a much stronger elicita- suggesting that the constitutive plant defense was gradually tion of glucosinolate volatiles than the heat shock stress and at replaced by induced plant defense as the treatment tempera- the highest long-term heat treatment temperature of 44 °C, ture raised. glucosinolate volatile production exceeded LOX volatile Heat shock treatment was associated with a significant in- production (Figs 3a,c & 8b). This evidence suggests that while crease of monoterpene emissions (Figs 3b & 8b) as has been the release of glucosinolate volatiles upon heat shock reflects observed in tomato (S. lycopersicum)(Copoloviciet al. 2012), a release of myrosinases upon disruption of plant cells, a but the mechanism of this increase is unclear. Enhanced sub- certain induction process is activated upon long-term heat strate availability for monoterpene synthases due to disruption stress. In fact, there is evidence that myrosinase expression of other metabolic pathways consuming isopentenyl diphos- can be enhanced by different biotic and abiotic stresses (Jost phate and dimethylallyl diphosphate such as carotenoid syn- et al. 2005; Pan et al. 2014; del Carmen Martinez-Ballesta & thesis could provide an explanation for the increase of Carvajal 2015). In addition, reactive oxygen species such as monoterpene emission. It can also reflect a certain storage ca- H2O2 can directly enhance myrosinase expression (Pan et al. pacity of monoterpenoids in idioblasts, also called the myrosin 2014). As heat stress leads to a major burst of H2O2 (Hüve cells or ‘mustard oil bombs’ (Ahuja et al. 2009; Borgen et al. et al. 2011), heat stress dependent enhancement of myrosinase 2012), or non-specific storage of monoterpenes in cellular activity is likely. On the other hand, there is still limited infor- membranes as is common in constitutive de novo monoterpene mation of tissue-specific expression of different isoforms of emitters (Niinemets & Reichstein 2002; Niinemets et al. 2010b). myrosinases as well as stress effects on the synthesis of gluco- Thus, the release of these compounds, especially the release of sinolates. For example, methanethiol, that has been previously α-pinene (Table 1) upon heat shock can occur due to cellular observed in Brassica upon tissue damage (Tulio et al. 2002; van damage. A positive correlation between LOX-compounds Dam et al. 2012) was mostly detected in long-term stress exper- and monoterpene emissions through the heat shock treatments iment, suggesting a certain reprogramming of glucosinolate (Fig. 6b) suggests that this is a plausible explanation, although synthesis. Furthermore, given the spatial separation of the correlation collapsed at 50 °C where the increase in LOX myrosinases and glucosinolates, any structural or physiological emissions vastly exceeded that in monoterpene emission. change that reduces the degree of separation, for example, ex- pression of a different myrosinase closer to the site of synthesis of glucosinolates or vice versa is also expected to enhance the Release of specialized brassicoid volatiles as a release of glucosinolate breakdown products. major trait differentiating among heat stress Clearly, the release of glucosinolate volatiles constitutes a treatments stress marker in Brassicaceae, but there is also evidence that glucosinolate volatiles may also play a key signalling role. In Synthesis of glucosinolates and formation of their volatile toxic particular, exposure to allyl isothiocyanate has been shown to hydrolysis products by myrosinases constitute the characteristic enhance thermotolerance of A. thaliana (Hara et al. 2013). Ap- defence system in Brassicales (Halkier & Du 1997; Raybould plication of allyl isothiocyanate has also been shown to lead to & Moyes 2001; Wang et al. 2011). Formation of glucosinolates reactive oxygen species formation and activation of a signalling primarily occurs in vascular tissues (Li et al. 2011), while cascade leading to the closure of stomata (Khokon et al. 2011; myrosinases are stored in myrosin cells diffusely distributed Hossain et al. 2013). In our study, stomatal conductance and through plant tissues (Kelly et al. 1998; Burow et al. 2007; Zhao glucosinolate release were correlated through the long-term et al. 2008; Misra et al. 2015). Thus, the release of myrosinases heat treatment (Fig. 7b), however, given that stomatal closure upon damage of myrosin cells is the first step required for the also occurred through temperature response curves where glu- formation of glucosinolate volatiles (Winde & Wittstock cosinolate volatile release was minimal (Fig. 3c), the correla- 2011), whereas the blend of volatiles released depends on the tion in Fig. 7b likely is not causal, but part of the heat stress mixture of structurally different glucosinolates, reaction syndrome. © 2016 John Wiley & Sons Ltd, Plant, Cell and Environment, 39,2027–2042 Responses of Brassica nigra to heat stress 2039

Heat stress effects on benzenoids and that did not occur upon short severe stress. As the result, differ- geranylgeranyl diphosphate pathway volatiles ent types of heat stress, mild, chronic and shock stress, are asso- ciated with different volatile fingerprints (Fig. 4). These In addition to LOX-pathway, GDP-pathway and glucosinolate different volatile blends could play important roles in heat- volatiles, heat treatments were associated with differences in elicited signalling responses as well as in multitrophic interac- emissions of two other volatile compound classes. The only tions in natural stressful environments (Hopkins et al. 2009; benzenoid, methyl salicylate (MeSA), was detected in long- Copolovici et al. 2014). Further work is needed to gain insight term stress treatment at temperatures 30, 40 and 44 °C into the role of induction of glucosinolate volatiles in heat resis- (Table 1). MeSA is a common plant stress volatile, activating tance and into how different types of heat stress affect plant– multiple biochemical pathways upon biotic and abiotic stresses insect interactions in Brassicaceae. (Arimura et al. 2005; Zhao et al. 2010), and its release has been observed in several cases upon long-term exposure to moder- ately high temperatures (Karl et al. 2008; Kleist et al. 2012). ACKNOWLEDGEMENTS MeSA can be de novo synthesized upon stress, but it can also The study was funded by the European Science Foundation be released from a glycosidically bonded form (Blažević & EUROCORES programme EuroVOL (project A-BIO- Mastelić 2009). Given that no MeSA release was observed VOC, T11070PKTF), the Estonian Ministry of Science and Ed- upon heat shock treatment, the release of MeSA upon long- ucation (IUT 8-3), and the European Research Council (ad- term stress suggests that it was de novo synthesized. vanced grant 322603, SIP-VOL+). We thank Dr E. Poelman In the case of geranylgeranyl diphosphate (GGDP) pathway (Department of Entomology, Wageningen University, the volatiles, we observed emissions of geranyl acetone and 6- Netherlands) for providing the seeds of B. nigra. methyl-5-hepten-2-one that are suggested to result from oxida- tive cleavage of carotenoids (Buttery et al. 1988; Goff & Klee 2006; Tieman et al. 2006). Both volatiles, geranyl acetone CONFLICT OF INTEREST (Taveira et al. 2009; Truong et al. 2014) and 6-methyl-5- The authors have no conflict of interest to declare. hepten-2-one (Geervliet et al. 1997) have been observed in emissions from Brassicaceae species. In B. nigra, only a reduc- tion of emission with raising temperature was observed in tem- REFERENCES perature response curve measurements similarly to changes in GDP-pathway volatiles (Fig. 3d). Analogously, geranyl ace- Agneta R., Lelario F., De Maria S., Möllers C., Bufo S.A. & Rivelli A.R. (2014) Glucosinolate profile and distribution among plant tissues and phenological tone emissions decreased in A. thaliana with increasing temper- stages of field-grown horseradish. Phytochemistry 106,178–187. ature (Truong et al. 2014). As both GDP-pathway and GGDP- Agrawal A.A. & Kurashige N.S. (2003) A role for isothiocyanates in plant resis- pathway are confined to plastids (Rajabi Memari et al. 2013), tance against the specialist herbivore Pieris rapae. Journal of Chemical Ecology 29,1403–1415. the release of these volatiles might be associated with turnover Ahuja I., Rohloff J. & Bones A.M. (2009) Defence mechanisms of Brassicaceae: of carotenoids that occurs as part of everyday plant metabolism implications for plant-insect interactions and potential for integrated pest man- (Beisel et al. 2010). If so, inhibition of the release of GGDP- agement. A review. Agronomy for Sustainable Development 30,311–348. volatiles with raising temperatures might indicate reversible in- Ameye M., Wertin T.M., Bauweraerts I., McGuire M.A., Teskey R.O. & Steppe K. (2012) The effect of induced heat waves on Pinus taeda and Quercus rubra hibition of carotenoid synthesis. seedlings in ambient and elevated CO2 atmospheres. The New Phytologist 196, 448–461. Andreou A. & Feussner I. (2009) Lipoxygenases – structure and reaction mech- CONCLUSIONS anism. Phytochemistry 70,1504–1510. Arimura G., Kost C. & Boland W. (2005) Herbivore-induced, indirect plant de- Overall, the results indicated that different types of heat treat- fences. Biochimica et Biophysica Acta-Molecular and Cell Biology of Lipids – ment are associated with major variation in photosynthetic and 1734,91 111. Arimura G., Matsui K. & Takabayashi K. (2009) Chemical and molecular ecology volatile responses in B. nigra (Figs 2–4 & 8). Temperature re- of herbivore-induced plant volatiles: proximate factors and their ultimate func- sponse curve measurements constituted a mild, physiological tions. Plant and Cell Physiology 50,911–923. stress that led to reductions in constitutively synthesized vola- Barth C. & Jander G. (2006) Arabidopsis myrosinases TGG1 and TGG2 have re- dundant function in glucosinolate breakdown and insect defense. The Plant tiles associated with immediate photosynthetic metabolism. Journal 46,549–562. Both long-term and heat shock stress resulted in elicitation of Behnke K., Ehlting B., Teuber M., Bauerfeind M., Louis S., Hänsch R. & Schnitz- lipoxygenase and glucosinolate volatiles once the threshold ler J.-P. (2007) Transgenic, non-isoprene emitting poplars don’t like it hot. The – heat dose was achieved. However, these two types of stresses Plant Journal 51,485 499. Beisel K.G., Jahnke S., Hofmann D., Köppchen S., Schurr U. & Matsubara S. primarily differed in the extent to which glucosinolate volatile (2010) Continuous turnover of carotenes and chlorophyll a in mature leaves 14 emission was induced relative to LOX product release (Fig. 8). of Arabidopsis revealed by CO2 pulse-chase labeling. Plant Physiology 152, In particular, long-term heat stress was associated with much 2188–2199. Bengtsson M., Bäckman A.C., Liblikas I., Ramirez M.I., Borg-Karlson A.K., stronger elicitation of glucosinolate emissions than the heat Ansebo L. & Witzgall P. (2001) Plant odor analysis of apple: antennal response shock response. In addition, methyl salicylate emissions were of codling moth females to apple volatiles during phenological development. only induced by long-term heat stress. Although both long- Journal of Agricultural and Food Chemistry 49,3736–3741. … term and short-term shock stress resulted in major raises of Betz G.A., Gerstner E., Stich S., Winkler B., Welzl G., Kremmer E., Ernst D. (2009) Ozone affects shikimate pathway genes and secondary metabolites in stress volatile emissions, sustained moderate heat stress re- saplings of European beech (Fagus sylvatica L.) grown under greenhouse con- sulted in the engagement of induced metabolic defense systems ditions. Trees 23,539–553. ©2016JohnWiley&SonsLtd, Plant, Cell and Environment, 39,2027–2042 2040 K. Kask et al.

Bidart-Bouzat M.G. & Imeh-Nathaniel A. (2008) Global change effects on plant Dicke M. & van Loon J.J.A. (2000) Multitrophic effects of herbivore-induced plant chemical defenses against insect herbivores. Journal of Integrative Plant Biol- volatiles in an evolutionary context. Entomologia Experimentalis et Applicata 97, ogy 50,1339–1354. 237–249. Bilger H.W., Schreiber U. & Lange O.L. (1984) Determination of leaf heat resis- Duke J.A. (1983) Handbook of energy crops. In: NewCROP.PurdueUniversity, tance: comparative investigation of chlorophyll fluorescence changes and tissue Department of Horticulture and Landscape Architecture, 625 Agriculture Mall necrosis methods. Oecologia 63,256–262. Drive West Lafayette, IN 47907-2010. Blažević I. & Mastelić J. (2009) Glucosinolate degradation products and other Fahey J.W., Zalcmann A.T. & Talalay P. (2001) The chemical diversity and distribu- bound and free volatiles in the leaves and roots of radish (Raphanus sativus tion of glucosinolates and isothiocyanates among plants. Phytochemistry 56,5–51. L.) Food Chemistry 113,96–102. Farré-Armengol G., Filella I., Llusià J., Niinemets Ü. & Peñuelas J. (2014) Bones A.M. & Rossiter J.T. (2006) The enzymic and chemically induced decom- Changes in floral bouquets from compound-specific responses to increasing position of glucosinolates. Phytochemistry 67,1053–1067. temperatures. Global Change Biology 20,3660–3669. Borgen B.H., Ahuja I., Thangstad O.P., Honne B.I., Rohloff J., Rossiter J.T. & Fatouros N.E., Lucas-Barbosa D., Weldegergis B.T., Pashalidou F.G., van Loon J. Bones A.M. (2012) ‘Myrosin cells’ are not a prerequisite for aphid feeding on J.A., Dicke M. & Huigens M.E. (2012) Plant volatiles induced by herbivore egg oilseed rape (Brassica napus) but affect host plant preferences. Plant Biology deposition affect insects of different trophic levels. PloS One 7, e43607. 14,894–904. Feussner I. & Wasternack C. (2002) The lipoxygenase pathway. Annual Review of Brilli F., Hörtnagl L., Bamberger I., Schnitzhofer R., Ruuskanen T.M., Hansel A., Plant Biology 53,275–297. Loreto F. & Wohlfahrt G. (2012) Qualitative and quantitative characterization Geervliet J.B.F., Posthumus M.A., Vet L.E.M. & Dicke M. (1997) Comparative of volatile organic compound emissions from cut grass. Environmental Science analysis of headspace volatiles from different caterpillar-infested or uninfested & Technology 46,3859–3665. food plants of Pieris species. JournalofChemicalEcology23,2935–2954. Bruinsma M., van Dam N.M., van Loon J.J.A. & Dicke M. (2007) Jasmonic acid- Goff S.A. & Klee H.J. (2006) Plant volatile compounds: sensory cues for health induced changes in Brassica oleracea affect oviposition preference of two spe- and nutritional value? Science 311,815–819. cialist herbivores. Journal of Chemical Ecology 33,655–668. Gomaa N.H., Al Sherif E.A., Hegazy A.K. & Hassan M.O. (2012) Floristic diver- Bruinsma M., Ijdema H., van Loon J.J.A. & Dicke M. (2008) Differential effects of sity and vegetation analysis of Brassica nigra (L.) Koch communities. Egyptian jasmonic acid treatment of Brassica nigra on the attraction of pollinators, para- Journal of Biology 14,63–72. sitoids, and butterflies. Entomologia Experimentalis et Applicata 128,109–116. Grote R., Monson R.K. & Niinemets Ü. (2013) Leaf-level models of constitutive Burow M., Rice M., Hause B., Gershenzon J. & Wittstock U. (2007) Cell- and and stress-driven volatile organic compound emissions. In Biology, Controls tissue-specific localization and regulation of the epithiospecifier protein in and Models of Tree Volatile Organic Compound Emissions (eds Niinemets Ü. Arabidopsis thaliana. Plant Molecular Biology 64,173–185. & Monson R.K.), pp. 315–355. Springer, Berlin. Buttery R.G., Guadagni D.G., Ling L.C., Seifert R.M. & Lipton W. (1976) Addi- Halkier B.A. & Du L.C. (1997) The biosynthesis of glucosinolates. Trends in tional volatile components of cabbage, broccoli, and cauliflower. Journal of Ag- Plant Science 2,425–431. ricultural and Food Chemistry 24,829–832. Hara M., Harazaki A. & Tabata K. (2013) Administration of isothiocyanates en- Buttery R.G., Teranishi R., Ling L.C., Flath R.A. & Stern D.J. (1988) Quantita- hances heat tolerance in Arabidopsis thaliana. Plant Growth Regulation 69, tive studies on origins of fresh tomato aroma volatiles. Journal of Agricultural 71–77. and Food Chemistry 36,1247–1250. Hartikainen K., Nerg A.-M., Kivimäenpää M., Kontunen-Soppela S., Mäenpää Ceuppens B., Ameye M., Van Langenhove H., Roldan-Ruiz I. & Smagghe G. M., Oksanen E., Rousi M. & Holopainen T. (2009) Emissions of volatile or- (2015) Characterization of volatiles in strawberry varieties ’Elsanta’ and ’So- ganic compounds and leaf structural characteristics of European aspen nata’ and their effect on bumblebee flower visiting. Arthropod-Plant Interac- (Populus tremula) grown under elevated ozone and temperature. Tree Physiol- tions 9,281–287. ogy 29, 1163–1173. Chen F., Tholl D., Bohlmann J. & Pichersky E. (2011) The family of terpene Havaux M. (1993) Characterization of thermal damage to the photosynthetic synthases in plants: a mid-size family of genes for specialized metabolism that electron transport system in potato leaves. Plant Science 94,19–33. is highly diversified throughout the kingdom. The Plant Journal 66,212–229. Hopkins R.J., Birch A.N.E., Griffiths D.W., Baur R., Städler E. & McKinlay R.G. Copolovici L., Kännaste A. & Niinemets Ü. (2009) Gas chromatography-mass (1997) Leaf surface compounds and oviposition preference of turnip root fly spectrometry method for determination of monoterpene and sesquiterpene Delia floralis: The role of glucosinolate and nonglucosinolate compounds. emissions from stressed plants. Studia Universitatis Babes-Bolyai, Chemia 54, Journal of Chemical Ecology 23,629–643. 329–339. Hopkins R.J., van Dam N.M. & van Loon J.J.A. (2009) Role of glucosinolates in Copolovici L., Kännaste A., Pazouki L. & Niinemets Ü. (2012) Emissions of green insect-plant relationships and multitrophic interactions. Annual Review of En- leaf volatiles and terpenoids from Solanum lycopersicum are quantitatively re- tomology 54,57–83. lated to the severity of cold and heat shock treatments. Journal of Plant Phys- Hossain M.S., Ye W., Hossain M.A., Okuma E., Uraji M., Nakamura Y.,Mori I.C. iology 169,664–672. & Murata Y. (2013) Glucosinolate degradation products, isothiocyanates, ni- Copolovici L., Kännaste A., Remmel T. & Niinemets Ü. (2014) Volatile organic triles, and thiocyanates, induce stomatal closure accompanied by peroxidase- compound emissions from Alnus glutinosa under interacting drought and her- mediated reactive oxygen species production in Arabidopsis thaliana. Biosci- bivory stresses. Environmental and Experimental Botany 100,55–63. ence, Biotechnology, and Biochemistry 77,977–983.

Copolovici L. & Niinemets Ü. (2010) Flooding induced emissions of volatile sig- Hu Z.-H., Leng P.-S., Shen Y.-B. & Wang W.-H. (2011) Emissions of saturated C6– nalling compounds in three tree species with differing waterlogging tolerance. C10 aldehydes from poplar (Populus simonii × P. pyramidalis ‘Opera 8277’)cut- Plant, Cell and Environment 33,1582–1594. tings at different levels of light intensity. Journal of Forestry Research 22,223–238.

Copolovici L. & Niinemets Ü. (2016) Environmental impacts on plant volatile Hu Z.-H., Shen Y.-B. & Su X.-H. (2009) Saturated aldehydes C6-C10 emitted from emission. In Deciphering Chemical Language of Plant Communication (eds ashleaf maple (Acer negundo L.) leaves at different levels of light intensity, O2, Blande J. & Glinwood R.), Springer, Berlin. and CO2. Journal of Plant Biology 52,289–298. Crespo E., Hordijk C.A., de Graaf R.M., Samudrala D., Cristescu S.M., Harren F. Hu Z., Zhang H., Leng P., Zhao J., Wang W. & Wang S. (2013) The emission of J.M. & van Dam N.M. (2012) On-line detection of root-induced volatiles in floral scent from Lilium ’siberia’ in response to light intensity and temperature. Brassica nigra plants infested with Delia radicum L. root fly larvae. Phytochem- Acta Physiologiae Plantarum 35,1691–1700. istry 84,68–77. Hüve K., Bichele I., Rasulov B. & Niinemets Ü. (2011) When it is too hot for pho- Cui L.J., Li J.L., Fan Y.M., Xu S. & Zhang Z. (2006) High temperature effects on tosynthesis: heat-induced instability of photosynthesis in relation to respiratory

photosynthesis, PSII functionality and antioxidant activity of two Festuca burst, cell permeability changes and H2O2 formation. Plant, Cell and Environ- arundinacea cultivars with different heat susceptibility. Botanical Studies 47,61–69. ment 34,113–126. de Gouw J.A., Howard C.J., Custer T.G. & Fall R. (1999) Emissions of volatile Hüve K., Bichele I., Tobias M. & Niinemets Ü. (2006) Heat sensitivity of photo- organic compounds from cut grass and clover are enhanced during the drying synthetic electron transport varies during the day due to changes in sugars process. Geophysical Research Letters 26,811–814. and osmotic potential. Plant, Cell and Environment 29,212–228. de Vos M., Kriksunov K.L. & Jander G. (2008) Indole-3-acetonitrile production Ishida M., Hara M., Fukino N., Kakizaki T. & Morimitsu Y. (2014) Glucosinolate from indole glucosinolates deters oviposition by Pieris rapae. Plant Physiology metabolism, functionality and breeding for the improvement of Brassicaceae 146,916–926. vegetables. Breeding Science 64,48–59. del Carmen Martinez-Ballesta M. & Carvajal M. (2015) Myrosinase in Brassica- Jansen R.M.C., Miebach M., Kleist E., van Henten E.J. & Wildt J. (2009) Release ceae: the most important issue for glucosinolate turnover and food quality. Phy- of lipoxygenase products and monoterpenes by tomato plants as an indicator of tochemistry Reviews 14,1045–1051. Botrytis cinerea-induced stress. Plant Biology 11,859–868.

© 2016 John Wiley & Sons Ltd, Plant, Cell and Environment, 39,2027–2042 Responses of Brassica nigra to heat stress 2041

Jost R., Altschmied L., Bloem E., Bogs J., Gershenzon J., Hänel U., … Hell R. Niinemets Ü., Kuhn U., Harley P.C., Staudt M., Arneth A., Cescatti A., … (2005) Expression profiling of metabolic genes in response to methyl jasmonate Peñuelas J. (2011) Estimations of isoprenoid emission capacity from enclosure reveals regulation of genes of primary and secondary sulfur-related pathways studies: measurements, data processing, quality and standardized measure- in Arabidopsis thaliana. Photosynthesis Research 86,491–508. ment protocols. Biogeosciences 8,2209–2246. Kadir S., Von Weihe M. & Al-Khatib K. (2007) Photochemical efficiency and re- Niinemets Ü., Monson R.K., Arneth A., Ciccioli P., Kesselmeier J., Kuhn U., … covery of photosystem II in grapes after exposure to sudden and gradual heat Staudt M. (2010b) The leaf-level emission factor of volatile isoprenoids: stress. Journal of the American Society for Horticultural Science 132,764–769. caveats, model algorithms, response shapes and scaling. Biogeosciences 7, Kännaste A., Copolovici L. & Niinemets Ü. (2014) Gas chromatography-mass 1809–1832. spectrometry method for determination of biogenic volatile organic com- Niinemets Ü. & Reichstein M. (2002) A model analysis of the effects of nonspe- pounds emitted by plants. In: Plant Isoprenoids: Methods and Protocols (ed cific monoterpenoid storage in leaf tissues on emission kinetics and composi- M. Rodríguez-Concepción), pp. 161-–169. Springer Science + Business Media, tion in Mediterranean sclerophyllous Quercus species. Global New York. Biogeochemical Cycles 16, 1110. Karban R. (2011) The ecology and evolution of induced resistance against herbi- Olivier C., Vaughn S.F., Mizubuti E.S.G. & Loria R. (1999) Variation in allyl iso- vores. Functional Ecology 25,339–347. thiocyanate production within Brassica species and correlation with fungicidal Karl T., Guenther A., Turnipseed A., Patton E.G. & Jardine K. (2008) Chemical activity. Journal of Chemical Ecology 25,2687–2701. sensing of plant stress at the ecosystem scale. Biogeosciences 5,1287–1294. Pan Y., Xu Y.-Y., Zhu X.-W., Liu Z., Gong Y.-Q., Xu L., Gong M.-Y. & Liu L.-W. Kelly P.J., Bones A. & Rossiter J.T. (1998) Sub-cellular immunolocalization of the (2014) Molecular characterization and expression profiles of myrosinase gene glucosinolate sinigrin in seedlings of Brassica juncea. Planta 206,370–377. (RsMyr2)inradish(Raphanus sativus L.) Journal of Integrative Agriculture Khaling E., Papazian S., Poelman E.H., Holopainen J.K., Albrectsen B.R. & 13,1877–1888. Blande J.D. (2015) Ozone affects growth and development of Pieris brassicae Pashalidou F.G., Fatouros N.E., van Loon J.J.A., Dicke M. & Gols R. (2015) on the wild host plant Brassica nigra. Environmental Pollution 199,119–129. Plant-mediated effects of butterfly egg deposition on subsequent caterpillar Khokon M.A.R., Jahan M.S., Rahman T., Hossain M.A., Muroyama D., Minami and pupal development, across different species of wild Brassicaceae. Ecolog- I., … Murata Y.(2011) Allyl isothiocyanate (AITC) induces stomatal closure in ical Entomology 40,440–450. Arabidopsis. Plant, Cell and Environment 34,1900–1906. Peñuelas J. & Llusià J. (2002) Linking photorespiration, monoterpenes and ther- Kleist E., Mentel T.F., Andres S., Bohne A., Folkers A., Kiendler-Scharr A., … motolerance in Quercus. The New Phytologist 155,227–237. Wildt J. (2012) Irreversible impacts of heat on the emissions of monoterpenes, Possell M. & Loreto F. (2013) The role of volatile organic compounds in plant re- sesquiterpenes, phenolic BVOC and green leaf volatiles from several tree spe- sistance to abiotic stresses: responses and mechanisms. In Biology, Controls cies. Biogeosciences 9,5111–5123. and Models of Tree Volatile Organic Compound Emissions (eds Niinemets Ü. Kos M., Houshyani B., Achhami B.B., Wietsma R., Gols R., Weldegergis B.T., … & Monson R.K.), pp. 209–235. Springer, Berlin. van Loon J.J.A. (2012) Herbivore-mediated effects of glucosinolates on differ- Rajabi Memari H., Pazouki L. & Niinemets Ü. (2013) The biochemistry and mo- ent natural enemies of a specialist aphid. Journal of Chemical Ecology 38, lecular biology of volatile messengers in trees. In Biology, Controls and Models 100–115. of Tree Volatile Organic Compound Emissions (eds Niinemets Ü. & Monson R. Li J., Kristiansen K.A., Hansen B.G. & Halkier B.A. (2011) Cellular and subcel- K.), pp. 47–93. Springer, Berlin. lular localization of flavinmonooxygenases involved in glucosinolate biosynthe- Rajamurugan R., Suyavaran A., Selvaganabathy N., Ramamurthy C.H., Reddy sis. Journal of Experimental Botany 62,1337–1346. G.P., Sujatha V. & Thirunavukkarasu C. (2012) Brassica nigra plays a remedy Liavonchanka A. & Feussner N. (2006) Lipoxygenases: occurrence, functions role in hepatic and renal damage. Pharmaceutical Biology 50,1488–1497. and catalysis. Journal of Plant Physiology 163,348–357. Raybould A.F. & Moyes C.L. (2001) The ecological genetics of aliphatic glucosin- Loreto F., Barta C., Brilli F. & Nogues I. (2006) On the induction of volatile or- olates. Heredity 87,383–391. ganic compound emissions by plants as consequence of wounding or fluctua- Redovniković I.R., Glivetić T., Delonga K. & Vorkapić-Furač J. (2008) Glucosin- tions of light and temperature. Plant, Cell and Environment 29,1820–1828. olates and their potential role in plant. Periodicum Biologorum 110,297–309. Loreto F., Förster A., Dürr M., Csiky O. & Seufert G. (1998) On the monoterpene Sharkey T.D. (2005) Effects of moderate heat stress on photosynthesis: impor- emission under heat stress and on the increased thermotolerance of leaves of tance of thylakoid reactions, rubisco deactivation, reactive oxygen species, Quercus ilex L. fumigated with selected monoterpenes. Plant, Cell and Envi- and thermotolerance provided by isoprene. Plant, Cell and Environment 28, ronment 21,101–107. 269–277. Lu C.M. & Zhang J.H. (2000) Heat-induced multiple effects on PSII in wheat Shen J., Tieman D., Jones J.B., Taylor M.G., Schmelz E., Huffaker A., … Klee H. plants. Journal of Plant Physiology 156,259–265. J. (2014) A 13-lipoxygenase, TomloxC, is essential for synthesis of C5 flavour Ludwig-Müller J., Krishna P. & Forreiter C. (2000) A glucosinolate mutant of volatiles in tomato. Journal of Experimental Botany 65,519–428. Arabidopsis is thermosensitive and defective in cytosolic Hsp90 expression af- Shinohara T. & Leskovar D.I. (2014) Effects of ABA, antitranspirants, heat and ter heat stress. Plant Physiology 123,949–958. drought stress on plant growth, physiology and water status of artichoke trans- Maccarrone M., Veldink G.A. & Vliegenthart J.F. (1992) Thermal injury and plants. Scientia Horticulturae 165,225–234. ozone stress affect soybean lipoxygenases expression. FEBS Letters 309, Singsaas E.L. & Sharkey T.D. (1998) The regulation of isoprene emission re- 225–230. sponses to rapid leaf temperature fluctuations. Plant, Cell and Environment Maffei M.E. (2010) Sites of synthesis, biochemistry and functional role of plant 21, 1181–1188. volatiles. South African Journal of Botany 76,612–631. Sønderby I.E., Geu-Flores F. & Halkier B.A. (2010) Biosynthesis of Matsui K. (2006) Green leaf volatiles: hydroperoxide lyase pathway of oxylipin glucosinolates-gene discovery and beyond. Trends in Plant Scince 15,283–290. metabolism. Current Opinion in Plant Biology 9,274–280. Staudt M. & Bertin N. (1998) Light and temperature dependence of the emission Matsui K., Sugimoto K., Mano J.i., Ozawa R. & Takabayashi J. (2012) Differen- of cyclic and acyclic monoterpenes from holm oak (Quercus ilex L.) leaves. tial metabolisms of green leaf volatiles in injured and intact parts of a wounded Plant, Cell and Environment 21,385–395. leaf meet distinct ecophysiological requirements. PloS One 7, e36433. Sun Z., Copolovici L. & Niinemets Ü. (2012) Can the capacity for isoprene emis- Misra B.B., Acharya B.R., Granot D., Assmann S.M. & Chen S. (2015) The sion acclimate to environmental modifications during autumn senescence in guard cell metabolome: functions in stomatal movement and global food secu- temperate deciduous tree species Populus tremula? Journal of Plant Research rity. Frontiers in Plant Science 6,1–13. 125,263–274. Najar-Rodriguez A.J., Friedli M., Klaiber J. & Dorn S. (2015) Aphid-deprivation Sung D.Y., Kaplan F., Lee K.J. & Guy C.L. (2003) Acquired tolerance to temper- from Brassica plants results in increased isothiocyanate release and parasitoid ature extremes. Trends in Plant Science 8,179–187. attraction. Chemoecology 25,303–311. Taveira M., Fernandes F., de Pinho P.G., Andrade P.B., Pereira J.A. & Valentão P. Niinemets Ü. (2010a) Mild versus severe stress and BVOCs: thresholds, priming (2009) Evolution of Brassica rapa var. rapa L. volatile composition by HS- and consequences. Trends in Plant Science 15,145–153. SPME and GC/IT-MS. Microchemical Journal 93,140–146. Niinemets Ü. (2010b) Responses of forest trees to single and multiple environ- Tholl D. (2006) Terpene synthases and the regulation, diversity and biological mental stresses from seedlings to mature plants: past stress history, stress inter- roles of terpene metabolism. Current Opinion in Plant Biology 9,297–304. actions, tolerance and acclimation. Forest Ecology and Management 260, Tieman D.M., Zeigler M., Schmelz E.A., Taylor M.G., Bliss P., Kirst M. & Klee 1623–1639. H.J. (2006) Identification of loci affecting flavour volatile emissions in tomato Niinemets Ü., Arneth A., Kuhn U., Monson R.K., Peñuelas J. & Staudt M. fruits. Journal of Experimental Botany 57,887–896. (2010a) The emission factor of volatile isoprenoids: stress, acclimation, and de- Truong D.-H., Delory B.M., Brostaux Y., Heuskin S., Delaplace P., Francis F. & velopmental responses. Biogeosciences 7,2203–2223. Lognay G. (2014) Plutella xylostella (L.) infestations at varying temperatures

©2016JohnWiley&SonsLtd, Plant, Cell and Environment, 39,2027–2042 2042 K. Kask et al.

induce the emission of specific volatile blends by Arabidopsis thaliana (L.) Wang H., Wu J., Sun S., Liu B., Cheng F., Sun R. & Wang X. (2011) Glucosinolate Heynh. Plant Signaling & Behavior 9,1–11. biosynthetic genes in Brassica rapa. Gene 487,135–142. Tulio A.Z., Yamanaka H., Ueda Y. & Imahori Y. (2002) Formation of Way D.A., Schnitzler J.-P., Monson R.K. & Jackson R.B. (2011) Enhanced methanethiol and dimethyl disulfide in crushed tissues of broccoli florets and isoprene-related tolerance of heat- and light-stressed photosynthesis at low,

their inhibition by freeze-thawing. Journal of Agricultural and Food Chemistry but not high, CO2 concentrations. Oecologia 166,273–282. 50,1502–1507. Wildt J., Kobel K., Schuh-Thomas G. & Heiden A.C. (2003) Emissions of oxy- Turk M.A. & Tawaha A.M. (2003) Allelopathic effect of black mustard (Brassica genated volatile organic compounds from plants – part II: Emissions of satu- nigra L.) on germination and growth of wild oat (Avena fatua L.) Crop Protec- rated aldehydes. Journal of Atmospheric Chemistry 45,173–196. tion 22,673–677. Winde I. & Wittstock U. (2011) Insect herbivore counteradaptations to the plant Usano-Alemany J., Palá-Paúl J. & Herráiz-Peñalver D. (2014) Temperature stress glucosinolate-myrosinase system. Phytochemistry 72,1566–1575. causes different profiles of volatile compounds in two chemotypes of Salvia Wittstock U. & Burow M. (2010) Glucosinolate breakdown in Arabidopsis: lavandulifolia Vahl. Biochemical Systematics and Ecology 54,166–171. mechanism, regulation and biological significance. The Arabidopsis Book 8, Vacca R.A., de Pinto M.C., Valenti D., Passarella S., Marra E. & De Gara L. 1–14. (2004) Production of reactive oxygen species, alteration of cytosolic ascorbate Wold S., Esbensen K. & Geladi P. (1987) Principal component analysis. peroxidase, and impairment of mitochondrial metabolism are early events in Chemometrics and Intelligent Laboratory Systems 2,37–52. heat shock-induced programmed cell death in tobacco bright-yellow 2 cells. Zhang R., Cruz J.A., Kramer D.M., Magallanes-Lundback M.E., DellaPenna D. Plant Physiology 134, 1100–1112. & Sharkey T.D. (2009) Moderate heat stress reduces the pH component of the van Dam N.M., Samudrala D., Harren F.J.M. & Cristescu S.M. (2012) Real-time transthylakoid proton motive force in light-adapted, intact tobacco leaves. analysis of sulfur-containing volatiles in Brassica plants infested with root- Plant, Cell and Environment 32,1538–1547. feeding Delia radicum larvae using proton-transfer reaction mass spectrometry. Zhang R. & Sharkey T.D. (2009) Photosynthetic electron transport and proton AoB 1–12. flux under moderate heat stress. Photosynthesis Research 100,29–43. Velikova V., Tsonev T., Barta C., Centritto M., Koleva D., Stefanova M., Busheva Zhao N., Guan J., Ferrer J.-L., Engle N., Chern M., Ronald P., Tschaplinski T.J. & M. & Loreto F. (2009) BVOC emissions, photosynthetic characteristics and Chen F. (2010) Biosynthesis and emission of insect-induced methyl salicylate changes in chloroplast ultrastructure of Platanus orientalis L. exposed to ele- and methyl benzoate from rice. Plant Physiology and Biochemistry 48,

vated CO2 and high temperature. Environmental Pollution 157,2629–2637. 279–289. von Caemmerer S. & Farquhar G.D. (1981) Some relationships between the bio- Zhao Z., Zhang W., Stanley B.A. & Assmann S.M. (2008) Functional proteomics chemistry of photosynthesis and the gas exchange of leaves. Planta 153, of Arabidopsis thaliana guard cells uncovers new stomatal signaling pathways. 376–387. Plant Cell 20,3210–3226. Wahid A., Gelani S., Ashraf M. & Foolad M.R. (2007) Heat tolerance in plants: an overview. Environmental and Experimental Botany 61,199–223. Received 11 March 2016; received in revised form 7 June 2016; accepted Wang C.L., Xing J.S., Chin C.K., Ho C.T. & Martin C.E. (2001) Modification of for publication 8 June 2016 fatty acids changes the flavor volatiles in tomato leaves. Phytochemistry 58, 227–232.

© 2016 John Wiley & Sons Ltd, Plant, Cell and Environment, 39,2027–2042 Science of the Total Environment 569–570 (2016) 489–495

Contents lists available at ScienceDirect

Science of the Total Environment

journal homepage: www.elsevier.com/locate/scitotenv

Induction of stress volatiles and changes in essential oil content and composition upon microwave exposure in the aromatic plant Ocimum basilicum

Ildikó Lung a, Maria-Loredana Soran a,⁎,OcsanaOpriş a,MihailRaduCătălin Truşcă a, Ülo Niinemets b, Lucian Copolovici b,c a National Institute for Research and Development of Isotopic and Molecular Technologies, 67-103 Donat Street, Cluj-Napoca 400293, Romania b Institute of Agricultural and Environmental Sciences, Estonian University of Life Sciences, 1 Kreutzwaldi Street, Tartu 51014, Estonia c Institute of Technical and Natural Sciences Research-Development of “Aurel Vlaicu” University, 2 Elena Drăgoi Street, Arad 310330, Romania

HIGHLIGHTS GRAPHICAL ABSTRACT

• Microwave irradiation represents a stress for the plants. • Microwave exposure leads to enhanced emissions of stress volatiles. • O. basilicum irradiation with microwaves increases the essential oil content. • Microwave pollution can constitute a threat to the environment.

article info abstract

Article history: Exposure to sustained low intensity microwaves can constitute a stress for the plants, but its effects on plant sec- Received 8 April 2016 ondary chemistry are poorly known. We studied the influence of GSM and WLAN-frequency microwaves on Received in revised form 17 June 2016 emissions of volatile organic compounds and content of essential oil in the aromatic plant Ocimum basilicum L. Accepted 19 June 2016 hypothesizing that microwave exposure leads to enhanced emissions of stress volatiles and overall greater in- Available online xxxx vestment in secondary compounds. Compared to the control plants, microwave irradiation led to decreased β α β α Editor: Elena Paoletti emissions of -pinene, -phellandrene, bornyl acetate, -myrcene, -caryophyllene and benzaldehyde, but in- creased emissions of eucalyptol, estragole, caryophyllene oxide, and α-bergamotene. The highest increase in Keywords: emission, 21 times greater compared to control, was observed for caryophyllene oxide. The irradiation resulted Constitutive emissions in increases in the essential oil content, except for the content of phytol which decreased by 41% in the case of Essential oils GSM-frequency, and 82% in the case of WLAN-frequency microwave irradiation. The strongest increase in re- Induced emissions sponse to WLAN irradiation, N17 times greater, was observed for hexadecane and octane contents. Comparisons Microwave irradiation of volatile compositions by multivariate analyses demonstrated a clear separation of different irradiance treat- Ocimum basilicum ments, and according to the changes in the volatile emissions, the WLAN-frequency irradiation represented a

⁎ Corresponding author. E-mail address: [email protected] (M.-L. Soran).

http://dx.doi.org/10.1016/j.scitotenv.2016.06.147 0048-9697/© 2016 Elsevier B.V. All rights reserved. 490 I. Lung et al. / Science of the Total Environment 569–570 (2016) 489–495

more severe stress than the GSM-frequency irradiation. Overall, these results demonstrating important modifications in the emission rates, essential oil content and composition indicate that microwave irradiation influences the quality of herbage of this economically important spice plant. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Lak, 2012) but given that plants play a key role as primary producers of food and oxygen, it is important to gain insight into their response Sweet basil (Ocimum basilicum L., Lamiaceae) is an economically im- to enhanced exposure to microwave-frequency irradiation. On the ef- portant aromatic annual herb that is cultivated and utilized throughout fect of GSM and UMTS, so far, only a limited number of studies have the world. Basil is a condimental plant, but it is also very well known in been carried out in plants. Plants of sweet basil (O. basilicum), celery folk medicine for its medicinal properties (Baratta et al., 1998; (Apium graveolens) and parsley (Petroselinum crispum) irradiated by Opalchenova and Obreshkova, 2003; Chiang et al., 2005). Essential oils GSM-frequency microwaves had greater essential oil (Soran et al., constitute a complex mixture of volatile and semi-volatile compounds 2014), and polyphenolics contents (Soran et al., 2016) than in the con- that determine the specific aroma of plants and the flavor of the condi- trol plants. Stan et al. (2014) further observed an increase in the content ment. The chemical composition and concentrations of individual com- of ascorbic acid in the leaves of A. graveolens and Anethum graveolens ponents are affected by plant developmental stage and/or by variations subjected to microwave irradiation. In several barley (Hordeum vulgare) in environmental and cultivation conditions (Jirovetz et al., 2003; Vina genotypes, microwave irradiation of seeds reduced plant chlorophyll and Murillo, 2003; Hassanpouraghdam et al., 2011; Sirousmehr et al., content (Creţescu et al., 2013). In mung bean (Vigna radiata), electro- 2014; Jiang et al., in press-a). The content of essential oil is often en- magnetic radiation at GSM-frequency (cellular phone) reduced seedling hanced by several environmental stresses as a defense response as length, dry weight, and the contents of proteins and carbohydrates. In well as the result of changes in the balance between carbon use in addition, the activities of stress marker enzymes, proteases, α-amylases, growth and secondary metabolite production (Letchamo et al., 1995; β-amylases, polyphenol oxidases, and peroxidases were enhanced in Munne-Bosch et al., 1999; Zabaras et al., 2002; Blanch et al., 2007; plants exposed to microwave irradiation (Sharma et al., 2010). This ev- Jiang et al., in press-a). However, severe environmental stresses, such idence collectively suggests that the plants were strongly influenced by as high and low temperature (Copolovici et al., 2012), drought stress due to the applied electromagnetic irradiation. In contrast, expo- (Copolovici et al., 2013; Timmusk et al., 2014), alkalinity, salinity, UV sure of corn (Zea mays) grains to microwaves resulted in improved stress and pathogen infection can overly reduce plant growth and pho- grain germination, leaves with greater midrib vascular bundle length, tosynthesis rates (Ramakrishna and Ravishankar, 2011) and ultimately thicker mesophyll, faster growth rate, higher contents of photosynthetic also the rate of production of secondary chemicals (Rhizopoulou and pigments, total soluble sugar and total carbohydrates (Khalafallah and Diamantoglou, 1991; Yani et al., 1993; Llusià and Peñuelas, 1998; Sallam, 2009). In fact, it has been suggested that low to moderate Blanch et al., 2007). power and exposure time of microwave radiation stimulate the germi- In particular, volatile organic compound (VOC) emissions can alter nation and seedling vigor of plants, while irradiation with increased fre- the success of a given plant species in its environment and influence quency and power density reduce the seed germination and seedling the ecosystem functioning due to toxicity of many VOCs for omnivorous vigor (Ragha et al., 2011). herbivores as well as due to their role as infochemicals within and Given the limited understanding of how plant secondary chemistry across trophic levels (Dicke et al., 2009; Ormeño et al., 2011). VOCs responds to a sustained stress generated by low power microwaves are divided in distinct groups: terpenes, phenolics, N- and S-containing we studied the essential oil content and volatile emissions in O. compounds (Mazid et al., 2011; Niinemets et al., 2013). Terpenes are the basilicum exposed to either microwaves corresponding to wireless rout- largest and most diverse class of volatile organic compounds (VOCs), er (WLAN) or mobile devices (GSM). Based on the results of our previ- whereas individual terpenoids can have overlapping or distinct biolog- ous experiment (Soran et al., 2014), we hypothesized that exposure to ical roles depending on the volatile compound and plant species microwaves results in overall increases in the emission of volatiles and (Peñuelas et al., 1995; Niinemets et al., 2013; Blande et al., 2014). essential oil content, and that the induction is stronger for WLAN irradi- VOCs can be immediately emitted after their synthesis or they can be ation due to higher frequency. As the stress volatiles have different func- first stored in specialized leaf storage structures and then slowly emit- tions than constitutively synthesized volatiles, it is important to gain ted from the storage (Niinemets et al., 2010b; Harrison et al., 2013). In insight into the share of constitutive vs. induced emission to assess the O. basilicum, the specialized storage tissues include oil glands and glan- quality of the plant for culinary and medicinal use. Chemometric tech- dular trichomes (Flinn et al., 1993; Gang et al., 2002; Ioannidis et al., niques were used as tools for the identification and characterization of 2002; de Almeida et al., 2010). When analyzing changes in the emission volatile compounds in the essential oil extracts and in volatile emissions composition upon stress, it is further important to consider that VOC based on their GC–MS (gas chromatograph coupled with mass spec- emissions can be stress-elicited (Niinemets et al., 2010a; Harrison et trometer) profiles. al., 2013). This distinction is especially relevant for aromatic plants with a strong constitutive VOC storage and emission capacity. In such 2. Materials and methods species, stress can elicit storage-independent induced emissions, and modify the storage-emissions by altering the rate of synthesis of stored 2.1. Plant material and stress application volatiles and/or alter the permeability of cell walls of storage structures (Niinemets et al., 2010a; Copolovici et al., 2012; Grote et al., 2013). Sweet basil (Ocimum basilicum L., Agrosel, Câmpia Turzii, Romania) In recent years, the electromagnetic pollution has increased due to seeds were sown in 150 mL plastic pots (height × diameter of enhanced spread of microwaves and radiofrequency radiation signals 8.5 × 6.5 cm) filled with commercial garden soil and grown in three from various sources, including GSM (Global System for Mobile com- fully-closed identical anechoic chambers (Surducan et al., 2012) with munications) and UMTS (Universal Mobile Telecommunications Sys- an isolation degree of 60 dB at the radiofrequency range between the tem)/3G signals from wireless local area networks, and wireless exterior and interior. One chamber was used for the untreated (control) personal area networks such as Bluetooth (Creţescu et al., 2013). Most plants, and the other two chambers were used for plant irradiation. In of the studies regarding the risk of microwave field have focused on each chamber, the same environmental conditions - light intensity of human health (Breckenkamp et al., 2003; Özdemir and Kargi, 2011; 300 μmol m−2 s−1 provided by four 4 W MR16 LED lamps (each lamp I. Lung et al. / Science of the Total Environment 569–570 (2016) 489–495 491 consisted of 26 warm white SMD 5050 LEDs, color temperature 3300 K), 200 mL min−1 for 20 min using a 1003-SKC constant flow sampling −1 air temperature of 25 °C, CO2 concentration of 385 ± 20 μmol mol pump (SKC Inc., Houston, TX, USA) at room temperature. The back- and humidity of 65% - were maintained. Plants were watered every ground air samples (blanks) were separately measured with the two days with 10 mL of bi-distilled water (bi-distiller AcquaMatic empty cuvette before and after the experiments. model AWC/4D, Hamilton Laboratory Glass Ltd., Kent, UK). Volatile analysis was performed using a combined automated car- The experiment was started in three weeks from sowing. The irradi- tridge desorber (Shimadzu TD20, Kyoto, Japan) and GC–MS instrument ation was performed with microwaves corresponding to mobile devices (Shimadzu QP2010 Plus GC–MS, Kyoto, Japan). The volatiles were ana- (GSM) using a modified generator AP5200 (D-LINK, China), operating in lyzed according to the detailed method described by Copolovici et al. four bands (860–910 MHz frequency range, Pout 29 dBm), and wireless (2009) and Toome et al. (2010) and identified. router (WLAN) using a D-LINK wireless router 802.11 g/2.4 GHz (2.412– 2.48 GHz frequency range, Pout 19 dBm). The power density was mea- 2.4. Essential oil sampling and analysis sured using a spectrum analyzer SPECTRAN HF 4060, AARONIA AG (Euscheid, Germany) at the base of the chamber. The irradiation levels Fresh leaf samples of O. basilicum were dried in an Ecocell oven at corresponded to those measured in open spaces for heavily used GSM 30 °C until the constant mass to preserve volatiles (Portillo-Estrada et networks (100 mW m−2) and for indoor WLAN networks upon inten- al., 2015). A sample of 0.5 g dry leaf material was ground in liquid nitro- sive data exchange (70 mW m−2). gen with a mortar and a pestle and essential oils were extracted in The plants were exposed to irradiation treatments for two weeks 25 mL of 1:1 mixture of HPLC grade diethyl ether and n-hexane until they had at least 10–15 leaves and were 20–25 cm tall, but still (Merck, Darmstadt, Germany). Initially, the ground plant material was lacked flowers. After the treatments, the plants were removed from soaked in the extraction solution, and then the essential oils were ex- the chambers for sampling of VOCs and for extraction of essential oils. tracted using an ultrasonic bath (Elmasonic S 15H, 37 kHz, Singen, Ger- All treatments were replicated three times. many) at 30 °C, for 30 min. The extracts were decanted and filtered through a 0.45 μm nylon syringe filter (VWR International, Radnor, PA, 2.2. Determination of photosynthetic parameters USA). Essential oil analysis was performed using the same GC–MS system

CO2 and H2O concentrations in the chamber inlet and outlet were employed for VOC analyses. The experimental conditions used for ana- measured with a portable gas exchange system GFS-3000 equipped a lyzing the essential oils are described in detail in Soran et al. (2014).The with leaf chamber fluorimeter (8 cm2 chamber cross-section area) for carrier gas was helium (purity 99.9999%, Elmer Messer Gaas AS, Tallinn, chamber illumination (Walz GmbH, Effeltrich, Germany). The standard Estonia) with a flow rate of 1 mL min−1. All the compounds analyzed in conditions used were: light intensity of 1000 μmol m−2 s−1, leaf tem- this study (VOCs and essential oils) were identified by comparison with −1 perature of 25 °C, chamber CO2 concentration of 385 μmol mol and the NIST spectral library and based on the retention time identity with humidity 65%. After enclosure of the leaf in the chamber, standard envi- authentic standards of GC–MS purity (Sigma-Aldrich, St. Louis, MO, ronmental conditions were established, and foliage gas exchange rates USA). were recorded after they had reached a steady state. The rates of net as- similation (A) and stomatal conductance to water vapor (gs) were calcu- 2.5. Data analyses and interpretation lated per unit enclosed leaf area according to von Caemmerer and Farquhar (1981). All analyses were carried out in triplicate and all results are expressed as mean ± standard error. The means of each detected com- 2.3. Volatile organic compound (VOC) sampling and analysis pound and photosynthetic characteristics were statistically compared among the treatments with one way ANOVA followed by post hoc GFS-3000 (Walz GmbH, Effeltrich, Germany) operated in the same Tukey's tests using ORIGIN 8 (Origin-Lab Corporation, Northampton, standard environmental conditions as for photosynthesis measure- MA, USA). ments was used for VOC sampling. After recording the steady-state pho- Volatile analysis by gas chromatography–mass spectrometry (GC– tosynthesis and transpiration rates, leaf volatiles from 4 L of air existing MS) used here is the routine analytical method that has high precision the cuvette were sampled in a multibed stainless steel cartridge and accuracy for major components of the emission and essential oil (10.5 × 3 cm, Supelco, Bellefonte, PA, USA) filled with Carbopack adsor- blends. However, it is difficult to detect the minor components of the bents (Carbopack C 20/40 and 40/60 mesh, and Carbopack × 20/ extracts that are present at low levels (Li et al., 2013), and GC–MS appli- 40 mesh). Chamber air sampling was performed at a flow rate of cation is restricted to compounds that are volatile and stable at the

Fig. 1. Changes in net assimilation rate (a) and stomatal conductance to water vapor per unit projected leaf area (b) in sweet basil (Ocimum basilicum) in response to microwave irradiation at frequency bands corresponding to mobile devices (GSM) and wireless router (WLAN). Each data point is the mean (±SE) of 8 independent replicate experiments with a different plant. Asterisks (*) demonstrate statistically significant differences between the microwave-irradiated and control plants (P b 0.05). 492 I. Lung et al. / Science of the Total Environment 569–570 (2016) 489–495 analysis temperatures. Another limitation is sometimes the similarity of Table 1 −2 −1 mass spectra, making it difficult to distinguish closely related isomers; Volatile organic compound (VOC) emissions (nmol m s ) (±SE) from Ocimum basilicum plants in response to microwave irradiation at bands corresponding to mobile lack of mass spectra of the compounds of interest can be another issue devices (GSM) and wireless router (WLAN). (Cai et al., 2006). Also, the peaks are often overlapping or embedded, even when the chromatographic/spectral conditions are optimized No. Compound Control GSM-irradiated WLAN-irradiated (Wang et al., 2010). Chemometric approaches can be applied as a pow- 1 1-Hexanol 0.012 ± 0.002 0.015 ± 0.002 0.02 ± 0.001# ⁎ erful method for characterizing botanical drugs of different origin and 2 3-(Z)-hexenol 0.042 ± 0.012 0.172 ± 0.013 0.24 ± 0.06 # ⁎# quality (Zhu et al., 2014). For pattern recognition, statistical approaches 3 2-(E)-hexenal 0.072 ± 0.011 0.183 ± 0.012 0.233 ± 0.042 4 Benzaldehyde 0.17 ± 0.08 0.053 ± 0.021 0.015 ± 0.006 commonly applied are: principal component analysis (PCA), hierarchi- 5 para-Cymene 0.053 ± 0.039 0.063 ± 0.012 0.061 ± 0.014 ⁎ cal cluster analysis (HCA), linear discriminant analysis (LDA), k-nearest 6 Estragole 0.034 ± 0.016 0.242 ± 0.012 0.31 ± 0.042 # neighbor (k-NN), and partial least squares-discrimination analysis (PLS- 7 D-Limonene 0.081 ± 0.011 0.092 ± 0.013 0.053 ± 0.013 ⁎ DA). PCA is a statistical method that reduces the dimensionality of the 8 Camphene 0.043 ± 0.011 0.142 ± 0.014 0.213 ± 0.010 # 9 α-Pinene 0.46 ± 0.08 0.453 ± 0.022 0.36 ± 0.13 original data set by constructing a smaller number of new orthogonal ⁎ 10 β-Pinene 0.373 ± 0.043 # 0.152 ± 0.012 0.122 ± 0.023 variables known as principal components (PCs). Each PC is built from ⁎ 11 β-Myrcene 0.183 ± 0.041 # 0.012 ± 0.006 n.d. ⁎ linear combinations of the original variables and is a useful tool to ex- 12 α-Phellandrene 0.073 ± 0.024 # 0.023 ± 0.006 0.016 ± 0.007 amine the relationships between the compounds as well as to detect 13 β-Phellandrene n.d. n.d. 0.011 ± 0.005 possible outliers (Sârbu et al., 2012; Casoni and Sârbu, 2014). HCA 14 Eucalyptol 0.67 ± 0.07 2.04 ± 0.05 2.68 ± 0.42# 15 α-Terpineol 0.041 ± 0.012 n.d. 0.010 ± 0.006 builds a hierarchy of clusters of data of varying similarity, typically rep- ⁎ 16 Bornyl acetate 0.360 ± 0.053 # 0.049 ± 0.011 0.061 ± 0.012 resented by a dendrogram. Application of PCA and HCA to chromato- 17 Isobornyl acetate 0.013 ± 0.006 0.019 ± 0.007 0.022 ± 0.006 ⁎ graphic data treatment has become popular in the last two decades 18 α-Caryophyllene 0.32 ± 0.06 # 0.032 ± 0.004 0.024 ± 0.005 ⁎ (Coltro et al., 2005), and these approaches are currently widely applied 19 Caryophyllene oxide 0.016 ± 0.007 0.161 ± 0.011 0.214 ± 0.033 # in plant compound analyses (Chun et al., 2011; Zhang et al., 2013; Gong 20 α-Bergamotene 0.114 ± 0.022 0.163 ± 0.014 0.172 ± 0.035 21 Geranyl acetone 0.042 ± 0.011 n.d. n.d. et al., 2014; Benevides et al., 2014). Thus, in addition to univariate anal- ⁎ 22 (E)-β-ocimene n.d. 0.054 ± 0.011 0.132 ± 0.022 # yses, we used multivariate analyses to gain insight into the similarity and dissimilarity of emission profiles among the treatments. First, a n.d. = not detected; ⁎ Statistically significant differences between the microwave-irradiated and control plants HCA was carried out with the data for volatile compounds and essential (P b 0.05). oils from irradiated and non-irradiated basil, using the Euclidian dis- # Statistically significant differences between GSM- and WLAN-irradiated plants tance as a measure of similarity. The distance between any two clusters (P b 0.05). was obtained with a complete linkage method. A PCA was further per- formed with the VOC emissions and essential oil contents from irradiat- considered to be one of the initial plant stress responses ed and non-irradiated basil plants to discriminate between treated and (Liavonchanka and Feussner, 2006). LOX emissions have been reported reference plants. Both HCA and PCA analyses were conducted with in response to drought (Permyakova et al., 2012), heat and cold Minitab 17 (Minitab Ltd., Coventry, UK). All statistical tests were consid- (Copolovici et al., 2012), high light (Loreto et al., 2006), ozone ered significant at P b 0.05. (Beauchamp et al., 2005), leaf herbivory or mechanical damage (Allmann and Baldwin, 2010; Holopainen, 2011), fungal infections 3. Results and discussion (Toome et al., 2010; Copolovici et al., 2014; Jiang et al., in press-b)and environmental pollutants including antibiotics (Opriş et al., 2013)and 3.1. Alteration of photosynthetic characteristics by microwave irradiation textile dyes (Copaciu et al., 2013). In our study, statistically significant increases in LOX compound emission were observed in the case of 3- Net assimilation rate and stomatal conductance to water vapor de- Z-hexenol and 2-E-hexenal (about 2.5 times higher than control), indi- creased in plants exposed to mild microwave irradiation, but there cating that microwave irradiation did lead to enhance stress status in were no significant differences between the WLAN- and GSM-irradiated the plants. plants (Fig. 1). However, the intercellular CO2 concentration (data not Among the stress-induced compounds in a plant, after the VOCs and shown) was similar among the treatments (384 ± 10 μmol mol−1 for LOX products are emitted the mono- and sesquiterpenes (Matsui, the control vs. 380 ± 12 μmol mol−1 for WLAN- and 378 ± 2006). The composition of constitutive and induced emissions varies, 15 μmol mol−1 for GSM-irradiated plants, means are not significantly N different, P 0.05), suggesting that the reduction in net assimilation Table 2 rate due to microwave irradiation is mainly resulting from reductions Essential oil content (mg g−1 DW) (±SE) in O. basilicum foliage in response to microwave in either photosynthetic capacity or mesophyll conductance to CO2 irradiations at bands corresponding to mobile devices (GSM) and wireless router (WLAN). (Niinemets et al., 2005; Flexas et al., 2012; Tomas et al., 2013). No. Compound Control GSM-irradiated WLAN-irradiated

1 Octane 0.0067 ± 0.0008 n.d. 0.141 ± 0.046 3.2. VOC emissions as altered by microwave irradiation ⁎ 2n-Decane 0.404 ± 0.042 0.62 ± 0.07 0.61 ± 0.06 # ⁎ 3 Undecane n.d. 0.050 ± 0.005 0.099 ± 0.007 # ⁎ Emission of altogether 22 VOCs belonging to three key compound 4 Dodecane 0.55 ± 0.04 1.05 ± 0.12 0.80 ± 0.07 # ⁎ classes was detected from the foliage of O. basilicum: lipoxygenase path- 5 Tetradecane 0.18 ± 0.07 0.51 ± 0.08 0.281 ± 0.012 # way products (LOX, also called green leaf volatiles), benzenoids and ter- 6 Hexadecane 0.012 ± 0.003 0.023 ± 0.0015 0.20 ± 0.06 7 Limonene 0.123 ± 0.025 0.24 ± 0.09 0.218 ± 0.021# penoids (including oxygenated mono and sesquiterpenes (Table 1)). 8 Eucalyptol 0.091 ± 0.001 0.11 ± 0.05 0.07 ± 0.001 ⁎ Although, p-cymene is chemically a benzenoid, it is suggested to be a 9 Phytol 0.17 ± 0.10 0.108 ± 0.014 0.035 ± 0.002 # conversion product of γ-terpinene, and thus, can be also classified as 10 Acetic acid n.d. n.d. 0.29 ± 0.05# ⁎# an aromatic terpene (Poulose and Croteau, 1978). 11 Butanal n.d. 0.047 ± 0.007 0.087 ± 0.009 12 n-Hexadecanoic acid n.d. 0.021 ± 0.005 0.29 ± 0.06 Typically, emissions of LOX volatiles are stress-elicited, while emis- ⁎ 13 Cytidine 0.16 ± 0.05 0.46 ± 0.04 0.28 ± 0.02 # sions of terpenoids and benzenoids can be constitutive or induced (Niinemets et al., 2010a; Joó et al., 2011; Harrison et al., 2013; Rajabi n.d. = not detected; ⁎ Statistically significant differences between the microwave-irradiated and control plants et al., 2013). LOX products are formed from polyunsaturated fatty (P b 0.05). acids that are released from plant membranes in different stress condi- # Statistically significant differences between GSM- and WLAN-irradiated plants tions (Feussner and Wasternack, 2002), and LOX emissions are (P b 0.05). I. Lung et al. / Science of the Total Environment 569–570 (2016) 489–495 493

Fig. 2. Hierarchical clustering dendrograms for volatile organic compounds (a) and essential oil (b), by complete linkage method. e.g., constitutive emissions are typically dominated by ubiquitous together, there was a negative relationship between induced and con- monoterpenes (α-pinene, β-pinene and limonene), while induced stitutive emissions. emissions are dominated by ocimenes, linalool and eucalyptol (e.g., Niinemets et al., 2010a). The emissions of these volatiles have been con- sidered as highly sensitive indicators of stress (Copolovici et al., 2012). 3.3. Microwave irradiation effects on essential oil Generally, important monoterpene emission can be induced in most plant species by different stress conditions (Loreto and Schnitzler, Essential oils consist of variable mixtures of terpenoids and benze-

2010; Niinemets, 2010). We observed major increases in the emissions noids, especially monoterpenes (C10), and sesquiterpenes (C15) as well of several terpenoids typically associated with stress such as: eucalyptol, as esterified or oxylated (typically methoxylated) benzenoids (between caryophyllene oxide, and α-bergamotene (Table 1). The highest in- ca. C8 and C15). Diterpenes (C20) may also be present along with a vari- crease was obtained for caryophyllene oxide (21 times greater content ety of other molecules (Calín-Sánchez et al., 2012; De Martino et al., compared to control). It currently remains unclear whether the greater 2015). The essential oil in O. basilicum consisted of saturated aliphatic content of caryophyllene oxide reflects enzymatic oxidation of α- compounds: octane; n-decane, undecane, dodecane, tetradecane, caryophyllene or autogenous oxidation due to greater oxidative stress hexadecane; monoterpenes: limonene and eucalyptol; oxygenated in microwave-irradiated plants. WLAN-frequency irradiation resulted compounds of varying chain length: phytol, acetic acid, and butanal, in a stronger induction of VOC emissions than the GSM-frequency irra- and n-hexadecanoic acid (Table 2). diation (Table 1), suggesting that it is a severe stress. Irradiation of O. basilicum with both types of microwaves used in this While stress generally leads to enhanced emissions, constitutive study led to an increase in the content of essential oils, except for phytol, emissions are often reduced in response to stress, especially upon se- whose content decreased by 4% in the case of GSM-frequency and 82% vere stress (Staudt et al., 2002; Lavoir et al., 2009; Copolovici et al., in the case of WLAN-frequency microwaves. Compared to the control 2014). In our study (Table 1), microwave-irradiated O. basilicum plants plants, the strongest increase in response to WLAN irradiation was ob- had lower emissions of β-pinene, α-phellandrene and bornyl acetate served for hexadecane and octane (N17 times greater content, Table (decreased N67% compared to control). In addition, major reductions 2). Plant irradiation with WLAN-frequency increased the content of lim- in the emissions of β-myrcene (94% in GSM-irradiated and complete onene and cytidine by N1.7 times compared with control plants. Five of absence in WLAN-irradiated plants), α-caryophyllene (90% in GSM-ir- the essential oil components, n-hexadecanoic acid, hexadecane, radiated and 93% in WLAN-irradiated plants) were also observed. undecane, acetic acid, and butanal, detected in the plants irradiated Among benzenoids, benzaldehyde emissions decreased by 70% in with GSM and WLAN microwaves were not detected in the control GSM-irradiated and 94% in WLAN-irradiated plants. Thus, taken plants (Table 2).

Fig. 3. A biplot representation of the volatile organic compounds (a) and essential oil (b) of the sweet basil plants according to the principal component analysis. In (a) each number corresponds to an individual organic compound as in Table 1 and in (b) corresponds as in Table 2. 494 I. Lung et al. / Science of the Total Environment 569–570 (2016) 489–495

3.4. Chemometric evaluation of microwave irradiance effects on secondary Beauchamp, J., Wisthaler, A., Hansel, A., Kleist, E., Miebach, M., Niinemets, Ü., Schurr, U., Wildt, J., 2005. Ozone induced emissions of biogenic VOCs from tobacco: relation- chemistry ships between ozone uptake and emission of LOX products. Plant Cell Environ. 28, 1334–1343. The hierarchical cluster analysis (HCA) of the GC–MS chromato- Benevides, C.M.J., Bezerra, M.A., Pereira, P.A.P., Andrade, J.B., 2014. HS-SPME/GC–MS anal- ysis of VOCs and multivariate techniques applied to the discrimination of Brazilian grams for the samples resulted in grouping all samples into three varieties of mango. Am. J. Anal. Chem. 5, 157–164. main clusters: control, GSM- and WLAN-irradiated plants. The same re- Blanch, J.-S., Peñuelas, J., Llusià, J., 2007. Sensitivity of terpene emissions to drought and sults were obtained for VOCs and essential oils (Fig. 2). fertilization in terpene-storing Pinus halepensis and non-storing Quercus ilex.Physiol. – Principal component analysis (PCA) was further applied to evaluate Plant. 131, 211 225. Blande, J.D., Holopainen, J.K., Niinemets, Ü., 2014. Plant volatiles in polluted atmospheres: microwave-driven modifications in the spectrum of VOCs and essential stress responses and signal degradation. Plant Cell Environ. 37, 1892–1904. oil in basil plants. In the case of the VOCs, the first principal component Breckenkamp, J., Berg, G., Blettner, M., 2003. Biological effects on human health due to ra- (PC1) had a variance (eigenvalue) of 18.54 and accounted for 84.3% of diofrequency/microwave exposure: a synopsis of cohort studies. Radiat. Environ. Biophys. 42, 141–154. the total variance. PC1 was represented by camphene, p-cymene, euca- Caemmerer, S., Farquhar, G.D., 1981. Some relationships between the biochemistry of lyptol,estragole,caryophylleneoxide,α-bergamotene, 3-Z-hexenol, 2- photosynthesis and the gas exchange of leaves. Planta 153, 376–387. E-hexenal, and E-β-ocimene. The second principal component (PC2) Cai, J.B., Lin, P., Zhu, X.L., Su, Q.D., 2006. Comparative analysis of clary sage (S. sclarea L.) oil volatiles by GC–FTIR and GC–MS. Food Chem. 99, 401–407. had an eigenvalue of 3.46 and accounted for 15.7% of the data variability. Calín-Sánchez, Á., Lech, K., Szumny, A., Figiel, A., Carbonell-Barrachina, Á.A., 2012. Volatile PC2 was represented by α-pinene and D-limonene. composition of sweet basil essential oil (Ocimum basilicum L.) as affected by drying For the essential oils, the eigenvalues for PC1 and PC2 were 8.17 and method. Food Res. Int. 48, 217–225. Casoni, D., Sârbu, C., 2014. Comprehensive evaluation of antioxidant activity: a chemo- 4.83, and these two components accounted for 62.9% (PC1) and 37.1% metric approach using principal component analysis. Spectrochim. Acta A Mol. (PC2) of the total variance. The components were represented by the Biomol. Spectrosc. 118, 343–348. following compounds: decane, limonene, hexadecane, octane, Chiang, L.C., Cheng, P.W., Chiang, W., Lin, C.C., 2005. Antiviral activity of extracts and se- hexadecanoic acid, undecane, acetic acid, and butanal for PC1, and dec- lected pure constituents of Ocimum basilicum. Clin. Exp. Pharmacol. Physiol. 32, 811–816. ane, limonene, eucalyptol, dodecane, tetradecane, and cytidine for PC2. Chun, M.H., Kim, E.K., Yu, S.M., Oh, M.S., Moon, K.Y., Jung, J.H., Hong, J., 2011. GC/MS com- This analysis further demonstrated that the plants were clearly separat- bined with chemometrics methods for quality control of Schizonepeta tenuifolia Briq: – ed into three groups: control plants, GSM- and WLAN-irradiated plants determination of essential oils. Microchem. J. 97, 274 281. Coltro, W.K.T., Ferreira, M.M.C., Macedo, F.A.F., Oliveira, C.C., Visentainer, J.V., Souza, N.E., (Fig. 3). Thus, both multivariate analysis methods were in agreement Matsushita, M., 2005. Correlation of animal diet and fatty acid content in young that volatile and essential oil compositions were different among the goat meat by gas chromatography and chemometrics. Meat Sci. 71, 358–363. ş three groups (Figs. 2 and 3). Such changes in the emission and essential Copaciu, F., Opri , O., Coman, V., Ristoiu, D., Niinemets, Ü., Copolovici, L., 2013. Diffuse water pollution by anthraquinone and azo dyes in environment importantly alters fo- oil blend can importantly alter the culinary properties and health effects liage volatiles, carotenoids and physiology in wheat (Triticum aestivum). Water Air of sweet basil. Soil Pollut. 224, 1478. Copolovici, L., Kännaste, A., Niinemets, Ü., 2009. Gas chromatography–mass spectrometry method for determination of monoterpene and sesquiterpene emissions from 4. Conclusions stressed plants. Stud. Univ. Babes-Bolyai Chem. LIV4, 329–339. Copolovici, L., Kännaste, A., Pazouki, L., Niinemets, Ü., 2012. Emissions of green leaf vola- Modifications in photosynthetic characteristics, VOC emissions and tiles and terpenoids from Solanum lycopersicum are quantitatively related to the se- verity of cold and heat shock treatments. J. Plant Physiol. 169, 664–672. essential oil content collectively demonstrate that microwave irradiation Copolovici, L., Kännaste, A., Remmel, T., Niinemets, Ü., 2013. Volatile organic compound represents a stress for the plants. WLAN-frequency irradiation resulted emissions from Alnus glutinosa under interacting drought and herbivory stresses. En- in a stronger induction of VOC emissions of O. basilicum than the GSM-fre- viron. Exp. Bot. 100, 55–63. Copolovici, L., Väärtnõu, F., Portillo Estrada, M., Niinemets, Ü., 2014. Oak powdery mildew quency irradiation, suggesting that WLAN irradiation constituted a more (Erysiphe alphitoides)-induced volatile emissions scale with the degree of infection in severe stress to the plants. On the other hand, irradiation of O. basilicum Quercus robur.TreePhysiol.34,1399–1410. with both types of microwaves used in this study led to an increase in Creţescu, I., Căpriţă, R., Velicevici, G., Căpriţă, A., 2013. Study regarding influence of micro- the content of essential oils, except for phytol. However, the composition wave irradiation on chlorophyll content at some barley (Hordeum vulgare L.) geno- types. Lucrări Ştiinţifice-Ser. Zootehnie 59, 270–273. of the studied plant was significantly altered, and it remains to investigate de Almeida, G.L., Alves, A.A., Campos, O.W., 2010. Characterization and ontogeny of the whether such changes alter culinary or health properties of the essential glandular trichomes of Ocimum selloi Benth. (Lamiaceae). Acta Bot. Bras. 24, 909–915. oil. The data presented here demonstrate that human-generated micro- De Martino, L., Nazzaro, F., Mancini, E., De Feo, V., 2015. Essential oils from Mediterranean aromatic plants. The Mediterranean Diet. An evidence-based approach, pp. 649–661. wave pollution can potentially constitute a threat to the environment, Dicke, M., van Loon, J.J.A., Soler, R., 2009. Chemical complexity of volatiles from plants in- or perhaps the microwave irradiation of the plants can be an alternative duced by multiple attack. Nat. Chem. Biol. 5, 317–324. way to enhance the culinary properties of basil plants, providing the fre- Feussner, I., Wasternack, C., 2002. The lipoxygenase pathway. Annu. Rev. Plant Biol. 53, 275–297. quency used not to be dangerous for both animals and human beings. Flexas, J., Barbour, M.M., Brendel, O., Cabrera, H.M., Carriqui, M., Diaz-Espejo, A., Douthe, C., Dreyer, E., Ferrio, J.P., Gago, J., Galle, A., Galmés, J., Kodama, N., Medrano, H., Acknowledgments Niinemets, Ü., Peguero-Pina, J.J., Pou, A., Ribas-Carbo, M., Tomas, M., Tosens, T., Warren, C.R., 2012. Mesophyll diffusion conductance to CO2: an unappreciated cen- tral player in photosynthesis. Plant Sci. 193, 70–84. This work was supported by Estonian Ministry of Science and Educa- Flinn, C.L., Murray, R., Simon, J.E., 1993. Anatomical investigation of essential oil glands tion (institutional grant IUT-8-3) and co-funded by the European Com- and their distribution in Ocimum basilicum varieties. HortSci. 28, 531. Gang, D.R., Simon, J., Lewinsohn, E., Pichersky, E., 2002. Peltate glandular trichomes of mission through European Regional Development Fund Structural Ocimum basilicum L. (sweet basil) contain high levels of enzymes involved in the bio- Operational Program “Increasing of Economic Competitiveness” Priority synthesis of phenylpropenes. J. Herbs Spices Med. Plants 9, 189–195. axis 2, operation 2.1.2. Contract Number 621/2014, and the European Gong, H.Y., Liu, W.H., Lv, G.Y., Zhou, X., 2014. Analysis of essential oils of Origanum vulgare – Regional Development Fund Centers of Excellence Program (Centre of from six production areas of China and Pakistan. Rev. Bras. Farmacogn. 24, 25 32. Grote, R., Monson, R.K., Niinemets, Ü., 2013. Leaf-level models of constitutive and stress- Excellence EcolChange). The authors thank Dr. E. Surducan team driven volatile organic compound emissions. In: Niinemets, Ü., Monson, R.K. (Eds.), (INCDTIM Cluj-Napoca) for all their experimental support with anecho- Biology, controls and models of tree volatile organic compound emissions. Springer, – ic chambers. Berlin, pp. 315 355. Harrison, S.P., Morfopoulos, C., Dani, K.G.S., Prentice, I.C., Arneth, A., Atwell, B.J., Barkley, M.P., Leishman, M.R., Loreto, F., Medlyn, B.E., Niinemets, Ü., Possell, M., Peñuelas, J., References Wright, I.J., 2013. Volatile isoprenoid emissions from plastid to planet. New Phytol. 197, 49–57. Allmann, S., Baldwin, I.T., 2010. Insects betray themselves in nature to predators by rapid Hassanpouraghdam, M.B., Gohari, G.R., Tabatabaei, S.J., Dadpour, M.R., Shirdel, M., 2011. isomerization of green leaf volatiles. Science 329, 1075–1078. NaCl salinity and Zn foliar application influence essential oil composition of basil Baratta, M.T., Dorman, H.J.D., Deans, S.G., Figueiredo, A.C., Barroso, J.G., Ruberto, G., 1998. (Ocimum basilicum L.). Acta Agric. Slov. 97 (2), 93–98. Antimicrobial and antioxidant properties of some commercial essential oil. Flavour Holopainen, J.K., 2011. Can forest trees compensate for stress generated growth losses by Fragr. J. 13, 235-234. induced production of volatile compounds? Tree Physiol. 31, 1356–1377. I. Lung et al. / Science of the Total Environment 569–570 (2016) 489–495 495

Ioannidis, D., Bonner, L., Johnson, C.B., 2002. UV-B is required for normal development of Permyakova, M.D., Permyakov, A.V., Osipova, S.V., Pshenichnikova, T.A., 2012. oil glands in Ocimum basilicum L. (sweet basil). Ann. Bot. 90, 453–460. Lipoxygenase from the leaves of wheat grown under different water supply condi- Jiang, Y., Ye, J., Li, S., Niinemets, Ü., 2016a. Regulation of terpenoid biosynthesis and emis- tions. Appl. Biochem. Microbiol. 48, 77–82. sion in flowers of sweet basil (Ocimum basilicum): controls by light and biotic stress Portillo-Estrada, M., Copolovici, L., Niinemets, Ü., 2015. Bias in leaf dry mass estimation elicited by methyl jasmonate. J. Plant Growth Regul. (in press). after oven-drying isoprenoid-storing leaves. Trees-Struct. Funct. 29, 1805–1816. Jiang, Y., Ye, J., Veromann, L.-L., Niinemets, Ü., 2016b. Scaling of photosynthesis and con- Poulose, A.J., Croteau, R., 1978. Biosynthesis of aromatic monoterpenes. Conversion of stitutive and induced volatile emissions with severity of leaf infection by rust fungus gamma-terpinene to p-cymene and thymol in Thymus vulgaris L. Arch. Biochem. (Melampsora larici-populina)inPopulus balsamifera Var. suaveolens.TreePhysiol.(in Biophys. 187, 307–314. press). Ragha, L., Mishra, S., Ramachandran, V., Bhatia, M.S., 2011. Effects of low-power micro- Jirovetz, L., Buchbauer, G., Shafi, M.P., Kaniampady, M.M., 2003. Chemotaxonomical anal- wave fields on seed germination and growth rate. J. Electromagn. Anal. Appl. 3, ysis of the essential oil aroma compounds of four different Ocimum species from 165–171. southern India. Eur. Food Res. Technol. 217 (2), 120–124. Rajabi, M.H., Pazouki, L., Niinemets, Ü., 2013. The biochemistry and molecular biology of Joó, E., Dewulf, J., Amelynck, C., Schoon, N., Pokorska, O., Šimpraga, M., Steppe, K., Aubinet, volatile messengers in trees. In: Niinemets, Ü., Monson, R.K. (Eds.), Biology, controls M., Van Langenhove, H., 2011. Constitutive versus heat and biotic stress induced and models of tree volatile organic compound emissions. Springer, Berlin, pp. 47–93. BVOCs emissions in Pseudotsuga menziesii. Atmos. Environ. 45, 3655–3662. Ramakrishna, A., Ravishankar, G.A., 2011. Influence of abiotic stress signals on secondary Khalafallah, A.A., Sallam, S.M., 2009. Response of maize seedlings to microwaves at metabolites in plants. Plant Signal. Behav. 6 (11), 1720–1731. 945 MHz. Rom. J. Biophys. 19 (1), 49–62. Rhizopoulou, S., Diamantoglou, S., 1991. Water stress-induced diurnal variations in leaf Lak, A., 2012. Human health effects from radiofrequency and microwave fields. J. Basic water relations, stomatal conductance, soluble sugars, lipids and essential oil content Appl. Sci. Res. 2 (12), 12302–12305. of Origanum majorana L. J. Hortic. Sci. 66, 119–125. Lavoir, A.-V., Staudt, M., Schnitzler, J.P., Landais, D., Massol, F., Rocheteau, A., Rodriguez, R., Sârbu, C., Nașcu-Briciu, R.D., Kot-Wasik, A., Gorinstein, S., Wasik, A., Namiesnik, J., 2012. Zimmer, I., Rambal, S., 2009. Drought reduced monoterpene emissions from the ever- Classification and fingerprinting of kiwi and pomelo fruits by multivariate analysis green Mediterranean oak Quercus ilex: results from a throughfall displacement exper- of chromatographic and spectroscopic data. Food Chem. 130, 94–1002. iment. Biogeosciences 6, 1167–1180. Sharma, V.P., Singh, H.P., Batish, D.R., Kohli, R.K., 2010. Cell phone radiations affect early Letchamo, W., Lu, H.L., Gosselin, A., 1995. Variations in photosynthesis and essential oil in growth of Vigna radiata (mung bean) through biochemical alterations. Z. Naturforsch. thyme. J. Plant Physiol. 147, 29–37. 65c, 66–72. Li, Y.Q., Kong, D.X., Wu, H., 2013. Analysis and evaluation of essential oil components of Sirousmehr, A., Arbabi, J., Asgharipour, M.R., 2014. Effect of drought stress levels and or- cinnamon barks using GC–MS and FTIR spectroscopy. Ind. Crop. Prod. 41, 269–278. ganic manures on yield, essential oil content and some morphological characteristics Liavonchanka, A., Feussner, N., 2006. Lipoxygenases: occurrence, functions and catalysis. of sweet basil (Ocimum basilicum). Adv. Environ. Biol. 8 (4), 880–885. J. Plant Physiol. 163, 348–357. Soran, M.L., Stan, M., Niinemets, Ü., Copolovici, L., 2014. Influence of microwave frequency Llusià, J., Peñuelas, J., 1998. Changes in terpene content and emission in potted Mediter- electromagnetic radiation on terpene emission and content in aromatic plants. ranean woody plants under severe drought. Can. J. Bot. 76, 1366–1373. J. Plant Physiol. 171, 1436–1443. Loreto, F., Schnitzler, J.P., 2010. Abiotic stresses and induced BVOCs. Trends Plant Sci. 15, Soran, M.L., Stan, M., Lung, I., Truşcă, M.R.C., 2016. Microwave field effect on polyphenolic 154–166. compounds from aromatic plants. J. Sustain. Dev. Energy Water Environ. Syst. 4 (1), Loreto, F., Barta, C., Brilli, F., Nogues, I., 2006. On the induction of volatile organic com- 48–55. pound emissions by plants as consequence of wounding or fluctuations of light and Stan, M., Soran, M.L., Varodi, C., Lung, I., 2014. Influence of microwave field on the ascorbic temperature. Plant Cell Environ. 29, 1820–1828. acid content in leaves of some common aromatic plants in Romania. Stud. Univ. Matsui, K., 2006. Green leaf volatiles: hydroperoxide lyase pathway of oxylipin metabo- Babes-Bolyai Chem. LIX 1, 125–133. lism. Curr. Opin. Plant Biol. 9, 274–280. Staudt, M., Rambal, S., Joffre, R., Kesselmeier, J., 2002. Impact of drought on seasonal Mazid, M., Khan, T.A., Mohammad, F., 2011. Role of secondary metabolites in defense monoterpene emissions from Quercus ilex in southern France. J. Geophys. Res. mechanisms of plants. Biol. Med. 3 (2), 232–249. D107, 4602. Munne-Bosch, S., Schwarz, K., Alegre, L., 1999. Enhanced formation of α-tocopherol and Surducan, E., Surducan, V., Halmagyi, A., 2012. Romanian Patent RO-125068B1. highly oxidized abietane diterpenes in water-stressed rosemary plants. Plant Physiol. Timmusk, S., El-Daim, I.A.A., Copolovici, L., Tanilas, T., Kännaste, A., Behers, L., Nevo, E., 121, 1047–1052. Seisenbaeva, G., Stenström, E., Niinemets, Ü., 2014. Drought-tolerance of wheat im- Niinemets, Ü., 2010. Mild versus severe stress and BVOCs: thresholds, priming and conse- proved by rhizosphere bacteria from harsh environments: enhanced biomass pro- quences. Trends Plant Sci. 15, 145–153. duction and reduced emissions of stress volatiles. PLoS One 9 (5), e96086. Niinemets, Ü., Cescatti, A., Rodeghiero, M., Tosens, T., 2005. Leaf internal diffusion conduc- Tomas, M., Flexas, J., Copolovici, L., Galmés, J., Hallik, L., Medrano, H., Ribas-Carbo, M., tance limits photosynthesis more strongly in older leaves of Mediterranean ever- Tosens, T., Vislap, V., Niinemets, Ü., 2013. Importance of leaf anatomy in determining green broad-leaved species. Plant Cell Environ. 28, 1552–1566. mesophyll diffusion conductance to CO2 across species: quantitative limitations and Niinemets, Ü., Arneth, A., Kuhn, U., Monson, R.K., Peñuelas, J., Staudt, M., 2010a. The emis- scaling up by models. J. Exp. Bot. 64, 2269–2281. sion factor of volatile isoprenoids: stress, acclimation, and developmental responses. Toome, M., Randjärv, P., Copolovici, L., Niinemets, Ü., Heinsoo, K., Luik, A., Noe, S.M., 2010. Biogeosciences 7, 2203–2223. Leaf rust induced volatile organic compounds signaling in willow during the infec- Niinemets, Ü., Monson, R.K., Arneth, A., Ciccioli, P., Kesselmeier, J., Kuhn, U., Noe, S.M., tion. Planta 232, 235–243. Peñuelas, J., Staudt, M., 2010b. The leaf-level emission factor of volatile isoprenoids: Vina, A., Murillo, E., 2003. Essential oil composition from twelve varieties of basil (Ocimum caveats, model algorithms, response shapes and scaling. Biogeosciences 7, spp.) grown in Columbia. J. Braz. Chem. Soc. 14 (5), 744–749. 1809–1832. Wang, Y., Jiang, Z.T., Jiang, S., 2010. Chemical composition of the essential oils of Niinemets, Ü., Kännaste, A., Copolovici, L., 2013. Quantitative patterns between plant vol- Cinnamomum loureirii Nees. from China obtained by hydrodistillation and micro- atile emissions induced by biotic stresses and the degree of damage. Front. Plant Sci. wave-assisted hydrodistillation. J. Essent. Oil Res. 22, 129–131. 4, 262. Yani, A., Pauly, G., Faye, M., Salin, F., Gleizes, M., 1993. The effect of a long-term water Opalchenova, G., Obreshkova, D., 2003. Comparative studies on the activity of basil–an es- stress on the metabolism and emission of terpenes of the foliage of Cupressus sential oil from Ocimum basilicum L. against multidrug resistant clinical isolates of the sempervirens. Plant Cell Environ. 16, 975–981. genera Staphylococcus, Enterococcus and Pseudomonas by using different test Zabaras, D., Spooner-Hart, R.N., Wyllie, S.G., 2002. Effects of mechanical wounding on con- methods. J. Microbiol. Methods 54, 105–110. centration and composition of essential oil from Melaleuca alternifolia leaves. Opriş, O., Copaciu, F., Soran, M.L., Ristoiu, D., Niinemets, Ü., Copolovici, L., 2013. Influence Biochem. Syst. Ecol. 30, 399–412. of nine antibiotics on key secondary metabolites and physiological characteristics in Zhang, L., Wang, X., Guo, J., Xia, Q., Zhao, G., Zhou, H., Xie, F., 2013. Metabolic profiling of Triticum aestivum: leaf volatiles as a promising new tool to assess toxicity. Ecotoxicol. Chinese tobacco leaf of different geographical origins by GC–MS. J. Agric. Food Chem. Environ. Saf. 87, 70–79. 61, 2597–2605. Ormeño, E., Goldstein, A., Niinemets, Ü., 2011. Extracting and trapping biogenic volatile Zhu, J., Fan, X., Cheng, Y., Agarwal, R., Moore, C.M.V., Chen, S.T., Tong, W., 2014. Chemo- organic compounds stored in plant species. Trends Anal. Chem. 30 (7), 978–989. metric analysis for identification of botanical raw materials for pharmaceutical use: Özdemir, F., Kargi, A., 2011. In: Zhurbenko, V. (Ed.), Electromagnetic waves and human a case study using Panax notoginseng.PLoSOne9(1),e87462. health, electromagnetic waves. InTech. ISBN: 978-953-307-304-0. Peñuelas, J., Llusià, J., Estiarte, M., 1995. Terpenoids: a plant language. Trends Ecol. Evol. 10, 289. Article

pubs.acs.org/est

Herbivory by an Outbreaking Moth Increases Emissions of Biogenic Volatiles and Leads to Enhanced Secondary Organic Aerosol Formation Capacity Pasi Yli-Pirila,̈*,†,∥ Lucian Copolovici,‡,§ Astrid Kannaste,̈ ‡ Steffen Noe,‡ James D. Blande,∥ Santtu Mikkonen,† Tero Klemola,⊥ Juha Pulkkinen,# Annele Virtanen,† Ari Laaksonen,†,∇ Jorma Joutsensaari,† Ülo Niinemets,‡,○ and Jarmo K. Holopainen∥

† Department of Applied Physics, University of Eastern Finland, P.O. Box 1626, 70211 Kuopio, Finland ‡ Institute of Agricultural and Environmental Sciences, Estonian University of Life Sciences, Kreutzwaldi 1, 51014 Tartu, Estonia § Institute of Technical and Natural Sciences, Research-Development of Aurel Vlaicu University, 2 Elena Dragoi Street, 310330 Arad, Romania ∥ Department of Environmental and Biological Sciences, University of Eastern Finland, P.O. Box 1627, 70211 Kuopio, Finland ⊥ Section of Ecology, Department of Biology, University of Turku, 20014 Turku, Finland # School of Pharmacy, University of Eastern Finland, P.O. Box 1627, 70211 Kuopio, Finland ∇ Finnish Meteorological Institute, P.O. Box 503, 00101 Helsinki, Finland ○ Estonian Academy of Sciences, Kohtu 6, 10130 Tallinn, Estonia

*S Supporting Information

ABSTRACT: In addition to climate warming, greater herbivore pressure is anticipated to enhance the emissions of climate-relevant biogenic volatile organic compounds (VOCs) from boreal and subarctic forests and promote the formation of secondary aerosols (SOA) in the atmosphere. We evaluated the effects of Epirrita autumnata, an outbreaking geometrid moth, feeding and larval density on herbivore-induced VOC emissions from mountain birch in laboratory experiments and assessed the impact of these emissions on SOA formation via ozonolysis in chamber experiments. The results show that herbivore-induced VOC emissions were strongly dependent on larval density. Compared to controls without larval feeding, clear new particle formation by nucleation in the reaction chamber was observed, and the SOA mass loadings in the insect-infested samples were significantly higher (up to 150-fold). To our knowledge, this study provides the first controlled documentation of SOA formation from direct VOC emission of deciduous trees damaged by known defoliating herbivores and suggests that chewing damage on mountain birch foliage could significantly increase reactive VOC emissions that can importantly contribute to SOA formation in subarctic forests. Additional feeding experiments on related silver birch confirmed the SOA results. Thus, herbivory-driven volatiles are likely to play a major role in future biosphere-vegetation feedbacks such as sun- screening under daily 24 h sunshine in the subarctic.

■ INTRODUCTION leaves, insect herbivores act as biotic stressors and induce a specific emission profile of VOCs which are also called Plant-generated volatile organic compound (VOC) emissions 7−10 have various protective functions against abiotic stresses at herbivore-induced plant volatiles (HIPV). These volatiles − cellular and whole plant levels.1 3 The rates of VOC emission have ecological effects in ecosystems ranging from plant to are controlled by compound physicochemical characteristics, plant communication to plant herbivore and plant carnivore − such as low volatility or diffusion, and the biochemical and interactions.11 14 However, specific herbivore-induced VOCs physiological controls over compound synthesis.4,5 Previous work on VOCs has mainly focused on constitutive emissions Received: June 5, 2016 under abiotic stress conditions, but different stresses alone and Revised: September 29, 2016 in combination can induce emission of many new compounds Accepted: October 5, 2016 and alter constitutive emissions.5,6 When feeding on plant Published: October 5, 2016

© 2016 American Chemical Society 11501 DOI: 10.1021/acs.est.6b02800 Environ. Sci. Technol. 2016, 50, 11501−11510 Environmental Science & Technology Article such as methyl salicylate (MeSa), numerous lipoxygenase VOC emissions may have an effect on the overall SOA particle (LOX) pathway products (e.g., (E)-2-hexenal, (Z)-3-hexenol, formation potential and the SOA formation may be even and (Z)-3-hexenyl acetate), homoterpenes (E)-4,8-dimethyl- suppressed by highly reactive VOCs with small individual SOA 1,3,7-nonatriene (DMNT) and (E,E)-4,8,12-trimethyl-1,3,7,11- mass yields.42,45,46 tridecatetraene (TMTT), monoterpene (MT) (E)-β-ocimene, Our aim was to assess the role of herbivore-induced and sesquiterpene (SQT) α-farnesene, can be induced in emissions on SOA formation under controlled laboratory − infections by fungal and bacterial plant pathogens as well.15 17 conditions to address the following questions: (1) how does E. Herbivore-feeding significantly increases the constitutive autumnata feeding at different densities on mountain birch emission of VOCs stored, e.g., in resin canals.18,19 Biotic affect herbivore-induced VOC emission rates from the foliage? infections can also reduce emissions of constitutively emitted (2) how does the composition of VOC emissions differ compoundssuchasisoprene16 or the proportion of between intact and herbivore-damaged plants? and (3) does constitutively emitted compounds can decrease in the total herbivore-induced VOC emissions stimulate formation and emissions.20,21 growth of SOA particles in reactions with ozone? The results Global warming is expected to most strongly affect the provide conclusive evidence of a major increase in SOA northern latitudes resulting in significantly increased VOC formation in herbivore-attacked plants. emissions from vegetation.2,22,23 Global warming is further predicted to induce large-scale insect outbreaks in the northern ■ EXPERIMENTAL SECTION 24−27 forests and recent observations from subarctic mountain Study Design. We performed two sets of experiments to birch forest indicate outbreak range expansion of the geometrid elucidate the effects of herbivory on plant emissions and 26,28−30 moths. Mountain birch (Betula pubescens ssp. czerepa- concurrent SOA formation. The first series of experiments novii (Orlova) Hamet-Ahti)̈ is a subspecies of downy birch and assessed the influence of varying degrees of Epirrita autumnata a dominating tree species in subarctic areas of Greenland, herbivory (0−16 larvae per small seedling) on the plant 31 Iceland, Norway, Sweden, Finland and Russia. Mountain physiological responses and induction of volatile emissions birch has been found to be a significant source of VOCs from Betula pubescens ssp. czerepanovii leaves. VOC emissions 32 dominated by sesquiterpenes and insect feeding on foliage were collected and analyzed by GC-MS. The second series of has induced typical emissions of herbivore-induced VOCs in experiments linked the herbivore-induced VOC emissions to 33 forest sites of northern Lapland. Mountain birch ramets SOA formation, using either 0 (control) or 6 larvae per (multiple stems) infested by foliage-feeding autumnal moth mountain birch seedling, and measuring the volatiles emitted (Epirrita autumnata Borkhausen (Lepidoptera: Geometridae)) from the plants at the inlet of the reaction chamber, larvae induced emissions of monoterpenes (E)-β-ocimene and representing the ingoing VOCs available for SOA production. linalool, homoterpene (E)-DMNT, and several sesquiter- The number of larvae (6) used in the second set of experiments 33 penes. Among the herbivore-induced plant VOCs, there are was optimized on the basis of the results of the first several highly reactive compounds which have shorter lifetimes experiments. than most of the compounds in the constitutive plant emissions Plant and Insect Material. Mountain birch, Betula 34−36 in reactions with atmospheric oxidants. pubescens ssp. czerepanovii (N. I. Orlova) Hamet-Ahti,̈ seedlings A large-scale data analysis of long-term atmospheric were grown from seed material originating from Swedish observations showed that VOCs are emitted from the Lapland (Kiruna, 67°51′ N20°13′ E). Seeds were sown in a 1:1 vegetation at higher rates under a warming climate.37 This mixture of sand and commercial fertilized peat (VAPO B2), will substantially stimulate oxidation of the VOCs to secondary and grown in 0.5 L pots in greenhouses. They were left organic aerosols (SOA); increase the number concentration of overwinter under snow for one winter before they were used in particles large enough to act as cloud condensation nuclei experiments. Autumnal moth (Epirrita autumnata (Lepidop- (CCN) and result in an increase of cloud albedo. This biogenic tera: Geometridae)) larvae were reared on silver birch (Betula VOC-based SOA growth mechanism produces roughly 50% of pendula) seedlings from eggs collected from E. autumnata aerosol particles at the size of CCN (30−100 nm38) across reared in Finnish Lapland (the Kevo Subarctic Research Europe.37 There is evidence of the existence of a negative Station, 69°45′N, 27°01′E). For feeding experiments on aerosol climate feedback mechanism in the continental mountain birch, larvae representing the third and fourth instars biosphere, which is directly dependent on biogenic VOC were used. emissions.37,39 A link between herbivore-induced VOCs and Plant Physiological Measurements. Whole plant was formation of SOA in the atmosphere was suggested by placed in a dynamic headspace sampling cuvette system with a Holopainen40 and demonstrated with elicitor-stimulated setup described in detail earlier.47,48 Light intensity was kept at μ −2 −1 ° herbivore-induced VOC emissions of cabbage plants by 800 mol m s , temperature at 28 C, and ambient CO2 Joutsensaari et al.41 Recently it has been suggested that climate concentration was 380−400 μmol mol−1. The air flow through warming may substantially promote the emission of reactive chamber was 1.4 L min−1. After hermetic installation of the HIPVs from boreal and subarctic forests through insect plants they were stabilized under chamber conditions for 1 h. ff outbreak e ects and promote SOA and cloud formation by CO2 and H2O concentrations at the chamber in- and outlets herbivore-induced VOC reactions with atmospheric oxi- were measured with an infrared dual-channel gas analyzer dants.14,36,42 Hence, a modeling analysis of bark beetle outbreak operated in differential mode (CIRAS II, PP-systems, Ames- areas in Northern America demonstrated that the substantial bury, MA, U.S.A.). The rates of net assimilation, transpiration, increase of reactive VOC emissions from bark beetle-attacked and stomatal conductance to water vapor were calculated from conifer forests may significantly increase the atmospheric SOA these measurements according to von Caemmerer and concentrations in the affected areas.43 Aerosol optical density Farquhar.49 By the end of the stabilization period (time 0), analyses of satellite data from Western Canada44 supported the the treatment was started by placing 5 larvae of E. autumnata modeling results of Berg et al.43 However, the composition of on the plants. Control plants were left untreated and measured

11502 DOI: 10.1021/acs.est.6b02800 Environ. Sci. Technol. 2016, 50, 11501−11510 Environmental Science & Technology Article in the same way. After introduction of the larvae, the plants target concentration set at 200 nmol mol−1) at the inlet of the were stabilized in the system in the dynamic air flow for 20 min reactor. The average ozone concentration inside the reaction before VOC sampling. The larvae were allowed to feed from chamber varied between 65 and 90 nmol mol−1 during the the leaves for 24 h. experiments. The total air flow into the reaction chamber was Herbivore-Induced VOC Emission Measurements. The 16 L min−1 giving an average residence time of 2 h in the impact of E. autumnata larval density on the emission rates was chamber. The chamber was kept at constant room temperature assessed using one-year-old nonbranching seedlings by (23 ± 1 °C) during the experiment. Relative humidity in the enclosing the above-ground part of each seedling together chamber varied between 30 and 35% depending on the water with various numbers (0, 4, 8, or 16) of E. autumnata third to vapor coming from the seedlings. Before each experiment, the fourth instar larvae inside 3 L glass chambers for GC-MS chamber was flushed continuously with purified dry air (dew measurements of emitted volatiles as described by Copolovici point −40 °C, filtered by active carbon and Purafil Select Media et al.8 In the 24-h experiment, larvae were allowed to feed for 7 pellets and HEPA filter) at least for 24 h to ensure minimal h after which they were removed. In the long term experiment contamination from previous experiments. (0−3 days), larvae were removed after 2.2 days and the light At the beginning of each experimental run, VOCs from cycle was 14 h light followed by 8 h darkness.VOC sampling seedlings were introduced to the reaction chamber for 90 min was performed via the outlets of each cuvette every hour with a before starting the ozone addition. The duration of the flow rate of 200 mL min−1 for 20 min by using a constant flow chamber experiment was ca. 21 h. The plants’ diurnal cycle was air sample pump (1003-SKC, SKC Inc., Houston, TX, U.S.A.). mimicked by turning off the lights over the plant chambers In addition, a sample was taken from the air inlet prior to the from 00:00 to 03:00. In all experiments, concentrations of cuvettes to measure the background VOC concentrations. ozone (measured with a DASIBI 1008-RS O3 analyzer, Dasibi Adsorbent cartridges were analyzed for LOX pathway products, Environmental Corp. Glendale, CA, U.S.A.), NOx (with mono-, homo- and sesquiterpene emissions with a combined Environnement AC 30 M NOx analyzer, Environnement S.A, Shimadzu TD20 automated cartridge desorber and Shimadzu Poissy, France) and SO2 (with Environnement AF21 M SO2 2010 plus GC-MS instrument (Shimadzu Corporation, Kyoto, analyzer, Environnement S.A., Poissy, France) were monitored. Japan) using a method as detailed in refs 16 and 50. The NOx and SO2 concentrations were below the analyzers background (blank) VOC concentrations were subtracted from quantification limits (3 nmol mol−1) during the experiments. emission samples of the seedlings. Separate samples were VOC concentrations in the ingoing (before the reaction collected from larvae without plants. The emissions of volatile chamber) and outgoing (after the reaction chamber) air were compounds from insects were below the detection limits of our measured by GC-MS. The VOCs were trapped on a Tenax TA system (see Copolovici et al.50). At the end of the experiment, (Supelco, mesh 60/80) cartridge with an air flow of 200 mL all the leaves were harvested and scanned at 300 dpi. Projected min−1 for 30 min. For the outgoing air, the sampled air passed leaf area was determined from the scanned images with custom- first through a potassium iodine-coated copper tube to scrub made software (see Copolovici et al.8). the ozone. The trapped compounds were desorbed from the SOA Formation Measurements. Six third to fourth instar collected VOC samples with a thermal desorption unit E. autumnata larvae were transferred to shoots of mountain (PerkinElmer ATD400 Automatic Thermal Desorption sys- birch seedlings and enclosed within mesh bags and kept in tem) and analyzed with a gas chromatograph−mass spec- growth chambers (Bioklim 2600T, Kryo-Service Oy, Helsinki) trometer (Hewlett-Packard GC 6890 and MSD 5973). A at 23 °C (day)/18 °C (night), 60−80% relative humidity, and detailed description of the VOC analysis can be found in Pinto 22 h photoperiod (350 μmol m−2 s−1 photosynthetically active et al.52 quantum flux density during the light period). Larvae were The ozonolysis of BVOCs resulted in a clear particle allowed to feed for 48 h before the start of the SOA nucleation and growth event which was monitored with a experiments to maximize the production of herbivore-induced scanning mobility particle sizer (SMPS, consisting of DMA VOCs.51 After the mesh bags were removed, eight control 3071 and CPC 3022A, TSI Inc., St. Paul, MN, U.S.A.) by seedlings or eight herbivore-damaged seedlings (each seedling measuring particle number size distribution in the diameter still having 6 feeding larvae) were transferred into each of two range of 15 to 700 nm. The total particle concentration 22.3 L glass vessels (Schott Duran, Schott AG, Mainz, (particle diameter >4 nm) was monitored with a condensation Germany). The plant pots were covered with aluminum foil particle counter (CPC 3775, TSI Inc.). The total mass to minimize root system emissions. Two lamps (Lival Shuttle concentration of the produced SOA was calculated from the Plus, Lival Oy, Sipoo, Finland, with OSRAM Dulux F 24W/ measured number size distribution using a particle density of 41−827 Fluorescent tubes, Osram-Melco Ltd., Japan) were 1.3 g cm−353 and corrected particle wall losses (see Supporting positioned above each vessel to ensure sufficient lighting level Information, SI). The SOA mass yields were estimated by (350 μmol m−2 s−1) for photosynthesis activity and VOC dividing the formed SOA mass (maximum SOA mass) by the emissions at laboratory conditions (24 °C). sum of reacted VOC concentrations (the total VOC The SOA formation experiments from the ozonolysis of concentration at the reactor inlet minus the concentration at mountain birch emitted VOCs were carried out in a rectangular the outlet).54 The experiments were conducted without seed continuous flow-through chamber made of fluorinated ethylene particles, which could lower SOA mass yields by increasing loss propylene (FEP) film. The experimental setup and the chamber of low volatile organics to the chamber walls.55 The effect of have been described elsewhere.44 Briefly, the reaction chamber VOC losses on SOA mass yields has been discussed in more volume was 2 m3. Air flow (8 L min−1) from the two glass detail in the SI. vessels (later on referred to as plant chambers), each having These exactly same experimental settings were repeated in either 8 control plants or 8 herbivore-damaged plants, was another set of experiments, where silver birch (Betula pendula) pooled to one line (Teflon tubing) and mixed with an ozone- seedlings were used as plant material. Results of these enriched air flow (generated with a UV lamp generator, with a experiments are shown in SI Tables S5−S7 and Figure S3.

11503 DOI: 10.1021/acs.est.6b02800 Environ. Sci. Technol. 2016, 50, 11501−11510 Environmental Science & Technology Article

Figure 1. Emission rate time series, 0−24 h, for different VOCs from herbivore-attacked and control plants. LOX products from herbivore-attacked (a) and control (b) B. pubescens ssp. czerepanovii plants, different monoterpenoids and homoterpene (E)-DMNT from herbivore-attacked (c) and control (d) plants and different sesquiterpenes from herbivore-attacked (e) and control (f) plants. Each data point is the mean (±SE) of 4 independent replicate experiments. At hour 0, the plants without the insects were measured. The insects were added at time t = 0.5 h and allowed to feed from the leaves for 7 h. After that larvae were removed and plants were measured without them.

Statistical Analysis. Plant physiological responses of especially α-farnesene, were induced after 3 h from the start of control and herbivore-damaged plants were compared by t the insect feeding (Figure 1e). No comparable increases in the tests and considered significant at p < 0.05 (Prism 5, GraphPad VOC emissions were observed in the control plants (Figure Software, Inc., La Jolla, CA, U.S.A.). Long-term (3 days) 1b,d,f). When plant physiological responses were monitored, variation in emission rates was studied with Mixed Model larval feeding on mountain birch leaves did not affect plant ANOVA (MMA) which takes into account the repetitive photosynthesis (see Figure S1). structure of the measurements and thus gives valid Over the longer term (3 days), the larval density was interpretations for the between group comparisons. Nonlinear positively correlated with the emission of LOX products, regression analysis was used to study the emission rates of monoterpenes and (E)-DMNT (Figure 2). Among the different compounds in relation to leaf area consumed by E. individual LOX products, emission rates of (Z)-3-hexenol, autumnata larvae. This analysis was made in R-software (R (E)-2-hexenal, and (Z)-3-hexenyl acetate were all dependent on Core Team, 2014). Concentrations of VOCs emitted by larval density (Figure 2a−c). LOX products are excellent mountain birch at the reaction chamber inlet in the SOA indicators of leaf mechanical damage, e.g., caused by chewing − formation experiments were compared with Mann Whitney mouth parts of herbivores.56,57 These emissions are short-lived − U-test. The Mann Whitney U-test and MMA analyses were and emerge immediately when the feeding starts and disappear performed in SAS 9.4 (SAS Institute, Cary NC). quickly when the feeding activity ends.57 However, emission of monoterpenes and the homoterpene ■ RESULTS AND DISCUSSION (E)-DMNT were more variable and were not quite as obviously We have shown that the VOC emission rate from herbivore- dependent on larval density as the LOX products (Figure 2d,e). damaged plants is related to the extent of the damaged leaf area For these compounds, eight larvae induced emissions nearly and the number of damaging herbivores. Aerosol measurement equal to the emission rates induced by 16 larvae. However, all ff fi in reaction chambers gave evidence that the quantitative and tested larval densities di ered statistically signi cantly from the qualitative changes in VOC emissions of mountain birch after control (Mixed Model ANOVA analysis pairwise comparisons herbivory are sufficient to promote secondary aerosol formation are shown in Table S1). In general, the emission rates could be in the atmosphere. expressed as a nonlinear function of leaf area eaten (see Figure Impact of Herbivore Density on VOC Emissions. In the S2). 24 h feeding data (Figure 1), emissions of different compounds Our results confirmed the fact that herbivore-damaged plants were induced with different time kinetics. The first major with clearly diminished photosynthetic leaf area are still capable emissions observed were LOX products, which were induced of emitting VOCs at a higher rate than healthy plants with immediately after the onset of larval feeding (Figure 1a) and intact foliage.19,58 Emission rates of the induced LOX stayed at higher levels for at least 7 h. Instead, the emissions of compounds, monoterpene (E)-β-ocimene and sesquiterpene monoterpenes did not show such a rapid response to larval (E,E)-α-farnesene were also positively correlated with the feeding; a slight increase in emission rates was detected just number of larvae (0 to 8) feeding on the foliage of Alnus after the start of feeding at t = 0.5 h, but the maximum emission glutinosa seedlings.8,59 In our experiments, B. pubescens ssp. rate was observed at t =5h(Figure 1c). Sesquiterpenes, czerepanovii seedlings gave very similar results, but increasing

11504 DOI: 10.1021/acs.est.6b02800 Environ. Sci. Technol. 2016, 50, 11501−11510 Environmental Science & Technology Article

birch seedlings, two with control plants and two with herbivore- damaged plants (Table 1). Our results show that the seedlings damaged by E. autumnata larvae had a significantly higher potential for SOA formation when compared with intact seedlings (Table 1, Figure 3). The VOCs (monoterpenes, a homoterpene, sesquiterpenes, LOX products, and methyl salicylate), acting as precursors to SOA formation, were emitted in higher rates in the herbivore-damaged mountain birch seedlings than in the control seedlings (Table 1). The same phenomenon was observed for a boreal forest species, silver birch (Betula pendula) seedlings in the same experimental settings (see Tables S5 and Figure S3). In addition to the greater emission strength, the emission profile was more varied in the damaged seedlings vs the controls, i.e., there were more individual terpenoid compounds identified in the emissions from damaged seedlings than in that of the intact seedlings (see Table S2). The greatest differences in the levels of VOCs emitted were seen in the LOX compounds: in the SOA formation experiments, LOXs were present only in the damaged samples. The main LOX compounds measured were the typical C6 aldehydes and alcohols, e.g., (E)-2-hexenal and (Z)-3-hexen-1-ol. In our control plants, VOC emissions were similar to those earlier reported from mountain birch; e.g., limonene, sabinene, and linalool of the MTs and β-caryophyllene and α-farnesene of the SQTs,32,33 while herbivore damage resulted in enhanced emissions of these ubiquitous terpenoids and additional emissions of (E)-β-ocimene, (E)-DMNT, and LOX com- pounds as also reported earlier by Mantylä ̈et al.33 In our small- Figure 2. Long-term, days 0−3, variation in emission rates of VOCs at varying herbivore damage severity. Figures represent LOX compounds sized herbivore-damaged seedlings, higher MT emissions than SQT emissions were detected, which is different from the (E)-2-hexenal (a), (Z)-3-hexenol (b), (Z)-3-hexenyl acetate (c), 33 homoterpene (E)-DMNT (d), and monoterpenes (e) from the foliage observation in older mountain birch ramets in a forest site. of B. pubescens ssp. czerepanovii at different feeding densities of E. While the overall VOC emissions were greater in the autumnata larvae. Sample size and measurement conditions as in damaged seedlings than in the control seedlings, there was a Figure 1. Larvae were removed after 2.2 days. The light cycle was 14 h difference between the two controls. Especially the SQT and light followed by 8 h darkness. MeSa emissions were higher in Control 2 than in either of the insect-damaged samples (see Table 1 and Table S2). Higher the larvae number from 8 to 16 per seedling did not SQT and MeSa emissions observed in Control 2 may indicate considerably increase emissions rates of VOCs. This might be unnoticed fungal pathogen attack or damage by sucking related to a very rapid decline in the whole plant leaf area in herbivores such as aphids62 in the seedlings or in their root these two treatments with high insect density. The presence of system. Our SI of B. pendula seedlings in similar conditions a quantitative relationship between VOC emissions and the (Tables S5−S7) suggests that intact Betula sp. seedlings do not degree of herbivore feeding is not only conceptually important, emit MeSA. However, the differences in SQT emissions could but also provides a basis for quantitative assessment of induced also have been due to natural variability as significant emissions in large-scale biosphere−atmosphere models.60,61 fluctuation in VOC emissions from visibly intact mountain Epirrita autumnata Feeding-Induced VOC Emissions birch in a natural growing site has been reported by Haapanala in Relation to SOA Formation. Four chamber experiments et al.32 The emissions of SQT and MeSa in Control 2 were conducted to see SOA forming potential of mountain experiments may have caused the higher SOA formation since

Table 1. Concentrations of Volatiles in the Reaction Chamber Inlet Emitted by B. pubescens ssp. czerepanovii Plants and Summary of the Resulting Secondary Organic Aerosol (SOA) Concentrations and Mass Yields

MTa SQTa LOXa MeSaa Δ(MT+SQT)a ΔLOXa ΔMeSaa SOA yielda exp. (μgm−3) (μgm−3) (μgm−3) (μgm−3) (μgm−3) (μgm−3) (μgm−3) SOAa (μgm−3) (%) control 1 1.1 ± 0.3 0.3 ± 0.02 0 0 1.2 ± 0.2 0 0 0.008 ± 0.004 0.6 ± 0.4 control 2 1.6 ± 0.3 2.2 ± 0.4 0 0.7 ± 0.3 3.5 ± 0.8 0 0.3 ± 0.1 0.168 ± 0.025 4.4 ± 1.2 herbivore 1 2.7 ± 0.4 0.5 ± 0.1 27 ± 17 0.1 ± 0.1 2.7 ± 0.4 24 ± 16 0.1 ± 0.1 0.431 ± 0.030 1.6 ± 1.0 herbivore 2 2.9 ± 0.2 1.1 ± 0.2 6.4 ± 2.4 0.3 ± 0.1 3.2 ± 0.5 5.9 ± 2.5 0.3 ± 0.1 1.248 ± 0.060 13.3 ± 4.4 aThe data for monoterpenes (MT), sesquiterpenes (SQT), lipoxygenase pathway products (LOX), and methyl salicylate (MeSa) are averages (±SE) for the duration of the experiment (3−5 samples in ca. 21 h). Homoterpene, (E)-4,8-dimethyl-1,3,7-nonatriene, was added to MTs for simplicity. The Δ(MT+SQT), ΔLOX, and ΔMeSa shows the concentration of the reacted MT, SQT, LOX, and MeSa compounds. The produced SOA was corrected with particle wall losses (see SI) and calculated using the maximum total particle volume and particle density of 1.3 g cm−3. The SOA yields are expressed as SOA divided by the sum of the Δ(MT+SQT), ΔLOX, and ΔMeSa.

11505 DOI: 10.1021/acs.est.6b02800 Environ. Sci. Technol. 2016, 50, 11501−11510 Environmental Science & Technology Article

Figure 3. Formation of secondary organic aerosols (SOA) from the ozonolysis of VOCs emitted from mountain birch control (a−d) and herbivore- attacked (e−h) seedlings. SOA formation is shown as particle size distribution (a,b; e,f); total particle number (c,g) and total particle mass concentration (d,h) as a function of time. Total particle mass was calculated from the measured size distributions using particle density of 1.3 g cm−3. VOCs were fed into the reaction chamber for 1.5 h before the onset of the experiment (approximately 15:00). Herbivore-attacked experiment 2 was started 1 h earlier than others (approximately 14:00). The plant chamber illumination was off from 00:00 to 03:00 (indicated by red vertical lines). these compounds have high SOA formation potential,42,63 as SOA formation petered out when the lights were off in the observed in the SOA formation experiment (discussed later). plant chamber (00:00−03:00 am), and commenced again when The clear differences in SOA production observed between the lights were turned on (Figure 3). Our diurnal experiment the controls and herbivory experiments reflected the differences thus demonstrates that the formation of SOA from mountain in the VOC emissions of the samples. While MT and SQT birch VOC emissions is not only dependent on the severity of emissions in the herbivore-damaged seedlings were approx- herbivore damage but it is also strongly light-dependent. This is imately twice those from intact seedlings, and greater perhaps not surprising as monoterpene emissions from broad- differences were seen in SOA formation (Table 1, Figure 3). leaved species lacking storage structures are in general typically The SOA mass loading was as much as 2 orders of magnitude photosynthesis- and light-dependent (e.g., refs 1 and 60). Even higher in herbivore-damaged experiments compared to Control in species having constitutive emissions from storage structures, − 1 where the SOA formation was very low (0.01 μgm 3). While the induced terpenoid emissions are characteristically depend- − − some SOA production (0.17 μgm 3) was also observed in the ent on light.6,69 71 However, in the case of compounds Control 2, the SOA mass loadings in the herbivory samples immediately released after herbivory such as LOX compounds, − were 3 to 7-fold higher (0.43−1.25 μgm3). The mass of the emissions can be importantly driven by insect feeding − produced SOA was moderate, less than 1.3 μgm3 for both habits. Thus, predicting SOA formation due to herbivory-driven control and herbivory samples. The main reasons for these low VOC emissions requires consideration of both plant physio- SOA mass loadings are low initial VOC concentrations and low logical responses to light and insect-specific feeding patterns. consumption of precursor VOCs. However, the VOC and Our chamber experiments showed that herbivore-damage organic particle mass concentrations used and observed in our induced by six E. autumnata larvae per seedling can increase the study, as well as particle formation and growth rates (see Table SOA formation potential of mountain birch due to increases in S3), are coincident with atmospheric concentrations in forested VOC emissions. Thus, our study expands the earlier findings − remote areas (e.g., in Hyytialä,̈ Finland).64 67 We suggest that which show that stress, both biotic and abiotic, can increase the our results better represent natural conditions than the results VOC emissions and concurrent SOA formation in from many other chamber studies, which have been conducted plants.41,42,44,52 However, in the view of several previous with much higher VOC and organic particulate mass levels. studies, the effect of the VOC composition on SOA production Furthermore, our SOA yields are in line with reported α-pinene is less clear.42,45,46 yields by Shilling et al.54 at low mass loadings (below 2 μg In an earlier study by Mentel et al.,72 SOA particle formation m−3). In fact, there is a need to collect SOA formation data in rate and particle growth rate were found to be more efficient atmospherically relevant organic aerosol mass loadings (ref 68 when the VOC emissions came from silver birch (B. pendula) and references therein). than when the VOC emissions originated from conifer trees.

11506 DOI: 10.1021/acs.est.6b02800 Environ. Sci. Technol. 2016, 50, 11501−11510 Environmental Science & Technology Article

Thus, they suggested that the oxidation of birch emitted SOA main oxidant in the current study, and it could be anticipated precursors can favor new particle formation. In their experi- that not all LOX products would have been oxidized. However, ments, an incremental mass yield of 11 ± 1% considering only OH radicals are formed in reactions between terpenes and the MT+SQT emissions from birch was observed. Our results ozone in yields between 0.3 to 0.9 in relation to the reacted presented in this study show yields of approximately the same alkene.79 Since we did not use any OH scavengers in our level in the herbivore-damaged seedling experiments, whereas experiments, OH radicals were therefore also available for the levels in our controls (especially control 1) were somewhat oxidation reactions in our experimental setting. Oxidation of lower. LOX products could also have a role in the SOA formation, as It has also been speculated that the mass yield can be both (Z)-3-hexen-1-ol and (Z)-3-hexenyl acetate have been significantly influenced by fractions of oxygenated VOCs (LOX shown to be consumed by reactions either with ozone or with compounds and MeSa) and SQT in the emissions. In the later OH-radicals formed in ozonolysis experiments. These LOX study, Mentel et al.42 studied the effect of VOC profiles on products are capable of forming oxidized compounds with SOA formation further and they came to somewhat differing vapor pressure low enough to take part of SOA formation and conclusion on the effect of LOX on SOA:LOX compounds growth, although the yields are rather low when compared to inhibit SOA formation. They suspected that the suppressive MT or SQT SOA forming potential.73,75 effect of LOX compounds on SOA formation is related to LOX To summarize, herbivore damage and the severity of the capacity to scavenge OH radicals. When comparing our damage intensity has strong influence on the quantity and herbivore-damaged replicates 1 and 2, we found that replicate composition of mountain birch VOC emissions and con- 1 had higher LOX emission rate but lower SOA mass loading. sequent SOA formation. Hence, in extensive E. autumnata or This finding could support the later hypothesis by Mentel et other birch-feeding geometrid outbreak areas in the subarctic, al.42 However, some other studies have found that LOX we may expect to find increased SOA formation rates in the compounds reacting with ozone can be a substantial source of atmosphere. Although there is evidence that LOX compounds 73−75 emitted after herbivore attack on plant foliage may induce SOA SOA globally. They have speculated that ozonolysis of − (E)-3-hexenol leads to significant formation of oligomers,73,75 formation,73 75 our results also gave support to the results of and this oligomerization mechanism has been shown to Mentel et al.42 that LOX emissions may suppress SOA contribute significantly to SOA formation and it could also formation in mixtures with terpenes. Future studies are needed have an effect on the chemical and physical properties of SOA to elucidate the role of LOX compounds in SOA formation in particles (ref 68 and references therein). In our experiments, field conditions. However, chewing herbivores have such a LOX emissions dominated the VOC composition in the strong impact on mono-, sesqui-, and homoterpene emissions herbivore-damaged samples, whereas they were almost absent and methyl salicylate emissions of mountain birch that higher a in the control experiments. If LOX compounds completely SOA formation rate is expected with outbreaks of defoliating inhibited the SOA formation as Mentel et al.42 suggested, then moths. We argue that additional SOA formation due to insect- we probably would not have been able to observe such a clear induced VOC emissions can affect the important sun-screening SOA formation in these samples. However, LOX compounds biosphere−atmosphere feedback system of aerosols under could have some suppressive effect on SOA formation as the subarctic 24-h sunshine periods. differences between our herbivory experiments 1 and 2 could be explained by the higher amount of produced and reacted ■ ASSOCIATED CONTENT LOX compounds. In conclusion, SOA formation yields from *S Supporting Information ozonolysis of LOX compounds, e.g., (E)-3-hexenol and (E)-3- The Supporting Information is available free of charge on the hexenyl acetate, are lower than yields reported for MT or SQT ACS Publications website at DOI: 10.1021/acs.est.6b02800. 73,74,76,77 only, which indicates that LOX compounds can partly Particle and VOC losses in the chamber, plant ff inhibit SOA formation and growth in mixtures of di erent physiological responses, emission rates of LOX com- VOCs. pounds and monoterpenoids in relation to leaf area Note that there were a variety of both MT and LOX ff ’ 42 consumed, pairwise comparisons of di erent VOC s compounds present in our experiment, whereas Mentel et al. between the larvae densities in different time points, α performed their experiments only with single compounds ( - concentrations of volatiles in the reaction chamber inlet pinene as a monoterpene and (Z)-3-hexenol as a LOX and outlet, and average particle growth and formation compound). It is thus likely that the reactant composition rates in the SOA formation experiments and results of ff has an e ect on the inhibitory potential and therefore the Betula pendula experiments (PDF) phenomenon is not quite as straightforward as suggested. On − average, 80% of the MT and 90 100% of the SQT were ■ AUTHOR INFORMATION consumed in the reaction chamber at the steady state Corresponding Author conditions in our experiments. In contrast, there were * fi substantial differences in the consumption of the individual E-mail: pasi.yli-pirila@uef. (P.Y.-P). LOX compounds; the consumption ranging from only 10% up Notes to 90% in the chamber (see Table S4). On average, the The authors declare no competing financial interest. consumption of the LOX products was lower than the consumption of other VOC groups measured (i.e., MT, SQT, ■ ACKNOWLEDGMENTS HT, MeSa). One reason for this could be the preferred oxidant The study was financially supported by the Academy of Finland of the LOX compounds; LOX undergo oxidation more readily (decision nos. 252908, 110763, and 133322), the strategic with OH radical than with ozone. For example, (E)-2-hexenal funding of the University of Eastern Finland, the Kone reaction with ozone is negligible but it reacts quickly with OH- Foundation, the Estonian Ministry of Science and Education radicals (ref 78 and references therein). Ozone was used as the (institutional grant IUT-8-3), the European Commission

11507 DOI: 10.1021/acs.est.6b02800 Environ. Sci. Technol. 2016, 50, 11501−11510 Environmental Science & Technology Article through the European Regional Fund (the Center of Excellence (19) Ghimire, R. P.; Markkanen, J. M.; Kivimäenpää,M.; in Environmental Adaptation), and the European Research Lyytikainen-Saarenmaa,̈ P.; Holopainen, J. K. Needle Removal by Council (advanced grant 322603, SIP-VOL+). Pine Sawfly Larvae Increases Branch-Level VOC Emissions and Reduces Below-Ground Emissions of Scots Pine. Environ. Sci. Technol. 2013, 47 (9), 4325−4332. ■ REFERENCES (20) Blande, J. D.; Tiiva, P.; Oksanen, E.; Holopainen, J. K. Emission (1) Loreto, F.; Schnitzler, J.-P. Abiotic stresses and induced BVOCs. of herbivore-induced volatile terpenoids from two hybrid aspen Trends Plant Sci. 2010, 15 (3), 154−166. (Populus tremula x tremuloides) clones under ambient and elevated (2) Peñuelas, J.; Staudt, M. BVOCs and global change. Trends Plant ozone concentrations in the field. Global Change Biology 2007, 13 − Sci. 2010, 15 (3), 133−144. (12), 2538 2550. (3) Possell, M.; Loreto, F. The Role of Volatile Organic Compounds (21) Ghirardo, A.; Heller, W.; Fladung, M.; Schnitzler, J. P.; in Plant Resistance to Abiotic Stresses: Responses and Mechanisms. In Schroeder, H. Function of defensive volatiles in pedunculate oak Biology, Controls and Models of Tree Volatile Organic Compound (Quercus robur) is tricked by the moth Tortrix viridana. Plant, Cell − Emissions, Niinemets, Ü.; Monson, R. K., Eds.; Springer: Berlin, 2013; Environ. 2012, 35 (12), 2192 2207. pp 209−235. (22) Tiiva, P.; Faubert, P.; Michelsen, A.; Holopainen, T.; ̈ Holopainen, J. K.; Rinnan, R. Climatic warming increases isoprene (4) Niinemets, U.; Loreto, F.; Reichstein, M. Physiological and − physicochemical controls on foliar volatile organic compound emission from a subarctic heath. New Phytol. 2008, 180 (4), 853 863. emissions. Trends Plant Sci. 2004, 9 (4), 180−186. (23) Faubert, P.; Tiiva, P.; Rinnan, A.; Michelsen, A.; Holopainen, J. (5) Niinemets, Ü.; Monson, R. K.; Arneth, A.; Ciccioli, P.; K.; Rinnan, R. Doubled volatile organic compound emissions from Kesselmeier, J.; Kuhn, U.; Noe, S. M.; Peñuelas, J.; Staudt, M. The subarctic tundra under simulated climate warming. New Phytol. 2010, 187 (1), 199−208. leaf-level emission factor of volatile isoprenoids: caveats, model ̈ algorithms, response shapes and scaling. Biogeosciences 2010, 7 (6), (24) Niemela, P.; Chapin, F. S., III; Danell, K.; Bryant, J. P. Herbivory-mediated responses of selected boreal forests to climatic 1809−1832. 2001 − (6) Niinemets, Ü.; Arneth, A.; Kuhn, U.; Monson, R. K.; Penuelas, J.; change. Clim. Change , 48 (2), 427 440. (25) Kurz, W. A.; Dymond, C. C.; Stinson, G.; Rampley, G. J.; Staudt, M. The emission factor of volatile isoprenoids: stress, Neilson, E. T.; Carroll, A. L.; Ebata, T.; Safranyik, L. Mountain pine acclimation, and developmental responses. Biogeosciences 2010, 7 (7), beetle and forest carbon feedback to climate change. Nature 2008, 452 2203−2223. (7190), 987−990. (7) Dicke, M.; van Loon, J. J. A.; Soler, R. Chemical complexity of (26) Jepsen, J. U.; Kapari, L.; Hagen, S. B.; Schott, T.; Vindstad, O. P. volatiles from plants induced by multiple attack. Nat. Chem. Biol. 2009, L.; Nilssen, A. C.; Ims, R. A. Rapid northwards expansion of a forest 5 (5), 317−324. insect pest attributed to spring phenology matching with sub-Arctic (8) Copolovici, L.; Kannaste,̈ A.; Remmel, T.; Vislap, V.; Niinemets, birch. Global Change Biology 2011, 17 (6), 2071−2083. Ü. Volatile Emissions from Alnus glutionosa Induced by Herbivory are (27) Ammunet,́ T. E. A.; Kaukoranta, T.; Saikkonen, K.; Repo, T.; Quantitatively Related to the Extent of Damage. J. Chem. Ecol. 2011, − Klemola, T. Invading and resident defoliators in a changing climate: 37 (1), 18 28. cold tolerance and predictions concerning extreme winter cold as a (9) Holopainen, J. K.; Blande, J. D. Where do herbivore-induced − − range-limiting factor. Ecological Entomology 2012, 37 (3), 212 220. plant volatiles go? Front. Plant Sci. 2013, 4 (185), 1 13. (28) Jepsen, J. U.; Hagen, S. B.; Ims, R. A.; Yoccoz, N. G. Climate (10) Holopainen, J. K.; Nerg, A.-M.; Blande, J. D. Multitrophic change and outbreaks of the geometrids Operophtera brumata and Signalling in Polluted Atmospheres. In Biology, Controls and Models of ̈ Epirrita autumnata in subarctic birch forest: evidence of a recent Tree Volatile Organic Compound Emissions; Niinemets, U.; Monson, R. − − outbreak range expansion. J. Anim. Ecol. 2008, 77 (2), 257 264. K., Eds.; Springer: Berlin, 2013; pp 285 314. (29) Jepsen, J. U.; Biuw, M.; Ims, R. A.; Kapari, L.; Schott, T.; (11) Kessler, A.; Baldwin, I. T. Defensive function of herbivore- Vindstad, O. P. L.; Hagen, S. B. Ecosystem Impacts of a Range induced plant volatile emissions in nature. Science 2001, 291 (5511), Expanding Forest Defoliator at the Forest-Tundra Ecotone. Ecosystems − 2141 2144. 2013, 16 (4), 561−575. (12) Blande, J. D.; Holopainen, J. K.; Li, T. Air pollution impedes (30) Vindstad, O. P. L.; Schott, T.; Hagen, S. B.; Jepsen, J. U.; Kapari, plant-to-plant communication by volatiles. Ecology Letters 2010, 13 L.; Ims, R. A. How rapidly do invasive birch forest geometrids recruit − (9), 1172 1181. larval parasitoids? Insights from comparison with a sympatric native (13) Dicke, M.; Baldwin, I. T. The evolutionary context for geometrid. Biological Invasions 2013, 15 (7), 1573−1589. ’ ’ herbivore-induced plant volatiles: beyond the cry for help . Trends (31) Hamet-Ahti,̈ L.; Palmen, A.; Alanko, P.; Tigersted, P. Woody − Plant Sci. 2010, 15 (3), 167 175. Flora of Finland; Yliopistopaino: Helsinki, 1992. (14) Holopainen, J. K.; Gershenzon, J. Multiple stress factors and the (32) Haapanala, S.; Ekberg, A.; Hakola, H.; Tarvainen, V.; Rinne, J.; − emission of plant VOCs. Trends Plant Sci. 2010, 15 (3), 176 184. Hellen, H.; Arneth, A. Mountain birch - potentially large source of ̈̈ (15) Vuorinen, T.; Nerg, A.-M.; Syrjala, L.; Peltonen, P.; Holopainen, sesquiterpenes into high latitude atmosphere. Biogeosciences 2009, 6 J. K. Epirrita autumnata induced VOC emission of silver birch differ (11), 2709−2718. from emission induced by leaf fungal pathogen. Arthropod-Plant (33) Mantylä ,̈ E.; Alessio, G. A.; Blande, J. D.; Heijari, J.; Holopainen, Interactions 2007, 1 (3), 159−165. ̈ J. K.; Laaksonen, T.; Piirtola, P.; Klemola, T. From Plants to Birds: (16) Toome, M.; Randjarv, P.; Copolovici, L.; Niinemets, U.; Higher Avian Predation Rates in Trees Responding to Insect Heinsoo, K.; Luik, A.; Noe, S. M. Leaf rust induced volatile organic Herbivory. PLoS One 2008, 3 (7), e2832. compounds signalling in willow during the infection. Planta 2010, 232 (34) Yuan, J. S.; Himanen, S. J.; Holopainen, J. K.; Chen, F.; Stewart, (1), 235−243. C. N. Smelling global climate change: mitigation of function for plant (17) Jansen, R. M. C.; Wildt, J.; Kappers, I. F.; Bouwmeester, H. J.; volatile organic compounds. Trends in Ecology & Evolution 2009, 24 Hofstee, J. W.; van Henten, E. J. Detection of Diseased Plants by (6), 323−331. Analysis of Volatile Organic Compound Emission. In Annual Review of (35) Arneth, A.; Niinemets, U. Induced BVOCs: how to bug our Phytopathology; VanAlfen, N. K.; Bruening, G.; Leach, J. E., Eds. 2011; models? Trends Plant Sci. 2010, 15 (3), 118−125. Vol. 49, pp 157−174.10.1146/annurev-phyto-072910-095227 (36) Holopainen, J. K. Can forest trees compensate for stress- (18) Heijari, J.; Blande, J. D.; Holopainen, J. K. Feeding of large pine generated growth losses by induced production of volatile weevil on Scots pine stem triggers localised bark and systemic shoot compounds? Tree Physiol. 2011, 31 (12), 1356−1377. emission of volatile organic compounds. Environ. Exp. Bot. 2011, 71 (37) Paasonen, P.; Asmi, A.; Petajä,̈ T.; Kajos, M. K.; Aijala, M.; (3), 390−398. Junninen, H.; Holst, T.; Abbatt, J. P. D.; Arneth, A.; Birmili, W.; van

11508 DOI: 10.1021/acs.est.6b02800 Environ. Sci. Technol. 2016, 50, 11501−11510 Environmental Science & Technology Article der Gon, H. D.; Hamed, A.; Hoffer, A.; Laakso, L.; Laaksonen, A.; (53) Kiendler-Scharr, A.; Zhang, Q.; Hohaus, T.; Kleist, E.; Mensah, Leaitch, W. R.; Plass-Dulmer, C.; Pryor, S. C.; Raisä nen,̈ P.; Swietlicki, A.; Mentel, T. F.; Spindler, C.; Uerlings, R.; Tillmann, R.; Wildt, J. E.; Wiedensohler, A.; Worsnop, D. R.; Kerminen, V. M.; Kulmala, M. Aerosol Mass Spectrometric Features of Biogenic SOA: Observations Warming-induced increase in aerosol number concentration likely to from a Plant Chamber and in Rural Atmospheric Environments. moderate climate change. Nat. Geosci. 2013, 6 (6), 438−442. Environ. Sci. Technol. 2009, 43 (21), 8166−8172. (38) Riipinen, I.; Yli-Juuti, T.; Pierce, J. R.; Petajä,̈ T.; Worsnop, D. (54) Shilling, J. E.; Chen, Q.; King, S. M.; Rosenoern, T.; Kroll, J. H.; R.; Kulmala, M.; Donahue, N. M. The contribution of organics to Worsnop, D. R.; McKinney, K. A.; Martin, S. T. Particle mass yield in atmospheric nanoparticle growth. Nat. Geosci. 2012, 5 (7), 453−458. secondary organic aerosol formed by the dark ozonolysis of α-pinene. (39) Kulmala, M.; Nieminen, T.; Chellapermal, R.; Makkonen, R.; Atmos. Chem. Phys. 2008, 8 (7), 2073−2088. Back,̈ J.; Kerminen, V.-M. Climate Feedbacks Linking the Increasing (55) Kokkola, H.; Yli-Pirila,̈ P.; Vesterinen, M.; Korhonen, H.; Atmospheric CO2 Concentration, BVOC Emissions, Aerosols and Keskinen, H.; Romakkaniemi, S.; Hao, L.; Kortelainen, A.; Clouds in Forest Ecosystems. In Biology, Controls and Models of Tree ̈ Joutsensaari, J.; Worsnop, D. R.; Virtanen, A.; Lehtinen, K. E. J. The Volatile Organic Compound Emissions; Niinemets, U.; Monson, R. K., role of low volatile organics on secondary organic aerosol formation. − Eds.; Springer: Berlin, 2013; pp 489 508. Atmos. Chem. Phys. 2014, 14 (3), 1689−1700. (40) Holopainen, J. K. Multiple functions of inducible plant volatiles. (56) Brilli, F.; Ruuskanen, T. M.; Schnitzhofer, R.; Muller, M.; − Trends Plant Sci. 2004, 9 (11), 529 533. Breitenlechner, M.; Bittner, V.; Wohlfahrt, G.; Loreto, F.; Hansel, A. ̈ (41) Joutsensaari, J.; Loivamaki, M.; Vuorinen, T.; Miettinen, P.; Detection of Plant Volatiles after Leaf Wounding and Darkening by Nerg, A. M.; Holopainen, J. K.; Laaksonen, A. Nanoparticle formation Proton Transfer Reaction ″Time-of-Flight″ Mass Spectrometry (PTR- by ozonolysis of inducible plant volatiles. Atmos. Chem. Phys. 2005, 5, TOF). PLoS One 2011, 6 (5), e20419. − 1489 1495. (57) Maja, M. M.; Kasurinen, A.; Yli-Pirila,̈ P.; Joutsensaari, J.; (42) Mentel, T. F.; Kleist, E.; Andres, S.; Dal Maso, M.; Hohaus, T.; Klemola, T.; Holopainen, T.; Holopainen, J. K. Contrasting responses Kiendler-Scharr, A.; Rudich, Y.; Springer, M.; Tillmann, R.; Uerlings, of silver birch VOC emissions to short- and long-term herbivory. Tree R.; Wahner, A.; Wildt, J. Secondary aerosol formation from stress- Physiol. 2014, 34 (3), 241−252. induced biogenic emissions and possible climate feedbacks. Atmos. − (58) Miresmailli, S.; Gries, R.; Gries, G.; Zamar, R. H.; Isman, M. B. Chem. Phys. 2013, 13 (17), 8755 8770. Population density and feeding duration of cabbage looper larvae on (43) Berg, A. R.; Heald, C. L.; Hartz, K. E. H.; Hallar, A. G.; tomato plants alter the levels of plant volatile emissions. Pest Manage. Meddens, A. J. H.; Hicke, J. A.; Lamarque, J. F.; Tilmes, S. The impact Sci. 2012, 68 (1), 101−107. of bark beetle infestations on monoterpene emissions and secondary (59) Copolovici, L.; Kannaste,̈ A.; Remmel, T.; Niinemets, Ü. organic aerosol formation in western North America. Atmos. Chem. − Volatile organic compound emissions from Alnus glutinosa under Phys. 2013, 13 (6), 3149 3161. interacting drought and herbivory stresses. Environ. Exp. Bot. 2014, (44) Joutsensaari, J.; Yli-Pirila,̈ P.; Korhonen, H.; Arola, A.; Blande, J. 100,55−63. D.; Heijari, J.; Kivimaenpä ä,̈ M.; Mikkonen, S.; Hao, L.; Miettinen, P.; (60) Grote, R.; Monson, R. K.; Niinemets, Ü. Leaf-Level Models of Lyytikainen-Saarenmaa,̈ P.; Faiola, C. L.; Laaksonen, A.; Holopainen, J. Constitutive and Stress-Driven Volatile Organic Compound Emis- K. Biotic stress accelerates formation of climate-relevant aerosols in sions. In Biology, Controls and Models of Tree Volatile Organic boreal forests. Atmos. Chem. Phys. 2015, 15 (21), 12139−12157. Compound Emissions; Niinemets, Ü.; Monson, R. K., Eds.; Springer: (45) Kiendler-Scharr, A.; Wildt, J.; Maso, M. D.; Hohaus, T.; Kleist, Berlin, 2013; pp 315−355. E.; Mentel, T. F.; Tillmann, R.; Uerlings, R.; Schurr, U.; Wahner, A. (61) Niinemets, Ü.; Kannaste,̈ A.; Copolovici, L. Quantitative New particle formation in forests inhibited by isoprene emissions. − patterns between plant volatile emissions induced by biotic stresses Nature 2009, 461 (7262), 381 384. − (46) Kiendler-Scharr, A.; Andres, S.; Bachner, M.; Behnke, K.; Broch, and the degree of damage. Front. Plant Sci. 2013, 4 (262), 1 15. S.; Hofzumahaus, A.; Holland, F.; Kleist, E.; Mentel, T. F.; Rubach, F.; (62) Blande, J. D.; Korjus, M.; Holopainen, J. K. Foliar methyl salicylate emissions indicate prolonged aphid infestation on silver birch Springer, M.; Steitz, B.; Tillmann, R.; Wahner, A.; Schnitzler, J. P.; − Wildt, J. Isoprene in poplar emissions: effects on new particle and black alder. Tree Physiol. 2010, 30 (3), 404 416. formation and OH concentrations. Atmos. Chem. Phys. 2012, 12 (2), (63) Bonn, B.; Moortgat, G. K. Sesquiterpene ozonolysis: Origin of − atmospheric new particle formation from biogenic hydrocarbons. 1021 1030. − (47) Rasulov, B.; Copolovici, L.; Laisk, A.; Niinemets, U. Geophys. Res. Lett. 2003, 30 (11), 1 4. Postillumination Isoprene Emission: In Vivo Measurements of (64) Hakola, H.; Tarvainen, V.; Laurila, T.; Hiltunen, V.; Hellen, H.; Keronen, P. Seasonal variation of VOC concentrations above a boreal Dimethylallyldiphosphate Pool Size and Isoprene Synthase Kinetics − in Aspen Leaves. Plant Physiol. 2009, 149 (3), 1609−1618. coniferous forest. Atmos. Environ. 2003, 37 (12), 1623 1634. (48) Copolovici, L.; Niinemets, Ü. Flooding induced emissions of (65) Zhang, Q.; Jimenez, J. L.; Canagaratna, M. R.; Allan, J. D.; Coe, volatile signalling compounds in three tree species with differing H.; Ulbrich, I.; Alfarra, M. R.; Takami, A.; Middlebrook, A. M.; Sun, Y. waterlogging tolerance. Plant Cell and Environment 2010, 33 (9), L.; Dzepina, K.; Dunlea, E.; Docherty, K.; DeCarlo, P. F.; Salcedo, D.; 1582−1594. Onasch, T.; Jayne, J. T.; Miyoshi, T.; Shimono, A.; Hatakeyama, S.; (49) von Caemmerer, S.; Farquhar, G. D. Some relationships Takegawa, N.; Kondo, Y.; Schneider, J.; Drewnick, F.; Borrmann, S.; between the biochemistry of photosynthesis and the gas exchange of Weimer, S.; Demerjian, K.; Williams, P.; Bower, K.; Bahreini, R.; leaves. Planta 1981, 153 (4), 376−387. Cottrell, L.; Griffin, R. J.; Rautiainen, J.; Sun, J. Y.; Zhang, Y. M.; (50) Copolovici, L.; Kannaste,̈ A.; Niinemets, Ü. Gas chromatog- Worsnop, D. R. Ubiquity and dominance of oxygenated species in raphy mass spectrometry method for determination of monoterpene organic aerosols in anthropogenically-influenced Northern Hemi- and sesquiterpene emissions from stressed plants. Studia Universitatis sphere midlatitudes. Geophys. Res. Lett. 2007, 34 (13), L13801. Babes-Bolyai Chemia 2009, 54 (4), 329−339. (66) Yli-Juuti, T.; Nieminen, T.; Hirsikko, A.; Aalto, P. P.; Asmi, E.; (51) Vuorinen, T.; Nerg, A.-M.; Ibrahim, M. A.; Reddy, G. V. P.; Hõrrak, U.; Manninen, H. E.; Patokoski, J.; Dal Maso, M.; Petajä,̈ T.; Holopainen, J. K. Emission of Plutella xylostella-induced compounds Rinne, J.; Kulmala, M.; Riipinen, I. Growth rates of nucleation mode from cabbages grown at elevated CO2 and orientation behavior of the particles in Hyytialä̈during 2003−2009: variation with particle size, natural enemies. Plant Physiol. 2004, 135 (4), 1984−1992. season, data analysis method and ambient conditions. Atmos. Chem. (52) Pinto, D. M.; Tiiva, P.; Miettinen, P.; Joutsensaari, J.; Kokkola, Phys. 2011, 11 (24), 12865−12886. H.; Nerg, A.-M.; Laaksonen, A.; Holopainen, J. K. The effects of (67) Nieminen, T.; Asmi, A.; Dal Maso, M.; Aalto, P. P.; Keronen, P.; increasing atmospheric ozone on biogenic monoterpene profiles and Petajä,̈ T.; Kulmala, M.; Kerminen, V.-M. Trends in atmospheric new- the formation of secondary aerosols. Atmos. Environ. 2007, 41 (23), particle formation: 16 years of observations in a boreal-forest 4877−4887. environment. Boreal Environment Research 2014, 19, 191−214.

11509 DOI: 10.1021/acs.est.6b02800 Environ. Sci. Technol. 2016, 50, 11501−11510 Environmental Science & Technology Article

(68) Hallquist, M.; Wenger, J. C.; Baltensperger, U.; Rudich, Y.; Simpson, D.; Claeys, M.; Dommen, J.; Donahue, N. M.; George, C.; Goldstein, A. H.; Hamilton, J. F.; Herrmann, H.; Hoffmann, T.; Iinuma, Y.; Jang, M.; Jenkin, M. E.; Jimenez, J. L.; Kiendler-Scharr, A.; Maenhaut, W.; McFiggans, G.; Mentel, T. F.; Monod, A.; Prevó̂t, A. S. H.; Seinfeld, J. H.; Surratt, J. D.; Szmigielski, R.; Wildt, J. The formation, properties and impact of secondary organic aerosol: current and emerging issues. Atmos. Chem. Phys. 2009, 9 (14), 5155−5236. (69) Gouinguene, S. P.; Turlings, T. C. J. The effects of abiotic factors on induced volatile emissions in corn plants. Plant Physiol. 2002, 129 (3), 1296−1307. (70) Hansen, U.; Seufert, G. Temperature and light dependence of beta-caryophyllene emission rates. Journal of Geophysical Research- Atmospheres 2003, 108 (D24), 1−7. (71) Niinemets, Ü. Mild versus severe stress and BVOCs: thresholds, priming and consequences. Trends Plant Sci. 2010, 15 (3), 145−153. (72) Mentel, T. F.; Wildt, J.; Kiendler-Scharr, A.; Kleist, E.; Tillmann, R.; Dal Maso, M.; Fisseha, R.; Hohaus, T.; Spahn, H.; Uerlings, R.; Wegener, R.; Griffiths, P. T.; Dinar, E.; Rudich, Y.; Wahner, A. Photochemical production of aerosols from real plant emissions. Atmos. Chem. Phys. 2009, 9 (13), 4387−4406. (73) Hamilton, J. F.; Lewis, A. C.; Carey, T. J.; Wenger, J. C.; Garcia, E. B. I.; Munoz, A. Reactive oxidation products promote secondary organic aerosol formation from green leaf volatiles. Atmos. Chem. Phys. 2009, 9 (11), 3815−3823. (74) Harvey, R. M.; Zahardis, J.; Petrucci, G. A. Establishing the contribution of lawn mowing to atmospheric aerosol levels in American suburbs. Atmos. Chem. Phys. 2014, 14 (2), 797−812. (75) Jain, S.; Zahardis, J.; Petrucci, G. A. Soft Ionization Chemical Analysis of Secondary Organic Aerosol from Green Leaf Volatiles Emitted by Turf Grass. Environ. Sci. Technol. 2014, 48 (9), 4835−4843. (76)Lee,A.;Goldstein,A.H.;Keywood,M.D.;Gao,S.; Varutbangkul, V.; Bahreini, R.; Ng, N. L.; Flagan, R. C.; Seinfeld, J. H. Gas-phase products and secondary aerosol yields from the ozonolysis of ten different terpenes. J. Geophys. Res. 2006, 111 (D7), 1−18. (77) Lee, A.; Goldstein, A. H.; Kroll, J. H.; Ng, N. L.; Varutbangkul, V.; Flagan, R. C.; Seinfeld, J. H. Gas-phase products and secondary aerosol yields from the photooxidation of 16 different terpenes. J. Geophys. Res. 2006, 111 (D17), 1−25. (78) Fukui, Y.; Doskey, P. V. Identification of nonmethane organic compound emissions from grassland vegetation. Atmos. Environ. 2000, 34 (18), 2947−2956. (79) Aschmann, S. M.; Arey, J.; Atkinson, R. OH radical formation from the gas-phase reactions of O3 with a series of terpenes. Atmos. Environ. 2002, 36 (27), 4347−4355.

11510 DOI: 10.1021/acs.est.6b02800 Environ. Sci. Technol. 2016, 50, 11501−11510

Environmental and Experimental Botany 138 (2017) 184–192

Contents lists available at ScienceDirect

Environmental and Experimental Botany

journal homepage: www.elsevier.com/locate/envexpbot

Research Paper

Disproportionate photosynthetic decline and inverse relationship

between constitutive and induced volatile emissions upon feeding of

Quercus robur leaves by large larvae of gypsy moth (Lymantria dispar)

a, b c a b

Lucian Copolovici *, Andreea Pag , Astrid Kännaste , Adina Bodescu , Daniel Tomescu ,

a d c,e

Dana Copolovici , Maria-Loredana Soran , Ülo Niinemets

a

Faculty of Food Engineering, Tourism and Environmental Protection, Research Center in Technical and Natural Sciences, “Aurel Vlaicu” University, 2 Elena

Dragoi, Arad 310330, Romania

b

Institute of Technical and Natural Sciences Research, Development of “Aurel Vlaicu” University, 2 Elena Dragoi, Arad 310330, Romania

c

Institute of Agricultural and Environmental Sciences, Estonian University of Life Sciences, Kreutzwaldi 1, Tartu 51014, Estonia

d

National Institute of Research and Development for Isotopic and Molecular Technologies, Cluj-Napoca 400293, Romania

e

Estonian Academy of Sciences, Kohtu 6, 10130 Tallinn, Estonia

A R T I C L E I N F O A B S T R A C T

Article history:

Received 7 January 2017 Gypsy moth (Lymantria dispar L., Lymantriinae) is a major pest of pedunculate oak (Quercus robur) forests

Received in revised form 21 March 2017 in Europe, but how its infections scale with foliage physiological characteristics, in particular with

Accepted 24 March 2017

photosynthesis rates and emissions of volatile organic compounds has not been studied. Differently from

Available online 27 March 2017

the majority of insect herbivores, large larvae of L. dispar rapidly consume leaf area, and can also bite

through tough tissues, including secondary and primary leaf veins. Given the rapid and devastating

Keywords:

feeding responses, we hypothesized that infection of Q. robur leaves by L. dispar leads to disproportionate

Green leaf volatiles

scaling of leaf photosynthesis and constitutive isoprene emissions with damaged leaf area, and to less

Induced emissions

prominent enhancements of induced volatile release. Leaves with 0% (control) to 50% of leaf area

Isoprene emission

removed by larvae were studied. Across this range of infection severity, all physiological characteristics

Large insect herbivores

were quantitatively correlated with the degree of damage, but all these traits changed disproportionately

Monoterpene emission

Photosynthesis with the degree of damage. The net assimilation rate was reduced by almost 10-fold and constitutive

Quantitative responses isoprene emissions by more than 7-fold, whereas the emissions of green leaf volatiles, monoterpenes,

Volatile organic compounds methyl salicylate and the homoterpene (3E)-4,8-dimethy-1,3,7-nonatriene scaled negatively and almost

linearly with net assimilation rate through damage treatments. This study demonstrates that feeding by

large insect herbivores disproportionately alters photosynthetic rate and constitutive isoprene

emissions. Furthermore, the leaves have a surprisingly large capacity for enhancement of induced

emissions even when foliage photosynthetic function is severely impaired.

© 2017 Elsevier B.V. All rights reserved.

1. Introduction Barbosa, 2016) and protection of plants against herbivore attacks

(Heil, 2014; Pichersky and Gershenzon, 2002;), against excess

In field environments, plants are frequently exposed to a temperatures (Becker et al., 2015; Possell and Loreto, 2013), and

multitude of abiotic and biotic stressors. To cope with environ- oxidative stress, e.g. that generated by ozone exposure (Loreto and

mental and biological stressors, more than 100,000 secondary Schnitzler, 2010; Possell and Loreto, 2013; Vickers et al., 2009).

chemical products are synthesized by plants and at least 1700 of While constitutive volatile emissions occur in only a limited

these are known to be volatile (Bauer et al., 1998; Copolovici and number of species (Fineschi et al., 2013), stress-driven volatile

Niinemets, 2016; Kesselmeier and Staudt, 1999; Pichersky and emissions can be induced by abiotic and biotic stresses in all plant

Gershenzon, 2002). These biogenic volatile organic compounds species (Copolovici and Niinemets, 2016; Harrison et al., 2013;

(BVOC) serve many functions such as pollinator attraction (Lucas- Niinemets, 2010). Key biotic stresses eliciting major volatile

emission responses are infestations by fungi, bacteria, herbivores

and insects (Copolovici and Niinemets, 2016; Niinemets et al.,

2013). Different biotic stressors elicit the same major classes of

* Corresponding author.

E-mail address: [email protected] (L. Copolovici). volatile compounds, green leaf volatiles (such as C5 and C6

http://dx.doi.org/10.1016/j.envexpbot.2017.03.014

0098-8472/© 2017 Elsevier B.V. All rights reserved.

L. Copolovici et al. / Environmental and Experimental Botany 138 (2017) 184–192 185

alcohols and aldehydes), ubiquitous (e.g., a-pinene, b-pinene, For example, in the Southeastern United States BVOC emissions

D-3-carene) and specific monoterpenes (e.g., b-ocimene, linalool), can have a significant influence on the total aerosol burden due to

sesquiterpenes (e.g. b-caryophyllene) and homoterpenes ((3E)- condensation on acidic sulfate seed aerosols and concomitant

4,8-dimethy-1,3,7-nonatriene (DMNT) and (3E,7E)-4,8,12-trime- growth of particles (Lee et al., 2012; Link et al., 2015). The current

thyltrideca-1,3,7,11-tetraene (TMTT)) (Blande et al., 2014; Copolo- estimates of BVOC emissions and their role in atmospheric

vici and Niinemets, 2016; Holopainen, 2011; Kleist et al., 2012). processes only consider constitutive emissions (Arneth et al.,

However, as comparative experiments among different stresses, 2011; Guenther, 2013; Guenther et al., 2012), but stress-dependent

e.g., among infested, mechanically damaged and jasmonic acid- elicitation of volatiles can significantly modify the overall volatile

treated plants (Giorgi et al., 2015), and among leaf wounding and blend and amount of volatiles released into the atmosphere (Hare,

darkening (Brilli et al., 2011), demonstrate, different biotic stresses 2011; Holopainen and Blande, 2013; Holopainen and Gershenzon,

result in different volatile emission fingerprints. 2010; Niinemets et al., 2010a). In particular, under major outbreaks

Stress-induced volatiles can directly protect plants against of feeding herbivores that regularly occur in nature (Abrams and

biotic stress by serving as repellents of herbivores (Lucas-Barbosa Orwig,1996; Dwyer et al., 2004; Mattson and Haack,1987) induced

et al., 2011) or inhibitors of fungal and bacterial growth (Schmidt volatiles can dominate the release of BVOC from vegetation over

et al., 2016). They can also confer indirect defense by attracting large areas, underscoring the importance of gaining a better

predators to their herbivore prey or oviposition host (Arimura knowledge of quantitative scaling of induced emissions with the

et al., 2000a; Dicke and Baldwin, 2010; Johnson and Gilbert, 2015; degree of herbivory damage.

Koski et al., 2015). For both functions, the composition of the Pedunculate oak (Quercus robur L.) is a widely distributed

emission blend as well as the total emission rates play important constitutively isoprene-emitting tree species that is one of the

roles, the first determining the specificity of the signal for repelling most economically important broad-leaved forest trees in Europe.

or attraction, and the second, determining the dispersal of the Oak forests currently exhibit declining productivity and crown

signal as well as its chemical and physiological activity. Herbivore- dieback in several sites throughout Europe. Various abiotic and

induced changes in emissions of volatiles have been shown in biotic factors, including herbivory by the larvae of phyllophagous

numerous papers (Dicke, 2016; Heil, 2014 for reviews), but only a insects have been shown to importantly contribute to the oak

few studies have focused on quantitative relationships between decline (Batos et al., 2014; Thomas et al., 2002; Tonioli et al., 2001).

the degree of damage and amount of compounds emitted (Bruce, The most important defoliating insects feeding on Q. robur are

2015; Copolovici et al., 2014a; Holopainen and Gershenzon, 2010; winter moth (Operophtera brumata L.), tortrix moth (Tortrix

Niinemets et al., 2013). Presence of correlations among the degree viridana L.) and European gypsy moth (Lymantria dispar dispar

of herbivory damage and emission of volatiles seems trivial, L.) (Thomas et al., 2002). Among these, L. dispar is a species with

however, it crucially depends on local and systemic emission large larvae that can grow to the size of 50–90 mm (Milanovic et al.,

2

responses. While local response in immediately impacted leaf 2014) and that are capable of consuming 10 cm leaf area per day

areas is induced rapidly, e.g., emissions of green leaves volatles per larva, including the second order veins and the terminal part of

(GLV) can occur within seconds due to constitutive activity of the mid-rib (Milanovic et al., 2014). In the current study, we used L.

lipoxygenases (Portillo-Estrada et al., 2015), induction of emissions dispar larvae as a model to characterize the influence of large insect

of mono- and sesquiterpenes that need de novo expression of herbivore on constitutive and induced volatile release in Q. robur.

responsible synthases is more time-consuming, from hours to days We tested the hypothesis that the physiological activity of Q. robur

(Copolovici et al., 2014a, 2011; Pazouki et al., 2016). Also, the leaves infected with the large herbivore L. dispar is quantitatively

induction response is quenched relatively rapidly, within a few associated with the degree of damage, in particular, that the

days upon relief of herbivory stress (Copolovici et al., 2014a, 2011). infection leads to a major decline in leaf photosynthetic activity

Thus, the whole leaf response is a complex amalgamate of local and and isoprene emissions and a modest increase in induced

systemic induction and quenching responses. This is biologically emissions.

highly relevant considering the huge diversity of herbivore impacts

plants encounter in the field. While the rate of leaf area 2. Materials and methods

consumption is low for small solitary herbivores, leaving plenty

of time for systemic responses in non-impacted leaf areas, 2.1. Plant material

simultaneous presence of many small herbivores and large solitary

herbivores can rapidly consume a major proportion of leaf area, The field measurements were performed in Lipova forest at

0 00 0 00

implying that the systemic emission response might not even Arad county, Romania (46 5 30 N, 21 41 30 E) in May 2014. The

occur in the major part of the leaf. In addition, while small site supports a broad-leaved deciduous forest that is mainly

herbivores mainly consume the intercostal leaf parts, large dominated by Quercus robur (canopy height 4–5 m), whereas Q.

herbivores can also consume major veins and lead to catastrophic petraea, Alnus glutinosa, and Q. rubra are minor canopy compo-

dysfunction of leaf water-, nutrient- and carbohydrate-conducting nents. The forest expands more than 6300 ha and more than half of

networks (Sack et al., 2003, 2004). Infections by small herbivores the trees (51%) were infected at the time of the study. The infection

can lead to compensatory increases of leaf photosynthetic activity was spatially highly heterogeneous, and forest patches with

in remaining leaf parts, but the damage of major veins could mean infested (more than 50 different patches with most trees infected

that feeding by large herbivores can lead to disproportionate observed) and non-infested patches with almost all trees lacking

reductions in foliage physiological activity, including photosyn- herbivores were interspersed. The weather conditions at the time

thetic activity and constitutive isoprenoid emissions as well as of measurements were: average air temperature of 26 2 C,

hindered elicitation of induced emissions (Copolovici et al., 2014a, relative humidity of 62% and atmospheric pressure of 102.4 kPa.

2011). The plants of Q. robur included for measurements were 10–12 years

Quantitative understanding of stress-dependent elicitation of old and 4–5 m tall. At the time of the measurements, the length of

volatile organic compound emission is further relevant for large- L. dispar larvae was about 30–40 mm. We took all measurements

scale processes in biosphere-atmosphere system. This is because with control leaves from the clean plots with healthy trees. These

BVOCs affect atmospheric OH radical and O3 concentrations and control leaves were further checked for presence of eggs and small

participate in the formation of secondary organic aerosols (SOA) larvae, and discarded if there was evidence of biotic interactions.

(Shen et al., 2013; Zhang et al., 2015; Ziemann and Atkinson, 2012). As all experiments were performed in natural conditions, the past

186 L. Copolovici et al. / Environmental and Experimental Botany 138 (2017) 184–192

and current larval damage as well as the actual number of 2.4. Leaf pigment analysis

herbivores that had been feeding in the leaf could not be

controlled. Therefore, we only report the relationships of leaf Pigment extraction was performed according to the method of

physiological traits with the degree of leaf damage. Opris et al. (2013) with minor modifications. Briefly, leaf samples of

2

4 cm were taken after leaf gas-exchange measurements and

2.2. Photosynthetic measurements volatile sampling and immediately frozen in liquid nitrogen. The

pigments were extracted in ice-cold 100% acetone with calcium

Foliage photosynthetic characteristics during larval feeding carbonate (Sigma-Aldrich, Steinheim, Germany), and centrifuged

were determined in the field with a portable gas exchange system with a Hettich centrifuge (Hettich 320 R Universal, Hettich GmbH,

GFS-3000 (Waltz, Effeltrich, Germany) as in Niinemets et al. Tuttlingen, Germany) at 0 C and 9500g for 3 min. Then the

(2010b), except for minor modifications in environmental con- supernatant was decanted, and the extraction was repeated till the

ditions as stated below. This system has an environmental- supernatant remained colorless, but at least three times. The

2

controlled cuvette with 8 cm window area and a full-window leaf extracts were pooled and brought to a final volume of 10 ml and

chamber fluorimeter for sample illumination and fluorescence then filtered through a 0.45 mm PTFE membrane filter (VWR

measurements. Each time, a leaf fully filling the cuvette area was International, Radnor, PA, USA). The obtained extracts were

enclosed and standard measurement conditions (light intensity of analyzed for carotenoid and chlorophyll pigments by a high

2 1

1000 nmol m s , leaf temperature of 25 C, chamber air humidi- pressure liquid chromatograph (HPLC–MS Shimadzu 2010, Kyoto,

1

ty of 70%, and CO2 concentration of 385 mmol mol ) were Japan) equipped with a diode array detector (DAD) according to

1

established. The leaf was stabilized under the standard conditions Niinemets et al. (1998) using a Lichrosorb RP-18 column (length

until stomata opened and steady-state CO2 and water vapor 125 mm, inner diameter 4 mm, film thickness 5 mm; Hichrom, UK).

exchange rates were reached. Steady-state values of net assimila- The column temperature was maintained at 10 C and the flow rate

1

tion (A) and transpiration (E) rates, and stomatal conductance to at 1.5 ml min . The solvents used for the chromatographic elution

water vapor (gs) were calculated according to von Caemmerer and consisted of ultra-pure water (A) and HPLC grade acetone (B)

Farquhar (1981). (Sigma-Aldrich, Steinheim, Germany). The chromatographic elu-

tion was started by isocratically running a mixture of 25% A and

2.3. Volatile sampling and GC–MS analyses 75% B for the first 7.5 min, followed by a 9.5 min linear gradient to

100% B, which was run isocratically for 3 min. Further, the eluent

Volatile organic compounds (VOC) were sampled via the outlet was changed to the initial composition of 25% A and 75% B by a

1

of the gas-exchange cuvette at a flow rate of 200 ml min for 2 min linear gradient. The HPLC was calibrated using commercially

20 min with a constant flow air sample pump 210-1003MTX (SKC available chlorophyll a, chlorophyll b, and b-carotene standards

Inc., Houston, TX, USA). The leaf chamber air was drawn through a (Sigma-Aldrich, Steinheim, Germany), and the calibration curves

multibed stainless steel cartridge (10.5 cm length, 4 mm inner were developed at corresponding spectral maxima (430 nm for

diameter, Supelco, Bellefonte, USA) filled with Carbotrap C 20/ chlorophyll a, and 455 nm for chlorophyll b and b-carotene).

40 mesh (0.2 g), Carbopack B 40/60 mesh (0.1 g) and Carbotrap X

20/40 mesh (0.1 g) adsorbents (Supelco, Bellefonte, USA) to 2.5. Estimation of the degree of leaf damage

quantitatively sample all volatiles in C5–C15 range (Kännaste

et al., 2014). Volatiles were also collected using L. dispar larvae The leaves were scanned at 200 dpi, and the total leaf area and

without plants in the laboratory conditions using a 3 l glass the leaf area damaged by L. dispar larvae were estimated with the

1

chambers with a flow rate of 2 l min similarly as in Copolovici “Leaf Area Measurement” software (www.plant-image-analysis.

et al. (2011). To estimate the background VOC concentrations org). Leaf dry mass was estimated after oven-drying at 70 C, and

(blank samples), air samples were taken from empty chambers leaf dry mass to leaf area was calculated. Foliage photosynthetic

before and after enclosure of leaves or larvae. rates, volatile emissions and pigment contents per unit leaf area

The adsorbent cartridges were analyzed for volatile lipoxyge- were estimated after correction for the consumed leaf area.

nase (LOX) pathway products (also called green leaf volatiles, GLV),

terpenes and methyl salicylate using a Shimadzu TD20 automated 2.6. Statistical analysis and data handling

cartridge desorber integrated with a Shimadzu 2010 Plus GC–MS

instrument (Shimadzu Corporation, Kyoto, Japan) according to the All measurements have been done in triplicates and data points

method of Toome et al. (2010) and Kännaste et al. (2014). Briefly, correspond to averages of three replicate leaves in each individual

1

for tube desorption, He purge flow was set at 40 ml min , primary plant (SE). The data were analyzed by linear and non-linear

desorption temperature at 250 C, and primary desorption time regression analyses, and the herbivory treatment effect at a certain

was 6 min. The mass spectrometer was operated in electron- level of damage severity relative to non-infected leaves was also

impact (EI) mode at 70 eV, with the transfer line temperature set at tested by ANOVA. All statistical analyses were conducted with

240 C and ion-source temperature at 150 C. The identification of ORIGIN 10.0 (OriginLab Corporation, MA, USA) and the statistical

terpenes, green leaf volatiles and isoprene was done using NIST tests were considered significant at P < 0.05.

spectral library ver. 14 and authentic standards (Sigma-Aldrich,

Taufkirchen, Germany). The absolute concentrations of com- 3. Results

pounds were calculated based on an external authentic standard

consisting of known amount of VOCs as described in full detail in 3.1. Effects of herbivory on foliage photosynthetic characteristics

Kännaste et al. (2014). Briefly, 1 ml of calibration sample has been

injected into the multi-bed tube followed by passing a nitrogen gas Herbivore feeding by L. dispar reduced leaf net CO2 assimilation

1

at 300 ml min through the tube for 5 min in order to evaporate rate (A), and even a moderate feeding, ca. 10% of leaf area removed,

the solvent and trap the volatiles on the adsorbent. Finally, the tube resulted in ca. 12% reduction of net assimilation rate, i.e. resulting

2 1

has been analyzed in GC–MS using the same program as for the in values of about 10 mmol m s (Fig. 1a). Net assimilation rate

samples. The background (blank) VOC concentrations were decreased with further increases in insect feeding, reaching values

2 1

subtracted from the concentrations with leaf samples and volatile of 1–2 mmol m s in leaves with ca. 50% of leaf area eaten

emission rates were calculated according to Niinemets et al. (2011). (Fig. 1a).

L. Copolovici et al. / Environmental and Experimental Botany 138 (2017) 184–192 187

30 a 25 ) -1

s 20 -2

15 (nmol m 10 r = 0.93 3

Isoprene emission rate 5

0

5 b 4 ) -1

s 3

-2 2 r = 0.79 5 (nmol m 1 GLV emission rate

0

6 c

) 4 -1 s -2 emission rate emission

2 (nmol m

Monoterpenes 0

Fig.1. Foliage net assimilation rate (a), stomatal conductance to water vapor (b) and 0 10 20 30 40 50 60

intracellular CO2 concentration per unit projected leaf area in Quercus robur plants

Leaf area damage (%)

in relation to the degree of damage by the larvae of the lymantriid moth Lymantria

dispar (percentage of leaf area consumed). Data points correspond to averages of

Fig. 2. Emissions rates of isoprene (a), green leaf volatiles (b; GLV, volatiles of

three replicate leaves in each individual plant (SE). Data were fitted by linear (a)

x lipoxygenase pathway, LOX volatiles) and monoterpenes (c) from Q. robur leaves

and non-linear regression in the form y = ab (b).

with different degrees of feeding by L. dispar larvae (replicates and data

presentation as in Fig. 1). Total LOX product emission was calculated as the sum

of emissions of 1-hexanol, (Z)-3-hexenol, (Z)-2-hexenal, and (Z)-3-hexenyl acetate

Differently from A, the stomatal conductance to water vapor (gs)

and the total monoterpene emission as the sum of emissions of a-pinene, b-pinene,

was only moderately affected by larval feeding (Fig. 1b). In fact, the

D b fi

2 1 camphene, limonene, -3-carene, p-cymene, and -phellandrene. Data were tted

average (SE) g of 109 20 mmol m s for strongly damaged x

s by non-linear regressions in the form of y = ab .

leaves with 30–35% leaf area eaten was not different from the

2 1

average g in non-damaged leaves (141 6 mmol m s ; P = 0.18

s induced emissions of ubiquitous monoterpenes (a-pinene, cam-

for the comparison among the means; Fig. 1b). Nevertheless, the

phene, D-3-carene, limonene and b-phellandrene) and typical

regression analysis indicated that through the entire damage range

stress-marker monoterpenes such as (E)-b-ocimene and linalool

of 0–50%, g was reduced by ca. 30%. Given the smaller g than A

s s (Table 1). The total emission rate of monoterpenes increased with

response to herbivory, the intercellular CO concentration (C )

2 i the degree of leaf damage from close to zero level in control leaves

increased with increasing the degree of herbivory damage (Fig. 1c). 2 1

to values as high as ca. 5.3 nmol m s in strongly infected leaves

(Fig. 2c). Herbivory feeding also led to low-level emissions of the

3.2. Elicitation of volatile emissions in herbivory-infected leaves

benzenoid methyl salicylate (MeSA) and the homoterpene (3E)-

4,8-dimethyl-1,3,7-nonatriene (DMNT), whereas the emission

Constitutive isoprene emission in herbivore-fed leaves was

rates increased curvilinearly with the percentage of leaf damage

2 1

reduced from 30.3 0.7 nmol m s (average SE) in healthy 2 1

to maximum values of 0.17 nmol m s for MeSA and 0.08 nmol

2 1

leaves to 4–5 nmol m s in heavily infected leaves (40–50% of 2 1

m s for DMNT (Fig. 3).

leaf area eaten by insects), and a strong negative correlation

between isoprene emission rate and percentage of leaf area eaten

3.3. Responses of foliage chlorophyll and b-carotene contents to

was observed (Fig. 2a). Larval feeding induced emissions of green herbivory

leaf volatiles [GLV; primarily, (Z)-3-hexenol, (E)-2-hexenal, (Z)-3-

hexenyl acetate and 1-hexanol), Fig. 2b], that increased with the

Both chlorophyll a and b contents decreased strongly with the

2

degree of damage, reaching very high values of 3.0–4.4 nmol m

area eaten by the larvae, e.g., the average chlorophyll a content

1

s (sum of all GLV) in heavily eaten leaves (Fig. 2b). Herbivory also 2

decreased from 231 2 mg m in non-infected leaves to 95 3

188 L. Copolovici et al. / Environmental and Experimental Botany 138 (2017) 184–192

Table 1

2 1

Average SD emission rates (nmol m s ) of different volatiles released from leaves of Q. robur in response to insect damage.

Compound Control Average (SD) degree of damage (%)

12.2 0.6 18.6 1.0 38.4 9.9

(Z)-3-hexenol nd 0.107 0.005 0.213 0.007 0.90 0.13

(E)-2-hexenal nd 0.100 0.003 0.320 0.010 0.82 0.17

(Z)-3-hexenyl acetate nd 0.054 0.007 0.080 0.010 0.223 0.023

1-hexanol nd 0.223 0.022 0.454 0.017 1.84 0.33

isoprene 30.3 3.2 19.1 3.4 14.0 1.7 3.9 1.4

apinene 0.0092 0.0012 0.622 0.011 0.848 0.027 1.58 0.11

camphene nd 0.033 0.002 0.043 0.001 0.111 0.010

D-3-carene 0.011 0.006 0.504 0.004 0.663 0.009 1.07 0.07

limonene 0.0094 0.003 0.305 0.032 0.412 0.008 1.02 0.05

b-phellandrene 0.0106 0.007 1.12 0.14 1.09 0.13 0.98 0.15

(E)-b-ocimene nd 0.052 0.001 0.098 0.001 0.297 0.044

linalool nd 0.029 0.001 0.053 0.002 0.128 0.024

DMNT nd 0.007 0.002 0.021 0.006 0.068 0.015

methyl salicylate nd 0.034 0.006 0.045 0.011 0.145 0.021

nd = not detected.

) 0.20 30

-1 DMNT

s a

-2 MeS A

r = 0.927 25 0.15 ) -1

s 20 -2 0.10

15

r = 0.91 9

(nmol m 0.05 r = 0.931 10 P < 0.0 5

Isoprene emission rate 5

Emission rate (nmol m 0.00

0 10 20 30 40 50 60 0

400

Leaf area damage (%) b )

-2

Fig. 3. Emissions of the benzenoid methyl salicylate (MeSA) and the homoterpene

300

(3E)-4,8-dimethyl-1,3,7-nonatriene (DMNT) from leaves of Q. robur in relation to

the degree of herbivory by L. dispar larvae. Data presentation and replication as in

Fig. 1. The relationships were fitted by linear regressions. 200

2

mg m in leaves with 30–50% damage (Fig. 4). The chlorophyll a/b r = 0.516

P = 0.0 6

ratio was on average 2.48 0.26 and did not depend on the degree

100

of leaf damage (r = 0.5, P > 0.05). b-Carotene content did not

significantly correlate with the degree of larval damage (Fig. 4). Chlorophyll a+b (mg m

0

3.4. Negative scaling among constitutive and induced isoprenoids 6

c

Through different damage severities, foliage net assimilation

rate scaled negatively with the emissions of induced volatiles

) 4 -1

(Fig. 5a for total monoterpenes, 5b for isoprene and 5c for total s -2 r = 0.953 P < 0.05 2 (nmol m

Monoterpenes emission rate Monoterpenes 0

0 10 20

-2 -1

Net assimilation rate (nmol m s )

Fig. 5. Correlations of net assimilation rate (a), isoprene emission rate (b) and

chlorophyll (a + b) content (c) with monoterpene emission rate in Q. robur leaves

with different degrees of damage by L. dispar larvae (Figs. 1 a, 2 a,c and 4 for the

correlations of given traits with the degree of damage). Data were fitted by non-

linear regressions.

Fig. 4. Effects of feeding by the larvae of L. dispar on the contents of chlorophyll a

and b, and carotene in Q. robur leaves. Statistical replicates and data presentation as

in Fig. 1. Data were fitted by linear regressions.

L. Copolovici et al. / Environmental and Experimental Botany 138 (2017) 184–192 189

chlorophylls, r = 0.83, P < 0.01 for GLV; r = 0.89, P < 0.01 for MeSA; activation of alternative carbon sources can have resulted in the

r = 0.87, P < 0.01 for DMNT, for linear regression used). Constitutive reduction of the pool size of the immediate isoprene precursor,

isoprene emissions were negatively correlated with induced dimethylallyl diphosphate (DMADP) in chloroplasts (Rasulov et al.,

emissions (r = 0.97, P < 0.001 for total monoterpenes; r = 0.88, 2011, 2009), thereby reducing the emission rate. In addition,

P < 0.01 for GLV; r = 0.93, P < 0.01 for MeSA; r = 0.91, P < 0.01 for simultaneous reduction of isoprene synthase activity and isoprene

DMNT). emission can indicate overall decreases in foliage primary

metabolism and constitutive isoprene synthesis in non-consumed

4. Discussion leaf areas. Such a reduction is supported by significant declines in

foliage chlorophyll contents in infected leaves (Fig. 4).

4.1. Changes in photosynthetic function in leaves infected by large On the other hand, chloroplastic isoprene and monoterpene

larvae syntheses rely on the same plastidic DMADP pool, but the in vivo

effective Michaelis-Menten constant for DMADP (Km) is much

Plant photosynthetic responses to biotic stresses such as smaller for monoterpenes than for isoprene (Rasulov et al., 2014).

herbivory can range from stimulating effects, no effect or even Lower Km for monoterpene synthesis implies that the competition

significant impairment (Attaran et al., 2014; Gog et al., 2005; for DMADP is one-sided, and thus, elicitation of monoterpene

Nabity et al., 2009, 2013; Roslin et al., 2006; Zhou et al., 2015). synthesis upon herbivory feeding, can also partly explain the

Especially, colonization by solitary small herbivores can lead to decline in isoprene emission rates (Fig. 5b).

compensatory enhancement of photosynthesis in remaining leaf

parts (Copolovici et al., 2014a, 2011; Delaney et al., 2008). In 4.3. Induction of green leaf volatile emissions in larval-eaten leaves

Quercus robur in our study, we observed a drastic, disproportionate

reduction in net assimilation rate (Fig. 1a), consistent with the As a result of an attack by insects, specific elicitor molecules are

hypothesis that herbivory by large larvae capable of biting through generated by chemical or physical damage to plant membranes

and consumption of also major veins can seriously impair (Heil, 2014). Our study demonstrated elicitation of all key stress-

photosynthesis. Yet, stomatal conductance was surprisingly little induced volatile compound classes, green leaf volatiles (GLV),

affected, seemingly inconsistent with this hypothesis (Fig. 1b). mono- and sesquiterpenes, homoterpenes and methyl salicylate in

However, the low change in calculated stomatal conductance herbivory-infected leaves (Table 1). GLV are synthesized in a

might be an artifact due to evaporation of water from free water process where free octadecanoid fatty acids (linoleic acid = 18:2

surfaces generated upon herbivory, as well as due to transient and linolenic acid = 18:3) are released from plant membranes by

increases in stomatal conductance upon relaxation of epidermal phospholipases. Upon release of these free fatty acids, lip-

tension due to damage and concomitant opening of stomata (so- oxygenases (LOX) then produce 9- or 13-hydroperoxylinoleic or

called Ivanov’s effect, Moldau et al., 1993). –linolenic acid or a mixture of both (Matsui, 2006). A hydroperox-

On the other hand, longer-term reductions in photosynthesis in ide lyase further catalyzes the breakdown of 13-hydroperoxylinole

herbivore-fed leaves have also been associated with impaired (n)ic acid to a C6-compound, (Z)-3-hexenal, and a C12-product (12-

electron transport rate (Nabity et al., 2013). In our study, the oxo-(Z)-9-dodecenoic acid). (Z)-3-Hexenal can further give rise to

chlorophyll content decreased with leaf damage severity (Fig. 4). (Z)-3-hexenol, (E)-2-hexenol, (E)-3-hexenol or (E)-2-hexenal in

Such a reduction has been observed in some studies looking at consequent reactions (Feussner and Wasternack, 2002; Matsui,

diffuse massive leaf infection, e.g. mustard (Brassica juncea) leaf 2006). The emission of green leaf volatiles is a reliable marker of

infection by phloem-sucking mustard aphid (Lipaphis erysimi) oxidative stress and membrane-level damage (Porta and Rocha-

(Rehman et al. (2014) and tomato (Solanum lycopersicum) infection Sosa, 2002). GLV are rapidly released upon herbivory feeding due

by cotton mealybug (Phenacoccus solenopsis) (Huang et al. (2013)), to constitutive activity of LOX, and their almost immediate release

but not necessarily upon infection by tissue-removing herbivores. has been typically associated with mechanical damage upon

However, damage of major veins and concomitant reduction in wounding (Matsui et al., 2012; Portillo-Estrada et al., 2015; Scala

water availability of isolated mesophyll areas can well lead to the et al., 2013). In our study, we observed a strong correlation

start of senescence processes (Munné-Bosch, 2007, 2008). Given between the emissions of green leaf volatiles and the percentage of

the reduction of leaf pigment content (Fig. 4), changes character- leaf area damaged (Fig. 2b). Similarly to our study, increases in the

istic to programmed cell death such as inhibition of photosynthetic emission of GLV with the degree of herbivore feeding were also

electron transport and reductions in the amount or activity of observed in experiments with Caberia pusaria feeding on grey alder

photosynthetic rate-limiting enzymes were also likely responsible (Alnus incana) (Copolovici et al., 2011) and in experiments with

for reduction in net-assimilation rate in larval-eaten leaves. Epirrita autumnata feeding on hybrid aspen (Populus tremula x P.

In contrast, b-carotene content was weakly affected by larval tremuloides) (Schaub et al., 2010). Given that in these studies and in

feeding (Fig. 4). Differently from chlorophylls, b-carotene primari- our study, the degree of damage quantified as the percentage of

ly functions as antioxidant and its sustained level might serve leaf area removed includes both fresh and somewhat older

protective function. Although carotenoids can be destroyed under damage, the question is how such a correlation can occur if GLV

severe abiotic stress conditions (Ashraf and Harris, 2013), release is exclusively associated with immediate damage. Howev-

carotenoids are maintained longer than leaf chlorophylls in leaf er, conversion of (Z)-3-hexenal to more reduced and less toxic

tissues though senescence Niinemets et al., 2012). volatiles can also occur in non-impacted leaf areas (Matsui et al.,

2012), and GLV release can continue for a certain period of time

4.2. Modification of constitutive isoprene emissions by herbivory after the immediate biotic impact (Copolovici et al., 2011; Jiang

et al., 2016). Thus, scaling of GLV emissions with the degree of

Concomitant reductions of net assimilation rates and constitu- damage (Fig. 2b) might be explained by the sustained GLV release

tive isoprene emissions as observed in herbivore-fed leaves in our from non-impacted leaf areas.

study (Fig. 1b) have been reported in several other herbivory There is evidence that GLV release is higher upon damage of

studies (Laothawornkitkul et al., 2008; Loivamaeki et al., 2008; veins than upon damage of intercostal tissues (Portillo-Estrada

Loreto et al., 2014) as well as in fungal-infected leaves (Copolovici et al., 2015). Interestingly, the increase of GLV release with the

et al., 2014b; Jiang et al., 2016). Such a simultaneous reduction degree of damage in leaves with minor to moderate damage of 5–

might indicate that limited plastidic carbon availability or delayed 20% was much less than in leaves with moderate to extensive

190 L. Copolovici et al. / Environmental and Experimental Botany 138 (2017) 184–192

damage of 20–50% (Fig. 2b). Such a difference might reflect the complex interplay between jasmonate- and salicylate-dependent

circumstance that in leaves with 5–20% damage, there was limited signalling pathways upon herbivore infestations.

big vein severance by herbivores, while herbivores also bit through

major veins in leaves with extensive damage. This result is in a 5. Conclusions

marked contrast with the linear relationship of GLV release vs. leaf

area damage in the study with small herbivores Caberia pusaria The results of the current study highlight a major negative

(Copolovici et al., 2011) and Monsoma pulveratum (Copolovici et al., scaling of foliage photosynthetic rates and constitutive isoprene

2014a) that are incapable of biting through major veins. emissions, and concomitant increase in induced volatile emissions

upon feeding by large herbivore larvae (Fig. 1,2,5). While the

damage-dependent increase of green leaf volatiles was dispropor-

4.4. Elicitation of emissions of terpenoids and MeSA tionately greater in leaves with extensive degree of damage than in

moderately damaged leaves (Fig. 2b), emission rates of terpenoids

As a widespread consensus, isoprene and monoterpenes are leveled off at higher degrees of leaf damage (Fig. 2c). This suggests

thought to be synthesized via 2-C-methyl-D-erythritol 4-phos- that faster and more severe damage, especially major vein

phate (MEP) pathway in plastids (Copolovici et al., 2014a; Fineschi severance by large herbivores can much more strongly influence

et al., 2013; Vranova et al., 2012) and sesquiterpenes via foliage physiological activity than feeding by smaller herbivores.

mevalonate (MVA) pathway in cytosol (Lombard and Moreira, These contrasting herbivore responses need consideration in

2011), although there is recent evidence of possible monoterpene modeling herbivore elicited emissions in large-scale biosphere-

synthesis in cytosol depending on substrate availability (Pazouki atmosphere models.

and Niinemets, 2016). DMNT is synthesized from the sesquiter-

pene (E)-nerolidol likely in the cytosol (Baldwin et al., 2006; Tholl Acknowledgements

et al., 2011). Although present in different cellular compartments,

both isoprenoid synthesis pathways are upregulated upon Funding for this study has been provided by the Estonian

herbivory stress and several terpene synthase genes involved Ministry of Science and Education (institutional grant IUT-8-3), the

fi

have already been identi ed (Arimura et al., 2000a, 2000b; European Commission through the European Regional Fund (the

Baldwin et al., 2006; Loreto et al., 2014). Thus, elicitation of Center of Excellence EcolChange), the European Research Council

emissions of mono-, sesqui-, and homoterpenes has been found in (advanced grant 322603, SIP-VOL+) and the European Commission

different deciduous trees under herbivore stress (Stam et al., 2014; and the Romanian Government (project POSCCE 621/2014). This

Zhu et al., 2014 for reviews). The interplay between different work was also supported by a grant of the Romanian National

terpene synthase pathways is still somewhat unclear, especially Authority for Scientific Research, CNCS – UEFISCDI (project

given that different compounds serve different ecological func- number PN-II-RU-TE-2011-3-0022). We thank the Forestry Direc-

tions. In Q. robur infected by the moth Tortrix viridana, Giraldo et al. torate Arad (Lipova branch) for all their support during the time of

(2012) demonstrated that the larvae were attracted to the plants samples collection.

releasing higher amounts of homoterpene DMNT and monoter-

pene (E)-b-ocimene, while sesquiterpenes a-farnesene and References

germacrene D acted as a repellent.

We observed that the monoterpene emission rate increased Abrams, M.D., Orwig, D.A., 1996. A 300-year history of disturbance and canopy

recruitment for co-occurring white pine and hemlock on the Allegheny Plateau,

with increasing leaf damage (Fig. 2c), indicating that the activity of

USA. J. Ecol. 84, 353–363.

monoterpene synthases increased upon herbivore feeding. In

Arimura, G., Ozawa, R., Shimoda, T., Nishioka, T., Boland, W., Takabayashi, J., 2000a.

addition, as discussed above, monoterpene synthesis could have Herbivory-induced volatiles elicit defence genes in lima bean leaves. Nature

406, 512–515.

been further favored by greater competitive capacity for chloro-

Arimura, G.I., Tashiro, K., Kuhara, S., Nishioka, T.O.R., Takabayashi, J., 2000b. Gene

plastic DMADP compared with isoprene synthesis (Fig. 5b) and

responses in bean leaves induced by herbivory and by herbivory-induced

inhibition of pigment synthesis as evident in the reduction in leaf volatiles. Biochem. Biophys. Res. Commun. 277, 305–310.

Arneth, A., Schurgers, G., Lathière, J., Duhl, T., Beerling, D.J., Hewitt, C.N., Martin, M.,

pigment content (Fig. 5c). Scaling of mono- and sesquiterpene

Guenther, A., 2011. Global terrestrial isoprene emission models: sensitivity to

emissions with the degree of damage has been observed in several

variability in climate and vegetation. Atmos. Chem. Phys. 11, 8037–8052.

experiments looking at lepidopteran larval feeding effects on Ashraf, M., Harris, P.J.C., 2013. Photosynthesis under stressful environments: an

overview. Photosynthetica 51, 163–190.

volatile release (for example Copolovici et al., 2014a, 2011). As we

Attaran, E., Major, I.T., Cruz, J.A., Rosa, B.A., Koo, A.J.K., Chen, J., Kramer, D.M., He, S.Y.,

hypothesized, there was evidence of leveling off of monoterpene

Howe, G.A., 2014. Temporal dynamics of growth and photosynthesis

and DMNT vs. damage severity relationships at higher severity of suppression in response to jasmonate signaling. Plant Physiol. 165, 1302–1314.

Baldwin, I.T., Halitschke, R., Paschold, A., von Dahl, C.C., Preston, C.A., 2006. Volatile

damage (Figs. 2 c, 3). In fact, the ratios of monoterpene to GLV

signaling in plant–plant interactions: talking trees in the genomics era. Science

emissions and DMNT to GLV emissions decreased with increasing

311, 812–815.

the degree of damage, indicating that the induction response was Batos, B., Seslija Jovanovic, D., Miljkovic, D., 2014. Spatial and temporal variability of

fl –

relatively less prominent in more severely damaged leaves owering in the pedunculate oak (Quercus robur L.). Sumarski List 138, 371 379.

Bauer, K., Garbe, D., Surburh, H., 1998. Flavors and fragrances, Ullmann's

compared with the immediate stress response (or with the rapidly

Encyclopedia of Industrial Chemistry. The CD-ROM Edition. 5th ed. Wiley-VCH,

induced stress response). This is different from infection by small

Verlag, Berlin.

larvae (Copolovici et al., 2014a, 2011) or from infection by slowly Becker, C., Desneux, N., Monticelli, L., Fernandez, X., Michel, T., Lavoir, A.-V., 2015.

Effects of abiotic factors on HIPV-mediated interactions between plants and

developing biotic stresses such as fungal infections (Copolovici

parasitoids. BioMed Res. Int. 2015 doi:http://dx.doi.org/10.1155/2015/342982

et al., 2014b; Jiang et al., 2016), where GLV and monoterpene and

Article ID 342982, 18 pages.

GLV emissions are almost proportional. Blande, J.D., Holopainen, J.K., Niinemets, U., 2014. Plant volatiles in polluted

atmospheres: stress responses and signal degradation. Plant Cell Environ. 37,

The emission of shikimate pathway derived compound, methyl

1892–1904.

salicylate (MeSA), is typically observed for sap-sucking herbivores

Brilli, F., Ruuskanen, T.M., Schnitzhofer, R., Mueller, M., Breitenlechner, M., Bittner,

fl

such as aphids or white ies (Li et al., 2006; Zarate et al., 2007) and V., Wohlfahrt, G., Loreto, F., Hansel, A., 2011. Detection of plant volatiles after leaf

wounding and darkening by proton transfer reaction time-of-flight mass

is not usually considered as part of chewing herbivore response.

spectrometry (PTR-TOF). PLoS One 6.

However, as in our study, several previous studies have demon-

Bruce, T.J.A., 2015. Interplay between insects and plants: dynamic and complex

strated the release of methyl salicylate upon chewing herbivore interactions that have coevolved over millions of years but act in milliseconds. J.

Exp. Bot. 66, 455–465. attacks (Cardoza et al., 2002; Dicke et al., 1999), indicating a

L. Copolovici et al. / Environmental and Experimental Botany 138 (2017) 184–192 191

Cardoza, Y.J., Alborn, H.T., Tumlinson, J.H., 2002. In vivo volatile emissions from birds use volatile organic compounds from plants as olfactory foraging cues?

peanut plants induced by simultaneous fungal infection and insect damage. J. Three experimental tests. Ethology 121, 1131–1144.

Chem. Ecol. 28, 161–174. Laothawornkitkul, J., Paul, N.D., Vickers, C.E., Possell, M., Taylor, J.E., Mullineaux, P.

Copolovici, L., Niinemets, Ü., 2016. Environmental impacts on plant volatile M., Hewitt, C.N., 2008. Isoprene emissions influence herbivore feeding

emission. In: Blande, J., Glinwood, R. (Eds.), Deciphering Chemical Language of decisions. Plant Cell Environ. 31, 1410–1415.

Plant Communication. Springer International Publishing, Berlin, pp. 35–59. Lee, A.K.Y., Hayden, K.L., Herckes, P., Leaitch, W.R., Liggio, J., Macdonald, A.M., Abbatt,

Copolovici, L., Kannaste, A., Remmel, T., Vislap, V., Niinemets, U., 2011. Volatile J.P.D., 2012. Characterization of aerosol and cloud water at a mountain site

emissions from Alnus glutionosa induced by herbivory are quantitatively during WACS 2010: secondary organic aerosol formation through oxidative

related to the extent of damage. J. Chem. Ecol. 37, 18–28. cloud processing. Atmos. Chem. Phys. 12, 7103–7116.

Copolovici, L., Kannaste, A., Remmel, T., Niinemets, U., 2014a. Volatile organic Li, Q., Xie, Q.G., Smith-Becker, J., Navarre, D.A., Kaloshian, I., 2006. Mi-1-mediated

compound emissions from Alnus glutinosa under interacting drought and aphid resistance involves salicylic acid and mitogen-activated protein kinase

herbivory stresses. Environ. Exp. Bot. 100, 55–63. signaling cascades. Mol. Plant Microbe Interact. 19, 655–664.

Copolovici, L., Vaartnou, F., Portillo Estrada, M., Niinemets, U., 2014b. Oak powdery Link, M., Zhou, Y., Taubman, B., Sherman, J., Morrow, H., Krintz, I., Robertson, L., Cook,

mildew (Erysiphe alphitoides)-induced volatile emissions scale with the degree R., Stocks, J., West, M., Sive, B.C., 2015. A characterization of volatile organic

of infection in Quercus robur. Tree Physiol. 34, 1399–1410. compounds and secondary organic aerosol at a mountain site in the

Delaney, K.J., Haile, F.J., Peterson, R.K.D., Higley, L.G., 2008. Impairment of leaf Southeastern United States. J. Atmos. Chem. 72, 81–104.

photosynthesis after insect herbivory or mechanical injury on common Loivamaeki, M., Mumm, R., Dicke, M., Schnitzler, J.-P., 2008. Isoprene interferes with

milkweed, asclepias syriaca. Environ. Entomol. 37, 1332–1343. the attraction of bodyguards by herbaceous plants. Proc. Natl. Acad. Sci. U. S. A.

Dicke, M., Baldwin, I.T., 2010. The evolutionary context for herbivore-induced plant 105, 17430–17435.

volatiles: beyond the ‘cry for help’. Trends Plant Sci. 15, 167–175. Lombard, J., Moreira, D., 2011. Origins and early evolution of the mevalonate

Dicke, M., Gols, R., Ludeking, D., Posthumus, M.A.,1999. Jasmonic acid and herbivory pathway of isoprenoid Biosynthesis in the three domains of life. Mol. Biol. Evol.

differentially induce carnivore-attracting plant volatiles in lima bean plants. J. 28, 87–99.

Chem. Ecol. 25, 1907–1922. Loreto, F., Schnitzler, J.-P., 2010. Abiotic stresses and induced BVOCs. Trends Plant

Dicke, M., 2016. Plant phenotypic plasticity in the phytobiome: a volatile issue. Curr. Sci. 15, 154–166.

Opin. Plant Biol. 32, 17–23. Loreto, F., Dicke, M., Schnitzler, J.-P., Turlings, T.C.J., 2014. Plant volatiles and the

Dwyer, G., Dushoff, J., Yee, S.H., 2004. The combined effects of pathogens and environment. Plant Cell Environ. 37, 1905–1908.

predators on insect outbreaks. Nature 430, 341–345. Lucas-Barbosa, D., van Loon, J.J.A., Dicke, M., 2011. The effects of herbivore-induced

Feussner, I., Wasternack, C., 2002. The lipoxygenase pathway. Annu. Rev. Plant Biol. plant volatiles on interactions between plants and flower-visiting insects.

53, 275–297. Phytochemistry 72, 1647–1654.

Fineschi, S., Loreto, F., Staudt, M., Peñuelas, J., 2013. Diversification of volatile Lucas-Barbosa, D., 2016. Integrating studies on plant-pollinator and plant-herbivore

isoprenoid emissions from trees: evolutionary and ecological perspectives. In: interactions. Trends Plant Sci. 21, 125–133.

Niinemets, Ü., Monson, R.K. (Eds.), Biology, Controls and Models of Tree Volatile Matsui, K., Sugimoto, K., Mano, J.i., Ozawa, R., Takabayashi, J., 2012. Differential

Organic Compound Emissions. Springer, Berlin. metabolisms of green leaf volatiles in injured and intact parts of a wounded leaf

Giorgi, A., Manzo, A., Nanayakkara, N.N.M.C., Giupponi, L., Cocucci, M., Panseri, S., meet distinct ecophysiological requirements. PLoS One 7, e36433.

2015. Effect of biotic and abiotic stresses on volatile emission of Achillea collina Matsui, K., 2006. Green leaf volatiles: hydroperoxide lyase pathway of oxylipin

Becker ex Rchb. Nat. Prod. Res. 29, 1695–1702. metabolism. Curr. Opin. Plant Biol. 9, 274–280.

Giraldo, A., Heller, W., Fladung, M., Schnitzler, J.-P., Schroeder, H., 2012. Function of Mattson, W.J., Haack, R.A., 1987. The role of drought in outbreaks of plant-eating

defensive volatiles in pedunculate oak (Quercus robur) is tricked by the moth insects. Bioscience 37, 110–118.

Tortrix viridana. Plant Cell Environ. 35, 2192–2207. Milanovic, S., Lazarevic, J., Popovic, Z., Miletic, Z., Kostic, M., Radulovic, Z., Karadzic,

Gog, L., Berenbaum, M.R., DeLucia, E.H., Zangerl, A.R., 2005. Autotoxic effects of D., Vuleta, A., 2014. Preference and performance of the larvae of Lymantria

essential oils on photosynthesis in parsley, parsnip, and rough lemon. dispar (Lepidoptera: Lymantriidae) on three species of European oaks. Eur. J.

Chemoecology 15, 115–119. Entomol. 111, 371–378.

Guenther, A.B., Jiang, X., Heald, C.L., Sakulyanontvittaya, T., Duhl, T., Emmons, L.K., Moldau, H., Wong, S.-C., Osmond, C.B., 1993. Transient depression of photosynthesis

Wang, X., 2012. The Model of Emissions of Gases and Aerosols from Nature in bean leaves during rapid water loss. Aust. J. Plant Physiol. 20, 45–54.

version 2.1 (MEGAN2.1): an extended and updated framework for modeling Munné-Bosch, S., 2007. Ageing in perennials. Crit. Rev. Plant Sci. 26, 123–138.

biogenic emissions. Geosci. Model. Dev. 5, 1471–1492. Munné-Bosch, S., 2008. Do perennials really senesce? Trends Plant Sci. 13, 216–220.

Guenther, A., 2013. Upscaling biogenic volatile compound emissions from leaves to Nabity, P.D., Zavala, J.A., DeLucia, E.H., 2009. Indirect suppression of photosynthesis

landscapes. In: Niinemets, Ü., Monson, R.K. (Eds.), Biology, Controls and Models on individual leaves by arthropod herbivory. Ann. Bot. 103, 655–663.

of Tree Volatile Organic Compound Emissions. Springer, Berlin, pp. 391–414. Nabity, P.D., Zavala, J.A., DeLucia, E.H., 2013. Herbivore induction of jasmonic acid

Hare, J.D., 2011. Ecological role of volatiles produced by plants in response to and chemical defences reduce photosynthesis in Nicotiana attenuata. J. Exp. Bot.

damage by herbivorous insects. Annu. Rev. Entomol. 56, 161–180. 64, 685–694.

Harrison, S.P., Morfopoulos, C., Dani, K.G.S., Prentice, I.C., Arneth, A., Atwell, B.J., Niinemets, U., Bilger, W., Kull, O., Tenhunen, J.D., 1998. Acclimation to high

Barkley, M.P., Leishman, M.R., Loreto, F., Medlyn, B.E., Niinemets, Ü., Possell, M., irradiance in temperate deciduous trees in the field: changes in xanthophyll

Peñuelas, J., Wright, I.J., 2013. Volatile isoprenoid emissions from plastid to cycle pool size and in photosynthetic capacity along a canopy light gradient.

planet. New Phytol. 197, 49–57. Plant Cell Environ. 21, 1205–1218.

Heil, M., 2014. Herbivore- induced plant volatiles: targets, perception and Niinemets, Ü., Arneth, A., Kuhn, U., Monson, R.K., Peñuelas, J., Staudt, M., 2010a. The

unanswered questions. New Phytol. 204, 297–306. emission factor of volatile isoprenoids: stress, acclimation, and developmental

Holopainen, J.K., Blande, J.D., 2013. Where do herbivore-induced plant volatiles go? responses. Biogeosciences 7, 2203–2223.

Front. Plant Sci. 4. Niinemets, Ü., Copolovici, L., Hüve, K., 2010b. High within-canopy variation in

Holopainen, J.K., Gershenzon, J., 2010. Multiple stress factors and the emission of isoprene emission potentials in temperate trees: implications for predicting

plant VOCs. Trends Plant Sci. 15, 176–184. canopy-scale isoprene fluxes. J. Geophys. Res. Biogeosci. 115, G04029.

Holopainen, J.K., 2011. Can forest trees compensate for stress-generated growth Niinemets, Ü., Kuhn, U., Harley, P.C., Staudt, M., Arneth, A., Cescatti, A., Ciccioli, P.,

losses by induced production of volatile compounds? Tree Physiol. 31, 1356– Copolovici, L., Geron, C., Guenther, A.B., Kesselmeier, J., Lerdau, M.T., Monson, R.

1377. K., Peñuelas, J., 2011. Estimations of isoprenoid emission capacity from

Huang, J., Zhang, P.-J., Zhang, J., Lu, Y.-B., Huang, F., Li, M.-J., 2013. Chlorophyll enclosure studies: measurements, data processing, quality and standardized

content and chlorophyll fluorescence in tomato leaves infested with an invasive measurement protocols. Biogeosciences 8, 2209–2246.

mealybug, Phenacoccus solenopsis (Hemiptera: Pseudococcidae). Environ. Niinemets, Ü., García-Plazaola, J.I., Tosens, T., 2012. Photosynthesis during leaf

Entomol. 42, 973–979. development and ageing. In: Flexas, J., Loreto, F., Medrano, H. (Eds.), Terrestrial

Jiang, Y., Ye, J., Veromann, L.-L., Niinemets, Ü., 2016. Scaling of photosynthesis and Photosynthesis in a Changing Environment. A Molecular, Physiological and

constitutive and induced volatile emissions with severity of leaf infection by Ecological Approach. Cambridge University Press, Cambridge, pp. 353–372.

rust fungus (Melampsora larici-populina) in Populus balsamifera var. suaveolens. Niinemets, Ü., Kännaste, A., Copolovici, L., 2013. Quantitative patterns between

Tree Physiol. 38, 856–872. plant volatile emissions induced by biotic stresses and the degree of damage.

Johnson, D., Gilbert, L., 2015. Interplant signalling through hyphal networks. New Frontiers in Plant Science. Front. Plant-Microb. Interact. 4, 262.

Phytol. 205, 1448–1453. Niinemets, Ü., 2010. Mild versus severe stress and BVOCs: thresholds, priming and

Kännaste, A., Copolovici, L., Niinemets, Ü., 2014. Gas chromatography mass- consequences. Trends Plant Sci. 15, 145–153.

spectrometry method for determination of biogenic volatile organic Opris, O., Copaciu, F., Soran, M.L., Ristoiu, D., Niinemets, U., Copolovici, L., 2013.

compounds emitted by plants. In: Rodríguez-Concepción, M. (Ed.), Plant Influence of nine antibiotics on key secondary metabolites and physiological

Isoprenoids: Methods and Protocols. Humana Press, New York, pp. 161–169. characteristics in Triticum aestivum: Leaf volatiles as a promising new tool to

Kesselmeier, J., Staudt, M., 1999. Biogenic volatile organic compounds (VOC): an assess toxicity. Ecotoxicol. Environ. Saf. 87, 70–79.

overview on emission, physiology and ecology. J. Atmos. Chem. 33, 23–88. Pazouki, L., Niinemets, Ü., 2016. Multi-substrate terpenoid synthases: their

Kleist, E., Mentel, T.F., Andres, S., Bohne, A., Folkers, A., Kiendler-Scharr, A., Rudich, Y., occurrence and physiological significance. Front. Plant Sci. 7, 1019.

Springer, M., Tillmann, R., Wildt, J., 2012. Irreversible impacts of heat on the Pazouki, L., Kanagendran, A., Li, S., Kännaste, A., Rajabi Memari, H., Bichele, R.,

emissions of monoterpenes, sesquiterpenes, phenolic BVOC and green leaf Niinemets, Ü., 2016. Mono- and sesquiterpene release from tomato (Solanum

volatiles from several tree species. Biogeosciences 9, 5111–5123. lycopersicum) leaves upon mild and severe heat stress and through recovery:

Koski, T.-M., Laaksonen, T., Mantyla, E., Ruuskanen, S., Li, T., Giron-Calva, P.S., from gene expression to emission responses. Environ. Exp. Bot. 132, 1–15.

Huttunen, L., Blande, J.D., Holopainen, J.K., Klemola, T., 2015. Do insectivorous

192 L. Copolovici et al. / Environmental and Experimental Botany 138 (2017) 184–192

Pichersky, E., Gershenzon, J., 2002. The formation and function of plant volatiles: Schmidt, R., Etalo, D.W., de Jager, V., Gerards, S., Zweers, H., de Boer, W., Garbeva, P.,

perfumes for pollinator attraction and defense. Curr. Opin. Plant Biol. 5, 237– 2016. Microbial small talk: volatiles in fungal-bacterial interactions. Front.

243. Microbiol. 6.

Porta, H., Rocha-Sosa, M., 2002. Plant lipoxygenases: physiological and molecular Shen, X., Zhao, Y., Chen, Z., Huang, D., 2013. Heterogeneous reactions of volatile

features. Plant Physiol. 130, 15–21. organic compounds in the atmosphere. Atmos. Environ. 68, 297–314.

Portillo-Estrada, M., Kazantsev, T., Talts, E., Tosens, T., Niinemets, Ü., 2015. Emission Stam, J.M., Kroes, A., Li, Y., Gols, R., van Loon, J.J.A., Poelman, E.H., Dicke, M., 2014.

timetable and quantitative patterns of wound-induced volatiles across different Plant interactions with multiple insect herbivores: from community to genes.

damage treatments in aspen (Populus tremula). J. Chem. Ecol. 41, 1105–1117. Annu. Rev. Plant Biol. 65, 689–713.

Possell, M., Loreto, F., 2013. The role of volatile organic compounds in plant Tholl, D., Sohrabi, R., Huh, J.-H., Lee, S., 2011. The biochemistry of homoterpenes

resistance to abiotic stresses: responses and mechanisms. In: Niinemets, Ü., common constituents of floral and herbivore-induced plant volatile bouquets.

Monson, R.K. (Eds.), Biology, Controls and Models of Tree Volatile Organic Phytochemistry 72, 1635–1646.

Compound Emissions. Springer, Berlin, pp. 209–235. Thomas, F.M., Blank, R., Hartmann, G., 2002. Abiotic and biotic factors and their

Rasulov, B., Hüve, K., Välbe, M., Laisk, A., Niinemets, Ü., 2009. Evidence that light, interactions as causes of oak decline in Central Europe. For. Pathol. 32, 277–307.

carbon dioxide and oxygen dependencies of leaf isoprene emission are driven Tonioli, M., Escarre, J., Lepart, J., Speranza, M., 2001. Facilitation and competition

by energy status in hybrid aspen. Plant Physiol. 151, 448–460. affecting the regeneration of Quercus pubescens Willd. Ecoscience 8, 381–391.

Rasulov, B., Hüve, K., Laisk, A., Niinemets, Ü., 2011. Induction of a longer-term Toome, M., Randjärv, P., Copolovici, L., Niinemets, Ü., Heinsoo, K., Luik, A., Noe, S.M.,

component of isoprene release in darkened aspen leaves: origin and regulation 2010. Leaf rust induced volatile organic compounds signalling in willow during

under different environmental conditions. Plant Physiol. 156, 816–831. the infection. Planta 232, 235–243.

Rasulov, B., Bichele, I., Laisk, A., Niinemets, Ü., 2014. Competition between isoprene Vickers, C.E., Gershenzon, J., Lerdau, M.T., Loreto, F., 2009. A unified mechanism of

emission and pigment synthesis during leaf development in aspen. Plant Cell action for volatile isoprenoids in plant abiotic stress. Nat. Chem. Biol. 5, 283–

Environ. 37, 724–741. 291.

Rehman, F., Khan, F.A., Anis, S.B., 2014. Assessment of aphid infestation levels in von Caemmerer, S., Farquhar, G.D., 1981. Some relationships between the

some cultivars of mustard with varying defensive traits. Arch. Phytopathol. biochemistry of photosynthesis and the gas exchange of leaves. Planta 153, 376–

Plant Protect. 47, 1866–1874. 387.

Roslin, T., Gripenberg, S., Salminen, J.P., Karonen, M., O'Hara, R.B., Pihlaja, K., Vranova, E., Coman, D., Gruissem, W., 2012. Structure and dynamics of the

Pulkkinen, P., 2006. Seeing the trees for the leaves – oaks as mosaics for a host- isoprenoid pathway network. Mol. Plant 5, 318–333.

specific moth. Oikos 113, 106–120. Zarate, S.I., Kempema, L.A., Walling, L.L., 2007. Silverleaf whitefly induces salicylic

Sack, L., Cowan, P.D., Holbrook, N.M., 2003. The major veins of mesomorphic leaves acid defenses and suppresses effectual jasmonic acid defenses. Plant Physiol. 43,

revisited: testing for conductive overload in Acer saccharum (Aceraceae) and 866–875.

Quercus rubra (Fagaceae). Am. J. Bot. 90, 32–39. Zhang, X., McVay, R.C., Huang, D.D., Dalleska, N.F., Aumont, B., Flagan, R.C., Seinfeld,

Sack, L., Streeter, C.M., Holbrook, N.M., 2004. Hydraulic analysis of water flow J.H., 2015. Formation and evolution of molecular products in alpha-pinene

through leaves of sugar maple and red oak. Plant Physiol. 134, 1824–1833. secondary organic aerosol. Proc. Natl. Acad. Sci. U. S. A. 112, 14168–14173.

Scala, A., Allmann, S., Mirabella, R., Haring, M.A., Schuurink, R.C., 2013. Green leaf Zhou, S., Lou, Y.-R., Tzin, V., Jander, G., 2015. Alteration of plant primary metabolism

volatiles: a plant’s multifunctional weapon against herbivores and pathogens. in response to insect herbivory. Plant Physiol. 169, 1488–1498.

Int. J. Mol. Sci. 14, 17781–17811. Zhu, F., Poelman, E.H., Dicke, M., 2014. Insect herbivore- associated organisms affect

Schaub, A., Blande, J.D., Graus, M., Oksanen, E., Holopainen, J.K., Hansel, A., 2010. plant responses to herbivory. New Phytol. 204, 315–321.

Real-time monitoring of herbivore induced volatile emissions in the field. Ziemann, P.J., Atkinson, R., 2012. Kinetics, products, and mechanisms of secondary

Physiol. Plant. 138, 123–133. organic aerosol formation. Chem. Soc. Rev. 41, 6582–6605. Diclofenac Influence on Photosynthetic Parameters and Volatile Organic Compounds Emission from Phaseolus vulgaris L. Plants

LUCIAN COPOLOVICI1, DANIELA TIMIS1, MONICA TASCHINA1, DANA COPOLOVICI1*, GABRIELA CIOCA2*, SIMONA BUNGAU3 1Aurel Vlaicu University, Faculty of Food Engineering, Tourism and Environmental Protection, Institute of Research, Innovation and Development in Technical and Natural Sciences, 2 Elena Dragoi Str., 310330, Arad, Romania 2Lucian Blaga University of Sibiu, Faculty of Medicine, 10 Victoriei Blvd., 550024, Arad, Romania 3University of Oradea, Faculty of Medicine and Pharmacy, 29 Nicolae Jiga Str., 410028, Oradea, Romania

Diclofenac, both from human and veterinary consumption, may arrive in landfills or in the wastewater treatment plants, becoming an environmental pollutant. Therefore, we aimed to study the influence of diclofenac on plants growth and development. We chose as model plant the bean (Phaseolus vulgaris L.) that was watered with different concentrations of aqueous diclofenac solutions (0-0.4 g/L). The plants exhibited linear decreased values of net assimilation rates and stomatal conductance to water vapors with increased diclofenac’s concentrations. Emission of 3-hexenol was determined to scale up with diclofenac concentration, therefore this compound may be proposed as stress marker. Also in the emission of bean plants were detected 3 different monoterpenes (α-pinene, camphene and 3-carene), their concentration increasing with elevated concentration of diclofenac. We can conclude that diclofenac may affect the plants photosynthetic parameters and also might disturb the methylerythritol phosphate pathway (MEP) in plastids. Keywords: diclofenac, photosynthetic parameters, volatile organic compounds emission, Phaseolus Vulgaris l. plants

2-(2-(2,6-dichlorophenylamino)phenyl)acetic acid treatment plants [7]. The medium concentration in (Diclofenac) (scheme 1) is one of the most popular non- German rivers for examples has been determined at 0.15 steroidal anti-inflammatory drugs along with ibuprofen, µg/L but concentration of 1.2µg/L has been determined aspirin, acetaminophen, ketoprofen, naproxen [1], and [29]. It has been demonstrated that living organisms could subject of a lot of studies [2-5]. be affected by pharmaceutical contamination [30] but they could be used as phyto-indicators [31]. Diclofenac has also been shown to induce cytological changes in rainbow and brown trout tissues after 21 days of exposure at 50 µg/L Scheme 1. Structure of Diclofenac diclofenac [32]. Furthermore, mussels could accumulate diclofenac from the water if they are exposed even at small concentrations [33]. The number of the studies regarding the diclofenac effects on the plants is very low. It has been demonstrated that diclofenac reduces the seedlings growth It is used to reduce pain, fever and inflammation, being and determined the increasing of enzymatic activity in sold without prescription in many countries worldwide [6]. three leguminous plants (lupin, pea, lentil) [34] but this In general, annual pharmaceuticals consumption could compound has less phytotoxic effect than different arrive at an average of 150 g per capita in many antibiotics. Biochemical processes of duckweed plants industrialized countries [7]. Diclofenac is prescribed as oral have been affected even at a very low concentration (10 tablets, topical gel, or suppositories, being sold under µg/L) of diclofenac [35]. different trade names. The total consumption of diclofenac In the present study we observed the influence of is almost impossible to be determined, but Intercontinental diclofenac on bean (Phaseolus vulgaris L.) as a model plant. Marketing Services (IMS) health data have been estimated Consequently, we determined the photosynthetic that around 940 tons of diclofenac is consumed in one parameters, the emission of volatile organic compounds year. Even more, the total sale of diclofenac in 2011 has and concentration of pigments of bean in the presence of been around 1.5 billion dollars [8]. According to IMS health diverse concentrations of diclofenac. data around 1500 tons of diclofenac is consumed globally only for human medicine [9]. In Europe, the most usage of Experimental part diclofenac has been reported in Germany with around 86 Materials and methods tons in 2001 [10], while in England the total consumption All commercial chemicals and solvents are reagent has been 26.13 tons per year [11]. grade and were used without further purification. Acetone Medicines are polluting the environment [12-15], mainly was purchased from Merck-Schuchardt, Germany. water sources [16-22] and soil [23-28], and affect the Bean plants (Phaseolus vulgaris L.) var. Ecaterina, plants growth and development. Studies regarding the Agrosel, Campia Turzii, Romania, were seeded in 3 L pots medicines concentrations that reach into the soil and their and grown for 3 weeks under artificial light at a rate of 300 impact on plants is scary. The main sources of diclofenac µmol m-2 s-1. The light / dark period was 12-12 h and pollution are the human and veterinary routes from which temperature was 25°C. The plants were watered every diclofenac ends up in landfills or in the wastewater

* email: [email protected]; [email protected] All the authors had equal contribution at this original article. 2076 http://www.revistadechimie.ro REV.CHIM.(Bucharest)♦68♦No. 9 ♦2017 second days. Four different diclofenac solution were used Such behavior of assimilation rate and stomatal for plant treatment: 0.1 g/L, 0.2 g/L, 0.3 g/L and 0.4 g/L. conductance indicated that the potential of photosynthetic A portable gas exchange system GFS-3000 (Waltz, activity of the plants decreased in all treatments compared Effeltrich, Germany) has been used for photosynthesis to control. Kummerova et al. have been shown the same measurement at following parameters: .air flow into the behavior in Lemna minor plants treated with 0.1 mg µ cuvette - 750 mol/s, CO2 - 385 ppm, PARtop 1000 mol diclofenac [35]. They have been demonstrated that values m-2 s-1, relative humidity 65%, leafs temperature 25oC. of non-photochemical quenching (NPQ) are increasing A part of the flow from the gas exchange system cuvette until more than 60 % relative to control for plants treated has been deviated via a T tube and sampled in a tube filled with diclofenac. Such behavior indicates an excess-radiant with adsorbent as described in [36]. energy dissipation to heat in photosynthetic system II (PSII) Tubes were desorbed using a TD 20 thermodesorber antenna complexes during light-adapted state. (Shimadzu, Japan) and injected in a gas chromatograph coupled with mass spectrometer (Shimatzu 2010 plus, Volatile organic compounds emission GCMSTQ8040, Japan). The carrier gas was helium and a Emission rates of lipoxygenase pathway volatiles capillary column (1 Accent MS column OPTIMA Germany) (named green leaf volatiles) and monoterpenes were (50 m × 0.2 mm, film thickness 0.33 µm) has been used analyzed in order to show the effects of diclofenac on bean for separation of the compounds. The program used to plants. The green leaf volatiles are emitted by plants in separate the volatile compounds is: 40 °C for 1 min-1, 9°C case of different abiotic or biotic stresses (see for review min -1 at 120°C, 2 °C min -1 at 190°C, 20°C min -1 at 240°C, [38]). Phaseolus vulgaris L. leaves emitted (Z)-3-hexenol 240°C for 5 min. (scheme 2) as a response of diclofenac stress. Even more, The mass spectrometer was operated in electron- the emission rates are scaling up with diclofenac impact mode (EI) at 70 eV, in the scan range m/z 48–400, concentrations as this compound could be used as stress the transfer line temperature was set at 250 °C and ion- marker. Our results are in good correlation with other source temperature at 200 °C. The compounds have been published papers which have been shown that different identified using their mass spectra by comparison with C6 aldehydes emission from tomato leaves scale increase the mass spectra from database (NIST 14.0). with stress temperatures [39]. For the pigments analysis we used the same method as described in [37]. Fig. 2. Emissions of Results and discussions (Z)-3-hexenol from The influence of diclofenac on photosynthetic parameters Phaseolus vulgaris L. Net assimilation rates and stomatal conductance to cv. Ecaterina plants in water vapors are significant different relative to control for response to plants treated even with 0.1 g/L diclofenac. For both diclofenac treatments parameters a linear decrease has been observed (fig. 1).

The emission of 3 different monoterpenes (α-pinene, camphene and 3-carene) was found in case of diclofenac stress. Phaseolus vulgaris L. is not a constitutive monoterpene emitter under non-stressed conditions, monoterpenes emission being typically induced in response to abiotic or biotic stresses [40, 41]. The α-pinene and camphene (scheme 2) emissions are statistically significant higher than control but there is

Fig. 1. Changes in net assimilation rate and, stomatal conductance to water vapor per unit projected leaf area in Phaseolus vulgaris L. Fig. 3. Emissions of monoterpenes from Phaseolus vulgaris L. cv. cv. Ecaterina plants treated with diclofenac Ecaterina plants in response to diclofenac treatments

Scheme 2. Structures of volatile organic compounds emitted in case of diclofenac stress on Phaseolus vulgaris L.: 1) 3-hexenol, 2) α-pinene, 3) camphene, and 4) 3-carene

REV.CHIM.(Bucharest)♦68♦No. 9 ♦2017 http://www.revistadechimie.ro 2077 no statistical difference in the emission in case of all 4 9.ACUNA, V., GINEBREDA, A., MOR, J. R., PETROVIC, M., SABATER, S., treatments with diclofenac. In contrast, 3-carene emission SUMPTER J., BARCELO, D., Environ. Internat., 85, 2015, p. 327. is increasing with diclofenac soil concentrations. The fact 10.HUSCHEK, G., HANSEN, P. D., MAURER, H. H., KRENGEL, D., KAYSER, could be explain by the disturbance in methylerythritol A., Environ. Tox. 19, 2004, p. 226. phosphate pathway (MEP) in plastids due to diclofenac. 11.JONES, O. A. H., VOULVOULIS, N., LESTER, J. N., Water Research, 36, 2002, p. 5013. Changes in chlorophylls and β-carotene composition 12.TIT, D. M., BUNGAU, S., NISTOR CSEPPENTO, C., COPOLOVICI, D. The roles of the carotenoids are to protect against M., BUHAS, C., J. Environ. Prot. Ecol., 17, nr. 4, 2016, p. 1425. oxidative damage and from photo inhibition but due to 13.POPESCU, D. E., BUNGAU, C., PRADA, M., DOMUTA, C., BUNGAU, oxidative severe stress conditions they could be rapidly S., TIT, D. M., J. Environ. Prot. Ecol., 17, nr. 3, 2016, p. 1011. destroyed [42]. The chlorophyll a and b concentrations 14.BUNGAU, S., BUNGAU, C, TIT, D. M., J. Environ. Prot. Ecol., 16, nr. 1, 2015, p. 56. decrease for plants leaves treated with diclofenac as can 15.BUNGAU, S, SUCIU, R., BUMBU, A., CIOCA, G., TIT, D. M., J. Environ. be seen in figure 4. Prot. Ecol., 16, nr. 3, 2015, p. 980. 16.DUMITREL, G.-A. , GLEVITZKY, M., POPA, M., VICA, M. L., J. Environ. Prot. Ecol., 16, nr. 3, 2015, p. 850. 17.VICA, M., POPA, M., DUMITREL, G.-A., GLEVITZKY, M., TODORAN, A., J. Environ. Prot. Ecol., 15, nr. 1, 2014, p. 64. 18.POPA, M., GLEVITZKY, M., POPA, D. M., DUMITREL, G.-A., J. Environ. Prot. Ecol., 15, nr. 4, p. 1543. 19.GLEVITZKY, M., VICA, M., POPA, M., AXINTE, R., VARVARA, S., J. Environ. Prot. Ecol., 12, nr. 3, 2011, p.1110. 20.POPA, M., VINTAN, D., ROXANA, B., POPA, D., J. Environ. Prot. Ecol., 15, nr. 1, 2014, p. 53. 21.ORBECI, C., NECHIFOR, G., STANESCU, R., Environ. Eng. Manag. J., 13, nr. 9, 2014, p. 2153. 22.DIACONU, I., GIRDEA, R., CRISTEA, C., NECHIFOR, G., RUSE, E., TOTU, E. E., Rom. Biotech. Lett., 15, nr. 6, 2010, p. 5702. Fig. 4. Effects of treatments with diclofenac on the chlorophyll a β 23.MASU, S., NICORESCU, V., POPA, M., J. Environ. Prot. Ecol., 16, nr. and b and -carotene contents of Phaseolus vulgaris L. cv. 3, 2015, p. 968. Ecaterina plants 24.DUMITREL, G.A., GLEVITZKY, M., POPA, M., CHIRILA, D., PALEA, It has been demonstrated that chlorophyll fluorescence A., J. Environ. Prot. Ecol., 18, nr. 1, 2017, p. 55. is not influenced by the very low concentration of diclofenac 25.MASU, S., DRAGOMIR, N., POPA, M., J. Environ. Prot. Ecol., 14, nr. (less than 0.1 mg/L) but for concentrations higher than 0.1 3, 2013, p. 1901. g/L both pigments concentration decreased, which means 26.POPA, M., GLEVITZKY, M., POPA, D., VARVARA, S., DUMITREL, G.-A., that there are exhibited important effects on PSII. J. Environ. Prot. Ecol., 13, nr. 4, 2012, p. 2123. The presence of diclofenac influences the β-carotene 27.BALINT, R., NECHIFOR, G., AJMONE-MARSAN, F., Environ. Sci.- content in the bean leaves. Even those carotenoids play an Processes & Impacts, 16, nr. 2, 2014, p. 211. 28.BALINT, R., ORBECI, C., NECHIFOR, G., PLESCA, M., AJMONE- important role in photo protection of photosynthetic MARSAN, F., Rev. Chim.(Bucharest), 64, no. 11, 2013, p. 1218.29. apparatus, in the presence of diclofenac their function is TERNES, T. A., JOSS A., SIEGRIST, H., Environ. Scie. & Tech. 38, 2004, not exhibited very clear. p. 392A. In some of our previous studies we have shown the 30.ZANURI, N. B. M., BENTLEY, M. G., CALDWELL, G. S., Marine same moderate effect on photosynthetic pigments of Environ.l Research, 127, 2017, p. 126. antibiotics and textile days [37, 41, 43] 31.CUNHA, S. C., PENA A., FERNANDES, J. O., Environ. Pollution, 225, 2017, p. 354. Conclusions 32.HOEGER, B., KOLLNER, B., DIETRICH, D. R., HITZFELD, B., Aquatic It has been shown that diclofenac could affect the plants Toxicology, 75, 2005, p. 53. photosynthetic parameters and could influence the 33.ERICSON, H., THORSEN G., KUMBLAD, L., Aquatic Toxicology, 99, metabolic pathways. Green leaf volatiles are scaling up 2010, p. 223. with the diclofenac concentration and could be used as a 34.ZIOLKOWSKA, A., PIOTROWICZ-CIESLAK, A. I., RYDZYNSKI, D., stress signals. ADOMAS, B., NALECZ-JAWECKI, G., Polish J. Environ. Studies, 23, The results attained in the study help us to better 2014, 263-269. estimate the influence of medicines on plants. 35.KUMMEROVA, M., ZEZULKA, S., BABULA P., TRISKA, J., J. Hazardous Materials, 302, 2016, p. 351. References 36.KANNASTE, A., COPOLOVICI L., NIINEMETS, U., Methods Mol. Biol., 1.HE, B. S., WANG, J., LIU, J., HU, X. M., Chemosphere, 181, 2017, p. 1153, 2014, p. 161. 178. 37.OPRIS, O., COPACIU, F., SORAN, M. L, RISTOIU, D., NIINEMETS, U., 2.TITA, B., MORGOVAN, C., TITA, D., NEAG, T. A., Rev. COPOLOVICI, L., Ecotox. Environ. Safety, 87, 2013, p. 70. Chim.(Bucharest), 67, no. 1, 2016, p. 38. 38.COPOLOVICI, L., NIINEMETS, U., Environmental impacts on plant 3.TITA, B., BANDUR, G., TITA, D., Rev. Chim.(Bucharest), 64, no. 6, volatile emission, Eds. J. D. Blande and R. Glinwood Springer, New 2013, p. 569. York, 2016. 4.TITA, D., FULIAS, A., TITA, B., J. Therm. Anal. Calorim., 111, nr. 3, 39.COPOLOVICI, L., KAENNASTE, A., PAZOUKI, L., NIINEMETS, U., J. 2013, p.1979. Plant Physiol., 169, 2012, p. 664. 5.TITA, B., STEFANESCU, M., TITA, D., Rev. Chim.( Bucharest), 62, no. 40.COPOLOVICI, L., KAENNASTE, A., REMMEL T., NIINEMETS, U., 11, 2011, p. 1060. Environ. Experim. Botany, 100, 2014, p. 55. 6.DIAZ-GONZALEZ, F., SANCHEZ-MADRID, F., European J. Immunol., 41.COPACIU, F., OPRIS, O., COMAN, V., RISTOIU, D., NIINEMETS, U., 45, 2015, p. 679. COPOLOVICI, L., Water Air and Soil Pollution, 2013, pp. 224. 7.LONAPPAN, L., BRAR, S. K., DAS, R. K., VERMA, M., SURAMPALLI, R. 42.MUNNE-BOSCH, S., PENUELAS, J., Planta, 217, 2003, p. 758. Y., Environ. Internat., 96, 2016, p. 127. 43.COPOLOVICI, L., NIINEMETS, U., Plant Cell Environ., 33, 2010, p. 8.ZHANG, Y. J., GEISSEN, S. U., GAL, C., Chemosphere, 73, 2008, p. 1582. 1151. Manuscript received: 10.04.2017 2078 http://www.revistadechimie.ro REV.CHIM.(Bucharest)♦68♦No. 9 ♦2017 FARMACIA, 2017, Vol. 65, 5 ORIGINAL ARTICLE THE INFLUENCE OF RESIDUAL ACETAMINOPHEN ON PHASEOLUS VULGARIS L. SECONDARY METABOLITES

MONICA TĂŞCHINĂ1#, DANA MARIA COPOLOVICI1#, SIMONA BUNGĂU2#, ANDREEA IOANA LUPITU1#, LUCIAN COPOLOVICI1#*, CIPRIAN IOVAN2#

1“Aurel Vlaicu” University of Arad, Faculty of Food Engineering, Tourism and Environmental Protection, Institute of Research-Development-Innovation in Technical and Natural Sciences, 2 Elena Drăgoi Street, 310330, Arad, Romania 2University of Oradea, Faculty of Medicine and Pharmacy, 10 Piaţa 1 Decembrie St., 410073, Oradea, Romania

*corresponding author: [email protected] #All the authors have equally contributed to this article. Manuscript received: January 2017

Abstract This study investigated the influence of paracetamol as a pollutant from wastewater on the beans plant (Phaseolus vulgaris L.). We used a portable gas exchange system to determine the photosynthetic parameters, gas chromatography-mass spectrometry (GC-MS) technique to determine the volatile organic compounds (VOC) produced by plants, and ultra-high performance liquid chromatography (UHPLC) to analyse the plants pigments (chlorophylls and β-carotene). The plants were grown in pots and treated with diverse concentrations of paracetamol. The residual concentration of paracetamol from the environment can be detrimental to the growth and development of the plants. Paracetamol influences the photosynthetic parameters (net assimilation rate and stomatal conductance), the emission rate of (Z)-3-hexenol, the concentration of 3-carene (one out the 3 monoterpenes identified in plant emission besides α-pinene and camphene), and β-carotene on Phaseolus vulgaris L. Chlorophyll was not influenced by paracetamol concentration.

Rezumat Acest studiu a investigat influența paracetamolului ca poluant din ape reziduale asupra plantei Phaseolus vulgaris L. Am utilizat un sistem portabil de schimb de gaze pentru a determina parametrii de fotosinteză, tehnica de cromatografie de gaze- spectrometrie de masă (GC-MS) pentru a determina compușii organici volatili (COV) ai plantelor și cromatografia de lichide de ultra înaltă performanță (UHPLC) pentru analiza pigmenților (clorofile și β-caroten). Plantele au fost crescute în ghivece și tratate cu soluții apoase cu diferite concentrații de paracetamol. Concentrația reziduală a paracetamolului din mediul înconjurător poate fi în detrimentul creșterii și dezvoltării plantelor. Paracetamolul influențează parametrii de fotosinteză (rata netă de asimilare și conductivitatea stomatală), rata de emisie a (Z)-3-hexenolului, concentrația 3-carenului și concentrația β-carotenului la Phaseolus vulgaris L. Clorofilele nu au fost influențate de concentrațiile de paracetamol.

Keywords: acetaminophen (paracetamol), Phaseolus vulgaris L., photosynthetic parameters, volatile organic compounds, pigments

Introduction population used paracetamol during six months [5, 32]. Acetaminophen (paracetamol) is an antipyretic and A great concern about the influence of different analgesic medicine, mostly used for mild to moderate pharmaceutical products and also of their metabolites pain and to reduce the fever. Paracetamol is the most on the environment has been rising due the fact that commonly prescribed drug for patients in case of different antidepressants, antibiotics, antihistamines, moderate fever [2, 21] and the first choice analgesic analgesics have been found in surface, ground, and including oral diseases [8, 13]. Due to the fact that drinking waters [6, 7, 12, 14, 15, 24, 27, 29]. The paracetamol is one of the most recommended over- concentrations of paracetamol in River Doubs and the-counter (OTC) drug, its market was valued at the River Lounge have been found to be 273 and around 801.3 million USD in 2014 and is expected 158 ng/L, respectively, being one of the most critical to reach 999.4 million USD in 2020 [33]. It has been pollutant after sulfamethoxazole and ofloxacine shown in a report that in the U.S. 36.5% of the inter- [9]. Previous studies have found paracetamol at viewed population had used paracetamol in two concentrations above 65 µg/L in the Tyne River, weeks [3], while in Germany 5.2% of a consulted UK [30]. In a recent study there were observed population used it in 7 days [31]. In Sweden, one concentrations of 12.42 µg/L paracetamol in the study reported that more than 70% of the population waste water treatment plants influents of several had used paracetamol during three months [20]. In rivers from Romania [17]. The ecological influence Romania, studies showed that self-medication of paracetamol has been demonstrated on bivalve as administrated by 46.6% from the interviewed 709 FARMACIA, 2017, Vol. 65, 5 Corbicula fluminea, crustaceans (Daphnia magna) from the cuvettes has been deviated via a T tube or fish (Oncorhynchus mykiss and Anguilla anguilla) and sampled in the tube. [16]. Paracetamol induced oxidative stress in the The air flow rate was 200 mL/min and the adsorption amphipod species of Hyalella Azteca [18]. The time was 20 minutes for each sample. The carbotrap published studies related to the influence of paracetamol tubes were initially conditioned with helium for 60 on plants are scarce. Anyway, the inhibition of Triticum minutes to remove possible contaminants. After aestivum seedlings growth and changes in photo- sampling, the tubes were stored at -24°C until analysis. synthetic pigments in the presence of paracetamol Tubes were desorbed using a TD 20 thermodesorber has been demonstrated [1]. The researchers have (Shimadzu, Japan). A gas chromatograph coupled shown that chlorophyll synthesis in wheat leaves with mass spectrometer (Shimadzu 2010 plus, has been inhibited by an exposure of 14 days at GCMSTQ8040, Shimadzu, Japan) was used for 2.8 mg/L paracetamol. Soluble protein peroxidase determination. The carrier gas was helium. VOCs and superoxide dismutase in wheat seedlings have were separated using a capillary column (1 Accent been affected only by exposure to a very high MS column OPTIMA, Germany) (50 m × 0.2 mm, concentration of paracetamol. In another study, the film thickness 0.33 µm). The program used to separate total polyphenols content has been increased for the volatile compounds was: 40°C for 1 min, 9°C/min Lemna minor plants treated with very low paracetamol at 120°C, 2°C/min at 190°C, 20°C/min at 240°C, concentration (0.1 mg/L) [23], the polyphenols as 240°C for 5 min. The mass spectrometer was operated well as other active substances content of different in electron-impact mode (EI) at 70 eV, in the scan plants varying obviously depending on a number of range m/z 48 - 400, the transfer line temperature other factors [4, 28]. In the present study we examined was set at 250°C and ion-source temperature at the influence of paracetamol on bean (Phaseolus 200°C. For the identification of the compounds in vulgarisL.) as a model plant. the samples, standard solutions and the mass spectra from NIST 14.0 database were used. Materials and Methods Pigments analysis Bean leaf samples of 4 cm2 were frozen in liquid All commercial chemicals and solvents were of nitrogen. The pigments were extracted in ice-cold reagent grade and were used without further acetone (100%) in the presence of calcium carbonate. purification. Acetone and paracetamol were purchased Then the extract was centrifuged with a Hettich from Merck, Germany. 380R Universal centrifuge (Hettich GmbH, Germany) Plant material at 0°C and 9500 g for 3 minutes. The supernatant Bean plants (Phaseolus vulgaris L.) var. Odir, GBBR, was removed and the extraction was repeated till Bekescsaba Hungary were seeded in 3 L pots and the supernatant became colourless. The pooled grown for 3 weeks under artificial light at a rate of supernatant was reduced to 1 mL volume and filtered 300 µmol/m2·s. The light/dark period was 12/12 through a 0.45 mm PTFE membrane filter (PALL, hours. The temperature was 25°C. The plants were USA). The UHPLC (NEXERA8030, Shimadzu, Japan) watered every two days. Four different paracetamol was calibrated using: chlorophyll a, chlorophyll b, and aqueous solutions were used for plant treatment: 1 β-carotene. The extracts were analysed for these g/L, 2 g/L, 3 g/L and 4 g/L. pigments by using a NUCLEOSIL 100-3 C18 Photosynthestic parameters determination reversed-phase column (Macherey-Nagel AG, Germany) Photosynthesis measurements were conducted with and the method described in literature by Opriș et a portable gas exchange system GFS-3000 (Waltz, al. [26]. Effeltrich, Germany). The system has an environmental- Statistics controlled cuvette with 8 cm2 window area. Each All experiments were performed in triplicates with time, a leaf fully filling the cuvette area was enclosed. independent samples of plants. The means were The following measurement parameters that have been statistically compared with ANOVA and then post set were kept constant during the measurements as hoc Tukey’s test was applied using ORIGIN8 (Origin follows: air flow into the cuvette: 750 µmol/s, CO : 2 Lab Corp., Northampton, MA, USA). All statistical 385 ppm, PARtop 1000 µmol/m2·s, relative tests were considered significant at p ˂ 0.05. humidity 65%, cell temperature 25°C. The bean leaves were stabilized under the standard conditions Results and Discussion until stomata opened and steady-state CO2 and water vapour exchange rates were reached. The influence of paracetamol on photosynthetic Volatile organic compounds (VOC) sampling and parameters GC–MS analysis Net assimilation rates and stomatal conductance to VOC sampling was performed on adsorbent cartridges water vapours were not significant different relative via the outlets of each cuvette, using steel tube with to control for plants treated with 1 g/L paracetamol. size ¼ filled with adsorbent [22]. A part of the flow Assimilation rate decreased with more than 20% in 710 FARMACIA, 2017, Vol. 65, 5 case of the plants treated with 2 g/L paracetamol. were no significant differences between plants treated The same trend has been observed for stomatal with higher concentration than 2 g/L paracetamol conductance to water vapour, but in this case there (Figure 1).

6 80 a ab 5 a a ab

) b ) 60 b -1 -1

s 4 b s -2

bc -2 c 3 40

mol m µ

( 2 (mmol m

to wa te r va pour 20 Net assimilation rate 1 Stomatal conductance

0 0 0 1 2 3 4 0 1 2 3 4 Paracetamol concentration (g/L) Paracetamol concentration (g/L) Figure 1. Changes in net assimilation rate and stomatal conductance to water vapour per unit projected leaf area in Phaseolus vulgaris L. cv. Odir plants treated with paracetamol

Such behaviour of assimilation rate and stomatal 0.05 conductance indicates that the potential photosynthetic activity of the plant decreased in all treatments 0.04 ) compared to control. Our results are consistent with -1 s literature [23] which has shown a decrease in the -2 0.03

ratio of chlorophyll fluorescence for plants treated with paracetamol. 0.02 r = 0.946 Volatile organic compounds emission (nmol m To gain further into the effects of paracetamol on 0.01 )-3-hexenol emission rate Z plants, emission rates of green leaf volatiles (known as ( 0.00 lipoxygenase pathway products) and monoterpenes 0 1 2 3 4 were analysed. The green leaf volatile compound Paracetamol concentration (g/L) emitted by Phaseolus vulgaris L. leaves identified in the samples was (Z)-3-hexenol (Figure 2). As well, Figure 3. three monoterpenes: α-pinene, camphene and 3- Emissions of (Z)-3-hexenol from Phaseolus vulgaris L. cv. Odir plants in response to carene (Figure 2) were also found in the emission. paracetamol treatments

H C 3 0.025 OH alpha-pinene (a) (b) (c) (d) camphene 0.020 Figure 2. 3-carene

Chemical structures of the three monoterpenes )

-1 0.015

found: 3-hexenol (a), α-pinene (b), camphene (c), s -2 and 3-carene (d) 0.010

Anyway, the concentrations of all emitted compounds (nmol m

Emission rate 0.005 were very low in the control plants. The stressed plants with low paracetamol concentrations exhibited 0.000 a very low emission (Figure 3), while the high 0 1 2 3 4 concentration of paracetamol induced quite high Paracetamol concentration (g/L) concentration emission of green leaf volatile. Figure 4. However, the emission rate for this lipoxygenase Emissions of monoterpenes from Phaseolus pathway product increased linearly with the vulgaris L. cv. Odir plants in response to concentration of the stress. paracetamol treatments Bean is not a constitutive monoterpene emitter under non-stressed conditions. Furthermore, monoterpenes In the present study, very low emissions of three are typically induced in response to abiotic or biotic monoterpenes have been found: α-pinene, camphene stresses [10, 11, 12]. and 3-carene (Figure 4). From them, only in the case of 3-carene there were different emissions for 711 FARMACIA, 2017, Vol. 65, 5 increasing stress concentrations. For α-pinene and The present study revealed the influence of low camphene the pathway flux going to monoterpene concentrations of paracetamol on growth and synthesis was not disturbed. development of bean plant that is an important food Changes in chlorophylls and β-carotene composition product. Nowadays, there are scarce information Carotenoids can be rapidly destroyed by oxidative about the influence of long-term exposure of small severe stress conditions therefore are no longer concentrations of pharmaceuticals on both plant available to protect against oxidative damage and and human health and development. There is an from photoinhibition [25]. In our study, paracetamol actual need for studies that approach the long-term did not influence significantly the chlorophyll effects of low doses of medicines and their concentrations (Figure 5). degradation products (from the environment and

food products exposed to them) on human health. A 500

) more challenging study will include the behaviour -2 C hlorophyll a of mixtures of medicines and also the interactions a C hlorophyll b 400 a of mixtures of drugs with other pollutants from the a a a environment on human health. 300

Conclusions 200 Paracetamol is influencing the growth and development of bean plant (Phaseolus vulgaris L.). Our data 100 a a a a a have shown that photosynthetic parameters, such as

Chlorophyll content (mg m net assimilation rate and stomatal conductance, 0 0 1 2 3 4 decreased when higher concentrations than 2 g/L of paracetamol were used. We have also determined that Paracetamol concentration (g/L) Figure 5. the emission of green leaf volatile scale strengthens Effects of treatments with paracetamol on the with the stress (paracetamol concentration), while chlorophyll a and b contents of Phaseolus vulgaris monoterpenes emission is not affected by paracetamol. L. cv. Odir plants Even more, β-carotene concentrations decrease with the drug concentrations. The β-carotene content in the bean leaves was significantly influenced by the concentration of Acknowledgement paracetamol (p < 0.05) (Figure 6). We observed a Authors acknowledge the POSCCE Nr. 621/2014 decreased quantity of β-carotene with more than project co-funded by European Union and Romanian 20% in the case of plants treated with 2 g/L Government. paracetamol and significant decreases at higher concentration of paracetamol (p < 0.05). References

100 1. An J., Zhou Q.X., Sun F.H., Zhang L., Eco- ) toxicological effects of paracetamol on seed -2 a germination and seedling development of wheat a (Triticum aestivum L.). J. Hazard Mater., 2009; 169: 751-757. a ab 2. Bailey B., Trottier E.D., Managing Pediatric Pain 50 b in the Emergency Department. Pediatr. Drugs, 2016; 18: 287-301. 3. Boudreau D.M., Wirtz H., Von Korff M., Catz S.L., St John J., Stang P.E., A survey of adult awareness and use of medicine containing acetaminophen. -carotene content (mg m

β Pharmacoepidemiol. Drug Saf., 2013; 22: 229-240. 0 0 1 2 3 4 4. Bungău S., Bâldea I., Copolovici L., Determination of ascorbic acid in fruit using a Landolt type method. Paracetamol concentration (g/L) Rev. Chim. Bucharest, 2003; 54(3): 213-216. Figure 6. 5. Bungău S., Bungău C., Ţiţ D.M., Pallag A., The Changes in β-carotene content in Phaseolus influence of specialized university studies on self- vulgaris L. cv. Odir plants in response to treatments medication. Rev. Rom. Bioet., 2015; 13(1): 153-158. with paracetamol 6. Bungău S., Bungău C., Ţiţ D.M., Studies about last stage of product lifecycle management for a The same effect on photosynthetic pigments of pharmaceutical product. J. Environ. Prot. Ecol., antibiotics and textile dyes has been found in 2015; 16(1): 56-62. previous studies [10, 26].

712 FARMACIA, 2017, Vol. 65, 5 7. Bungău S., Suciu R., Bumbu A., Cioca G., Ţiţ 20. Hedenrud T., Hakonse H., Purchase habits, use of D.M., Study on hospital waste management in paracetamol, and information sources on a reregulated medical rehabilitation clinical hospital, Băile Felix. Swedish pharmacy market: A population-based study. J. Environ. Prot. Ecol., 2015; 16(3): 980-987. Health Policy, 2017; 121: 35-41. 8. Cazacu I., Leucuta D.C., Baciut M., Baciut G., 21. Kanabar D.J., A clinical and safety review of Mogosan C., Haramburu F., Fourrierréglat A., paracetamol and ibuprofen in children. Inflammo- Loghin F., Acute postoperative pain and analgesics pharmacology, 2017; 25: 1-9. consumption in oral and maxillofacial surgery. 22. Kannaste A., Copolovici L., Niinemets U., Gas Farmacia, 2016; 64(4): 539-543. Chromatography-Mass Spectrometry method for 9. Chiffre A., Degiorgi F., Bulete A., Spinner, L., Badot determination of biogenic volatile organic compounds P.M., Occurrence of pharmaceuticals in WWTP emitted by plants. Pl. Isoprenoids: Meth. Prot., effluents and their impact in a karstic rural catchment 2014; 1153: 161-169. of Eastern France. Environ. Sci. Pollut. Res., 2016; 23. Kummerova M., Zezulka S., Babula P., Triska J., 23: 25427-25441. Possible ecological risk of two pharmaceuticals 10. Copaciu F., Opriş O., Coman V., Ristoiu D., Niinemets diclofenac and paracetamol demonstrated on a model U., Copolovici L., Diffuse water pollution by plant Lemna minor. J. Hazard Mater., 2016; 302: anthraquinone and azo dyes in environment 351-361. importantly alters foliage volatiles, carotenoids and 24. Mompelat S., Thomas O., Le Bot B., Contamination physiology in wheat (Triticum aestivum). Water Air levels of human pharmaceutical compounds in Soil Pollut., 2013; 224: 1478. French surface and drinking water. J. Environ. 11. Copolovici L., Kaennaste A., Remmel T., Niinemets Monit., 2011; 13: 2929-2939. U., Volatile organic compound emissions from Alnus 25. Munné-Bosch S., Peñuelas J., Photo- and anti- glutinosa under interacting drought and herbivory oxidative protection, and a role for salicylic acid stresses. Environ. Exp. Bot., 2014; 100: 55-63. during drought and recovery in field-grown Phillyrea 12. Copolovici L., Niinemets U., Flooding induced angustifolia plants. Planta, 2003; 217: 758-766. emissions of volatile signalling compounds in three 26. Opriş O., Copaciu F., Soran M.L., Ristoiu D., tree species with differing waterlogging tolerance. Niinemets U., Copolovici L., Influence of nine Plant Cell Environ., 2010; 33: 1582-1594. antibiotics on key secondary metabolites and 13. Dumitrache M.A., Ionescu E., Sfeatcu R., Ginghina physiological characteristics in Triticum aestivum: O., Burcea Dragomiroiu G.T.A., Petre A., The Leaf volatiles as a promising new tool to assess pharmacist's role in preventive and pharmaceutical toxicity. Ecotox. Environ. Saf., 2013; 87: 70-79. treatment for oral diseases. Farmacia, 2016; 64(6): 27. Orbeci C., Nechifor G., Stănescu R., Removing 966-969. toxic compounds from wastewater. Environ. Eng. 14. Fodor A., Petrehele A., Bungău S., Ţiţ D.M., Manag. J., 2014; 13(9): 2153-2158. Tetracycline and metabolites in agricultural 28. Pallag A., Bungău S.G., Ţiţ D.M., Jurca T., Sîrbu farmland soil and Petroselinum Crispum var. V., Honiges A., Horhogea C., Comparative study of Neapolitanum roots and leafs. Annals Oradea polyphenols, flavonoids, and chlorophylls in Equisetum Univ., Fasc. Environ. Prot., 2013; XX(A): 1-6. Arvense T. populations. Rev. Chim. Bucharest, 15. Fodor A., Petruş-Vancea A., Petrehele A.I.G., Bungău 2016; 67(3): 530-533. S.G., Accumulation Properties and Foliar Limb 29. Popescu D.E., Bungău C., Prada M., Domuţa C., Effects of Tetracycline and Metabolites in Lactuca Bungău S., Ţ iţ D.M., Waste management strategy sativa L., Eruca Sativa L. and Spinacia oleracea at a public university in Smart City context. J. L.. Nat. Res. Sust. Develop., 2015; 2015: 57-66. Environ. Prot. Ecol., 2016; 17(3): 1011-1020. 16. Freitas R., Coelho D., Pires A., Soares A., Figueira 30. Roberts P.H., Thomas K.V., The occurrence of selected E., Nunes B., Preliminary evaluation of Diopatra pharmaceuticals in wastewater effluent and surface neapolitana regenerative capacity as a biomarker waters of the lower Tyne catchment. Sci. Total Environ., for paracetamol exposure. Environ. Sci. Pollut. 2006; 356: 143-153. Res., 2015; 22: 13382-13392. 31. Sarganas G., Buttery A.K., Zhuang W., Wolf I.K., 17. Gheorghe S., Petre J., Lucaciu I., Stoica C., Nita- Grams D., Rosario A.S., Scheidt-Nave C., Knopf Lazar M., Risk screening of pharmaceutical H., Prevalence, trends, patterns and associations of compounds in Romanian aquatic environment. analgesic use in Germany. BMC Pharmacol. Toxicol., Environ. Monit. Assess., 2016; 188(6): 379. 2015; 16: 1-13. 18. Gomez-Olivan L.M., Neri-Cruz N., Galar-Martinez M., 32. Ţiţ D.M., Bungău S., Nistor Cseppento C., Copolovici Vieyra-Reyes P., Garcia-Medina S., Razo-Estrada C., D.M., Buhaş C., Disposal of unused medicines Dublan-Garcia O., Corral-Avitia A.Y., Assessing resulting from home treatment in Romania. J. the oxidative stress induced by paracetamol spiked Environ. Prot. Ecol., 2016, 17(4): 1425-1433. in artificial sediment on Hyalella azteca. Water Air 33. ***www.atozresearch.com Soil Pollut., 2012; 223: 5097-5104. 19. Grigore R.A., Mihai A.M., Ciolan D.F., Tarmure V., Ionescu E., Pain therapy associated with orthodontic treatment. Farmacia, 2014; 62(6): 1062-1071.

713

View publication stats Phytochemistry 147 (2018) 80e88

Contents lists available at ScienceDirect

Phytochemistry

journal homepage: www.elsevier.com/locate/phytochem

Diterpenoid fingerprints in pine foliage across an environmental and chemotypic matrix: Isoabienol content is a key trait differentiating chemotypes

* Astrid Kannaste€ a, , Lauri Laanisto a, Leila Pazouki a, b, Lucian Copolovici a, c, Marina Suhorutsenko a, Muhammad Azeem d, e, Lauri Toom f, Anna-Karin Borg-Karlson a, e, f, Ülo Niinemets a, g a Institute of Agricultural and Environmental Sciences, Estonian University of Life Sciences, Kreutzwaldi 1, Tartu 51014, Estonia b Department of Biology, University of Louisville, Louisville, KY 40292, USA c Institute of Technical and Natural Sciences Research-Development of “Aurel Vlaicu” University, 2 Elena Dragoi St., Arad 310330, Romania d Department of Chemistry, COMSATS Institute of Information Technology, Abbottabad 22060, Pakistan e Ecological Chemistry Group, Department of Chemistry, KTH, Royal Institute of Technology, 100 44 Stockholm, Sweden f Institute of Technology, University of Tartu, Nooruse 1, Tartu 50411, Estonia g Estonian Academy of Sciences, Kohtu 6, 10130 Tallinn, Estonia article info abstract

Article history: Diterpenoids constitute an important part of oleoresin in conifer needles, but the environmental and Received 20 March 2017 genetic controls on diterpenoid composition are poorly known. We studied the presence of diterpenoids Received in revised form in four pine populations spanning an extensive range of nitrogen (N) availability. In most samples, iso- 12 December 2017 abienol was the main diterpenoid. Additionally, low contents of (Z)-biformene, abietadiene isomers, Accepted 14 December 2017 manoyl oxide isomers, labda-7,13,14-triene and labda-7,14-dien-13-ol were quantified in pine needles. According to the occurrence and content of diterpenoids it was possible to distinguish ‘non diterpenoid pines’, ‘high isoabienol pines’, ‘manoyl oxide e isoabienol pines’ and ‘other diterpenoid pines’. ‘Non Keywords: ’ ‘ ’ ‘ ’ Chemotypic variability diterpenoid pines , high isoabienol pines and other diterpenoid pines were characteristic to the dry > ‘ ’ ‘ Diterpenoids forest, yet the majority of pines ( 80%) of the bog Laeva represented high isoabienol pines . Manoyl Isoabienol oxide e isoabienol pines’ were present only in the wet sites. Additionally, orthogonal partial least- Pinus sylvestris L. squares analysis showed, that in the bogs foliar nitrogen content per dry mass (NM) correlated to Terpenoid signatures diterpenoids. Significant correlations existed between abietadienes, isoabienol and foliar NM in ‘manoyl N content oxide e isoabienol pines’, and chemotypic variation was also associated by population genetic distance estimated by nuclear microsatellite markers. Previously, the presence of low and high D-3-carene pines has been demonstrated, but the results of the current study indicate that also diterpenoids form an independent axis of chemotypic differentiation. Further studies are needed to understand whether the enhanced abundance of diterpenoids in wetter sites reflects a phenotypic or genotypic response. © 2017 Elsevier Ltd. All rights reserved.

1. Introduction (Pinus sylvestris L.), have characteristically high longevity, which is associated with multiple traits, in particular, with presence of Conifer trees, including the extremely wide-spread Scots pine oleoresin in their woody tissues and leaves (Lewinsohn et al., 1991a,b). Oleoresin is contained in the resin ducts and it protects trees against a wide range of herbivores, fungal and bacterial pathogens, providing a very successful broad-spectrum defense * Corresponding author. system (Croteau et al., 1987; Lewinsohn et al., 1991a,b; Phillips and € E-mail addresses: [email protected] (A. Kannaste), [email protected] Croteau, 1999). The biosynthetic pathways of oleoresin constituents (L. Laanisto), [email protected] (L. Pazouki), [email protected] (L. Copolovici), [email protected] (M. Suhorutsenko), mazeemfsd@ have been widely investigated (Bohlmann and Keeling, 2008; hotmail.com (M. Azeem), [email protected] (L. Toom), [email protected] (A.-K. Borg- Keeling et al., 2010; Memari et al., 2013; Rosenkranz and Karlson), [email protected] (Ü. Niinemets). https://doi.org/10.1016/j.phytochem.2017.12.007 0031-9422/© 2017 Elsevier Ltd. All rights reserved. A. Kannaste€ et al. / Phytochemistry 147 (2018) 80e88 81

Schnitzler, 2013). Oleoresin is composed of turpentine fraction severity of simultaneous abiotic stressors in different ecosystems consisting of mono- and sesquiterpenoids and rosin fraction con- (Niinemets, 2010). Although the genetic variability within and sisting of diterpenoids and their derivatives, in particular diterpe- among pine populations has been studied to some extent (Pazouki noid acids (Phillips and Croteau, 1999). Oleoresin can effectively et al., 2016; Dering et al., 2017), our understanding of the associa- seal the injuries resulting from direct mechanical stresses such as tion between chemotypes and environment of pine populations is wind damage, but also from browsing by large herbivores, from still limited (Kannaste€ et al., 2013; Pazouki et al., 2016). Pinus spe- insect feeding and pathogen damage (Phillips and Croteau, 1999; cies including P. sylvestris can occupy highly varying habitats Loreto et al., 2000). including fertile well-drained mineral soils, excessively drained Diterpenoids are toxic to microorganisms and arthropods, and sandy soils or infertile waterlogged peat soils. Given this high thus play an important role in plant defense (Bailey et al., 1974; plasticity in coping with habitat conditions, it is surprising that the Ulubelen et al., 1994; Jasinski et al., 2001; Falara et al., 2010; environmentally-driven variations in key stress-dependent Koutsaviti et al., 2011; Chaturvedi et al., 2012). For example, specialized metabolites have been poorly examined across pine labdane-type diterpene (11E,13E)- labda-11,13-diene-8a,15-diol populations (Hornoy et al., 2015). Understanding such associations named as WAF-1 contributes to the activation of pathogen- or is particularly relevant as long-term stress-induced changes at the wound-induced reactions in tobacco leaves (Seo et al., 2003). molecular level can lead to formation of novel pine chemotypes Additionally, labdane-type diterpenoids (Fig. 1) participate in the (phenotypes with different chemical composition) (Jablonka, biosynthesis of diterpenoid acids, also called resin acids (Phillips 2013). To our knowledge, chemotypic differences associated with and Croteau, 1999; Lee et al., 2001; Keeling and Bohlmann, 2006), stress-dependent changes in diterpenoids have been only studied composition of which is significantly modified by biotic stress in Norway spruce (Picea abies) clones exposed to ozone stress (Kainulainen et al., 1993; Miller et al., 2005). Due to their toxicity, (Kainulainen et al., 1995). resin acids inhibit mycelial growth of pathogenic fungi carried by Although there are basic controls on specialized metabolism, bark beetles (Kopper et al., 2005). In Sitka spruce (Picea sitchensis), pathway fluxes to different terpenoid classes are often regulated accumulation of dehydroabietic acid is associated with tree resis- differently (Dudareva et al., 2004; Tholl, 2006; Chen et al., 2011). tance against white pine weevil (Pissodes strobe)(Robert et al., This is particularly true for differences among cytosol-synthesized 2010). Hence, obtaining detailed insight about the factors control- sesquiterpenoids and plastid-synthesized mono- and diterpe- ling the content and composition of diterpenoids in oleoresin will noids, but even within plastids, mono- and diterpene biosynthesis help to understand the overall role of diterpenoids in induced de- can be differently controlled by the availability of C10 substrate fense reactions in gymnosperms. geranyl diphosphate (GDP) and C20 substrate geranylgeranyl Pinus species are widely known for their extensive genetic diphosphate (GGDP) (Orlova et al., 2009; Rajabi Memari et al., variance (Dering et al., 2017). For example, pines of D-3-carene, a- 2013). In our previous study, which was based on pines from pinene (Hiltunen et al., 1975; Sjodin€ et al., 1996, 2000; Back€ et al., three wet sites and one dry site, we reported a large variation in 2012; Kannaste€ et al., 2013) and isoabienol chemotypes (Gref, mono- and sesquiterpenoid signatures over short spatial distances 1981) are known. Moreover, the resin composition of pines is in P. sylvestris populations spanning wet organic soils and dry strongly influenced by environmental factors (Hiltunen et al., 1975; mineral soils (Kannaste€ et al., 2013). Specifically, D-3-carene was Taft et al., 2015), especially due to the variability in timing and the main monoterpene effectively differentiating the populations (Kannaste€ et al., 2013). To further differentiate among the P. sylvestris chemotypes, we studied here the content and compo- sition of diterpenoids in the same four populations dispersed over a large gradient of soil conditions (Kannaste€ et al., 2013). We hy- pothesized that in addition to D-3-carene, diterpenoids form an additional axis of chemical differentiation for P. sylvestris pop- ulations. Based on the nutrient-dependent variation in mono- and sesquiterpenoid content across the sites (Kannaste€ et al., 2013), we suggested that diterpenoid contents in pine needles scale nega- tively with site N availability. In addition, given that the genetic distance among the studied populations has been estimated on the basis of microsatellite markers (Pazouki et al., 2016), we further hypothesized that chemotype-level differentiation is associated with genetic distance among the studied pine populations.

2. Results and discussion

2.1. Variations of diterpenoid contents and compositions

Diterpenoids were present in more than 90% of 218 analyzed samples (Table 1). High isoabienol content was characteristic to the majority of pine samples. Moreover, isoabienol contents were similar in the bog samples, but foliage isoabienol content was greater in dry forest samples than in bogs Laeva and Parika (Fig. 2). Isoabienol has been previously found in the needles of Juniperus communis (Adams et al., 2010) and in the wood of Cunninghamia ko nishii (Cheng et al., 2012); but also in P. sylvestris samples from Sweden and Finland (Gref, 1981; Manninen et al., 2002). In the Finnish and Swedish samples, it was only found in a few cases (Gref, Fig. 1. Molecular structures of diterpenoids quantified in Pinus sylvestris L. needles. 1981; Manninen et al., 2002). Thus, widespread and high rate of 82 A. Kannaste€ et al. / Phytochemistry 147 (2018) 80e88

Table 1 Distribution of non- and diterpenoid containing Pinus sylvestris L. samples across the sites.

Population Number of samples Non-diterpenoid pinesa Diterpenoid containing pines (%) (%) ‘Manoyl oxide - isoabienol pines’b ‘High isoabienol pines’b ‘Other diterpenoid pines’b (%) (%) (%)

Laeva (L) 69 4 6 81 9 Parika (P) 47 2 26 53 19 Kakaerdaja (K) 70 7 17 61 15 Manniku€ (M) 32 341 0 44 22

The first three sites are temperate raised bogs, while the Manniku€ site is a dry pine forest on sandy soil. a Samples were grouped according to the presence of diterpenoids to either `non-diterpenoid pines’ or ‘diterpenoid-containing pines’. b According to the content of isoabienol and manoyl oxide pines grouped as follows: needle extracts of ‘manoyl oxide - isoabienol pines’ contained 8.9 ± 1.3 mg g 1 of isoabienol; needle extracts of ‘high isoabienol pines’ contained 27.3 ± 2.1 mg g 1 of isoabienol; finally ‘other diterpenoid pines’ contained isoabienol at very low level of 0.068 ± 0.026 mg g 1 (average ± SE).

13-ol or vice versa seems to be related to hydration/dehydration of the respective diterpenoid (Fig. 1). In fact, biochemical hydration/ dehydration would also explain high correlation values between isoabienol and (Z)-biformene (Table 2). The coexistence of isoabienol and manoyl oxide revealed two different positive correlations among these compounds (Fig. 3). The first correlation (Fig. 3A) was characteristic only to ‘manoyl oxide e isoabienol samples’ from the wet sites (r ¼ 0.62, P < .001), while the second correlation appeared within ‘high isoabienol samples’ from both the wet sites and the dry forest (r ¼ 0.89, P < .001) (Fig. 3B, pine chemotypes discussed below). Hence, manoyl oxide and iso- abienol may be biosynthesized via different pathways. Cyclization/ decyclization would explain high correlation values between iso- abienol and manoyl oxides. However, in the perfume industry, ambreinolide, which has similar structure as manoyl oxide, is synthesized via oxidation of isoabienol (Kukovinets et al., 2001). Moreover, in a recent genetic engineering study, the diterpenoid profile was shown to depend on the substrate. For example, the Fig. 2. Total contents of isoabienol and other diterpenoids (mg g 1, mean ± SE) in combination of Z-abienol synthase NtABS from Nicotiana tabacum P. sylvestris needles of Parika (P), Kakerdaja (K), Laeva (L) and Manniku€ (M) populations (Sallaud et al., 2012) and copalyl diphosphate synthase from Salvia (Table 3 and Fig. S1 for the sites). Lowercase letters indicate significant differences fruticosa yielded (Z)-biformene, while the combination of NtABS across the sites (P < .05 according to one-way ANOVA followed by a Tukey's test). and 8-hydroxy copalyl diphosphate synthase from Cistus creticus yielded (Z)-abienol (Ignea et al., 2015). A recent review demon- this diterpenoid in Estonian samples was surprising. strates that the capacity for multi-substrate use is widespread Besides isoabienol, P. sylvestris foliage contained abietadiene, among species (Pazouki and Niinemets, 2016). abieta-8(14),13(15)-diene, (Z)-biformene, labda-7,14-dien-13-ol, labda-7,13,14-triene, manool, manoyl oxide and 13-epi-manolyl 2.2. Separation of pines according to diterpene profiles oxide (Fig. 1). The total content of these diterpenoids remained 1 below 1 mg g (Fig. 2). Generally, diterpenoids are not routinely Isoabienol and manoyl oxide were the dominant diterpenoids, analyzed together with mono- and sesquiterpenoids in conifer which affected the grouping of pines into clusters. Hence, based on studies. This is partly because of different extraction procedures the isoabienol and manoyl oxide contents per needle dry mass, the required for diterpenoids, and thus, diterpenes not normally pre- ‘diterpenoid containing pines’ were, in turn, classified into ‘manoyl sent in the essential oil fraction obtained by steam distillation oxide e isoabienol pines’, ‘high isoabienol pines’ and ‘other diter- (Ormeno~ et al., 2011), and also because different GC-MS analysis penoid pines’ in PCA plots (Fig. 4A and B). Normalization of diter- conditions required for diterpenes. We argue that future studies penoid contents per total diterpenes showed that the cluster of should also include detailed analysis of all fractions of conifer ‘other diterpenoid pines’ consisted of ‘abieta-8(14),13(15)-diene e oleoresin. manoyl oxide pines’ (Fig. 5A) and ‘abietadiene pines’ (Fig. 5B). Diterpenoids with similar molecular structures (Fig. 1) may be ‘Manoyl oxide e isoabienol pines’ contained on average 93% of biosynthesized via the common pathway. In addition to the total isoabienol and 6% of manoyl oxide (Fig. 5C), but among ‘high iso- contents of diterpenoids, it was possible to evaluate the total leaf abienol pines’ it was again possible to distinguish two different dry mass investment in diterpenoids biosynthesis (Table 2, Figs. 3 chemotypes (Fig. 5D and E). Both chemotypes were dominated by and 4). Correlation analyses (Table 2) demonstrated strong co- isoabienol, and the content of manoyl oxide remained below 1%. variances between isoabienol and (Z)-biformene contents per dry Nevertheless, the diterpenoid profiles of those chemotypes differed mass (r ¼ 0.81, P < .001), isoabienol and labda-7,14-dien-13-ol considerably (Monte Carlo test, P < .05). (r ¼ 0.71, P < .001), (Z)-biformene and labda-7,14-dien-13-ol (r ¼ 0.61, P < .001), abieta-8(14),13(15)-diene and labda-7,13,14- 2.3. Diterpenoids profiles in relation to site triene (r ¼ 0.62, P < .01), labda-7,13,14-triene and manoyl oxide (r ¼ 0.62, P < .01), and manool and 13-epi-manoyl oxide (r ¼ 0.81, In our previous study (Kannaste€ et al., 2013) the distribution of P < .05). The change from isoabienol synthesis to labda-7,14-dien- pine chemotypes according to the presence or absence D-3-carene A. Kannaste€ et al. / Phytochemistry 147 (2018) 80e88 83

Table 2 Correlation coefficients between diterpenoids (mg g 1)inP. sylvestris samples.

Diterpenoids Diterpenoids

Abietadiene Abieta-8(14),13(15)- (Z)- Isoabienol Labda-7,14-dien- Labda-7,13,14- Manool Manoyl 13-epi-Manoyl diene Biformene 13-ol triene oxide oxide

Abietadiene 1 0.43** 0.49** 0.47*** 0.40** 0.006 0.70 0.18 0.51*** Abieta-8(14),13(15)- 1 0.50** 0.20 0.26 0.62** e 0.045 0.17 diene (Z)-Biformene 1 0.81*** 0.61*** 0.14 0.12 0.33** 0.40** Isoabienol 1 0.71*** 0.002 0.36 e 0.40*** Labda-7,14-dien-13-ol 1 0.18 0.36 0.12 0.33* Labda-7,13,14-triene 1 0.54 0.62*** 0.20 Manool 1 0.30 0.81* Manoyl oxide 1 0.13 13-epi-Manoyl oxide 1

Level of significance at P-value of 0.05 (*), 0.01 (**) or 0.001 (***).

Fig. 3. Correlation between isoabienol and manoyl oxide expressed per unit dry mass Fig. 4. Loading-plot (A) and score-plot (B) of PCA analysis based on the diterpenoids 1 in P. sylvestris samples of ‘manoyl oxide - isoabienol pines’ and ‘high isoabienol pines’. per dry mass (mg g ) in the needles of P. sylvestris called ‘high isoabienol pines’ Correlation was significant at level of P < .001 (***). (yellow circle), ‘manoyl oxide e isoabienol pines’ (pink circle) and ‘other diterpenoid pines’ (green circle) (see Table 1). In the score plot, each symbol represents one sample of one tree of Parika (triangle symbols), Laeva (diamond symbols), Kakerdaja (circle symbols) and Manniku€ (red star symbols) sites. In the loading plot the quantified amounts of diterpenoids (see Table 4) increase with the distance from the origin of the was not related to the provenance of samples, although across the coordinate system. In total nearly 100% of data variance was explained by the first and bogs monoterpenoid profiles from Laeva samples differed consid- the second principal components (PCA 1 97%, PCA 2 2.9%). erably from those from Kakerdaja and Parika (Kannaste€ et al., 2013). In the present study, majority of bog samples contained high fi amounts of isoabienol (Table 1, Fig. 5) and were classi ed either as (r ¼ 0.49, P < .05, Fig. 6) showed that the distribution of pines' ‘ e ’ ‘ ’ manoyl oxide isoabienol pines or high isoabienol pines samples across the sites (Table 1) could be related to the high (Table 1). Conifer resin composition is mainly controlled geneti- acclimatization ability of pines to local environment via epigenetic fl cally, but it is also in uenced by environmental factors (Hiltunen factors (Bussotti et al., 2015). According to Baradat and Yazdani et al., 1975; Holliday et al., 2010; Taft et al., 2015). Pazouki et al. (1988) and Taft et al. (2015) the impact of environmental factors (2016) concluded that the high genetic variation in pines was on terpenoids is selective. Hence, taking into account the age of related to pollen dispersal in bogs. However, based on nuclear trees, site differences (see Kannaste€ et al., 2013), and the results of microsatellite markers (nSSR markers), pines samples from dryer earlier genetic studies (Baradat and Yazdani, 1988; Jablonka, 2013; and wetter areas of various bogs clustered differently (Pazouki et al. Hornoy et al., 2015; Taft et al., 2015; Pazouki et al., 2016), we (2016). Moreover, Mantel test based on nSSR markers (Pazouki conclude that the genetic factors have been involved in pine et al., 2016) and diterpenoid contents per dry mass (Mantel test 84 A. Kannaste€ et al. / Phytochemistry 147 (2018) 80e88

chemotype diversification (Table 1, Fig. 5) by either suppressing or upregulating the biosynthesis pathways of diterpenoids (Table 2, Figs. 3e5). In wet sites nitrogen (N) availability is one of the main stress factors (Niinemets and Lukjanova, 2003; Portsmuth et al., 2005; Niinemets, 2010; Kannaste€ et al., 2013). It is well known, that limited N availability slows down the normal growth of plants (Ingestad and Kahr,€ 1985; Kainulainen et al., 2000), but N fertil- ization increases the amount of resin cells and the content of resin acids in the mature needles of P. sylvestris (Kainulainen et al., 1996). When the relationships between foliar diterpenoids (mg g 1) and NM (%) were evaluated (Orthogonal partial least-squares (OPLS) algorithm, Fig. 7), the R2Y- and Q2Y-values (model fitting co- efficients indicating it’s predictive performance) for each bog were above 0.9 (Fig. 7). Additionally, the impact of NM on diterpenoid content was tested with Pearson correlation. In ‘manoyl oxide e isoabienol pines’ from Kakerdaja and Parika, NM correlated nega- tively with the sum of abietadiene and abieta-8(14),13(15)-diene (r ¼0.79, P < .01, n ¼ 11); and positively with isoabienol (r ¼ 0.49, P < .05, n ¼ 13). In previous studies, NM correlated also with foliar mono- and sesquiterpenoids (Kainulainen et al., 2000; Kannaste€ et al., 2013). In addition, Laeva samples had the lowest NM and the lowest content of other diterpenoids (Fig. 2), while the dry forest samples contained the highest amount of isoabienol € (Fig. 2) and NM (Kannaste et al., 2013). Finally, abietadienes, manoyl oxide and isoabienol may play essential role in pine defense reaction (Figs. 2 and 5) and their co- existence may be even synergistic (Fig. 2). For example, isoabienol is an antimicrobial (Koutsaviti et al., 2011) and antifungal agent (Cheng et al., 2012). In plant leaves infected by fungi, dehy- droabietinal, which may originate from abietadiene (Keeling and Bohlmann, 2006), activates a systemic acquired resistance mecha- nism (Chaturvedi et al., 2012; Dempsey and Klessig, 2012). In co- nifers, the quantity of abietic acid, which is biosynthesized from abietadiene (LaFever et al., 1994; Phillips and Croteau, 1999), de- termines the tree ability to overcome the mass-attack of mountain bark beetles (Dendroctonus ponderosae)(Boone et al., 2013). The induction of manoyl oxide may be partly related to the fact that diterpenoid is the precursor of forskolin (Garcia-Granados et al., 2007; Pateraki et al., 2014). The latter metabolite increases the content of intracellular cyclic adenosine monophosphate (cAMP), which in turn is known to be associated with a group of root pathogen induced resistance genes involved in the defense mechanism of Norway spruce (Jøhnk et al., 2005). In higher plants þ cAMP controls K -flow in the mesophyll protoplasts (Gehring, þ 2010). In flooded areas the ability of plants to keep K -flow under control is essential in order to survive (Wang et al., 2013).

2.4. Conclusion

Isoabienol dominated in the needles of P. sylvestris, although the blends of diterpenoids varied strongly across pine chemotypes. Similarly to mono- and sesquiterpenoids (Kannaste€ et al., 2013), variations in diterpenoid profiles were related to foliar NM content. Additionally, distribution of pine chemotypes across the sites re- flected significant environmental plasticity. Taken together, based on earlier studies (Baradat and Yazdani, 1988; Jablonka, 2013;

Fig. 5. Diterpenoid profiles (%, mean ± SE) in the needles of P. sylvestris of ‘abieta-8(14), 13(15)-diene e manoyl oxide type’ (A), ‘abietadiene type’ (B), ‘manoyl oxide e iso- abienol type’ (C) and ‘high isoabienol type’ (D and E). On PCA plot ‘abieta-8(14),13(15)- diene e manoyl oxide pines’ and ‘abietadiene pines’ formed a single cluster called ‘other diterpenoid pines’ (see Table 1 and Fig. 4A and B). The diterpenoid profiles of ‘high isoabienol pines’ (D and E) differed considerably (Monte Carlo permutation test, p < .05). A. Kannaste€ et al. / Phytochemistry 147 (2018) 80e88 85

Table 4 Diterpenoids in P. sylvestris samples.

No Diterpenoid Retention index

RI RIc

1 Abietadiene 2088 a 2102 2 Abieta-8(14),13(15)-diene 2154 a 2160 2152 b 3(Z)-Biformene 1988 b 1942 4 Isoabienol 2124 b 2121 5 Labda-7,14-dien-13-ol 2096 b 2097 6 Labda-7,13,14-triene 1978 b 1983 7 Manool 2057 a 2064 Fig. 6. Relationship between genetic and chemical distances according to nSSR 2070 b markers and diterpenoids (mg g 1). Relationship was significant at P < .05 (*). 8 Manoyl oxide 1998 a 2009 2007 b 9 13-epi-Manoyl oxide 2017 a 2030 2023 b

a Retention indices according to Adams (2001). b Retention indices according to software MassFinder 3 library. c Retention indices calculated by injecting the hydrocarbon standard of C7 to C30 (Sigma- Aldrich, St Louis, MO, USA) to GC-MS.

the adaptation of pines to local environment by developing unique diterpenoid profiles.

3. Experimental

3.1. GC-MS analysis

Diterpenoids were quantified with a b-DEX 120 (0.25 mm i.d. 30 m, film thickness 0.25 mm, Supelco, Bellefonte, PA, USA) chiral column with the following GC oven program: 60 C for 5 min, 5 C min 1 to 80 C for 20 min, 2 C min 1 to 200 C for 1 min, and 10 C min 1 to 220 C for 10 min. The MS operated in electron- impact mode (EI) at 70 eV, and in the mass scan range of m/z 33e400 in total ion currant (TIC) mode. Transfer line temperature € Fig. 7. OPLS model based on foliar diterpenoids (mg g 1) and nitrogen contents (%) of was set at 240 C and ion-source temperature at 150 C(Kannaste P. sylvestris samples of bog Kakerdaja. et al., 2013).

3.2. Identification and quantification of diterpenoids Hornoy et al., 2015; Taft et al., 2015; Pazouki et al., 2016), we sug- gest that the growth environment of pines plays essential role in Due to the high content in pine needles (Fig. S1) isoabienol was

Table 3 Location, tree characteristics and flora at sites of studied populations of P. sylvestris.

Population Altitude Growth conditions GPS Characteristics of trees Flora (m) coordinates height stem diameter m cm

Laeva (L) 44 transition mire/raised N 58 26.102 1.4e63e20.5 tree layer e P. sylvestris, Betula pubescens; bog E 26 16.124 shrub/dwarf-shrub layer characteristic to transitional bog - Salix lapponum, Oxycoccus palustris; shrub/dwarf-shrub layer characteristic to raised bog- Chamaedaphne calyculata and Ledum palustre; grass/moss layer- Phragmites communis, Eriophorum vaginatum, Carex spp and various Sphagnum spp Parika (P) 46 raised bog N 58 30.091 0.8e6.5 3e17 tree layer e P. sylvestris, B. pubescens, Betula E 25 46.823 nana; shrub/dwarf-shrub layer e O. palustris, Empetrum nigrum and Andromeda polifolia; grass/moss layer - Eriophorum spp., Carex spp. and Sphagnum spp. Kakaerdaja 80 raised bog N 59 11.321 0.4e4.5 3.5e20 tree layer - P. sylvestris; shrub/dwarf shrub layer (K) E 25 30.613 - E. nigrum, A. polifolia and L. palustre; grass/ moss layer - Scheuchzeria palustris, Rubus chamaemorus and various Sphagnum spp. Manniku€ 57 dry forest/heath N 59 21.555 8e15 9e40 tree layer - P. sylvestris and Picea abies; shrub (M) E 24 42.605 layer - Sorbus aucuparia and Frangula alnus; grass/moss layer - Calluna vulgaris, Thymus serpyllum, Pleurozium schreberi 86 A. Kannaste€ et al. / Phytochemistry 147 (2018) 80e88 extracted and identified by NMR analysis (see supporting mixed, then left at 6 C for extraction (Kannaste€ et al., 2013). The information). Other diterpenoids (Table 4) were identified by extraction protocol was optimized by extracting the diterpenoids analyzing the needle extracts and a C7 -C30 hydrocarbon standard for 1, 3, 5, 24 and 48 h (Fig. S2). The analysis showed, that the (Sigma Aldrich. St. Louis, MO, USA) with GC-MS using a ZB5-MS content and composition of diterpenoids did not alter during the column (0.25 mm i.d. 30 m, 0.25 mm film Zebron, Phenomenex, sub-sample removal and rapid determination of needle mass for Torrance, CA, USA). The GC oven program was: 40 C for 1 min, 5 C analyses (Fig. S4). Hence, diterpenoids were extracted by following min 1 to 220 C for 5 min. The use of the hydrocarbon standard an extraction protocol of 3 h. enabled the calculation of retention indices (RIs), which were After the extraction procedure supernatant was immediately compared to published data (Adams, 2001; NIST library and injected into Shimadzu QP2010 Plus (GC-MS) using an auto- MassFinder 3 database) in addition to the mass spectrums of injector AOC-20i (Shimadzu Corporation, Kyoto, Japan). Samples diterpenoids (Fig. S2). were analyzed with a split ratio of 5:1. Repeated analyses of Isoabienol was used as a diterpenoid standard for calibrating the different replicates taken randomly from the same samples GC-MS. Calibration samples were prepared by diluting known demonstrated that the variation among diterpenoid concentrations quantities of isoabienol in hexane, then analyzing 1 ml of each caused by within-sample variability and slight inaccuracies during sample to create a calibration factor that was used for quantifying the extraction (mass determination, imperfect extraction, potential the peak areas of diterpenoids in pine samples. Contents of diter- loss of sample during the extraction) was less than 10%. penoids were calculated according to the linear regression (y ¼ 18,897,009 x) by dividing the peak areas of diterpenoids with 3.6. Molecular analysis the calibration factor and dry weight of pine needles. Detection limit of diterpenoids was 10 pmol/ul and the linear range of the MS Details of nSSR analysis are described in a previous study detector was up to 100 mg. (Pazouki et al., 2016).

3.3. Study sites 3.7. Foliar N analysis

The study was conducted in four naturally-regenerated The needle N content per dry mass was estimated with Vario P. sylvestris dominated sites in Estonia. Pine samples from the MAX CNS analyzer (Elementar Analysensysteme GmbH, Hanau, bogs of Parika, Kakerdaja, and Laeva were compared to the ones Germany). from the dry forest of Manniku€ (Table 3). Location of populations, as well as tree characteristics and main floral species at tree-, shrub- 3.8. Data handling and statistical analysis and grass layer are shown in Table 3. For a detailed overview about the sites together with soil pH-values is presented in our previous The composition of diterpenoids in pine samples was evaluated study (Kannaste€ et al., 2013). by principal component analysis (PCA) (Wold et al., 1987), where the diterpenoid dataset was mean-centered and log-transformed. 3.4. Foliage collection for chemical analyses The differences in diterpenoid profiles were tested by Monte- Carlo permutation test and redundancy data analysis (RDA). The sampling in all sites was conducted during June 2e25, 2010. Multivariate analyses were conducted by Canoco 5.0 software (ter Sampled trees were chosen randomly. Samples were collected not Braak and Smilauer, Biometris Plant Research International, The further than 50 m from the trajectory (Fig. S3). In the bog sites, Netherlands). All statistical tests were considered significant at sampling started from the drier and less acidic edges, and pro- P < .05. gressed towards the wetter and more acidic bog interior in equi- Mantel test with 1000 random permutations was carried out for distal intervals. In the bogs of Kakerdaja and Parika sampling studying the correlation between the matrix of pairwise genetic stopped near bog lakes (Fig. S3). In the Laeva bog, the transition differentiation and a matrix of diterpenoid distance using GenAlEx mire gradually transformed into raised bog between samples 45 v6 (Peakall and Smouse, 2006). and 60 (Fig. S3). In the dry forest, the samples were taken randomly. In Pearson correlation analyses diterpenoid contents per dry Terminal branches were taken from open-growing trees on the weight were used to gain insights on the relationships among southern side of the crown. One-year-old foliage from multiple diterpenoids (Faldt€ et al., 2003; Semiz et al., 2007; Kannaste€ et al., shoots was collected. Finally, one pine sample analyzed by GCMS 2013). These analyses were conducted with R (R Foundation for corresponded to one pine tree. 32 Samples from dry forest of Statistical Computing, Vienna, Austria). Additionally, orthogonal Manniku€ to 70 samples from bog of Kakerdaja were collected from partial least-squares (OPLS) analysis (Trygg and Wold, 2002)was each site. The plant material was placed in sealed plastic bags and used to analyse the relationship between diterpenoids and NM. kept in an ice-box until transported to the laboratory (within 2 h OPLS allows to model separately the variations of the predictors after collecting). The samples for terpene analyses were kept at e correlated and orthogonal to the response (Thevenot et al., 2015). 80 C until analyzed. OPLS analysis was carried out using ropls R package, function ‘opls’ (Thevenot et al., 2015). 3.5. Sample homogenization for diterpenoid analyses Conflicts of interest For each chemical analyses, a sample of 140 ± 20 mg of fresh plant material from randomly chosen needles was weighted into a The authors declare that they have no competing interests. 2 ml tube with 2.8 mm stainless steel beads (Bertin Technologies, Aix-en-Provence, France). Acknowledgements The tubes filled with liquid nitrogen and plant material were immediately homogenized with Precellys 24 homogenizer (Bertin This study was funded by the gs1:European Social Fund (post- Technologies, Aix-en-Provence France) at 6500 rotation per minute doctoral grant MJD 14), the Estonian Ministry of Science and Edu- for three 60 s periods at 20 C. After homogenization, 1 ml of GC- cation (IUT 8-3, PUT 1409), European Union through the European grade hexane (Sigma Aldrich. St. Louis, MO, USA) was added and Regional Fund (The Centre of Excellence in Environmental A. Kannaste€ et al. / Phytochemistry 147 (2018) 80e88 87

Adaptation TK107 and The Centre of Excellence, EcolChange, 91e103. € TK131) and Mobilitas Program Top Researcher Grant No. MTT2 Ingestad, T., Kahr, M., 1985. Nutrition and growth of coniferous seedlings at varied relative nitrogen addition rate. Physiol. Plantarum 65, 109e116. “Chemical Ecology” (to LT and AKBK) and the European Research Jablonka, E., 2013. Epigenetic inheritance and plasticity: the responsive germline. Council (advanced grant 322603, SIP-VOLþ). Finally, we would like Prog. Biophys. Mol. Biol. 111, 99e107. to acknowledge prof. Thomas D. Sharkey for linguistic corrections. Jasinski, M., Stukkens, Y., Degand, H., Purnelle, B., Marchand-Brynaert, J., Boutry, M., 2001. A plant plasma membrane ATP binding cassette-type transporter is involved in antifungal terpenoid secretion. Plant Cell 13, 1095e1107. Appendix A. Supplementary data Jøhnk, N., Hietala, A.M., Fossdal, C.G., Collinge, D.B., Newman, M.A., 2005. Defense- related genes expressed in Norway spruce roots after infection with the root rot pathogen Ceratobasidium bicorne (anamorph : Rhizoctonia sp.). Tree Physiol. 25, Supplementary data related to this article can be found at 1533e1543. https://doi.org/10.1016/j.phytochem.2017.12.007. Kainulainen, P., Holopainen, J.K., Oksanen, J., 1995. Effects of gaseous air pollutants on secondary chemistry of Scots pine and Norway spruce seedlings. Water Air Soil Pollut. 85, 1393e1398. References Kainulainen, P., Holopainen, J., Palomaki, V., Holopainen, T., 1996. Effects of nitrogen fertilization on secondary chemistry and ectomycorrhizal state of Scots pine Adams, R.P., 2001. Identification of Essential Oil Components by Gas Chromatog- seedlings and on growth of grey pine aphid. J. Chem. Ecol. 22, 617e636. raphy/Quadrupole Mass Spectroscopy. Allured Publishing Corporation, Carol Kainulainen, P., Satka, H., Mustaniemi, A., Holopainen, J.K., Oksanen, J., 1993. Conifer Stream, USA. aphids in an air-polluted environment. II. Host plant quality. Environ. Pollut. 80, Adams, R.P., Beauchamp, P.S., Dev, V., Bathala, R.M., 2010. The leaf essential oils of 193e200. Juniperus communis L. varieties in North America and the NMR and MS data for Kainulainen, P., Utriainen, J., Holopainen, J.K., Oksanen, J., Holopainen, T., 2000. isoabienol. J. Essent. Oil Res. 22, 23e28. Influence of elevated ozone and limited nitrogen availability on conifer seed- Bailey, J.A., Vincent, G.G., Burden, R.S., 1974. Diterpenes from Nicotiana glutinosa and lings in an open-air fumigation system: effects on growth, nutrient content, their effect on fungal growth. J. Gen. Microbiol. 85, 57e64. mycorrhiza, needle ultrastructure, starch and secondary compounds. Global Baradat, P., Yazdani, R., 1988. Genetic expression for monoterpenes in clones of Change Biol. 6, 345e355. Pinus sylvestris grown on different sites. Scand. J. For. Res. 3, 25e36. Keeling, C.I., Bohlmann, J., 2006. Diterpene resin acids in conifers. Phytochemistry Bohlmann, J., Keeling, C.I., 2008. Terpenoid biomaterials. Plant J. 54, 656e669. 67, 2415e2423. Boone, C.K., Keefover-Ring, K., Mapes, A.C., Adams, A.S., Bohlmann, J., Raffa, K.F., Keeling, C.I., Dullat, H.K., Yuen, M., Ralph, S.G., Jancsik, S., Bohlmann, J., 2010. 2013. Bacteria associated with a tree-killing insect reduce concentrations of Identification and functional characterization of monofunctional ent-copalyl plant defense compounds. J. Chem. Ecol. 39, 1003e1006. diphosphate and ent-kaurene synthases in white spruce reveal different pat- Bussotti, F., Pollastrini, M., Holland, V., Brüggemannb, W., 2015. Functional traits and terns for diterpene synthase evolution for primary and secondary metabolism adaptive capacity of European forests to climate change. Environ. Exp. Bot. 111, in gymnosperms. Plant Physiol. 152, 1197e1208. 91e113. Kopper, B.J., Illman, B.L., Kersten, P.J., Klepzig, K.D., Raffa, K.F., 2005. Effects of Back,€ J., Aalto, J., Henriksson, M., Hakola, H., He, Q., Boy, M., 2012. Chemodiversity of diterpene acids on components of a conifer bark beetle-fungal interaction: a Scots pine stand and implications for terpene air concentrations. Bio- tolerance by Ips pini and sensitivity by its associate Ophiostoma ips. Environ. geosciences 9, 689e702. Entomol. 34, 486e493. Chaturvedi, R., Venables, B., Petros, R.A., Nalam, V., Li, M., Wang, X., Takemoto, L.J., Koutsaviti, A., Milenkovic, M., Tzakou, O., 2011. Antimicrobial activity of the Shah, J., 2012. An abietane diterpenoid is a potent activator of systemic acquired essential oil of Greek endemic Stachys spruneri and its main component, iso- resistance. Plant J. 71, 161e172. abienol. Nat. Prod. Commun. 6, 277e280. Chen, F., Tholl, D., Bohlmann, J., Pichersky, E., 2011. The family of terpene synthases Kukovinets, O.S., Zainullin, R.A., Odinokov, V.N., Kislitsyn, M.I., Roshchin, V.I., in plants: a mid-size family of genes for specialized metabolism that is highly Galin, F.Z., Tolstikov, G.A., 2001. Ozonolysis of alkenes and study of reactions of diversified throughout the kingdom. Plant J. 66, 212e229. polyfunctional compounds: LXIV. Synthesis of ambreinolide and 8a,13-Epoxy- Cheng, S.-S., Chung, M.-J., Lin, C.-Y., Wang, Y.-N., Chang, S.-T., 2012. Phytochemicals 14,15,16-trisnorlabd-12-ene proceeding from isoabienol ozonolysis. Russ. J. Org. from Cunninghamia konishii Hayata act as antifungal agents. J. Agric. Food Chem. Chem. 37 (2), 235e237. € 60, 124e128. Kannaste, A., Copolovici, L., Pazouki, L., Suhhorutsenko, M., Niinemets, Ü., 2013. Croteau, R., Gurkewitz, S., Johnson, M.A., Fisk, H.J., 1987. Biochemistry of oleoresi- Highly variable chemical signatures over short spatial distances among Scots nosis: monoterpene and diterpene biosynthesis in Lodgepole Pine saplings pine (Pinus sylvestris) populations. Tree Physiol. 33, 374e387. infected with Ceratocystis clavigera or treated with carbohydrate elicitors. Plant LaFever, R.E., Vogel, B.S., Croteau, R., 1994. Diterpenoid resin acid biosynthesis in Physiol. 85, 1123e1128. conifers: enzymatic cyclization of geranylgeranyl pyrophosphate to abietadiene, Dempsey, D.A., Klessig, D.F., 2012. SOS - too many signals for systemic acquired the precursor of abietic acid. Arch. Biochem. Biophys. 313, 139e149. resistance? Trends Plant Sci. 17, 538e545. Lee, H.-J., Ravn, M.M., Coates, R.M., 2001. Synthesis and characterization of abie- Dering, M., Kosinski, P., Wyka, T.P., PerseKamczyc, E., Boratynski, A., Boratynska, K., tadiene, levopimaradiene, palustradiene, and neoabietadiene: hydrocarbon _ Reich, P.B., Romo, A., Zadworny, M., Zytkowiak, R., Oleksyn, J., 2017. Tertiary precursors of the abietane diterpene resin acids. Tetrahedron 57, 6155e6167. remnants and Holocene colonizers: genetic structure and phylogeography of Lewinsohn, E., Gijzen, M., Croteau, R., 1991a. Defense mechanisms of conifers: dif- Scots pine reveal higher genetic diversity in young boreal than in relict Medi- ferences in constitutive and wound-induced monoterpene biosynthesis among terranean populations and a dual colonization of Fennoscandia. Divers. Distrib. species. Plant Physiol. 96, 44e49. 23, 540e555. Lewinsohn, E., Gijzen, M., Savage, T.J., Croteau, R., 1991b. Defense mechanisms of Dudareva, N., Pichersky, E., Gershenzon, J., 2004. Biochemistry of plant volatiles. conifers: relationship of monoterpene cyclase activity to anatomical speciali- Plant Physiol. 135, 1893e1902. zation and oleoresin monoterpene content. Plant Physiol. 96, 38e43. Falara, V., Pichersky, E., Kanellis, A.K., 2010. A Copal-8-ol diphosphate synthase from Loreto, F., Nascetti, P., Graverini, A., Mannozzi, M., 2000. Emission and content of the angiosperm Cistus creticus subsp creticus is a putative key enzyme for the monoterpenes in intact and wounded needles of the Mediterranean pine, Pinus formation of pharmacologically active, oxygen-containing labdane-type diter- pinea. Funct. Ecol. 14, 589e595. penes. Plant Physiol. 154, 301e310. Manninen, A.M., Tarhanen, S., Vuorinen, M., Kainulainen, P., 2002. Comparing the Faldt,€ J., Martin, D., Miller, B., Rawat, S., Bohlmann, J., 2003. Traumatic resin defense variation of needle and wood terpenoids in Scots pine provenances. J. Chem. in Norway spruce (Picea abies): methyl jasmonate-induced terpene synthase Ecol. 28, 211e228. gene expression, and cDNA cloning and functional characterization of (þ)-3- Memari, H.R., Pazouki, L., Niinemets, Ü., 2013. The biochemistry and molecular carene synthase. Plant Mol. Biol. 51, 119e133. biology of volatile messingers in trees. In: Niinemets, Ü., Monson, R.K. (Eds.), Garcia-Granados, A., Martinez, A., Parra, A., Rivas, F., 2007. Manoyl-oxide bio- Biology, Controls and Models of Tree Volatile Organic Compound Emissions. transformations with filamentous fungi. Curr. Org. Chem. 11, 679e692. Springer, Printforce, the Netherlands, p. 547. Gehring, C., 2010. Adenyl cyclases and cAMP in plant signaling - past and present. Miller, B., Madilao, L.L., Ralph, S., Bohlmann, J., 2005. Insect-induced conifer defense. Cell Commun. Signal. 8, 1e5. White pine weevil and methyl jasmonate induce traumatic resinosis, de novo Gref, R., 1981. Variation in isoabienol content in Pinus sylvestris needles. Can. J. Bot. formed volatile emissions, and accumulation of terpenoid synthase and puta- 59, 831e835. tive octadecanoid pathway transcripts in Sitka spruce. Plant Physiol. 137, Hiltunen, R., Tigerstedt, P.M.A., Juvonen, S., Pohjola, J., 1975. Inheritence of 3-carene 369e382. quantity in Pinus sylvestris L. Farmaseuttinen Aikakauslehti 84, 69e72. Niinemets, Ü., 2010. Responses of forest trees to single and multiple environmental Holliday, J.A., Ritland, K., Aitken, S.N., 2010. Widespread, ecologically relevant ge- stresses from seedlings to mature plants: past stress history, stress interactions, netic markers developed from association mapping of climate-related traits in tolerance and acclimation. For. Ecol. Manag. 260, 1623e1639. Sitka spruce (Picea sitchensis). New Phytol. 188, 501e514. Niinemets, Ü., Lukjanova, A., 2003. Needle longevity, shoot growth and branching Hornoy, B., Pavy, N., Gerardi, S., Beaulieu, J., Bousquet, J., 2015. Genetic adaptation to frequency in relation to site fertility and within-canopy light conditions in Pinus climate in white spruce involves small to moderate allele frequency shifts in sylvestris. Ann. For. Sci. 60, 195e208. functionally diverse genes. Genome Biol. Evol. 7, 3269e3285. Orlova, I., Nagegowda, D.A., Kish, C.M., Gutensohn, M., Maeda, H., Varbanova, M., Ignea, C., Ioannou, E., Georgantea, P., Loupassaki, S., Trikka, F.A., Kanellis, A.K., Fridman, E., Yamaguchi, S., Hanada, A., Kamiya, Y., Krichevsky, A., Citovsky, V., Makris, A.M., Roussis, V., Kampranis, C.S., 2015. Reconstructing the chemical Pichersky, E., Dudareva, N., 2009. The small subunit of snapdragon geranyl diversity of labdane-type diterpene biosynthesis in yeast. Metab. Eng. 28, diphosphate synthase modifies the chain length specificity of tobacco 88 A. Kannaste€ et al. / Phytochemistry 147 (2018) 80e88

geranylgeranyl diphosphate synthase in planta. Plant Cell 21, 4002e4017. 2012. Characterization of two genes for the biosynthesis of the labdane diter- Ormeno,~ E., Goldstein, A., Niinemets, Ü., 2011. Extracting and trapping biogenic pene Z-abienol in tobacco (Nicotiana tabacum) glandular trichomes. Plant J. 72, volatile organic compounds stored in plant species. Trends Anal. Chem. 30, 1e17. 978e989. Semiz, G., Heijari, J., Isik, K., Holopainen, J.K., 2007. Variation in needle terpenoids Pateraki, I., Andersen-Ranberg, J., Hamberger, B., Heskes, A.M., Martens, H.J., among Pinus sylvestris L. (Pinaceae) provenances from Turkey. Biochem. Sys- Zerbe, P., Bach, S.S., Moller, B.L., Bohlmann, J., 2014. Manoyl oxide (13R), the temat. Ecol. 35, 652e661. biosynthetic precursor of forskolin, is synthesized in specialized root cork cells Seo, S., Seto, H., Koshino, H., Yoshida, S., Ohashi, Y., 2003. A diterpene as an in Coleus forskohlii. Plant Physiol. 164, 1222e1236. endogenous signal for the activation of defense responses to infection with Pazouki, L., Shanjani, P.S., Fields, P.D., Martins, K., Suhhorutsenko, M., Viinalass, H., tobacco mosaic virus and wounding in tobacco. Plant Cell 15, 863e873. Niinemets, Ü., 2016. Large within-population genetic diversity of the wide- Sjodin,€ K., Persson, M., Borg-Karlson, A.-K., Norin, T., 1996. Enantiomeric composi- spread conifer Pinus sylvestris at its soil fertility limit characterized by nuclear tions of monoterpene hydrocarbons in different tissues of four individuals of and chloroplast microsatellite markers. Eur. J. For. Res. 135, 161e177. Pinus sylvestris. Phytochemistry 41, 439e445. Pazouki, L., Niinemets, Ü., 2016. Multi-substrate terpenoid synthases: their occur- Sjodin,€ K., Persson, M., Faldt,€ J., Ekberg, I., Borg-Karlson, A.-K., 2000. Occurence and rence and physiological significance. Front. Plant Sci. 7, 1e16. correlations of monoterpene hydrocarbon enantiomers in Pinus sylvestris and Peakall, R., Smouse, P.E., 2006. GENALEX V. 6: genetic analysis in excel. Population Picea abies. J. Chem. Ecol. 26, 1701e1720. genetic software for teaching and research. Mol. Ecol. Notes 6, 288e295. Taft, S., Najar, A., Godbout, J., Bousquet, J., Erbilgin, N., 2015. Variations in foliar Phillips, M.A., Croteau, R.B., 1999. Resin-based defenses in conifers. Trends Plant Sci. monoterpenes across the range of jack pine reveal three widespread chemo- 4, 184e190. types: implications to host expansion of invasive mountain pine beetle. Front. Portsmuth, A., Niinemets, Ü., Truus, L., Pensa, M., 2005. Biomass allocation and Plant Sci. 6. growth rates in Pinus sylvestris are interactively modified by nitrogen and Thevenot, E.A., Roux, A., Xu, Y., Ezan, E., Junot, C., 2015. Analysis of the human adult phosphorus availabilities and by tree size and age. Can. J. For. Res. 35, urinary metabolome variations with age, body mass index, and gender by 2346e2359. implementing a comprehensive workflow for univariate and OPLS statistical Rajabi Memari, H., Pazouki, L., Niinemets, Ü., 2013. The biochemistry and molecular analyses. J. Proteome Res. 14 (8), 3322e3335. biology of volatile messengers in trees. In: Niinemets, Ü., Monson, R.K. (Eds.), Tholl, D., 2006. Terpene synthases and the regulation, diversity and biological roles Biology, Controls and Models of Tree Volatile Organic Compound Emissions. of terpene metabolism. Curr. Opin. Plant Biol. 9, 297e304. Springer, Berlin, pp. 47e93. Trygg, J., Wold, S., 2002. Orthogonal projections to latent structures (O-PLS). Robert, J.A., Madilao, L.L., White, R., Yanchuk, A., King, J., Bohlmann, J., 2010. J. Chemometr. 16 (3), 119e128. Terpenoid metabolite profiling in Sitka spruce identifies association of dehy- Ulubelen, A., Topcu, G., Eris¸ , C., Sonmez,€ U., Kartal, M., Kurucu, S., Bozok- droabietic acid, (þ)-3-carene, and terpinolene with resistance against white Johansson, C., 1994. Terpenoids from Salvia sclarea. Phytochemistry 36, pine weevil. Bot. Botanique 88, 810e820. 971e974. Rosenkranz, M., Schnitzler, J.-P., 2013. Genetic engineering of BVOC emissions from Wang, M., Zheng, Q., Shen, Q., Guo, S., 2013. The critical role of potassium in plant trees. In: Niinemets, Ü., Monson, R.K. (Eds.), Biology, Controls and Models of stress response. Int. J. Mol. Sci. 14, 7370e7390. Tree Volatile Organic Compound Emissions. Springer, Berlin, pp. 95e118. Wold, S., Esbensen, K., Geladi, P., 1987. Principal component analysis. Chemometr. Sallaud, C., Giacalone, C., Toepfer, R., Goepfert, S., Bakaher, N., Roesti, S., Tissier, A., Intell. Lab. Syst. 2, 37e52.