**4. Discussion**

In this study, 38 out of 67 tested spices displayed various degrees of antibacterial activity against antibiotic-resistant strain of *S. aureus*, while only four were effective against drug-resistant strain of *S. enteritidis*. The antibacterial activity seemed to be bacteria-dependent, and Gram-positive bacteria were more susceptible to the tested spice extracts than Gram-negative bacteria, which was in accordance with many previous studies [31,32]. Different from Gram-positive bacteria, Gram-negative bacteria have an outer membrane rich in lipopolysaccharides, as well as a unique periplasmic space. The complex composition and spatial structure of lipopolysaccharides form a barrier for penetration of antimicrobial agents, besides, the presence of enzymes in periplasmic space may break down intrusive molecules, preventing the antibacterial drugs entering intracellular environment [29]. Additionally, the antibacterial activity of certain spice extracts tested in our study was also reported by previous studies. For instance, chilli, lemongrass, bay leaf, cumin, cinnamon, clove, parsley, basil, sage, thyme, rosemary, and mint, were all demonstrated to show antibacterial capacity against *S. aureus* [33–36]. However, considering the difference in extraction solvent, extraction method, and dosage of samples, it is difficult to directly compare these results with the results of our present study. More importantly,

the inhibitory effects of spice extracts on multi-drug resistant bacteria were relatively less reported. Gull et al. revealed that eight drug-resistant bacteria were inhibited by ginger extract at a concentration of 100 mg/mL, with the DIZ ranging from 11 to 15 mm [37]. Mandal et al. reported that the DIZ value obtained from ethanol extracts (20 μL, 10 mg/mL) of cinnamon, clove, and cumin against methicillin-resistant *S. aureus* was in the range of 22–27 mm, 19–23 mm, and 9–15 mm, respectively [38]. Similarly, Revati et al. found that high level gentamicin-resistant enterococci isolates were sensitive to ethanol extracts (50 μL, 100 mg/mL) of cinnamon, ginger, clove, and cumin, with the DIZ values of 31–34, 27–30, 25–26, and 19–20 mm, respectively [14]. Even though, most of the previous investigations were carried out with a limited number of antibiotic-resistant bacterial isolates as well as the tested spice samples, thus the broad antibacterial spectra of spice extracts could not be demonstrated. In addition, we did not set a positive control, such as antibiotics, mainly with two reasons. On the one hand, the antibiotic resistance of bacterial strains used in our study was determined using 11 different antibiotics (Table 2). On the other hand, we were not intended to compare the antibacterial activity of these 67 spice extracts with antibiotics, since the effects of the crude extracts were generally not comparable to pure antibiotics. Besides, in our study, we used the stock concentration of extracts at a relatively high concentration, 100 mg/mL, for the DIZ evaluation, since our samples were dissolved in DMSO, which also possessed an antibacterial effect at a relatively high concentration, such as more than 5%. To rule out the interference of DMSO during subsequent MIC and MBC assays, it was necessary to increase the stock concentration of extracts to reduce the concentration of DMSO in the final working solution of samples.

In our present study, we further chose 11 spices whose DIZ values were higher than 15 mm to verify their antibacterial effects on another ten antibiotic-resistant strains of *S. aureus*, since tested spice extracts exhibited much better antibacterial activity on Gram-positive *S. aureus* than Gram-negative *S. enteritidis*, and found that galangal, fructus galangae, cinnamon, yellow mustard seed, and rosemary overall had the best antibacterial effect, and could probably be developed into antimicrobial agents. Our study may be the first large-scale investigation on the antibacterial effect of spice hydrophilic extracts on antibiotic-resistant bacteria. Therefore, this study can give a clear comparison of the antibacterial activity of spice extracts, especially against antibiotic-resistant bacteria. To provide useful information like safety for further use of these spice extracts, HFF cells were used to evaluate the cytotoxicity of them by MTT assays. All spices except galangal, rosemary, and sage were with low toxicity with LC50 values higher than 100 μg/mL. It was worth noting that galangal, which exhibited excellent antibacterial activity among tested spices was also found to show some cytotoxicity against HFF cells in vitro, while its toxicity should also be evaluated in in vivo studies in the future before reaching the conclusion on its toxicity. Discarding crude extracts with good antimicrobial activity only based on the in vitro cytotoxic experiments should involve caution, since cytotoxic compounds might not necessarily be the same antibacterial compounds in some cases [27]; therefore, the main antibacterial and cytotoxic compounds of galangal ethanol extracts should be further isolated and identified in the future before a final conclusion can be made.

In addition to microbial contamination, lipid oxidation is another major cause of food spoilage, therefore, we also measured the antioxidant capacity of 67 spice extracts. The antioxidant activity of tested 67 spice extracts determined by both FRAP and TEAC assays were in the range of 50.3–6682 mmol Fe(II)/g DW and 17.4–3415 mmol trolox/g DW extract powder, respectively. Among them, clove showed the highest antioxidant capacity, even comparable to butylated hydroxyanisole (BHA), an antioxidant commonly applied in food industry preservation due to its excellent hydrogen-donating capacity and metal-chelation ability [39]. Additionally, the results of PCA analysis showed that the extract of clove (both fruit and flower) and cinnamon were spotlighted as potential good candidates as natural food preservatives due to their excellent antibacterial and antioxidant properties. Several other spice extracts like coriander, cinnamon, oregano, mustard, holy basil, and green pepper were also reported to be potent food preservatives [40–44]. Indeed, some studies demonstrated the potential application of clove extracts in raw chicken meat and raw pork during storage to extend shelf-life, in terms of

reducing microbes, maintaining natural color, and retarding lipid oxidation [40]. The antimicrobial and antioxidant activities of clove were mainly attributed to the presence of secondary metabolites. A study conducted by Suleiman et al. revealed that the ethanolic extract of clove flower bud appeared to be rich in flavonoids (26.8%), phenolic acid (20.8%), and tannins (4.9%) [45], whose antioxidant effects were already well-known, similar to another phytochemical screening of clove made by Upadhyaya et al. [46]. In addition, the extract of clove flower bud with stronger antimicrobial capacity was also found to exhibit higher phenolic content [47], indicating that phenolic compounds that contributed to the antioxidant activity also displayed antibacterial capacity. Moreover, some components mainly existing in volatile oil also participated in the contribution of antibacterial activity, such as eugenol, isoeugenol, eugenyl acetate, caryophyllene, and humulene. Eugenol was even classified as a substance generally regarded as safe by Food and Drug Administration (FDA). Compared with male clove (flower bud), there were limited studies on female clove (fruit). Although they were derived from the same plant, chemical components were significantly different, and the phytochemicals in clove fruit were identified as eugenol, 2-hydroxy-4, 6-dimethoxy-5-methylacetophenone, and cyclohexane, which might exert antibacterial and antioxidant effects [48].

The antimicrobial activity of spice extracts is mainly attributed to their phytochemicals. Phenolic compounds, such as phenolic acids, flavonoids, and tannins are among the most abundant and widely distributed groups of secondary metabolites in edible plants [49,50]. Moreover, phenolic compounds have been reported to be highly responsible for the antioxidant activity in spices [9], which is also agreement with our results, showing strong correlation between TPC and FRAP/ABTS values (*r* = 0.918 and 0.931, respectively, *p* < 0.01). Thus, TPC can serve as a bridge connecting the antibacterial and antioxidant activity of spice extracts. In a previous study, Shan et al. showed that there was a strong positive linear relationship among antibacterial activity, antioxidant activity, and TPC values in spices [29]. Indeed, in some spices like sage, higher antibacterial activity could be observed in spices containing higher TPC [51]. Moreover, some phenolic compounds identified in spices showed good bacterial inhibitory efficiency. Taking oregano as an example, its antibacterial activity was strongly linked to the presence of phenolic compounds like carvacrol and thymol [52]. Besides, the phenolic compounds identified in many spices like curcumin in turmeric, eugenol in cloves, thymol in thyme, and gingerol in ginger, as well as caffeic acids and ferulic acids in thyme, cinnamon, and galangal have also been demonstrated to exhibit evident antibacterial capacity [8,50,53–56]. Moreover, the number and position of phenolic hydroxyl groups are also considered to be tightly related to the toxicity towards microorganisms [6]. The antibacterial activity of these phenolic compounds involves many modes of action, such as destroying cell membrane morphology, altering membrane fatty acids, depleting proton motive force, causing reactive oxygen damage, impairing enzymatic mechanisms for energy production and metabolism, disrupting normal functionality of proteins, and inhibiting nucleic acid synthesis [6,29,57].

In our study, however, we found a significant but weak correlation of antibacterial activity with TPC and antioxidant activity, indicating that polyphenols were only partially contributed to the antibacterial activity of spice extracts. The Pearson correlation coefficient (*r* = 0.541) tested between TPC and DIZ values of *S. aureus* in our study was overall consistent with a previous study [24], reporting that the TPC of 28 pigmented edible bean coats were weakly correlated (*r* = 0.540) with DIZ values of *S. aureus*. In addition, Weeakkody et al. found a similarly poor correlation (*r<sup>2</sup>* < 0.30) between the antimicrobial activity of seven edible spice extracts and phenolic compound levels [58]. Our study and these studies suggest that in addition to polyphenols, there should be other substances responsible for the overall antibacterial activity of spice extracts. For instance, in our study, although the TPC of galanga was lower than some other spices, its antibacterial activity was highest among tested spices, indicating that other nonphenolic constituents, like 5-hydroxymethyl furfural (accounting for 59.9% in methanol extract), might have the capacity to act as antimicrobial agents [59,60]. Besides, alkaloids, such as piperine from black pepper, were also found to be effective against *Escherichia coli*, *Klebsiella*

*penumonia*, *Salmonella enterica*, and *S. aureus* [61]. Therefore, polyphenols combined with other bioactive compounds should contribute to the overall antibacterial activity of spice extracts.
