**Fractionation of Biomolecules in** *Withania coagulans* **Extract for Bioreductive Nanoparticle Synthesis, Antifungal and Biofilm Activity**

**Murtaza Hasan 1,2,\*,**†**, Ayesha Zafar 2,**†**, Irum Shahzadi 2, Fan Luo 1, Shahbaz Gul Hassan 3, Tuba Tariq 2, Sadaf Zehra 4, Tauseef Munawar 5, Faisal Iqbal <sup>5</sup> and Xugang Shu 1,\***


Academic Editor: Giovanni Benelli

Received: 5 June 2020; Accepted: 20 July 2020; Published: 31 July 2020

**Abstract:** *Withania coagulans* contains a complex mixture of various bioactive compounds. In order to reduce the complexity of the plant extract to purify its phytochemical biomolecules, a novel fractionation strategy using different solvent combination ratios was applied to isolate twelve bioactive fractions. These fractions were tested for activity in the biogenic synthesis of cobalt oxide nanoparticles, biofilm and antifungal activities. The results revealed that plant extract with bioactive fractions in 30% ratio for all solvent combinations showed more potent bioreducing power, according to the observed color changes and the appearance of representative absorption peaks at 500–510 nm in the UV-visible spectra which confirm the synthesis of cobalt oxide nanoparticles (Co3O4 NPs). XRD diffraction was used to define the crystal structure, size and phase composition of the products. The fractions obtained using 90% methanol/hexane and 30% methanol/hexane showed more effectiveness against biofilm formation by *Pseudomonas aeruginosa* and *Staphylococcus aureus* so these fractions could potentially be used to treat bacterial infections. The 90% hexane/H2O fraction showed excellent antifungal activity against *Aspergillus niger* and *Candida albicans*, while the 70% methanol/hexane fraction showed good antifungal activity for *C. albicans*, so these fractions are potentially useful for the treatment of various fungal infections. On the whole it was concluded that fractionation based on effective combinations of methanol/hexane was useful to investigate and study bioactive compounds, and the active compounds from these fractions may be further purified and tested in various clinical trials.

**Keywords:** fractionation; reducing activity; biomolecules; antibiofilm; microbial infection

### **1. Introduction**

Adverse increases in the rates of microbial, fungal and viral infections worldwide prompted by compromised and human immunity are due in part to the indiscriminate use of antibiotics that enhances resistance in microbial communities against the corresponding antigens [1]. The generation of biofilms by microbes, which root in a self-produced matrix on living and non-living surfaces [2], is a peculiar behavior of microbes in inducing and producing resistance. Biofilm affinity is associated to a firm attachment of the microbe and biofilm-forming microbes have a great tendency to stick permanently to the large variety of surfaces [3]. These tiny creatures' biofilms are protected by a layer of exopolysaccharides, which can be up to 1000 times more resistant to antimicrobials, which has increased exponentially the rate of chronic infections caused by increased resistance against the host immune system and antibiotics [4,5]. Among such microbes is *Candida albicans*, a well-known resistant nosocomial bacterium primarily known for being the main cause of infectious diseases [6] such as oral thrush [7], vaginitis [8] organ transplant recipients [9] and forms of cancer in HIV/AIDS patients. Besides resistance, the limited availability of commercial drugs effective against bacteria and the resulting toxicity has increased the global rate and effects of infections in people. This severe problem has driven the interest of researchers in developing less toxic, herbal bioactive compounds that could work against such strains of microbes. Similarly, the commonly known resistant fungus, *Aspergillus* species, responsible for pulmonary diseases, has also acquired resistance to many common drugs [10]. In order to overcome these biofilm-producers alternative treatments include the use of antibiofilm agents produced by medicinal plants as this mode of action reduces the resistance susceptibility [11]. Plants, being an enriched source of naturally occurring biologically active components, play a vital role in the prevention and treatment of diseases by boosting immunity and reducing toxicity [12,13]. Ancient plants like *W. coagulans* contain many useful bioactive molecules such as withanolide, withaferin, withacoagin [14], etc., that have been used to synthesize therapeutic drugs for the prevention and treatment of various diseases due to their reduced side effects [15,16]. *W. coagulans* belongs to the Solanaceae, a family of common traditional therapeutic plants with wide range of pharmacological applications [17], including antimicrobial, anti-inflammatory [18], antitumor [19], antihyperglycemic [20], cardiovascular, and immunosuppressive properties [21]. The constituents of *W. coagulans* include free amino acids, essential oils, steroidal lactones and esterases, widely used for their pharmacological activities [22]. A few studies have also recommended the use of withanolide, withaferin and other biological entities found in *W. coagulans* for their bioreducing potential in the synthesis of nanoparticles [23,24], and studies have reported the eco-friendly and less toxic preparation of nanoparticles and pharmacological studies using *W. coagulans* components [25].

So far, all these biological activities were tested using crude extracts containing complex mixtures of active biomolecules and the solvents-based screening, fractionation and functionalization of bioactive compounds has not been previously reported. The development of antibiofilm strategies is a major interest and also the basis of an important field of investigation that is the development of premium, environmentally friendly antibiofilm biomolecules. The present work was focused on investigate the functional role of fractions obtained using methanol and hexane with water and mixtures of methanol and hexane to purified active biomolecules from *W. coagulans* extract. For this purpose, *W. coagulans* fractions were extracted with mixtures containing different ratios of methanol and *n*-hexane and water and methanol and *n*-hexane mixtures in order to evaluate the bioactivities such as bioreducing potential for the synthesis of cobalt nanoparticles, and antioxidant, anti-biofilm and antifungal activities.

### **2. Results**

The increasing resistance of microbes against antibiotics calls for the urgent discovery of unique biomolecules from extracts of plants like *W. coagulans* that are of potential interest for their antibiofilm and antifungal activity and as bioreducing agents for the synthesis of cobalt oxide nanoparticles (Co3O4 NPs). The species *W. coagulans* is highly acclaimed in the Indian ayurvedic system of medicine, where it is known for its medicinal significance in promoting physical and mental health [26,27]. Its active components include alkaloids, steroidal compounds, lactones, withaferin a [28], withanoloids [29], withanone [30], etc. that act as anti-inflammatory, anticancer, chemoprotective, hepatoprotective, immune modulatory, antifungal, antibacterial, hypocholestroemic, and radical scavenging agents [31]. The complex bioactive extract of *W. coagulans* contains potent and functional molecules that must be fractionated to simplify the complexity and provide separate bioactive molecules that can exhibit their functionalities efficiently. Different fractions of plant extract obtained using different solvents and mixtures of solvents were used to resolve the complexity of the biological entities of *W. coagulans* used as bioreducing, antibacterial, antifungal agents [32]. This fractionation route provided a means to separate, simplify and unveil the hidden active molecules in the complex. Initially using a *W. coagulans* extract, 12 different methanol, hexane and their mixture fractions in ratios of 30%, 50%, 70%, 90% were made (Figure S1) and their bio-reducing, antibiofilm and antifungal potential in vitro evaluated (Figure 1).

**Figure 1.** Schematic illustration of *W. coagulans* biomolecules and their applications.

### *2.1. Green Synthesis of Co3O4 NPs*

Pink coloured cobalt chloride solution was mixed individually with all 12 different solvent-based plant extract fractions that turned to a dark brown color upon addition and continuous magnetic stirring at 90 ◦C for five h. As the chemical reaction proceeded the color changed from dark brown to light brown indicating the synthesis of Co3O4 NPs (Figure S2).

### 2.1.1. Characterization of Green Synthesized Co3O4 NPs

Monitoring the reduction potential of synthesized Co3O4 NPs by UV spectroscopy using the *Withania*-based fractions showed different peaks within the 500–510 nm range for different solvent fractions [33]. The methanol and water ratio results conclusively indicated that 30% methanol/H2O (3:1) showed the highest peak, indicating that the 30% fraction was a more active bioreducing fraction than 50% methanol/H2O (5:5), 70% methanol/H2O (7:3), or 90% methanol/H2O (9:1), as they all showed less bioreducing activity [34,35] (Figure 2a). Among the next four fractions based on hexane and water ratio 30% hexane/H2O (3:1) and 90% hexane/H2O (9:1) showed almost same highest peak which indicated that these fractions have more bioreducing potential than 50% hexane/H2O (5:5) and 70% hexane/H2O (7:3). Furthermore 70% hexane/H2O showed a much lower peak with no bio-reducing potential [36,37] (Figure 2b). Similarly, the four methanol/hexane-based fractions with different ratios (30%, 50%, 70%, 90%) were evaluated next for bioreducing potential and was indicated that 30% fraction mixture of methanol/hexane (3:1) showed a much sharper peak indicating better bioreducing potential than 50%, 70%, 90% methanol/hexane fraction mixtures [23,38] (Figure 2c).

**Figure 2.** Bioreducing potential of *W. coagulans* based on: (**a**) methanol (**b**) hexane (**c**) methanol/hexane (mixture) fractions for Co3O4 NPs synthesis.

For optimizing the results a comparative analysis was done between 30% fraction of methanol/H2O, 30% hexane/H2O and 30% methanol/hexane and the results demonstrated that out of all mixtures the 30% methanol/hexane (3:7) fraction mixture showed a much sharper peak. meaning it had a higher bioreducing ability than 30% methanol/H2O and 30% hexane/H2O fraction, as seen in Figure 3a.

**Figure 3.** Bioreducing potential of *Withania coagulans* based on methanol, hexane and methanol/hexane (mixtures) using (**a**) 30% fraction, (**b**) 50% fraction, (**c**) 70% fraction, (**d**) 90% fraction.

Among the 50% fractions, 50% methanol/hexane (5:5) fraction mixture showed higher peaks corresponding to a higher bioreducing potential than 50% methanol/H2O and 50% hexane/H2O, but 50% hexane/H2O and 50% methanol/H2O showed almost the same peak and almost the same bioreducing potential (Figure 3b).

Next, among the different 70% fractions of *W. coagulans*, the 70% methanol/hexane (7:3) mixture fraction showed a high peak with higher bioreducing potential than 70% methanol/H2O and 70% hexane/H2O. Here 70% methanol/H2O showed a much sharper peak (indicating better bioreducing potential) than the 70% hexane/H2O fraction, as illustrated in Figure 3c. Finally, out of all the 90% fractions of *W. coagulans*, 90% methanol/hexane (9:1) fraction mixture showed the highest peak indicating a higher bioreducing potential than 90% methanol/H2O and 90% hexane/H2O. Here different results were observed because 90% hexane/H2O shows a much sharper peak than 70% methanol/H2O meaning that 70% hexane/H2O fraction has higher bioreducing ability than 70% methanol/H2O (Figure 3d).

### 2.1.2. XRD Analysis of Co3O4 NPs

XRD diffraction was used to define the crystal structure and phase composition of the produced NPs. The XRD patterns of the samples obtained with different solvent fraction ratios are presented in Figure 4a–c. The observable diffraction pattern of materials obtained using methanol (fraction (a)), hexane (fraction (b)) and methanol/hexane (fraction (c)) were well-matched with Co3O4. The diffraction patterns of the methanol fraction were thus consistent with JCPDS Card No. 01-080-1534, hexane fraction (b) with JCPDS Card No. 01-074-1657, and methanol/hexane fraction (c) with JCPDS Card No. 01-076-1802, respectively. The peaks and related planes are indicated in Figure 4. The XRD results show that none of the samples have any characteristic peaks due to impurities, which shows that the grown samples have outstanding crystalline nature. The lattice parameters (*a*) and unit cell volume (*v*) of the samples were calculated using the following formula:

$$\frac{1}{d^2} = \frac{h^2 + k^2 + l^2}{a^2} \tag{1}$$

$$v = a^3 \tag{2}$$

**Figure 4.** XRD analysis of Co3O4 nanoparticles based on solvent fractions: (**a**) methanol fraction (**b**) hexane fraction, (**c**) methanol/hexane fraction.

Where (hkl) are the miller index, '*d*' is d-spacing, and '*a*' is lattice constant. The calculated values are listed in Table 1. The average crystallite size (*D*) of all synthesized samples was determined by using the well-known Debye–Scherer Formula [39,40]:

$$D = \frac{K\lambda}{\beta \cos \theta} \tag{3}$$

**Table 1.** Structural parameters of grown samples.


In these equations *K* is the shape factor having value (0.94), λ is the wavelength of X-ray (1.5406 Å), β is the full width at half maxima. From the results, it can be concluded that the crystallite size follows the trend b (59 nm) > a (50 nm) > c (49 nm) (Table 1). The dislocation density (δ) and d-spacing can be calculated by:

$$
\delta = 1/D^2\tag{4}
$$

$$2d\sin\theta = n\,\lambda\tag{5}$$

where 'λ' is the wavelength of X-rays in Å, 'θ' is the diffraction angle (Bragg angle) in degrees, *n* is the order of diffraction which is the spacing between adjacent crystal planes. The calculated values are listed in Table 1. The results show that d-spacing varies directly with crystallite size while dislocation density varies as square inverse of crystallite size.

Furthermore, compound microscopy results (Figure 5a–c) show that changing the nature of the solvent influenced on the shape of Co3O4 NPs. Figure 5a shows bead-shaped Co3O4 NPs obtained using methanol solvent extract as reducing agent [41] while in Figure 5b the shape of Co3O4 NPs obtained with hexane was different because of the different biomolecules present as compared to methanol solvent [42]. In the case of a mixed ratio of methanol and hexane solvents (Figure 5c), the Co3O4 NPs were cube-shaped, most probably because of the action of different active biomolecules in this fraction when they reduce the cobalt nanoparticles [27,43].

From the above results, we can conclude that the active biomolecules exhibiting reducing potential found in the methanol/hexane fraction were proven to have the best bioreducing potential in the synthesis of Co3O4 NPs. Among the different fractional concentrations of similar solvents 30% fraction showed the best bioreducing efficiency. This means that when preparing fractions with these three solvents, and running a separate solvent fraction-based reaction, the 30% fraction will provide more significant results as previously reported [44,45]. It shows a well-defined sharp peak for every solvent containing a 30% solvent faction. The exposed binding sites for the binding of cobalt precursors and saturating the metal by biochemical agents in order to provide stability was done by solvent-based fractionation of *Withania* extract as reported earlier [46].

The scheme (Figure 1) shows a double dip strategy where the nature and concentration of a solvent reduce the complexity, provide active sites and finally highlight the functional activity of the biomolecules. This solvent fractionation actually works similarly to an enzyme substrate reaction, as active sites are provided as product gets generated. Here the fractionation helps expose and present the active sites by reducing the complexity and generating Co3O4 NPs. In the next level of optimization, the concentration was kept constant and the solvent was altered. The results showed that the mixture of methanol/hexane was a hybrid solvent that reinforced the characteristic properties of each solvents. Conclusively in order to optimize our study, mixtures of methanol/hexane, at all concentrations provide the best reduction capacity. Thus, to reduce complexity, unlocking the bioactive molecules in methanol/hexane mixtures of 30% fraction should provide an excellent lead for identifying compounds good at reducing cobalt to Co3O4 NPs.

### *2.2. Biofilm Activity of Prepared W. coagulans Fractions*

Bioactive fractions from *W. coagulans* (12 different fractions) were evaluated for antibiofilm activity against the drug sensitive strains *Pseudomonas aeruginosa* and *Staphylococcus aureus* in 96 well micro-titer plates. The purpose was to evaluate the potential of the 12 different fractions to inhibit the growth of a preformed biofilm already established in the wells of the micro-titer plate [47]. In anti-biofilm assay biofilm was induced to grow on 96 well micro-titer plates by adding 100 μL nutrient broth, 100 μL plant extract and 20,100 μL bacterial culture in each well and incubating for 24 h at 37 ◦C then staining the next day with crystal violet (dye) give a dark blue color to the well where biofilm formation took place (Figure S3). Crystal violet is a dye that binds non-specifically to negatively charged surface molecules such as the polysaccharide matrix of biofilms and stains them with a blue color so it is generally used to estimate biofilm biomass [48], so a reduction in blue color indicates biofilm inhibition by different tested plant fractions.

### *2.3. Antibiotic Selectivity*

First an effective positive control for *P. aeruginosa* and *S. aureus* (drug sensitive strains) was established by treating with four different antibiotics (clindamycin, moxifloxacin, penicillin and ciprofloxacin). The results showed that moxifloxacin and ciprofloxacin were more active drugs against the *P. aeruginosa* strain as indicated by a larger zone of inhibition shown by the drugs (Figure 6a,b) but ciprofloxacin was a more effective antibiotic against *S. aureus* as shown by its larger inhibition zone (Figure 6c). Thus, the strong antibiotic ciprofloxacin was selected to test the *W. coagulans*-based 12 different fractions of methanol and hexane and their mixtures to evaluate the biofilm inhibition potential against *P. aeruginosa* and *S. aureus* at concentrations of 5 mg/mL and 100 mg/mL.

**Figure 6.** Antibiotic selectivity of (**a**) clindamycin, moxifloxacin (**b**) penicillin, ciprofloxacin against *P. aeruginosa* (**c**) penicillin, ciprofloxacin (**d**) clindamycin, moxifloxacin against *S. aureus.*

### 2.3.1. Biofilm Inhibition Potential of *W. coagulans* Fraction against *P. aeruginosa*

Ciprofloxacin, being a positive control against *P. aeruginosa*, shows a reduction of dark blue color of the dye (crystal violet) in the first well and solvent blank without bacterial strain marked as first negative control that does not contain bacteria so no biofilm formation occurred there, thus no crystal violet dye staining was observed (Figure S3), leaving a colorless well indicating the absence of bio-film formation. As a second negative control a well was loaded with 55 mg/mL of *P. aeruginosa* without the plant extract and blue colored biofilm was observed. Color reduction of the dark blue dye in the micro-titer plate well gave a rapid qualitative analysis of biofilm inhibition potential by the crystal violet staining technique that was measured as a percentage inhibition of biofilm formation. With the positive control, ciprofloxacin, the percentage inhibition against *P. aeruginosa* was found to be 50%, and it was 0.7% with the negative control.

After running the successful controls, the *Withania*-derived solvent-based fractions were assessed. For 30% methanol (Meth.I ) the inhibition was 0.6%, for 50% methanol (Meth.II) it was 0.5%, for 70% methanol (Meth.III) it was 29% and for 90% methanol (Meth.IV) the inhibition reached 50%. Hexane was next and 30% hexane (Hex.I ) exhibited 29% inhibition, 50% hexane (Hex.II) showed 30% inhibition, 70% hexane (Hex.III) showed 31% and 90% hexane (Hex.IV) gave about 24% inhibition.

The third series includes mixtures of methanol and hexane, among which 30% methanol-hexane (M−HI ) showed 49% inhibition, 50% methanol-hexane (M−HII) 43%, 70% methanol-hexane (M−HIII) 42% and 90% methanol-hexane (M−HIV) showed only 20% inhibition of biofilm formation. Overall Meth.IV exhibited a 100% percentage inhibition of biofilm production with respect to control. On average Meth. inhibited 40%, Hex. inhibited 57% and methanol-hexane mixture inhibited 77% with respect to control. Hence the solvent mixture super-combination showed superior results on average at all concentrations by decoding the complexity with the hybrid mixture of solvents. Biofilm formation by dye degradation and calculated inhibitions are shown in Figure 7a. These results are relevant to previous work done using plant extracts against the biofilm activity [49].

**Figure 7.** Biofilm activity of *W. coagulans* 12 fraction against *P. aeruginosa* with concentration (**a**) 5 mg/mL (**b**) 10 mg/mL.

Similarly, when using the 10 mg/mL extract against *P. aeruginosa* where the positive control showed 26% inhibition of biofilm and 0.7% of inhibition for the negative control, 0.5% > 35% > 0.6%> 1.5% inhibition was seen for Meth.I > Meth.II > Meth.III > Meth. IV. Moving to the next solvent fraction Hex.<sup>I</sup> > Hex.II > Hex.III > Hex.IV (11% > 0.3%> 49%> 48%) and lastly, for the mixture fraction <sup>M</sup>−H<sup>I</sup> <sup>&</sup>gt; <sup>M</sup>−HII <sup>&</sup>gt; <sup>M</sup>−HIII <sup>&</sup>gt; <sup>M</sup>−HIV (98% >33% > 22%) as depicted in Figure 6b along with biofilm formation (Figure S4). The best fractions M−H<sup>I</sup> , Hex.III, Hex.IV, Meth.II and M−HII provided an outstanding inhibition representing 277%, 88%, 85%, 34% and 26% more than the control. As a result, Meth provided 36% inhibition with respect to control, Hex exhibited 4% more inhibition with respect to control whereas the excelling M−H mixture exhibited 68% more inhibition with respect to the control on average. Some fractions had previously shown significant inhibition with 5 mg/mL *Withania* solution against *P. aeruginosa* [50] but changing the concentration to 100 mg/mL the bio-film percentage inhibition increased even above the control level, showing higher antibacterial activity as shown in Figure 7b.

### 2.3.2. Biofilm Inhibition Potential of *W. coagulans* Fractions against *S. aureus*

The activity of concentrations of each fraction up to 10 mg/mL against *S. aureus* was observed. Biofilm formation against *S. aureus* strain was done with ciprofloxacin as positive control which was found to be active against the drug sensitive *S. aureus* strain as shown by the white colour of wells. A negative control was also added (Figure S5).

The controls gave 55% and 0.4% inhibition, respectively. For the other 12 fractions a concentration of 55 g/mL was used that provided no significant or results as shown in Figure 8a where the positive control inhibition was 55% and that of the negative control was 0.4%. Meth.I > Meth.II > Meth.III > Meth.IV values were 1.2% > 1.8% > 1.9% > 48%. For hexane, i.e., Hex.<sup>I</sup> > Hex.II > Hex.III > Hex.IV the inhibition was 3.1% > 36% > 2.1% > 3.2% and for mixtures M−HI > M−HII > M−HIII > M−HIV, percentage inhibitions of 0.8% > 20% > 17.5% > 12.5% were exhibited which were quite insignificant against such a resistant strain and at a such minute concentration.

Next the change in concentration up to 10 mg/mL against *S. aureus* showed significant results, whereby the positive control showed 40% inhibition and the negative one showed 0% inhibition. In the first fraction series Meth.I > Meth.II > Meth.III > Meth.IV the inhibition was 71% > 65% > 28% > 24%. For hexane fractions, i.e., Hex.<sup>I</sup> > Hex.II > Hex.III > Hex.IV the results showed 3.5% > 20% > 19% > 3.7% inhibition and the percentage inhibition was calculated as 2.7% <sup>&</sup>gt; 62%<sup>&</sup>gt; 72% <sup>&</sup>gt; 72% for <sup>M</sup>−H<sup>I</sup> <sup>&</sup>gt; <sup>M</sup>−HII <sup>&</sup>gt; <sup>M</sup>−HIII <sup>&</sup>gt; <sup>M</sup>−HIV as shown in Figure 8b. Compared to the control M−HIII <sup>&</sup>gt;M−HIV <sup>&</sup>gt; Meth.I <sup>&</sup>gt; Meth.II <sup>&</sup>gt; <sup>M</sup>−HII exhibited 80% <sup>&</sup>gt; 80% <sup>&</sup>gt; 78% <sup>&</sup>gt; 62% <sup>&</sup>gt; 55% more biofilm formation indicating an outstanding result at the particular dilutions that revealed the presence of antibacterial biomolecules in such fractions. On average Meth. showed 18% more inhibition, Hex. showed only 28% inhibition with respect to control and M−H was superior, exhibiting more film formation with 30% inhibition. The present observations regarding bacterial biofilm formation match the work rewported by previous researchers [51].

**Figure 8.** Biofilm activity of *W. coagulans* 12 fraction against *S. aureus* with concentration (**a**) 5 mg/mL (**b**) 10 mg/mL.

### *2.4. Antifungal Activity of Prepared W. coagulans Fractions*

### 2.4.1. Antifungal Activity of Prepared *W. coagulans* against *A. Niger*

The antifungal activity was evaluated using all 12 different fractions extracts of methanol and hexane and their mixtures using plant extract of *W. coagulans* against *A. niger* and *C. albicans* by the disc method [52]. The active biomolecules were resolved into simple *W. coagulans* plant molecules that exhibited antifungal activity. The active principal molecules were measure and made visible by the zone of inhibition produced by the fraction molecules against the specific strains (Figure 9a–h).

**Figure 9.** Antifungal activity of 12 *W. coagulans* fractions against *A. niger* measured by zone of inhibition: (**a**) Positive control; (**b**) negative control; (**c**) methanol fractions; (**d**) hexane fraction 30%, 50%; (**e**) hexane fraction 70%, 90%; (**f**) methanol/hexane fraction 30%, 50% (**g**); methanol/hexane fraction 70%, 90% (**h**).

The tested fractions provided significant results as follows: amphotericin B at concentration (10 mg/mL) was used as standard for both fungus strains that were pathogenic [53]. The positive control shows good antifungal activity against *A. niger* as indicated by the large clear inhibition zone (20 mm) whereas the negative control exhibited no clear zone of inhibition as shown in Figure 9b,c. The tested concentrations beginning with Meth.<sup>I</sup> > Meth.II > Meth.III > Meth.IV exhibited 10 mm > 16 mm > 10 mm > 6 mm inhibition, with an average of 50% antifungal activity compared to the control (Figure 9d). Next is the hexane fractions, Hex.I > Hex.II > Hex.III > Hex.IV showing 10 mm >14 mm >18 mm > 24 mm zone of inhibition with 83% agreement with the control (Figure 9e,f). Finally M−HI <sup>&</sup>gt; <sup>M</sup>−HII <sup>&</sup>gt; <sup>M</sup>−HIII <sup>&</sup>gt; <sup>M</sup>−HIV where the zone of inhibition provided 62% similar result on average with control and 12 mm > 10 mm > 14 mm > 14 mm inhibition zones, respectively (Figure 9g,h). Surprisingly Hex.IV showed 20% more antifungal activity than the control against *A. niger*.

### 2.4.2. Antifungal Activity of Prepared *W. coagulans* against *C. albicans*

*C. albicans* showed a 20 mm zone of inhibition with the positive control amphotericin B, an effective drug against this strain (Figure 10a–h). The negative control provided no zone of inhibition indicating no antifungal activity (Figure 10c). On further treatment the 12 fractions provided significant results, where Meth. provided 6% antifungal activity, Hex. 70% activity and M−H 83% activity with respect to the control.

**Figure 10.** Antifungal activity of 12 *W. coagulans* fractions against *C. albicans* (zone of inhibition): (**a**) Positive control (**b**) negative control (**c**) methanol fractions (**d**) hexane fraction 30%, 50% (**e**) hexane fraction 70%, 90% (**f**) methanol/hexane fraction 30%, 50% (**g**) methanol/hexane fraction 70%, 90% (**h**).

Individually Meth.I > Meth.II > Meth.III > Meth.IV provided 4 mm > 1 mm > 0 mm > 0 mm of inhibition zone (Figure 10d), Hex.<sup>I</sup> > Hex.II > Hex.III > Hex.IV had 10 mm > 12mm > 10 mm > 24 mm inhibition (Figure 10e,f). Finally 16 mm > 14 mm > 24 mm > 12 mm zones of inhibition were measured for M−H<sup>I</sup> > M−HII > M−HIII > M−HIV fractions (Figure 10g,h). With *C. albicans* Hex.IV and M−HIII exhibited 20% more antifungal activity that the control.

The antibacterial and antifungal activity using *W. coagulans* was proven to be significant because of the biomolecules initially present in complex form that were resolved into simple and more functionally active groups by the solvent-based fractionation method. Owing to such a strategy and the significant activity this set of optimizations can be incorporated in the medicinal field in order to combat bacterial and fungal infections. Plant extracts have shown a variety of potentials such as reducing, antioxidant, synthetic, and medicinal activities due to the presence of numerous bio-molecules that exist in different parts of the plant. Depending upon the nature each show different extents of variation in their capabilities due to the presence of some additional biomolecules and the varying concentrations of those biomolecules. Considering *Withani*, it is truly rich in phenols, flavonoids, alkaloids, steroids and

other complex structures that provide reducing, antibacterial and antifungal activities. The results of this work show that the separation of these components using different solvents such as water, hexane, methanol, acetone, etc., enhanced the activities by aiding in resolving the complexity, dissolving components of different nature according to their solubility in different solvents, combining the biomolecules for effective interaction and thus showing their potentials at their maximum level. Similarly, *Withania* had shown antibacterial activity against *Salmonella typhi, Klebsiella pneumonia, S. aureus* with percentage antibacterial activities as 43%, 0%, 73% respectively. *Withania*-decorated iron rods enhanced the activity up to 30% for *S. aureus* and *P. aeruginosa,* whereas *Withania* showed less inhibition against a *Brucella* strain. Multiple examples have shown that the bacterial inhibition of crude extracts was not so high as that achieved by using solvent- based fractionation methods that enhance the values and activity to a significant level. Antifungal activity was exhibited against various strains such as *A. flavus*, *A. niger, Penicillium* and *Alternaria alternate*, where a significant 6–10 mm zone was measured. Thus, the addition of solvents, mixtures of solvents, and the concentration help in simplifying the complex structure of the plant extracts that displayed much higher activities, including bioreduction, antibacterial and antifungal properties.

### **3. Materials and Methods**

### *3.1. Plant Material*

Plant of *W. coagulans* was obtained from a local market in Bahawalpur, Pakistan in September 2018. Fresh plant was washed three time with distilled H2O and kept in the shade until it was completely dried, then it was crushed into powder form for further use.

### *3.2. Preparation of Plant Extract*

Whole plant was dried and crushed using a pestle and mortar to obtain a fine powder, then 10 g of extract powder was dipped in different concentrations of methanol and hexane to make 12 different fractions with ratios of 90%, 70%, 50%, 30% (final volume 200 mL). After overnight incubation the extracts were filtered and the filtrates were dried in an incubator at 37 ◦C. These powder extracts then used to check the bioreducing, antifungal and biofilm activities.

### *3.3. Synthesis of Cobalt Oxide Nanoparticles*

For the synthesis of cobalt oxide nanoparticles, a 0.5 M solution of cobalt chloride was prepared Flasks containing 40 mL cobalt chloride solution and 10 mL plant extract (90%, 70%, 50%, 30%) were prepared, put on a magnetic stirrer (150 rpm) and kept there for 4 h at 90 ◦C as a reaction occurred indicated by a change in color confirming the synthesis of nanoparticles. After this the mixture was centrifuged at 6000 rpm for 10 min., the pellet was separated and dried for characterization.

### *3.4. UV-Vis Spectroscopy*

All 12 fractions were subjected to UV-Vis spectroscopy (Instrument model VT05404-0998, Biotek, Winooski, VT, USA) at predetermined time intervals to confirm the formation of cobalt nanoparticles and the wavelength was noted. Peaks between 550–510 nm give a positive indication of nanoparticle synthesis. Also, the color changes of reaction mixtures were used as evidence of cobalt oxide nanoparticle formation

### *3.5. Morphology Analysis of via Compound Microscope*

The dried form of the cobalt oxide nanoparticles was uniformly distributed in Petri plates with relevant solvent and allowed to dry. Later compound microscopy (model IM-850, IRMECO GmbH, Hamburg, Germany) was used to observe the morphological variations in all three fractions.

### *3.6. Biofilm Assay of W. coagulans Fraction*

Biofilm assays were performed by a crystal violet staining assay. The effect of extracts on biofilm formation was evaluated in 96-well polystyrene plates. Firstly, the 96-well micro-titer plates were washed with sterile distilled water, air dried and then oven-dried at 60 ◦C for 45 min. Briefly, nutrient broth, standard drug (ciprofloxacin) and bacterial culture were used as positive control while nutrient broth, distilled water and bacterial culture were used as negative control. Nutrient broth, plant fractions and bacterial culture were added to each micro-plate and incubated at 37 ◦C for 24 h. After that staining with 0.1% crystal violet was performed and the OD was recorded at 630 nm using an ELISA reader (model IM-850, IRMECO GmbH, Hamburg, Germany) and % inhibition was calculated by following formula:

$$\% \text{ inhibition} = (\text{A}\_0 - \text{A}\_1) / \text{A}\_0 \times 100 \tag{6}$$

where A0 is absorbance of negative control and A1 is the absorbance of the plant fractions

### *3.7. Antifungal Activity of W. coagulans Fraction*

Fresh plant was washed two times with distilled water and allowed to dry at room temperature for 3 to 4 days. The dried material was ground and extracted separately by making different methanol and hexane fractions. The extracts were filtered and the filtrate was dried. All extracts fractions were stored at 4 ◦C and used for the bioassays. The plant extracts were tested against two important fungal pathogens, *C. albicans* and *A. niger,* obtained from the laboratory of the Department of Biochemistry and Biotechnology (Islamia University Bahawalpur). All cultures were maintained on SDA agar at 37 ◦C. Overnight cultures on SGA slants at 37 ◦C were used to prepare the fungal inoculum to be used in the antimicrobial assays. The antifungal activity of *W. coagulans* methanolic and hexane extracts was measured according to the disc diffusion method. Sterile blank discs of 6 mm diameter were soaked with the prepared *W. coagulans* extracts to give a final concentration of 10 mg/mL, respectively. The discs were then placed firmly on a SDA surface which has been previously seeded with *C. albicans* strain suspension. The same steps were repeated for the A. *niger* strain. All plates were incubated overnight at 37 ◦C. Throughout this experiment, a blank disc impregnated with sterile distilled water represented as negative control while a disc soaked with 100 μL of amphotericin B was the positive control. The susceptibility of each *Candida* spp. was determined by the diameter of the growth inhibited zone surrounding the disc.

### **4. Conclusions**

Twelve different *W. coagulans*-based fractions prepared using combinations of different solvents (methanol, hexane) and their mixture were used to study the effect of different solvent combinations on various biological activities. Plant fractions of different concentration (30%, 50%, 70%, 90%) were used. These fraction were used to investigate the bioreducing potential of the plant extracts containing complex biomolecule mixtures, it was found that collectively 30% fraction of methanol. hexane, and mixture of methanol-hexane provided the highest reducing potential for the synthesis of cobalt oxide nanoparticles. Results also showed that 90% methanol/hexane and 30% methanol/hexane were more active against biofilm formation of *P. aeruginosa* and *S. aureus* so these fractions could be used for treatment of various drug resistance-related bacterial infections. A 90% fraction of hexane/H2O showed excellent antifungal activity against *P. niger* and *C. albicans*, while 70% methanol/hexane show good antifungal activity for *C. albicans,* so these fractions are potentially useful for the treatment of various fungal infections. This solvent-based fractionation method provides a direct means to reduce the complexity of the *W. coagulans* extracts and reveal the strong bioreducing, antifungal and antibiofilm activities and optimize the particular activity for practical applications. This provides a cost-effective, ecofriendly, non-toxic and effective source for medicinal and synthetic applications.

*Molecules* **2020**, *25*, 3478

**Supplementary Materials:** The following are available online, Figure S1 Screening strategy for exploring bioactive fraction of *W. coagulans*; Table S1 Different solvent fractionation of *Withania coagulans*.

**Author Contributions:** Conceptualization, M.H., S.G.H. and X.S.; methodology, S.Z.; software, S.G.H.; validation, T.M., F.L. and F.I.; formal analysis, S.Z.; investigation, A.Z.; resources, T.T.; data curation, S.G.H.; writing—original draft preparation, M.H.; writing—review and editing, M.H.; visualization, X.S.; supervision, M.H.; project administration, X.S.; funding acquisition, X.S.; Experimentation, I.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no special external funding.

**Acknowledgments:** The authors express their gratitude for the financial support from the Provincial Education Department Project (Natural Science, 2017KZDXM045), the Agriculture and Rural Department Project of Guangdong Province, the Guangzhou Foreign Cooperation Project (201907010033), the Graduate Technology Innovation Fund (KJCX2019004), and the Undergraduate Innovation and Entrepreneurship Training Program (S201911347028). The authors would also like to thank, The Islamia University Bahawalpur, Pakistan, National Research Program for University (NRPU) for Higher Education Commission (9458).

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

### **References**


**Sample Availability:** Samples of the compounds are available from the authors.

© 2020 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 (http://creativecommons.org/licenses/by/4.0/).

### *Article* **E**ff**ect of Naringenin and Its Derivatives on the Probing Behavior of** *Myzus persicae* **(Sulz.)**

**Katarzyna Stec 1, Joanna Kozłowska 2, Anna Wróblewska-Kurdyk 1, Bo ˙zena Kordan 3, Mirosław Anioł <sup>2</sup> and Beata Gabry´s 1,\***


Academic Editors: George Grant and Angelo Canale Received: 28 May 2020; Accepted: 8 July 2020; Published: 13 July 2020

**Abstract:** Substances that alter insect behavior have attracted a lot of attention as potential crop protection agents. Naringenin (5,7,4- -trihydroxyflavanone) is a naturally occurring bioactive flavanone. We evaluated the influence of naringenin on aphid activities during individual phases of probing and feeding and the effect of structural modifications of naringenin on its activity towards aphids. We monitored the probing behavior of *Myzus persicae* (Sulz.) (Hemiptera: Aphididae) using the Electrical Penetration Graph (EPG) technique. The chemical modifications were the substitution of hydrogen atoms with methyl, ethyl or pentyl groups and the replacement of the carbonyl group in naringenin and its derivatives with an oxime moiety. Depending on the substituents, the activity of naringenin-derived compounds varied in potency and mode of action. Naringenin was an attractant of moderate activity, which enhanced sap ingestion. The naringenin derivative with two methyl groups—7,4- -di-*O*-methylnaringenin—was a deterrent, which hindered aphid probing in non-phloem tissues. Naringenin oxime derivatives with methyl substituents—7,4- -di-*O*-methylnaringenin oxime, 7-*O*-methylnaringenin oxime, and 5,7,4- -tri-*O*-methylnaringenin oxime—and the derivative with a pentyl substituent—7-*O*-pentylnaringenin oxime—were strong attractants which stimulated aphid probing in non-phloem tissues and the ingestion of phloem sap.

**Keywords:** electrical penetration graph; peach potato aphid; antifeedants; attractants; structureactivity relationships

### **1. Introduction**

Naringenin (5,7,4- -trihydroxyflavanone) is a natural flavonoid, most common in *Citrus* fruits, known to have bioactive effects on human health, such as antidiabetic, antidepressant, immunomodulatory, antitumor, anti-inflammatory, DNA protective, and antioxidant effects [1,2]. Various effects of naringenin on insect development and behavior were also reported; naringenin inhibited the feeding of adult Japanese beetles *Popillia japonica* (Newman) (Coleoptera: Scarabaeidae) [3], caused a reduction in larval growth and development in the common cutworm *Spodoptera litura* (Fabricius) (Lepidoptera: Noctuidae) [4], stimulated oviposition in the spotted pink ladybeetle *Coleomegilla maculata* De Geer (Coleoptera: Coccinellidae) [5], and impaired the learning abilities of the honey bee *Apis mellifera* L. (Hymenoptera: Apidae) [6].

Aphids (Hemiptera: Aphididae) are responsible for at least 2% of losses caused by insect feeding in the world's crops [7]. In addition to the removal of assimilates from plant phloem transporting vessels, aphids transfer viral diseases from infected to healthy plants. The extremely polyphagous peach potato aphid *Myzus persicae* (Sulz.) can transmit over 100 plant viruses among plants within over 40 families [8]. To our knowledge, there exists only one published study that reports the effect of naringenin on aphids. Goławska et al. [9] showed that the addition of naringenin into a sucrose–agarose diet caused an increase in the duration of the pre-reproductive period and mortality, as well as a decrease in fecundity and the intrinsic rate of natural increase in the pea aphid *Acyrthosiphon pisum* (Harris). In the same study, the authors demonstrated that high concentrations of naringenin inhibited the passive ingestion (analogous to passive ingestion of phloem sap on plants and represented by EPG waveform g-E2) of the naringenin-supplemented sucrose–agarose diet, but stimulated the active ingestion (analogous to active ingestion of xylem sap on plants and represented by EPG waveform g-G) of the diet [9].

Nowadays, aphid control relies mainly on neurotoxic insecticides. However, several aphid species, especially *M. persicae*, evolved diverse mechanisms of resistance to various insecticides [10,11]. At the same time, a global trend for the reduction in insecticide use is observed in response to environmental issues. In recent years, serious restrictions in neonicotinoid use have been established in the European Union [12]. Therefore, there is a growing demand for the replacement of traditional insecticides, at least in part, by natural product-based insect control agents. Specifically, behavior modifying substances (repellents, antifeedants, attractants, etc.), which may cause the withdrawal of the herbivore from the plant or other substrates, are searched for [13–15]. The exogenous application of xenobiotics may alter aphid response to otherwise acceptable host plants, which has been shown in studies on aphid antifeedants involving different chemical groups, including terpenoids, quassinoids, flavonoids, and cyanogenic glycosides [9,16–20]. At the same time, aphids may be attracted to other areas, such as trap crops or barrier crops, in 'push–pull' strategies [21–23]. Unfortunately, the application of natural compounds for the protection of plants is limited. The main constraints are the low content in natural sources and usually complicated structures, which make their synthesis complex and expensive. Therefore, the synthesis of natural compound analogues is one of the most promising ways leading to their practical use in insect pest population control [24]. Structural transformations of the natural molecule usually change the mode of action and potency of its activity [25,26]. The possibility of reducing aphid infestation of crop plants by naringenin and its analogues application has never been explored.

The aim of the present study was to assess the influence of naringenin on aphid activities during individual phases of probing and feeding and evaluate the effect of structural modifications of naringenin on its activity towards aphids. The chemical modifications were the substitution of hydrogen atoms with methyl, ethyl or pentyl groups and the replacement of the carbonyl group in naringenin and its derivatives with an oxime moiety. We monitored aphid probing with the Electrical Penetration Graph (EPG) technique, which visualizes the movements of aphid mouthparts within individual plant tissues. The values of parameters derived from EPG recordings are reliable and accurate indicators of aphid behavioral responses to alteration in plant suitability due to exogenous application of xenobiotics [19–24].

### **2. Results**

The typical behavior of *M. persicae* on control untreated plants consisted of non-probing (11% time of the 8 h experiment), probing in non-phloem tissues (34%), and probing in phloem tissues (55%). Sap ingestion occupied 95% of the phloem phase. Aphid probing activities were divided into 19.8 (±10.0) probes on average, and these probes were approximately 0.6 (±0.6) hours long. Nearly 10% of these probes contained a phloem phase (Table 1). *M. persicae* needed approximately 2.0 (±1.3) hours to reach phloem vessels and commence sap ingestion. In that time, 25% were non-probing activities. Probing activities in non-phloem tissues before the first phloem phase were divided into 12.7 (±8.8)

events (probes) (Table 2). The phloem phase in *M. persicae* consisted of 4.4 (±2.8) separate periods and almost all of these phloem phases included sap ingestion. Of these sap ingestion periods, 60% were longer than 10 min and were 2.1 (±2.2) hours long on average. The first contact with sieve elements (the first phloem phase) was 1.7 (±2.4) hours long (Table 3).

*M. persicae* probing behavior on plants treated with naringenin and its derivatives (**2**–**17**) (Figures 1 and 2) was significantly different from the aphid behavior on the control plants. Non-probing activities were significantly reduced on plants treated with (**8**) and (**17**). The total duration of the non-phloem phase was significantly longer on (**2**)- and (**12**)-treated plants, and the phloem phase was significantly shorter on (**2**)-treated plants than on the control plants. The total duration of sap ingestion was longer on (**13**)-treated plants than on the control. The number of probes was lower, and the probes were longer on (**3**)-, (**4**)-, (**8**)-, (**13**)-, (**15**)-, and (**16**)-treated plants (Table 1). During the pre-phloem phase, i.e., the period before the first phloem phase occurred, the total duration of non-probing was longer on (**7**)-treated plants but shorter on (**3**)-, (**8**)-, and (**13**)-treated plants than on the control. The duration of probing in non-phloem tissues, the total time to the first phloem phase and the first phloem sap ingestion phase from the onset of probing were shorter on (**13**)-treated plants than on the control. The total number of probes and the number of short probes before the first phloem phase were lower on (**3**)-, (**4**)-, (**8**)-, (**13**)-, (**15**)-, and (**16**)-treated plants (Table 2). The number of phloem phases and the number of sap ingestion phases were lower on naringenin, (**3**)-, (**7**)-, (**8**)-, (**10**)-, (**15**)-, and (**17**)-treated plants than on the control. On these plants, the first phloem phases were significantly longer than on the control plants. The mean duration of phloem sap ingestion phase was longer on naringenin-, (**8**)-, and (**17**)-treated plants than on the control (Table 3).

The comparison within the group of aphids on treated plants showed that the highest number of probes occurred in aphids on (**7**)-treated plants (16.7 ± 9.7 probes) in contrast to aphids on (**3**)- and (**16**)-treated plants (4.4 ± 4.2 and 5.2 ± 7.3 probes, respectively), the durations of periods before the first phloem phase and phloem sap ingestion phase were longest in aphids on (**7**)-treated plants (2.8±1.4 h) in contrast to aphids on (**13**)-treated plants (1.0 ± 0.8 h) (Tables 1 and 2). The mean duration of the phloem phase was the shortest in aphids on (**2**)-treated plants (1.1 ± 2.0 h) in contrast to aphids on (**15**)-treated plants (5.0 ± 2.7 h) (Table 3).

**Figure 1.** Naringenin and its derivatives. Naringenin: R1=H, R2=H, R3=H **(1)**; 5,7,4- -tri-*O*-methylnaringenin: R1=CH3, R2=CH3, R3=CH3 **(2)**; 5,7,4- -tri-*O*-ethylnaringenin: R1=CH3CH2, R2=CH3CH2, R3=CH3CH2 **(3)**; 7-*O*-ethylnaringenin: R1=CH3CH2, R2=H, R3=H **(4)**; 7-*O*-pentylnaringenin: R1=CH3(CH2)4, R2=H, R3=H **(5)**; 7,4- -di-*O*-ethylnaringenin: R1=CH3CH2, R2=H, R3=CH3CH2 **(6)**; 7,4- -di-*O*methylnaringenin: R1=CH3, R2=H, R3=CH3 **(7)**; 7-*O*-methylnaringenin: R1=CH3, R2=H, R3=H **(8)**.


2)-treated plants (means ± SD). The mean

non-parametric

 tests, in which all individual data were

 stylet movements;

 G—xylem sap ingestion.

**Table 1.** General aspects of *Myzus persicae* probing behavior on naringenin and naringenin derivatives (**1**–**17**, Figures 1 and

and SD values given are a

included;

*n*—number

 of replications;

C—pathway;

E1—phloem

 salivation;

E2—phloem

 sap ingestion; F—derailed

representation

 of

non-Gaussian

 data, but the statistical analysis was done by


non-parametric

 tests,

**Table 2.** *Myzus persicae* behavior in non-phloem treated plants (means ± SD); the mean and SD values given are a

 tissues prior to the first phloem phase during probing on naringenin and naringenin derivatives (**1**–**17**, Figures 1 and 2)-

representation

 of

non-Gaussian

 data, but the statistical analysis was done by


**Table 3.** *Myzus persicae* (means ± SD); the mean and SD values given are a

behavior associated with probing in sieve elements on naringenin

representation

 of

non-Gaussian

 data, but the statistical analysis was done by

 and naringenin

 derivatives

 (**1**–**17**, Figures 1 and

non-parametric

 tests, in which all

2)-treated plants

**Figure 2.** Naringenin oxime and its derivatives. Naringenin oxime: R1=H, R2=H, R3=H **(9)**; 5,7,4- -tri-*O*-methylnaringenin oxime: R1=CH3, R2=CH3, R3=CH3 **(10)**; 5,7,4- -tri-*O*-ethylnaringenin oxime: R1=CH3CH2, R2=CH3CH2, R3=CH3CH2 **(11)**; 7-*O*-ethylnaringenin oxime: R1=CH3CH2, R2=H, R3=H **(12)**; 7-*O*-pentylnaringenin oxime: R1=CH3(CH2)4, R2=H, R3=H **(13)**; 7,4- -di-*O*-ethylnaringenin oxime: R1=CH3CH2, R2=H, R3=CH3CH2 **(14)**; 7,4- -di-*O*-methylnaringenin oxime: R1=CH3, R2=H, R3=CH3 **(15)**; 7-*O*-methylnaringenin oxime: R1=CH3, R2=H, R3=H **(16)**; 7,4- -di-*O*-pentylnaringenin oxime: R1=CH3(CH2)4, R2=H, R3=CH3(CH2)4 **(17)**.

### **3. Discussion**

The pre-phloem and phloem phases of aphid probing in plant tissues are two crucial steps in the chemosensory-based host plant selection and host plant acceptance processes. Allelochemicals are the main cues used by aphids for host plant selection during either the pathway or phloem phase [16]. The long duration of probing time in non-phloem tissues as compared to total penetration time, the relatively long time to the 1st phloem phase within a probe, and a failure in finding sieve elements may be interpreted as pre-ingestive effects of antifeedants that restrain aphid probing at the level of non-phloem tissues [27]. In contrast, the reduction in the number of probes and the elongation of these probes indicates the attractant character of chemical factors. The long total and mean durations of phloem sap ingestion may point to the ingestive mode of feeding stimulatory activity [28,29]. The interpretation of aphid behavior in response to the chemical properties of plant tissues is based on studies on plant resistance mechanisms. Aphids feeding on susceptible plant genotypes have a significantly greater duration of sieve element phase than when feeding on resistant genotypes and the time taken to reach the first sieve element phase in resistant genotypes is significantly greater than in susceptible genotypes [30–33].

Based on the comparison of the EPG-monitored *M. persicae* behavior on naringenin and naringenin derivatives-treated plants to control and the overall trends for each compound, it is possible to group the studied compounds according to their potential to modify aphid probing activities: (i) strong attractants (**8**), (**13**), (**15**), and (**16**), which stimulated aphid activities in the non-phloem as well as in the phloem tissues. In comparison to control untreated plants, on treated plants, aphids rarely withdraw stylets from plant tissues (**8, 13, 15, 16**) and the non-probing time was significantly reduced (**8, 13**), which caused a significant reduction in time to reach phloem vessels from the onset of probing (**13**), the individual phloem sap ingestion periods were long and rarely interrupted (**8**, **15**, **16**), and the total duration of sap ingestion was longer (**13**); (ii) moderate attractants naringenin, (**3**), (**10**), and (**17**)—naringenin, (**10**), and (**17**) had no effect on aphid behavior during the pre-phloem phase but encouraged sap ingestion, and (**3**) caused a slight reduction in the number of phloem phases as well as the number of probes and a decrease in the duration of non-probing before the phloem phase; (iii) weak attractant (**4**) caused a slight reduction in the number and a slight increase in the duration of probes; (iv) weak deterrent (**7**) caused aphid restlessness by hindering pre-phloem pathway activities, which was manifested in the frequent withdrawals of stylets from plant tissues, an increased time of

non-probing, and, in consequence, a delay in reaching phloem vessels; (v) inactive compounds (**2**), (**5**), (**6**), (**9**), (**11**), (**12**), (**14**). In comparison to control, aphid behavior was not altered on plants treated with these compounds.

The biological activity of a given compound is species-specific and depends on its structural characteristics. Variations, such as incorporation of functional groups, epoxidation, or lactonization, can produce radical changes in activity [29]. In our previous studies, we determined that chemical modifications of naturally occurring terpenoids, e.g., incorporation of functional groups, epoxidation or lactonization, evoked significant changes in their activity profiles. We have established that the potency and persistence of behavioral effects on aphid probing of piperitone-, β-damascone-, xanthohumol-, isoxanthohumol-, and *cis*-jasmone-derived compounds depended on their substituents. Certain modifications caused shifts from attractant to deterrent properties, or vice versa [24–26,29]. In the present study, we also revealed specific structure–activity relationships. Naringenin appeared to be an attractant of moderate activity. Three ethyl groups incorporated at positions 5,7, and 4' in (**3**) did not significantly alter the naringenin activity. The compound with one methyl group in the position 7 (**4**) was a weak attractant. However, the compound with two methyl groups in the positions 7 and 4' (**7**) was a weak deterrent. The incorporation of three methyl groups at positions 5,7, and 4' (**2**), two ethyl groups at positions 7, and 4' (**6**), or one pentyl group at position 7 (**5**) caused a loss of naringenin activity towards *M. persicae*. Naringenin oxime (**9**) was inactive towards *M. persicae*. However, the substitution of hydrogen atoms with methyl, ethyl or pentyl groups, in addition to oxime moiety, caused a significant rise in the activity of some of the derived compounds. All naringenin oxime (**9**) derivatives with methyl substituents at positions 7 (**15**), 7 and 4' (**16**), and 5,7,4' (**10**) and the derivative with a pentyl substituent at position 7 (**13**) appeared strong attractants. Pentyl groups at positions 7 and 4' made the compound (**17**) a weak attractant. The ethyl group substituents did not improve the activity of naringenin oxime (**9**); all derivatives with one (**12**), two (**14**), or three (**11**) ethyl substituents remained inactive towards *M. persicae*.

The results of the experiments in the present work illustrate two major aspects of the biological activity of naringenin and its derivatives that depend on their substituents: (1) the variation in the potency of the behavioral effect and (2) a switch from attractant to deterrent properties. In summary, the most effective transformations of the naringenin molecule were the substitutions of hydrogen atom(s) in hydroxyl group(s) with methyl or pentyl group(s) in combination with the replacement of the carbonyl group with an oxime moiety. The behavioral effects of these transformations were manifested mainly in the stimulation of probing in non-phloem tissues as well as the ingestion of phloem sap.

The results of the present study could be applied towards modifying aphid attraction or deterrence to plants in the field using genetic modification or topical application of naringenin or naringenin-derived analogues [1,34]. This is especially important in the context of virus transmission. Aphids may acquire and inoculate viruses during various stages of plant penetration with sucking–piercing mouthparts. During brief intracellular probes in the epidermis and parenchyma (mesophyll in leaves) that precede feeding in phloem vessels, aphids may transmit nonpersistent and semi-persistent viruses. When aphid stylets reach sieve elements, persistent viruses may be transmitted [35–38]. It is crucial then, to deter aphid probing or at least prevent feeding, to protect plants from pathogen infection and limit the virus spread within the field crops. Besides the direct negative effect on aphid feeding, a deterrent that impedes activities during pre-phloem and phloem stylet penetration should also prevent the transmission of non-persistent and persistent viruses, respectively. Considering the activities of naringenin and its derivatives revealed here, the strong attractants (**8**), (**10**), (**13**), (**15**), and (**16**) have the highest potential for practical applications in 'push–pull' strategies. As the probing and feeding stimulants, they can be applied topically to any species of barrier plant to pull *M. persicae* out of the protected crop. By making barrier plants more attractive to aphids, virus spread within the crop may be reduced [22]. In addition, the weak deterrent (**7**) that makes aphids restless can be applied on the crop plant to push *M. persicae* out of the crop plant stand.

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

### *4.1. Naringenin and Naringenin Derivatives*

Naringenin (5,7,4- -trihydroxyflavanone) was purchased from SIGMA (W530098). The naringenin derivatives (**2–17**) were prepared as described previously by Kozłowska et al. [39]. Briefly, monoand di-*O*-alkyl compounds (**4–8**) were obtained by stirring anhydrous potassium carbonate and significant excess of appropriate alkyl iodide in the solution of naringenin in anhydrous acetone at room temperature for 24–96 h. After solvent evaporation and washing with a saturated brine, the products were extracted with diethyl ether, dried, concentrated and separated by column chromatography. Tri-*O*-alkyl derivatives of naringenin (**2**, **3**) were obtained similarly to the method mentioned above, but dimethylformamide was used instead of acetone; after stirring for 7–24 h at room temperature, the reaction mixtures were neutralized with 1 M HCl and extracted with methylene chloride. The reaction yields were in the range of 20–72%.

Syntheses of oximes (**9**–**17**) were performed by stirring the *O*-alkyl derivatives of naringenin (1.0 eq.) (**2**–**8**), hydroxylamine hydrochloride (1.5 eq.), and anhydrous sodium acetate (1.5 eq.) in anhydrous ethanol at 40–50 ◦C. The reaction mixtures were poured into ice water. The precipitated crystals were collected, dried in a vacuum, and purified by column chromatography. The reaction yields were in the range of 81–99%.

The purity of obtained compounds was monitored using thin layer chromatography, highperformance liquid chromatography and proton nuclear magnetic resonance.

### *4.2. Aphid and Plant Cultures*

Laboratory culture of the peach potato aphid *Myzus persicae* (Sulz.), kept as a multiclonal colony (i.e., deriving from different parthenogenetically reproducing females), was maintained on *Brassica rapa* L. ssp. *pekinensis* L. in the laboratory at 21 ◦C, 65% r.h., and L16:8D photoperiod. *M. persicae* had originally been collected in the greenhouse and kept on *B. rapa* ssp. *pekinensis* in the laboratory since 2000. The plants for the EPG experiments were *B. rapa* ssp. *pekinensis* and were grown under similar laboratory conditions as aphid cultures, at 21 ◦C, 65% r.h., and L16:8D photoperiod. Plants were grown in plastic pots (0.33 L) filled with fine garden soil commonly used for greenhouse experiments. Plants were watered regularly, and no additional nutrients were supplied.

### *4.3. Preparation and Application of Compounds*

To mimic the natural environment under laboratory conditions, naringenin and its analogues were offered to aphids by application through their host plants. Preparation and application of the compounds followed the procedure described by Polonsky et al. [40], later modified by Gabry´s et al. [26]. Briefly, each compound was dissolved in 70% ethanol to obtain a 0.1% solution [40]. All compounds were applied on the adaxial and abaxial leaf surfaces by immersing one leaf of the experimental plant in the ethanolic solution of a given compound for 30 s. Leaves of similar size of the control plants were immersed in 70% ethanol that was used as a solvent for the studied compounds. There was no effect of ethanol application on aphid probing behavior and plant condition [41]. Treated and control leaves were allowed to dry for 1 h before the start of the experiment to permit the evaporation of the solvent.

### *4.4. Aphid Probing Behavior*

*Myzus persicae* probing behavior was monitored by using the Electrical Penetration Graph (EPG) technique. The EPG technique provides a unique opportunity to reveal aphid mouthparts stylets activities in plant tissues [41–43]. The parameters describing aphid behavior during probing and feeding, such as total time of probing, duration and frequency of sap ingestion periods, number of probes, etc., are good indicators of plant suitability or interference of probing by chemical or physical factors in individual plant tissues [16–20]. In this experimental setup, aphid and plant are made parts of an electric circuit, which is completed when the aphid inserts its stylets into the plant. Weak voltage

is supplied in the circuit, and all changing electric properties are recorded as EPG waveforms that have been correlated with aphid activities and stylet position in plant tissues [42,43].

In the present study, one- to seven day old adult apterous females of *M. persicae* and three week old plants with four to five fully developed leaves were used for all experiments according to the standard procedure applied in similar studies [17–20,24–27]. Aphids were attached to a golden wire electrode with conductive silver paint and starved for 1 h prior to the experiment. Probing behavior of apterous *M. persicae* was monitored for 8 h continuously with a Giga-8 DC EPG with 1 GΩ of input resistance (EPG Systems, Wageningen, The Netherlands) and Stylet+ software (www.epgsystems.eu). Each aphid was given access to a freshly prepared leaf of an unused plant, which means that each plant and each aphid were used only once. One aphid–plant combination was considered a replication. Two rounds of 8 replications (*n* = 16) were carried out for each studied substance and control. Giga-8 DC EPG allows the recording of 8 samples simultaneously. Incomplete EPG recordings, i.e., those that were prematurely ended due to the aphid falling off the plant or other incidents, were discarded from analysis. All experiments were carried out under the same conditions of temperature, relative humidity, and photoperiod, as described for the rearing of plants and aphids. The bioassays started at 10–11 a.m. MEST (Middle European Summer Time).

The following aphid behaviors related to mouthparts positions in or out of the plant tissues were distinguished: non-probing, which represents aphid stylets outside the plant tissues, pathway phase 'C', which represents the movement of aphid stylets within the epidermis and mesophyll; phase 'F', which represents unidentified ('derailed') stylet movements within apoplast; xylem phase 'G', which represents active xylem sap uptake; phloem phase consisting of watery salivation E1 and passive ingestion of phloem sap 'E2'. 'F' and 'G' occurred sporadically irrespective of a treatment, therefore, these activities were analyzed together with phase 'C' and referred to as the 'non-phloem' phase of probing.

### *4.5. Statistical Analysis*

The EPG parameters describing aphid probing behavior were calculated manually and individually for every aphid and the means and standard deviations were subsequently calculated using the EPG analysis Excel worksheet created for this study. Two comparative analyses were carried out. First, aphid behavior on control plants was compared to aphid behavior on naringenin- and naringenin derivatives-treated plants individually for each compound/treatment. This comparison (Mann–Whitney U-test) was performed and the results were interpreted to reveal the mode of action of a given compound (deterrent, attractant, neutral), which allowed grouping of the studied compounds according to their similarity in the effect they had on aphid behavior. A second comparison was carried out to determine the effect of structural modifications in the naringenin molecule on the aphid behavior modifying activity. For this purpose, aphid behavior on only the treated plants was compared. Due to failure to meet the assumptions of analysis of variance, the obtained data were analyzed by the Kruskal–Wallis test and post hoc multiple comparisons of mean ranks for all groups (Dunn's test). The Kruskal–Wallis test is a non-parametric alternative to the one-factor ANOVA test for independent measures and it is commonly used to analyze data deriving from EPG recordings of aphid probing [32]. The mean and SD values given in Tables 1–3 are a representation of non-Gaussian data, but the statistical analysis was done by non-parametric tests, in which all individual data were included. All statistical calculations were performed using StatSoft, Inc. (2014) STATISTICA (data analysis software system, version 12, www.statsoft.com).

**Author Contributions:** Conceptualization, B.G. and M.A.; Methodology, B.G., M.A. and B.K.; Formal Analysis, A.W.-K.; Investigation, K.S.; A.W.-K.; B.K.; M.A. and J.K.; Data Curation, B.G. and M.A.; Writing-Original Draft Preparation, B.G. and M.A.; Writing-Review & Editing, B.G.; Visualization, K.S. and M.A.; Supervision, B.K.; B.G. and M.A.; Funding Acquisition, M.A. and B.K. All authors have read and agreed to the published version of the manuscript.

**Funding:** Publication was supported by the National Science Centre, Grant No. 2016/21/B/NZ9/01904. The project was financially supported by Minister of Science and Higher Education in the range of the program entitled "Regional Initiative of Excellence" for the years 2019-2022, Project No. 010/RID/2018/19, amount of funding 12.000.000 PLN.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and in the decision to publish the results.

### **References**


**Sample Availability:** Not available.

© 2020 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 (http://creativecommons.org/licenses/by/4.0/).

*Article*
