All chemicals were HPLC grade (Sigma, Egypt). Uniform R. officinallis L. plants that have 16 cm stems were obtained from local commercial nurseries. The plants were identified and voucher specimen were maintained at Alexandria University, Egypt. The plants were cultivated in a greenhouse and the experiments were repeated for two successive seasons in January 2016 and 2017. Plants were transplanted into 3.1 L pots containing coarse sand and 2 g L⁻¹ media of commercial compound fertilizer (Crystalon®, 19% N: 19% P: 19% K). Plants were maintained under moderate temperatures (15.1–24.5°C); relative humidity (68–81%) and photosynthetically active radiation (900 W m⁻² at 12.00 p.m.). The plants were irrigated with 250–300 mL plant⁻¹ of saline water every 2 days (640, 2,000, and 4,000 ppm NaCl) and 4 biweekly sprays of salicylic acid (SA) at 0, 100, 200, and 300 ppm until drop off for 8 weeks. Plants subjected to salinity of 640 ppm NaCl (tap water) and 0 ppm salicylic acid sprays were considered as control. The plants were distributed on three blocks and each treatment was represented by six replicates with a total number of plants of 216. After 8 weeks, plant heights (cm) were measured then all plants were harvested. Number of branches per plant was counted and the fresh weight (g) was measured then the dry weight (g) was obtained following drying at room temperature (20°C) as reported by Elansary and Mahmoud (2015).

For essential oil extraction, a Clevenger-type apparatus was used for the hydrodistillation of dried ground leaf samples for 1 h, then the oil was dried, filtered, and finally kept at 4°C in dark conditions (Elansary et al., 2016). The essential oils were subjected to Thermo Scientific Gas Chromatograph Mass Spectrometer equipped with TG-1MS column. The carrier gas was helium, and temperature ramped from 45–165°C at 4°C min⁻¹ for 2 min, then gradual increase to end temperature of 280°C for 15 min. The same column and temperatures were used for GC-FID analysis. Chemical compounds were identified using their retention time and retention indices of n-alkanes (C₁₀–C₃₆) and computer matching using NIST mass spectra ver. 2.0 and WILEY libraries as well as literature references (Chalchat et al., 1993; Adams, 2007). For the preparation of leaf extracts, ground dried leaves were dissolved in methanol (99%, 3 mL), maintained on a shaker in dark conditions for 1 h, then centrifuged for 5 min (4°C, 7,000x gr) and the supernatant was kept at −80°C for further analyses.

2,2′-diphenypicrylhydrazyl (DPPH) and β-carotene-linoleic acid assays were used following Elansary et al. (2017) to determine essential oils free radical scavenging activity of each sample which was expressed as IC₅₀ (μg mL⁻¹) or the sample concentration required to scavenge 50% of DPPH/β-carotene-linoleic acid. Total and free ascorbate (non-enzymatic antioxidants) were quantified in ground frozen leaves following Elansary et al. (2017).

The enzymes activities of ascorbate peroxidase (APX), catalase (CAT), and superoxide dismutase (SOD) were quantified in leaves tissues according to Zhang and Kirkham (1996) and Elansary et al. (2017). ROS accumulation was determined by quantifying H₂O₂ composition in leaves following Elansary et al. (2017) and using the Beyotime Colorimetric Assay Kit (Beyotime, China).

Total phenolics were quantified in leaves according to Amerine and Ough (1988) and Singleton and Rossi (1965) using a Folin-Ciocalteau method. The proline was quantified in 0.5 g (% of dry matter) following Bates et al. (1973) spectrophotometrically at 520 nm. The chemical constituents of the soil (sandy) used were quantified following Jackson (1958) (Table S1). Total chlorophyll content was determined following Moran and Porath (1980). Total carbohydrates (%) in the leaves was determined according to Dubios et al. (1956). Sodium and chloride contents (mg g⁻¹) in the leaves were determined according to Piper (1947).

Quantitative real-time PCR was conducted to evaluate the expression levels of APX gene, 3 SOD genes (FeSOD, Cu/ZnSOD, MnSOD) and 4 genes conferring tolerance to salinity (bZIP62, DREB2, ERF3, and OLPb) in the rosemary plants subjected to different SA treatments (0, 100, 200, 300 ppm) under different saline conditions (640, 2,000, 4,000 ppm). RNeasy Plant Mini kit (Qiagen) was used to isolate the total RNA from these plant tissues. Contaminating DNA was then removed. cDNA was synthesized by using Reverse Transcription kit, and qRT-PCR was then carried out in triplicates using QuantiTect SYBR Green PCR kit. The conditions for PCR amplification were 95°C for 10 min; 50 cycles of 95°C for 30 s, 62°C for 30 s, and 72°C for 2 min; then at 72°C for 3 min. Melting curve analysis was applied to test the amplification specificity. Gene-specific primers of Radyukina et al. (2011), Elansary et al. (2017), and Kim et al. (2017) were used for genes amplification. Actin (Radyukina et al., 2011) was used as a housekeeping gene, and the relative expression levels were determined using 2^−ΔΔCt method.

To explore the antibacterial activities of SA treated compared to control plants under saline conditions, Gram-negative and Gram-positive bacterial were screened against essential oils of different treatments (season 2016). Listeria monocytogenes (clinical isolate), Bacillus cereus (ATCC 14579), Staphylococcus aureus (ATCC 6538), Micrococcus flavus (ATCC 10240), Escherichia coli (ATCC 35210), and Pseudomonas aeruginosa (ATCC 27853) were examined against essential oils of different treatments using the micro-dilution method (Elansary et al., 2016). In microtiter plates, known concentration of essential oil was mixed with bacterial inoculum in Triptic Soy broth, then kept at 37°C for 24 h in a rotary shaker to determine the minimum inhibitory and bactericidal concentrations (MICs and MBCs). The MBC was determined using serial sub-cultivation of each essential oil (2 μL) with each bacterium then incubated 24 h at 37°C and the MBC was defined as the minimum concentration of the essential oil indicated killing off 99.5% of the of the inoculum. The density was determined at 655 nm and all experiments were performed twice in triplicates. Positive control (streptomycin and ampicillin, 0.01–10 mg/ mL) as well as the DMSO (5%) negative controls were used.

Salinity was the main plot and Salicylic acid was the subplot in split plot experimental design. Experiments were repeated twice in January 2016 and 2017. The data of each season were presented as means and subjected to Least significant differences (LSD) test in ANOVA test using SPSS (PASW Ver. 21).

The use of irrigation water (tap water) that has 640 ppm NaCl and 3 doses of SA (100, 200, and 300 ppm) indicated that SA had significant effects on enhancing all studied morphological parameters in both seasons compared to the 640 ppm NaCl and 0 ppm SA (control) as shown in Table 1. Increasing NaCl concentration from 640 to 2,000 and 4,000 ppm significantly reduced number of branches, plant height, fresh and dry weights in plants received 0 ppm SA (Table 1). For example, plant height was reduced from 31.16 to 19.33 cm when increasing NaCl concentration from 640 to 4,000 ppm without using SA (0 ppm SA). On the other hand, SA significantly alleviated the morphological stress effects of salinity by increasing number of branches, plant height and fresh and dry weights in both seasons. Higher SA doses (200 and 300 ppm) were more effective in enhancing the overall morphological parameters during salinity stress (2,000 and 4,000 ppm NaCl). For example, plant heights increased from 22.33 to 28.33 cm and from 19.33 to 27.33 cm when treated with 300 ppm SA whereas 100 ppm SA showed lower values in 2016 season. Additionally, under saline conditions, the number of branches per plants showed significant increases following SA in both seasons. The fresh and dry weights showed parallel pattern for that found in number of branches and plant height.

Ten major oil constitutes were found in the essential oils of rosemary leaves such as cineole, α-pinene, camphor, linalool, borneol, and verbenone (Table 2). Plants irrigated with tap water (640 ppm NaCl) and sprayed with different doses of SA (100–300 ppm) showed significant differences in all of the 10 essential oil constitutes compared to SA-untreated plants. It was noted that specific oil constitutes decreased following spraying with 100–300 ppm SA in the first season such as that found in α-pinene which decreased from 8.56% in control plants to 4.02% in 640 ppm NaCl and 100 ppm SA treated plant in the first season. However, in the 640 ppm NaCl and 100–300 ppm SA treated plants, there were increases in several compounds such in linalool which increased from 5.43% in control plants to 6.59% in 640 ppm NaCl and 200 ppm SA treated plant in 2016. Similar increases were found in camphor, borneol, and caryophyllene oxide.

Compared to control, salinity conditions of 2,000 ppm NaCl and 0 ppm SA showed apparent changes in many major oil constitutes such as the great reductions in α-pinene, β-pinene, and cineole along sharp increases in linalool, camphor, borneol, and verbenone in both growing seasons. Interestingly, the applications of SA at 100–300 ppm largely ameliorated the effects of salinity at 2,000 ppm NaCl. The notable saline mitigation effects were found in plants treated with SA at 100 ppm such as that found in α-pinene which increased from 0.9% in 2,000 ppm NaCl and 0 ppm SA treated plants to 8.26% in 2,000 NaCl and 100 ppm SA treated plants in 2016. The value of 8.26% α-pinene is comparable to 8.56% found in control plants in 2016. Both seasons of 2016 and 2017 had comparable values.

Under 4,000 ppm NaCl and 0 ppm SA, the reduction in specific oil constitutes including α-pinene and linalool were comparable to that found in plants treated with 2,000 ppm NaCl and 0 ppm SA in both seasons. However, SA sprays at 100–300 ppm had significant effects in reversing reductions in specific essential oil compositions such as α-pinene, β-pinene, camphen, cineol, and terpinolen. SA applications reduced specific essential oils constitutes such as the linalool (from 7.89% in 4,000 ppm NaCl and 0 ppm SA treated plants to 5.91% in 4,000 NaCl and 200 ppm SA treated plants in 2016) and the verbenone (from 12.68% in 4,000 ppm NaCl and 0 ppm SA treated plants to 3.95% in 4,000 NaCl and 200 ppm SA treated plant in 2016). Essential oil ratio showed significant variations related to saline irrigation and SA sprays (Table 2). Significant increases were recorded in the essential oil ratio following 640 ppm NaCl and SA sprays at 100–300 ppm compared to SA-untreated plants. Under saline conditions of 2,000–4,000 ppm NaCl, there were significant reduction in the essential oil ratio in 2,000 and 4,000 ppm NaCl and 0 ppm SA treatments compared to 2,000 and 4,000 NaCl and 100–300 ppm SA plants.

Leaves total phenolic, chlorophyll, carbohydrate, and proline contents varied among treatments (Table 3). Under 640 ppm NaCl, SA sprays at 100–300 ppm significantly increased the phenolic composition of treated plants compared to 0 ppm SA in both seasons. Saline treatment at 2,000 ppm NaCl increased the phenolic composition compared to control plants. Also, under 2,000 ppm NaCl, SA treatments at 100–300 ppm significantly increased the phenolic compositions compared to 2,000 ppm NaCl and 0 ppm salicylic acid treated plants. Under 4,000 ppm NaCl, SA sprays at 100–300 ppm increased the phenolic composition compared to 4,000 ppm NaCl and 0 ppm SA treated plants.

Total chlorophyll contents showed significant increases following SA sprays at 100–300 ppm under 640, 2,000, and 4,000 ppm NaCl compared to SA-untreated plants (Table 3). The highest chlorophylls were found in 640 ppm NaCl and 200–300 ppm SA treated plants in both growing seasons. In addition, 0 ppm SA treatments showed the lowest values of chlorophyll contents compared to SA treated plants under 640, 2,000, and 4,000 ppm. For example, the plants treated with 4,000 ppm NaCl and 0 ppm SA had 18.93 mg g⁻¹ FW of total chlorophyll compared to 21.75 mg g⁻¹ FW in 200 ppm SA treatment. The total carbohydrates showed significant increases in the leaves of SA treated plants with 640, 2,000, and 4,000 ppm NaCl and 200–300 ppm SA compared to 0 ppm SA treated plants in both seasons. The proline composition of leaves showed significant increases associated with increased salinity levels in both season (Table 3).

Leaves chloride and sodium compositions showed significant variations among treatments in the two seasons (Table 4). There was significant reduction in Na⁺ and Cl⁻ compositions in treated plants at 640 ppm NaCl and 200–300 ppm SA compared to control treatments in both seasons. Under saline conditions of 2,000 and 4,000 ppm NaCl, SA treatments at 200–300 ppm significantly reduced Na⁺ and Cl⁻ compositions in the leaves in both seasons compared to 0 and100 ppm SA treatments.

The total antioxidant capacity of essential oils as well as leaf extracts were determined in all treated plants, and significant differences were found among plants as responses to different levels of salinity (640–4,000 ppm NaCl) and SA treatments (0–300 ppm) (Table 5). The plants treated with saline conditions (2,000 and 4,000 ppm) without SA treatment showed significantly higher antioxidant capacity compared to control plants. For example, the IC₅₀ in the DPPH method was reduced from 0.26 to 0.23 μg mL-1 when increasing the salinity from 640 to 4,000 ppm in the first season. The plants treated with SA at 100–300 and 640–4,000 ppm NaCl showed higher antioxidant capacity than 0 ppm SA treated plants. Using both methods of DPPH and linoleic acid assays, the highest antioxidant capacity among essential oils was found in plants treated with 4,000 ppm NaCl and 100–300 ppm SA in both seasons of 2016 and 2017.

In the DPPH assay, leaf extracts of plants treated with 640 ppm NaCl and 100–200 ppm SA showed higher antioxidant capacity than 0 ppm SA-treated plants in the two seasons. Furthermore, linoleic acid antioxidant assay pattern was similar to that found in the DPPH.

Significant increases were found in activities of CAT, SOD, and APX following SA sprays at 100–300 ppm. SA sprays at 300 ppm showed the highest activities of the enzymes compared to lower doses (Figure 1). There were significant reductions in H₂O₂ compositions in leaves following SA sprays compared to non-sprayed plants (Figure 2). Additionally, free ascorbate showed significantly higher increases following SA treatments at 200–300 ppm, and slight increases in the total ascorbate were also found (Figure 3).

The expression levels of APX gene, 3 SOD genes and 4 genes conferring tolerance to salinity (bZIP62, DREB2, ERF3, and OLPb) were assessed in the rosemary plants subjected to different SA treatments (0, 100, 200, 300 ppm) under different saline conditions (640, 2,000, 4,000 ppm). The results showed that the APX and 3 SOD genes revealed higher levels in SA-treated rosemary under saline, with respect to non-sprayed plants (Figure 4), indicating the important roles of salicylic acid and antioxidant enzymes under abiotic stresses. SA treatment at 3,000 ppm showed the highest expression level for all the antioxidant enzymes genes.

Moreover, the expression levels of the genes conferring tolerance to salinity (bZIP62, DREB2, ERF3, and OLPb) were evaluated using qRT-PCR. All the genes revealed higher levels in SA-treated rosemary under salt stress, as compared to non-sprayed plants (Figure 5). SA treatment at 3,000 ppm showed the highest expression level for all the stress-related genes. This indicates that SA treatment modulated the abiotic stress responsive gene expression in rosemary which in turn enhanced its tolerance to salinity stress.

The antibacterial activities of 2016 season essential oils treatments were compared as shown in Table 6. There were increases in the antibacterial activities with increasing either salinity or SA treatment concentration. Essential oils MIC were in the range of 0.0001–0.011 mg mL⁻¹, whereas the MBC were in the range of 0.0004–0.030 mg mL⁻¹. The highest antibacterial activities were found against S. aureus compared to other bacteria. The essential oil of plants treated with 4,000 ppm NaCl and 300 ppm SA showed the highest antibacterial activity ranges with MIC and MBC of 0.0010–0.0001 and 0.0021–0.0004 mg mL⁻¹, respectively. The antibacterial activities of rosemary essential oils were much higher than antibiotics.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.