Rubia tinctorum root extracts: chemical profile and management of type II diabetes mellitus

The chemical and biological profiling of the root extracts of Rubia tinctorum was performed. The activities of different extracts were determined considering the antidiabetic effect against type II diabetes mellitus together with anti-obesity and hepatoprotective effects and lipid profile. The methanolic extract of Rubia tinctorum exhibited significant results in decreasing body weight, improving lipid profile, normalizing hyperglycaemia, insulin resistance, hyperinsulinemia. Additionally, it showed enhancement of liver tissue structure and function. The methanolic extract, being the most significant one, was subjected to LC-HRMS analysis to determine its chemical constituents. Finally, the chemical constituents were evaluated by molecular docking study that was carried out to identify the interaction of a panel of 45 compounds in silico and to correlate the structures to their anti-diabetic activity. Among the tested compounds, 1-hydroxy-2-hydroxymethyl anthra-quinone and naringenin-7-O-glucoside showed the most potent activity as α-amylase inhibitors.


Introduction
Diabetes mellitus and obesity are important health problems worldwide. They contribute to the development of various pathological processes such as hypertension, cardiovascular diseases, hyperlipidaemia, certain types of cancer and even Alzheimer's. 1,2 There is a strong correlation between diabetes and obesity, where adipose tissue has an important role in diabetes, a disease characterized by hyperglycemia, insulin hyposecretion, and insulin resistance. 3,4 The exploitation of phyto-medicine as a therapy for diabetes as well as obesity is a crucial issue for the development of safer alternatives to pharmaceuticals which not only decrease blood glucose but also enhance the antioxidant system. 5 The genus Rubia belongs to the family Rubiaceae and comprises about 70 species. 6 Previous research on Rubia species yielded the isolation and chemical identication of about 250 compounds with different chemical classes which possess diverse pharmacological activities including anti-bacteria, antithrombic, anticancer, anti-inammatory and anti-oxidant. [7][8][9][10] Rubia tinctorum L. (Madder root) was used as a source of red dyes in ancient Egypt. 11 In addition, several ethnobotanical surveys have reported its use for treatment of various ailments, such as cardiovascular disease, 12,13 liver pain, diarrhea, 14,15 rheumatism 16 and kidney stones. 17,18 Moreover numerous biological studies have been extensively conducted on R. tinctorum and proved its therapeutic potential as antiplatelet aggregation, 19 antitumor, 20 hepatoprotective, 21 vasoconstriction and protective effect on aorta. 22 The current work was designed to evaluate the antiobesity, antidiabetic and hepatoprotective efficacy of different extracts of R. tinctorum along with proling of potent bioactive components responsible for the antidiabetic activity. In our strategy, we used LC-HRMS technique to identify the phytochemicals in the root extract of R. tinctorum then to determine the binding mode of the tested compounds with 1HX0 as a-amylase inhibitor by using molecular docking tool.

Materials and methods
2.1. In vitro determination of hypoglycemic activity, liver enzymes and lipid prole 2.1.1. Plant material. The roots of Rubia tinctorum were obtained from the Egyptian market and the identity was conrmed in the Faculty of Science, Suez Canal University. A specimen was deposited at Pharmacognosy Department, Faculty of Pharmacy, Suez Canal University, with a code number 2019-RT. Two kilograms of the roots were dried, powdered and extracted with methanol. The extract was dried under vacuum using rotary evaporator to give 250 g of brownish-red methanolic extract (RM). A weight of 160 g of the methanol extract were withdrawn and suspended in 200 mL of distilled water then fractionated with 3 L of each of hexane, chloroform and ethyl acetate successively. Each of the three extracts was dried under vacuum using rotary evaporator to yield 30, 60 and 40 g of hexane (RH), chloroform (RC) and ethyl acetate (RE) extracts respectively.
2.1.2. Experimental animals. Thirty-ve male Wistar rats were used in the current study. The base line body weight was in the range of 118-152 g. They were kept in clean cages with temperature equals 21 AE 6 C and normal light-dark cycle. They had free permission to water and regular diet or HFD. Study protocol was approved by the Committee of Research Ethics at Faculty of Pharmacy in Suez Canal University (license number 202004R2).
2.1.3. Experimental design. Wistar male rats were distributed into seven groups; with 5 rats in every group. First group of rats was assigned as the normal group and received normal chow diet during the experimentation period. Diabetes mellitus type II was induced in the other six groups by modied high fat diet model. [23][24][25] The other ve groups were fed with HFD (87.7% standard diet, 10% lard fat, 30% glucose) for 7 weeks followed by small dose of streptozotocin (STZ 1% g L À1 acetic acid; 30 mg kg À1 , S. C). Aer further ve days, blood glucose levels are determined to insure the incidence of diabetes mellitus. Aer that, rats in Group II (diabetic control group) were given distilled water (1 mL per kg per day, p. o.) till the end of experiment. Rats in group III were given pioglitazone (10 mg per kg per day, p. o.) while rats in group IV, V, VI and VII were given the following extracts RM (200 mg kg À1 ), RH (200 mg kg À1 ), RC (200 mg kg À1 ) and RE (200 mg kg À1 ), respectively for further four weeks. Aer completion of the treatment regimens, nal body weight was recorded. The change in body weight was calculated from the following equation: Dbody weight ¼ (the nal body weight À the initial body weight)/initial body weight Â 100. Similarly, obesity index was calculated by: obesity index ¼ weight of total adipose tissue/nal body weight Â 100.
2.1.4. Fasting blood glucose determination. Rats were subjected to overnight fasting. Blood specimens were gathered from each rat's tail tip, and fasting glucose was recorded by the use of an automated blood glucometer (Super Glucocard, Japan).
2.1.5. Liver processing. The rats were sacriced under anaesthesia. Each rat's liver was quickly dissected and washed out of blood with cold saline solution. The weight of livers was measured and the following formula was used for determining the liver index: (liver weight/body weight Â 100).
Portion of liver tissue was removed from the biggest hepatic lobe, xed in formaldehyde and nally stained with hematoxylin and eosin (H&E).
2.1.6. Measurement of serum biochemical parameters 2.1.6.1. Liver enzymes. Spectrophotometrically method was done with marketable kits (Biocon Diagnostic, Germany) to evaluate serum activity of liver enzymes; alanine transaminase enzyme (ALT) (EC 2.6.1.2), aspartate transaminase enzyme (AST) (EC 2.6.1) in accordance with the protocol reported by the manufacturer. 26 2.1.6.2. Lipid prole. A spectrophotometric assay kits (Spinreact, Spain) were used to measure serum total cholesterol (TC)(CHOD-POD), triglycerides (TGs)(GPO-POD, Líquido), highdensity lipoprotein (HDL) (HDLc-P) and low-density lipoprotein (LDL)(LDLc-D) according to the manufacturer's protocol. 27,28 2.1.6.3. Insulin & leptin ELISA kits. The level of serum insulin and leptin were determined by a rat insulin and leptin ELISA kits (PELOBIOTECH GmbH-Am Klopferspitz 19-82152 planning-Germany) following the manufacturer's protocol. Insulin resistance was determined using the homeostasis model assessment index for insulin resistance (HOMA-IR) index. 29 2.1.7. Statistical analysis for the data. Results obtained from the current study were expressed as mean AE S. E. M and analysed with the version 16 of SPSS program. A one-way analysis of variance (ANOVA) was used to analyse quantitative variables, followed by the multiple comparison test of Bonferroni. Signicant variations were measured at p # 0.05.

Preparation of the sample and LC-HRMS analysis
The mobile phase working solution (MP-WS) was prepared from DI-water : methanol : acetonitrile (50 : 25 : 25). One mL of MP-WS was added to 50 mg weighted dry methanolic extract, vortex for 2 min. This step was followed by ultra-sonication for 10 min then centrifugation for 10 min at 10 000 rpm. 20 mL stock (50/1000 mL) was diluted with 1000 mL reconstitution solvent. Finally, the injected concentration was 1 mg mL À1 where 10 mLs were injected on positive mode. Also, 10 mL MP-WS were injected as a blank sample. The used mobile phase consisted of (A): 5 mM ammonium formate buffer pH 3 containing 1% methanol and (B): 100% acetonitrile. The ow rate was 0.3 mL min À1 . The used pre-column was in-line lter disks (Phenomenex, 0.5 mm Â 3.0 mm) and the column was X select HSS T3 (Waters, 2.5 mm, 2.1 Â 150 mm). Data processing was via MS-DIAL3.52. Master view was used for feature (peaks) extraction from total ion chromatogram based on the following criteria: features should have signal-to-noise greater than 5 (nontargeted analysis) and features intensities of the sample-toblank should be greater than 5.

Molecular docking
Molecular modelling study was carried out to study the interaction of a panel of 28 anthraquinone and 17 avonoids in silico and to correlate the structures to their anti-diabetic activity. Molecular docking study was conducted on a computational soware basis using the Molecular Operating Environment (MOE 2014.09 Chemical Computing Group, Canada). The threedimensional structures of 1HX0 completed with AC1 as alphaamylase inhibitor was freely accessible from the protein data bank (https://www.rcsb.org/structure/1HX0). 30 Compounds were chemically optimized and energetically minimized, while the receptor was prepared and manipulated using routine protocol according to Nae et al. 31 The active sites were dened using grid boxes of appropriate sizes around the co-crystallized ligands. These compounds were docked into the receptor active site, each ligand-receptor complex was tested for binding energy using MOE and interaction analysis using Chimera as a visualizing soware.

Results and discussion
3.1. Effect of different R. tinctorum extracts and pioglitazone (10 mg kg À1 ) on percent change in body weight and obesity index on type II diabetic rats Treatment with high fat diet followed with STZ (30 mg kg À1 ) in diabetic group resulted in a signicant increase in nal body weight (352 AE 7.5), % change in body weight (152 AE 14.1) and obesity index (5.9 AE 0.5) versus normal group (210 AE 10) (42.5 AE 9) and (0.83 AE 0.07) respectively at p # 0.05 (Table 1). Treatment with pioglitazone (10 mg kg À1 ) for four weeks aer induction of diabetes induced signicantly decrease in nal body weight, % change in body weight and obesity index when compared with diabetic group at p # 0.05. Pioglitazone is one member from thiazolidines, as it is PAPR-gamma agonist in particular is known to favorably inuence the majority of the components of Table 1 Effect of different R. tinctorum extracts and pioglitazone (10 mg kg À1 ) on percent change in body weight and obesity index in the experimental groups of type II diabetic rats a a Results are expressed as mean AE S. E. M. and analysed using one-way ANOVA followed by Bonferroni's test for multiple comparisons. a P # 0.05 versus normal group. b P # 0.05 versus diabetic group. c P # 0.05 versus diabetic + pioglitazone (10 mg kg À1 ) group. d P # 0.05 diabetic + RM (200 mg kg À1 ) group. n ¼ 5. insulin resistance characteristic of type 2 diabetes mellitus including adiposity, dyslipidaemia, hyperglycaemia and changes in liver and ovaries. 32 However, its effect in weight gain was previously discussed. 33,34 On the other hand, its role in decreasing weight gain, enhancing lipid prole and stimulation of lipid mobilization from visceral part to the lower part of body was also reported. 35,36 The current results are in agreement with these articles, as pioglitazone treatment reduced total body weight and decreased liver-fat resulting in elevation of insulin sensitivity in these tissues. Additionally, the effect of pioglitazone is related to the correct choice of its dose as lower and higher doses of pioglitazone may exert no or adverse action like sodium water retention and weight gain. 37 So, the dose of pioglitazone should be monitored and well selected. 23 In the current study the selected dose was (10 mg kg À1 ) which is considered to be medium dose and has signicant effect in all measured parameters.
Similarly, diabetic rats treated with RM, RC or RE (each 200 mg kg À1 ) extracts showed a signicant decrease in nal body weight and % change in body weight, however, only diabetic rats treated with RM (200 mg kg À1 ) and RC (200 mg kg À1 ) showed a signicant improvement in obesity index compared with diabetic group. On the other hand, the group treated with the extract RH (200 mg kg À1 ) couldn't show any signicant enhancement in nal body weight, % change in body weight or obesity index in comparison with diabetic group at p # 0.05. Moreover, the results achieved by the treatment with the extract RM (200 mg kg À1 ) were the best in improving the decrease % change in body weight and obesity index (Table 1).

3.2.
Effect of different R. tinctorum extracts and pioglitazone (10 mg kg À1 ) on blood glucose level, serum insulin, insulin resistance and serum leptin level on type II diabetic rats The current results showed signicant increases in blood glucose level (mM L À1 ), serum insulin level (ng L À1 ), HOMA-IR and serum leptin level (ng L À1 ) in diabetic group in comparison with normal group at p # 0.05 (Table 2). However, treatment with pioglitazone (10 mg kg À1 ) signicantly decreased blood glucose level (mM L À1 ), serum insulin level (ng L À1 ), HOMA-IR and serum leptin level (ng L À1 ) when compared to diabetic group at p # 0.05. Additionally, treatment with RM, RH, RC and RE (each of 200 mg kg À1 ) signicantly induced a decrease in blood glucose level (mM L À1 ), serum insulin level (ng L À1 ), HOMA-IR and serum leptin level (ng L À1 ) when compared to diabetic group at p # 0.05. However, the most signicant results were obtained from the group treated with RM (200 mg kg À1 ) concerning normalization of the level of serum leptin (ng L À1 ) at p # 0.05 (Table 2). Table 4 Effect of different R. tinctorum extracts and pioglitazone (10 mg kg À1 ) on lipid profile, serum triglycerides (TG), total cholesterol (TC), high-density lipoprotein (HDL) and low-density lipoprotein (LDL) in the experimental groups of type II diabetes in rats a  (Table 3). Diabetic group showed elevation in serum liver enzymes AST and ALT in comparison with normal group at p # 0.05. Furthermore, treatment of diabetic rats with only pioglitazone (10 mg kg À1 ) or the extract RM (200 mg kg À1 ) for four weeks could signicantly normalize the liver index in comparison with diabetic group at p # 0.05. Additionally, treatment with pioglitazone (10 mg kg À1 ), RM, RH, RC and RE (each 200 mg kg À1 ) signicantly induced a decrease in both two serum liver enzymes ALT and AST in comparison with diabetic group at p # 0.05. Finally, diabetic group showed evidence of injury; hydropic degeneration (black arrows) and a signicant increase in percent of steatosis (red arrows) (H&E, 40Â) when compared to normal group at p # 0.05 (Fig. 1). Treatment with either pioglitazone (10 mg kg À1 ) or any of the Rubia extracts showed enhancement in liver architecture and a signicant decrease in percent of steatosis with a signicant reduction in hydropic degeneration of hepatocytes (black arrows) and many hepatocytes show uniform morphology (red arrows) (H&E, 40Â) in comparison with diabetic group at p # 0.05 (Fig. 1). However, the results obtained by treatment with the extract RM were the best and closer to the normal group (Table 3). 3.4. Effect of different R. tinctorum extracts and pioglitazone (10 mg kg À1 ) on lipid prole, serum triglycerides (TG), total cholesterol (TC), high-density lipoprotein (HDL) and lowdensity lipoprotein (LDL) on type II diabetic rats Treatment with high fat diet followed with STZ (30 mg kg À1 ) in diabetic group resulted in a signicant increase in serum triglycerides (TG) (mg dL À1 ), total cholesterol (TC) (mg dL À1 ) and low-density lipoprotein (LDL) (mg dL À1 ) and a signicant decrease in high-density lipoprotein (HDL) (mg dL À1 ) in comparison with normal group at p # 0.05 (Table 4). Treatment with pioglitazone (10 mg kg À1 ) resulted in a signicant decrease in serum triglycerides (TG) (mg dL À1 ), total cholesterol (TC) (mg dL À1 ) and low-density lipoprotein (LDL) (mg dL À1 ) and a signicant increase in high-density lipoprotein (HDL) (mg dL À1 ) in comparison with diabetic group at p # 0.05. Furthermore, all extracts induced a signicant decrease in serum triglycerides (TG) (mg dL À1 ), total cholesterol (TC) (mg dL À1 ) and low-density lipoprotein (LDL) (mg dL À1 ) at p # 0.05 (Table 4).
Moreover, treatment with RM (200 mg kg À1 ) resulted in a signicant improvement and normalization in high-density lipoprotein (HDL) serum level (mg dL À1 ) in comparison with diabetic group at p # 0.05 (Table 4).
Accordingly, the results obtained from treatment of the diabetic rats with the methanolic extract of Rubia tinctorum were the best and more close to normal group and pioglitazone treated group either in decreasing body weight, obesity, improving lipid prole, normalization hyperglycaemia, insulin resistance, hyperinsulinemia or in enhancing liver tissue structure and function.

LC-HRMS analysis
Based on the results of the biological activities of different extracts of R. tinctorum which revealed that the best extract was the methanolic one, accordingly, the methanolic extract was subjected to LC-HRMS analysis to detect its chemical constituents (Fig. 2). The results showed 45 hits as indicated in Tables 5 and 6. The anthraquinones were previously reported to be major components of Rubia species specially and family Rubiaceae generally. 38 Table 5 revealed the presence of 28 anthraquinones. All of the detected compounds were previously reported to be isolated from the Rubia tinctorum. [38][39][40][41] On the other hand, a number of anthraquinones were previously reported to be isolated from the same plant as for example nor damnacanthal and lucidin methyl ether 42 were not recorded her in the present study. Additionally, family Rubiaceae revealed the presence of a number of avonoids of which derivatives of quercetin, rhamnetin, isorhamnetin, apigenin and kaempferol are the most common. 42 In this work a number of 17 avonoids were recorded as minor constituents of Rubia tinctorum (Table 6). On the other hand, rutin which was previously reported to be isolated from R. tinctorum 38 was not detected here, and instead, quercetrin (which could be considered as a secondary glycoside of rutin) was recorded.

Docking study
a-Amylase has been considered as an important therapeutic target for the management of type 2 diabetes mellitus, hence, we aimed to elucidate the binding mode of the tested compounds with 1HX0 as a-amylase inhibitor. 43 We performed induced t molecular docking studies with the compounds under investigation. The docking results with docking scores, and the hydrogen bonded residues are given in Table 7. Additionally, 3D representative images of one of the high binding affinities of both anthraquinone and avonoids compared to AC1 as the co-crystallized ligand are shown in Fig. 3.
As shown in Table 7, most of the tested derivative were docked and bound to amino acids of the receptor binding site with high and mild binding affinities (energies), while other derivatives couldn't be docked. In reference to the cocrystallized ligand (AC1) forms two major interactions with the Val 163 and Gly 106 as the key amino acids residues, we found two of the anthraquinones with high binding affinity (À13.92-21.03 kcal mol À1 ), and four avonoids (À16. [16][17][18][19][20][21][22][23].56 kcal mol À1 ) towards a-amylase inhibition by forming the same key interactions. Nine of the anthraquinones, and twelve of avonoids showed mild binding affinities (À9.05-15.45 kcal mol À1 ) and (À16.59-27.45 kcal mol À1 ), respectively, as they form only one hydrogen bond with either Val 163 or Gly 106.
As shown in Fig. 3, three-dimensional representation of two highly docked compounds as two active leads relative to the AC1 with moieties of ligand and receptor involved in the interaction, interaction-type, bond-length for each docking procedure. Among anthraquinone, 1-hydroxy-2-hydroxy-methyl anthraquinone forms two hydrogen bonds through the hydroxyl groups as Table 7 Docking results of the tested compounds with high and mild inside 1HX0 binding site as a-amylase inhibitor compared to AC1 as the cocrystallized ligand. Co-crystallized ligand (AC1) inside the binding site of 1HX0 forms 2 HB with the key amino acids Gly 106 and Val 163 a

Group
Binding affinity Identied compound Binding energy (kcal mol À1 ) Ligand-receptor interactions with Highly-bonded interactive docked compounds in the same way like AC1. *The rest of compounds of both groups weren't able to bind with the receptor pocket.
H-donor with Val 163 with bonds length 2.05 A, and H-acceptor with Gly 106 with bonds length 1.74 A. Among avonoids, naringenin-7-O-glucoside forms two hydrogen bonds through the hydroxyl groups as H-donors with Val 163 and Gly 106 with bonds length 1.59 and 1.53Å.
From docking study, we conclude the good affinity of compounds under investigation through their hydroxyl and carbonyl active groups that bind inside the tested target (a-amylase inhibition compared to AC1), which is correlated to anti-diabetic activity.
In addition to a-amylase inhibition activity, most of the compounds detected in the extract proved to possess antioxidant and anti-inammatory activities. [44][45][46] It is well known that both activities play an important role in treatment of diabetes mellitus. 47,48 This also can justify the activity of the methanolic extract of Rubia tinctorum which accumulate a great number of compounds showing antioxidant, anti-inammatory and a-amylase inhibition activities.

Conclusions
The current study proved that the methanolic extract of Rubia tinctorum showed signicant results in decreasing body weight, improving lipid prole, normalizing hyperglycaemia, insulin resistance, hyperinsulinemia in addition to enhancing liver tissue structure and function. There activities could be attributed to its chemical constituents that exert antioxidant, anti-inammatory and a-amylase inhibition activities. As indicated by the docking study, 1-hydroxy-2-hydroxymethyl anthraquinone and naringenin-7-O-glucoside were the most potent as aamylase inhibitors.

Conflicts of interest
There are no conicts to declare.