Antimicrobial activities of Asian plant extracts against pathogenic and spoilage bacteria

Abstract

This study was conducted to investigate the antimicrobial effects of 300 Asian plant extracts (PEs) against pathogenic and spoilage bacteria. The antimicrobial activities were examined using agar well or agar disc diffusion, and micro-titer methods. Results revealed that PEs exhibited higher antimicrobial effects against Gram-positive bacteria compared than against Gram-negative bacteria. With few exceptions, PEs delayed the lag time (LT) of pathogenic bacteria (1.17–3.75 times). Among PEs tested, Alchornea trewioides (AT) and Erodium stephanianum (ES) were the most effective in inhibiting pathogenic and spoilage bacteria. In the study evaluating the effect on the growth inhibition in the broth, Acetobacter aceti was inhibited at 2.77 and 3.02 log CFU/mL by the combination treatment of AT+nisin and ES+nisin after storage for 7 days, respectively. Although further investigations are needed to clarify the antimicrobial mechanism of PEs, this study demonstrated that antimicrobial efficacy varied with PE types, solvents, and bacteria.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10068-022-01182-0.

Introduction

As one of the commonly used strategies to ensure food safety, various preservatives have been used to inactivate microorganisms and prolong the shelf-life of food products. However, using antibacterial agents often causes antimicrobial resistance among microorganisms that are sensitive to certain antimicrobial agents (Aminov and Mackie, 2007; HPA, 2011). For example, pathogens such as Escherichia coli, Salmonella typhi, and Staphylococcus aureus are resistant to synthetic preservatives (Akreyi et al., 2018; Hussein et al., 2018). Moreover, synthetic preservatives adversely affect human health. Yoon and Yoon (2018) demonstrated that exposure to antimicrobial agents can alter the balance of intestinal microbiota and consequently increase the risk of intestinal and immunity-related disorders. With these concerns and potential risks, studies have focused on plant-derived natural compounds as substitutes for synthetic preservatives to inactivate microorganisms (Cetin-Karaca, 2011).

Plant extracts (PEs) have demonstrated their antimicrobial potential against pathogenic bacteria (Abu-shanab et al., 2006; Cock, 2017; Ikon et al., 2020). They contain phytochemical and phenolic compounds, such as tannins, lignins, thiosulfinates, saponin, and flavonoids, which are important to inactivate pathogens (Cock, 2017; Ceruso et al., 2020; Ikon et al., 2020). Tajkarimi et al. (2010) demonstrated the antimicrobial and antifungal activities of glucosinolates and described the effectiveness of thiosulfinates in eliminating Gram-negative bacteria. The antimicrobial mechanisms of plant-derived active compounds are associated with cytoplasmic membrane disintegration, proton motive force (PMF) destabilization, electron flow, active transport, and cell content coagulation (Silva & Fernandes, 2010). Ceruso et al. (2020) evaluated the antimicrobial activity of 780 PEs against L. monocytogenes F 2365. Among them, 12 PEs caused morphological disruption, leakage of intracellular components, and loss of flagella in treated samples compared with those in normal cells. In addition, PEs inhibit quorum sensing-regulated virulence traits. For instance, 0.50 mg/mL ethanol extract from Syzygium jambos has an anti-quorum sensing activity against Pseudomonas aeruginosa PAO1 (Rajukumari et al., 2018). Furthermore, plant-derived antimicrobials are used in food industries. Tajkarimi et al. (2010) reviewed the application of plant compounds as natural preservatives of meat products, fish, vegetables and fruits, rice, dairy products, and animal feed. L. monocytogenes, Aeromonas hydrophila, and spoilage bacteria in meat products are susceptible to eugenol and coriander, clove, oregano, and thyme oil (Burt, 2004). Abdalla et al. (2007) showed that mango seed extract effectively inactivates E. coli in pasteurized cow milk.

Although antimicrobial efficacy has been widely, many PEs remain unexplored. Therefore, this study was performed to screen the antimicrobial activities of 300 Asian PEs against pathogenic and spoilage bacteria and confirm their possibility as natural food preservatives. A predictive model was applied to describe the growth characteristic of foodborne pathogens in the presence of PEs. Finally, this study evaluated the combined effect of PEs with antimicrobial substances in laboratory media.

Materials and methods

Bacterial strains

Several food spoilage and pathogenic bacteria were tested in this study. Six Gram-positive bacteria (Bacillus cereus ATCC 10876, Lactobacillus plantarum ATCC 14917, Listeria monocytogenes ATCC 7644, Pediococcus acidilactisi ATCC 33314, Staphylococcus aureus ATCC 6538 and 49444, and Streptococcus thermophilus ATCC 19258), seven Gram-negative bacteria (Acetobacter aceti KCCM 12634, Enterobacter sakazakii ATCC 29004, Escherichia coli O157:H7 ATCC 43895, Klebsiella pneumoniae ATCC 13883, Pseudomonas aeruginosa ATCC 10145, Salmonella Typhimurium ATCC 19585, and Vibrio parahaemolyticus ATCC 43996), and Saccharomyces cerevisiae EC 1118 were obtained from the bacterial culture collection of Chung-Ang University (Anseong-si, Korea). Their stock cultures were stored in 1.0 mL of tryptic soy broth (TSB; Difco Laboratories, Detroit, MI, USA) with 50% glycerol at − 80 °C. Gram-positive and Gram-negative strains were activated by cultivation in TSB, and lactic acid bacteria were activated by cultivation in lactobacilli MRS broth (Difco Laboratories, Becton Dickinson, Sparks, MD, USA) at 37 °C overnight.

Preparation of PEs

A total of 300 PEs were tested in this study (Table S1). They were obtained and identified at the International Material Research Center of Korea Research Institute of Bioscience & Biotechnology (Daejeon, Korea). They were chosen on the basis of their treatment efficacy against tuberculosis and pneumonia.

Commercially available Alchornea trewioides leaves and branch (Yedeognamu, Jherb, Korea), Lagerstroomia floribunda branch (Crape myrtle, Korea), Erodium stephanianum leaves and branch (Geranium, Seohyeonyakcho, Korea), and Ilex umbellulata tea (Mate-cha in Argentina, E-le, Korea) were used to examine their antimicrobial activity depending on extraction solvent. Each sample (25 g) was shaken vigorously in a blender with 250 mL of 80% ethanol or methanol for 24 h and boiled in 250 mL of distilled water for 10 min. After decantation, the obtained extract was filtered through Whatman no. 1 filter paper (GE Healthcare Life Science, UK). Finally, the used solvents were removed under reduced pressure in a rotary evaporator (Eyela cca-1110, Eyela, Japan), and the concentrated extracts were dissolved in each solvent (ethanol, methanol, or water) prior to antimicrobial assays.

Agar well diffusion assay

Agar well diffusion assay (Balouiri et al., 2016) was conducted with some modifications to screen 300 Asian plant extracts and investigate the antimicrobial activity of eight selected PEs: Alchornea trewioides (Benth.) Mull. Arg., Bischofia polycarpa (H. Lev.) Airy shaw, Erodium stephanianum Willd., Ilex umbellulata (Wall.) Loes., Lagerstroemia cochinchinensis Pierre, Lagerstroemia floribunda Jack, Phyllanthus urinaria L., and Syzygium jambos (L.) Alston. A 100 µL fresh bacterial culture containing 7–8 log CFU/mL taken in 10 mL of soft agar was poured on agar plates. In each plate, 4 × 4 holes (2 mm diameter) were made using a sterile borer and filled with 3 µL of sample PEs. The plates were then incubated at 37 °C overnight. Their antimicrobial activity was evaluated by estimating the diameter of the inhibition zone.

Agar disc diffusion assay

An agar disk diffusion assay was used to examine the antimicrobial efficacy of different solvent extracts of the selected plants (Salie et al., 1996). A 100 µL fresh bacterial culture containing 7–8 log CFU/mL taken in 10 mL of soft agar was poured on agar plates and allowed to solidify. Afterward, a filter paper disc (6 mm in diameter) containing the test compound was placed on the agar surface. The plates were incubated at 37 °C overnight. The antimicrobial activity was determined by estimating the diameter of inhibition zones.

Micro-titer assay

A micro-titer assay was performed in accordance with the method of Khurram et al. (2012) to determine the minimum inhibitory concentration (MIC) and minimum bacterial concentration (MBC). Each 100 µL of 1:1 (v/v) TSB (2×) and PE solution were mixed in the first row of 96-well plates (SPL Gyeonggi-do, Korea) to obtain final concentrations of 0.195%. Twofold serial dilutions were then performed using 100 µL of TSB. Then, each well was inoculated with 10 µL of the cultured strains of six pathogens and incubated at 37 °C for 24 h. After incubation, the optical densities (OD) of the cultures at 595 nm were measured. The MIC and MBC were determined at the lowest concentration based on the turbidity of two subjects at the same concentration at which optical densities at 595 nm were less than 0.5 and 0.1, respectively (Al-Bayati & Al-Jarjry, 2007).

Predictive modeling

The modified Gompertz equation (Eq. 1; Gibson et al., 1988) was used to develop the primary model of the growth of foodborne pathogens under different treatments. Growth parameters were estimated by fitting the objected data to the primary models by using GraphPad Prism version 4.0 (GraphPad Software, San Diego, CA, USA).

$$\text{Y}=\text{N}0+\text{C}·\text{exp}[-\text{exp}\{(2.718·\mu /\text{C})·(\text{lag}-\text{X})+1\}],$$ 1

where Y is the log-transformed cell number (log CFU/g), N0 is the log initial number of cells (log CFU/g), C is the difference between the initial and final cell numbers, X is the incubation time (h), μ is the maximum growth rate (GR/h), and lag is the lag time (LT) before growth (h). All samples were assessed in triplicate.

Effects of the combination of PEs and antimicrobial substances

The effects of PEs combined with nine antimicrobial substances against A. aceti, S. aureus, and V. parahaemolyticus were examined using agar well diffusion assay. Two kinds of PEs and nine antimicrobial substances (three essential oils, four organic acids, EDTA, and nisin) were used in the combination assay. The MICs determined by the time-kill test (NCCLS, 1999) were as follows: 0.3125 µL/mL A. trewioides (AT), 0.625 µL/mL E. stephanianum (ES), 1.25% cavacrol (Sigma Aldrich, Louis, MO, USA), 2.5% eugenol (Wako Pure Chemical Industries, Ltd., Japan), 1.25% thymol (Samchun Pure Chemical Co., Ltd.), 1.25% acetic acid (Duksan Pure Chemical Co., Ansan-si, Korea), 5% ascorbic acid (Kanto Chemical Co. Inc., Tokyo, Japan), 1.25% caprylic acid (Sigma Aldrich, Louis, MO, USA), 1.25% lactic acid (Duksan Pure Chemical Co., Ansan-si, Korea), 0.28125% EDTA (Sigma Aldrich, Louis, MO, USA), and 0.625% nisin (Sigma Aldrich, Louis, MO, USA).

The growth inhibition of A. aceti and S. cerevisiae was examined in single or combination treatments of PEs (AT and ES) with nisin. This test was performed in mannitol broth (0.5% yeast extract, 0.3% peptone, and 2.5% mannitol) and yeast extract-glucose broth (1% yeast extract and 1% glucose) by using 14 tubes containing A. aceti and S. cerevisiae, respectively. Thirteen tubes were treated with single (4 MIC, 2 MIC, and 1 MIC) or combined (2 MIC × 2 MIC and 1 MIC × 1 MIC) treatments of PEs and nisin, and the remainder was set as the control. All the treated tubes were incubated at 30 °C for 7 days. Then, each culture of A. aceti and S. cerevisiae was enumerated onto yeast extract peptone dextrose agar (YPD; 1% yeast extract, 1% glucose, and 2% peptone) and potato dextrose agar (PDA; Difco Laboratories, Detroit, MI, USA), respectively.

Statistical analysis

The combination experiments of PEs and nisin were repeated thrice with duplicate plates, and the average of the duplicate plate counts from three replications were converted to log colony-forming units per milliliter. Data were processed with the Statistical Analysis System (version 9.1; SAS Institute, Cary, NC, USA) or Excel (Microsoft Office XP; Microsoft, Redmond, WA, USA) via analysis of variance (ANOVA) and Duncan’s multiple range tests or Students t-test to determine whether significant differences (P ≤ 0.05) were observed in the mean of the treatment groups.

Results and discussion

Antimicrobial activity of PEs

A total of 300 PEs were screened to determine their antimicrobial activities against E. coli and S. aureus via the agar well diffusion assay. As a result, 8 out of the 300 PEs (2.67%) showed the antimicrobial effect against both pathogens (data not shown). The antimicrobial effects of the eight PEs against various pathogens were evaluated via the screening assays (Table 1), and Gram-positive bacteria were significantly affected by PEs. B. cereus was the most inhibited by the PEs except ES methanol extract, and all inhibition zones were more than 6.5 mm. Among the Gram-negative bacteria, P. aeruginosa and S. Typhimurium were susceptible to PEs. The most resistant pathogen was V. parahaemolyticus producing inhibition zones only in three PEs (ES, I. umbellulata, and S. jambos). The extracts of S. jambos and I. umbellulata displayed inhibitory effects against all tested pathogenic bacteria.

Table 1.

Antimicrobial activities of eight PEs against pathogens by the agar well diffusion assay

PEs	Gram-positive bacteria		Gram-negative bacteria
	B. cereus	S. aureus	E. coli O157:H7	P. aeruginosa	S. Typhimurium	V. parahaemolyticus
A. trewioides	+++	+++	+	+++	+	−
B. polycarpa	+++	++	++	+	++	−
E. stephanianum	−	++	−	+	+	++
I. umbellulata	+++	++	++	++	++	+
L. cochinchinensis	+++	+++	++	++	++	−
L. floribunda J.	+++	++	+	++	+	−
P. urinaria	+++	+++	−	+++	+	−
S. jambos	+++	+	++	+	++	+

− not detected; + low inhibition (≤ 5.5 mm);++ medium inhibition (> 5.5 mm but ≤ 6.5 mm); + ++ high inhibition (> 6.5 mm)

The antimicrobial activities of PEs were more effective against Gram-positive bacteria than against Gram-negative bacteria (Table 1). These results were consistent with previous studies (Cock, 2017; Costa et al., 2008; Derbal & Niar, 2019; Moreno et al., 2006) possibly because of differences in cell structures between these bacteria. Gram-positive bacteria only have a peptidoglycan layer, while lipopolysaccharides (LPS) in the outer membrane of Gram-negative bacteria inhibit the permeability of hydrophobic compounds as a preventive barrier (Derbal & Niar, 2019; Tajkarimi et al., 2010). Furthermore, phenolic compounds affect membrane permeability by inactivating cellular enzymes and alter cellular integrity, leading to cell death (Moreno et al., 2006).

MICs and MBCs of PEs

MICs and MBCs are important parameters to confirm the antimicrobial potential. Therefore, Table 2 shows the MICs and MBCs of the eight PEs for Gram-positive and Gram-negative bacteria. Plant-based antimicrobials such as herbs and spice essential oils have been effective against pathogens at concentrations in the range of 500–1000 μg/mL (Tajkarimi et al., 2010), while the PEs tested in this study exhibited various activities at concentrations in the range of 67.9–16,720 μg/mL. The methanol extracts of AT and P. urinaria exhibited significant antimicrobial efficacy against B. cereus. The MICs of AT and P. urinaria were 95.4 and 112.0 μg/mL, respectively. However, the MICs of other PEs could not be defined. The overall MICs of S. aureus were higher than those of B. cereus. The I. umbellulata extract indicated the lowest MIC (2930.0 μg/mL) and followed by AT (3052.5 μg/mL) and P. urinaria (3582.5 μg/mL). As for Gram-negative bacteria, the L. cochinchinensis extract was the most effective at low concentrations except for V. parahaemolyticus. The MICs were 225.9, 113.0, and 225.9 μg/mL for E. coli O157:H7, P. aeruginosa, and S. Typhimurium, respectively. Conversely, the MICs and MBCs could not be measured against V. parahaemolyticus. Interestingly, V. parahaemolyticus was the most susceptible to ES, and the MIC was 67.9 μg/mL.

Table 2.

MIC and MBC (μg/mL) of eight PEs against pathogens

	PEs	Gram-positive bacteria		Gram-negative bacteria
		B. cereus	S. aureus	E. coli O157:H7	P. aeruginosa	S. Typhimurium	V. parahaemolyticus
MIC	A. trewioides	95.4	3052.5	–	> 12,210.0	–	–
	B. polycarpa	–	4157.5	2078.8	–	1039.4	–
	E. stephanianum	–	> 8690.0	–	–	–	67.9
	I. umbellulata	–	2930.0	> 16,720.0	> 16,720.0	2930.0	–
	L. cochinchinensis	–	> 14,460.0	225.9	113.0	225.9	–
	L. floribunda J	–	> 12,910.0	–	> 12,910.0	–	–
	P. urinaria	112.0	3582.5	–	223.9	–	–
	S. jambos	–	> 8830.0	551.9	–	1103.8	–
MBC	A. trewioides	> 12,210.0	> 12,210.0	–	> 12,210.0	–	–
	B. polycarpa	–	> 8315.0	> 8315.0	–	> 8315.0	–
	E. stephanianum	–	> 8690.0	–	–	–	> 8690.0
	I. umbellulata	–	> 16,720.0	> 16,720.0	> 16,720.0	> 16,720.0	–
	L. cochinchinensis	–	> 14,460.0	> 14,460.0	> 14,460.0	> 14,460.0	–
	L. floribunda J.	> 12,910.0	> 12,910.0	–	> 12,910.0	–	–
	P. urinaria	> 14,330.0	> 14,330.0	–	> 14,330.0	–	–
	S. jambos	–	> 8830.0	> 8830.0	–	> 8830.0	–

– not tested

Because the PEs used in this study were rarely explored, experimental information is limited. Therefore, the MICs and MBCs of most PEs used in this study have not been reported, and this study presented helpful experimental data for the first time. Nevertheless, some similar results showed that Lagerstroemia species (Family: Lythraceae) inhibits various microorganisms. The antimicrobial activities of the methanol extract of L. indica were tested against S. aureus, S. enteritidis, E. coli, L. monocytogenes, and C. albicans (Diab et al., 2012). The active compound 4-methoxy apigenin-8-C-β-D-glucopyranoside was purified from L. indica leaves with minimum lethal concentrations (MLCs) of 32 μg/mL for C. albicans and 16 μg/mL for S. aureus, S. enteritidis, E. coli, and L. monocytogenes. Although L. cochinchinensis and L. floribunda J. have no antimicrobial effect against V. parahaemolyticus (Tables 1, 2), the methanol extract of L. speciosa has a well-defined zone of inhibition (9.0 ± 1.56 mm) against V. parahaemolyticus (Sharmin et al., 2018). Discrepancy is related to various factors, such as microbial strains, plant parts, PE concentrations, active compounds, and extract methods (Cock, 2017; Ikon et al., 2020; Moreno et al., 2006). Tajkarimi et al. (2010) described the antimicrobial activities of natural compounds are not consistent with bacterial characteristics though Gram-positive bacteria are more generally sensitive than Gram-negative bacteria. The methanol extract of Lagerstroemia exhibited antimicrobial activities against two fungi, five Gram-positive bacteria, and seven Gram-negative bacteria except S. typhi. However, the hexane soluble fraction of the same plant is effective only in B. megaterium (Sharmin et al., 2018). Hyun et al. (2015) showed that the susceptibilities of essential oils vary depending on the type of pathogens. For instance, lemongrass-1 showed the lowest MIC against B. cereus (313 mg/mL), followed by E. coli (625 mg/mL) and S. aureus (1,250 mg/mL). Conversely, the lowest MIC was detected in lemongrass-2 against E. coli (1250 mg/mL), and the values were the same in B. cereus, S. aureus, and S. Typhimurium as 2500 mg/mL. Since the antibacterial activities of plant-based antimicrobials did not elicit the same effect in all cases, the mechanism of PEs used in this study should be discussed to clarify their antimicrobial effectiveness.

Predictive modeling

To our knowledge, this study is the first to provide a scientific predictive estimate of the microbial growth kinetics treated with PEs, and the values of GR and LT were determined using the Gompertz model (Table S2). The OD of the samples treated with L. cochinchinensis and L. floribunda J. extracts was not determined because of their specific colors. The GR of pathogens differed depending on the PEs though the values of LT had a similar tendency to previous results. The LT values of Gram-positive bacteria were more affected than those of Gram-negative bacteria, indicating that bacterial growth was delayed by the PEs. Among the eight pathogens, B. cereus had the most delayed LT by PEs, excluding ES. The methanol extract of P. urinaria increased the LT values of B. cereus from 2.760 to 10.350 (3.75-fold). For S. aureus, PEs delayed the LT from 1.74-fold to 2.40-fold, and the most effective PE was S. jambos ethanol extract.

For Gram-negative bacteria, the LTs of all strains were increased from 1.17-fold to 2.37-fold regardless of the type of PE. The LT of S. Typhimurium was the most affected by the PEs, and the methanol extract of I. umbellulata increased the LT by 2.37-fold (2.410–5.700) compared with that of the control. Then, P. aeruginosa was greatly influenced by B. polycarpa (2.04-fold), V. parahaemolyticus by ES (1.68-fold), and E. coli by S. jambos (1.63-fold).

Antimicrobial activity of selected PEs extracted by various solvents

As the extracts of AT, ES, I. umbellulata, and L. floribunda J. revealed a high antimicrobial activity, the efficacy of PEs depending on various solvents was evaluated via the agar disc diffusion assay (Table 3). The ethanol or methanol extracts of PEs showed high inhibitory effects against the tested bacteria compared with those of the aqueous extract. The size of the clear zone varied with bacteria although AT and ES extract produced a larger average appearance of clear zones than other PEs did. In particular, the ethanol extracts of AT and ES were the most effective against P. aeruginosa, producing the largest clear zones of 21 and 23 mm, respectively. Conversely, I. umbellulata ethanol extract and L. floribunda J. methanol extract displayed the smallest inhibition zone (8 mm) for both B. cereus and S. aureus, respectively.

Table 3.

Inhibition zone (mm) around sterile discs (6 mm) loaded with four plant extracts against six tested bacteria

	Strains	PEs
		A. trewioides			E. stephanianum			I. umbellulata			L. floribunda J.
		AEa	EE	ME	AE	EE	ME	AE	EE	ME	AE	EE	ME
Gram- positive	B. cereus	–b	13	11	–	11	10	–	8	10	–	–	–
	S. aureus	12	11	13	12	11	9	–	–	–	–	–	8
Gram-negative	E. coli O157:H7	12	10	12	13	12	–	–	–	–	–	–	–
	P. aeruginosa	13	21	16	16	23	18	12	19	17	9	14	9
	S. Typimurium	–	10	–	15	–	14	–	–	–	–	–	–
	V. parahaemolyticus	–	13	14	–	–	–	–	–	–	–	–	–

^aAE aqueous extract, EE ethanol extract, ME methanol extract

^b– not detected

The antimicrobial efficacy of ethanol extracts is better than that of aqueous extracts. Ikon et al. (2020) demonstrated that the ethanol or methanol extract of PEs included more phytochemical constituents such as saponin, flavonoids, and polyphenols. For instance, the antibacterial activity of the ethanol extracts of four PEs (Althaea officinalis, Mentha longifolia, Melissa officinalis, and Rosa damascena) against methicillin-resistant S. aureus presents larger inhibition zones than that of their aqueous extracts (Abu-shanab et al., 2006).

Table 4 describes the MICs of ethanol extracts of AT and ES, which showed strong antimicrobial effects in previous experiments. In general, the ethanol extract of AT had lower MICs for the tested bacteria than ES, except for P. aeruginosa. AT extract significantly inhibited V. parahaemolyticus with a low MIC (1.5625 μg/mL), followed by S. aureus, B. cereus, and A. aceti (3.125 μg/mL). ES extract showed the lowest MIC for both P. aeruginosa and V. parahaemolyticus (3.125 μg/mL), followed by S. aureus, B. cereus, and A. aceti (6.25 μg/mL). However, neither of these extracts showed any antibacterial effect against L. plantarum, P. acidilactici, and S. thermophiles at the tested concentrations.

Table 4.

MIC (μg/mL) of ethanol extracts derived from A. trewioides and E. stephanianum

	Strains	A. trewioides	E. stephanianum
Gram-positive	B. cereus ATCC 10876	3.125	6.250
	L. monocytogenes ATCC7644	6.250	25.000
	L. plantarum ATCC 14917	–a	–
	P. acidilactici ATCC 33314	–	–
	S. aureus ATCC 49444	3.125	6.250
	S. thermophilus ATCC 19258	–	–
Gram-negative	A. aceti KCCM 12634	6.250	12.500
	E. coli O157:H7 ATCC 43895	6.250	12.500
	E. sakazakii ATCC 29004	12.500	3.125
	K. pneumoniae ATCC 13883	1.5625	3.125
	P. aeruginosa ATCC 10145	6.250	12.500
	S. Typhimurium ATCC 19585	12.500	12.500
	V. parahaemolyticus ATCC 27969	6.250	12.500

^a– no inhibition zone

Martínez et al. (2017) reported that Alchornea (Euphorbiaceae) contains secondary metabolites such as alkaloids, terpenes, steroids, phenolic acid, and saponins. In addition, a new phenolic acid, namely, 1-O-galloyl-6-O-vanilloyl-β-glucose, was found in the ethanol extract of AT bark (Qin et al., 2012). Costa et al. (2008) reported that the ethanol extract of A. castaneifolia exhibited inhibitory effects against S. epidermidis at 5000 μg/mL, containing bioactive compounds such as phenolic compounds, flavonoids, saponins, steroids, tannins, and triterpenes. Likewise, several studies have demonstrated that Erodium spp. induce microbial inactivation. Quave et al. (2008) reported that the ethanol extract of E. ciconium did not produce any inhibition zone against S. aureus; however, Munekata et al. (2019) reviewed that the antimicrobial activities of Erodium spp. are dependent on Erodium species, plant parts, solvents, and bacteria. The MICs of Erodium spp. varied from 0.039 mg/mL to 64 mg/mL for various microorganisms. The ethanol extract of E. glaucophyllum elicits inhibitory effects against C. albicans, E. coli, and S. aureus with MICs of 1.99, 2.50, and 3.16 mg/mL, respectively (Gohar et al., 2003).

Antimicrobial activity of the combined antimicrobial substances

The combined effects of PEs and various antimicrobial substances, including organic acids and essential oils, are shown in Table 5. The antimicrobial activities of the combination of AT and ascorbic acid (AT+ASA) were the strongest among all others and increased depending on the PE concentration. The antimicrobial effects of the combination of AT and acetic acid (AT+ACA) or lactic acid (AT+LA) were higher at 2 MIC of against S. aureus; however, they had no distinctive antimicrobial activity against V. parahaemolyticus and A. aceti. In addition, ES combined with thymol (ES+T) or eugenol (ES+E) inhibited S. aureus to a greater extent than other combination treatments. For V. parahaemolyticus, the antimicrobial properties were not enhanced in most of the treatments although AT+ASA, ES+ASA, and ES+E presented antimicrobial activities. For A. aceti, no antimicrobial effect was observed in any single treatment with ASA, EDTA, nisin, or caprylic acid (CA). However, increased antimicrobial activities were observed in the combination treatments with Pes (AT+ASA, AT+EDTA, AT/ES+nisin, and AT/ES+CA). In particular, the combinations with nisin (AT+nisin and ES+nisin) showed the highest antimicrobial effect depending on the concentration of PEs although the treatment with AT or ES alone also showed similar antimicrobial effects.

Table 5.

Effectiveness of 2 MIC, 1 MIC, and 1/2 MIC of the combination of plant extracts with antimicrobial activities

			AT/ESa	ACA	LA	ASA	EDTA	Nisin	CA	T	E	C
S. aureus		Controlb		−c	−	+	−	+	−	+	+	+
	AT	1/2 MIC	−	−	−	−	−	−	−	−	−	−
		MIC	+	−	−	++	+	−	−	−	−	−
		2 MIC	+	++	++	++	+	−	−	+	+	+
	ES	1/2 MIC	−	+	−	+	++	−	+	+	++	−
		MIC	+	−	−	−	−	−	−	++	++	+
		2 MIC	++	−	−	+	+	−	−	++	+	++
V. parahaemolyticus		Control		−	−	−	−	−	−	+	−	−
	AT	1/2 MIC	−	−	−	−	−	−	−	−	−	−
		MIC	−	+	−	++	−	−	−	−	−	−
		2 MIC	+	−	−	++	−	−	+	−	−	−
	ES	1/2 MIC	−	−	−	+	−	−	−	−	+	−
		MIC	−	−	−	++	−	−	−	−	+	−
		2 MIC	+	−	−	+++	−	−	−	−	+	−
A. aceti		Control		−	−	−	−	−	−	+	+	+
	AT	1/2 MIC	+	−	−	−	−	+	−	−	−	+
		MIC	++	−	−	+	+	++	+	++	++	++
		2 MIC	+++	−	−	++	+++	+++	+	++	++	+++
	ES	1/2 MIC	+	−	−	−	−	+	+	+	−	+
		MIC	++	−	−	−	−	++	+	+	+	+
		2 MIC	++	−	−	−	+	+++	+	+	++	+++

^aAT, A. trewioides; ES, E. stephanianum, ACA, acetic acid; LA, lactic acid; ASA, ascorbic acid; CA, caprylic acid; T, thymol; E, eugenol; C, cavacrol

^bMIC value

^c− not detected; + low inhibition (≤ 5.5 mm); ++ medium inhibition (> 5.5 mm but ≤ 6.5 mm); +++ high inhibition (> 6.5 mm)

Their inhibitory effects on A. aceti and S. cerevisiae were evaluated over 7 days to obtain more details regarding the PE and nisin combination treatments (Fig. 1). The average of the initial population of A. aceti was 5.73 ± 0.98 log CFU/ml, and the viability of A. aceti decreased as the concentrations of PEs and nisin increased by 1 MIC to 4 MIC in single treatments after 7 days of storage. At a concentration of 1 MIC, A. aceti in the single treatment of AT, ES, and nisin was not inhibited, and their levels were 7.70, 7.20, and 6.96 log CFU/mL after 7 days of storage, respectively (Fig. 1A–C). Conversely, when PE and nisin were combined at the same MIC, the antimicrobial activity increased (Fig. 1D). The reduction levels of A. aceti were 2.77 and 3.02 CFU/mL in AT+nisin and in ES+nisin after 7 days of storage, respectively. The antimicrobial efficacies of the combined treatments at 2 MIC were higher than those of single treatments at the same concentrations of PEs (Fig. 1).

Fig. 1.

Effect of combination of PEs with nisin on the growth inhibition against A. aceti (A–D) and S. cerevisiae (E–H) for 7 days. Data are expressed as mean ± standard deviation (n = 3). AT (A, E); ES (B, F); nisin (C, G); AT or ES+nisin (D and H). In single treatments (A–C and E–G): control (●); 1 MIC (△); 2 MIC (▼); 4 MIC (○). In the commination treatments (D, H); control (●); AT+nisin (○, 2 MIC × 2 MIC); AT+nisin (▼, 1 MIC × 1 MIC); ES+nisin (△, 2 MIC × 2 MIC); ES+nisin (■, 1 MIC × 1 MIC)

Conversely, the survival of S. cerevisiae was not affected by single treatments with PEs except at 4 MIC (Fig. 1E, F). Similarly, nisin had no inhibitory effect against S. cerevisiae at any of the tested concentrations (Fig. 1G). The combination of PEs with nisin slightly affected the inactivation of S. cerevisiae for 7 days regardless of the concentration (Fig. 1H). The population reduced by 1.67 log CFU/mL in AT+nisin and 1.42 log CFU/mL in ES+nisin at 2 MIC after 7 days of storage. Interestingly, the initial population of S. cerevisiae in the treatment with 4 MIC of AT or ES was significantly lower than the populations of others before storage possibly because the high concentration (4 MIC) of AT or ES affected the growth of S. cerevisiae when it was plated on PDA. These similar patterns were described in other studies investigating the antimicrobial activities of high-concentration chemicals (De Oliveira et al., 2011; Marchante et al., 2019).

As shown in Table 5 and Fig. 1, nisin was the most effective antimicrobial substance tested in this study, and synergistic effects were observed when it was combined with PEs. Similar results were also reported in previous research (Rojo-Bezares et al., 2007). In the combination of nisin and metabisulfite, the synergistic antimicrobial effects were found against AAB, LAB, and yeast. In particular, the MIC₅₀ of metabisulfite for LAB notably decreased from 200 to 25 mg/L in combination with 0.39 mg/L nisin. Several studies have evaluated plant properties as alternative preservatives for foods (Chavan & Tupe, 2014; Santos et al. 2012). The essential oils of Citrus medica and Cinnamomum zeylanicum delay the growth of spoilage microorganisms, Dekkera, Candida, Hanseniaspora, and Zygosaccharomyces at 18–20 °C or 4 °C (Mitropoulou et al., 2020). Marchante et al. (2019) assessed the antimicrobial effects of grape seed and stem aqueous extract in white wine. For AAB, grape seed extracts (0.5 g/L) showed similar inhibitory effects to those of SO₂ (50 mg/L), and higher antimicrobial activities were observed in 0.5 g/L of grape stem extracts.

Recently, consumers have preferred natural antimicrobials to artificial antibiotics because of side effects. Thus, this study screened a large collection of Asian PEs, including various plant parts, which have not extensively been studied. As the results, the ethanol extract of AT and ES exhibited high antimicrobial activity compared to other solvents and the combination with nisin considerably enhanced their efficacy on inhibiting the growth of A. aceti in laboratory broth. These results suggested the potential of PEs as natural preservatives though individual properties of PEs were not identified in this study. Hence, further studies on bioactive compounds in PEs may be needed to elucidate their antimicrobial mechanisms and development as alternative natural antimicrobial agents.

Supplementary Information

Below is the link to the electronic supplementary material.

Declarations

Conflict of interest

The authors declare no conflicts of interest.