Leaves of R. officinalis from flowering plants in Huntsville, Alabama (34°38′27.2″ N, 86°33′44.4″ W, elevation 183 m above sea level (asl)) were identified by W.N. Setzer, and collected on 4 August 2016. The fresh plant material (55.44 g) was hydrodistilled for 4 h using a Likens-Nickerson apparatus with continuous extraction with dichloromethane to give 950 mg (1.7% yield) colorless essential oil, which was stored at −20 °C until analysis.

The leaves of R. officinalis were collected during the flowering stage in May 2013, from Dhamar province, Yemen. The plant was identified by Dr. Hassan M. Ibrahim of the Botany Department, Faculty of Sciences, Sana’a University. A voucher specimen of the plant material (YMP-lam-31) has been deposited at the Pharmacognosy Department, Sana’a University, Yemen. The dried leaves were hydrodistilled for 3 h in a Clevenger type apparatus according to the European Pharmacopoeia [42]. The obtained oil was subsequently dried over anhydrous Na₂SO₄ and kept at 4 °C until analysis. After filtration, the yield of the oil was 1.1% w/w.

R. officinalis, in the flowering stage, from Jumla, Nepal (29°16′28.99″ N, 82°11′1.79″ E, elevation 2500 m asl), was identified by Prasun Satyal, and collected on 2 July 2016. The fresh plant material (100 g) was hydrodistilled for 4 h using a Clevenger-type apparatus and collected to give 500 mg (0.5% yield) colorless essential oil after drying with Na₂SO₄.

R. officinalis from Ribeeck Kasteel, Western Cape, South Africa (33°23′7.21″ S, 18°53′54.65″ E, elevation 300 m asl), was identified by Prabodh Satyal, and collected on 20 August 2016. The fresh plant material (500 g) was subjected to steam distillation for 4 h using a Clevenger-type apparatus and collected to give 4 g (0.8% yield) colorless essential oil after drying with Na₂SO₄.

Flowering R. officinalis from Lancefield, Victoria, Australia (37°16′ S, 144°43′ E, elevation 495 m asl), was identified by Chris Burder, and collected on 20 December 2015. The fresh plant material (1 kg) was subjected to steam distillation for 3 h using a Clevenger-type apparatus and collected to give 9 g (0.9% yield) colorless essential oil after drying with Na₂SO₄.

R. officinalis from Thika, Kenya (1°1′ S, 37°5′ E, elevation 1500 m asl), was identified by Aaron Sorensen, and collected on 8 July 2016. The fresh plant material (1 kg) was subjected to steam distillation for 3.5 h using a Clevenger-type apparatus and collected to give 10 g (1.0% yield) pale yellow essential oil after drying with Na₂SO₄.

The essential oils of R. officinalis were analyzed by GC-MS using a Shimadzu GCMS-QP2010 Ultra operated in the electron impact (EI) mode (electron energy = 70 eV), scan range = 40–400 atomic mass units, scan rate = 3.0 scans/s, and GC-MS solution software. The GC column was a ZB-5 fused silica capillary column with a (5% phenyl)-polymethylsiloxane stationary phase and a film thickness of 0.25 μm. The carrier gas was helium with a column head pressure of 552 kPa and flow rate of 1.37 mL/min. Injector temperature was 250 °C and the ion source temperature was 200 °C. The GC oven temperature program was programmed for 50 °C initial temperature, temperature increased at a rate of 2 °C/min to 260 °C. A 5% w/v solution of the sample in CH₂Cl₂ was prepared and 0.1 μL was injected with a splitting mode (30:1). Identification of the oil components was based on their retention indices determined by reference to a homologous series of n-alkanes, and by comparison of their mass spectral fragmentation patterns with those reported in the literature [43], and stored in our in-house MS library.

Chiral analysis of the essential oils was performed on a Shimadzu GCMS-QP2010S operated in the EI mode (electron energy = 70 eV), scan range = 40–400 amu, scan rate = 3.0 scans/s. GC was equipped with a Restek B-Dex 325 capillary column (30 mL × 0.25 mm ID × 0.25 μm film). Oven temperature was started at 50 °C, and then gradually raised to 120 °C at 1.5 °C/min. The oven was then raised to 200 °C at 2 °C/min and held for 5 min. Helium was the carrier gas and the flow rate was maintained at 1.8 mL/min. Samples were diluted 3% w/v with CH₂Cl₂ and then a 0.1 μL sample was injected in a split mode with a split ratio of 1:45.

A total of 72 R. officinalis essential oil compositions from the published literature [17,18,19,20,28,29,30,31,32,33,34,35,36,37,39,41,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81] in addition to the six samples from this study were treated as operational taxonomic units (OTUs). The percentage composition of 23 major essential oil components (1,8-cineole, α-pinene, camphor, verbenone, borneol, camphene, myrcene, bornyl acetate, β-pinene, limonene, α-terpineol, linalool, β-caryophyllene, p-cymene, terpinen-4-ol, γ-terpinene, geraniol, α-phellandrene, terpinolene, α-terpinene, α-humulene, sabinene, and caryophyllene oxide) was used to determine the chemical relationship between the various R. officinalis essential oil samples by agglomerative hierarchical cluster (AHC) analysis using the XLSTAT software, version 2015.4.01 (Addinsoft™, New York, NY, USA). Pearson correlation was selected as a measure of similarity, and the unweighted pair-group method with arithmetic average (UPGMA) was used for cluster definition.

Antifungal activity was carried out as previously described [82]. Briefly, minimum inhibitory concentrations were determined using microdilution methods for Candida albicans (American Type Culture Collection, ATCC #18804), Cryptococcus neoformans (serotype D or var. neoformans) (ATCC #24067), and Aspergillus niger (ATCC #16888). Single colonies from potato dextrose agar plates were grown in potato dextrose broth for three days. Cells were diluted to 2 × 10³ cells/mL using MOPS (3-(N-morpholino)propanesulfonic acid) buffered RPMI (Roswell Park Memorial Institute) medium and aliquoted into sterile 12 × 75 mm tubes (900 μL). Essential oil (100 μL) was added to each tube followed by incubation at 37 °C for three days in a shaking incubator (175 rpm). Minimum inhibitory concentrations (MICs) were calculated from microdilution in 96-well plates, performed in triplicate. Serial dilution was performed by adding 50 μL of RPMI to each well, then an equal volume of essential oil to be tested to the first row. After mixing, 50 μL was removed and added to the next row. The procedure was repeated for each row, discarding the final 50 μL removed. To the mixture in each well, 50 μL of cells were added. The plates were incubated for two days at 37 °C before growth was quantitated from turbidity.

Using xanthine as substrate, the xanthine oxidase (XO) activity of R. officinalis oil from Yemen was assayed spectrophotometrically according to Apaya and Hernandez [83]. A mixture containing 1 mL of 100 μg/mL of R. officinalis oil or allopurinol, 1.9 mL of 50 mM potassium phosphate buffer, and 1 mL of xanthine substrate (0.6 mM) was preincubated for 10 min at 25 °C, and the reaction was started by the addition of 0.1 mL of XO enzyme (0.1 U/mL in phosphate buffer). The reaction was incubated at 25 °C for 30 min, and the absorbance was measured against phosphate buffer as blank at 295 nm using quartz cuvettes. Allopurinol was used as standard enzyme inhibitors. The percent xanthine oxidase inhibition was calculated according to the following formula: % inhibition = 100 − (A₁ − B) × 100/(A_o − B)(1) where A₁ is the activity of the enzyme in presence of the oil, B is the absorbance in absence of the enzyme, and A_o is the absorbance in absence of the oil or inhibitor.

In vitro mushroom tyrosinase inhibitory activity of R. officinalis oil from Yemen was determined by a spectrophotometric approach using l-tyrosine as the substrate [84]. In a total volume of 200 µL, the enzyme activity was measured in buffer containing 50 mM phosphate buffer, pH 6.5, 50 U/mL mushroom tyrosinase, and 50 µg/mL l-tyrosine. The reaction (conversion of l-tyrosine to DOPAchrome) was conducted at 37 °C for 30 min, and absorbance was then measured at the wavelength of 490 nm using a microplate reader. Blanks were run for the same concentration of essential oil in the absence of the enzyme. The inhibition assays were carried out in the presence of 100 µg/mL of the essential oil. Kojic acid and arbutin were used as positive control inhibitors in these assays. The absorbance of the same mixture without the essential oil was used as the negative control. The percent inhibition of tyrosinase activity was calculated as follows: % inhibition = ((A − B)/A) × 100(2) where A is the absorbance difference at 490 nm of the negative control (enzyme activity without oil)—the absorbance of the blank (no enzyme, no oil); B is the absorbance of test sample (with enzyme)—the absorbance of test sample blank (no enzyme).

Human colorectal cancer cell lines (SW480 and HCT116) were generously provided by Dr. Rick F. Thorne (University of Newcastle, Australia) and were cultured in Dulbecco’s Modified Eagle Medium (DMEM) containing 10% fetal calf serum (Bio Whittaker, Verviers, Belgium). The rosemary oil from Yemen was tested for acute cytotoxic effect; the essential oil was dissolved in dimethylsulfoxide (DMSO) and different dilutions in culture buffer were prepared. The acute cytotoxic effect of the essential oil on colorectal cancer cells was determined using the MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay [85]. Briefly, cells were seeded at 5000 cells per well onto flat-bottomed 96-well culture plates and allowed to grow for 24 h before the desired treatment. Cells were then incubated with 200 µL of essential oil concentrations (0–200 μg/mL) for 72 h. Cells were then labeled with MTT from the Vybrant MTT Cell Proliferation Assay Kit (Molecular Probes, Eugene, OR, USA) according to the manufacturer’s instruction and resulting formazan was solubilized with DMSO. Absorbance was read in a microplate reader at 540 nm.

The chemical compositions of six Rosmarinus officinalis essential oils are compiled in Table 1. α-Pinene and 1,8-cineole dominated the essential oils of all six samples (13.5%–38.1% and 16.3%–29.4%, respectively). Verbenone was also relatively abundant in the samples from Alabama (USA), Kenya, and Yemen. The sample from Yemen, with its relatively high verbenone content (18.6%) and relatively low α-pinene content (13.5%), along with 1,8-cineole (20.6%) and camphor (7.0%), firmly places it in the verbenoniferum chemotype, type IIIA, according to Napoli et al. [39]. The other five samples in this study have higher concentrations of α-pinene than 1,8-cineole or verbenone, and can be characterized, based on Napoli et al., as cineoliferum type VIA [39].

The enantiomeric distributions of monoterpenoids of the rosemary oils, determined by chiral GC-MS, are summarized in Table 2. (+)-α-Pinene was the predominant enantiomer (97%–99%) in the five samples analyzed in this study. β-Pinene, on the other hand, showed 29%–33% (+)-enantiomer, in general agreement with that reported by Presti and co-workers, 36%–40% (+)-β-pinene [86]. The enantiomeric distributions of limonene (40%–54% (+)-limonene), linalool (95%–97% (−)-linalool), terpinen-4-ol (69%–70% (+)-terpinen-4-ol), and α-terpineol (66%–75% (+)-α-terpineol), are also in qualitative agreement with those reported by Presti et al., 48%–53% (+)-limonene, 99% (−)-linalool, 73%–76% (+)-terpinen-4-ol, and 66%–67% (+)-α-terpineol [86]. Presti and co-workers found (+)-sabinene to be dominant in rosemary oil (96%–98%) [86], but sabinene concentrations were too low to determine the enantiomeric distribution in the samples in this study. König and co-workers found the enantiomeric distribution of borneol in rosemary oils to show significant differences; the (−)-enantiomer generally dominated, but was much larger in rosemary oil originating in Spain [87]. In this current study, (−)-borneol also dominated (95%–98%). Furthermore, bornyl acetate in this study was enantiomerically pure with 100% (−)-bornyl acetate. Verbenone was also enantiomerically pure, 100% (+)-verbenone, in all rosemary oils examined. Ravid and co-workers also found high enantiomeric purity (96%–100% (+)-verbenone) in their R. officinalis essential oil samples [88]. Thus, we can conclude that the (+)/(−) ratios of each of the monoterpenoids remains relatively constant in rosemary oils, regardless of geographical location of the oil source.

In order to provide additional insight into the chemotypes of rosemary essential oils, we have carried out a hierarchical cluster analysis based on the chemical compositions of the six oils in this study along with 72 additional rosemary oil chemical compositions from the literature. The dendrogram of the analysis is shown in Figure 1. Based on this analysis, there are five different chemotypes: (1) α-pinene/1,8-cineole, (2) verbenone/α-pinene/camphor/1,8-cineole, (3) myrcene/1,8-cineole/camphor, (4) 1,8-cineole/camphor/α-pinene, and (5) α-pinene/β-pinene/camphene. Chemotype 1, dominated by α-pinene [36,37,89,90,91,92], is a cluster made up of 32 samples, including those from Victoria (Australia), Nepal, Kenya, Western Cape (South Africa), and Alabama (USA). Chemotype 2, dominated by verbenone (i.e., the verbenoniferum chemotype) [39,74], has eight samples, including the sample from Yemen. Chemotype 3, dominated by myrcene [89], has four samples in this analysis. Chemotype 4 is dominated by 1,8-cineole [36,38,93,94,95] and is comprised of 33 samples, and can be considered comparable to the cineoliferum chemotype described by Napoli and co-workers [39]. The fifth chemotype, represented by only one sample from Mexico [35], has nearly equal quantities of α-pinene, β-pinene, and camphene; the relatively high β-pinene concentration (12%) separates this sample from chemotype 1.

Antifungal activity was tested against three common opportunistic fungal pathogens. Yeast-like C. albicans and mold-like A. niger are Ascomycota while C. neoformans is a Basidiomycota. Overall, no inhibitory activity at or below 2500 ppm for the R. officinalis essential oil chemotypes was overserved against either ascomycete. Inhibitory activity was observed against C. neoformans for the majority of R. officinalis chemotypes tested. Though not exceedingly promising, R. officinalis from Australia did demonstrate an MIC of 625 ppm. Previous investigations of antifungal activity rosemary oil against A. niger [76,96,97] and C. albicans [19,98,99] have also found rosemary oil to be inactive, while against C. neoformans it had been found to be moderately active (MIC = 156 μg/mL) [100]. Thus, rosemary oil cannot be considered a worthwhile antifungal agent.

The rosemary oil from Yemen was screened for xanthine oxidase inhibitory, tyrosinase inhibitory, and cytotoxic (SW480 and HCT116 human colorectal carcinoma cells) activity, but showed no activity in any of these assays. Yemeni rosemary oil showed only 8.6% ± 4.7% inhibition of xanthine oxidase. Against tyrosinase, Yemeni rosemary oil showed only 3.3% ± 1.5% inhibition at 100 μg/mL. In contrast, rosemary oil from cultivated plants from Mauritius did show tyrosinase inhibitory activity (median inhibitory concentration, IC₅₀ = 97 μg/mL) [101]. Rosemary oil was found to be relatively non-toxic to SK-OV-3, HO-8910, Bel-7402 (IC₅₀ ≥250 μg/mL) [102], MCF-7, LNCaP, and NIH-3T3 cells (IC₅₀ >180 μg/mL) [32], but active against A549 cells (IC₅₀ = 80 μg/mL) [103]. In summary, the enzyme inhibitory potential of rosemary oil is fairly low. In high concentrations, however, rosemary oil may be harmful to some mammalian tissues.