Potential Mechanisms Involved in the Protective Effect of Dicaffeoylquinic Acids from Artemisia annua L. Leaves against Diabetes and Its Complications

Diabetes mellitus is a chronic disease affecting the globe and its incidence is increasing pandemically. The use of plant-derived natural products for diabetes management is of great interest. Polar fraction of Artemisia annua L. leaves has shown antidiabetic activity in vivo. In the present study, three major compounds were isolated from this polar fraction; namely, 3,5-dicaffeoylquinic acid (1); 4,5-dicaffeoylquinic acid (2), and 3,4- dicaffeoylquinic acid methyl ester (3), using VLC-RP-18 and HPLC techniques. The potential protective effects of these compounds against diabetes and its complications were investigated by employing various in vitro enzyme inhibition assays. Furthermore, their antioxidant and wound healing effectiveness were evaluated. Results declared that these dicaffeoylquinic acids greatly inhibited DPPIV enzyme while moderately inhibited α-glucosidase enzyme, where compounds 1 and 3 displayed the most prominent effects. In addition, compound 3 showed pronounced inhibition of α-amylase enzyme. Moreover, these compounds markedly inhibited aldose reductase enzyme and exerted powerful antioxidant effects, among which compound 3 exhibited the highest activity implying a notable potentiality in impeding diabetes complications. Interestingly, compounds 2 and 3 moderately accelerated scratch wound healing. Our findings suggest that these dicaffeoylquinic acids can be promising therapeutic agents for managing diabetes and its complications.


Introduction
Diabetes mellitus is a chronic metabolic disorder defined by elevated blood glucose level due to defective insulin secretion [1]. This disease is a cardinal global health condition, and its incidence is increasing progressively in a pandemic way. Despite the presence of high number of marketed pharmaceuticals for diabetes treatment [2], there is still a lack of satisfactory effective and safe therapy with the existing agents. In recent years, there has been an expanded interest in the use of therapeutically active herbal products for diabetes management due to their low cost, higher accessibility and relative safety [3,4].
Interestingly, several species of Artemisia are used traditionally in different countries for the treatment of diabetes and their extracts have shown antidiabetic activities in different in vivo animal models of diabetes [5]. Artemisia annua L., family Asteraceae is an annual herb native to China and is known traditionally for treating malaria, and as a source of artemisinin. This plant started to attract much attention due to its wide spectrum of biological activities including antimalarial, antimicrobial, anti-inflammatory, antioxidant,

Dipeptidyl Peptidase IV (DPPIV) Inhibition
Incretins are gut hormones that play an important role in glucose homeostasis by stimulating glucose-dependent insulin secretion, together with their anti-apoptotic and proliferation-enhancement effects on pancreatic β-cells [13]. Therefore, as these hormones are inactivated by dipeptidyl peptidase IV (DPPIV) enzyme, inhibition of this enzyme has emerged as a promising approach for controlling blood glucose [14,27]. Based on these facts, the in vitro DPPIV inhibition capability of the polar fraction of A. annua extract (ART-CQ) and its isolated compounds was assessed as a potential antidiabetic mechanism. Interestingly, as shown in Figure 3 the ART-CQ and the isolated compounds 1, 2, 3 have displayed potent DPPIV inhibitory activities with IC50 values of 1.53 ± 0.08, 0.225 ± 0.04, 2.27 ± 0.17, 0.571 ± 0.03 µg/mL, respectively. Notably, compound 1 approached the activity of vildagliptin (IC50 0.154 ± 0.02 µg/mL).

Dipeptidyl Peptidase IV (DPPIV) Inhibition
Incretins are gut hormones that play an important role in glucose homeostasis by stimulating glucose-dependent insulin secretion, together with their anti-apoptotic and proliferation-enhancement effects on pancreatic β-cells [13]. Therefore, as these hormones are inactivated by dipeptidyl peptidase IV (DPPIV) enzyme, inhibition of this enzyme has emerged as a promising approach for controlling blood glucose [14,27]. Based on these facts, the in vitro DPPIV inhibition capability of the polar fraction of A. annua extract (ART-CQ) and its isolated compounds was assessed as a potential antidiabetic mechanism. Interestingly, as shown in Figure 3 the ART-CQ and the isolated compounds 1, 2, 3 have displayed potent DPPIV inhibitory activities with IC 50 values of 1.53 ± 0.08, 0.225 ± 0.04, 2.27 ± 0.17, 0.571 ± 0.03 µg/mL, respectively. Notably, compound 1 approached the activity of vildagliptin (IC 50 0.154 ± 0.02 µg/mL).

Intestinal Enzymes Inhibition
The intestinal enzymes, α-amylase and α-glucosidase, play an essential role in the hydrolysis of polysaccharides and disaccharides into glucose. Targeting these enzymes is one of the current strategies for diabetes management [16]. Thus, the effects of ART-CQ and its isolated compounds on the α-amylase and α-glucosidase activities were measured in the present study.

Intestinal Enzymes Inhibition
The intestinal enzymes, α-amylase and α-glucosidase, play an essential role in the hydrolysis of polysaccharides and disaccharides into glucose. Targeting these enzymes is one of the current strategies for diabetes management [16]. Thus, the effects of ART-CQ and its isolated compounds on the α-amylase and α-glucosidase activities were measured in the present study.

Aldose Reductase Inhibition
Increased activity of aldose reductase is associated with the development of different diabetes complications such as cataract, neuropathy, nephropathy and cardiovascular

Aldose Reductase Inhibition
Increased activity of aldose reductase is associated with the development of different diabetes complications such as cataract, neuropathy, nephropathy and cardiovascular complications. Therefore, approaches aiming at repressing this enzyme are of great value in managing diabetes complications [19]. In this study, we have investigated the in vitro aldose reductase inhibition potential of the combined polar fraction of A. annua extract and its isolated compounds. Results depicted in Figure 5 unveiled that the ART-CQ and the isolated compounds have displayed excellent aldose reductase inhibitory activities. Remarkably, compound 3 exhibited the most prominent activity (IC 50 8.90 ± 0.61 µg/mL) which was comparable to the standard quercetin (IC 50   Moreover, our data implied that the three dicaffeoylquinic acid derivatives could also mediate the antidiabetic potential of A. annua polar fraction via inhibiting the α-glucosidase enzyme, thus impeding glucose absorption and preventing postprandial hyperglycemia. Notably, compound 1 displayed the highest α-glucosidase inhibitory activity in accordance with previous literatures illustrating very potent α-glucosidase inhibitory activities of 3,5-dicaffeoylquinic acid from Cichorium glandulosum [28] and from Geigeria alata to which its antihyperglycemic effect was attributed [30]. Our results are in agreement as well with the reported α-glucosidase inhibitory activity of 4,5-dicaffeoylquinic acid and 3,4-dicaffeoylquinic acid methyl ester (compound 2 and 3) from Gynura divaricata [33]. Furthermore, other compounds were isolated for the leaves and stems of A. annua that showed α-glucosidase inhibitory effectiveness [34]. This, together with our finding, would suggest that inhibition of α-glucosidase enzyme might play a significant role in the antidiabetic activity of A. annua. On the other hand, compound 3 exhibited a substantial αamylase inhibition activity while compound 2 showed moderate activity. This is in accordance with the previously described α-amylase inhibitory activity for 4,5dicaffeoylquinic acid [35,36]. Interestingly, molecular docking has illustrated that dicaffeoylquinic acid exhibited the most fitting for the active sites of α-glucosidase and αamylase enzymes among the major compounds isolated from Chasmanthe aethiopica extract which justified its antihyperglycemic and insulin increasing effects [37].
Intriguingly, the combined polar fraction of A. annua together with its three major biologically active dicaffeoylquinic acid derivatives potentially offer protection against diabetes complications by efficiently inhibiting aldose reductase enzyme activity. Our findings were in harmony with the study by Kuroda and his colleagues who have reported a marked aldose reductase inhibitory activity by the same dicaffeoylquinic acid derivatives isolated from the leaves of Tussilago farfara [38]. In addition, the same study Various studies have reported that dicaffeoylquinic acids exerted antihyperglycemic effect and preserved pancreatic cell function and structure in various models of diabetes [28][29][30]. Indeed, the aforementioned results proved that dicaffeoylquinic acids might contribute to the antidiabetic effect of A. annua polar fraction through various mechanisms. Intriguingly, the three dicaffeoylquinic acid derivatives greatly inhibited the DPPIV enzyme activity, an effect that could potentially convey the antidiabetic effect of A. annua. DPPIV enzyme inhibitors have emerged as efficient antidiabetic medications that are already available in the market and their usage has expanded markedly with the advantages of low risk of hyperglycemia and being orally active [31]. Inhibition of DPPIV enzyme could modulate hyperglycemia by increasing endogenous incretin hormones level with the resultant increment in insulin secretion and enhancement of pancreatic β-cell survival [13]. Our findings might thus explain the reported increase in plasma insulin level and pancreatic islet mass by the administration of a butanol fraction of Cichorium glandulosum seeds, which was rich in dicaffeoylquinic acids, to streptozotocin-induced diabetic mice with the consequent improvement in glucose tolerance [28]. This fraction contained 3,5-dicaffeoylquinic acid as the major component in addition to the presence of moderate amounts of 3,4-and 4,5dicaffeoylquinic acids. Consistently, a methanolic extract of Gynura divaricata enriched in 3,5-and 4,5-dicaffeoylquinic acids decreased blood glucose level and pancreatic apoptosis while enhanced β-cell regeneration in diabetic mice [32].
Moreover, our data implied that the three dicaffeoylquinic acid derivatives could also mediate the antidiabetic potential of A. annua polar fraction via inhibiting the α-glucosidase enzyme, thus impeding glucose absorption and preventing postprandial hyperglycemia.
Notably, compound 1 displayed the highest α-glucosidase inhibitory activity in accordance with previous literatures illustrating very potent α-glucosidase inhibitory activities of 3,5-dicaffeoylquinic acid from Cichorium glandulosum [28] and from Geigeria alata to which its antihyperglycemic effect was attributed [30]. Our results are in agreement as well with the reported α-glucosidase inhibitory activity of 4,5-dicaffeoylquinic acid and 3,4-dicaffeoylquinic acid methyl ester (compound 2 and 3) from Gynura divaricata [33]. Furthermore, other compounds were isolated for the leaves and stems of A. annua that showed α-glucosidase inhibitory effectiveness [34]. This, together with our finding, would suggest that inhibition of α-glucosidase enzyme might play a significant role in the antidiabetic activity of A. annua. On the other hand, compound 3 exhibited a substantial α-amylase inhibition activity while compound 2 showed moderate activity. This is in accordance with the previously described α-amylase inhibitory activity for 4,5-dicaffeoylquinic acid [35,36]. Interestingly, molecular docking has illustrated that dicaffeoylquinic acid exhibited the most fitting for the active sites of α-glucosidase and α-amylase enzymes among the major compounds isolated from Chasmanthe aethiopica extract which justified its antihyperglycemic and insulin increasing effects [37].
Intriguingly, the combined polar fraction of A. annua together with its three major biologically active dicaffeoylquinic acid derivatives potentially offer protection against diabetes complications by efficiently inhibiting aldose reductase enzyme activity. Our findings were in harmony with the study by Kuroda and his colleagues who have reported a marked aldose reductase inhibitory activity by the same dicaffeoylquinic acid derivatives isolated from the leaves of Tussilago farfara [38]. In addition, the same study demonstrated that the methanolic extract of different species of Artemisia; A. absinthium, A. dracunculus and A. vulgaris exerted potent inhibition of aldose reductase enzyme. Moreover, 4,5dicaffeoylquinic acid isolated from an antidiabetic methanolic extract of A. dracunculus showed good aldose reductase inhibitory efficacy [39].

In Vitro Antioxidant Activity
One of the important molecular mechanisms implicated in the development of diabetes complications is enhanced oxidative stress [20] and the use of antioxidants is considered as one of the useful measures for protecting against these complications [22]. Based on the abovementioned facts, the antioxidant and free radical scavenging potential of the tested samples was assessed using the Ferric Reducing Antioxidant Power (FRAP) and 2, 2-Diphenyl-1-picrylhydrazil (DPPH) assays. Results of the DPPH assay declared that the ART-CQ and its isolated compounds (1-3) possessed potent free radical scavenging capability where 50% scavenging was attained at concentrations (IC 50 ) of 36.75 ± 0.015, 30.94 ± 4.61, 14.94 ± 0.60 and 7.78 ± 0.24 µg/mL, respectively (Figure 6a). Compound 3 demonstrated the highest activity that was comparable to the reference ascorbic acid (IC 50 5.50 ± 0.30 µg/mL) (Figure 6a). Amazingly, on the other hand, all samples showed powerful antioxidant activity in the FRAP assay that greatly surpassed that of the ascorbic acid positive control where the FRAP values of the ART-CQ and compounds (1,2) were about 8 times that of the ascorbic acid (Figure 6b). In agreement with the DPPH results, compound 3 had the most prominent antioxidant activity with tremendous FRAP value of more than 30 folds of that of ascorbic acid (Figure 6b).
Our findings are in agreement with the previously shown in vitro [10] and in vivo [9] antioxidant activity of the ART-CQ. The polar fraction of A. Annua greatly scavenges the 2,2 -azino-bis (3-ethylbenzothiazoline-6-sulphonic acid) (ABTS) free radicles in vitro [10] and normalized blood glutathione level of alloxan diabetic rats [9]. Furthermore, these data are in line with the previously reported potent antioxidant activities of 3,5-and 4,5-dicaffeoylquinic acids that were delineated as the major antioxidant metabolites isolated from different herbs of family Asteraceae [40] and also from Scabiosa comosa and S. tschilliensis [41]. The combination of the catechol group and three unsaturated ester moieties in the structure of compound 3 (3,4-dicaffeoylquinic acid methyl ester) accounted for the marked increase in its antioxidant activity [42]. In solidarity, in vivo administration of 3,5-dicaffeoylquinic acid in streptozotocin-induced diabetic rats reduced blood glucose level and displayed antioxidant activity where it reduced malondialdehyde level and enhanced glutathione level and the endogenous antioxidant enzymes activities [29,30]. Treatment of diabetic mice with dicaffeoylquinic acids-enriched extract of Gynura divaricata increased glutathione peroxidase and superoxide dismutase activities while mitigating malondialdehyde level [32]. The impressive antioxidant properties of A. annua polar fraction and its dicaffeoylquinic acid derivatives enable them to be potential candidates for protection against diabetes complications by combating the hyperglycemia-provoked oxidative stress. and normalized blood glutathione level of alloxan diabetic rats [9]. Furthermore, these data are in line with the previously reported potent antioxidant activities of 3,5-and 4,5dicaffeoylquinic acids that were delineated as the major antioxidant metabolites isolated from different herbs of family Asteraceae [40] and also from Scabiosa comosa and S. tschilliensis [41]. The combination of the catechol group and three unsaturated ester moieties in the structure of compound 3 (3,4-dicaffeoylquinic acid methyl ester) accounted for the marked increase in its antioxidant activity [42]. In solidarity, in vivo administration of 3,5dicaffeoylquinic acid in streptozotocin-induced diabetic rats reduced blood glucose level and displayed antioxidant activity where it reduced malondialdehyde level and enhanced glutathione level and the endogenous antioxidant enzymes activities [29,30]. Treatment of diabetic mice with dicaffeoylquinic acids-enriched extract of Gynura divaricata increased glutathione peroxidase and superoxide dismutase activities while mitigating malondialdehyde level [32]. The impressive antioxidant properties of A. annua polar fraction and its dicaffeoylquinic acid derivatives enable them to be potential candidates for protection against diabetes complications by combating the hyperglycemia-provoked oxidative stress.

In Vitro Cell Viability Cytotoxicity
Cell viability assay was conducted to evaluate the effect of the ART-CQ and its major biologically active compounds on the viability of human skin fibroblast (HSF) cells and to select the appropriate concentrations to be used in the wound healing assay. Incubating the HSF cells with the ART-CQ and the isolated compounds for 72 h had no significant cytotoxic effect and more than 80% of cells remain viable up to the upper concentration, 100 µg/mL. The IC 50 values were calculated as 1905 µg/mL, 3393 µg/mL, 3577 µg/mL and 8575 µg/mL for the combined polar fraction (ART-CQ), compound 1, 2 and 3, respectively. This indicated the high safety of the polar fraction of the A. annua and its major isolated compounds.

In Vitro Wound Healing
Diabetes mellitus is accompanied by defective wound healing that can lead to one of the serious complications of diabetes, the diabetic foot ulcer [23]. In the current study, we examined the effect of ART-CQ and its major isolated compounds on wound healing in layers of HSF cell lines using the in vitro scratch wound healing model. Results declared that compounds 2 and 3 improved the wound healing at a concentration of 10 µg/mL. Treating the HSF with compound 2 enhanced cell migration and accelerated the scratch wound healing where wound width being reduced to 40% of the original wound width as compared to only 56% in the control at 24 h (Figure 7a,c). Furthermore, wounds were completely healed by in 48 h while wounds in the control plates were healed only in 72 h (Figure 7a,c). Compound 3 showed increased cell migration only at 48 h with decreased wound width to 23% compared to 37% in the control and both wounds were healed completely in 72 h (Figure 7b,d). However, the combined polar fraction of A. annua did not accelerate wound healing at any of the tested concentrations while compound 1 inhibited cell migration and wound healing at all tested concentrations.   Cell migration is one of the important events driving wound closure and healing [24]. Compound 2, and to a lesser extent compound 3, showed enhanced cell migration sug-gesting a potential efficacy in diabetic wound healing. A methanolic extract of Artemisia annua was used for the first time for the composition of a wound dressing that was shown to increase proliferation of L929 fibroblast cells but only after 4 days of incubation and to exert potent antibacterial activity [43]. Furthermore, accelerated wound healing was described in other Artemisia species [44,45]. Rhimi et al. (2019) has reported enhanced wound healing in mice by using an ointment of the ethanolic extract of Dittrichia viscosa, family Asteraceae [46]. For our concern, this extract was enriched with different dicaffeoylquinic acid derivatives including 4,5-and 3,4-dicaffeoylquinic acids. In line with our elucidated wound healing promoting effect of compound 2 (4,5-dicaffeoylquinic acid), Kurisu and his coworkers demonstrated the induction of the mitogenic factor hepatocyte growth factor by this compound in Neonatal Normal Human Dermal Fibroblasts [47]. Additionally, 4,5dicaffeoylquinic acid from Artemisia absinthium potentiated the effect of antibacterial agents against bacteria that could potentially infect the wounds as Straphylococcus aureus [48]. Taken together, 4,5-dicaffeoylquinic acid (compound 2) could be a promising agent that might promote diabetic wound healing.

Plant Material
A. annua leaves were obtained from the Experimental Station of Medicinal Plants, Faculty of Pharmacy, Cairo University in July 2012. This time was selected as this is the season where the highest content of quinic acid derivatives can be obtained at the preflowering stage of the plant, which is the same stage where the artemisinin content is the highest [9]. The plant was authenticated by Prof. Dr. Ibrahim El-Garf, Professor of Botany, Department of Science, Cairo University. A voucher specimen (13 April 2014) was deposited at the herbarium at the Pharmacognosy Department, Faculty of Pharmacy, Cairo University, Egypt.

General
Alpha-amylase (porcine pancreas), 2-chloro-4-nitrophenyl-D-maltotrioside (CNPG3) and tetramethylsilane were from Sigma-Aldrich (St. Louis, MO, USA) while the remaining chemicals were from UFC Biotechnology. Spectrophotometric readings were performed using the microplate reader (Anthos Zenyth 200RT, Cambridge, UK). 1 H-NMR (400 MHz) analysis was carried out using Bruker NMR Spectrometer (Bruker, Yokohama, Japan). Spectra were recorded in CD 3 OD or DMSO-d 6 with tetramethylsilane being employed as an internal standard, and chemical shift values were expressed in ppm.

Extract Preparation
The air-dried leaves of A. annua (1 kg) were extracted with 70% ethanol by sonication till exhaustion, yielding a green residue (240 g). The alcoholic extract (200 g) was extracted successively with hexane, chloroform, ethyl acetate and n-butanol [10]. The combined polar fraction (ART-CQ, 49 g) was used for the current study.

HPLC/RP-18 Analysis of ART-CQ
HPLC apparatus: Agilent Technologies 1100 series, HPLC system (Agilent Technologies, Palo Alto, CA, USA), equipped with a quaternary pump and degasser G1322A series 1200 was used. Agilent Chem Station software was used for data acquisition and processing [9].

Isolation of the Major Compounds from ART-CQ
Two parts of the polar fraction (ART-CQ) each of 1.2 g was dissolved in 10% methanol/water (M/W), applied on VLC-RP-18 column (RP-18 silica, 42 gm, 5.5 cm L × 4 cm D), and elution was started with 10% M/W, with gradual increase of methanol in water till 25% M/W. Fractions of 100 mL each, were collected and monitored by HPLC to give three major collective fractions; I-III. Collective fraction I (eluted with 10-12% M/W, 600 mL), showed major peak (65% purity) at Rt 8.5 min. It was further purified using VLC-RP-18 column giving compound (1) (74 mg). Collective fraction II (eluted with 14-16% M/W, 500 mL), showed major peak (68% purity) at Rt 9.5 min, which was purified by repeated VLC-RP-18 columns giving compound (2) (62 mg). While collective fraction III (eluted with 19-22% M/W, 700 mL), showed major peak (54% purity) at Rt 12.7 min, and was further purified on VLC-RP-18 columns giving compound (3) (38 mg). The dipeptidyl peptidase IV (DPPIV) inhibition effectiveness of the polar fraction of Artemisia annua (ART-CQ) and the isolated compounds was determined fluorometrically [49] by using the DPPIV Inhibitor Screening Kit (Abnova, Taipei, Taiwan) following the manufacturer's protocol. Briefly, different concentrations of either ART-CQ (10, 50, 100, 200 µg/mL) or the pure compounds (5, 20, 50, 100 µg/mL) were preincubated in a plate with the enzyme solution at 37 • C for 10 min. Then the enzyme substrate solution was added, and the absorbance was measured at Ex/Em = 360/460 nm. The plate was reincubated for 30 min at 37 • C and the absorbance was recorded again. Vidagliptin was used as the reference standard. The difference between the two readings was estimated and the IC 50 was determined.

α-Glucosidase Inhibition Assay
The in vitro α-glucosidase inhibitory potential of ART-CQ and its isolated compounds was evaluated using the α-Glucosidase Inhibitor Screening Kit (Biovision, Milpitas, CA, USA). In a 96 well plate, 10 µL of the test samples and the standard acarbose was incubated with 10 µL Diluted Enzyme Solution for 20 min at room temperature. After incubation, 20 µL of the α-Glucosidase Substrate in buffer was added and the absorbance was measured kinetically at 410 nm for 60 min. The polar fraction was tested in concentrations (10, 50, 100, 200 µg/mL) while the concentrations of the isolated compounds were (5, 20, 50, 100 µg/mL). Acarbose was the standard used in the assay and an enzyme control was carried out in which the samples were replaced by the assay buffer.
The change in absorbance is directly proportional to enzyme activity and the presence of enzyme inhibitor reduces the formation of the end product with the resultant decrease in absorbance. The concentration required to inhibit 50% of enzyme activity, IC 50 for test compounds as well as the standard was determined.

α-Amylase Inhibition Assay
The inhibitory effect of the polar fraction (ART-CQ) and isolated compounds from Artemisia Annua on the activity of α-amylase enzyme was assessed kinetically as previously described [27]. This assay depends on the hydrolysis of 2-chloro-4-nitrophenyl maltotrioside (CNPG3) by the α-amylase enzyme with the production of 2-chloro-4-nitrophenol that can be measured spectrophotometrically at 405 nm for 3 min. The change in absorbance per min is proportional to the enzyme activity. Various concentrations of ART-CQ (10, 50, 100, 200 µg/mL) and isolated compounds (5,20,50, 100 µg/mL) were preincubated with the enzyme for 15 min at 37 • C before the addition of CNPG3. A standard acarbose and a sample-free control were used in the assay and the 50% inhibitory concentrations (IC 50 ) were calculated for all tested samples and the standard.

Aldose Reductase Inhibition Assay
Aldose reductase inhibitory activity was determined kinetically by measuring the decrease in the absorption of NADPH at 340 nm in a reaction mixture containing, in addition to NADPH, the enzyme, DL-glyceraldehyde and the samples [50]. This was accomplished using the Aldose Reductase Inhibitor Screening Kit (Biovision, Milpitas, CA, USA), in accordance with the manufacturer's instructions. In brief, 10 µL of either the Enzyme Assay Buffer, ART-CQ (10, 50, 100, 200 µg/mL) or the isolated compounds (5, 20, 50, 100 µg/mL) were placed in a 96 well plate. Then 60 µL of NADPH and 90 µL of enzyme solution were added and the plate was incubated in dark for 20 min at 37 • C. Subsequently, 40 µL of DL-glyceraldehyde was pipetted and the absorbance was determined at 340 nm at 37 • C for 90 min. Quercetin was utilized as the standard and the IC 50 for the investigated samples and the standard was calculated.

Ferric Reducing Antioxidant Power (FRAP) Assay
The DetectX ® Ferric Reducing Antioxidant Power (FRAP™) Detection Kit (Arbor assays, Ann Arbor, MI, USA) was used to estimate the antioxidant and ferric reducing activity of the polar fraction of A. annua and the isolated compounds in accordance with the method of Benzie and Strain (1996) [51]. ART-CQ was utilized in a concentration of (50 µg/mL) while the isolated compounds were used in a concentration of (25 µg/mL). Ferric chloride standard and ascorbic acid control were used in the assay and the FRAP values were estimated and presented as µM FeCl 2 .

In Vitro Cell Culture
Human Skin Fibroblasts (HSF) cell line was obtained from Nawah Scientific Inc. (Mokatam, Cairo, Egypt). Cells were kept in DMEM media with 100 mg/mL of strep-tomycin, 100 units/mL of penicillin and 10% of heat-inactivated fetal bovine serum in humidified, 5% (v/v) CO 2 atmosphere at 37 • C.

In Vitro Cytotoxicity
Effect of the alcoholic extract and the isolated compounds on cell viability was evaluated to assess their safety utilizing SRB assay [53]. An aliquot of 100 µL cell suspension (5 × 10 3 cells) was incubated with 100 µL media containing the test samples at concentrations (0.01, 0.1, 1, 10, 100 µg/mL) for 72 h. Media was removed and 150 µL of 10% TCA was added for fixing the cells that was further incubated at 4 • C for 1h. Cells were then washed with distilled water and 70 µL 0.4% SRB solution was added, and plates were placed in dark for 10 min. Plates were then washed with 1% acetic acid and left to dry overnight. 150 µL of 10 mM TRIS was then added and the absorbance was measured at 540 nm using a BMG LABTECH ® -FLUOstar Omega microplate reader (Ortenberg, Germany). Experiments were done in triplicates and the IC 50 was calculated.

Scratch Wound Healing Assay
The wound healing assay was conducted to evaluate the effect of different concentrations of the polar fraction and isolated compounds from Artemisia Annua on the migration ability of HSF cell line using the scratch wound healing assay as previously described [54]. In a coated 12-well plates, cells were seeded at density of 2 × 10 5 /well in 5% FBS-DMEM and cultured overnight at 37 • C and 5% CO 2 . On the following day, horizontal scratches were made into the confluent monolayer and the plates were washed with PBS and treated with fresh media containing the combined polar extract and the isolated compounds in concentrations of 10, 100, 1000 µg/mL. Control wells were done using only plain medium. The plates were incubated at 37 • C and 5% CO 2 and images were captured using an inverted microscope at 24, 48 and 72 h time intervals. The acquired images were analyzed by MII ImageView software version 3.7. Experiments were performed in triplicates. Wound width was calculated as the average distance between the edges of the scratches. As cell migration enhanced, the wound width diminishes.

Conclusions
Results of this study introduced different mechanisms of action behind the antidiabetic effect of the combined polar fraction of the leaves of A. annua (ART-CQ), as well as three major dicaffeoylquinic acid derivatives contributing to the activity of this biologically active fraction. The dicaffeoylquinic acid derivatives were capable of inhibiting the activities of the DPPIVI, α-amylase and α-glucosidase enzymes in vitro. These actions could, at least partially, account for the well-recognized antidiabetic activity of Artemisia annua by increasing glucose-dependent insulin release, protecting pancreatic β-cells and delaying glucose absorption, thus reducing postprandial glucose excursion. Moreover, these compounds exhibited potential effectiveness in preventing diabetes complications via inhibiting aldose reductase enzyme and alleviating oxidative stress. Interestingly, some of these compounds promoted wound healing. Our findings gave new insights into the mechanisms of the antidiabetic effect of A. annua and its bioactive compounds that could represent promising lead compounds for the management of diabetes. However, additional future studies are needed to investigate other mechanisms of action of the antidiabetic effect of A. annua and these active metabolites and to exploit their efficacy in vivo as well. Human Skin Fibroblast IC 50 Inhibitory concentration 50