Antihyperglycemic Effect of Lavandula pedunculata: In Vivo, In Vitro and Ex Vivo Approaches

Lavandula pedunculata (Mill.) Cav. (LP) is one of lavender species traditionally used in Morocco to prevent or cure diabetes, alone or in the form of polyherbal preparations (PHP). Therefore, the primary objective of this study was to test the antihyperglycemic effect of the aqueous extract of LP, alone and in combination with Punica granatum L. (PG) and Trigonella foenum-graecum L. (FGK). The secondary objective was to explore some mechanisms of action on the digestive functions. The antihyperglycemic effect of the aqueous extract of LP, alone and in combination with PG and FGK, was studied in vivo using an oral glucose tolerance test (OGTT). In addition, LP extract was tested on the activities of some digestive enzymes (pancreatic α-amylase and intestinal α-glucosidase) in vitro and on the intestinal absorption of glucose ex vivo using a short-circuit current (Isc) technique. Acute and chronic oral administration of LP aqueous extract reduced the peak of the glucose concentration (30 min, p < 0.01) and the area under the curve (AUC, p < 0.01). The effect of LP + PG was at the same amplitude to that of the positive control Metformin (MET). LP aqueous extract inhibited the pancreatic α-amylase with an IC50 almost identical to acarbose (0.44 ± 0.05 mg/mL and 0.36 ± 0.02 mg/mL, respectively), as well as the intestinal α-glucosidase, (IC50 = 131 ± 20 µg/mL) and the intestinal glucose absorption (IC50 = 81.28 ± 4.01 µg/mL) in concentration-dependent manners. LP aqueous extract exhibited potent actions on hyperglycemia, with an inhibition on digestive enzymes and glucose absorption. In addition, the combination with PG and FGK enhanced oral glucose tolerance in rats. These findings back up the traditional use of LP in type 2 diabetes treatment and the effectiveness of the alternative and combinative poly-phytotherapy (ACPP).


Plant Material
The

Preparation of the Aqueous Extracts
Air dried flowering tops of LP or PG pericarps were mixed with distilled water (1:20; w/v) and heated for one hour at 70 to 80 • C. The mixtures were then filtered and the obtained filtrates were dried in the oven at 70 • C until obtaining the dry extract powder. The extracts were put in closed flasks away from light and humidity until further use.
The aqueous extracts were solubilized in a methanol / water mixture (1:1, v/v), so as to obtain a concentration of 1 mg/mL, then they were filtered through 0.2 µm PTFE filter. For each analysis, 4 µL of the extract were injected. The temperature was set at 30 • C and flow rate was set at 0.3 mL/min.
This analysis was carried out on few standards chosen according to bibliographic data, and which were injected under the same conditions as those of the extracts. These standards are cinnamic acid, luteolin, apigenin, myricetin, ferulic acid, protocatechuic acid, vanillic acid, chlorogenic acid, caffeic acid, rosmarinic acid, gallic acid, herniarine, and coumarin.
Compounds were identified by matching their retention time, UV spectrum and molecular weight to those of the used standards.

Animals
For OGTT and toxicological studies, healthy adult Wistar rats (150-250 g body weight, 2-3 months of age) and Swiss albino mice (20-30 g body weight) were provided by the animal house of the Faculty of Science, Mohammed First University, Oujda-Morocco.
For the intestinal glucose absorption studies, 7 week old male mice (Black mice C57BL/6JRj, 20-25 g) were purchased from Janvier SASA (Route des chênes, Le Genest-st-Isle, St Berthevin, France) and were acclimated for a week in the animal house conditions of the Faculty of Pharmacy, Lille University, France (approval N • D5935010). The animals were grouped in polycarbonate cages with free access to food and water, in an environmentally controlled room (22)(23)(24)(25)(26) • C, ventilation, 12/12 h light/dark cycle). Experiments on animals were carried out in accordance with the internationally approved "Guide for the care and use of laboratory animals" prepared by the National Academy of Sciences and published by the National Institutes of Health [22]. All efforts were made to minimize animal suffering and the number of animals used. For OGTT studies, ethical approval was obtained from Mohammed First University, Oujda-Morocco (31/2019/LBBEH-05 and 21/011/2019).

Inhibition Assay of Pancreatic α-Amylase Enzyme Activity
The inhibitory effect of LP aqueous extract on the enzymatic activity of pancreatic α-amylase was studied in vitro according to the method described by  with some modifications [23]. Then, 200 µL of the α-amylase enzyme solution (13 IU) and 200 µL of phosphate buffer (0.02 M; pH = 6.9) were placed in different tubes. On one hand, 200 µL of the LP extract were added at different concentrations: 2.27, 1.82, 1.36, 0.91, 0.45, and 0.23 mg/mL. On the other hand, 200 µL of acarbose were used as a positive control at several concentrations: 2.27, 0.91, 0.45, 0.23, 0.11, and 0.06 mg/mL. All the reagents were dissolved in phosphate buffer and 200 µL of this solution were used as a control. The mixtures were pre-incubated for 10 min at 37 • C before adding 200 µL of starch (1%) to each tube. The different solutions were then incubated for 20 min at 37 • C. Then, 600 µL of dinitrosalicylic acid (DNSA) color reagent (2.5%) was added to stop the enzymatic reaction. The tubes were then incubated for 8 min at 100 • C, and they were put in an ice-water bath for few minutes. At the end, the mixtures were diluted by the addition of 1 mL of distilled water and the absorbance of each tube was measured at 540 nm. The inhibition percentage was then calculated using the formula below: where IP: inhibition percentage (%); A control : absorbance of the final mixture without inhibitor; A test : absorbance of the final mixture in the presence of LP extract or acarbose. The concentration of the tested substance (LP extract/acarbose) inhibiting 50% (IC 50 ) of the enzymatic activity of α-amylase was determined graphically.

Inhibition Assay of Intestinal α-Glucosidase Enzyme Activity
The inhibitory effect of LP aqueous extract on the enzymatic activity of intestinal α-glucosidase was evaluated in vitro according to the method described by  with some modifications [23]. The method consists of the quantification of D-glucose released from sucrose degradation. 10 µL of different concentrations of acarbose solutions (41, 82, 165, 328 and 656 µg/mL) or LP extract solutions (80, 170, 250, 330 and 650 µg/mL) were added to a mixture containing 100 µL of sucrose (50 mM), 100 µL of α-glucosidase enzyme solution (10 IU), and 1000 µL of phosphate buffer (50 mM; pH = 7.5). All the reagents were dissolved in phosphate buffer and 10 µL of this solution was used as a control. The mixtures were then incubated in a water bath at 37 • C for 25 min. Immediately after incubation, the different solutions were heated for 5 min at 100 • C to stop the enzymatic reaction. The D-glucose oxidase method using GOD/POD kit was used to determine the released D-glucose. The absorbance was measured at 500 nm and the inhibition percentage was calculated using the formula below: where IP: inhibition percentage (%); A control : absorbance of the final mixture without inhibitor; A test : absorbance of the final mixture in the presence of LP extract or acarbose. The concentration of the tested substance (LP extract/acarbose) inhibiting 50% (IC 50 ) of the enzymatic activity of α-glucosidase was determined graphically.

Intestinal Glucose Absorption
D-glucose absorption through the intestinal tissue was studied ex vivo using Ussing chambers. The Black mice were left on empty stomachs with free access to drinking water for 18 h before the experiments. After killing the animals by CO 2 inhalation, their intestines were removed and emptied of their content using a cold Ringer's solution. Then the jejunum was cut into small fragments (1-1.5 cm) and opened along the mesenteric border. Every fragment was placed as a flat sheet between the two sides of the plexiglas Ussing chambers (exposed area: 0.3 cm 2 ), allowing a mucosal compartment (luminal side) and a serosal compartment (serosal side) to be determined (blood side). Every half-chamber was filled in with 3 mL of Ringer's solution (pH = 7.4) thermostatically controlled at 37 • C. The oxygen saturation of this solution was constantly maintained by a permanent bubbling with carbogen (95% O 2 /5% CO 2 ) circulation the chambers [24].
The spontaneous transmural potential difference (PD) was measured using agar bridges containing KCl solution (3 M) in agar (4 percent, w/v) to represent asymmetry of electrical charges between the mucosal and serosal sides of the gut. These bridges were placed on both sides of the tissue and connected to calomel half-cells, connected to a highimpedance voltmeter. The PD was short-circuited and maintained at 0 mV by a short-circuit current (I sc ) throughout the experiment via two stainless steel 316L working electrodes directly placed in each compartment [25], and linked to a voltage-clamp system (JFD-1V, Laboratoires TBC, France). The delivered I sc (in µA/cm 2 ), that was recorded continuously using Biodaqsoft software and corrected for fluid resistance. It represents the sum of the net ion fluxes transported across the epithelium in the absence of an electrochemical gradient (mainly Cl − , Na + and HCO 3 − ). LP extract solutions were prepared extemporaneously when added to the mucosal compartment (prewarmed at 37 • C) in a volume of 250 µL, 5 min before adding the glucose (10 mM) to obtain final concentrations of 0.01 to 1000 µg/mL (glucose was replaced with mannitol at 10 mM in the serosal compartment).
Results are expressed as the difference between the I sc plateau measured after 20 min of the glucose addition (without extract), and the I sc plateau measured after the glucose addition preceded by the addition of LP extract. The principle is to stimulate the intestinal absorption of glucose (increase of I sc ) in the presence and absence of LP extract at different concentrations to detect any blocking of this absorption.

Antihyperglycemic Study in Healthy Rats Acute Oral Glucose Tolerance Test
Acute OGTT was carried out using normal Wistar rats, left on empty stomachs for 14 h with free access to water, and which were grouped in six groups of four rats each (two males and two females). The control group (CT) was orally administered distilled water (10 mL/kg b.w.). The five test groups (LP, PG, FGK, GO, MET) were administered the aqueous extracts at 1 g/kg b.w. The rats were force-fed with the different products immediately after the measurement of their glycaemia; 30 min later, another measurement of the glycaemia was carried out, and then the rats were overloaded with D-glucose at a dose of 2 g/kg b.w. Subsequently, the variation in blood sugar was monitored every 30 min for two hours [23].

Chronic Oral Glucose Tolerance Test
To assess the effect of the repeated extracts administration, a chronic OGTT was performed using the same groups of rats used above (Acute oral glucose tolerance test). The testing products were orally administered to rats once a day for four weeks (chronic treatment). After the experimental period, the rats were left on empty stomachs for 14 h, and the glycaemia measurement was carried out according to the same procedure described above (Acute oral glucose tolerance test).

Acute Oral Glucose Tolerance Test for Plant Mixtures
The principle of this test is similar to the one used for individual plants which is described above (Acute oral glucose tolerance test). Four groups were studied; starting with the control group which was administered distilled water orally (10 mL/kg b.w.). Test group (LP + PG) was administered a mixture containing 50% of LP extract and 50% of PG extract at a final dose of 1 g/kg b.w. Test group (LP + FGK) was administered a mixture containing 50% of LP extract and 50% of FGK extract at a final dose of 1 g/kg b.w. MET group was administered MET (1 g/kg b.w.).

Acute Toxicity
Swiss albino mice were used to test the acute toxicity of lavender aqueous extract (LP). Mice were left on empty stomachs for 16 h before the test, and then they were randomly divided into five groups of six mice each (three males and three females). The control group was force-fed with distilled water. Test groups 1, 2, 3, 4, and 5 were administered different concentrations of LP extract (1, 3, 5, 7, and 10 g/kg b.w., respectively). The mice were given the extract only once, and then they were put under observation for 15 days. Any clinical signs and/or behavioral changes would have been considered signs of acute toxicity.

Subchronic Toxicity
Subchronic toxicity test was performed using the same groups of rats used for the chronic OGTT (Chronic oral glucose tolerance test). The body weights of the rats were measured weekly as they were force-fed with the plant extracts. After the experimental period, the rats were left on empty stomachs for 14 h before being sacrificed to collect their blood and organs. The biochemical parameters of the collected blood were then measured in plasma: Alanine aminotransferase (ALT), aspartate aminotransferase (AST), direct bilirubin, total bilirubin, albumin, total cholesterol, triglycerides, high-density lipoproteins (HDL-c), low-density lipoproteins (LDL-c), plasma glucose, lipase, total protein, creatinine, urea, uric acid, calcium, and phosphate. All tests were performed with the COBAS INTEGRA ® 400-Plus analyzer using standard clinical diagnostic kits. Also, the relative weights of the livers and the kidneys were measured.

Histology
For histological examination, different tissues (stomach, small intestine, liver, spleen, and kidneys of rats) were fixed in 10% neutral formalin, embedded in paraffin, sectioned to a thickness of approximately 5 µm, stained with hematoxylin and eosin, and examined for histopathological changes under the microscope (Olympus, Tokyo, Japan).

Statistical Analysis
Pharmacological responses for separate experiments using n tissues are presented as means ± SEM (standard error of the mean). Graphs of the concentration-response curves were determined using nonlinear regression and were fitted to the Hill equation by an iterative least-squares method (GraphPad Prism version 8.0 for Windows, GraphPad Software, San Diego, CA, USA). One-way analysis of variance (ANOVA) was performed for the comparison of the different effects with the control (Dunnett) and for the multiplegroup comparisons (Newman-Keuls). GraphPad Prism computed the AUC using the trapezoid rule. Statistical significance was set as p < 0.05.

UHPLC Analysis
The chemical composition of LP aqueous extract was determined using UHPLC-MS. The results (Table 1) show that the most abundant compound in this extract is rosmarinic acid. Figure 1 shows the UHPLC-MS chromatograms of rosmarinic acid detected in LP aqueous extract. Moreover, coumarin, protocatechuic acid, herniarin, caffeic acid, apigenin, luteolin, myricetin, and chlorogenic acid are present in this extract but with a lower abundance, as well as gallic acid and cinnamic acid that are present as traces. We can notice that vanillic acid and ferulic acid are absent in LP aqueous extract.

Oral Glucose Tolerance Test
To evaluate the effect of LP aqueous extract on the systemic glucose homeostasis, an OGTT was performed in conscious fasted rats after acute and chronic oral administration of this extract (LP), and we compared its effect with other plant aqueous extracts such as PG, FGK, and GO.
During the OGTT, acute oral dose of LP decreased the peak glucose concentration (30 min, p < 0.01, Figure 2A) and the AUC (p < 0.01, Figure 2B). We noticed that this effect was lower when compared to that of GO and the positive control MET (p < 0.001), but remained close to that of PG (p < 0.01).

Oral Glucose Tolerance Test
To evaluate the effect of LP aqueous extract on the systemic glucose homeostasis, an OGTT was performed in conscious fasted rats after acute and chronic oral administration of this extract (LP), and we compared its effect with other plant aqueous extracts such as PG, FGK, and GO.
During the OGTT, acute oral dose of LP decreased the peak glucose concentration (30 min, p < 0.01, Figure 2A) and the AUC (p < 0.01, Figure 2B). We noticed that this effect was lower when compared to that of GO and the positive control MET (p < 0.001), but remained close to that of PG (p < 0.01).
The same findings were observed after the chronic OGTT with LP when compared to the control (p < 0.01, Figure 2C,D). Finally, no statistical differences between male and female animals were observed (data not illustrated).

Oral Glucose Tolerance Test
To evaluate the effect of LP aqueous extract on the systemic glucose homeostasis, an OGTT was performed in conscious fasted rats after acute and chronic oral administration of this extract (LP), and we compared its effect with other plant aqueous extracts such as PG, FGK, and GO.
During the OGTT, acute oral dose of LP decreased the peak glucose concentration (30 min, p < 0.01, Figure 2A) and the AUC (p < 0.01, Figure 2B). We noticed that this effect was lower when compared to that of GO and the positive control MET (p < 0.001), but remained close to that of PG (p < 0.01).
The same findings were observed after the chronic OGTT with LP when compared to the control (p < 0.01, Figure 2C,D). Finally, no statistical differences between male and female animals were observed (data not illustrated).

Oral Glucose Tolerance Test with Combinations of Plants
To evaluate the effect of plants aqueous extracts combined with LP aqueous extract on the systemic glucose homeostasis. After acute treatment of the combination of LP + PG and LP + FGK, an OGTT was conducted in awoke fasting rats. We compared the obtained The same findings were observed after the chronic OGTT with LP when compared to the control (p < 0.01, Figure 2C,D). Finally, no statistical differences between male and female animals were observed (data not illustrated).

Oral Glucose Tolerance Test with Combinations of Plants
To evaluate the effect of plants aqueous extracts combined with LP aqueous extract on the systemic glucose homeostasis. After acute treatment of the combination of LP + PG and LP + FGK, an OGTT was conducted in awoke fasting rats. We compared the obtained effect with the positive control (MET). Acute oral administration of LP + PG and LP + FGK reduced the peak of the glucose concentration and the AUC (p < 0.001 with LP + PG and p < 0.01 with LP + FGK, Figure 3A,B). The effect of LP + PG was similar to that of the positive control (MET).

Oral Glucose Tolerance Test with Combinations of Plants
To evaluate the effect of plants aqueous extracts combined with LP aqueous extract on the systemic glucose homeostasis. After acute treatment of the combination of LP + PG and LP + FGK, an OGTT was conducted in awoke fasting rats. We compared the obtained effect with the positive control (MET). Acute oral administration of LP + PG and LP + FGK reduced the peak of the glucose concentration and the AUC (p < 0.001 with LP + PG and p < 0.01 with LP + FGK, Figure 3A,B). The effect of LP + PG was similar to that of the positive control (MET).

Inhibition of Pancreatic α-Amylase Activity
The enzymatic activity of pancreatic α-amylase, incubated with the substrate (starch) in the presence of LP extract, was inhibited in concentration-dependent manner ( Figure 4) and this inhibition was close to that observed with acarbose with IC 50 respectively of 0.44 ± 0.05 mg/mL for LP and 0.36 ± 0.02 mg/mL for acarbose.

Inhibition of Pancreatic α-Amylase Activity
The enzymatic activity of pancreatic α-amylase, incubated with the substrate (starch) in the presence of LP extract, was inhibited in concentration-dependent manner ( Figure 4) and this inhibition was close to that observed with acarbose with IC50 respectively of 0.44 ± 0.05 mg/mL for LP and 0.36 ± 0.02 mg/mL for acarbose.

Inhibition of Intestinal α-Glucosidase Activity
The enzymatic activity of intestinal α-glucosidase, incubated with the substrate (sucrose) in the presence of LP extract, was inhibited in a concentration-dependent manner ( Figure 5). The effect of Acarbose was more pronounced than that of LP extract. The IC50 was respectively of 131 ± 20 µg/mL for LP, whereas that of Acarbose was 52 ± 2

Inhibition of Intestinal α-Glucosidase Activity
The enzymatic activity of intestinal α-glucosidase, incubated with the substrate (sucrose) in the presence of LP extract, was inhibited in a concentration-dependent manner ( Figure 5). The effect of Acarbose was more pronounced than that of LP extract. The IC 50 was respectively of 131 ± 20 µg/mL for LP, whereas that of Acarbose was 52 ± 2 µg/mL. . Inhibiting effect of the LP aqueous extract and acarbose on the pancreatic α-amylase enzyme activity. α-amylase was mixed and incubated with starch in the presence or absence of LP extract. Arcabose (medicine) was used as positive control.

Inhibition of Intestinal α-Glucosidase Activity
The enzymatic activity of intestinal α-glucosidase, incubated with the substrate (sucrose) in the presence of LP extract, was inhibited in a concentration-dependent manner ( Figure 5). The effect of Acarbose was more pronounced than that of LP extract. The IC50 was respectively of 131 ± 20 µg/mL for LP, whereas that of Acarbose was 52 ± 2 µg/mL.

Lavandula Pedunculata Inhibits Absorption of Sodium-Dependent D-Glucose
The impact of LP on the intestinal transit of glucose was studied using a short-circuit current method in Ussing chambers.
After a stabilization time (minimum 40 min), 10 mM D-glucose were added to the mucosal side of the mouse jejunum and the activity of the sodium-dependent glucose transporter-1 (SGLT-1) was followed as the rise in the I sc sodium-dependent ( Figure 6A).

Lavandula Pedunculata Inhibits Absorption of Sodium-Dependent D-Glucose
The impact of LP on the intestinal transit of glucose was studied using a short-circuit current method in Ussing chambers.
After a stabilization time (minimum 40 min), 10 mM D-glucose were added to the mucosal side of the mouse jejunum and the activity of the sodium-dependent glucose transporter-1 (SGLT-1) was followed as the rise in the Isc sodium-dependent ( Figure 6A).
The introduction of LP aqueous extract into the mucosal compartment, 5 min before adding the D-glucose (Figure 5), induced a significant and concentration-dependent inhibition of the Isc (IC50 = 81.28 ± 4.01 µg/mL) as shown in Figure 6B. The maximum inhibition of the Isc reached almost 60% of the inhibition induced by phloridzin (0.5 mM) and was obtained with a concentration of 500 µg/mL.
The inhibitory impact of LP aqueous extract on mucosal intestinal glucose absorption was not shown to be glucose concentration dependent.

Acute Toxicity
Oral administration of the LP aqueous extract to mice at different doses (1, 3, 5, 7, and 10 g/kg b.w.) did not cause any mortality, and no signs of toxicity were noticed (diarrhea, vomiting, abnormal mobility, etc.) after a follow-up of 15 days.

Effect on Body Weights and Relative Weights of Livers and Kidneys
The safety of the LP, FGK, PG, and GO aqueous extracts was evaluated by moni- The introduction of LP aqueous extract into the mucosal compartment, 5 min before adding the D-glucose (Figure 5), induced a significant and concentration-dependent inhibition of the I sc (IC 50 = 81.28 ± 4.01 µg/mL) as shown in Figure 6B. The maximum inhibition of the I sc reached almost 60% of the inhibition induced by phloridzin (0.5 mM) and was obtained with a concentration of 500 µg/mL. The inhibitory impact of LP aqueous extract on mucosal intestinal glucose absorption was not shown to be glucose concentration dependent.
3.6. Toxicological Studies 3.6.1. Acute Toxicity Oral administration of the LP aqueous extract to mice at different doses (1, 3, 5, 7, and 10 g/kg b.w.) did not cause any mortality, and no signs of toxicity were noticed (diarrhea, vomiting, abnormal mobility, etc.) after a follow-up of 15 days.

Subchronic Toxicity Effect on Body Weights and Relative Weights of Livers and Kidneys
The safety of the LP, FGK, PG, and GO aqueous extracts was evaluated by monitoring the variation in the body weight of normal rats during four weeks of the extract consumption at a dose of 1 g/kg b.w. (Figure 7). Also, this safety was assessed by measuring the relative weights of the livers and the kidneys at the end of the experimental period ( Table 2). The obtained results showed that the variation in body weights and relative weights of organs of the treated rats is not significant in comparison with the control.  Biochemical parameters of the studied rats that were administered the LP, FGK, PG, and GO aqueous extracts for four weeks is shown in Table 3. The daily administration of these extracts at a dose of 1 g/kg b.w. did not cause any significant variation in these parameters in comparison with the control group.

Effect on the Blood Biochemical Parameters
Biochemical parameters of the studied rats that were administered the LP, FGK, PG, and GO aqueous extracts for four weeks is shown in Table 3. The daily administration of these extracts at a dose of 1 g/kg b.w. did not cause any significant variation in these parameters in comparison with the control group. Direct bilirubin (mg/mL) 1.00 ± 0.00 1.00 ± 0.00 1.00 ± 0.00 1.00 ± 0.00 1.00 ± 0.00     . Histological analysis of tissues from control or Lavandula pedunculata Rats were given LP or water every day for 4 weeks by gavage. Liver, spleen, and analyzed from control (a,c,e, respectively) or LP-treated rats (b,d,f, respectively) ifications were observed in the rats receiving LP in comparison with the control. Figure 9. Histological analysis of tissues from control or Lavandula pedunculata (LP) treated rats. Rats were given LP or water every day for 4 weeks by gavage. Liver, spleen, and kidney slices were analyzed from control (a,c,e, respectively) or LP-treated rats (b,d,f, respectively). No cellular modifications were observed in the rats receiving LP in comparison with the control.

Discussion
UHPLC analysis of LP aqueous extract studied in this work demonstrated its high content in rosmarinic acid. Moreover, coumarin, protocatechuic acid, herniarin, caffeic acid, apigenin, luteolin, myricetin, and chlorogenic acid were also identified in this plant extract. Lopes and his collaborators identified some phenolic compounds of LP aqueous extract from different populations in Portugal using HPLC analysis. The results showed that phenolic acids are the most abundant, especially salvianolic acid B and rosmarinic acid present in large concentrations, and caffeic acid present in smaller ones. As for flavonoids, luteolin-7-O-glucuronide was the major compound in the studied extracts [26]. Another study was also conducted on different extracts of LP from Portugal that were analyzed using HPLC. The results showed high concentrations of rosmarinic acid and smaller ones of luteolin. Moreover, apigenin was present in the ethanolic and hydroethanolic extracts and absent in the aqueous ones [27].
Different studies demonstrated the inhibitory effects exerted by rosmarinic acid on the enzymatic activity of α-amylase and α-glucosidase. Other studies showed that this com-pound is able to improve insulin sensitivity and glucose uptake in animal models [28,29]. Furthermore, the antidiabetic activity of luteolin was assessed in mice and the results showed its potential in improving diabetes [30].
To our knowledge, this is the first study demonstrating that LP directly inhibits the pancreatic α-amylase and intestinal α-glucosidase enzyme activities in vitro, as well as the electrogenic intestinal glucose absorption in vitro, and that it improves the glucose tolerance in rats after acute and chronic oral administration in vivo. Although this inhibition of the intestinal glucose absorption remains low compared to that of Arbutus unedo root crude aqueous extract [31] and Boscia senegalensis seeds [32], the maximum effect is close to 60%. This work shows that LP could join the list of natural α-amylase, α-glucosidase and SGLT1 inhibitors. Acarbose, Phlorizin (a glucoside of phloretin) and other natural α-amylase, α-glucosidase, and SGLT1 inhibitors have been utilized as pharmacological agents or in the treatment of type 2 diabetes [33,34]. Moreover, LP could join the list of natural α-glucosidase inhibitors.
LP aqueous extract exhibited potent actions on hyperglycemia. It ameliorates the oral glucose tolerance in rats, and it exhibits an inhibiting effect on digestive enzymes and glucose absorption. Although there are no available studies on the antidiabetic activity of LP, there are few limited studies concerning other lavender species like L. stoechas and L. officinalis. Indeed, Bint Mustafa and her collaborators have shown in a previous study that the treatment of alloxan-induced diabetic mice with a hydroalcoholic (7:3) maceration of L. stoechas roots showed significant (p < 0.05) effects on fasting mice's blood glucose levels in a dose-dependent manner. The results were similar to the ones obtained by the reference drug (pioglitazone; 1 mg/kg b.w. intraperitoneal). The prepared plant extract was used at different concentrations (50, 100, and 150 mg/kg b.w.). It was used for intraperitoneal injection into the experimental mice at a dose of 0.1 mL/kg b.w. [15]. The same lavender species (L. stoechas) was also studied by Sebai et al. (2013). These researchers used its essential oil that they had obtained by hydrodistillation of the aerial parts. After 15 consecutive days of treatment with the essential oil (50 mg/kg b.w.), the obtained chronic effect showed that L. stoechas essential oil treatment significantly protected against the increase of blood glucose induced by alloxan treatment of Wistar rats [35]. In another research, the effect of L. officinalis aerial parts macerated in ethanol (96%) was studied in alloxan-induced diabetic Wistar rats. Two doses of the extract (100 and 200 mg/kg b.w.) were used during 21 days of treatment. The results showed that the two doses of the L. officinalis ethanolic extract significantly (p < 0.01) diminished the blood glucose level in comparison with the diabetic non-treated group [36].
In addition, the combination of LP with PG and FGK improved the oral glucose tolerance in rats after acute oral administration in vivo and it exhibited a higher antihyperglycemic activity in comparison with the individual plant extracts. Moreover, the activity obtained after combining LP with PG was judged to be similar to the reference drug MET. This amplifier effect obtained after combining the plant extracts could be either the result of the accumulation of their common active compounds, or a synergic reaction between their different compounds. The obtained results confirm and support the traditional use of LP in ACPP in type 2 diabetes therapy. In fact, LP is one of lavender species widely used by traditional healers in Morocco to prevent or cure diabetes in combination with other antidiabetic plants. This traditional method consists of combining plant extracts in the same homemade beverage to ameliorate or reduce the effect of healer potion. In Africa, it is one of the methods used to personalize treatment of patients, according to their own physical condition (child, elderly, pregnant woman) [21].
It was found in another study that the polyherbal preparation of L. stoechas, Curcuma longa, Aegle marmelos, and Glycyrrhiza glabra hydroalcoholic extracts at a dose of 150 mg/kg b.w. is significantly (p < 0.05) more antihyperglycemic than individual extracts [15]. Likewise, it was cited in an ethnobotanical study realized in Morocco that the population use a mixture of L. stoechas, Artemisia herba alba, and Mentha rotundifolia inflorescences in a decoction to treat diabetes [10].
Pharmaceutics 2021, 13, x. https://doi.org/10.3390/xxxxx www.mdpi.com/journal/pharmaceutics directly inhibited the electrogenic intestinal glucose absorption ex vivo. Moreover, LP improved the antihyperglycemic effect of PG and FGK in vivo in rats after acute oral administration ( Figure 10). These findings back up the traditional use of LP in type 2 diabetes treatment and the effectiveness of the alternative and combinative poly-phytotherapy (ACPP). Still, more studies are needed to better understand all pharmacological parameters of this plant extract to be used as alternative medicine to treat hyperglycemia.

Acknowledgments:
The authors acknowledge the generous support from researchers supporting project-under the group number (RSP-2021/132), King Saud University, Riyadh, Saudi Arabia.

Conflicts of Interest:
The authors declare no conflict of interest.