*Article* **Inhibitory Effect of Coumarins and Isocoumarins Isolated from the Stems and Branches of** *Acer mono* **Maxim. against** *Escherichia coli β***-Glucuronidase**

**Nguyen Viet Phong 1,2 , Byung Sun Min <sup>3</sup> , Seo Young Yang 4,\* and Jeong Ah Kim 1,2,\***


**Abstract:** We isolated eight known secondary metabolites, including two isocoumarins and six coumarins, from the stems and branches of *Acer mono* Maxim. Their structures were confirmed using nuclear magnetic resonance spectroscopy and by comparing the data to published reports. The inhibitory effects of all compounds (**1**−**8**) on *Escherichia coli β*-glucuronidase were evaluated for the first time using in vitro assays. 3-(3,4-Dihydroxyphenyl)-8-hydroxyisocoumarin (**1**) displayed an inhibitory effect against *β*-glucuronidase (IC<sup>50</sup> = 58.83 ± 1.36 µM). According to the findings of kinetic studies, compound **1** could function as a non-competitive inhibitor. Molecular docking indicated that compound **1** binds to the allosteric binding site of *β*-glucuronidase, and the results corroborated those from kinetic studies. Furthermore, molecular dynamics simulations of compound **1** were performed to identify the behavioral and dynamic properties of the protein–ligand complex. Our results reveal that compound **1** could be a lead metabolite for designing new *β*-glucuronidase inhibitors.

**Keywords:** *Acer mono* Maxim.; isocoumarins; *β*-glucuronidase; non-competitive inhibitor; allosteric binding site; molecular dynamics

## **1. Introduction**

*β*-Glucuronidase is a well-known enzyme that hydrolyzes conjugated compounds, including *β*-glucuronic acid, into their derivatives and free glucuronic acid [1]. *β*-Glucuronidase is frequently found in unicellular microorganisms, such as *Escherichia coli* and *Peptostreptococcus* species, and multicellular organisms, including plants and mammals [2,3]. In particular, this enzyme can be detected in several human body organs, such as the kidney, liver, lung, and digestive system [4]. In 2010, Wallace et al. first reported the crystal structure of *E. coli β*-glucuronidase [5]. The *β*-glucuronidase asymmetric unit (139 kDa) comprises two monomers (597 residues) and is organized into the following three areas: the N-terminal contains 180 residues and is similar to the carbohydrate-binding domain of the glycoside hydrolase 2 members; the C-terminal contains residue 274 to residue 603 and comprises an αβ loop; and the region between terminals N and C contains an Ig-like *β*-sandwich domain, as is the case with other members of the glycoside hydrolase 2 family [5,6]. Inhibiting *β*-glucuronidase can be beneficial for preventing and treating various diseases [7]. *β*-Glucuronidase is generated in the synovial fluid under inflammatory conditions, such as rheumatoid arthritis [8]. Moreover, an increased risk of colon cancer has been associated with the role of *β*-glucuronidase in the disease, as well as enhanced intestinal

**Citation:** Phong, N.V.; Min, B.S.; Yang, S.Y.; Kim, J.A. Inhibitory Effect of Coumarins and Isocoumarins Isolated from the Stems and Branches of *Acer mono* Maxim. against *Escherichia coli β*-Glucuronidase. *Appl. Sci.* **2022**, *12*, 10685. https://doi.org/ 10.3390/app122010685

Academic Editors: Luca Mazzoni, Maria Teresa Ariza Fernández and Franco Capocasa

Received: 4 October 2022 Accepted: 19 October 2022 Published: 21 October 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

enzyme levels [9]. Increased levels of *β*-glucuronidase in the blood, owing to liver injury, can lead to liver cancer [7,10]. Thus, the discovery and development of *β*-glucuronidase inhibitors may be valuable in reducing these carcinogenic risks.

*Acer* L., commonly known as maple, is a diverse genus in the Aceraceae family, which comprises over 125 species with several infraspecific taxa, and is widely distributed in northern temperate regions worldwide [11]. Many *Acer* species provide products of economic value, including furniture, lumber, horticultural plants, and herbal medicines, especially gamma-linolenic acid, a dietary supplement that can help treat cancer and cardiovascular diseases [12]. *A. mono* Maxim., a deciduous tree of the *Acer* genus, is commonly found in Korea, Japan, and Northeast China [13,14]. Over the years, this plant has attracted considerable attention from both chemists and pharmacists, given its noteworthy traditional uses and pharmacological activities. In Korea, the leaves of *A. mono* Maxim. have long been employed as a material for hemostasis, whereas the roots have been used in traditional Korean medicine to treat arthralgia and cataclasis [13–15]. The sap of *A. mono* Maxim. has been used to treat gout, neuralgia, urinary hesitancy, constipation, and other gastroenteric conditions [15,16]. *A. mono* Maxim. reportedly possesses various pharmacological properties, including hepatoprotective, antioxidant, and osteoporotic inhibitory effects [13,14]. Phytochemical constituents of *A. mono* Maxim. include stilbenes, flavonoids, and their derivatives [14]. Among them, 5-*O*-methyl-(*E*)-resveratrol 3-*O*-*β*-D-glucopyranoside and 5-*O*-methyl-(*E*)-resveratrol 3-*O*-*β*-D-apiofuranosyl-(1→6)-*β*-D-glucopyranoside, two stilbene glycosides isolated from the leaves of *A. mono* Maxim., could reduce the levels of DPPH radicals at the concentrations of 100 µM, thereby exhibiting significant free-radical scavenging effects, with IC<sup>50</sup> values of 103.6 and 80.5 µM, respectively [13].

As part of our ongoing investigation of the secondary metabolites and respective biological effects from herbal and medicinal plants in Korea, we isolated and structurally elucidated eight compounds, including two isocoumarins and six coumarins, from the stems and branches of *A. mono* Maxim. in the present study. To the best of our knowledge, this is the first report regarding the isolation of isocoumarins from the *Acer* genus. In vitro assays were performed to determine the *β*-glucuronidase inhibitory activity of all compounds isolated from *A. mono* Maxim. In addition, kinetic analysis studies, molecular docking, and molecular dynamics (MD) simulations were performed to comprehensively clarify the inhibition mode, critical amino acid interactions, and the protein–ligand binding mechanism of active compound **1** with *β*-glucuronidase proteins.

#### **2. Results**

#### *2.1. Isolated Compounds from the Stems and Branches of A. mono Maxim.*

Methanol residue from the stems and branches of *A. mono* Maxim. was suspended in distilled water and partitioned with *n*-hexane, CH2Cl2, and EtOAc to obtain four extracts. The CH2Cl<sup>2</sup> extract (77.8 g) and water layer (170 g) were separated by repeated column chromatography (CC) on silica gel, RP-18 gel, and Diaion resins, followed by preparative high-performance liquid chromatography (HPLC) to isolate and purify two isocoumarins (**1** and **2**) and six coumarins (**3**−**8**) (Figure 1). Based on an in-depth analysis of nuclear magnetic resonance (NMR) results and comparison with corresponding data from previous reports, the structures of all the compounds isolated were elucidated as 3-(3,4-dihydroxyphenyl)-8-hydroxyisocoumarin (**1**) [17], hydrangenol (**2**) [18], scopoletin (**3**) [19], isoscopoletin (**4**) [19], isofraxidin (**5**) [20], fraxidin 8-*O*-*β*-D-glucopyranoside (**6**) [21], aquillochin (**7**) [22], and cleomiscosin D (**8**) [23].

**Figure 1.** Chemical structures of isolated compounds from *Acer mono* Maxim. (**1**−**8**). **Figure 1.** Chemical structures of isolated compounds from *Acer mono* Maxim. (**1**−**8**). **Figure 1.** Chemical structures of isolated compounds from *Acer mono* Maxim. (**1**−**8**).

#### *2.2. Inhibitory Activity of Isolated Compounds against β-Glucuronidase 2.2. Inhibitory Activity of Isolated Compounds against β-Glucuronidase 2.2. Inhibitory Activity of Isolated Compounds against β-Glucuronidase*

All isolated compounds (**1**−**8**) were examined for their ability to inhibit *β*-glucuronidase using D-saccharic acid 1,4-lactone (DSA), a well-known inhibitor of *β*-glucuronidase, as a positive control [24]. The results are displayed in Figure 2A and Table 1, which revealed that 3-(3,4-dihydroxyphenyl)-8-hydroxyisocoumarin (**1**) exhibited *β*-glucuronidase inhibitory activity, with 69.85% of inhibition at 100 μM and an IC<sup>50</sup> value of 58.83 ± 1.36 μM. Compound **2** with the absence of the C-3−C-4 double bond in the lactone ring did not exhibit *β*-glucuronidase inhibitory activity (15.89% of inhibition at 100 μM). Coumarin and its derivatives (**3**−**8**) failed to exhibit inhibitory activity against *β*-glucuronidase (IC<sup>50</sup> >100 µM). These results suggested that the different positions of an oxygen atom and carbonyl group in the isocoumarin structure, compared with those of coumarin, could play an important role in *β*-glucuronidase inhibition. All isolated compounds (**1**−**8**) were examined for their ability to inhibit *β*-glucuronidase using D-saccharic acid 1,4-lactone (DSA), a well-known inhibitor of *β*-glucuronidase, as a positive control [24]. The results are displayed in Figure 2A and Table 1, which revealed that 3-(3,4-dihydroxyphenyl)-8-hydroxyisocoumarin (**1**) exhibited *β*-glucuronidase inhibitory activity, with 69.85% of inhibition at 100 µM and an IC<sup>50</sup> value of 58.83 ± 1.36 µM. Compound **2** with the absence of the C-3−C-4 double bond in the lactone ring did not exhibit *β*glucuronidase inhibitory activity (15.89% of inhibition at 100 µM). Coumarin and its derivatives (**3**−**8**) failed to exhibit inhibitory activity against *β*-glucuronidase (IC<sup>50</sup> >100 µM). These results suggested that the different positions of an oxygen atom and carbonyl group in the isocoumarin structure, compared with those of coumarin, could play an important role in *β*-glucuronidase inhibition. All isolated compounds (**1**−**8**) were examined for their ability to inhibit *β*-glucuronidase using D-saccharic acid 1,4-lactone (DSA), a well-known inhibitor of *β*-glucuronidase, as a positive control [24]. The results are displayed in Figure 2A and Table 1, which revealed that 3-(3,4-dihydroxyphenyl)-8-hydroxyisocoumarin (**1**) exhibited *β*-glucuronidase inhibitory activity, with 69.85% of inhibition at 100 μM and an IC<sup>50</sup> value of 58.83 ± 1.36 μM. Compound **2** with the absence of the C-3−C-4 double bond in the lactone ring did not exhibit *β*-glucuronidase inhibitory activity (15.89% of inhibition at 100 μM). Coumarin and its derivatives (**3**−**8**) failed to exhibit inhibitory activity against *β*-glucuronidase (IC<sup>50</sup> >100 µM). These results suggested that the different positions of an oxygen atom and carbonyl group in the isocoumarin structure, compared with those of coumarin, could play an important role in *β*-glucuronidase inhibition.

**Figure 2.** Inhibitory activity of isolated compounds **1**−**8** at 100 µM (**A**) and determination of IC<sup>50</sup> value by compound **1** on *β*-glucuronidase (**B**). **Figure 2.** Inhibitory activity of isolated compounds **1**−**8** at 100 µM (**A**) and determination of IC<sup>50</sup> value by compound **1** on *β*-glucuronidase (**B**). **Figure 2.** Inhibitory activity of isolated compounds **1**−**8** at 100 µM (**A**) and determination of IC<sup>50</sup> value by compound **1** on *β*-glucuronidase (**B**).


**Table 1.** Inhibition of compounds **1**–**8** against *β*-glucuronidase. **Table 1.** Inhibition of compounds **1**–**8** against *β*-glucuronidase. **Compounds Inhibition against** *β***-Glucuronidase**

*Appl. Sci.* **2022**, *12*, x FOR PEER REVIEW 4 of 13

<sup>1</sup> The values (µM) represent 50% inhibition of *<sup>β</sup>*-glucuronidase. Results are presented as the mean <sup>±</sup> standard error of triplicate experiments. <sup>2</sup> Positive control. <sup>1</sup> The values (μM) represent 50% inhibition of *β*-glucuronidase. Results are presented as the mean ± standard error of triplicate experiments. <sup>2</sup> Positive control.

#### *2.3. Enzyme Kinetics of Compound* **1** *against β-Glucuronidase 2.3. Enzyme Kinetics of Compound 1 against β-Glucuronidase*

To determine the type of inhibition and the inhibitory constant *K<sup>i</sup>* of active compound **1**, enzyme kinetics were conducted at different concentrations of the substrate 4-nitrophenyl *β*-D-glucuronide (PNPG) and inhibitor [7]. In the Lineweaver−Burk plot, the x-axis is the reciprocal of the substrate concentration, or 1/(S), and the y-axis is the reciprocal of the reaction velocity (1/V). The non-competitive or competitive inhibition mode is indicated by the family of straight lines that intersect at the same point on the 1/(S) or 1/V axis, respectively, whereas mixed inhibition is represented by the straight lines of the inhibitor that intersect at the xy region. As depicted in Figure 3A, the Lineweaver−Burk plot revealed intersecting lines on the 1/(S) axis, indicating that compound **1** inhibited *β*-glucuronidase in a non-competitive inhibition mode. In addition, a Dixon plot was used to determine the *K<sup>i</sup>* value between the inhibitor and the enzyme, where the intersection value on the x-axis implies *K<sup>i</sup>* . The *K<sup>i</sup>* is the concentration of an inhibitor needed to decrease the maximum rate of the reaction by 50% [25]. As presented in Figure 3B, the *K<sup>i</sup>* value of **1** was calculated to be 40.9 µM. To determine the type of inhibition and the inhibitory constant *K<sup>i</sup>* of active compound **1**, enzyme kinetics were conducted at different concentrations of the substrate 4-nitrophenyl *β*-D-glucuronide (PNPG) and inhibitor [7]. In the Lineweaver−Burk plot, the x-axis is the reciprocal of the substrate concentration, or 1/(S), and the y-axis is the reciprocal of the reaction velocity (1/V). The non-competitive or competitive inhibition mode is indicated by the family of straight lines that intersect at the same point on the 1/(S) or 1/V axis, respectively, whereas mixed inhibition is represented by the straight lines of the inhibitor that intersect at the xy region. As depicted in Figure 3A, the Lineweaver−Burk plot revealed intersecting lines on the 1/(S) axis, indicating that compound **1** inhibited *β*-glucuronidase in a non-competitive inhibition mode. In addition, a Dixon plot was used to determine the *K<sup>i</sup>* value between the inhibitor and the enzyme, where the intersection value on the x-axis implies *Ki*. The *K<sup>i</sup>* is the concentration of an inhibitor needed to decrease the maximum rate of the reaction by 50% [25]. As presented in Figure 3B, the *K<sup>i</sup>* value of **1** was calculated to be 40.9 μM.

**Figure 3. Figure 3.** Lineweaver−Burk plot ( Lineweaver−Burk plot ( **A A** ) and Dixon plot ( ) and Dixon plot ( **B B** ) analyses ) analyses using active compound using active compound **1 1**. .

#### *2.4. Molecular Docking Studies*

Molecular docking simulations were performed using AutoDock 4.2 to predict how compound **1** behaves in the binding site of the target protein and to clarify the fundamental biochemical processes of **1** with the *β*-glucuronidase enzyme. The results were analyzed and displayed using PyMOL and BIOVIA Discovery Studio (Figure 4). According to

the kinetic analysis results, compound **1** displayed a non-competitive inhibition mode, suggesting that **1** could precisely bind to a specific region of *β*-glucuronidase, named the allosteric binding site, which differed from the active site (orange spheres) that binds to the substrate PNPG (red). The allosteric binding site of *β*-glucuronidase was predicted using the AlloSite 2.10 web server (green spheres) [26] and protein allosteric sites server PASSer2.0 (rainbow spheres) [27] (Figure 4A). To optimize the docking procedure, the substrate, PNPG, was docked as a native ligand into *β*-glucuronidase (PDB ID: 6LEL) (Figure 4B). to the kinetic analysis results, compound **<sup>1</sup>** displayed a non-competitive inhibition mode, suggesting that **<sup>1</sup>** could precisely bind to a specific region of *β*-glucuronidase, named the allosteric binding site, which differed from the active site (orange spheres) that binds to the substrate PNPG (red). The allosteric binding site of *β*-glucuronidase was predicted using the AlloSite 2.10 web server (green spheres) [26] and protein allosteric sites server PASSer2.0 (rainbow spheres) [27] (Figure 4A). To optimize the docking procedure, the substrate, PNPG, was docked as a native ligand into *β*-glucuronidase (PDB ID: 6LEL) (Figure 4B).

Molecular docking simulations were performed using AutoDock 4.2 to predict how compound **1** behaves in the binding site of the target protein and to clarify the fundamental biochemical processes of **1** with the *β*-glucuronidase enzyme. The results were analyzed and displayed using PyMOL and BIOVIA Discovery Studio (Figure 4). According

*Appl. Sci.* **2022**, *12*, x FOR PEER REVIEW 5 of 13

*2.4. Molecular Docking Studies*

**Figure 4.** Predicted binding sites ((**A**); orange spheres: active site, green spheres: possible allosteric binding site predicted by AllositePro, and rainbow spheres: possible allosteric binding site predicted by PASSer2.0) and docking pose results of the substrate and compound **1** with the *β*-glucuronidase enzyme ((**B**); substrate: red; compound **1**: blue). **Figure 4.** Predicted binding sites ((**A**); orange spheres: active site, green spheres: possible allosteric binding site predicted by AllositePro, and rainbow spheres: possible allosteric binding site predicted by PASSer2.0) and docking pose results of the substrate and compound **1** with the *β*-glucuronidase enzyme ((**B**); substrate: red; compound **1**: blue).

The docking results revealed that compound **1** could bind to the allosteric binding site of *β*-glucuronidase with a binding energy value of −8.35 kcal/mol (Figure 5 and Table 2). The carbonyl group and oxygen atom of **1** establish hydrogen bonding interactions with the amino acid residues, Thr191 and Thr438, whereas the hydroxy groups of **1** interact with Arg272, Leu435, and Asp436. Two benzene rings of **1** bind with residues Lys400 and Pro403 through π−alkyl interactions and Thr188 via π−σ interactions. The lactone ring displayed π−σ interactions with Val190, π−cation interactions with Arg439, and π−lone pair interactions with Val189. The other residues, including Gly270, Asn401, and Pro437, from different sites of the *β*-glucuronidase enzyme, bound to compound **1** via van der Waals interactions. The docking results revealed that compound **1** could bind to the allosteric binding site of *β*-glucuronidase with a binding energy value of −8.35 kcal/mol (Figure 5 and Table 2). The carbonyl group and oxygen atom of **1** establish hydrogen bonding interactions with the amino acid residues, Thr191 and Thr438, whereas the hydroxy groups of **1** interact with Arg272, Leu435, and Asp436. Two benzene rings of **1** bind with residues Lys400 and Pro403 through π−alkyl interactions and Thr188 via π−σ interactions. The lactone ring displayed π−σ interactions with Val190, π−cation interactions with Arg439, and π−lone pair interactions with Val189. The other residues, including Gly270, Asn401, and Pro437, from different sites of the *β*-glucuronidase enzyme, bound to compound **1** via van der Waals interactions.

**Table 2.** Docking energies and binding site interactions of compound **1** with the *β*-glucuronidase enzyme.


**Comp. Binding Energy**

**1** −8.35

**(kcal/mol)**

**Figure 5. The** 3D docking poses (**A**) and 2D interaction diagrams (**B**) of *β*-glucuronidase inhibition mediated by compound **1**. Green: hydrogen bonds, violet and pink: hydrophobic, orange: electrostatic, and light green: van der Waals interactions. **Figure 5. The** 3D docking poses (**A**) and 2D interaction diagrams (**B**) of *β*-glucuronidase inhibition mediated by compound **1**. Green: hydrogen bonds, violet and pink: hydrophobic, orange: electrostatic, and light green: van der Waals interactions.

#### **Table 2.** Docking energies and binding site interactions of compound **1** with the *β*-glucuronidase *2.5. Molecular Dynamics Simulation of β-Glucuronidase Inhibition*

enzyme. **Hydrogen Bonds van der Waals Interactions Hydrophobic Interactions Electrostatic Interactions Others** To examine the structural stability and its variations of the **1**-6LEL complex, MD was performed after docking computations with a period of 100 ns, using the Desmond package (Schrödinger 2020-1, New York, NY, USA).

Thr191 Arg272 Leu435 Thr438 Gly270 Asn401 Pro437 Thr188 (π−σ) Val190 (π−σ) Lys400 (π−alkyl) Pro403 (π−alkyl) Arg439 (π−cation) Val189 (π−lone pair) *2.5. Molecular Dynamics Simulation of β-Glucuronidase Inhibition* To examine the structural stability and its variations of the **1**-6LEL complex, MD was performed after docking computations with a period of 100 ns, using the Desmond package (Schrödinger 2020-1, New York, NY, USA). The general conditions of the simulation and its stabilization are obtained by rootmean-square deviation (RMSD) analysis [28]. The higher stability of the protein–ligand complex was demonstrated by a lower RMSD value, whereas decreased stability was indicated by an increased RMSD value [28]. The RMSD value of the **1**-6LEL complex increased rapidly during the initial equilibration fluctuation for 5 ns, slowly increasing from 5 to 38 ns, after which the RMSD was maintained between 2.2 and 2.4 Å , until the simulation was completed (Figure 6A). A maximum RMSD of 2.6 Å was observed for the target protein of *β*-glucuronidase, indicating that the **1**-6LEL complex maintained stability throughout the MD period. The flexibility and fluctuation of each residue in *β*-glucuronidase over the 100 ns simulation were represented by the root-mean-square fluctuation The general conditions of the simulation and its stabilization are obtained by rootmean-square deviation (RMSD) analysis [28]. The higher stability of the protein–ligand complex was demonstrated by a lower RMSD value, whereas decreased stability was indicated by an increased RMSD value [28]. The RMSD value of the **1**-6LEL complex increased rapidly during the initial equilibration fluctuation for 5 ns, slowly increasing from 5 to 38 ns, after which the RMSD was maintained between 2.2 and 2.4 Å, until the simulation was completed (Figure 6A). A maximum RMSD of 2.6 Å was observed for the target protein of *β*-glucuronidase, indicating that the **1**-6LEL complex maintained stability throughout the MD period. The flexibility and fluctuation of each residue in *β*-glucuronidase over the 100 ns simulation were represented by the root-mean-square fluctuation (RMSF) value used to predict the ligand binding-induced structural alterations in the protein structure [29]. Higher RMSF values represent greater flexibility in MD simulations [30]. The RMSF plot of the complex between compound **1** and *β*-glucuronidase is displayed in Figure 6B, where the peaks represent the *β*-glucuronidase regions that fluctuated the most during the 100 ns simulation. The RMSF values of amino acid residues in the allosteric site of *β*-glucuronidase were less than 1.3 Å. In contrast, the RMSF values of residues in the active site of *β*-glucuronidase fluctuate slightly more, between 2.0 and 3.5 Å, suggesting that the protein structure in ligand-bound conformations is stable during MD simulation. Hence, compound **1** may function as a non-competitive inhibitor of *β*-glucuronidase, which is consistent with the results of the kinetic study.

(RMSF) value used to predict the ligand binding-induced structural alterations in the protein structure [29]. Higher RMSF values represent greater flexibility in MD simulations [30]. The RMSF plot of the complex between compound **1** and *β*-glucuronidase is displayed in Figure 6B, where the peaks represent the *β*-glucuronidase regions that fluctuated the most during the 100 ns simulation. The RMSF values of amino acid residues in the allosteric site of *β*-glucuronidase were less than 1.3 Å . In contrast, the RMSF values of residues in the active site of *β*-glucuronidase fluctuate slightly more, between 2.0 and 3.5 Å , suggesting that the protein structure in ligand-bound conformations is stable during MD simulation. Hence, compound **1** may function as a non-competitive inhibitor of *β*-

glucuronidase, which is consistent with the results of the kinetic study.

**Figure 6.** Molecular dynamics simulation of active compound **1** and *β*-glucuronidase (PDB ID: 6LEL) complex: RMSD ((**A**), protein RMSD: green line and RMSD of **1**: red line) and RMSF (**B**). **Figure 6.** Molecular dynamics simulation of active compound **1** and *β*-glucuronidase (PDB ID: 6LEL) complex: RMSD ((**A**), protein RMSD: green line and RMSD of **1**: red line) and RMSF (**B**). **Figure 6.** Molecular dynamics simulation of active compound **<sup>1</sup>** and *β*-glucuronidase (PDB ID: 6LEL) complex: RMSD ((**A**), protein RMSD: green line and RMSD of **1**: red line) and RMSF (**B**).

Figure 7 presents the protein–ligand contact diagram between compound **1** and *β*glucuronidase. The hydroxy groups of **1** established hydrophobic contact with Leu435, polar interactions with Thr191 and His192, a positive charge with Lys400, and a negative charge with Asp436; these residues of the *β*-glucuronidase allosteric site contributed to the binding interactions of 24, 17, 13, 28, and 98%, respectively. In addition, the lactone ring of **1** was linked with Thr438 via polar interactions; Arg272 and Arg439 through π−cation interactions, accounting for 12 and 41% of the contribution, respectively. The MD results revealed the crucial role of the lactone ring in the *β*-glucuronidase inhibitory activity mediated by active compound **1**. Figure 7 presents the protein–ligand contact diagram between compound **1** and *β*glucuronidase. The hydroxy groups of **1** established hydrophobic contact with Leu435, polar interactions with Thr191 and His192, a positive charge with Lys400, and a negative charge with Asp436; these residues of the *β*-glucuronidase allosteric site contributed to the binding interactions of 24, 17, 13, 28, and 98%, respectively. In addition, the lactone ring of **1** was linked with Thr438 via polar interactions; Arg272 and Arg439 through π−cation interactions, accounting for 12 and 41% of the contribution, respectively. The MD results revealed the crucial role of the lactone ring in the *β*-glucuronidase inhibitory activity mediated by active compound **1**. Figure 7 presents the protein–ligand contact diagram between compound **1** and *β*glucuronidase. The hydroxy groups of **1** established hydrophobic contact with Leu435, polar interactions with Thr191 and His192, a positive charge with Lys400, and a negative charge with Asp436; these residues of the *β*-glucuronidase allosteric site contributed to the binding interactions of 24, 17, 13, 28, and 98%, respectively. In addition, the lactone ring of **1** was linked with Thr438 via polar interactions; Arg272 and Arg439 through π−cation interactions, accounting for 12 and 41% of the contribution, respectively. The MD results revealed the crucial role of the lactone ring in the *β*-glucuronidase inhibitory activity mediated by active compound **1**.

**Figure 7.** Protein–ligand contact analysis between compound **1** and *β*-glucuronidase complex ((**A**), timeline and (**B**), bar chart presentations) and 2D interaction diagram (**C**). **Figure 7.** Protein–ligand contact analysis between compound **1** and *β*-glucuronidase complex ((**A**), timeline and (**B**), bar chart presentations) and 2D interaction diagram (**C**). **Figure 7.** Protein–ligand contact analysis between compound **1** and *β*-glucuronidase complex ((**A**), timeline and (**B**), bar chart presentations) and 2D interaction diagram (**C**).

The ligand torsion plot that characterizes the conformational evolution of each rotatable bond (RB) in compound 1 during the 100 ns simulation trajectory is displayed in Figure 8A, where the 2D chemical structure of compound **1** with colored RB is followed by a same-colored dial and bar plots. A total of four RBs were observed in compound **1**, with the potential values of 3.36 kcal/mol and 7.16 kcal/mol for hydroxy groups and the linkage between the phenol ring and lactone ring, respectively. The ligand torsion plot that characterizes the conformational evolution of each rotatable bond (RB) in compound 1 during the 100 ns simulation trajectory is displayed in Figure 8A, where the 2D chemical structure of compound **1** with colored RB is followed by a same-colored dial and bar plots. A total of four RBs were observed in compound **1**, with the potential values of 3.36 kcal/mol and 7.16 kcal/mol for hydroxy groups and the linkage between the phenol ring and lactone ring, respectively.

**Figure 8.** Ligand torsion profile (**A**) and ligand properties trajectory (**B**) of the **1**-6LEL complex throughout the simulation trajectory (0 to 100 ns). **Figure 8.** Ligand torsion profile (**A**) and ligand properties trajectory (**B**) of the **1**-6LEL complex throughout the simulation trajectory (0 to 100 ns).

Considering ligand properties, we evaluated the ligand RMSD, the radius of gyration (rGyr), intramolecular hydrogen bonds (intraHB), molecular surface area (MolSA), solvent-accessible surface area (SASA), and polar surface area (PSA) (Figure 8B). The RMSD of compound **1,** with respect to the reference conformation, ranged from 0.2 to 1.5 Å , and its balance was approximated at 0.45 Å . rGyr was analyzed to examine the stability of compound **1** in the allosteric binding site of *β*-glucuronidase during the 100 ns simulation. The **1**-6LEL complex exhibited an average rGyr value of 3.70 Å . No significant fluctuations were observed in rGyr, suggesting that the **1**-6LEL complex exhibited steady behavior. IntraHB refers to the number of internal hydrogen bonds within the ligand. The constant intraHB value for ligand **1** indicated the consistency of **1** during the simulation process. The MolSA value was calculated using a 1.4 Å probe radius and was equivalent to the van der Waals surface area. The MolSA value for ligand **1** was slightly altered, from 237.5 to 245.0 Å , throughout the 100 ns MD simulation. The SASA value represents the surface area of a molecule that can be accessed by a water molecule. The SASA value of the **1**- 6LEL complex significantly increased from 25 to 60 Å until 4 ns of the MD simulation, Considering ligand properties, we evaluated the ligand RMSD, the radius of gyration (rGyr), intramolecular hydrogen bonds (intraHB), molecular surface area (MolSA), solventaccessible surface area (SASA), and polar surface area (PSA) (Figure 8B). The RMSD of compound **1,** with respect to the reference conformation, ranged from 0.2 to 1.5 Å, and its balance was approximated at 0.45 Å. rGyr was analyzed to examine the stability of compound **1** in the allosteric binding site of *β*-glucuronidase during the 100 ns simulation. The **1**-6LEL complex exhibited an average rGyr value of 3.70 Å. No significant fluctuations were observed in rGyr, suggesting that the **1**-6LEL complex exhibited steady behavior. IntraHB refers to the number of internal hydrogen bonds within the ligand. The constant intraHB value for ligand **1** indicated the consistency of **1** during the simulation process. The MolSA value was calculated using a 1.4 Å probe radius and was equivalent to the van der Waals surface area. The MolSA value for ligand **1** was slightly altered, from 237.5 to 245.0 Å, throughout the 100 ns MD simulation. The SASA value represents the surface area of a molecule that can be accessed by a water molecule. The SASA value of the **1**-6LEL complex significantly increased from 25 to 60 Å until 4 ns of the MD simulation, followed by gradual stabilization at 56 Å. PSA is the SASA of a molecule, provided only by oxygen

and nitrogen atoms. The PSA of compound **1** fluctuated at a consistent rate throughout the 100 ns MD simulation, while the PSA value ranged between 185 and 195 Å.

#### **3. Discussion**

All the secondary metabolites were isolated from *A. mono* Maxim. (**1**−**8**) and evaluated to determine their *β*-glucuronidase inhibitory activity. Compound **1** exhibited a *β*-glucuronidase inhibitory effect, with an IC<sup>50</sup> value of 58.83 µM, whereas the remaining compounds demonstrated no inhibitory activity against *β*-glucuronidase. The results suggest that the different positions of an oxygen atom and carbonyl group in the lactone ring of the isocoumarin structure, as well as the presence of a double bond between C-3 and C-4, could contribute to *β*-glucuronidase inhibition. Kinetic analysis indicated that active compound **1** displayed non-competitive inhibition. Thus, molecular docking studies were employed to determine the binding position and critical amino acid interactions of compound **1** with the allosteric binding site of the *β*-glucuronidase protein. The docking results indicated that compound **1** could tightly bind to the allosteric binding site of *β*glucuronidase, with a binding energy of −8.35 kcal/mol. The lactone ring of compound **1** displays hydrogen bonds with Thr191 and Thr438, π−σ interactions with Val190, π−cation interactions with Arg439, and π−lone pair interactions with Val189, which might explain their inhibitory activity against *β*-glucuronidase. In addition, the conformational stability of compound **1** complexed with the *β*-glucuronidase protein and its variations were investigated by performing MD simulation trajectories (100 ns). According to the MD results, all ligand properties fluctuated during the initial simulation period, gradually reaching equilibrium and a steady state as the simulation was completed. This indicated that compound **1** was stable in the allosteric binding site of *β*-glucuronidase. Based on the predicted pharmacokinetic properties (Supplementary Material), compound **1** exhibited no blood–brain barrier penetration and behaved as a safe drug candidate, given that it adhered to Lipinski's rule of five.

#### **4. Materials and Methods**

#### *4.1. Experimental Procedures*

<sup>1</sup>H and <sup>13</sup>C-NMR spectra were acquired on a Bruker Advance Digital 500 MHz instrument (Bruker, Karlsruhe, Germany). CC was conducted on silica gel 60 (230–400 mesh) and Cosmosil C<sup>18</sup> reversed phase gel (Nacalai Tesque, Kyoto, Japan). HPLC was conducted using a Waters HPLC system with a 2996 PDA detector (Waters, Milford, MA, USA). Thinlayer chromatography was conducted on pre-coated glass plates (silica gel 60 F<sup>254</sup> and RP-18 F254s; Merck, Darmstadt, Germany). Plates were checked under ultraviolet radiation (254 and 365 nm), followed by spraying with sulfuric acid 10% and heating at 100–110 ◦C.

#### *4.2. Chemicals and Reagents*

HPLC solvents were supplied from Fisher Scientific Korea Ltd. (Seoul, Korea). *E. coli β*-glucuronidase enzyme (EC 3.2.1.31, G7396) and DSA (S0375) were provided by Sigma-Aldrich (St. Louis, MO, USA). PNPG (N0618) was provided by Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). All chemicals used in the experiments were provided by Duksan Pure Chemicals Inc. (Ansan, Gyeonggi, Korea).

#### *4.3. Plant Material*

Stems and branches of *A. mono* Maxim. were collected from the Medicinal Herb Garden of Daegu Catholic University in August 2021 and identified by Professor Byung Sun Min at the College of Pharmacy, Daegu Catholic University, Korea. A voucher specimen of *A. mono* Maxim. (23A-AM) was deposited at the Laboratory of Pharmacognosy, College of Pharmacy, Kyungpook National University, Korea.

#### *4.4. Extraction and Isolation*

Dried stems and branches of *A. mono* Maxim. (5.4 kg) were cut into small pieces and extracted with methanol (4 × 20 L) under reflux. The MeOH solvent was evaporated under reduced pressure to obtain the MeOH residue (350.0 g), which was subsequently suspended in H2O (3 L) and then partitioned with *n*-hexane, CH2Cl2, and EtOAc to obtain the *n*-hexane extract (35.2 g), CH2Cl<sup>2</sup> extract (77.8 g), EtOAc extract (62.4 g), and water layer after removing the chemical solvents.

The CH2Cl<sup>2</sup> extract was separated by CC on silica gel using a stepwise eluent of *n*-hexane–acetone (100:1–1:100, *v*/*v*) and CH2Cl2–MeOH (30:1–1:1, *v*/*v*) to yield seven fractions, 1A–G. Fraction 1D (8.1 g) was chromatographed on silica gel CC, eluted using a gradient mixture of CH2Cl2–acetone (100:1–1:100, *v*/*v*) and CH2Cl2–MeOH (20:1–1:1, *v*/*v*), to yield six subfractions, 1D1–6. Subfraction 1D2 (49.7 mg) was separated by RP-18 CC, using a mixture of acetone and H2O (1.5:1, *v*/*v*) as the eluent to obtain compound **3** (8.9 mg). Compounds **4** (5.8 mg) and **5** (2.0 mg) were purified from subfractions 1D3 (140.1 mg) and 1D4 (92.2 mg), respectively, by RP-18 CC elution with MeOH–H2O (1:1.5, *v*/*v*). Compounds **1** (4.9 mg) and **2** (8.9 mg) were isolated from subfraction 1D6 (223.2 mg) by HPLC using an isocratic mixture of MeOH and distilled H2O (40:60, *v*/*v*). Fraction 1E (9.5 g) was separated by silica gel CC, using CH2Cl2–acetone (gradient 100:1–1:100, *v*/*v*) and then CH2Cl2–MeOH (gradient 30:1–1:1, *v*/*v*) as the eluent to yield eight subfractions, 1E1−8. From subfraction 1E7 (239.2 mg), compounds **7** (44.5 mg) and **8** (17.4 mg) were isolated by HPLC, using 52% MeOH in distilled H2O as the eluent.

The water layer was chromatographed by Diaion CC using MeOH–H2O (gradient 0:1−1:0, *v*/*v*) as the eluent to yield three fractions, 2A−C. Fraction 2B (43.7 g) was chromatographed on silica gel and eluted with CH2Cl2–MeOH (gradient 100:1–1:100, *v*/*v*), followed by RP-18 CC using a MeOH–H2O mixture (1.5:1, *v*/*v*) as the mobile phase to obtain four fractions, 2B1−4. Compound **6** (7.2 mg) was isolated from fraction 2B1 (3.8 g) by HPLC, using an isocratic mixture of MeOH and H2O (50:50, *v*/*v*) as the eluent.

3-(3,4-Dihydroxyphenyl)-8-hydroxyisocoumarin (**1**), yellow needles, <sup>1</sup>H-NMR (500 MHz, CD3OD-*d*4): *δ*<sup>H</sup> 7.56 (1H, t, *J* = 7.9 Hz, H-6), 7.46 (1H, d, *J* = 2.0 Hz, H-20 ), 7.32 (1H, d, *J* = 7.6 Hz, H-5), 7.11 (1H, dd, *J* = 8.2, 2.0 Hz, H-60 ), 6.87 (1H, d, *J* = 8.1 Hz, H-7), 6.80 (1H, d, *J* = 7.6 Hz, H-50 ), 6.46 (1H, s, H-4); <sup>13</sup>C-NMR (125 MHz, CD3OD-*d*4): *δ*<sup>C</sup> 167.9 (C-1), 144.1 (C-3), 108.6 (C-4), 116.6 (C-5), 137.8 (C-6), 116.4 (C-7), 158.3 (C-8), 109.9 (C-9), 143.8 (C-10), 126.9 (C-10 ), 111.7 (C-20 ), 146.5 (C-30 ), 147.5 (C-40 ), 117.7 (C-50 ), 124.2 (C-60 ).

#### *4.5. β-Glucuronidase Inhibition Assay*

The *β*-glucuronidase inhibition assay was evaluated as previously described [7].

#### *4.6. β-Glucuronidase Kinetics Assay*

The *β*-glucuronidase kinetics assay was performed as previously described [7].

#### *4.7. Molecular Docking*

Docking simulations were conducted using AutoDock 4.2, following our previously described protocol [7]. The crystallographic structure of *β*-glucuronidase was retrieved from the RCSB PDB website (PDB ID: 6LEL) [31]. The allosteric site of *β*-glucuronidase was predicted and generated using the AllositePro method provided by the AlloSite 2.10 web server and PASSer2.0 protein allosteric sites server [26,27].

#### *4.8. Molecular Dynamics Simulation*

MD simulations were performed using the Desmond package (Schrödinger 2020- 1, New York, NY, USA). The protein–ligand complex was prepared in a 10.0 × 10.0 × 10.0 Å orthorhombic box (simple point-charge solvation model) [32]. Next, a 0.15 M NaCl solution and counter-ions were added to the system for neutralization. The solvated system was energy-minimized, and its position was restrained with the OPLS3e force field. The minimized system was implemented in an NPT ensemble at 300 K and 1 atm. Finally, the

MD simulation was conducted to run for 100 ns, and 1000 frames were generated, with a recording interval of 100 ps.

#### *4.9. Statistics*

All results are expressed as the mean ± standard error (SEM) of the three independent experiments. Statistical significance was analyzed using one-way analysis of variance (ANOVA) and Duncan's test (*p*-value < 0.05).

#### **5. Conclusions**

In the present study, we performed a chemical investigation of the stems and branches of *A. mono* Maxim., which resulted in the purification and structural elucidation of eight known compounds (two isocoumarins and six coumarins). To the best of our knowledge, our study is the first report that isolated isocoumarins from an *Acer* species. In addition, we, for the first time, determined the inhibitory activity of the isolated molecules against *β*-glucuronidase. Our results revealed that 3-(3,4-dihydroxyphenyl)-8-hydroxyisocoumarin (**1**) inhibited *β*-glucuronidase activity. The results of the kinetic analysis were consistent with the molecular docking studies, suggesting that compound **1** could bind to the *β*glucuronidase allosteric site. This result was supported by MD studies up to 100 ns, which confirmed the binding stability of the protein–ligand complex in the trajectory analysis. These findings imply that the complex of *β*-glucuronidase and compound **1** is quite stable in biological systems. Moreover, the pharmacokinetic properties of compound **1** were analyzed, and the results suggested that compound **1** could be a promising drug candidate, given that no violations of the drug-likeness rules were observed.

**Supplementary Materials:** The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/app122010685/s1, Figure S1. Bioavailability radar of compound **1**; Table S1. Drug-likeness properties of compound **1** [33–37].

**Author Contributions:** Conceptualization and methodology, S.Y.Y.; investigation, and data curation, N.V.P.; formal analysis, writing—original draft preparation, N.V.P. and S.Y.Y.; writing—review and editing, B.S.M. and J.A.K.; supervision, J.A.K. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (No. NRF-2020R1A5A2017323 and NRF-2022R1C1C1004636).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Inah Gu , Luke Howard and Sun-Ok Lee \***

Department of Food Science, University of Arkansas, Fayetteville, AR 72704, USA **\*** Correspondence: sunok@uark.edu; Tel.: +1-479-575-6921

**Abstract:** Volatile compounds in fruits are responsible for their aroma. Among fruits, berries contain many volatile compounds, mainly esters, alcohols, terpenoids, aldehydes, ketones, and lactones. Studies for volatile compounds in berries have increased extensively as the consumption of berry products rapidly increased. In this paper, we reviewed biosynthesis and profiles of volatiles in some berries (strawberry, blueberry, raspberry, blackberry, and cranberry) and their bioavailability and health benefits, including anti-inflammatory, anti-cancer, anti-obesity, and anti-diabetic effects in vitro and in vivo. Each berry had different major volatiles, but monoterpene had an important role in all berries as aroma-active components. Volatile compounds were nonpolar and hydrophobic and rapidly absorbed and eliminated from our body after administration. Among them, monoterpenes, including linalool, limonene, and geraniol, showed many health benefits against inflammation, cancer, obesity, and diabetes in vitro and in vivo. More research on the health benefits of volatile compounds from berries and their bioavailability would be needed to confirm the bioactivities of berry volatiles.

**Keywords:** berry volatiles; biosynthesis; chemical composition; bioavailability; health benefits

**Citation:** Gu, I.; Howard, L.; Lee, S.-O. Volatiles in Berries: Biosynthesis, Composition, Bioavailability, and Health Benefits. *Appl. Sci.* **2022**, *12*, 10238. https:// doi.org/10.3390/app122010238

Academic Editors: Luca Mazzoni, Maria Teresa Ariza Fernández and Franco Capocasa

Received: 30 August 2022 Accepted: 5 October 2022 Published: 12 October 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

#### **1. Introduction**

Berries, one of the most common fruits in the human diet (strawberry, blueberry, red raspberry, black raspberry, blackberry, and cranberry in the United States), are rich in minerals, vitamins, dietary fibers, and especially polyphenols and volatiles [1–3]. Volatile compounds in berries are responsible for the unique aroma of berries [4]. Fruit volatile compounds are mainly comprised of diverse chemicals, including esters, alcohols, terpenoids, aldehydes, ketones, and lactones [5]. The volatile composition of berries is complex and different by many factors, including the cultivar, ripeness, pre- and post-harvest storage conditions, fruit samples, temperature, and experimental conditions [6–11]. Blueberries (*Vaccinium ashei*) showed linalool increasing and α-terpineol and β-caryophyllene decreasing during the maturation of the blueberry [6]. Full-red harvested strawberries contained more volatile compounds than <sup>3</sup> ⁄4-red harvested strawberries, regardless of the storage duration [12]. Raspberries (raw, frozen, or frozen for a year) were examined to compare the long-term frozen storage. The changes in volatile composition during long-term frozen storage were negligible except for an increase in α-ionone and caryophyllene [8].

Volatile compounds are small and light molecules (below 250–300 Da) with low polarity and high vapor pressure [13]. Plants synthesize and release volatile compounds to communicate and interact with environments, compensating for the immobility of plants [14]. Volatile compounds play an important role in pollination by attracting pollinators, protecting from pathogens and herbivores, and even communicating with inter- and intra-plants [15]. Volatiles can be divided into primary compounds and secondary compounds. Primary compounds are synthesized during maturation by anabolic or catabolic pathways of the plant [16]. Secondary volatile compounds are produced from tissue disruption by autoxidation or enzyme catalyzing reactions [17–19].

A mixture of many different volatile compounds makes a unique aroma. Although a lot of compounds were found as volatile compounds in fruits, only a few compounds have been identified as aroma compounds of fruit flavor based on their quantitative abundance and olfactory thresholds [20]. With the increasing consumption of berries and berry products such as fruits, juice, puree, jams, and other berry ingredients, there have been many studies conducted about identifying aroma volatile compounds of berries for developing consumer acceptability [21] and the studies about the health beneficial effect of berries, especially berry polyphenols. Berry polyphenols are composed of flavonoids, phenolic acids, tannins, stilbenes, lignans, and others and have been shown to possess many health effects, such as antioxidant, anti-inflammatory, and anti-cancer activities [22–24]. Unlike berry polyphenols, although there have been extensive analyses of volatile berry composition, there is still a very limited number of studies on the bioavailability and health benefits of berry volatiles. Recently, volatile compounds in plants have been reported to have health-promoting activities such as anti-inflammatory effects [25,26]. Many review articles mainly focused on the composition of berry volatiles and affecting factors such as different locations, ripeness, cultivar/genotypes, harvest and storage conditions, etc., but there is a lack of information about the bioavailability and biological activities of berry volatiles in our body. In this article, the biosynthesis of volatiles in plants, the chemical composition of some berries commonly consumed in the U.S. (blackberry, blueberry, cranberry, raspberry, and strawberry), bioavailability, and the health benefits of volatile compounds that are rich in berries were reviewed.

#### **2. Biosynthesis of Volatiles in Plants**

Plant volatiles can be grouped into terpenoids, phenylpropanoids/benzenoids, and fatty acid derivatives based on chemical structure and biochemical synthesis [14,27,28]. Terpenes/terpenoids are the most abundant and diverse family of secondary plant metabolites and essential oils, including more than 30,000 compounds [29,30]. Terpenes are classified by the amount of C5 isoprene units in the structure: isoprene (C5), monoterpenes (C10), sesquiterpenes (C15), diterpenes (C20), and so on, with polyterpenes (C5n where n can be ~30,000) [30]. Monoterpenes and sesquiterpenes are the most abundant terpenes found in essential oils [31]. Although they have diverse chemical structures, they all share common biosynthesis pathways, and they are synthesized in all parts of the plants, such as the leaves, fruits, flowers, stems, and roots [15]. Since terpenes/terpenoids are the major class of berry volatiles, we focused more on the biosynthesis of terpenes, such as monoterpenes and sesquiterpenes, in this review.

#### *2.1. Biosynthesis of Terpenes*

All terpenes are synthesized from two universal precursors (C5), isopentenyl diphosphate (IPP) and dimethylallyl pyrophosphate (DMAPP) [32]. A series of condensation reactions of IPP and DMAPP produce prenyl diphosphates, which are precursors of terpenes. Condensation of IPP and DMAPP produces geranyl diphosphate (GPP, C10), the precursor of monoterpenes, by catalyzation of geranyl diphosphate synthase (GDS). Two IPPs and a DMAPP are condensed to the precursor of sesquiterpenes, farnesyl diphosphate (FPP, C15), by farnesyl diphosphate synthase (FDS) [33]. Three IPPs and a DMAPP also produce geranylgeranyl pyrophosphate (GGPP, C20), the precursor of diterpenes. Geranylfarnesyl diphosphate (GFPP, C25) is the recently discovered precursor of sesterterpenes [33]. These prenyl precursors of terpenes are converted to terpenes (iso-, mono-, sesquiterpenes, and so on) by terpene synthases (TPS). Monoterpenes (C10) are one of the major groups of terpenes in essential oils and berries [34]. During the conversion of prenyl precursors to monoterpenes, many different catalytic reactions, including hydroxylation, oxidation, reduction, acetylation, methylation, glycosylation, isomerization, conjugation, and others, occur to modify the structure of monoterpenes and produce many different compounds of terpenes [30,35,36].

#### *2.2. MVA and MEP Pathways*

The two precursors of all terpenes, IPP and DMAPP, are generated from two different pathways in different subcellular compartments: the mevalonic acid (MVA) pathway in the cytosol and the 2-methylerythritol 4-phosphate (MEP) pathway (1-deoxy-xylulose-5 phosphate (DOXP) pathway) in plastids (Figure 1) [37]. The MVA pathway generates IPP from acetyl-CoA, while the MEP pathway produces IPP and DMAPP from pyruvate and glyceraldehyde-3-phosphate (GA-3P) [38,39]. In the MVA pathway, there are six enzymatic reactions to generate IPP. Three acetyl-CoA are sequentially condensed, reduced to mevalonate (MVA), then form IPP through two phosphorylations and decarboxylation by mevalonate kinase, phosphomevalonate kinase, and mevalonate diphosphate decarboxylase, respectively [40,41]. Produced IPP in the MVA pathway further forms its allylic isomer, DMAPP, by isopentenyl diphosphate isomerase [42]. In the MEP pathway, seven enzymatic actions are involved in generating IPP and DMAPP [43]. Condensation of pyruvate and GA-3P generates DOXP, and DOXP synthesizes MEP by DOXP reductoisomerase (DXR). Further transformations form 1-hydroxy-2-methyl-2-(E)-butenyl 4-diphosphate (HMBPP) and (E)-4-hydroxy-3-methlbut-2-enyl diphosphate reductase (HDR) catalyzes the conversion of HMBPP to IPP and DMAPP [44]. MVA pathway mainly synthesizes sesquiterpenes, which account for about 28% of total flower terpenes, while the MEP pathway synthesizes more monoterpenes and diterpenes, accounting for about 53% and 1% of total flower terpenes, respectively [45,46]. However, metabolic crosstalk exists between the MVA and MEP pathways, especially from plastids to cytosol [47–49].

#### *2.3. Biosynthesis of Other Plant Volatiles*

#### 2.3.1. Phenylpropanoids/Benzenoids

Phenylpropanoids/benzenoids are synthesized from the shikimate pathway [31]. The shikimate pathway starts with the condensation of phosphoenolpyruvate and erythrose 4-phosphate [51]. Chorismic acid is formed by the elimination of ring alcohol from shikimic acid, and this forms the phenylpropionic acid skeleton [29]. Cinnamic acid and phydroxycinnamic acid undergo many enzymatic reactions to produce volatile compounds, including eugenol and benzyl benzoate [52].

#### 2.3.2. Volatile Fatty Acid Derivatives

Fatty acid-derived volatiles are synthesized via the lipoxygenase pathway [53]. C18 fatty acids, including linoleic acid and linolenic acid, are catalyzed by lipoxygenases and generate 9-hydroperoxy and 13-hydroperoxy derivatives of fatty acids [54]. These two intermediates turn into fatty acid derivatives, including methyl jasmonate and green leaf volatiles. The lipoxygenase pathway also can synthesize oxylipins, isoprene, carotenoid derivatives, indoles, phenolics, methyl salicylate, and aromatic volatile organic compounds [55,56].

#### *2.4. Application of Plant Volatile Biosynthesis*

Volatile compounds that are emitted from plants have an important role in many different functions, such as pollinator attraction, direct and indirect defenses against herbivores, insects, and microorganisms, and communication between and within plants [27,57,58]. In addition, natural volatile compounds such as methyl jasmonate, allyl isothiocyanate, and tea tree oil have been used for modulating volatile biosynthesis and controlling the pre- and post-harvest quality of berries [9,59]. Those volatile compounds also suppressed the decay in strawberries and blackberries stored at 10 ◦C [60]. Sangiorgio et al. found a positive correlation between *Lactobacillus*, *Paenibacillus* spp., and norisoprenoids and a negative correlation between Enterobacteriaceae and monoterpenes [61]. From these results, the raspberry microbiome can be selectively chosen for the overall better quality of fruit, including its aroma, shelf-life, and safety [61]. Although there is still more research on the mechanisms required, accumulated results in metabolomic and genomic approaches can be used for making advances in fruit ripeness, quality, and consumer acceptability [62–64].

tween the MVA and MEP pathways, especially from plastids to cytosol [47–49].

**Figure 1.** Biosynthetic pathways of volatile terpenes in plants (adopted and modified from Nagegowda [50]). AACT = acetoacetyl-CoA thiolase; AcAc-CoA = acetoacetyl-CoA; CDP-ME = 4- (cytidine 50-diphospho)-2-C-methyl-D-erythritol; CDP-ME2P = 4-(cytidine 50-diphospho)-2-C-methyl-D-erythritol phosphate; CMK = CDP-ME kinase; DMAPP = dimethylallyldiphosphate; DOXP=1-deoxy-D-xylulose 5-phosphate; DXR = DOXP re-ductoisomerase; DXS=DOXP synthase; FDS = farnesyl diphosphate synthase; FPP = farnesyl diphosphate; GA-3P = glyceraldehyde-3-phosphate; GDS = geranyl diphosphate synthase; GGDS = geranyl geranyl diphosphate synthase; GGPP **Figure 1.** Biosynthetic pathways of volatile terpenes in plants (adopted and modified from Nagegowda [50]). AACT = acetoacetyl-CoA thiolase; AcAc-CoA = acetoacetyl-CoA; CDP-ME = 4-(cytidine 50-diphospho)-2-C-methyl-D-erythritol; CDP-ME2P = 4-(cytidine 50-diphospho)-2-C-methyl-Derythritol phosphate; CMK = CDP-ME kinase; DMAPP = dimethylallyldiphosphate; DOXP = 1-deoxy-D-xylulose 5-phosphate; DXR = DOXP re-ductoisomerase; DXS = DOXP synthase; FDS = farnesyl diphosphate synthase; FPP = farnesyl diphosphate; GA-3P = glyceraldehyde-3-phosphate; GDS = geranyl diphosphate synthase; GGDS = geranyl geranyl diphosphate synthase; GGPP = geranyl geranyl di-phosphate; GPP = geranyldiphosphate; HDR = (E)-4-hydroxy-3-methylbut-2-enyl di-phosphate reductase; HDS = (E)-4-hydroxy-3-methylbut-2-enyl diphosphate synthase; HMBPP = (E)-4-hydroxy-3 methylbut-2-enyl diphosphate; HMG-CoA = 3-hydroxy-3-methylglutaryl-CoA; HMGR = HMG-CoA reductase; HMGS = HMG-CoA synthase; IDI = isopentenyl diphosphateisomerase; IPP = isopentenyl diphosphate; ISPS = isoprene synthase; MCT = 2-C-methyl-D-erythritol 4-phosphate cytidylyltransferase; MDS = 2-C-methyl-D-erythritol 2,4-cyclodiphosphate synthase; ME-2,4cPP = 2-C-methyl-D-erythritol 2,4-cyclodiphosphate; MEP = 2-C-methyl-D-erythritol 4-phosphate; MVD = mevalonate diphosphate decar-boxylase; MVK = mevalonate ki-nase; PMK = phosphomevalonate kinase; TPS=terpene synthase. Names of the enzymes are in gray.

and others, occur to modify the structure of monoterpenes and produce many different

The two precursors of all terpenes, IPP and DMAPP, are generated from two different pathways in different subcellular compartments: the mevalonic acid (MVA) pathway in the cytosol and the 2-methylerythritol 4-phosphate (MEP) pathway (1-deoxy-xylulose-5-phosphate (DOXP) pathway) in plastids (Figure 1) [37]. The MVA pathway generates IPP from acetyl-CoA, while the MEP pathway produces IPP and DMAPP from pyruvate and glyceraldehyde-3-phosphate (GA-3P) [38,39]. In the MVA pathway, there are six enzymatic reactions to generate IPP. Three acetyl-CoA are sequentially condensed, reduced to mevalonate (MVA), then form IPP through two phosphorylations and decarboxylation by mevalonate kinase, phosphomevalonate kinase, and mevalonate diphosphate decarboxylase, respectively [40,41]. Produced IPP in the MVA pathway further forms its allylic isomer, DMAPP, by isopentenyl diphosphate isomerase [42]. In the MEP pathway, seven enzymatic actions are involved in generating IPP and DMAPP [43]. Condensation of pyruvate and GA-3P generates DOXP, and DOXP synthesizes MEP by DOXP reductoisomerase (DXR). Further transformations form 1-hydroxy-2-methyl-2-(E)-butenyl 4-diphosphate (HMBPP) and (E)-4-hydroxy-3-methlbut-2-enyl diphosphate reductase (HDR) catalyzes the conversion of HMBPP to IPP and DMAPP [44]. MVA pathway mainly synthesizes sesquiterpenes, which account for about 28% of total flower terpenes, while the MEP pathway synthesizes more monoterpenes and diterpenes, accounting for about 53% and 1% of total flower terpenes, respectively [45,46]. However, metabolic crosstalk exists be-

compounds of terpenes [30,35,36].

*2.2. MVA and MEP Pathways* 

#### **3. The Chemical Composition of Volatile Compounds in Berries**

The major volatile compounds identified from five common berries in the U.S. (strawberry, blueberry, raspberry, blackberry, and cranberry) were summarized (Table 1).


**Table 1.** The major volatile compounds in berries.

#### **Table 1.** *Cont.*


*Furanones*

Mesifurane [65–67,69]


#### **Table 1.** *Cont.*

101

Cis-1,5-octadien-3-one [71]

Furaneol [65,66] [74] [76] [71]


**Table 1.** *Cont.*

#### *3.1. Strawberry*

Strawberries (*Fragaria* spp.) are the most consumed berry fruit for their sweet taste and unique aroma [86]. The consumption of strawberry products, such as jams, juices, and puree, has significantly increased [87,88]. Strawberries are rich in volatile compounds responsible for the strawberry flavor and aroma [89,90]. Volatile compounds in strawberries have been extensively studied, and more than 360 volatile compounds have been identified [91]. These compounds include esters, which were qualitatively and quantitatively dominant, terpenes, furanones, sulfur, lactones, alcohols, and carbonyls. In the study of Lu et al., a total of 42 volatiles were detected, with 19 esters, 10 alcohols, and 6 terpenes being the most abundant in the strawberry samples analyzed [65].

Esters are the most abundant and major aroma volatiles affecting the aroma of strawberries [92]. Among 19 ester compounds, methyl butanoate, methyl hexanoate, ethyl acetate, and hexyl acetate were the major ester compounds [65]. Terpenes are important compounds to the flavor and possess many preferable aroma profiles [93]. Among six terpenes, limonene and α-terpinene were the main compounds in strawberries [65]. Although strawberries contained many alcohol compounds, alcohols did not affect the strawberry flavor notably. In addition, c-decalactone (peach-like aroma) contributes significantly to the strawberry flavor. Hexanal, trans-2-hexenal, and cis-3-hexen-1-ol are responsible for the green, unripe aroma of the strawberry. The furanones 2,5-dimethyl-4-hydroxy-3(2H) furanone (furaneol) and 2,5-dimethyl-4-methoxy-3(2H)-furanone (mesifurane), which have the sweet, floral, and fruity aroma, are major furans found in strawberries [65].

Summarizing some strawberry volatile studies, volatile components that are consistently considered important aroma compounds include ethyl butanoate, ethyl hexanoate, methyl butanoate, ethyl 3-methylbutanoate, fureneol, and linalool [66,85,94]. Additional compounds that have been reported include 2-heptanone, mesifurane, cis-3-hexenal, ethyl 2 methylpropanoate, 2,3-butanedione, 3-methylbutyl acetate, methyl hexanoate, and ethyl 2 methylbutanoate [67]. Eight different strawberry varieties (Sabrosa, Albion, Sweet Ann, Festival, Fortuna, Ventana, Camarosa, and Rubygem) were examined to investigate the volatile composition [72]. With headspace solid-phase micro-extraction gas chromatography-mass spectrometry (HS-SPME/GC-MS), esters (31 volatiles), especially ethyl hexanoate were detected as major compounds [72]. The most detected compounds were esters (31 compounds), which give fruity and floral characteristics to strawberries [72]. The SPME/GC-MS is one of the most common and effective methods for volatile analysis. SPME fiber samples the volatile compounds from air and thermally desorbs the sample in the injection port of a GC system [95,96]. In Gu et al., 55 volatiles were found in the strawberry extract, with monoterpene being the predominant volatiles (43% of total volatile concentration) [69]. It was followed by acids, esters, furan, aldehydes, alcohol, ketones, and alkylbenzene. Predominant compounds in strawberry extract were myrtenol, butanoic acid, mesifuran, ethyl butanoate, and hexyl butanoate.

#### *3.2. Blueberry*

Blueberries (*Vaccinium* spp.) are the second most popular berry fruit in the U.S. after strawberries [97]. The blueberry market has increased 10–20% annually over the last 5 years [98]. The increased blueberry consumption is due to its well-known health benefits and flavor [99]. Highbush blueberries mainly contained ethyl acetate, (E)-2-hexenal, (E)-2 hexenol, hexanal, (Z)-3-hexenol, linalool, and geraniol [68]. Others, including citronellol, α-terpineol, 2-phenylethanol, and vanillin, were also considered to have the highbush blueberry aroma [81]. A total of 38 aroma volatiles were detected from southern highbush blueberry [74]. There were nine aldehydes, eight esters, seven terpenes, five ketones, two alcohols, two acids, two sulfurs, and three miscellaneous compounds. Aldehydes were the most abundant chemical group within southern highbush blueberry and a major volatile compound group to the blueberry aroma. Hexanal, (Z)-3-hexenal, (E)-2-hexenal, (E,Z)-2,6-nonadienal, and (E,E)-2,4-nonedienal had "fresh green", "grassy", and "fruity" aroma characteristics, whereas pentanal, octanal, (E,E)-2,4-hexadienal, and decanal had "fatty" and "citrus" [74]. Esters were the second most abundant in southern highbush blueberry, and they included ethyl propanoate, methyl 2-methylbutanoate, methyl 3 methylbutanoate, ethyl 2-methylbutanoate, ethyl 3-methylbutanoate, (Z)-3-hexyl acetate, (E)-2-hexyl acetate, and geranyl acetate, with "green", "sweet", "fruity", "apple", "banana", "pear", and "floral" aromatic characteristics. Unlike strawberries, apples, and bananas, where esters are the main contributors to the aroma, fewer esters were found in highbush blueberry [66,100]. Terpenes including linalool, citronellol, nerol, and geraniol showed "sweet", "floral", "fruity", "citrus", and "berry-like", while 1,8-cineole, dihydrolinalool oxide, and α-terpineol had "woody", "herbaceous", and "piney" characteristics. Linalool was one of the major volatiles in southern highbush blueberry. Two alcohols, including (Z)- 3-hexenol and 2-heptanol, were aroma active. Only three ketones, including 2-heptanone, 1-octen-3-one, and 2-nonanone with "fruity" "mushroom", "earthy", and "cheese-like", were found to have aroma activity. Furaneol was found in southern highbush blueberries. Furaneol had "sweet", "candy", and "caramel" characteristics. In another study, five *Vaccinium* cultivars ("Biloxi", "Brigitta Blue", "Centurion", "Chandler", and "Ozark Blue") were found to have 106 volatile organic compounds [78]. Esters, 25 compounds, were the major compounds and were followed by 18 aldehydes, 16 alcohols, 14 monoterpenes, 7 ketones, 4 acids, 4 hydrocarbons, 3 sesquiterpenes, 1 lactone, and 1 norisoprenoid. Aldehydes, including (E)-2-hexenal, hexanal, (Z)-3-hexenal, hexadienal, or heptenal, are the most abundant, accounting for almost half of the *Vaccinium* volatile composition. Monoterpenes, including 1,8-cineole and linalool, were also important compounds in the volatile blueberry profile [78]. Even though esters had smaller content compared to others, esters had a unique aroma that characterizes the aroma of blueberry. Among seven ketones identified, 2-heptanone and 6-methyl-5-hepten-2-one were the ones with the highest contents. Octane, ethyl benzene, p-xylene, and mxylene were identified as hydrocarbons. Other volatile compounds in the blueberry aroma profile were hexanoic acid, octanoic acid, nonanoic acid, decanoic acid (acids), d-elemene, (E)-caryophyllene, and caryophyllene oxide (sesquiterpenes), b-damascenone (norisoprenoid), and butyrolactone (lactone). Dymerski et al. conducted identification of volatile blueberry compounds, and alcohol (51.8%), ester (32.8%), and carboxylic acid (6.9%) were mainly detected [101]. Fortysix blueberry volatiles were identified by Gu et al. by using GC-MS [69]. Monoterpenes accounted for 45% of total volatile concentration, with alcohols (17%), aldehydes (8%), C13 norisoprenoids and esters (each 7%), furans (5%), ketones (4%), and others. The major individual blueberry volatiles were linalool, linalool oxide, phenylethyl alcohol, 2-ethylhexanol, α-terpineol, and β-ionone.

#### *3.3. Raspberry*

In the Rosaceae family, raspberry (*Rubus spp.*) is a fruit with an attractive appearance and unique flavor [102,103]. There are red raspberry (*Rubus idaeus*) and black raspberry (*Rubus occidentalis*). Different cultivars and varieties of raspberries are grown worldwide,

in Europe, North America, and Asia [88]. Raspberries have been reported to contain an aroma impact compound, which is a single compound that has an odor characteristic of raspberry [67]. This compound has been identified as 1-(phydroxyphenyl)-3-butanone and is referred to as raspberry ketone. Approximately 200 volatile compounds were detected in raspberries [76]. Raspberry ketone, α-ionone, β-ionone, linalool, (Z)-3-hexenol, geraniol, nerol, α-terpineol, furaneol, hexanal, β-ocimene, 1-octanol, β-pinene, β-damascenone, ethyl 2-methylpropanoate, (E)-2-hexenal, heptanal, and benzaldehyde have been identified as the raspberry aroma. Among them, α-ionone, β-ionone, geraniol, nerol, linalool, and raspberry ketone especially contributed to the red raspberry aroma. Monoterpene is an important class of fruit volatile organic compounds (VOCs) [104]. This class contains some of the most aroma-active compounds, such as citronellol, nerol, geraniol, α-terpineol, and linalool. The volatile composition of raspberries was identified with 30 compounds, including (Z)-hexenol, hexanal, (E)-2-hexenal, 2-heptanone, δ-octalactone, δ-decalactone, geraniol, α-ionone, β-ionone, and terpinen-4-ol [79]. The main volatile compounds in raspberries include monoterpenes (20%), acids (14%), alcohols (12%), esters (12%), aldehydes (8%), ketones (7%), C-13 norisopernoids (6%), hydrocarbons (6%), lactones (4%), sesquiterpenes (4%), furans (3%), sulfur (3%), and phenols (1%) [105]. Gu et al. identified 78 and 73 volatiles from black and red raspberry extracts, respectively [69]. The major chemical class was monoterpene (61% in black raspberries, 47% in red raspberries) in both raspberries. In black raspberries, (−)-myrtenol, linalool, α-terpineol, 2-ethylhexanol, cuminaldehyde, hexanoic acid, ethyl acetate, and (+)-myrtenol were the major compounds. In red raspberries, myrtenol, butanoic acid, linalool, eugenol, 3-methylbutanoic acid, α-ionone, and vanillin were found as major compounds.

#### *3.4. Blackberry*

Blackberries (*Rubus* spp.) are produced worldwide and consumed mostly as fresh but also as frozen, preserves, jelly, wine, dietary supplements, and others [106]. There have been studies identifying and analyzing the volatile composition of blackberries, but they are still limited. In D'Agostino et al., thirteen *Rubus ulmifolius* schott blackberries from different locations in Italy and Spain were used to identify the volatile composition by using SPME and GC-MS [80]. They identified a total of 74 volatiles from blackberry samples. Esters and aliphatic alcohols were the major classes, and methylbutanal, ethanol, 2,3-butanedione, trans-2-hexenal, 3-hydroxy-2-butanone, 1-hexanol, 1-octanol, and methylbutanoic acid were mainly found in all samples, which were 76.4% and 65.1% of volatile blackberry profiles from Italy and Spain, respectively. Wang et al. compared the aroma compositions of Chickasaw blackberries grown in Arkansas and Oregon [71]. A total of 84 compounds, including 19 esters, 18 terpenes and terpenoids, 15 alcohols, 13 aldehydes, 4 ketones, 4 acids, 4 lactones, 2 furans, 2 sulfur-containing compounds, 1 pyrazine, and 2 miscellaneous compounds, were identified. Even though they were the same cultivar, climate difference in the two regions strongly affected their blackberry aroma. The most attributing aromas of Chickasaw from Oregon were ethyl butanoate, linalool, methional, trans,cis-2,6-nonadienal, cis-1,5-octadien-3-one, and 2,5-dimethyl-4-hydroxy-3(2H)-furanone. However, the most potent aromas in Chickasaw from Arkansas were ethyl butanoate, linalool, methional, ethyl 2-methylbutanoate, β-damascenone, and geraniol. In sensory evaluations, Oregon samples were evaluated to have green, fruity, citrus, and watermelon aromas, while Arkansas samples were to have cinnamon, piney, floral, sweet, and caramel aromas [71]. Qian and Wang also investigated the volatile compositions of Marion and Thornless Evergreen blackberries by using GC-MS [73]. Acids (53.83%) and alcohols (24.25%) were the most abundant compounds in Marion, while alcohols (46.62%) were the most abundant in Thornless Evergreen. Thornless Evergreen blackberries showed much more amounts of volatiles (27.33 ppm) compared to Marion blackberries (8.62 ppm). The most abundant individual volatiles were acetic, hexanoic, decanoic, and 2/3-methylbutanoic acids, ethanol, and linalool for Marion, and 2-heptanol, octanol, α-pinene, hexanol, p-cymen-8-ol, and nopol for Thornless Evergreen. Based on odor activity values (OAVs), the most potent odorants were ethyl hexanoate, β-ionone, linalool, 2-heptanone, 2-undecanone, α-ionone, and hexanal for Marion, and ethyl hexanoate, 2-heptanone, ethyl 2-methylbutanoate, 2-heptanol, 3-methylbutanal, α-pinene, limonene, p-cymene, linalool, t-2-hexenal, myrtenol, hexanal, 2-methylbutanal, and sabinene [73]. Sixty-one volatiles were identified from volatile blackberry extract: 24 monoterpenes, 12 alcohols, 6 esters, 4 ketones, 4 C13 norisoprenoids, 3 furans, 2 acids, a lactone, and a phenolic [69]. Acids (57%) accounted for the highest concentration, followed by alcohols (18%), esters (10%), monoterpenes (10%), and others. Major individual compounds were butanoic acid, hexanoic acid, 4-methyl-1-pentanol, myrtenol, 2-ethylhexanol, isophorone, limonene, and 4-terpineol. Morin et al. identified a total of 80 volatiles in volatile extracts from three blackberry genotypes: Natchez and two University of Arkansas breeding lines, A2528T and A2587T [70]. Monoterpenes, alcohols, and esters were the predominant chemical classes in the three genotypes. As individual volatile compounds, ethyl acetate and α-terpineol were found to be the major volatiles in all three genotypes.

#### *3.5. Cranberry*

Cranberries (*Vaccinium* spp.) are native to North America, and production has been highly increased due to their well-known health benefits of cranberries [107]. The unique cranberry aroma is developed during ripening [75]. In 1981, Hirvi et al. detected 70 volatile compounds from European (*Vaccinium oxycoccus*, L.) and American cranberries (*Vaccinium macrocarpon*, Ait.) [82]. In this study, benzyl alcohol accounted for 29.2% and 21.6% in European and American cranberries, respectively. α-terpineol was 13% and 9.7% of total volatiles in European and American cranberries, respectively. Ruse et al. identified 21 volatiles from fresh cranberries [77]. Common volatile compounds detected from wild (*Vaccinium oxycoccus* L.) and different cultivars of cranberries (*Vaccinium macrocarpon* Ait., 'Early Black', 'Ben Lear', 'Steven', Bergman' and 'Pilgrim') were 4-penten-2-ol, 3-cis-hexenyl formate, benzaldehyde, α-1-terpineol, butyric acid, and benzyl alcohol. Zhu et al. analyzed the cranberry (*Vaccinium macrocarpon* Ait.) volatile composition of four cultivars ('Early Black', 'Howes', 'Searles', and 'McFarlin') by using GC-MS and GC-olfactometry (GC-O) [75]. A total of 33-36 volatiles were detected as odor-active compounds by GC-O. Hexanal, pentanal, (E)-2-heptenal, (E)-2-hexenal, (E)-2-octenal, (E)-2-nonenal, ethyl 2 methylbutyrate, β-ionone, 2-methylbutyric acid, and octanal were mainly contributing to the cranberry aroma. Khomych et al. detected 54 aromatic compounds in cranberry juice [84]. Twenty-three alcohols (41.2% of total concentration) were predominant in cranberry juice, followed by eight acids (40.7%), ten aldehydes (1.7%), five ketones (2.2%), five ethers (1.4%), three lactones, and each heterocyclic and unidentified compound (less than 1% each). Among alcohols, benzyl alcohol was the major volatile, accounting for 23.1% of total volatile concentration. Moore et al. detected 23 cranberry volatiles by using GC-MS [83]. In terms of total volatile concentration, Monoterpene (84%) was predominantly contained in cranberries, followed by aldehyde (8%) and alcohols (3%). The major volatile compounds were α-terpineol, linalool oxide, eucalyptol, trans-2-decanal, and 2-octanal. In cranberry volatile extract, 35 volatiles were found: 16 monoterpenes, 8 alcohols, 6 aldehydes, 2 esters, 2 ketones, and an acid [81]. Monoterpenes (60%) were predominant in total volatile concentration. α-terpineol, eucalyptol, 2-methylbutyric acid, ethyl benzoate, citronellol, and linalool were the major individual volatile compounds in cranberry.

#### **4. Bioavailability of Berry Volatiles**

The definition of bioavailability by the U.S. Food and Drug Administration (FDA) is "the rate and extent to which the active ingredient or active moiety is absorbed from a drug product and becomes available at the site of action" [108]. More proper meaning is the part of ingested compound reaching the systemic circulation and specific site where it is available in the body [109]. Investigating the bioavailability of a compound is important to find the clinical relevance of the health-promoting activities of the bioactive compound in our body [110]. Thus, it is necessary to study the absorption, distribution, metabolism, and excretion of bioactive compounds.

Bioavailability can be variable due to many different factors, including chemical structure, physical state, solubility, route of administration, and distribution via biotransformation and excretion [111,112]. Based on solubility, volatile compounds and essential oils are relatively more nonpolar hydrophobic compounds than polyphenols that are more polar hydrophilic nutraceuticals [113,114]. Regarding the route of administration, volatile compounds are more suitable for pulmonary administration through inhalation, while polyphenols are normally administrated orally [115]. Oral administration generally takes, on average, 30–90 min of action, while inhalation of gaseous compounds takes, on average, only 2–3 min [116]. The bioavailability of volatiles is largely affected by volatility, instability, and hydrophobicity [117].

Although there are many types of research conducted about the identification and quantification of berry volatiles, information is still lacking on the bioavailability of berry volatiles in animals and humans [118]. In this review, bioavailability studies of essential oils and herbal medicinal products that contain volatiles commonly present in berries were selected to estimate the bioavailability of volatile berry compounds. Unfortunately, the studies found are limited and mostly include animal models.

Most of the bioavailability studies of essential oils showed that the volatile compounds in essential oils are rapidly absorbed and eliminated after pulmonary, dermal, and oral administration [115]. The compounds were mostly metabolized and eliminated within an hour of elimination half-life through the kidney after phase-II conjugation or CO<sup>2</sup> exhalation.

In Igimi et al., the absorption, distribution, and excretion of d-limonene, a monoterpene in many essential oils but also found in black raspberries and blackberries, were investigated in rats [119]. <sup>14</sup>C-labeled d-limonene was orally administered to 21 male Wistar rats. The maximum radioactivity was obtained 2 h after administration in blood and 1–2 h after administration in tissues. High radioactivity in the liver, kidney, and blood became not significant after 48 h. The excretion of d-limonene was 60% in urine, 5% in feces, and 2% in expired CO<sup>2</sup> in 48 h. About 25% of administered d-limonene was eliminated in bile in 24 h in bile duct cannulated rats. Biotransformation studies of (+)-limonene in humans showed that the main metabolites of biotransformation were perillic acid, dihydroperillic acid, and limonene-10-ol, and their glucuronides, perillyl alcohol, p-mentha-1,8-dien-carboxylic acid, cis- and trans-dihydroperillic acid, limonene, 1,2-diol and limonene-8,9-diol [120].

Dermal application of α-pinene on humans (ointment) and mice (bath) resulted in rapid absorption in plasma, reaching maximum plasma levels in 10 min of application [115]. Inhalation of α-pinene in humans resulted in 61% absorption of α-pinene [115]. However, only 4–6% of α-pinene was found to be absorbed in the blood. In α-pinene dermal and pulmonary administration studies [115], the half-life was short in the α-phase (5 min) and longer in the β-phase (26–38 min). α- and β-pinene in humans were metabolized to transand cis-verbenol, respectively, and they were further hydroxylated to diols. In rabbits, transverbenol was major, and myrtenol and myrtenic acid were minor metabolites of α-pinene, while cis-verbenol was the major metabolite of β-pinene [121]. In a recent open-label, singlearm study, ten male subjects consumed Mastiha oil (1 mL) containing rich monoterpenes, and blood samples were collected at 0–24 h after Mastiha oil administration [122]. Mastiha oil contained α-pinene (82.2%), myrcene (8.5%), and β-pinene (2.4%) as the major terpenes and also had linalool and limonene (0.8% each). In subjects' blood samples, the three major terpenes were detected. Myrcene reached its peak at 2.2 h (966.6 µg/L), and α-pinene and β-pinene reached their peaks at 3.8 (914.8 µg/L) and 3.6 h (18 µg/L), respectively [122].

Linalool, one of the major berry volatile compounds, was metabolized to dihydrolinalool, tetrahydrolinalool, and 8-hydroxylinalool, then further oxidized to 8-carboxylinalool by cytochrome P450 (CYP). Metabolites derived by CYP formed glucuronide conjugates [123]. Oral administration of α-terpineol to rats resulted in the metabolization of alpha-terpineol to p-menthane-1,2,8-triol. The major biotransformation occurred in 1,2-double bound with allylic methyl oxidation and reduction [124].

#### **5. Health Benefits of Berry Volatiles 5. Health Benefits of Berry Volatiles**

(18 μg/L), respectively [122].

Despite the sensory properties of fruits and vegetables, there are studies demonstrating that the role of aroma compounds is more than their odor impact [18,125]. Recently, volatile compounds in plants have been reported to have health-promoting activities, including anti-inflammatory [25,126], anti-cancer [26], anti-obesity, and anti-diabetic effects [127]. However, since there are not many studies on the health-promoting effects of berry volatiles, studies of volatile compounds from essential oils and other fruits and plants that berries commonly contain were used to review the potential health benefits of volatile compounds in berries (Figure 2). Despite the sensory properties of fruits and vegetables, there are studies demonstrating that the role of aroma compounds is more than their odor impact [18,125]. Recently, volatile compounds in plants have been reported to have health-promoting activities, including anti-inflammatory [25,126], anti-cancer [26], anti-obesity, and anti-diabetic effects [127]. However, since there are not many studies on the health-promoting effects of berry volatiles, studies of volatile compounds from essential oils and other fruits and plants that berries commonly contain were used to review the potential health benefits of volatile compounds in berries (Figure 2).

Linalool, one of the major berry volatile compounds, was metabolized to dihydrolinalool, tetrahydrolinalool, and 8-hydroxylinalool, then further oxidized to 8-carboxylinalool by cytochrome P450 (CYP). Metabolites derived by CYP formed glucuronide conjugates [123]. Oral administration of α-terpineol to rats resulted in the metabolization of alpha-terpineol to p-menthane-1,2,8-triol. The major biotransformation occurred in 1,2-

[115]. Inhalation of α-pinene in humans resulted in 61% absorption of α-pinene [115]. However, only 4–6% of α-pinene was found to be absorbed in the blood. In α-pinene dermal and pulmonary administration studies [115], the half-life was short in the α-phase (5 min) and longer in the β-phase (26–38 min). α- and β-pinene in humans were metabolized to trans- and cis-verbenol, respectively, and they were further hydroxylated to diols. In rabbits, trans-verbenol was major, and myrtenol and myrtenic acid were minor metabolites of α-pinene, while cis-verbenol was the major metabolite of β-pinene [121]. In a recent open-label, single-arm study, ten male subjects consumed Mastiha oil (1 mL) containing rich monoterpenes, and blood samples were collected at 0–24 h after Mastiha oil administration [122]. Mastiha oil contained α-pinene (82.2%), myrcene (8.5%), and β-pinene (2.4%) as the major terpenes and also had linalool and limonene (0.8% each). In subjects' blood samples, the three major terpenes were detected. Myrcene reached its peak at 2.2 h (966.6 μg/L), and α-pinene and β-pinene reached their peaks at 3.8 (914.8 μg/L) and 3.6 h

double bound with allylic methyl oxidation and reduction [124].

*Appl. Sci.* **2022**, *12*, x FOR PEER REVIEW 12 of 27

**Figure 2.** Potential bioactivities of berry volatiles. COX-2 = cyclooxygenase-2; ERK = extracellular signal-regulated kinase; IL-1β = interleukin-1β; IL-6 = interleukin-6; JNK = c-Jun N-terminal kinase; MAPK = mitogen-activated protein kinase; NF-κB = nuclear factor kappa B; NO = nitric oxide; Nrf2 = nuclear factor erythroid 2-related factor 2; PGE2 = prostaglandin 2; TNF-α = tumor necrosis factor-α; ↓ = decrease. **Figure 2.** Potential bioactivities of berry volatiles. COX-2 = cyclooxygenase-2; ERK = extracellular signal-regulated kinase; IL-1β = interleukin-1β; IL-6 = interleukin-6; JNK = c-Jun N-terminal kinase; MAPK = mitogen-activated protein kinase; NF-κB = nuclear factor kappa B; NO = nitric oxide; Nrf2 = nuclear factor erythroid 2-related factor 2; PGE<sup>2</sup> = prostaglandin 2; TNF-α = tumor necrosis factor-α; ↓ = decrease.

#### *5.1. Inflammation*

Infection, inflammation, or any cellular damage/stimuli are detected by macrophages and dendritic cells through pattern recognition receptors (PRRs) with pathogen-associated molecular patterns and danger-associated molecular patterns [128,129]. Toll-like receptors and intracellular nucleotide-binding domain leucine-rich-repeat-containing receptors recognize these stimuli and stimulate signal transductions, mitogen-activated protein kinases (MAPKs) [130–132]. MAPK signal transduction pathways include extracellular signalregulated kinase (ERK), c-Jun N-terminal kinase (JNK), and p38 and regulate downstream protein kinases and transcription factors [133]. Stimulated signal transductions activate pro-inflammatory transcription factors, such as nuclear factor kappa-B (NF-κB) and nuclear factor erythroid 2-related factor 2 (Nrf2) [132]. NF-κB regulates the production of pro-inflammatory cytokines and chemokines, including interleukin (IL)-1ß, IL-6, nitric oxide (NO), prostaglandin (PGE) 2, and tumor necrosis factor (TNF)-α [134,135]. NF-κB is activated by IκB kinase (IKK) phosphorylating NF-κB-inhibitory protein (IκBα), and translocating into the nucleus, promoting transcription of pro-inflammatory mediators [136]. Nrf2 is another transcription factor related to oxidative damage and inflammation [137]. During the cascade of inflammatory responses, reactive oxidative stress (ROS) increases oxidative stress on cells, leading to the autophagy of cells [138,139]. However, recent studies found that volatile compounds in plants, especially terpenes, mitigate inflammation by suppressing many different inflammatory processes [140,141]. In this section, the effects of volatile compounds rich in berries against inflammation in different steps of inflammatory processes were summarized (Table 2).


**Table 2.** The effect of volatile compounds rich in berries on inflammation models.

= interferon-γ; IκBα = IκB kinase phosphorylating NF-κB-inhibitory protein; IL-18 = interleukin-18; IL-1β = interleukin-1β; IL-6 = interleukin-6; iNOS = inducible nitric oxide synthase; JNK = c-Jun N-terminal kinase; MAPK = mitogen-activated protein kinase; MCP-1 = monocyte chemoattractant protein-1; NF-κB = nuclear factor kappa B; NO = nitric oxide; Nrf2 = nuclear factor erythroid 2-related factor 2; PGE<sup>2</sup> = prostaglandin 2; ROS = reactive oxidative stress; TLR4 = toll-like receptor 4; TNF-α = tumor necrosis factor-α; ↑ = increase; ↓ = decrease.

#### 5.1.1. Modulation of Pro-Inflammatory Mediators

Many volatile compounds in plants showed anti-inflammatory effects by reducing the level of pro-inflammatory mediators and cytokines, such as NO, PGE2, cyclooxygenase (COX-2), TNF-α, and interleukins [69,83,142,145–147,150–154,159,160]. Amorim et al. indicated that the essential oils obtained from Citrus species and limonene demonstrated a significant anti-inflammatory effect by reducing cytokine production, including NO, TNF-α, and IL-1β [142]. The essential oils of citrus fruit peel contain abundant monoterpenes such as limonene, geranial, β-pinene, and γ-terpinene, which are one of the major volatile compounds in berries [159]. In Rehman et al., d-limonene at 5% and 10% doses mixed in a diet given to rats for 20 days effectively reduced doxorubicin-induced COX-2, inducible nitric oxide synthase (iNOS), and NO [160]. In an ulcerative colitis rat model, rats (n = 8/group) fed d-limonene (50 and 100 mg/kg) for 7 days showed anti-inflammatory effects by suppressing the level of iNOS, COX-2, and PGE<sup>2</sup> [145]. In lipopolysaccharide (LPS)-induced RAW264.7 cells, 125–1000 µg/mL of rheosmin (raspberry ketone) isolated from pine needles exerted an anti-inflammatory activity with reduced NO, PGE2, iNOS, and COX-2 production [147]. In the LPS-induced murine macrophage RAW264.7 cell model, α-terpineol treatment (1.16 µg/mL) before and after LPS stimulation showed significant inhibition on the level of NO [83]. In a mouse model of carrageenan-induced peritonitis, oral administration of γ-terpinene 1 h before intraperitoneal carrageenan injection significantly attenuated the TNF-α and IL-1β production [146]. In a triple transgenic Alzheimer's mouse model, oral administration of 25 mg/kg linalool, every 48 h for 3 months, markedly decreased the production of iNOS, COX-2, and IL-1β [150]. Mice administered 2.6 and 5.2 mg/kg linalool before injecting endotoxin remarkably suppressed the nitrate/nitrite, IL-1β, IL-18, TNF-α, and interferon (IFN)-γ production [151]. Oral linalool administration (15 and 30 mg/kg) also lowered iNOS levels in lung tissues in mice with allergic asthma [152]. Intraperitoneal injection of 10, 20, and 40 mg/kg linalool two hours before cigarette smoke exposure for five days ameliorated the lung inflammation by suppressing the level of proinflammatory cytokines (TNF-α, IL-6, IL-1β, IL-8, and monocyte chemoattractant protein (MCP)-1) [153]. Linalool (40–120 µg/mL) attenuated the LPS-induced inflammation on RAW264.7 cells with the suppressed level of TNF-α and IL-6 [154]. In the LPS-stimulated mouse macrophage RAW264.7 cell model, 1 h pretreatment of volatile extracts (50-fold dilution) from blackberry, black raspberry, blueberry, cranberry, red raspberry, and strawberry significantly reduced production in NO, PGE2, and COX-2 [69]. Blackberry, blueberry, cranberry, and volatile strawberry extracts also effectively suppressed LPS-induced TNF-α and IL-6 production.

#### 5.1.2. Regulation of Inflammatory Transcription Factors and Signal Transduction

NF-κB is a crucial target for anti-inflammation since it is one of the main transcription factors regulating pro-inflammatory mediators [161]. Linalool significantly reduced the NF-κB activation in mice with endotoxin injection [151], cigarette smoke-induced acute lung inflammation [153], and airway allergic inflammation [152]. Limonene (25–75 mg/kg) intraperitoneal injection 1 h before LPS administration down-regulated the phosphorylation of IκBα, NF-κB p65, p38 MAPK, JNK, and ERK in LPS-induced acute lung injury mice [143]. In Rehman et al., d-limonene also suppressed NF-κB activation in a doxorubicin-stimulated inflammation rat model [160]. Limonene and myrcene attenuated the IL-1β-induced inflammation by suppressing NF-κB and JNK activation in human chondrocytes [144]. Terpinen-4-ol inhibited NF-κB in the dextran sulfate sodium (DSS)-increased experimental colitis in mice [148]. Terpinen-4-ol also attenuated the LPS-stimulated IκBα and NF-κB p65 phosphorylation in acute lung injury mice [149]. In Wu et al., linalool increased nuclear translocation of Nrf2 in mice with pneumonia infected by *Pasteurella multocida* [155]. Linalool (162–648 µM), one of the major berry volatiles, reduced LPS-stimulated inflammation on BV2 microglia cells through Nrf2/HO-1 signaling pathway [156]. α-Pinene in coniferous trees and rosemary oils suppressed the MAPK and NF-κB activation in LPS-induced macrophages [157]. In LPS-induced RAW264.7 murine macrophage cells, volatile extract (50-fold dilution) from blackberry, black raspberry, blueberry, and cranberry significantly suppressed the NF-κB activation by down-regulating phosphorylation of NF-κB p65 and IκBα [69].

#### 5.1.3. Attenuation of Oxidative Stress and Autophagy

A disparity between the production and elimination of ROS causes oxidative stress. Excessively produced ROS can damage tissues, increasing inflammatory responses and leading to cell death, such as necrosis and apoptosis [162]. There have been many examinations of the antioxidant activities of volatile compounds in plants against oxidative stress in vitro. α-terpinene, γ-terpinene, and linalool showed antioxidant activities in 2,20 azino-bis-3-ethylbenzthiazoline-6-sulphonic acid (ABTS), chelating power, 2,2-diphenyl-1-picrylhydrazyl (DPPH) and oxygen radical absorbance capacity (ORAC) assays [163]. α-pinene, 1,8-cineole, and d-limonene remarkably ameliorated the formation of ROS in H2O2-stimulated oxidative stress [158].

#### *5.2. Cancer*

Kim et al. demonstrated that geraniol inhibits human prostate cancer cell PC-3 proliferation in in vitro and in vivo xenograft mice models [164]. Geraniol at 0.5 and 1 mM significantly suppressed the cell growth of PC-3 by increasing cell cycle arrest and apoptosis. Balb/C nude mice inoculated with PC-3 cells took intratumoral geraniol injection daily for 38 days at 0, 12, 60, or 300 mg/kg. The mice treated with 60 or 300 mg/kg geraniol showed significantly decreased tumor volume and weight. Injection of geraniol at 20 mg/kg also sensitized the chemotherapeutic agent Docetaxel (2 mg/kg) in the xenograft mice model. In Lee et al., geraniol inhibited prostate cancer growth by a down-regulating E2F8 transcription factor and inducing G2/M phase cell cycle arrest [165]. In gastric adenocarcinoma AGS cells, geraniol showed cytotoxicity by inhibiting the JNK/ERK signaling pathway [166]. Lavender essential oil, its active compounds linalool and linalyl acetate [167], and ethyl acetate fraction of Ajwa dates [168] also inhibited PC-3 cell proliferation by increasing apoptosis and cell cycle arrest. Linalool also reduced tumor growth in PC-3 xenograft mice [167] and the 22Rv1 xenograft mice model [169]. Ethyl acetate exerted an anti-proliferative activity on human breast cancer MCF7 and SKBR3 cells [170] and human cervical cancer HeLa cells [171]. α-terpineol [172] and linalool [173] also showed strong cytotoxicity on HeLa cells with apoptosis and cell cycle arrest. In human acute myeloid leukemia U937 cells [173], human oral cancer cells [174,175], and lung adenocarcinoma A549 cells [176], linalool significantly suppressed the cell growth. D-limonene [177] and d-limonene-rich blood orange volatile oils [178] inhibited the proliferation of lung cancer A549 and H1299 cells and human colon adenocarcinoma cells SW480 and HT-29 cells, respectively. Limonene (9 µM) significantly reduced the proliferation of human bladder cancer cell T24 after 24 h, showing induced apoptosis with increased G2/M cell cycle arrest and apoptotic markers (Bax, and cleaved caspase-3, 8, and 9) [179]. In Yu et al., d-limonene induced apoptosis and autophagy-related genes in a lung cancer model [177]. In an acetic acid-induced gastric ulcer rat model, 7-day oral administration of (-)-myrtenol at 50–100 mg/kg increased the healing of the ulcer [180]. Myrcene640 µM) significantly decreased the proliferation of SCC9 oral cancer cells after 24 h [181]. Myrcene also showed increased apoptosis with the concentration of 5–20 µM and significantly suppressed the migration of SCC9 cells at 10 µM myrcene treatment. The effects of volatile compounds rich in berries on cancer models were summarized in Table 3.

**Table 3.** The effect of volatile compounds rich in berries on cancer models.



**Table 3.** *Cont.*

ERK = extracellular signal-regulated kinase; JNK = c-Jun N-terminal kinase; MAPK = mitogen-activated protein kinase; ROS = reactive oxidative stress; ↑ = increase; ↓ = decrease.

#### *5.3. Obesity*

A high-fat diet for obese humans and animals can increase endothelial dysfunction. It can lead to many other severe cardiovascular diseases and metabolic disorders. In Wang et al., geraniol was examined for the effect on endothelial function in high-fat diet (HFD)-fed mice [182]. Forty mice were fed HFD for 8 weeks, while 20 mice had a normal diet. Then, HFD-fed mice were randomly assigned to intraperitoneal geraniol treatment (20 mice) or vehicle treatment (20 mice) group for 6 weeks. As a result, geraniol protected and improved HFD-induced endothelial dysfunction in HFD-fed mice by reducing aortic NADPH oxidases and ROS production. In Sousa et al., α-terpineol enantiomers were exam-

ined for their effect on the biological markers in HFD-induced obese rats [183]. Six weeks of daily α-terpineol supplementation (50–100 mg/kg of diet) suppressed pro-inflammatory cytokines (TNF-α and IL-1β), serum thiobarbituric acid reactive substances (TBARS), and recovered insulin sensibility. In Li et al., Microcapsules of d-limonene-rich sweet orange essential oil (SOEO) were orally administered to HFD-induced obese rats for 15 days [184]. SOEO microcapsules decreased the body weight in obese rats by protecting the gut barrier, increasing *Bifidobacterium*, and reducing low-grade inflammation. While white adipocytes store the excessive energy in triglyceride forms, brown adipocytes burn the calories in heat form by non-shivering thermogenesis [185]. Lone and Yun showed that limonene increased 3T3-L1 adipocytes browning through the activation of the β3-adenergenic receptor and ERK signaling pathway [185]. In Ayala-Ruiz et al., male Wister rats (*n* = 6 per group) were administered with control, high-fat-sucrose diet (HFSD) and 0.6 mL of corn oil, HFSD with 1,8-cineole (0.88 mg/kg), limonene (0.43 mg/kg), α-terpineol (0.32 mg/kg), or the mixture of three terpenes per gavage for 15 weeks [186]. Rats fed with HFSD with terpenes significantly reduced weight gain compared to the ones with only HFSD. In addition, all terpenes suppressed the fat deposition, serum glucose levels, and triacylglycerol levels. The effects of volatile compounds rich in berries against obesity were summarized (Table 4).

**Table 4.** The effect of volatile compounds rich in berries on obesity models.


IL-1β = interleukin-1β; NADPH = nicotinamide adenine dinucleotide phosphate; ROS = reactive oxidative stress; TBARS = thiobarbituric acid reactive substances; TNF-α = tumor necrosis factor-α; ↑ = increase; ↓ = decrease.

#### *5.4. Diabetes*

In Bacanlı et al., streptozotocin (STZ) (60 mg/kg) was injected into Wistar rats to induce type 1 diabetes [187]. Diabetic rats were orally treated with d-limonene (50 mg/kg body weight) for 28 days. D-limonene treatment significantly reduced DNA damage and induced the level of antioxidant enzymes (catalase, superoxide dismutase, and total glutathione). D-limonene also altered hepatic enzyme and lipid profile, suggesting the potential of d-limonene being protective against diabetes in the liver and kidney in rats. D-limonene showed potential antihyperglycemic activities [188,189] and reduced lipid peroxidation with increased antioxidant activity [190]. In El-Bassossy et al., geraniol (150 mg/kg) was orally treated in STZ-induced obese rats for 7 weeks [191]. Geraniol significantly reduced systolic cardiac function related to diabetes by alleviating oxidative stress. Geraniol treatment also reduced GLUT 2 transporter [192], hyperglycemia [193], diabetic nephropathy [194], and improved impaired vascular reactivity [195] in STZ-induced diabetic rats. Linalool also exerted a reduction in fasting blood glucose level, insulin resistance, glycation oxidative stress [196], and nephropathic changes in kidneys [197] on STZ-induced diabetic rats. Xuemei et al. investigated the effect of myrtenol on STZ-induced gestational diabetes mellitus (GDM) in rats [198]. GDM is diabetes that occurs only during pregnancy. Twenty-five mg/kg of STZ was injected into the pregnant rats to induce GDM. Myrtenol (50 mg/kg) was orally administered for 2 weeks. Myrtenol oral administration helped

decrease blood glucose levels and pro-inflammatory markers. It also increased high-density lipoprotein (HDL) and antioxidant status in diabetic pregnant rats. The effects of volatile compounds rich in berries against diabetes were summarized in Table 5.


HDL = high-density lipoprotein; NF-κB = nuclear factor kappa B; TGF-β<sup>1</sup> = Transforming growth factor beta-1; ↑ = increase; ↓ = decrease.

#### **6. Conclusions**

As berry consumption through the fruit and products, including berry-flavored water, juice, and others, have rapidly increased, many studies about monitoring and improving overall berry quality, including flavor, aroma, appearance, shelf-life, and safety, were conducted to target consumer acceptability. However, the beneficial health effects of berry volatiles have not been extensively studied. In this article, we looked into the biosynthesis of plant volatiles, volatile composition, and possible bioavailability and health benefits of some berry volatiles were reviewed. Major terpene volatiles were synthesized via MVA and MEP pathways. Major chemical classes in berries were esters, alcohols, terpenoids, aldehydes, ketones, and lactones. Berries had different profiles of volatiles, but monoterpene showed a crucial role in characterizing the unique berry aroma in all five berries. Volatile compounds were nonpolar and hydrophobic and rapidly absorbed and eliminated from our body after administration. Among them, monoterpenes, including linalool, limonene, and geraniol, showed many health benefits associated with inflammation, cancer, obesity, and diabetes in vitro and in vivo, suggesting potential health beneficial effects of berry volatiles. More research on animal and human models of the health benefits of berry volatiles and bioavailability would be needed to confirm their bioactivities.

**Author Contributions:** Conceptualization, L.H. and S.-O.L.; writing—original draft preparation, I.G.; writing—review and editing, S.-O.L., L.H. and I.G.; supervision, S.-O.L. and L.H.; funding acquisition, L.H. and S.-O.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** The work was supported by the Arkansas Biosciences Institute.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

## **References**

