Introduction

Nowadays, studies aim to assess new edible seed oils recovered from plant sources1,2. Seed oils are mainly composed of triglycerides with other minor compounds including phenols, phytosterols, tocopherols, carotenoids, and phospholipids1. They are used in food, cosmetics, pharmaceutical, and chemical industries for different biological and technological purposes3.

Antimicrobial Resistance (AMR) is a major problem in public health, posing a potential threat to living organisms and causing death4. It happens as a consequence of antibiotics misuse without consulting healthcare and agricultural professionals5,6.

In this regard, plant extracts represent unlimited sources of bioactive chemicals that can be used as antimicrobial agents7. Oils represent an important source of natural antimicrobials allowing their implementation as natural antimicrobials in food8,9. For example, chitosan films enriched with fig seed oil alone or in combination with apricot and plum seed oils were found to protect against microbial spoilage when applied on fresh lemon and banana slices9.

Prunus armeniaca L., known as apricot, is a member of Rosaceae family10. Apricots are widely consumed and used in food manufacturing like juices and jams, which in turn produce large masses of seeds that need further exploitation11. Apricot seeds contain oil content up to 50% for each seed, making it possible to produce tons of ASO each year12. ASO is an edible oil and serves as a functional ingredient in different industries including the food sector3. Apricot seeds possess several pharmacological functions including antioxidant activity13, antitubercular activity, cardioprotective effect, and hepatoprotective effect14. In addition, seeds showed antibacterial and antifungal activities. Apricot seed extracts showed antimicrobial activity against E. coli, P. mirabilis, S. aureus, Candida albicans, C. glabrata, and C. parapsilosis13. Apricot seed essential oil showed an activity against Bacillus cereus, Enterococcus faecalis, K. pneumoniae, and Serratia marcescens15.

Vitis vinifera, known as grape, has a valuable nutritional content and important pharmaceutical properties of its derivatives such as seed and peel extracts16. The winemaking process produces about 0.2 kg of grape pomace for each kg of crushed grapes, where seeds represent about 25% of total pomace17. Grape seed is rich in cellulose, protein, bioactive compounds, and has up to 20% oil content17,18. GSO can be used for culinary purposes, such as frying food, and as an emulsifier in sauces and dressings production19. It can also be used as an ingredient in pharmaceutical, cosmetics, and biodiesel production19. GSO antibacterial activity was reported against S. aureus and E. coli16. Grape seed extracts showed antibacterial activity against S. aureus, K. pneumoniae, P. aeruginosa, and E. coli20 as well as antifungal activity21. Furthermore, an antioxidant activity, cardioprotective effect16, hepatoprotective effect22, neuroprotective effect23, and anti-inflammatory effect24 were also observed.

Phoenix dactylifera, known as date or date palm, is a nutritious fruit recording an annual increase in its production25,26. Date seed has a valuable chemical content; however, it is considered as an industrial waste that requires further exploitation26. Date seed has 5% to 13% oil content that can be an excellent ingredient in different industries due to its essential functions and edibility26,27. For example, DSO can be applied as a natural ingredient in UV protector formulas for its ability to absorb light in UV range, and as a natural colorant in butter and margarines for its yellow color28. Date seed extracts showed an important antibacterial activity against different strains including E. coli, K. pneumoniae, P. mirabilis, P. aeruginosa, S. aureus, Bacillus subtilis, and Streptococcus pyogenes29,30. In addition, DSO showed an antioxidant activity31, a chemopreventive effect32, and a photoprotective effect33.

Nigella sativa, an annual flowering plant belonging to Ranunculaceae family, produces numerous seeds commonly known as black seeds or black cumin. Black seed contains fixed oil, essential oil, protein, saponin, and alkaloid in addition to important nutritional components such as minerals, carbohydrates, fats, and vitamins34. The seed contains 30% to 44.2% fixed oil, which is extracted for several purposes such as culinary uses, food flavorings, food preservatives for confectioneries, fats stabilizers, and pharmaceutical and therapeutic applications35. Many studies were reported on the therapeutic applications of BSO. For example, antibacterial activity36, antifungal activity34, anti-inflammatory activity, anti-histaminic effect35, antioxidant activity37, antiviral effect38, anticancer activity36, gastroprotective effect39, and hepatoprotective effect were reported34.

This study focused on ASO, DSO, and GSO since these oils have limited studies on their antibacterial activity. BSO was also added for its well-known therapeutic potency36, making it of interest to test for possible synergism when combined with other oils.

Synergy improves biological properties where mixing oils can increase the components’ diversity and their structures’ complexity, which in turn provide multiple sites of actions and affect multiple biochemical processes in bacteria40. In addition, pathogens cannot develop resistance to multiple components present in the oil mixtures leading to an amplified antimicrobial activity40. The innovation of this study is screening for synergistic antibacterial activity of ASO, DSO, GSO, and BSO when they are applied in binary mixtures. Fatty acid profile and total phenol content of oils and their mixtures were evaluated as well. To our knowledge, no previous studies were done to test for possible synergism between these oils. Oils and oil mixtures with important antibacterial activity can be then tested for their ability to serve as natural antimicrobials when applied in films or nanoparticles in different food, pharmaceutical, and other manufactured products.

Materials and methods

Seed oils

Commercially available ASO, DSO, GSO, and BSO were bought from a local shop, originating from France. Oils were extracted from seeds by cold-pressing. All oils were filtered using sterile 0.22 µm pore syringe filters prior to experiments. Oil mixtures were prepared as follow: AB (ASO+BSO), AG (ASO+GSO), AD (ASO+DSO), BD (BSO+DSO), BG (BSO+GSO), and GD (GSO+DSO). All mixtures were prepared in 1:1 (v/v).

Polyphenol extraction and quantification

Polyphenols were first extracted according to41. 2 mL of hexane were added to 1 gram of each oil, then 5 mL of 80% methanol/water solution were added. The mixture was vortexed and then centrifuged (5000 g, 20 min). Aqueous phase was analyzed.

Total phenolic content was measured by Folin–Ciocalteau method according to42. 1 mL of diluted Folin solution (1:10) was added to 0.2 mL of aqueous phase extracts. Then 0.8 mL of sodium carbonate solution (7.5%) was added. Solutions were incubated at room temperature for 1 h, and then absorbance was read at 765 nm. Samples were measured in triplicates and mean value was reported. Calibration curve of gallic acid was prepared, and results were expressed as mg GAE/g of oil.

Fatty acid analysis

To determine fatty acid composition of oils and oil mixtures, fatty acid methyl esters were prepared by derivatization reaction using trimethylsulfonium hydroxide (TMSH)43. 20 μL of oil sample were dissolved in 2 mL of tert-butyl methyl ether. Then, 20 μL of this solution were removed and derivatized using 20 μL TMSH. Samples were incubated for 1 h, and then derivatized samples were measured using GC-MS (Shimadzu GCMS-QP2010 Plus)44. Helium was used as a carrier gas with split ratio 1:20 and flow rate 2.06 mL/min. Injection volume was 1 μL. Injection temperature and interface temperature were 230 °C. Scanning by MS was set at 35–250 m/z, ionization energy of 70 eV, and ion source temperature of 200 °C. Programming temperature was set as follow: 120 °C to 190 °C at 10 °C/min, 190 °C to 220 °C at 5 °C/min (hold 2 min).

Antibacterial assessment

Bacterial strains

Five bacterial strains were used in the study due to their clinical significance. Four gram-negative strains including E. coli (ATCC 25922), three clinical isolates of K. pneumoniae, P. aeruginosa, P. mirabilis, and one gram-positive strain of S. aureus clinical isolate were provided from Beirut Arab University laboratory. Fresh cultures were prepared on blood agar and MacConkey agar incubated overnight at 37 °C.

Disk diffusion for single oils

Kirby–Bauer method was used to perform agar disk diffusion for the antibacterial screening of oils alone and in mixtures. Mueller Hinton agar plates were inoculated with bacterial suspensions with load of 1.5 × 108 CFU per mL. 6 mm disks of Whatman filter paper were placed on the agar, and 20 µL of each oil were added on each disk45. Blank disks were used as a negative control. After incubating the plates at 37 °C for 16–18 h, diameters of inhibition zones (IZs) were measured. Each experiment was repeated three times. Results were reported as the mean of three trials.

Disk diffusion for oil mixtures

Same method was used to screen the synergistic antibacterial effect of mixtures. 20 µL of each mixture were added on each disk. Tests were repeated three times, and zones were reported as the mean of three trials.

Data were interpreted according to46 by comparing experimental IZs obtained from screening oil mixtures with theoretical sum of IZs of separate oils; interactions were interpreted as: synergy if experimental value is higher than the theoretical, antagonism if experimental is lower than theoretical, and indifference if theoretical is approximately equal to the experimental value. Theoretical IZs of mixtures were calculated according to47.

Statistical analysis

Data in all figures were reported as mean ± standard deviation of three trials. Analysis of Variance (ANOVA) was conducted followed by post-hoc test to compare and check for significance between groups. Significance level was set at p < 0.05 for all tests. SPSS Statistics 20.0 was used to conduct statistical analysis.

Results and discussion

Fatty acid composition

Seed oils are rich in several important bioactive compounds including polyphenols, flavonoids, carotenoids, and fatty acids. Oils and oil mixtures were analyzed for their fatty acids (FA) composition (Fig. 1a). Unsaturated FA (oleic and linoleic acid) were dominating (up to 57%) compared to saturated FA (palmitic and stearic acid) (up to 12%). Fatty acid profiles of BSO, DSO, and GSO were similar (Fig. 1b–f). In these three oils, major FA found was linoleic acid (52%–57%, Fig. 1e) followed by oleic acid (26%–30%, Fig. 1d), palmitic acid (9%–12%, Fig. 1b), stearic acid (4%–5%, Fig. 1c), and finally linolenic acid (2%–3%, Fig. 1f).

Figure 1
figure 1

Fatty acid composition in oil samples in % reported as the mean ± standard deviation of three trials. Different letters indicate statistical significance (p < 0.05). ASO-Apricot seed oil, BSO-Black seed oil, DSO-Date seed oil, GSO-Grape seed oil, AB-ASO + BSO, AG-ASO + GSO, AD-ASO + DSO, BD-BSO + DSO, BG-BSO + GSO, GD-GSO + DSO.

Similar order of FA was reported for BSO and GSO48,49. Previous studies on DSO showed that oleic acid is the dominating FA41,50, while our results showed that linoleic acid is the dominating FA in this oil. This difference could be due to variations in cultivation practices and processing techniques of DSO51. ASO showed different percentages, where oleic acid was the major FA (27%), followed by linoleic acid (16%), linolenic acid (8%), palmitic acid (4%), and stearic acid (4%). ASO showed similar results in a previous study12. FA profile of oil mixtures was also rich in unsaturated FA with linoleic acid dominating in all samples. Blending oils showed significant changes in saturated and unsaturated FA in certain mixtures. Palmitic acid and oleic acid were significantly increased in AG compared to ASO and GSO alone. Similar findings were reported by Ramadan and coworkers, where they noticed an increase in palmitic acid and stearic acid when blending corn oil with black cumin seed oil52. In addition, Kaseke and coworkers reported similar results for oil blends53. On the other hand, some mixtures showed non-significant change compared to each oil alone. This was found for stearic acid in BD, BG, and GD, oleic acid in GD, and linolenic acid in BG and GD. Resulting amounts of fatty acids upon blending oils could be influenced by alterations in fatty acids, efficiency of oil blend, and matrix of oil mixtures53.

Total phenolic content (TPC)

Oils are considered an important source of phenolic compounds8. However, their phenolic content is less than seed extracts, and this could be explained by the hydrophilic nature of phenolic compounds making them less soluble in the lipid fraction of the seed16.

Figure 2 shows total phenolic content (TPC) in oils and oil mixtures. GSO significantly showed the highest TPC (0.42 mg GAE/g of oil) while ASO significantly showed the lowest amount (0.03 mg GAE/g of oil). Similar polyphenol content were reported by49,54. BSO and DSO showed 0.09 and 0.08 mg GAE/g of oil, respectively with no significant difference between them. These results were slightly higher than those reported previously41,48. This could be due to differences in cultural conditions, extraction methods, and geographical locations41.

Figure 2
figure 2

Total phenolic content (mg GAE/g of oil) reported as the mean ± standard deviation of three trials. Different letters indicate statistical significance (p < 0.05). ASO-Apricot seed oil, BSO-Black seed oil, DSO-Date seed oil, GSO-Grape seed oil, AB-ASO + BSO, AG-ASO + GSO, AD-ASO + DSO, BD-BSO + DSO, BG-BSO + GSO, GD-GSO + DSO.

For mixtures, GD (0.32 mg GAE/g of oil) showed the highest phenolic content while AB (0.05 mg GAE/g of oil) showed the lowest. TPC of GSO was significantly higher than ASO, BSO, and DSO. Blending the oils increased TPC significantly up to four times in BG, AD, AG, and GD. Comparable results were reported by Kaseke and coworkers; where an increase in TPC content was found after blending sunflower oil with pomegranate seed oil53. BD showed non-significant difference in TPC compared to BSO and DSO.

Antibacterial activity of single oils

Antibacterial activity of oils obtained from apricot seeds, black seeds, grape seeds, and date seeds (Fig. 3a) was evaluated against five bacterial strains using disk diffusion method (Fig. 3b–f). Against P. mirabilis (Fig. 3b), ASO showed significantly highest IZ (15.1 mm). Smaller IZs were observed for GSO (10.6 mm) followed by BSO (9.8 mm) with no significant difference between them (p > 0.05), and the lowest IZ was recorded for DSO (8.6 mm). On the other hand, no significant difference was found between tested oils against P. aeruginosa (Fig. 3c) (p > 0.05). IZs recorded against this strain were 10.5 mm for BSO, 9.5 mm for ASO, 9.3 mm for GSO, and 9 mm for DSO. Absence of inhibition was found only in ASO against S. aureus (Fig. 3d). GSO and DSO showed significantly the highest IZs of 12.3 mm and 11.5 mm respectively, with no significant difference between them, whereas DSO (9.83 mm) significantly showed the lowest IZ against S. aureus. While assessing the activity against E. coli (Fig. 3e), BSO significantly showed the highest IZ (12.3 mm) followed by ASO (11.3 mm) with no significant difference between them. Lower IZs were recorded for DSO (11.5 mm) and GSO (9.8 mm). While assessing the activity against K. pneumoniae (Fig. 3f), DSO significantly showed the lowest IZ of 9.33 mm. Higher IZs were obtained from BSO (11.33 mm), GSO (11 mm), and ASO (10.5 mm) with no observed significant difference between their zones.

Figure 3
figure 3

Antibacterial activity of (a) ASO, BSO, GSO and DSO against (b) P. mirabilis, (c) P. aeruginosa, (d) S. aureus, (e) E. coli, (f) K. pneumoniae, obtained by disk diffusion method. Results of inhibition zones (mm) are given as the mean ± standard deviation of three trials. Different letters indicate statistical significance (p < 0.05). ASO-Apricot seed oil, BSO-Black seed oil, DSO-Date seed oil, GSO-Grape seed oil.

The hydrophobic nature of oils facilitates their penetration through the lipid layers of bacterial cell membrane disturbing their structure, and thus increasing their permeability where ions and cell content can leak out55. Furthermore, antibacterial effect of the oil depends on the bacterial strain used as well as the oil tested where growth inhibition could be a result of synergistic interactions between certain components56.

ASO showed antibacterial effect against all strains, except for S. aureus. Limited studies were done on ASO. Previous studies focused on apricot’s seed extracts13 and essential oil15. Those studies showed antibacterial effect against similar strains but with higher activity, which agrees with those in earlier studies57,58, showing that essential oils and seed extracts could have higher antibacterial activity than the cold-pressed fixed oil extracted from the same seed. On the other hand, total absence of inhibition was found against S. aureus which could be because of the significant low amount of linoleic acid in ASO compared to the other oils. Important antibacterial activity was reported for linoleic acid against S. aureus by inhibiting its FabI, a protein responsible for fatty acid biosynthesis in bacteria59. It was also reported that when S. aureus is exposed to linoleic acid, fermentative and glycolic metabolic pathways are affected by altering their gene expression, leading to loss of its energy production60.

BSO showed an effect against all strains where E. coli had significantly the highest IZ. This is in agreement with previous studies where61 reported highest zone for E. coli compared to S. aureus. In addition, Arici and coworkers found also an effect for this oil against K. pneumoniae and P. aeruginosa along with the other strains; however, with higher IZs62. This could be explained by the fact that BSO bought from distinct commercial sources or extracted using different extraction methods from the same type of seeds have shown significant content variation in their major antibacterial chemical, thymoquinone63. Moreover, different storage conditions such as temperature and light could affect quinone content of the oil that is directly related to the antibacterial activity63.

GSO showed an effect against all the tested strains where S. aureus is significantly the most susceptible strain. This was in accordance with64, who reported the highest antibacterial effect of grape seed extract against S. aureus. More studies on grape seed extracts showed an effect against similar strains but with zones of inhibition higher than those produced in our study20. This is in accordance with the results of58 who found that GSO had the least antibacterial effect when they compared it with other different grape seed extracts. Studies conducted on GSO are limited. Among these studies,65 showed an antibacterial effect of GSO against S. aureus and E. coli.

DSO had shown inhibition against all the strains with zones ranging between 8.6 and 11.5 mm, where S. aureus and E. coli were significantly the most susceptible strains. This is in accordance with a study done by25 on the antibacterial effect of date seed extracts and found that S. aureus and E. coli were highly sensitive to extracts. Also, Aljazy et al. studied date seed extracts and found IZs29 against almost all the strains analyzed in our study, but with higher IZs that could be explained by the low solubility of polyphenols in oil16. However, only one study was found on the assessment of antibacterial effect of DSO, similarly showing an effect against S. aureus and E. coli66.

Antibacterial activity of oils mixtures

Oils were mixed in binary mixtures and tested for possible synergistic antibacterial activity as mentioned in Fig. 4a. Figures 4b,c, 5a–c showed antibacterial activity of oil mixtures against each tested bacterial strain.

Figure 4
figure 4

(a) Method illustration. (b, c) Theoretical IZ, experimental IZ, and interpreted effect of the antibacterial activity of AB, AG, AD, BD, BG, and GD mixtures obtained by disk diffusion method. Results of inhibition zones (mm) are given as the mean ± standard deviation of three trials. Different letters indicate statistical significance (p < 0.05). ASO-Apricot seed oil, BSO-Black seed oil, DSO-Date seed oil, GSO-Grape seed oil, AB-ASO + BSO, AG-ASO + GSO, AD-ASO + DSO, BD-BSO + DSO, BG-BSO + GSO, GD-GSO + DSO.

Figure 5
figure 5

(ac) Theoretical IZ, experimental IZ, and interpreted effect of the antibacterial activity of AB, AG, AD, BD, BG, and GD mixtures obtained by disk diffusion method. Results of inhibition zones (mm) are given as the mean ± standard deviation of three trials. Different letters indicate statistical significance (p < 0.05). ASO-Apricot seed oil, BSO-Black seed oil, DSO-Date seed oil, GSO-Grape seed oil, AB-ASO + BSO, AG-ASO + GSO, AD-ASO + DSO, BD-BSO + DSO, BG-BSO + GSO, GD-GSO + DSO.

Experimental IZs of oil mixtures, theoretical IZs, as well as interpreted effects were reported. Type of interaction obtained when combining oils is a result of the interactions between their components47. Synergy occurs when a combination of two oils produces an antimicrobial effect greater than the sum of the individual oils. Antagonism occurs when the combination produces an antimicrobial effect lower than the sum of the individual oils. Indifference occurs when the combination produce an effect equal to the sum of the individual oils67.

Against P. mirabilis, synergistic interactions were found in AG and GD mixtures (Fig. 4b). The best effect was observed in AG mixture where its diameter of IZ measured 18.6 mm, while theoretically it should be 12.85 mm. Three antagonistic interactions were found in AB, AF, and BG mixture. One indifferent interaction was shown in BD mixture. A study was conducted on the antibacterial interaction of ferulic acid with other phenolic acids against P. mirabilis68. They found that combining caffeic acid and ferulic acid induce synergy against P. mirabilis; ferulic acid found in ASO, DSO, and GSO31,69,70 along with caffeic acid found in GSO and DSO31,69 could explain our results.

P. aeruginosa was the only strain that showed no synergism at all against all the mixtures tested (Fig. 4c). AB, BD, and GD showed antagonistic interaction while AG, AD, and BG showed indifferent interaction. Similar results were reported in previous studies when testing other oil mixtures against this strain71,72. This could be explained by the low permeability of its outer membrane. A study conducted by Mann and coworkers found that permeabilizing P. aeruginosa’s membrane increases the antibacterial activity of tea tree oil against it73.

ASO showed no antibacterial activity against S. aureus. However, when combined with BSO, GSO, and DSO an antibacterial activity was recorded (Fig. 5a). This could be explained by the increase in linoleic acid content of oil mixtures compared to ASO alone inducing an antibacterial activity against this strain. BD also showed synergy against S. aureus which could be due to gallic acid found also in DSO26 and thymol in BSO36,74. GD showed antagonistic effect while BG showed indifferent effect. Highest number of synergistic interactions was recorded against S. aureus, the only gram-positive strain investigated in this study. This could be due to the hydrophobic nature of oils that are not able to penetrate well through the outer hydrophilic membrane present in gram negative bacteria making them less susceptible to oils75.

One synergistic interaction was recorded for AG mixture against E. coli (Fig. 5b). This could be due to the interaction of gallic acid and ferulic acid found in both ASO and GSO17,69,70 that induces synergy once they are combined together76. Antagonism was found for AD, BD, and BG mixtures whereas indifference was found in GD and AB.

Against K. pneumoniae, synergy was found in AG and BD (Fig. 5c). Two antagonistic interactions in AB and BG, as well as two indifferent interactions in AD and GD were also recorded. Synergy in AG and BD could be due to poorly understood interactions between their components against K. pneumoniae, or due to synergistic interactions obtained in another gram negative bacteria belonging to the same family Enterobacteriaceae, Salmonella Typhi, after combining ferulic acid that is present in DSO and ASO31,70 with hydroxybenzoic acid found in BSO77 as well as caffeic acid found in GSO68,69.

AG showed synergistic interactions against all strains tested except P. aeruginosa. This is comparable to the results of FA and TPC where it shows a significant increase in two FA as well as TPC content. On the contrary, BG mixture was the only one with no synergistic interaction recorded. Overall, it is a complex mechanism when it comes to the antibacterial activity of single oils and oils mixtures against distinct bacterial strains. Different interactions between major and minor compounds, ability of compounds present in trace amount to alter activity of major ones, specificity of some compounds to certain bacteria, and susceptibility of tested bacterial strains can all influence the final effect produced78. More importantly, the cell penetration of different components, whether it is hydrophilic or lipophilic attraction, oils fixation on cell membranes and cell walls, and resistance mechanisms exerted by bacterial cells could strongly influence the type of interaction produced79. Hence, it is hard to predict what exact compounds are interacting together to produce synergistic or antagonistic antibacterial activity78.

Conclusions

Using seed oils as antimicrobial agents is an important natural approach. This is also a remarkable way to utilize the oils extracted from fruit seeds instead of discarding them as industrial wastes. ASO and BSO were more effective against gram-negative bacteria while GSO and DSO were more effective against gram-positive bacterium. Blending oils improved FA profile of oils and their antibacterial activity. AG was the most efficient blend showing improved antibacterial activity. ASO showed the highest effect against P. mirabilis, and AG showed the highest number of synergistic interactions against all strains except P. aeruginosa. The results of this study showed that the oils studied have important antibacterial activity that was improved and strengthened when applied together in different mixtures. This was the first study revealing the nature of interactions of the antibacterial effect obtained from binary mixtures of ASO, DSO, GSO, and BSO. Further studies could be done to test their possible application as antibacterial compounds when applied using different methods in both in vivo and in vitro models.