Quality Control, Anti-Hyperglycemic, and Anti-Inflammatory Assessment of Colvillea racemosa Leaves Using In Vitro, In Vivo Investigations and Its Correlation with the Phytoconstituents Identified via LC-QTOF-MS and MS/MS

Colvillea racemosa is a cultivated ornamental plant that is a monotypic genus of Fabaceae. It is native to Madagascar, with limited studies. For the first time, the leaf quality control parameters, the anti-hyperglycemic and anti-inflammatory in vitro activity of Colvillea racemosa ethanol extract (CRE) and its fractions of petroleum ether (CRP), methylene chloride (CRMC), ethyl acetate (CREA), n-butanol (CRB), and methanol (CRME) were evaluated. It exhibited significant inhibition against α-amylase, α-glucosidase and membrane stabilization. CRB was the most active fraction, and in vivo studies revealed that oral treatment with CRB of STZ-induced diabetic rats efficiently lowered blood glucose by 67.78%, reduced serum nitric oxide and lipid peroxide levels by 41.23% and 38.45%, respectively, and increased the GSH level by 90.48%. In addition, compared with the diabetic group, there was a 52.2% decrease in serum VCAM, a 55.5% increase in paraoxonase, an improved lipid profile, and improved liver and kidney functions for a treated diabetic group with CRB. Metabolite profiling of CRB was determined by UPLC-ESI-QTOF-MS and tandem MS/MS. Twenty-three chromatographic peaks were identified, which were classified into phenolic compounds and amino acids. The characterized flavonoids were apigenin and luteolin derivatives.


Introduction
One of the main flowering plant families is Fabaceae. It includes 727 genera and 19,327 species, divided into six recently reclassified subfamilies (Caesalpinioideae, Dialioideae, Detarioideae, Cercidoideae, Duparquetioideae, and Papilionoideae) [1][2][3]. The genus Colvillea includes only one species, Colvillea racemosa Bojer ex Hook (Colville's Fame Tree, Whip Tree), native to Madagascar. It is a cultivated ornamental plant, whereas in some tropical regions, it is used as a decorative shade tree [4]. However, to date, biological and chemical investigations of the leaves are rare. Different parts of the Colvillea racemose plant showed significant biological activity, as the HPLC analysis of leaves showed flavonoids and phenolic contents which could be responsible for its antioxidant, cytotoxicity, and anti-microbial activity [5], while the isolated compounds from the stem, lupeol, kaempferol, genkwanin, vitexin, aromadendrin, naringenin, isoliquiritigenin, R-liquiritigenin, isovitexin, vicenin 2 and α,β-dihydroxydihydrochalcones (colveol A, colveol B), had high activ- The results reveal that Colvillea racemosa leaves are a valuable source of mineral elements due to large quantities of ash. The moisture percentage is less than 15%, implying that leaves are well suited for formulation, since a low moisture content inhibits microbial growth [15]. Leaves are an abundant source of proteins, carbohydrates, and fats, which are the main components of the diet [16]. Because the nutritious value of leaves is considerably higher, they could also give a high energy content. Generally, proximate analysis shows that Colvillea racemosa leaves have the possibility to be a dietary supplement.

Minerals
From the results of the analysis presented in Table 2, the levels of calcium (Ca), potassium (K), and magnesium (Mg) content were the highest, at about 2426. 67, 2001.9, and 344.01 mg/100 g, respectively. However, other elements were found in small amounts, while selenium (Se), lead (Pb), and cadmium (Cd) were not detected, but when our results were compared with RDI, we found that Ca and manganese (Mn) were above RDI.  [17,18]. * MDI: maximum daily intake for adult [17,18]. N.D: not detected.
The findings reveal that Colvillea racemosa leaves have a high concentration of components (Mg, Mn, Ca, and K) that may perform an essential part in the management and control of diabetes mellitus. Because metabolic syndromes, besides type II diabetes, are linked disorders defined by chronic, low-grade inflammation, caused in part by Mg deficiency [19], Mg supplementation has been revealed to increase insulin secretion [20], enhance insulin sensitivity and be a glycemic regulator in Mg-deficient persons [21].
Mn is a cofactor of the antioxidant enzyme catalase and its absence can also lead to reduced activity, leading to free radical accumulation [22]. Mn is a vital element which plays a role in glucose regulation and lipid metabolism in humans. Furthermore, it is required for Mn-superoxide dismutase, which is responsible for ROS in mitochondrial oxidative stress [23].
Epidemiological investigations have also revealed that dietary Ca deficiency is combined with a higher incidence of type II diabetes mellitus, while six weeks of Ca-vitamin D co-supplementation may decrease biomarkers of oxidative stress and inflammation [24]. In addition, K intake affected the incidence of diabetes mellitus [25].

In Vitro Anti-Inflammatory Activity Evaluation by Membrane Stabilization % Assay
CRE fractions also exhibited a membrane stabilization effect via preventing hypotonicityinitiated erythrocyte membrane lysis. The erythrocyte membrane is like the lysosomal membrane, so the stabilization of the erythrocyte membrane is related to the stabilization of the lysosomal membrane [28]. The stabilization of the lysosomal membrane is important to prevent the release of chemical mediators and lysosomal components of activated neutrophils, which restrict inflammatory progression [29]. Hypotonicity-induced hemolysis can result from cell shrinkage because of intracellular electrolyte and fluid component osmotic loss. Our extract may inhibit the mechanisms that induce or increase the efflux of certain intracellular components [28]. The effect of CRE and fractions of Colvillea racemosa leaves on RBC membrane stability was investigated. It was found that CRE and fractions except for CRP at different concentrations can stabilize the RBC membrane in hypotonic solution and inhibit hemolysis. CRB had similar activity to indomethacin, a reference standard, with an IC 50 estimated at 23.8 ± 0.14 and 17.02 ± 0.23, respectively. CRE, CRMC, CREA, and CRME, on the other hand, displayed membrane-stabilizing action, with IC 50 values of 39.9 ± 0.23, 72.14 ± 0.34, 25.5 ± 0.23, and 120.7 ± 0.45, respectively. Figure 1c shows the percent inhibition of different fraction concentrations versus membrane stability.
In the current research, only CRB was evaluated for further studies (in vivo and LC-MS/MS analysis) because the CRB fraction showed the highest inhibitory effect against α-glucosidase, α-amylase, and hypotonicity-induced membrane lysis, which could be attributed to high phenolic and flavonoid contents extracted in n-butanol solvent saturated with water [30,31].
2.2.3. In Vivo Anti-Hyperglycemic, Anti-Inflammatory, and Antioxidant Activity Assessment In Vivo Pancreatic Activity Evaluation Using Streptozotocin (STZ)-Induced Hyperglycemia in Rats Determination of Fasting Blood Glucose (FBG) and α-Amylase Activity STZ was used to induce hyperglycemia in rats to investigate anti-hyperglycemic activity in vivo. It acts as a diabetogenic agent by increasing the production of ROS in the pancreatic cell [32]. Pancreatic α-amylase is one of the most common amylase enzymes. Suppressing α-amylase enzyme activity helps to reduce hyperglycemia [33].
As shown in Figure 2, the diabetic group had a significant rise in fasting blood glucose levels of 296.04% compared with the untreated normal group (p < 0.05), with a concurrent increase in α-amylase activity of 28.88% compared with the normal group. In Vivo Lipid Profile Assessment Using STZ-Induced Diabetic Rats Short-term initiation of diabetes using STZ in rats significantly changed the lipid profile when compared with healthy rats. We examined total lipid, triglyceride, cholesterol, and HDL-cholesterol levels to assess the effect of CRB in the lipid profile.
The diabetic group displayed a rise in the levels of total lipid, triglyceride, and cholesterol by 58.54%, 53.7%, and 56.42%, respectively, compared with the control group (p < 0.05), with a related reduction in HDL-cholesterol by 39.92%, as shown in Table 3.
In comparison to diabetic rats, the oral administration of GLB effectively improves the lipid profile, as evidenced by a significant reduction in total lipid, triglyceride, and cholesterol levels of 26.31%, 25.11%, and 35.12%, respectively (p < 0.05), in addition to a significant increase in the serum HDL-cholesterol level of 76.41%. Meanwhile, the treatment of diabetic rats orally with CRB efficiently decreased serum total lipid, triglyceride, The oral administration of glibenclamide (GLB) ameliorates hyperglycemia, manifested as an effective lowering of FBG of approximately 73.19% (p < 0.05), with a significant decrease in serum α-amylase activity of 17.66% concerning STZ-induced diabetic rats. Similarly, the treatment of diabetic rats orally with CRB successfully reduced serum blood glucose by 67.78%. This concurrently accompanied a substantial decrease in serum αamylase activity by 19.39%, greater than that of the GLB-treated group, which reflects the effectiveness of CRB in improving pancreatic activity and hyperglycemia.

In Vivo Lipid Profile Assessment Using STZ-Induced Diabetic Rats
Short-term initiation of diabetes using STZ in rats significantly changed the lipid profile when compared with healthy rats. We examined total lipid, triglyceride, cholesterol, and HDL-cholesterol levels to assess the effect of CRB in the lipid profile.
The diabetic group displayed a rise in the levels of total lipid, triglyceride, and cholesterol by 58.54%, 53.7%, and 56.42%, respectively, compared with the control group (p < 0.05), with a related reduction in HDL-cholesterol by 39.92%, as shown in Table 3. Table 3. Effect of the n-butanol fraction of Colvillea racemosa total extract (CRB) orally administered on lipid profile (mg/dL) in Streptozotocin (STZ)-induced diabetic rats.

Groups
Total Lipid Triglycerides Total Cholesterol HDL-Cholesterol In comparison to diabetic rats, the oral administration of GLB effectively improves the lipid profile, as evidenced by a significant reduction in total lipid, triglyceride, and cholesterol levels of 26.31%, 25.11%, and 35.12%, respectively (p < 0.05), in addition to a significant increase in the serum HDL-cholesterol level of 76.41%. Meanwhile, the treatment of diabetic rats orally with CRB efficiently decreased serum total lipid, triglyceride, and cholesterol levels by 32.86%, 20.89%, and 29.71%, respectively, compared with the STZinduced diabetic group, which significantly increased serum HDL-cholesterol by 72.76%.

In Vivo Liver Function Assessment Using STZ-Induced Diabetic Rats
Diabetes has a significant impact on various endogenous organs, with the liver being one of the most critical [34]. The most prominent causes of liver damage in diabetic individuals are hyperglycemia-initiated oxidative stress and consequent disturbances in carbohydrate, protein, and lipid metabolisms [35].
The diabetic group showed a rise in ALT, AST, ALP, and bilirubin levels by 110.63%, 59.88%, 72.57%, and 100%, respectively, compared with the control group (p < 0.05), with a reduction in total protein level by 34.35% compared with the control group, as shown in Table 4. Table 4. Effect of n-butanol fraction of Colvillea racemosa total extract (CRB) orally administered on liver function in Streptozotocin (STZ)-induced diabetic rats. In comparison to diabetic rats, the administration of GLB orally successfully ameliorated liver function, manifested as an effective reduction in ALT, AST, ALP, and bilirubin by 47.52%, 39.27%, 41.24%, and 45.83%, respectively (p < 0.05), conveyed by a significant increase in serum total protein level of 40.59%. Otherwise, the treatment of diabetic rats orally with CRB effectively decreased serum ALT, AST, ALP, and bilirubin by 43.41%, 31.12%, 39.13, and 41.67%, respectively, compared with the diabetic group, which showed a significantly increased serum total protein level by 43.05%.

In Vivo Kidney Function Assessment Using STZ-Induced Diabetic Rats
The diabetic group's creatinine and urea levels increased by 88.46% and 85.29%, respectively (p < 0.05), as shown in Table 5, compared with a control group. In comparison to diabetic rats, the oral administration of GLB effectively improves kidney function, as evidenced by a significant reduction in creatinine and urea of 48.97% and 44.63%, respectively (p < 0.05), while treatment of STZ-induced diabetic rats orally with CRB decreased serum creatinine and urea levels by 40.82% and 40.22%, respectively. Table 5. Effect of n-butanol fraction of Colvillea racemosa total extract (CRB) orally administered on kidney function (mg/dL) in Streptozotocin (STZ)-induced diabetic rats.

Groups
Creatinine Level Urea Activity

In Vivo Oxidative Stress Markers Assessment Using STZ-Induced Diabetic Rats
The raised levels of oxidative stress in diabetic rats are a result of glucose autoxidation, protein glycation, the peroxidation of lipid, and antioxidant enzyme-lowering activities [36]. The diabetic group showed an increase in the levels of nitric oxide, glutathione peroxidase (GPX), glutathione-s-transferase (GST), catalase, and lipid peroxide by 70.75%, 64.09%, 30.16%, 30.93%, and 67.41%, respectively, and a decline in the level of GSH by 61.82% compared with the control group (p < 0.05) Figure 3.  In contrast, the oral administration of GLB effectively alleviates oxidative stress, as evidenced by a potent decrease in nitric oxide, GPX, GST, catalase, and lipid peroxide by 36.52%, 23.44%, 16.45%, 20.19%, and 41.06%, respectively, and an increase in GSH level by 67.62% (p < 0.05) compared with diabetic rats. Otherwise, the treatment of diabetic rats orally with CRB decreased serum nitric oxide, GPX, GST, catalase, and lipid peroxide In contrast, the oral administration of GLB effectively alleviates oxidative stress, as evidenced by a potent decrease in nitric oxide, GPX, GST, catalase, and lipid peroxide by 36.52%, 23.44%, 16.45%, 20.19%, and 41.06%, respectively, and an increase in GSH level by 67.62% (p < 0.05) compared with diabetic rats. Otherwise, the treatment of diabetic rats orally with CRB decreased serum nitric oxide, GPX, GST, catalase, and lipid peroxide levels by 41.23%, 25%, 16.09%, 22.44%, and 38.45%, respectively, and increased the GSH level by 90.48%.
In Vivo Inflammatory Markers Assessment Using (STZ) Induced Diabetic Rats TNF-α was the first factor at the nexus of inflammation and metabolic illness. Insulin resistance was the cause of increased TNF-α generation in tissue [37]. Compared with the control group, the diabetic showed higher levels of VCAM, ICAM-1, TNF-α, IL1-β, and TGF-β by 130.6%, 308.4%, 240.8%, 52.6%, and 52.6%, respectively, and a decrease in paraoxonase by 50% (p < 0.05), as shown in Figure 4.

Liver
The present histopathological examination, as shown in Figure 5, shows liver from a photomicrograph of a control healthy rat (a), while in photomicrograph (b) is shown liver from a positive diabetic rat with diffused fatty degenerated hepatocytes, which appeared as circumscribed vacuolated cells accompanied with the appearance of a signet ring (H&E × 400) (Lesion Score: ++++). However, the photomicrograph of diabetic rats treated with CRB (c) showed the complete disappearance of diabetic lesions, with apparently healthy hepatic parenchyma with normal hepatocytes, blood sinusoids, and portal area (H&E × 400) (Lesion Score: 0). Nevertheless, in other liver sections of diabetic rats treated with CRB, photomicrograph (d) shows the congestion of the central vein (Lesion Score: +). Furthermore, photomicrograph (e) displays the moderate loss of diabetic lesions and small numbers of vacuolated hepatocytes in diabetic rats treated with GLB (Lesion Score: ++). By comparing liver photomicrographs of CRB-and GLB-treated groups, we found a similar effect. Previous research is consistent with the current findings [38,39].

Pancreas
Similarly, histopathological observation, as shown in Figure 6, shows a healthy control pancreas photomicrograph (a). However, photomicrograph (b) shows a diabetic pancreas with hyperplasia in the pancreatic duct, accompanied by congestion in the blood

Pancreas
Similarly, histopathological observation, as shown in Figure 6, shows a healthy control pancreas photomicrograph (a). However, photomicrograph (b) shows a diabetic pancreas with hyperplasia in the pancreatic duct, accompanied by congestion in the blood vessels (H&E X 200) (Lesion Score: ++++). Otherwise, in diabetic rats treated with CRB, as shown in photomicrograph (c), there was complete fading of the diabetic lesions with apparently healthy pancreatic acini, islets, and ducts (H&E X 400) (Lesion Score: 0). Meanwhile, other pancreas sections from diabetic rats treated with CRB, as shown in photomicrograph (d), showed a loss of the diabetic lesions with slight hyperplasia of the pancreatic islets (arrowhead) (H&E X 400) (Lesion Score: +). However, diabetic rats treated with GLB, as shown in photomicrograph (e), showed moderate fading of the diabetic lesions with dilated pancreatic ducts (H&E X 200) (Lesion Score: ++). By comparing pancreas photomicrographs of CRB-and GLB-treated groups, we found that CRB improves the diabetic features of the pancreas. Prior research has found vacuolation of cells in the islets of Langerhans and focal bleeding in the pancreas of diabetic rats, in addition to hyperplasia in the pancreatic duct in STZ-induced diabetic rats, as well as significantly dilated and congested blood vessels (Lesion Score: ++++), which are compatible with the current findings [38,39].

Kidney
In addition, histopathological examination of the kidney, as shown in Figure 7, shows congestion in the interstitial blood vessel with a thickened wall (arrowhead), together with degenerative changes in the renal tubules (H&E X 400) (as shown in photomicrograph (c) of the renal diabetes-positive group) (Lesion score +++). Additionally, photomicrograph (b) of the kidney cells of diabetic rats in our CRB-treated group shows congestion in the glomerular capillaries (arrows) in addition to the vacuolated renal tubular epithelium (arrowhead) (H&E X 400) (Lesion Score: +). Photomicrograph (d) of renal diabetic rats treated with GLB displays vacuolated glomerular epithelium (arrows) (H&E X 400) (Lesion score+). In parallel results, previous studies are consistent with the current findings [38,39].

Kidney
In addition, histopathological examination of the kidney, as shown in Figure 7, shows congestion in the interstitial blood vessel with a thickened wall (arrowhead), together with degenerative changes in the renal tubules (H&E X 400) (as shown in photomicrograph (c) of the renal diabetes-positive group) (Lesion score +++). Additionally, photomicrograph (b) of the kidney cells of diabetic rats in our CRB-treated group shows congestion in the glomerular capillaries (arrows) in addition to the vacuolated renal tubular epithelium (arrowhead) (H&E X 400) (Lesion Score: +). Photomicrograph (d) of renal diabetic rats treated with GLB displays vacuolated glomerular epithelium (arrows) (H&E X 400) (Lesion score+). In parallel results, previous studies are consistent with the current findings [38,39]. Accordingly, the results obtained from the treatment of the diabetic rats with CRB are close to the normal group and the STD-treated diabetic group. It is obvious that CRB effectively ameliorates hyperglycemia and its complications, via inducing pancreatic activity, improving the lipid profile, controlling liver functions, enhancing kidney functions, decreasing oxidative stress, and controlling inflammation, which is the most relied upon for its phytoconstituents. It is the first time for Colvillea racemosa leaves to be reported as antihyperglycemic and anti-inflammatory. This could be attributed to its phenolic content [5]. Although Colvillea racemosa stems were previously demonstrated to contain naringenin and fisetin flavonoids, nothing was reported about its antihyperglycemic activity [6,40]. However, the Delonix species, which belongs to the same subfamily, showed antihyperglycemic, antioxidant, and anti-inflammatory activity [41,42].

Phytoconstituents Investigation Using UPLC-ESI-QTOF-MS of CRB
Metabolites profiling of the CRB was performed by UPLC-ESI-QTOF-MS and tandem MS/MS. Twenty-three chromatographic peaks were identified, where Figure 8 represents the base peak chromatogram of CRB with the characterized metabolites' peak numbers. Accordingly, the results obtained from the treatment of the diabetic rats with CRB are close to the normal group and the STD-treated diabetic group. It is obvious that CRB effectively ameliorates hyperglycemia and its complications, via inducing pancreatic activity, improving the lipid profile, controlling liver functions, enhancing kidney functions, decreasing oxidative stress, and controlling inflammation, which is the most relied upon for its phytoconstituents. It is the first time for Colvillea racemosa leaves to be reported as antihyperglycemic and anti-inflammatory. This could be attributed to its phenolic content [5]. Although Colvillea racemosa stems were previously demonstrated to contain naringenin and fisetin flavonoids, nothing was reported about its antihyperglycemic activity [6,40]. However, the Delonix species, which belongs to the same subfamily, showed antihyperglycemic, antioxidant, and anti-inflammatory activity [41,42]. Metabolite assignments were achieved by the observation of the retention time (RT) for each candidate alongside the mass to charge (m/z) observed in the positive mode [M + H] + , molecular formula, mass score, error, and major MS/MS fragments. The observed values were compared with databases [43,44] and the literature through the Egyptian Knowledge Bank [45], as shown in Table 6. In this context, 23 metabolites were detected. They were classified into phenolic compounds (13) and amino acids (10).  [43,44] and the literature through the Egyptian Knowledge Bank [45], as shown in Table 6. In this context, 23 metabolites were detected. They were classified into phenolic compounds (13) and amino acids (10).

Phenolic Compounds
The detected phenolic compounds were grouped into flavonoids and phenolic acids. The characterized flavonoids (10) belonged to the flavones subclass and they were of apigenin and luteolin derivatives. As for apigenin derivatives, peak 23 expressed m/z 271.06 with a molecular formula of C 15 H 10 O 5 and was characterized as apigenin. Besides, peaks 15 and 18 revealed the m/z 433.11 and molecular formula C 21 H 20 O 10 . They were described as apigenin hexoside. The fragmentation patterns of both isomers were different as the first one showed the neutral loss of two water moieties followed by the consecutive loss of two CHOH moieties (m/z 367.08 and 337.07) representing the typical fragmentation pattern of C-glycosides [48] and it was described as apigenin C-hexoside as shown in Figure S1 which was described in the stems of Colvillea racemosa as vitexin and isovitexin [6]. Where the seconded isomer exerted the neutral loss of 162 Da complying with the loss of a hexosyl moiety and reflecting O-glycosylation and hence it was described as apigenin O-hexoside which was described before in Fabaceae [43]. It bears noting that both isomers represent 26% of the total characterized phenolics in terms of relative abundance. Moreover, peaks 13 and 17 expressed m/z of 579.17 and a molecular formula of C 27 H 30 O 14 with the sequential loss of a deoxyhexose and a hexose followed by the fragment m/z 153 accounting for the ion ( 1,3+ A) + . They are described as apigenin O-deoxyhexoside hexoside (I-II) as shown in Figure S1.
With regards to luteolin derivatives, peak 22 showed an m/z 287.06 and a molecular formula of C 15 H 10 O 6 with the ion m/z 153, 135, and 121 expressing the ions ( 1,3+ A) + and its dehydrated form followed by ( 1,3+ B) + , respectively. It was characterized as luteolin [48]. In this line, peak 16 showed a similar fragmentation pattern to luteolin with an additional loss of a hexosyl moiety and was described as luteolin O-hexoside [43,48]. In addition, peaks 12 and 14 showed an m/z 595.17 and molecular formula C 27 H 30 O 15 . The first isomer expressed the typical fragmentation of C-glycosides Table 6 and hence it was described as luteolin C-deoxyhexoside C-hexoside as shown in Figure S1 whereas the former isomer expressed the neutral loss of 146 and 162 Da, respectively representing the O-glycosylation with a deoxyhexoside and a hexose and it was described as luteolin O-deoxyhexoside O-hexoside They represent the major phenolic compounds with a relative abundance of 31.31% of the total characterized phenolics. In this sense, peak 11 showed an m/z 611.16 and a molecular formula of C 27 H 30 O 16 , exerted a neutral loss of a water moiety (m/z 593.15) as well as CHOH moieties (m/z 575.14, 515.12, 473.11 and 353.06), indicating the presence of C-glycosides [44,48], and was characterized as luteolin di C-hexoside.
Concerning phenolic acids, three derivatives were observed in the characterized metabolites of the CRB of Colvillea racemosa leaves extract. Briefly, peak 19 expressed an m/z of 285.0947 and a molecular formula of C 13 H 16 O 7 . It exerted the neutral loss of a hexosyl moiety (162 Da) with the appearance of the molecular ion of benzoic acid (m/z 123), followed by its decarboxylation (m/z 79). Consequently, it was characterized as benzoyl hexoside, which was found for the first time in the Fabaceae family in accordance with the Reaxys database [44].
Moreover, peak 20, with m/z 373.15 and molecular formula C 17 H 24 O 9 , expressed decarboxylation followed by the neutral loss of two methyl moieties, accompanied by the loss of a hexosyl moiety, as shown in Figure S2. It was characterized as a syringin. Additionally, peak 21 showed a loss of hexosyl moiety followed by the loss of ethyl and water moieties, as shown in Table 6. It was characterized as ethyl ferulate hexoside, which was reported for the first time in the family Fabaceae according to the Reaxys database [44].

Amino Acids
Regarding amino acids, ten amino acids were observed, as shown in Table 6. Peaks 3, 4, and 5 displayed m/z 118.09 and a molecular formula of C 5 H 11 NO 2 and exerted the neutral loss of water, CO 2 and methyl. They were characterized as hydroxymethyl hydroxypyrrolidine I-III. They were found before in Fabaceae according to the KNApSAcK-Core-System database [43]. In this sense, Figure S2 describes the fragmentation pattern of hydroxymethyl hydroxypyrrolidine III. Peaks 2 and 6 expressed m/z 146.08 and molecular formula C 6 H 11 NO 3 , and they were characterized as hydroxypipecolic acid I-II. Peak 1 was characterized as proline (m/z 116.06, C 5 H 9 NO 2 ) with a fragment of m/z 99.05 accounting for deamination [47]. In addition, peaks 8 and 9 were characterized as leucine/isoleucine I-II with a common fragment of m/z 86 accounting for the neutral loss of CO 2 + 2H. In addition, peak 7 expressed a similar fragmentation pattern of leucine/isoleucine with a former loss of a hexosyl moiety, and hence was described as leucine/isoleucine hexoside. Peak 10 with m/z 166.09 and molecular formula C 9 H 11 NO 2 applied a neutral loss of NH 3 and CO 2 , expressing the typical fragmentation of amino acids [47]. Therefore, it was characterized as phenylalanine, as shown in Figure S2.
Finally, the metabolic profiling of CRB may explain the antidiabetic and anti-inflammatory activity of CRB, in which flavonoids are antioxidant molecules found in nature that have anti-diabetic properties. Likewise, growing scientific evidence suggests that they have anti-inflammatory and anti-oxidant properties [51]. The anti-diabetic activity of apigenin and luteolin derivatives was previously assessed through a reduction in gluconeogenic and lipogenic capacity, even with the inhibition of the PKB/AKT pathway [52]. Although apigenin and its derivatives reduce cellular free radicals, prevent cell destruction in pancreatic β-cells, and promote GLUT4 translocation, lowering the glucose level, luteolin and its derivatives increase insulin secretion by suppressing Maf A through the NF-κB signaling cycle and activate PPAR-γ, which targets adiponectin, leptin and GLUT4 genes [53]. In addition, the antidiabetic effect of proline and phenylalanine was revealed by improved glucose uptake through the AMPK pathway [54]. Leucine administration improves glycemic control in humans and rodents with type II diabetes and decreases oxidative stress. CRB's hypoglycemic, anti-inflammatory, and antioxidant therapeutic properties could be linked to the metabolites discovered in this study. The leaves were air-dried in a dark dry room with humidity control at room temperature, then powdered using a herbal grinding machine.

Preparation of Colvillea racemosa Leaf Extract
The dried powdered leaves (1100 g) were macerated in 70% ethanol (to enhance the extraction of many high-polar or medium-to-low-polar compounds by increasing cell wall permeability), then divided into 22 muslin bags, each bag containing 50 g. Two bags each were put in a jar containing 500 mL of 70% ethanol for 5 days under ultrasonic irradiation 1 h per day, for a period of extraction of 20 days till exhaustion. Extracts were collected, filtered, pooled, then evaporated under vacuum (at 50 • C) to give a dark brown residue (110 g). A total of 80 g of extract was fractionated using different solvents of increasing polarity (petroleum ether, methylene chloride, ethyl acetate, n-butanol saturated with water, and methanol) to ensure extraction of many compounds.

Proximate Composition and Mineral Content
Proximate analysis of the powdered leaves was performed by adopting the procedures of the A.O.A.C. [55], which include: determination of moisture content, total ash content, total lipid, and crude fiber. Total protein was detected by the Kjeldahl method [56]. Meanwhile, the content of total carbohydrate for the leaves was determined according to Kostas [57]. The nutritive value was stated in kilocalories per 100 g of the dry weight of air-dried leaves, measured by Formula (1) [58].
Nutritive value = (4 × %protein) + (9 × %crude fat) + (3.75 × %total carbohydrate). (1) The macro-and micro-minerals were estimated by inductively coupled plasma atomic emission spectroscopy (ICP-AES), Thermo Sci, model: iCAP6000 series [59]. Na and K were determined by atomic absorption spectrometry [60]. The powdered sample was introduced into pressure vessels and digested under a controlled temperature and pressure. The concentrations of Ca, Mg, Cd, Cu, Pb, Mn, Se, and Zn were detected. Argon gas was used for the excitation of the element atoms.

In Vitro Anti-Hyperglycemic Assessment via Inhibition of α-Amylase Activity Assay
The 3,5-dinitrosalicylic acid (DNSA) technique was used to perform the α-amylase inhibition assay [61]. A total of 1ml of the extract at various concentrations (1000−7.81 µg/mL) was added to 1mL of enzyme solution (which was prepared by dissolving enzyme in a 20 mM phosphate buffer (pH = 6.9) at the concentration of 0.5 mg/mL), then mixed and incubated at 25 • C for 10 min. This was followed by the addition of 0.5% starch solution (1 mL) to the mixture, then incubated for 10 min at 25 • C. Finally, 2 mL of dinitro salicylic acid (DNS, color reagent) was added to terminate the reaction, then the mixture was heated in a boiling water bath (5 min). The absorbance was detected colorimetrically at 565 nm after cooling. The inhibition % was calculated by Formula (2).
Ac is the absorbance of the control reaction and As is the absorbance of the test sample. Acarbose was the standard drug.

In Vitro Anti-Hyperglycemic Assessment Using Inhibition of α-Glucosidase
Activity Assay α-Glucosidase inhibitory activity was detected according to the reference technique with slight adjustment [62]. In total, 10 µL of the α-glucosidase enzyme at a concentration of 1 U/mL was added to 50 µL phosphate buffer (100 mM, pH = 6.8), and 20 µL of different concentrations of samples (1000 to 7.81 µg/mL) were preincubated at 37 • C for 15 min. At the same time, 20 µL P-NPG (5 mM) was added as a substrate and incubated at 37 • C for 20 min. Finally, the reaction was terminated by 50 µL Na 2 CO 3 (0.1 M). The absorbance of the released p-nitrophenol was detected at 405 nm using a Multiplate Reader. Acarbose was encompassed as a standard. The percentage inhibition was calculated using Formula (2).
The IC 50 value was defined as the concentration of α-amylase/α-glucosidase inhibitor required to inhibit 50% of its activity under the assay conditions. 3.4.3. In Vitro Anti-Inflammatory Evaluation Using Membrane Stabilization % Assay Erythrocyte suspension preparation: total blood was taken from rats through cardiac puncture with heparinized syringes. Three washes with 154 mM NaCl isotonic buffered solution in 10 mM sodium phosphate buffer (pH 7.4) were performed on blood. Finally, blood was centrifuges for 10 min at 3000× g.
Hypotonic-solution-induced erythrocyte hemolysis was used to assess the samples' membrane-stabilizing activity. The test sample was composed of 0.5 mL of stock erythrocyte suspension combined with 5 mL of hypotonic solution (50 mM NaCl) in 10 mM sodium phosphate-buffered saline (pH 7.4) containing the extract (1000-7.81 g/mL) or indomethacin. The control consisted of 0.5 mL of RBC mixed with hypotonic-buffered saline solution on its own. The mixtures were incubated at room temperature for 10 min before being centrifuged at 3000× g for 10 min. The absorbance of the supernatant was measured at 540 nm in 96-well plates. According to the modified approach presented, the % suppression of hemolysis or membrane stabilization was calculated using Formula (3) [63].
Inhibition of hemolysis (membrane stabilization %) = 100 × (OD1 − OD2/OD1% (3) where OD1 is the optical density of hypotonic-buffered saline solution alone, and OD2 is the optical density of the test sample in a hypotonic solution.
The IC 50 value was defined as the concentration of the sample to inhibit 50% RBCs hemolysis under the assay conditions.

In Vivo Anti-Hyperglycemic Activity Evaluation in Rats Rats and Rat Treatment
For this study, fifty male albino rats weighing from 200 ± 50 g were divided into 5 groups of 10 rats each. Normal control rats, group 1, were merely given a citrate buffer. Normal healthy rats, group 2, were orally administrated CRB (0.5 g/kg bodyweight every day for 30 days; according to LD 50 study).
STZ diabetic rats comprised group 3, injected intraperitoneally with one dose of 60 mg/kg of STZ. Group 4 comprised STZ diabetic rats given CRB orally at a dose 0.5 g/kg body weight for 30 days; each rat received 55 mg/0.5 mL distilled water [64]. Group 5 was the positive control, where diabetic rats were administered 10 mg/kg body weight orally (each rat received 1.5 mg/0.5 mL distilled water) of GLB. The study was carried out in accordance with the Canadian Council on Animal Care's guidelines. The ethics committee at the Faculty of Pharmacy, Suez Canal University, Egypt approved the experimental protocol (Code # 201903PHDA1).

Induction of Hyperglycemia in Rats
STZ was used to produce type II diabetes in rats, which was given intraperitoneally with a single dose of STZ (60mg/kg body weight), dissolved before use in 0.01 M citrate buffer [65]. Rats had unrestricted access to food and water following injection. After two hours each rat was given 2.5 mL of 40% glucose solution in addition to 5% glucose solution orally to consume overnight to counter hypoglycemic shock [66]. For the experiment, hyperglycemic rats had blood glucose levels ≥ 250. GLB, standard oral hypoglycemic, was used as a positive control [67].

Preparation of Sample for Analysis
-Serum First, the weight of each rat was measured, then a blood sample taken in a dry, clean test tube via puncturing the sublingual vein, which was allowed to clot for 10 min, then centrifuged at 3000 rpm to separate the serum, which was kept at −80 • C until analysis. -Tissue First, the weight of the liver tissue was measured, then the tissue was homogenized using a polytron homogenizer with 10 volumes of a suitable medium. Then, the supernatant was kept at −80 • C for assessment. The extraction technique used 1 g of tissue that was homogenized in 9 mL physiological saline (0.9N). The tissue sample was centrifuged for 10 min at 3000 rpm and a collection of supernatant was used for various biochemical tissue analysis [68].
Detection of the Biochemical Parameters 1.

Assessment of Lipid Profile
Serum total lipid and triglyceride were determined colorimetrically [71,72], while total cholesterol and HDL-cholesterol were investigated by enzymatic procedures [73,74].

3.
Assessment of liver function Serum AST and ALT were detected colorimetrically [75]. ALP activity was determined according to the modified method [76]. Total bilirubin and total protein were determined according to spectrophotometric procedures [77,78].

4.
Assessment of kidney function Serum content of urea and creatinine were estimated [79,80].

Assessment of antioxidants and oxidative stress biomarkers
Reduced glutathione concentrations in the samples were determined spectrophotometrically [81]. Nitric oxide was measured using the colorimetric method in tissue liver homogenate [82]. GPX was assayed [83]. GST activity was determined with 1-chloro-2,4dinitrobenzene (CDNB) at 340 nm [84]. Catalase activity was determined by the colorimetric method with hydrogen peroxide as substrate [85]. Serum lipid peroxides were calorimetrically measured using the thiobarbituric acid reaction method [86].

7.
Histopathological examination The sacrificed rats' organs (pancreas, liver, and kidneys) were removed and immersed in a solution of 10% formalin. Then, the specimens were cut, rinsed, and dehydrated in a progressively higher quality of alcohol. For histopathological analysis, dehydrated specimens were cleansed in xylol, embedded in paraffin, sectioned at 4-6 µm thickness, and stained with hematoxylin and eosin [88].

UPLC-ESI-QTOF-MS Analysis
LC/QTOF (6530, Agilent Technologies, Santa Clara, CA, USA) was accompanied with an autosampler-G7129A, a quat. pump (G7104C), and a column comp (G7116A) at Fayoum University, Faculty of Pharmacy. The LC-ESI-QTOF-MS investigation was performed on a Poroshell 120 EC-C18, 1000 bar, RP-18 column from Agilent Technologies, dimensions: 150 mm × 3 mm, dp = 2.7 µm. In a flow rate of 0.2 mL /min, a combination of solvent A and solvent B was used as the mobile phase, A: water + 0.1 percent formic acid, and B: acetonitrile + 0.1 percent formic acid, in which isocratic elution from 0 to 2 min, 10% B, linear elution (2-5 min) from 10 to 30% B, and gradient elution (5-15 min) from 30 to 70% B, were followed with gradient elution (15-18 min) from 70 to 100% B and finally, gradient elution (18-20 min) from 100 to 10% B. The sample was ready at a concentration of 1 mg/mL in HPLC-grade methanol, and the injection volume was 10 µL. The metabolites characterization was according to the strategy proposed by several studies [47,49,89]. Metabolite assignments were achieved by the observation of the retention time (RT) for each candidate alongside the mass to charge (m/z) observed in the positive mode [M+H]+, molecular formula, mass score not less than 80%, error ± 3, and major MS/MS fragments. -

Statistical analysis
The results were statistically analyzed using one-way analysis of variance (ANOVA), in which our findings were displayed as mean ± standard deviation of the mean (SD). Values with p < 0.05 were considered to be significantly different [90].

Conclusions
According to the results of this study, Colvillea racemosa leaves have high nutritional value. In addition, CRB showed promising anti-hyperglycemic effects, as shown in in vitro and in vivo investigations. This could contribute to the CRB content of polyphenolic compounds. This could be appreciated by a group of diabetic patients due to the naturally occurring leaves. Nevertheless, extra in vivo investigations and clinical trials for the plant and identified compounds are strongly suggested to authenticate its therapeutic activity as an anti-hyperglycemic and anti-inflammatory agent. Additionally, various pharmaceutical dosage forms containing Colvillea racemosa leaves, either in the form of crude extracts or herbal teas, could be formulated after validating their beneficial potential.

Conflicts of Interest:
There are no conflict of interest declared by the authors.

CRE
Colvillea racemosa ethanol extract CRP petroleum ether fraction of Colvillea racemosa ethanol extract CRMC methylene chloride fraction of Colvillea racemosa ethanol extract