Introduction

The medicinal properties of plants have been widely known and used since ancient times due to their beneficial and curative effects. A medicinal plant’s distinct pharmacological qualities stem from the combined effects of its many natural elements. The whole expression of a medicinal plant’s bioactivity is found in its plant complex, which is a biochemical entity (Bouras et al., 2024). In fact, its synergistic and additive effects, increased pharmacological efficacy, or even decreased toxicity, are often not related to a single active component (Severina et al., 2013). Essential oils (EOs) are volatile compounds with a low molecular weight and biological activities synthesized in different plant organs, especially flowers, buds, leaves, branches, stems, seeds, fruits, woods, and roots (Jain et al., 2022). Furthermore, an abundance of active chemicals, including alkaloids, tannins, steroids, glycosides, resins, phenols, volatile oils, and flavonoids, are present in EOs (Amirifar et al., 2022; Asgari et al., 2017).

Nowadays, plant extracts and EOs are sources of advantageous chemical compounds with potential uses in the food, cosmetics, pharmaceutical, and agricultural industries. They are becoming popular natural alternatives to synthetic antioxidants to address consumer concerns about the adverse effects of synthetic antioxidants that have toxic effects on consumers and can lead to various cancers. Therefore, new and affordable sources of natural antioxidants are becoming more widely available in order to preserve and improve customers’ health and ensure food safety (Nastaran et al., 2021; Zahnit et al., 2022). Due to the significant amounts of volatile, aromatic, and bioactive chemicals they contain, EOs and extracts are highly valued in various sectors (Messaoudi et al., 2022; Samadi et al., 2021). Additionally, these effective compounds have inherent antioxidant, enzymatic, and antibacterial characteristics that are crucial for daily life (Bolouri et al., 2022).

Around the world, Juniperus species are frequently used in traditional medicine for a variety of purposes. Among their many uses, they are used as insect repellent, hypoglycemic, carminative, diuretic, antibacterial, antitussive, antifertility, and stomachic. However, they are also favored as treatments for urticaria, rheumatoid arthritis, TB, leukorrhea, fever, diarrhea, jaundice, and urinary tract infections. Berries of the J. phoenicea plant are the source of juniper oil, which is included in various pharmacopeias, including the 8th edition of the European Pharmacopoeia (Ph. Eur. 8). For centuries, J. phoenicea berries and their EOs have been used for both medicinal and cosmetic purposes, due to their complex combination of chemical compounds, mostly aromatics and terpene hydrocarbons, in addition to terpenoids, polyphenolic compounds, alcohols, myrcene, thujone, glycoside acetate-ethers, citronellol, curcumin, geraniol, terpineol, and ketones. According to Amalich et al. (2015), the primary chemical found in the EO of Moroccan J. phoenicea was α-pinene (78.8%), germacrene D (5.42%), 4 trans-data-4(11),7-diene (2.98%), E-caryophyllene (2.77%), α-himachalene (0.9%), and δ- cadinene (0.71%). Also, in the study by Mehira et al. (2021), which compared EO extractions from leaves and a combination of leaves/berries, a significant increase in the amount of chemical composition was found for the combination of leaves/berries. In the same context, El-Sawi et al. (2007) reported, in their study, the chemical composition of Egyptian J. phoenicea berries oil, where they found the major components were α-pinene (39.30%) followed by sabinene (24.29%), trans-pinocarveol (4.27%), β-phellandrene (4.13%), α-terpinyl acetate (3.36%), p-mentha-1,5-diene-8-ol (2.88%), and β-pinene (2.45%). Additionally, Medini et al. (2011) examined the chemical composition and antioxidant activity of the EO of Tunisian J. phoenicea L ripe and unripe berries. They indicated that the major components of the oils were α-pinene (58.61%–77.39%), camphene (0.67%–9.31%), δ-3-carene (0%–10.01%), and trans-verbenol (0%–5.24%). Also, Ennajar et al. (2009) analyzed the chemical composition of J. phoenicea plants collected from the southeastern region of Tunisia. They found that the main components were α-pinene (80.7%), δ-3-carene (4.5%), γ-cadinene (5.1%), β-caryophyllene (2.9%), β-myrcene (1.8%), and germacrene B (1.5%). Another study conducted by Bouyahyaoui et al. (2016) investigated the chemical composition of Algerian J. phoenicea EO from berries. They found that the main constituents were α-pinene (56.6%), α-terpineol, myrcene (2%), verbenone (1.9%), germacrene D (1.5%), β-caryophyllene (1.2%), camphor (1.9%), p-mentha-1,5-dien-8-ol (1.6%), terpinen-4-ol (1.1%), trans-pinocarveol (1.1%), α-phellandrene (1%), and β-pinene (1%). Harhour et al. (2018) analyzed the composition of the EOs of wild J. phoenicea from northern Algeria, including berries and branches. They found that the major components were α-p (40.3%), δ-3-carene (20.1%), α-cedrol (7%), α-terpinolene (4.5%), β-phellandrene (4.1%), β-myrcene (2.8%), α-terpinylacetate (2.7%), β-caryophyllene (1.9%), α-fenchene (1.9%), α-humulene (1.6%), β-pinene (1.5%), and elemol (1.2%).

The novelty of the present study is the enhancement of the chemical data from two non-polar crude extracts (cyclohexane and ethyl acetate). This information was obtained for the first time using GC-MS techniques to the best of our knowledge. The main purpose of this study was to characterize the chemical composition of the EO and non-polar extracts (cyclohexane and ethyl acetate) isolated from J. phoenicea berries. These berries were harvested in the Aflou region of Laghouat (Algerian Sahara). The study aims to assess their biological activities, including anti-diabetic effects (α-amylase activity) and antioxidative properties (DPPH, ABTS, and phenanthroline activities). Additionally, outcomes were further analyzed with molecular docking and absorption, distribution, metabolism, elimination, and toxicity (ADMET) studies. These analyses aimed to investigate the binding affinity of all compounds identified through GC-MS analysis to the active sites of α-amylase and α-glucosidase. Furthermore, the pharmacokinetic properties of each extract were determined through ADMET experiments. Ultimately, the data synthesized through these experiments provided insight into the best-docked molecules.

Materials and Methods

Plant material

Ripe berries of Juniperus phoenicea were collected in March 2023 from the region of Aflou (Djebel Amour), located at the center of the Saharan Atlas and at the edge of the Sahara Desert (34°12' N; 2°10' E, 1300–1470 m) in the Laghouat province of southern Algeria. The plant’s identity was verified by the Department of Biological Sciences, Faculty of Science at Laghouat University in Algeria. Voucher specimens were deposited at the Research Unit of Medicinal Plant at the University of Laghouat, Algeria with the identification numbers URPM. Jp.03.23.

The J. phoenicea ripe berries were separated and placed on paper. The samples were dried in two steps: shade-drying and oven-drying.In the case of shade-drying, initially, 700 g of fresh sample was spread over 1 m² of area and dried in a ventilated place at a temperature of 25 ± 2°C for 3 weeks. The same quantity was then transferred to an oven and dried for 4 days at a temperature of 50°C with 40% ventilation. The plant materials were ground using a basic grinder (IKA-10) with parameters of 3000rpm and a 1 mm sieve to get a homogenous powder.

To calculate the moisture content of the sample, 5 mg of ground fruit was placed in a porcelain mortar in the oven at a temperature of 105°C without ventilation for 24 h. After that, the moisture content (TH%) was determined to be 5.3%.

Preparation of plant extracts

Essential oil extraction

Juniperus phoenicea ripe berries (300 g) were ground into small pieces and subjectedto hydro-distillation for 3 h in a Clevenger-type apparatus, using 1 L of deionized water. The transparent oil was collected, dried over anhydrous Na₂SO₄, and stored at 4°C in dark sealed vials for further analysis.

Non-polar crude extracts

Forty grams (40 g) of dried berry powder was divided into two equal parts (20 g each) and separately extracted by maceration with 100 mL of solvents (cyclohexane and ethyl-acetate) in a 1:5 ratio. This extraction process was repeated three times with fresh solvent each time. The mixture was then filtered through Whatman filter paper no.1. The filtrate was evaporated at 40°C using a rotary evaporator and stored in a refrigerator at 4°C for further analysis.

Reagents and chemicals used

At the National Center for Biotechnology Research, we used a 96-well microplate reader (PerkinElmer Multimode Plate Reader EnSpire) to measure the activities of various substances. These substances included ascorbic acid, quercetin, α-tocopherol, butylated hydroxyanisole (BHA), butylated hydroxytoluene (BHT), Folin–Ciocalteu reagent (FCR), 1,1-diphenyl-2-picrylhydrazyl (DPPH), 2,20-azino-bis(3-ethylbenzothiazoline-6-sulfonicacid)-diammonium salt (ABTS), phenanthroline, pancreatic α-amylase enzyme (1U), acarbose, starch, sodium phosphate dibasic heptahydrate, sodium phosphate monobasic monohydrate, and hydrochloric acid, all obtained from Sigma-Aldrich. Additionally, we used sodium carbonate, aluminum chloride (AlCl₃), iron (III) chloride (FeCl₃), sodium bicarbonate (NaHCO₃), potassium iodide (KI), potassium persulfate (K₂S₂O₈), and potassium acetate (CH₃CO₂K), obtained from Biochem Chemopharma. All other chemicals and solvents used were of analytical grade.

Gas chromatography-mass spectrometry (GC-MS)

The composition of J. phoenicea berries’ EO and crude extracts was determined using a gas chromatography-mass spectrometer system (GC-MS; GC-2010 Shimadzu Corporation, Kyoto, Japan) coupled to a mass spectrometer detector (QP-2020 A, Shimadzu). The GC–MS system utilized an Rxi®-5ms capillary column (Phase: Crossbond® 5% diphenyl/95% dimethyl polysiloxane) with dimensions of 30 m × 0.25 mm and a 0.25 µm film thickness. This column is similar in phase to HP-1ms, HP-1msUI, DB-1ms, DB-5ms, DB-1msUI, Ultra-1, VF-1ms, ZB-1, ZB-1ms and is considered equivalent to USP G1, G2, and G38 phases.

Analyses were conducted using helium (99.995% purity) as a carrier gas at a column flow rate of 1 mL/min with 1 µL of the sample injected in split mode at a ratio of 1:50. The oven program was as follows: injector and interface temperatures were maintained at 250°C and 310°C, respectively. The column temperature was initially set at 50°C fixed for 2 min, then increased to 310°C at a rate of 3°C/min, and then maintained at 310°C for 2 min.

The mass spectrometer conditions included an ionization voltage of 70 eV, ion source temperature of 200°C, and electron ionization mass spectra acquired over the mass range of 45–600 m/z. Component identification was determined by matching mass spectra with Wiley and NIST library data from 2017, standards of the main components, comparing Kovats Retention Indices (KRI) with reference libraries, and relevant literature (Adams, 2001; Barrero et al., 2005). Component concentrations were determined by semi-quantification through peak area integration from GC peaks and applying the correction factors of homologs series of C₉-C₁₀ n-alkanes and available authentic standards. For both EO and crude extract, a similar method was used, with only difference being in the sample volume and split mode (0.5 µL for the crude extract and a split ratio of 1:10).

Colorimetric total phenolic and flavonoid content

Colorimetric total phenolic (CTP)

The Folin–Ciocalteu method, slightly modified by Ostadi et al. (2020), was used to assess the total phenolic content, with results expressed in µg of gallic acid per mg of extract.

Colorimetric total flavonoid (CTF)

Following the study by Singleton et al. (1999), total flavonoids were calculated using the aluminum chloride method, with results presented in µg of quercetin per mg of extract.

Colorimetric antioxidant capacity

Antioxidants exists in various classes and forms in plants, including flavonoids, coumarins, benzoic acid derivatives, phenolic compounds, and carotenoids. Multiple spectrophotometric methods are used to measure the total antioxidant capacity and content, providing a comprehensive profile of the antioxidant levels and capacity of the substances being studied. This study evaluated the antioxidant potential of EOs and non-polar extracts using six different methods, such as DPPH^•, phenanthroline, and ABTS^•+, compared to five standards: BHA, BHT, ascorbic acid, quercetin, and α-tocopherol.

Colorimetric DPPH assay

The DPPH assay was determined using the methodology outlined by Moreno et al. (2000). The results were compared to positive standards including BHA, ascorbic acid, quercetin, α-tocopherol, and BHT, with absorbance readings taken at 517 nm. The concentration at which the DPPH radical could be 50% extinguished (IC₅₀) was determined by plotting percentage inhibitions of extinction, as shown in Equation (1). The results indicated that the tested sample exhibited a greater capacity for scavenging free radicals, as evidenced by the decrease in absorbance of the solution:

where, A_control is the absorbance of blank and the A_sample is that of extract or standard sample.

Colorimetric phenanthroline assay

Oyaizu (1986) utilized the phenanthroline technique to determine the activity, with the findings reported as A_0.50 (µg/mL). In order to compare the results obtained, positive standards such as BHA, ascorbic acid, quercetin, α-tocopherol, and BHT were used (Oyaizu, 1986).

Colorimetric ABTS assay

The technique used to assess the ABTS^•+ activity, and the findings were expressed as IC₅₀, based on the work of Özyürek et al. (2012). BHA, ascorbic acid, quercetin, α-tocopherol, and BHT were utilized as positive standards for comparing the acquired findings.

Anti-diabetic assay

Anti-α-amylase assay

This method was performed using the iodine/potassium iodide (IKI) method described by Zengin et al. (2014), with a few minor adjustments. A 96-well microplate was incubated for 10 min at 37°C with 50 μL of α-amylase solution (1U) prepared in phosphate buffer [PBS (pH 6.9 supplemented with 6 mM sodium chloride)] combined with 25 μL of each extract dissolved in MeOH, or acarbose as a positive control, at varying concentrations (6.25–400 μg/mL). Subsequently, 50 μL of 0.1% starch solution were added to initiate the reaction. The reaction mixture was then incubated at 37°C for 10 min. The reaction was stopped by adding HCl (25 μL, 1 M) and 100 μL of IKI solution. Similarly, a blank was prepared by adding the sample solution to all reagents without the enzyme α-amylase solution, and acarbose was used as a positive control. Absorbances were measured at 630 nm, with the absorbance of the blank subtracted from that of the samples (Zengin et al., 2014). The α-amylase inhibitory effect was calculated according to Equation (2):

where,

A_c: Absorbance of the mixture of starch + IKI + HCl + MeOH + phosphate buffer solutions

A_e: Absorbance of the mixture of the enzyme + starch + IKI + HCl+ MeOH solutions

A_s: Absorbance of the mixture of the enzyme + sample + starch + IKI + HCl solutions

A_b: Absorbance of the mixture of the sample + IKI + phosphate buffer solutions.

Molecular Docking

Selection and preparation of protein targets

The crystallographic 3D structures of the studied enzymes: human pancreatic alpha-amylase in complex with montbretin A (PDB ID: 4W93) and sugar beet alpha-glucosidase with acarbose (PDB ID: 3W37) were retrieved from the Protein Data Bank website (http://www.rcsb.org) in “.PDB” format. The native ligands and water molecules were removed, and polar hydrogen atoms and Kollman charges were added using MGL-AutoDockTools (ADT, v1.5.7) (Morris et al., 2009). This procedure was applied to both proteins and then saved as the dockable protein databank extension “.pdbqt” for docking simulations.

Retrieval and preparation of ligands

The 3D structures of bioactive compounds, detected by GC-MS analysis, were retrieved from the PubChem compound database (https://pubchem.ncbi.nlm.nih.gov/) in “.SDF” format. The collected ligands were then converted to the “.pdbqt” format using AutoDock Tools for molecular docking simulation.

Molecular docking execution

Autodock Vina was used for assessing binding affinity. The grid box centers, X, Y, and Z dimensions, used for the studied targets were (–9.632488, 4.340907, –23.107256) (20x20x20) and (0.108773, –1.916977, –23.053182) (20x20x20) for α-amylase (4W93) and α-glucosidase (3W37), respectively. The best binding pose with the lowest binding energy (kcal/mol) was chosen, and the interaction bonds were visualized in 3D and 2D images using Discovery Studio software. These parameters were applied to all studied compounds, and the top three interacted ones for each extract were graphically presented. The docking results were validated by redocking the native inhibitor into the enzyme’s active region (Trott and Olson, 2010).

ADMET properties

The study of drug absorption, distribution, metabolism, excretion, and toxicity (ADMET) is known as pharmacokinetics. Eliminating poor drug candidates is a crucial concept that not only determines the availability of a drug but also helps to prevent issues during in vivo research (Cheng et al., 2012; Yang et al., 2019). In this study, ADMET characteristics were determined using ADMETSAR.

Results and Discussion

Chemical composition

Essential oil

Hydrodistillation was used to extract the EO of J. phoenicea berries, allowing the extraction of 3.33% of the total mass of crude dry plant material., The yellow pale EO of J. phoenicea was compared to yields found in the literature, and Table 1 summarizes the different yields found. The chemical composition of the EO was analyzed by GC-MS, with a total of 50 compounds identified (Table 2) totalling 98%. Oxygenated monoterpenes were the most abundant components, despite the main components being derived from the monoterpenes hydrocarbons α-pinene (76.03%), β-pinene (2.75%), myrcene (1.73%), 3-carene (1.15%), and D-limonene (1.32%). Other compounds from sesquiterpene hydrocarbons such as caryophyllene (3.10%), α-humulene (1.25%), and germacrene D (1.7%) were also present. Most of the compounds identified in the EO of J. phoenicea berries in this study have been previously described for the same subspecies, although in different amounts.

The berries used in this study were in the same stage of maturation as those reported by the previous authors. α-Pinene predominated in all examined EOs, with levels ranging from 27% to 87.5%, depending on the local, harvesting season, and amount of plant material. As mentioned in the literature, the phoenicea group of juniper significantly supported these findings. These results are in agreement with those reported by research communities in this field, especially those focusing on this genre in Algeria. The majority of compounds remain the same, with variationsmainly in percentage. Slight differences were reported for some compounds like Dauca-4(11),7-diene that were reported in a study conducted by Amalich et al. (2015).

Non-polar extracts

In the present study, and for the first time to our knowledge, we provide a report on the phytochemical composition of cyclohexane and ethyl acetate extracts from Phoenician juniper. According to a study by Byrne et al. (2016) which categorized green solvents for extractions, separations, formulations, and reaction chemistry, the two solvents selected for this study were the most appropriate based on environmental, health, and safety (EHS) criteria as outlined by the three big pharmaceutical companies Pfizer, GlaxoSmithKline, and Sanofi (Byrne et al., 2016).

Classical maceration was used for 3 days to prepare the crude extract of non-polar solvents (cyclohexane and ethyl acetate) from Phoenician juniper berries. This process allowed the extraction of 7.95% and 8.7,% respectively, of the total mass of the crude dry plant material. The variation in yield was attributed to the polarities of the different compounds in the berries, a variation that has been noted in the literature, as seen in the study by Ennajar et al. (2009). The total flavonoid content was reported as micrograms of quercetin equivalents per milligram of extract (µg QE/mg) and the total phenolic content was reported as micrograms of gallic acid equivalents per milligram of extract (µg GAE/mg). The results are summarized in Table 2.

The chemical composition of cyclohexane and ethyl acetate extracts was analyzed using the same GC-MS. A total of 89 and 82 compounds were identified (Table 3), with a total percentage of identification of 67% and 68%, respectively. Despite the EO, the most abundant components for the non-polar extract are the oxygenated diterpenes, with amounts of 21.34% and 21.95% respectively. The analysis reveals the appearance of new compounds such as sterol (gamma.-Sitosterol), lignan (Anthricin), ether (Methyl tetratriacontyl ether) and ester (Acetic acid, 2,6,6-trimethyl-3-methylene-7-(3-oxobutylidene)oxepan-2-yl ester). For both extracts (cyclohexane and ethyl acetate), the main components were derived from the oxygenated diterpenes such as communic acid (23.66% and 22.38%), abitale (110.75% and 13.22%), abitole (4.25% and 4.60%), and abieta-7,13-diene-3-one (4.02% and 4.55%), respectively. Despite the existence of α-pinene with significant amounts for both extracts with percentages of 13.01% and 16.21 %, which proves the presence of J. phoenicea. The GC-MS analysis of the non-polar extracts reveals chemical variations in the content compared with the EO, noting the presence of diterpene hydrocarbons, oxygenated diterpenes in significant amounts, and sesquiterpene oxygen, which is present in significant quantities as in the EO. The total ion chromatogram of representative samples in which the main compounds are pointed out is displayed in Figure 1. The complete EO composition is shown in Table 2.

Figure 1. GC-MS total ion chromatogrammes from a representative sample of berries of Juniperus phoenicea L. (A) Essential oil. (B) Cyclohexane extract. (C) Ethyl acetate extract.

The preliminary study examined the chemical composition, antioxidants, antidiabetic effects, and in silico modeling of the EO and non-polar extracts of Algerian J. phoenicea berries. The results of this study showed that the EO was characterized by α-pinene, caryophyllene, β-pinene, myrcene, and germacrene D. These results, although slightly different in profile from earlier research, are generally consistent with previous findings (Amalich et al., 2015; Angioni et al., 2003; Medini et al., 2011; Menaceur et al., 2013, Abdelli et al. 2018). The minor differences in EO composition could be attributed to genotype-environment interactions and the phenological stage of the species under study. Plants may produce certain chemical biomolecules as an adaptive response to changing environmental conditions. The results also indicated that the principal elements mentioned above were present in the non-polar extracts. While there were no previous studies mentioning the composition of non-polar extracts, especially for cyclohexane and ethyl acetate, the current outcome was consistent with the literature. The main constituents of most J. phoenicea species include abitale, abitole abieta-7,13-diene-3-one, and α-pinene (Boudiba et al., 2021 Bouyahyaoui et al., 2016; Ennajar et al., 2010).

According to literature data, extracts such as ethyl acetate, ethanol, and methanol solvents had the highest content of total phenols and flavonoids (Bouyahyaoui et al., 2016). This could be explained by the fact that polar solvents have a higher affinity for these components than non-polar solvents. On the other hand, variations in phenolic content might be attributed to different extraction methods, the standard solution used, geographical location, and climate conditions. Overall, the study reported that total phenolic and total flavonoids levels were relatively low.

Antioxidant activities

A total of three in vitro antioxidant tests (DPPH, ABTS, and phenanthroline) were conducted to compare with five standards: BHA, BHT, ascorbic acid, quercetin, and α-tocopherol. The results were obtained using linear regression analysis, and the IC₅₀ and A_0.5 values are presented in Table 3 as the mean values ± SD of three measurements. In most tests, the IC₅₀ and A_0.50 values differed significantly from the standard ones.

Antioxidants operate through two distinct processes: electron transfer (single electron transfer, SET) and hydrogen donation (hydrogen atom transfer, or HAT). Our investigation utilized several techniques (DPPH^•, ABTS^•+, and phenanthroline) to gather more data on the antioxidant activity of the examined extracts. The differences between these tests lie in their modes of action and reaction mechanisms.

In the DPPH assay, all samples showed lower activity (IC₅₀ > 800 g/mL) compared to the standards. The lower activity exhibited by all samples was likely due to the high concentration of terpene hydrocarbons (53%) in the EO. Similarly, the non-polar extracts showed lower activity despite containing a high concentration of oxygenated terpene (40.23%–42.68%). This can be attributed to the fact that the inhibitory activity of the DPPH radical is not solely dependent on the total polyphenol content, but rather on specific polyphenols with unique chemical structures (Sánchez-Vioque et al., 2013). In other words, DPPH is highly selective and only reacts with flavonoids that contain hydroxyl groups in the B cycle (Yokozawa et al., 1998), as well as aromatic acids with multiple hydroxyl groups, which were not observed in our case. These results were in agreement with existing literature on similar species, such as the antioxidant activity studies conducted by Ennajar et al. (2009) and Taviano et al. (2011) on methanol and water extracts of five different Turkish Juniperus species, which showed weaker activity compared to the extracts examined in this study (Ennajar et al., 2009; Taviano et al., 2011). On the other hand, compared to the polar extract case by Ghouti et al. (2018) where hydroethanolic and infusion extracts were used, our study demonstrated interesting DPPH activity that was two to three times more effective than Trolox. Additionally, research by Keskes et al. (2014) reported that the methanol extract exhibited a stronger ability to scavenge DPPH radicals (EC₅₀ value = 2 μg/mL).

Regarding the phenanthroline assay, the EO and non-polar extracts of J. phoenicea showed a weak ability to reduce iron with A_0.50 value of >200 µg/mL for the high concentration. This is in contrast to the results of standards such as ascorbic acid, quercetin, BHA, BHT, and α-tocopherol, which demonstrated a substantial reducing impact on iron. Our results indicated a low absorbance for both extracts (200 μg/L – 0.35), which is comparable to findings in the literature (Byrne et al., 2016; Menaceur et al., 2013). The phenolic content of the extracts was closely correlated with their poor activity in the DPPH radical scavenging and phenanthroline tests. Numerous investigations have shown a strong positive association between total phenolic content, phenanthroline, and anti-DPPH action.

In the ABTS assay, it was observed that the extracts of J. phoenicea berries exhibited a stronger free radical scavenging capacity compared to DPPH for all samples. This can be attributed to the versatility of the radical cation ABTS, which is soluble in both water and organic solvents. This allows evaluation of anti-radical activity for both hydrophilic and lipophilic compounds. Additionally, the planar form of ABTS makes it more susceptible to attack compared to DPPH, which is more complex (Gülçin, 2010). When compared to the five positive standards, ABTS still showed minimal activity. The percentage of ABTS-scavenging ability of the EO from Tunisian J. pheonicea berries was determined to be 128.7 ± 3.8 μg/mL compared to vitamin C at 1.9 ± 0.9 μg/mL. These results were slightly lower than previous investigations. Despite their low concentration, no correlation was found between phenolic compounds and ABTS^•+ activity. Certain monoterpene alcohols, ethers, ketones, aldehydes, and non-polar bioactive substances like pigments, were associated with ABTS^•+ activity (Edris, 2007; Merola et al., 2017).

In general, the antioxidant capacity of both EO and non-polar extracts (cyclohexane and ethyl acetate) of J. phoenicea berries was tested in three different assays (DPPH, ABTS, and phenanthroline). It can be concluded that the EO exhibits no antioxidant activity despite increases in concentration. This could be explained by the complex mixture of EOs that may exhibit antioxidant activity different from that of their major components tested alone. Literature reports suggest that the antioxidant activity of EO may result from a complex interaction between different constituents, producing additive, synergistic, or antagonistic effects, even at low concentrations (Ennajar et al., 2010). The non-polar extracts show some antioxidant activities but are still insignificant compared to five standards. This could be due to the low phenolic content and the mixture of extracts preventing the reduction of both DPPH and ABTS, and even forming a complex with phenanthroline for reducing the ferric ion. Table 3 presents the different values obtained in the antioxidant activities.

Anti-diabetic activity

One of the most crucial approaches to treating type 2 diabetes (T2D) is controlling postprandial hyperglycemia. This is achieved by inhibiting the enzymes that hydrolyze carbohydrates, primarily intestinal α-glucosidase (EC 3.2.1.21.), and pancreatic α-amylase (EC 3.2.1.1.1.). This process may reduce blood levels of glucose in circulation by delaying the gastrointestinal absorption of dietary carbohydrates (Al-Ishaq et al., 2019; Deo et al., 2016; Priscilla et al., 2014). Due to these factors, the present investigation examined our plant’s potential as an antidiabetic by evaluating its inhibitory effect on the enzymatic activity of α-amylase. The primary enzyme in humans, α-amylase, breaks down starch into simple sugars such as dextrin, maltotriose, maltose, and glucose. α-amylase inhibitors are believed to enhance glucose tolerance in diabetic patients, although the activity of the enzyme has not been directly linked to the genesis of diabetes. To improve diabetes management, great effort has been put forward in recent years to discover α-amylase inhibitors that are clinically effective (Loizzo et al., 2007). The results are expressed as mean ± SD and provided in Table 4.

As demonstrated by the results obtained, both extracts showed a strong dose-dependent inhibitory effect against α-amylase, with maximum inhibition rates >75.0% at the high tested dose of 400 μg/mL. Table 4 shows the significant increase in the α-amylase inhibitory effect with an IC₅₀ of 186.91 ± 5.74 μg/mLfor the cyclohexane extract compared to 351.48 ± 0.17 μg/mL for the ethyl acetate extract. This difference is nearly 20 times greater than that of acarbose, used as a positive control. Interestingly, acarbose’s inhibitory efficacy against α-amylase is about 15 times less (3650.9 ± 1.7 μg/mL) than the ethyl acetate extract. However, pure quercetin demonstrated effective inhibition of α-amylase with the lowest IC₅₀ value of 4.3 ± 0.2 μg/mL, surpassing the activity observed in our investigation against α-amylase. These results suggest the strong anti-diabetic properties of both non-polar extracts (cyclohexane and ethyl acetate) of J. phoenicea berries. In contrast, the EO showed a maximum inhibition of 33% at a concentration of 16 mg/L, providing a reliable value compared to the extracts.

Regrettably, the literature lacked significant information about the potential inhibitory effects of Juniperus species against α-amylase, while few reports on J. phoenicea berries’ α-amylase inhibitory activities were published. Interestingly, Loizzo et al. (2007) investigated the α-amylase inhibitory activities of J. oxycedrus berries EO. The bio-assay was adopted and modified based on the work of Conforti et al. (2005). Their results indicated moderate activities of berries EO with IC₅₀ value of >25 µL/mL (Loizzo et al., 2007). In the same context, the work of Orhan et al. (2012) evaluated the hypoglycemic and antidiabetic activity of J. oxycedrus subsp . Oxycedrus berries. They studied the oral administration of berry extracts using in vivo models in normal and glucose-hyperglycemic rats. They indicated that ethanol extract showed a higher and continuous hypoglycemic effect (7.7%–23.6%) in STZ-diabetic rats. For normal rats, the ethanol extract exhibited a moderate hypoglycemic effect (12.8%–13.0%). The n-ButOH extract gives better results, especially for fraction C, which exhibited the most effective antidiabetic activity compared with others between 18.6% and 26.4%. This activity was caused by the presence of bioactive compounds such as shikimic acid (Orhan et al., 2012).

Banerjee et al. (2013) studied the impact of a methanolic extract of J. communis (100 mg/kg, 200 mg/kg) administered orally to streptozotocin nicotinamide-induced diabetic rats. They found that the extract had dose-dependent and significant anti-diabetic and antihyperlipidemic properties, supporting its potential as an effective drug for type-2 diabetes. In another study conducted by Sánchez de Medina et al. (1994) and Swanston-Flatt et al. (1990), it was shown that decoction and infusion of J. communis fruits led to significant decrease in blood glucose levels in both normoglycemic rats and STZ-induced diabetic rats and mice.

Overall, research on the antidiabetic effects of Juniperus species, both in vitro and in vivo, suggests strong inhibitory activity against α-amylase and α-glucosidase. These findings have the potential to improve hyperglycemia in individuals with type 2 diabetics, highlighting the need for further investigation into the mechanisms of action and active components of these plants.

Molecular docking

Molecular docking is an increasingly common method in structure-based drug design (SBDD) because it can accurately predict the binding conformation of small molecules on their corresponding target proteins (Wang et al., 2020). This technique facilitates the analysis and visualization of significant intermolecular biological processes, including the binding of ligands to their corresponding receptors and interactions that maintain the stability of the ligand-receptor system. Furthermore, it is possible to rank the “docked” conformations of the simulated ligands by using their binding free energies since molecular docking techniques quantify receptor-ligand binding energies (Huang and Zou, 2010; Kadi et al., 2023).

To understand the binding efficacy of compounds from J. phoenicea L. berries’ EOl and non-polar extracts (cyclohexane and ethyl acetate) to the selected proteins associated with diabetes (α-amylase 4W93 and α-glucosidase 3W37), molecular docking simulations were conducted using Autodock Vina. The native ligands montbretin A and acarbose were used for comparison, and their binding energies were -11.0 and -10.0 kcal/mol, respectively. Among those results, binding energy values varied from -4.6 to -10.0 kcal/mol. The best result in this range was measured in terms of the score value of the native ligand, and the three best compounds from each extract composition were selected.

Our findings showed potent molecules such as Humulene epoxide II, Caryophyllene, and cedrene against α-amylase, valencene, γ-muurolene, and δ-cadinene for α-glucosidase in the EO composition. In the cyclohexane extract, the most potent ones were androst-5-ene-3,17-diol, 3-acetate (3.β.,17.β.), γ-sitosterol and abieta-8,11,13-trien-7-one, for α-amylase and androst-5-ene-3,17-diol,3-acetate (3.β.,17.β.), neoabietyl acetate, and γ-Sitosterol for α-glucosidase. In the ethyl acetate extract, γ-Sitosterol, Neoabietol, and Aromadendrene epoxide showed a potent affinity against α-amylase while abitatriene, γ-sitosterol, and neoabietol for α-glucosidase.

Major compounds for each extract showed good results but this varied for each protein, possibly due to the amino acid construction of each protein. Although there have been fluctuations in this range, the result of diterpene and sesquiterpene is still high compared to monoterpene, reflecting the fact that their structure has many hydroxyl groups that facilitate ligands in forming hydrogen bonds with a free residue of the receptor. Additionally, Table 4 shows the best receptor of these bioactive compounds in J. phoenicea L. berries, α-amylase 4W93 followed by α-glucosidase 3W37. Interaction analysis in 3D and 2D images was performed to understand the binding mode of the studied compounds in interaction with α-amylase 4W93 and α-glucosidase 3W37. Figures 2 and 3 show the different interactions.

Figure 2. The molecular interactions in 2D and 3D images between the studied compounds and the main α-amylase 3W93 using Autodock Vina. Essentail oil: (A) Cedrene, (B) Caryophyllene, (C) Humulene epoxide II. Cyclohexane extract: (D) Androst-5-ene-3,17-diol, 3-acetate, (E) g-Sitosterol, (F) Abieta-8,11,13-trien-7-one. Ethyl acetate extract: (G) g-Sitosterol, (H) Neoabietol, (I) Aromadendrene epoxide.

Figure 3. The molecular interactions in 2D and 3D images between the studied compounds and the main α-glucosidase 4W37 using Autodock Vina. Essential oil: (A) γ-Muurolene, (B) δ-Cadinene, (C) Valencene. Cyclohexane extract : (D) Androst-5-ene-3,17-diol, 3-acetate, (E) Neoabietyl acetate, (F) γ-Sitosterol. Ethyl acetate extract: (G) Abitatriene, (H) Neoabietyl acetate, (I) γ-Sitosterol.

Our results showed that in the case of ethyl acetate extract, γ-Sitosterol is the most interactive ligand with α-amylase 4W93. This interaction is facilitated through two positive ionizable interactions formed with the catalytic triad of hydrogen-donor, with ASP197 and GLU233, five interactions (one π-sigma and four π-alkyl) with TRP59, and one alkyl interaction observed with LEU162. On the other hand, in the case of α-glucosidase 3W37, the most interactive ligand was abitatriene. This interaction involved three interactions with the first ring, one π-anion with ASP568, one intriguing π-sulfur interaction with MET470, and one π-π T-shaped interaction with TRP432.

For the cyclohexane extract, the most interactive ligand in this study, for α-amylase was androst-5-ene-3,17-diol,3-acetate (3.β.,17.β.). It showed two hydrogen donors intracting with HIS201 and GLN63, two π-π interactions with TRP59, and three hydrophobic alkyl interactions, two with ALA198, and one with LEU162. In the case of EO, we observed lower values for α-amylase activity, which was confirmed by molecular docking. Lower values were also registered for both proteins α-amylase 4W93 and α-glucosidase 3W37. Cedrene was identified as the best compound with the highest binding energy with α-amylase, forming two π-sigma interactions with TYR62 and TRP59, and two π-alkyl interactions with TRP58 and TYP92. On the other hand, valencene showed better binding energy with α-glucosidase 3W37, with four interactions with the same amino acid TRP59, two being π-sigma and two being π-alkyl. Table 5 summarizes the different values of binding energies, type, and distances of each interaction for both proteins α-amylase 4W93 and α-glucosidase 3W37.

Table 5. Different values of banding energies, type and distances of each interaction for both proteins α-amylase 4W93 and α-glucosidase 3W37.

Human pancreatic α-amylase is 56 kD protein composed of 496 amino acids as single peptide chains divided into three domains: domain A (residues 1–99 and 169–404), the largest part where the active site is located as a V-shaped cleft in the catalytic domain A that binds to the substrate, domain B (residues 100–168) emerges from domain A and is located between the third β-strand and the α-helix of the β-barrel of the catalytic domain, maintaining the conformation of the enzyme, and domain C (residues 405–496) is an anti-parallel β-sheet of domain A loosely associated with other domains (Brayer et al., 1995; Mótyán et al., 2011; Nahoum et al., 2000; Williams et al., 2012). According to the results of our research, domain A accounts for the majority of contacts with a particular level of efficacy such as those of ASP197 and GLU233 except for the interaction of LEU162 which belongs to domain B and presents a lower hydrophobic interaction.

In the same context, the structures of α-glucosidases are complex. There are four primary domains in this structure: two C-terminal domains, a catalytic domain of the (β/α) 8-barrel, and an N-terminal domain. The inserts 1 and 2 of the catalytic domain are situated next to β-stands 3 and 4, respectively. Except for insert 1, the overall architectures of these subunits are nearly identical. The β-barrel loops generate the active site pocket (Subsite-1) in the catalytic domain, and the residues involved in subsite-1 formation have been extensively conserved throughout the subunits of α-glucosidases (Kashtoh and Baek, 2022). In our study, the most interesting interaction that contributes to enzyme/ligand binding was established with SER497, ARG552, and LYS506, which provide hydrogen-donor interactions in the case of Androst-5-ene-3, 17-diol, 3-acetate, (3.β.,17.β.), Neoabietyl acetate, and γ-Sitosterol.

The compounds (best docked) from each extract were further investigated using the online software ADMETSAR, which provides more knowledge about pharmacokinetic, physiochemical behavior, and druglikeness. ADMET properties of those bioactive compounds, which include Caco-2 cell permeability, brain/blood barrier, human intestinal absorption, AMES mutagenesis, and carcinogenicity, were elucidated in the present study. Results of ADMETSAR were analyzed and tabulated in Tables 6 and 7. The molecular weight of all the bioactive compounds was found to be less than 500, anticipating their easy transportation, absorption, and diffusion. Hydrogen bonding describes drug permeability; poor permeation correlates to more than 5 H-bond donors and 10 H-bond acceptors, which were not detected in our results, indicating the best permeation for all tested compounds. Despite that, not all our tested compounds are in the range of acceptable ALog P value (≤ 5), but still orally absorbed based on the Lipinsky rule, an important consideration of the drug’s likeness. Furthermore, a drug molecule is expected to be in an aqueous solubility range of –1 to –5, and the Log S values of all the selected compounds fall within the range except for the case of Abitatriene, which was observed with a value of -6.20 (Bergenhem, 2011). The different results are shown in Table 6. The determination of pharmacological qualities is crucial to identify bioactive molecules with acceptable pharmaceutical properties and then address their drug availability. This involves researching absorption, distribution, metabolism, excretion, and toxicity (Ghannay et al., 2020).

ADMET pharmacokinetic properties reveal that all selected compounds had better Human Intestinal Adsorption (HIA) with a score of > 99%. Greater HIA indicates that the compound could be better absorbed from the intestinal tract upon oral administration. In addition to the HIA, all selected compounds had a positive value of Caco-2 permeability and blood-brain barrier BBB. It is observed that the lower value is registered for both γ-Sitosterol and Androst-5 -ene-3,17-diol, 3-acetate,(3.β.,17.β.). The admetSAR online tool was used to evaluate the toxicological properties of the selected compounds, as the safety of the compounds is an important parameter for becoming a good drug (Cheng et al., 2012), Table 7 presents the different values of the pharmacokinetic profile. In this study, no AMES toxicity and carcinogens were identified as a threat. An exception was observed for respiratory toxicity, where some of those compounds posed a positive threat against the respiratory system, such as cedrene, humulene epoxide II, androst-5-ene-3,17-diol,3-acetate (3.β.,17.β.), abieta-8,11,13-trien-7-one, neoabietol, neoabietyl acetate, and especially γ-sitosterol.

Conclusions

The medicinal plants include natural products such as phytochemicals, used in various medical applications including their antidiabetic effects. Traditionally, Juniperus phoenicea berries extracts are considered promising natural medicinal agents and widely used for the treatment of type 2 diabetes (T2D). The findings obtained in this study are divided into three main axes:

• Analysis of the chemical composition of EO, cyclohexane extract, and ethyl acetate extract allowed the identification of50, 60, and 56 compounds, respectively, representing over 98% of the essential oil. Oxygenated monoterpenes are the major components, representing 68% of both non-polar extracts, with oxygenated diterpenes being the major components of the Juniperus phoenicea berries from the Aflou region.

• In vitro assays for antidiabetic effects, based on the α-amylase inhibitory capability of both non-polar extracts, exhibited a strong inhibitory effect against the α-amylase enzyme in a dose-dependent manner. Maximum inhibition rates of over 75.0% were observed at the high tested dose of 400 mg/mL against α-amylase, with good IC₅₀ values for both cyclohexane and ethyl acetate extracts (IC₅₀ = 186.91 ± 5.74 mg/mL, IC₅₀ = 351.48 ± 0.17 mg/mL) respectively. In comparison, the EO showed a maximum inhibition rate of 33.0% at the high tested dose.

• The study was further enhanced with molecular docking analysis to get a comprehensive understanding of the phenomenon. Various compounds showed promising binding affinity to specific proteins (α-amylase and α-glucosidase), with, all compounds displaying a favorable ADMET profile, indicating their potential to serve as lead compounds in drug discovery. Comprehensive in silico studies, including molecular docking studies, were conducted to evaluate and support the in vitro findings. However, further studies using different antidiabetic models, especially in vivo models, are needed to confirm the beneficial qualities of these extracts.