HPLC profiling for the simultaneous estimation of antidiabetic compounds from Tradescantia pallida

1 Introduction

Diabetes is a multifaceted chronic disease that necessitates glycemic control and ongoing medical attention in order to implement risk-reduction strategies (Association, 2022). In 2021, diabetes impacted 537 million individuals globally, and the worldwide prevalence has surpassed 10 %, as reported by the 10th Edition of IDF Diabetes Atlas. Oral antidiabetic medications like biguanides, sulfonylureas, thiazolidinediones, and digestive enzyme inhibitors are the first-line treatment for managing diabetes, with insulin acting as a last desperate measure (Mohamed, 2021). The development of a safer, more affordable, and less toxic medication is required due to the common side effects of synthetic antidiabetics, which include hepatic dysfunction, fluctuation in body weight, vision problems, peripheral arterial disease, and gastrointestinal distress (Asghar, 2018; Marín-Peñalver, 2016). Due to their efficient, secure, and sustained mode of operation, medicinal plants have attracted the interest of researchers worldwide in the pursuit of finding a remedy for diabetes (Rahman, 2022). These medicinal plants contain biologically active compounds possessing valuable therapeutic value (Mohamed, 2021; Abid et al., 2022).

Tradescantia pallida, also known as Purple Heart, is a member of the Commelinaceae family. This enduring herb has a history of application within the food sector, where it has been conventionally utilized for its preservative, coloring, and additive properties (Tan and Kwan, 2020). The plant has been employed medicinally to treat inflammation (Li and Taiwanese native medicinal plants: phytopharmacology and therapeutic values., 2006), bacterial infections (Tan, 2014) and to treat sore eyes (Ragragio et al., 2013). Several species of the Tradescantia genus have manifested antidiabetic activity. Mexicans and Puerto Ricans have used infusions made from Tradescantia zebrina and Tradescantia spathacea to treat diabetes (Andrade-Cetto and Heinrich, 2005; Gavillán-Suárez, 2015). In our previous studies, the extract of Tradescantia pallida has been investigated for its antidiabetic potential in vitro as well as in vivo (Imtiaz, 2023; Imtiaz, 2023). An analysis of Tradescantia pallida's composition demonstrated a richness in polyphenols, tannins, flavonoids, steroids, and alkaloids (Huq, 2016). However, as of now, there is a lack of published quantitative data on these components and the bioactive compounds responsible for antidiabetic activity.

High-Performance Liquid Chromatography (HPLC) stands out as a powerful tool, recognized for its precision, sensitivity, and capacity to dissect and quantify complex mixtures of compounds within a sample. The primary objectives of method development are to identify and purify drugs. Additionally, it provides crucial information regarding the drug’s stability and bioavailability (Tartaglia, 2022). The simultaneous estimation is a rapid and cost-effective method which increases pharmaceutical sector productivity by performing a single procedure to analyze the mixture of two or more compounds in an extract or a pharmaceutical dosage form (Habib, 2020). The adoption of HPLC profiling in our study stems from the technique's well-established reputation for its analytical versatility.

Natural products have more drug-like properties concerning structural intricacy and functional groups, surpassing those derived from combinatorial chemistry in these aspects (Karageorgis, 2020). The critical need to extract and isolate bioactive compounds from natural sources for the advancement of human health underscores the significance of ongoing research. Despite the documented antidiabetic activity of various Tradescantia species, there remains a notable gap in the understanding of the antidiabetic evaluation of Tradescantia pallida leaves. This gap in knowledge prompted our investigation into the compounds responsible for the antidiabetic activity of Tradescantia pallida, particularly considering its native presence in diverse regions, including Pakistan, India, West Indies, Africa, and the United States of America. In this context, the principal objective of this investigation involves the isolation of targeted bioactive compounds from Tradescantia pallida for comprehensive analysis and characterization and evaluate their antidiabetic potential through in vitro and in silico studies. Furthermore, this study aims to establish and validate an efficient and expeditious method for the simultaneous estimation of these isolated antidiabetic compounds.

2 Materials and methods

2.4

2.4 Isolation of bioactive compounds

Fraction ε underwent normal phase silica column chromatography, involving a silica gel column (200 g) and elution with methanol:dichloromethane mixtures in varying ratios, ranging from 100:1 to 1:100, yielding 16 sub-fractions ranging from ε1 to ε16. While fraction θ provided six sub-fractions (θ1 to θ6). TLC was used to examine the fractions.

Five sub-fractions, ε13a to ε13e, were obtained after further fractionating sub-fraction ε13, employing a silica gel column (50 g), the sub-fraction was subjected to elution using a mixture ranging from 90:10 to 50:50 methanol and dichloromethane. Bioactive compound SACL4 was isolated from fraction ε13c (262.5 mg). Bioactive compound PACL7 was isolated from fraction ε13e (385 mg). Sephadex LH-20 was used to purify the subfraction ε15 using a water:methanol ratio of 10:90 to 0:100, yielding a total of 227.5 mg of the MACL6.

Sub-fraction θ3 was purified from fraction θ using Sephadex LH-20. Water: methanol was used to elute the swelled beads (10:90 to 0:100). A 220 mg of P42 was isolated from the subfractions θ31 to θ33.

The determination of the compounds' structure (Fig. 1) was achieved through NMR spectroscopy (Bruker AVANCE NEO spectrometer) comprising ¹H NMR (600 MHz), ¹³C NMR (150 MHz) and DEPT-135 with a triple resonance probe in addition to several other spectroscopic analysis mainly ATR-FTIR (Cary 630, Agilent, USA), UV–Visible spectroscopy (2550 UV–Vis, Shimadzu, Japan), and LC-MS (1260 Infinity II and 6470 LC/TQ, Agilent, USA) (Imtiaz, 2023).

Close

Fig. 1

Structures of antidiabetic compounds isolated from Tradescantia pallida leaves. where SACL4-syringic acid, PACL7-p-coumaric acid, MACL6-morin and P42-catechin.

Export to PPT

SACL4 (Syringic acid) (262.5 mg) A white, amorphous solid with a pale hue, melting point ∼ 207 °C. λ_max 216 nm (4.33). ATR-FTIR; 3373 cm⁻¹, 2843 cm⁻¹, 1694 cm⁻¹, 1455 cm⁻¹, 1316 cm⁻¹, 1194 cm⁻¹, 1174 cm⁻¹, 1101 cm⁻¹, 1038 cm⁻¹, 907 cm⁻¹, 862 cm⁻¹, 767 cm⁻¹, 687 cm⁻¹ and 669 cm⁻¹. ESI-MS [M]- 182.00 m/z and [M - H]^- 197.03 m/z denoted to the molecular formula C₉H₁₀O₅. ¹H NMR (CD₃OD) (s, 6H and 2H) 7.31 δppm and (s, 5OCH3 and 3OCH3 − 6H) 3.82 δppm.¹³C and DEPT-135 (CD₃OD) (6OCH3) 56.69 δppm, (6C and 2C) 108.15 δppm, (1-C) 121.83 δppm, (5C and 3C) 141.64 δppm, (4C) 148.79 δppm, (–COOH) 169.93 δppm.

PACL7 (p-coumaric acid) (385 mg) pale yellow solid, melting point ∼ 211 °C. λ_max 310 nm (3.49). ATR-FTIR; 3345 cm⁻¹, 2815 cm⁻¹, 1666 cm⁻¹, 1625 cm⁻¹, 1600 cm⁻¹, 1507 cm⁻¹, 1446 cm⁻¹, 1420 cm⁻¹, 1310 cm⁻¹, 1280 cm⁻¹, 1241 cm⁻¹, 1211 cm⁻¹, 1172 cm⁻¹, 1103 cm⁻¹, 976 cm⁻¹, 937 cm⁻¹, 911 cm⁻¹, 827 cm⁻¹, 797 cm⁻¹, 689 cm⁻¹ and 659 cm⁻¹. ESI-MS [M]^- 119.00 m/z and [M - H]^-163.00 m/z corresponded to the molecular formula C₉H₈O₃. ¹H NMR (CD₃OD) (d, 1H − 8H, J = 15.89 Hz) 6.28 δppm, (d, 2H − 3/5H, J = 8.65 Hz) 6.80 δppm, (d, 2/6H − 2H, J = 8.54 Hz) 7.43 δppm, (d, 7H, 1H, J = 15.92 Hz) 7.61 δppm.¹³C and DEPT-135 (CD₃OD) (8-C) 115·52 δppm, (3/5C) 116·77 δppm, (1C) 127·16 δppm, (2/6C) 131·09 δppm, (7C) 146·67 δppm, (4C) 161·14 δppm and, (9C) 171·07 δppm.

MACL6 (Morin) (227.5 mg) light yellow solid, melting point range 302–304 °C. λ_max 207 nm (4.31). ATR-FTIR; 3499 cm⁻¹, 1647 cm⁻¹, 1602 cm⁻¹, 1170 cm⁻¹, 1138 cm⁻¹, 1101 cm⁻¹, 1086 cm⁻¹, 788 cm⁻¹, and 683 cm⁻¹. ESI-MS [M]^- 124.90 m/z and [M - H]^- 301.06 m/z represented the molecular formula C₁₅H₁₀O₇. ¹H NMR (CD₃OD) (d, 6-H, 8-H, J = 1.81, 1.80 Hz) 6.16 and 6.32 δppm, (d, 1H − 3′-H, J = 2.14 Hz) 6.42 δppm, (dd, 1H − 5′-H, J = 2.18, 8.59 Hz) 6.47 δppm, and (d, 1H − 6′-H, J = 8.58 Hz) 7.41 δppm.¹³C and DEPT-135 (CD₃OD) Ring A; (6-C) 99.27 δppm, (8-C) 94.56 δppm, (5-C) 157.53 δppm, (7-C) 162.78 δppm, (9-C) 165.56 δppm. While for ring B; (C-3′) 105.09 δppm, (C-5′) 111.38 δppm, (C-6′) 132.27 δppm, (C-1′) 136.39 δppm, (2′-C) 149.93 δppm and, (4′-C) 158.97 δppm.

P42 (Catechin) (220 mg) cream white solid, melting point ∼ 211 °C. [α]²⁵_D + 19.25. λ_max 280 nm (4.44). ATR-FTIR; 3233 cm⁻¹, 1608 cm⁻¹, 1518 cm⁻¹, 1459 cm⁻¹, 1364 cm⁻¹,1282 cm⁻¹, 1239 cm⁻¹, 1142 cm⁻¹, 1077 cm⁻¹ and, 711 cm⁻¹. ESI-MS [M]^- 245.09 m/z and [M - H]^- 289.06 m/z correlated to the molecular formula C₁₅H₁₄O₆. ¹H NMR (CD₃OD); (dd, 1H − 10-H, J = 8.21 Hz) 2.50 δppm, (dd, 1H − 4-H, J = 5.41 Hz) 2.83 δppm, (m, 1H − 3-H, J = 5.49, 5.23, 5.44 Hz) 3.96 δppm, (d, 1H − 7-H, J = 7.55 Hz) 4.55 δppm, (d, 1H − 8-H, J = 2.29 Hz) 5.84 δppm, (d, 1H − 6-H, J = 2.30 Hz) 5.92 δppm, (dd, 1H − 6′-H, 1.95) 6.71 δppm, (d, 1H − 5′-H, J = 1.95 Hz) 6.72 δppm, (d, 1H − 2′-H, J = 1.97 Hz) 6.83 δppm.¹³C and DEPT-135 (CD₃OD); (4-C) 28.51 δppm, (3-C) 68.77 δppm, (2-C) 82.82 δppm, (6-C) 95.39 δppm, (8-C) 96.16 δppm, (10-C) 100.71 δppm, (2′-C) 115.17 δppm, (5′-C) 119.99 δppm, (1′-C) 132.13 δppm, (3′-C) 146.19 δppm, (4′-C) 146.22 δppm, (9-C) 156.88 δppm, (5-C) 157.56 δppm and (7-C) 157.81 δppm.

3 Results

3.1

3.1 In vitro analysis

3.1.1

3.1.1 α-amylase inhibition

The outcome of compounds SACL4, PACL7, MACL6 and P42 demonstrated that with the increase in the concentrations of these compounds, the enzymatic inhibition increased which eventually reached a steady state. The assessment of the dose–response relationship indicated compound MACL6 possesses the most significant antienzyme potential in comparison to the other isolated compounds (Fig. 2).

Close

Fig. 2

Percentage inhibition of α-amylase (a) standard and isolated compounds (b) SACL4 (c) PACL7 (d) MACL6 and (e) P42 from the leaves of Tradescantia pallida. where standard-acarbose, P42-catechin, SACL4-syringic acid, PACL7-p-coumaric acid and MACL6-morin.

Export to PPT

After normalizing the data and calculating the log values of percentage inhibition, the median inhibitory concentration was determined (Fig. 3a-e).

Close

Fig. 3

α-amylase median inhibitory concentration of (a) standard and isolated compounds (b) SACL4 (c) PACL7 (d) MACL6 and (e) P42 from the leaves of Tradescantia pallida. where standard-acarbose, P42-catechin, SACL4-syringic acid, PACL7-p-coumaric acid and MACL6-morin.

Export to PPT

Close

Fig. 3

α-amylase median inhibitory concentration of (a) standard and isolated compounds (b) SACL4 (c) PACL7 (d) MACL6 and (e) P42 from the leaves of Tradescantia pallida. where standard-acarbose, P42-catechin, SACL4-syringic acid, PACL7-p-coumaric acid and MACL6-morin.

Export to PPT

Close

Fig. 3

α-amylase median inhibitory concentration of (a) standard and isolated compounds (b) SACL4 (c) PACL7 (d) MACL6 and (e) P42 from the leaves of Tradescantia pallida. where standard-acarbose, P42-catechin, SACL4-syringic acid, PACL7-p-coumaric acid and MACL6-morin.

Export to PPT

3.1.2

3.1.2 Glycosylation of hemoglobin

This assay revealed that compound P42 had the highest antidiabetic potential among the isolated compounds, inhibiting 71.66 % glycosylation non-enzymatically, followed by MACL6, SACL4, and PACL7 (Fig. 4). At 750 µg/mL concentration, the results of P42 were comparable with standard (p = 0.0016). Fig. 5a-e displays the IC₅₀ values of standard acarbose and isolated compounds. In silico studies.

Close

Fig. 4

Percentage inhibition of non-enzymatic glycosylation of hemoglobin assay of (a) standard and compounds (b) SACL4 (c) PACL7 (d) MACL6 and (e) P42 from the leaves of Tradescantia pallida. where standard-acarbose, P42-catechin, SACL4-syringic acid, PACL7-p-coumaric acid and MACL6-morin.

Export to PPT

Close

Fig. 5

Median inhibitory concentration of non-enzymatic glycosylation of hemoglobin assay of (a) standard and compounds (b) SACL4 (c) PACL7 (d) MACL6 and (e) P42 from the leaves of Tradescantia pallida. where standard-acarbose, P42-catechin, SACL4-syringic acid, PACL7-p-coumaric acid and MACL6-morin.

Export to PPT

Close

Fig. 5

Median inhibitory concentration of non-enzymatic glycosylation of hemoglobin assay of (a) standard and compounds (b) SACL4 (c) PACL7 (d) MACL6 and (e) P42 from the leaves of Tradescantia pallida. where standard-acarbose, P42-catechin, SACL4-syringic acid, PACL7-p-coumaric acid and MACL6-morin.

Export to PPT

Close

Fig. 5

Median inhibitory concentration of non-enzymatic glycosylation of hemoglobin assay of (a) standard and compounds (b) SACL4 (c) PACL7 (d) MACL6 and (e) P42 from the leaves of Tradescantia pallida. where standard-acarbose, P42-catechin, SACL4-syringic acid, PACL7-p-coumaric acid and MACL6-morin.

Export to PPT

3.1.3

3.1.3 Molecular docking and PRIME MM-GBSA analysis

Among the four ligands evaluated (Fig. 6a-d), catechin demonstrated a notable interaction with 5U3A, forming a total of three bonds in the ligand–protein complex, all of which were hydrogen bonds (Fig. 7a-d). The first hydrogen bond interaction occurred between ASP197 and the phenolic ring, with a bond distance of 1.65 Å. The second and third hydrogen bonds were observed between ASP300 and THR163, with bond distances of 1.70 and 1.67 Å, respectively. The XP docking score and MM-GBSA ΔG bind were determined to be −5.612 kcal/mol and −25.82 kcal/mol, respectively. Details of the docking and MM-GBSA scores for other ligands are provided in Table 2.

Close

Fig. 6

Molecular docking of isolated compounds from Tradescantia pallida with human pancreatic α-amylase protein 5U3A (a) MACL6 (b) P42 (c) PACL7 (d) SACL4. where MACL6-morin, P42-catechin, PACL7-p-coumaric acid and SACL4-syringic acid.

Export to PPT

Close

Fig. 6

Molecular docking of isolated compounds from Tradescantia pallida with human pancreatic α-amylase protein 5U3A (a) MACL6 (b) P42 (c) PACL7 (d) SACL4. where MACL6-morin, P42-catechin, PACL7-p-coumaric acid and SACL4-syringic acid.

Export to PPT

Close

Fig. 7

2D LID of (a) MACL6 (b) P42 (c) PACL7 (d) SACL4 with 5U3A. where MACL6-morin, P42-catechin, PACL7-p-coumaric acid and SACL4-syringic acid.

Export to PPT

Table 2 Molecular docking and Prime MM-GBSA analysis of isolated compounds from Tradescantia pallida.

PDB ID	Ligands	Interacting Residues	Distances Å	Types of interactions	XP Docking scores	Glide Energy Scores kcal/mol	MM-GBSA ΔG Bind
2DN1	MACL6	HIS87, TYR42, HIE58	4.19, 5.04, 5.42, 1.96	1H-Bond, 3 hydrophobic interactions	−9.779	−44.113	−41.08
	P42	TYR42	5.17	1 hydrophobic interaction	−8.906	−44.220	−20.14
	PACL7	ASN97	2.00	1H-Bond	−8.992	−49.118	5.62
	SACL4	No interacting residues	----		−6.572	−31.021	−25.76
5U3A	MACL6	HIS201, GLH233, ASP197	5.14, 1.58, 2.37, 2.2, 1.97	4H-Bond and 1 hydrophobic interaction	−8.805	−48.497	−4.08
	P42	ASP197, ASP300, THR163	1.65, 1.70, 1.67	3H-Bonds	−5.612	−46.133	−25.82
	PACL7	GLH233, LYS200, TYR151	1.79, 4.22, 2.02, 2.17	2H-bonds, 1 salt bridge and 1 hydrophobic interaction	−3.664	−29.515	−2.35
	SACL4	LYS200, TYR151, HIP305	1.78, 2.91, 2.13	2H-Bonds and 1 salt bridge	−3.172	−19.315	−2.29

where P42-catechin, SACL4-syringic acid, PACL7-p-coumaric acid and MACL6-morin

Fig. 8(a-d) depicts the docked compounds with 2DN1. Fig. 9a illustrates the Local Interaction Density (LID) of morin and 2DN1, showcasing the establishment of four noteworthy interactions as compared to others (Fig. 9b-d). Among these, three were identified as hydrophobic interactions, and one was characterized as a hydrogen bond. Specifically, a hydrogen bond was observed between HIE58 and the hydroxyl residue of morin, with a distance of 1.96 Å. The first hydrophobic interaction occurred between TYR42 and the phenolic ring of the ligand at a bond distance of 5.42 Å. The second and third hydrophobic interactions were noted between HIS87 and the additional phenolic ring of morin at bond distances of 4.19 and 5.04 Å, respectively. These interactions indicated the formation of stable connections, supported by an XP docking score of −9.779 kcal/mol and an MM-GBSA ΔG Bind of −41.08 kcal/mol. Notably, among the four ligands assessed, morin exhibited the most substantial docking and MM-GBSA scores.

Close

Fig. 8

Molecular docking of isolated compounds from Tradescantia pallida with human pancreatic hemoglobin protein 2DN1 (a) MACL6 (b) P42 (c) PACL7 (d) SACL4. where MACL6-morin, P42-catechin, PACL7-p-coumaric acid and SACL4-syringic acid.

Export to PPT

Close

Fig. 8

Molecular docking of isolated compounds from Tradescantia pallida with human pancreatic hemoglobin protein 2DN1 (a) MACL6 (b) P42 (c) PACL7 (d) SACL4. where MACL6-morin, P42-catechin, PACL7-p-coumaric acid and SACL4-syringic acid.

Export to PPT

Close

Fig. 9

2D LID of (a) MACL6 (b) P42 (c) PACL7 (d) SACL4 with 2DN1. where MACL6-morin, P42-catechin, PACL7-p-coumaric acid and SACL4-syringic acid.

Export to PPT

As depicted in Table 2, Morin exhibited significant interactions and binding free energy with 2DN1, while catechin demonstrated noteworthy interactions with 5U3A. Subsequently, both Morin-2DN1 and Catechin-5U3A complexes underwent Molecular Dynamics (MD) simulations to further validate their interactions.

3.1.4

3.1.4 MD simulations

To assess the stability and intermolecular interactions within the ligand–protein complex, Molecular Dynamics (MD) simulations were conducted on the top-docked compounds. Each compound and its respective protein underwent MD simulations for a duration of 100 ns (ns). The resulting trajectories were scrutinized using standard metrics, including Root Mean Square Deviation (RMSD) for α-carbons and the examination of interactions between ligands and proteins. Fig. 10a–b depict the RMSD plots for the ligand-5U3A and ligand-2DN1 complexes. The RMSD graphs illustrate the atomic displacement over the 100 ns simulation period for the 5U3A and 2DN1 systems. The RMSD trajectories of the proteins revealed that, within the processing frame, atomic fluctuations peaked at a maximum of 2.4 Å. Equilibration of the simulated systems was confirmed by the MACL6 complex with 2DN1, demonstrating a stable complex with minimal fluctuations. However, in the case of 5U3A, the P42 component stabilized within 24 ns, experienced destabilization at 65 ns, and then re-engaged with the protein around 70 ns.

Close

Fig. 10

Root mean square deviations of (a) P42 simulated with human pancreatic protein 5U3A and (b) MACL6 simulated with human hemoglobin protein 2DN1. where P42-catechin and MACL6-morin.

Export to PPT

Fig. 11(a-b) presents an overview of the intermolecular interactions and key residues involved in hydrophilic and hydrophobic interactions, water bridges, and ionic interactions. In the simulation with 5U3A, P42 exhibited hydrophilic linkage with residues ASP197 and ASP356, accounting for almost 100 % of the interactions. Similarly, in the simulation with 2DN1, MACL6 demonstrated approximately 100 % and 50 % of hydrophilic contacts attributed to SER102, HIS87, and PHE108. These findings highlight the specific residues and nature of interactions contributing to the stability and behavior of the ligand–protein complexes.

Close

Fig. 11

Intermolecular interactionsof (a) P42 simulated with human pancreatic protein 5U3A and (b) MACL6 simulated with human hemoglobin protein 2DN1. where P42-catechin and MACL6-morin.

Export to PPT

3.2

3.2 HPLC method development

3.2.1

3.2.1 Standard and sample preparation

The standard was injected into HPLC in a 10 µL volume.The samples were processed using the same method and injected with a volume of 10 μL. Fig. 12 presents chromatograms of standard and samples. Table 3 entails the retention times for each component.

Close

Fig. 12

HPLC chromatogram of (a) standards and samples (b) isolated compounds (c) extract of Tradescantia pallida. where P42-catechin, SACL4-syringic acid, PACL7-p-coumaric acid and MACL6-morin.

Export to PPT

Table 3 Retention times of standard and samples.

No.	Name	Retention time
		Standard	Sample 1	Sample 2
1	P42	11.577	11.522	11.547
2	SACL4	13.350	13.327	13.339
3	PACL7	14.312	14.300	14.311
4	MACL6	15.800	15.797	15.815

where P42-catechin, SACL4-syringic acid, PACL7-p-coumaric acid and MACL6-morin

3.3

3.3 Method validation

3.3.1

3.3.1 System suitability parameters

This test was carried out to ensure that the chromatographic conditions were effective and fit for use in HPLC analysis. The parameters considered were the separation factor, capacity factor, resolution, tailing factor, and number of theoretical plates. The parameters were calculated automatically by the HPLC system's software. Table 4 displays the results of the analysis.

Table 4 System suitability parameters for HPLC method development.

Compound	Number of Theoretical Plates (N)	Tailing Factor (T)	Resolution (Rs)	Capacity Factor (K)	Separation Factor
P42	40,721	0.973	---	---	---
SACL4	91,243	1.164	8.728	0.163	---
PACL7	112,380	1.206	5.533	0.251	1.537
MACL6	118,558	1.221	8.402	0.384	1.534

where P42-catechin, SACL4-syringic acid, PACL7-p-coumaric acid and MACL6-morin

3.3.2

3.3.2 Linearity

A slope was drawn between the area of each dilution and the concentrations of 60, 80, 100, 120, and 140 ppm to test for linearity. For each compound, a linear curve was obtained, as shown in Fig. 13. Each curve was assigned a correlation coefficient. The correlation coefficient values for syringic acid, p-coumaric acid, morin, and catechin were 0.9994, 0.9996, 0.9993, and 1.0000, respectively. The results were confirmed by applying the statistical analysis (t-test) which presented good correlation and p value 0.0019.

Close

Fig. 13

Linearity curve of the isolated compounds (a) P42 (b) SACL4 (c) PACL7 (d) MACL6. where P42-catechin, SACL4-syringic acid, PACL7-p-coumaric acid and MACL6-morin.

Export to PPT

Close

Fig. 13

Linearity curve of the isolated compounds (a) P42 (b) SACL4 (c) PACL7 (d) MACL6. where P42-catechin, SACL4-syringic acid, PACL7-p-coumaric acid and MACL6-morin.

Export to PPT

3.3.3

3.3.3 LOD/LOQ

The calibration equations' determination coefficients (R²) were greater than 0.999 for all analytes. lists the calculated LOQ and LOD values for the compounds (Table 5). LOD and LOQ were in the range of 3.37 – 4.98 ppm and 10.23 and 15.10 ppm, respectively.

Table 5 LOD and LOQ values of the phenolic compounds.

Compound	Concentration (ppm)	Linear Equation	R2	LOD (ppm)	LOQ (ppm)	Precison (RSD %)
						Intraday	Interday
P42	40–160	6274.3x + 23416	0.9983	4.98	15.10	0.13	0.70
SACL4	40–160	26788x + 79426	0.9987	4.26	12.92	0.07	0.62
PACL7	40–160	19493x + 1422.2	0.9992	3.37	10.23	1.25	0.68
MACL6	40–160	63408x + 120568	0.9985	4.63	14.04	0.30	0.52

where P42-catechin, SACL4-syringic acid, PACL7-p-coumaric acid and MACL6-morin

3.3.4

3.3.4 Precision

Analyses of standard compounds were replicated (n = 6) to determine instrumental precision. Table 5 shows the precision results for individual compounds. The RSD values of the intra- and inter-day precisions of each compound were < 2 %.

3.3.5

3.3.5 Accuracy

The results showed that the recovered amount ranged between 98 and 102 % of all compounds (Table 6). The statistical analysis was carried out using GraphPad Prism 8.0 and applying Residual Analysis. The t-test was applied and the values obtained was; t = 133.1, df = 5.000 and p < 0.0001.

Table 6 Percentage recovery of the phenolic compounds.

Compound	Spiked concentration	Recovery
	%
P42	60	101.58
	100	100.43
	140	98.40
SACL4	60	101.54
	100	100.33
	140	98.85
PACL7	60	100.11
	100	100.71
	140	99.76
MACL6	60	100.59
	100	100.33
	140	99.01

where P42-catechin, SACL4-syringic acid, PACL7-p-coumaric acid and MACL6-morin

3.3.6

3.3.6 Robustness

Table 7 compared the resolution for normal parameters and modified parameters. The RSD of retention time and area of standard were calculated and displayed in Table 8. Each parameter change was thoroughly monitored for resolution between drug peaks. The chromatograms under normal conditions, at wavelength 252 nm, at wavelength 256 nm, at column oven 28 °C, at column oven 32 °C, at flow rate 0.9 mL/min, and at flow rate 1.1 mL/min are presented in Fig. 14a-f.

Table 7 Comparison of the resolutions of the compounds under normal and robust conditions.

Chromatographic Parameters	Resolution
	Normal Condition	Wavelength nm		Flow Rate mL/min		Oven Temperature °C
		252	256	0.9	1.1	28	32
P42	–	–	–	–	–	–	–
SACL4	8.728	8.781	8.682	8.657	9.008	8.71	8.817
PACL7	5.533	5.599	5.456	5.350	6.057	5.633	5.594
MACL6	8.402	8.414	8.301	8.057	8.929	8.373	8.514

where P42-catechin, SACL4-syringic acid, PACL7-p-coumaric acid and MACL6-morin

Table 8 Comparison of the RSD and retention times under robust conditions.

Compound	Chromatographic parameters
	Wavelength				Flow rate				Oven temperature
	252 nm		256 nm		0.9 mL/min		1.1 mL/min		28 °C		32 °C
	Area	RT	Area	RT	Area	RT	Area	RT	Area	RT	Area	RT
	(RSD)
P42	0.49	0.87	0.53	0.74	1.24	0.95	1.14	1.98	1.04	0.92	0.84	0.88
SACL4	0.49	0.41	0.57	0.34	1.21	0.29	1.10	0.85	0.99	0.40	0.78	0.25
PACL7	0.49	0.28	0.71	0. 14	1.21	0.29	1.23	0.51	1.01	0.30	0.75	0.27
MACL6	0.38	0.20	0.90	0.28	1.21	0.37	1.08	0.31	1.15	0.34	0.78	0.40

Close

Fig. 14

Standard chromatogram at various chromtographic conditions (a) 252 nm (b) 256 nm (c) 0.9 mL/min (d) 1.1 mL/min (e) 28 °C (f) 32 °C where P42-catechin, SACL4-syringic acid, PACL7-p-coumaric acid and MACL6-morin.

Export to PPT

Close

Fig. 14

Standard chromatogram at various chromtographic conditions (a) 252 nm (b) 256 nm (c) 0.9 mL/min (d) 1.1 mL/min (e) 28 °C (f) 32 °C where P42-catechin, SACL4-syringic acid, PACL7-p-coumaric acid and MACL6-morin.

Export to PPT

Close

Fig. 14

Standard chromatogram at various chromtographic conditions (a) 252 nm (b) 256 nm (c) 0.9 mL/min (d) 1.1 mL/min (e) 28 °C (f) 32 °C where P42-catechin, SACL4-syringic acid, PACL7-p-coumaric acid and MACL6-morin.

Export to PPT

3.3.7

3.3.7 Analytical solution stability

After 24 h, the precision assay samples were analyzed again with freshly formulated standard solutions. Table 9 displays the results of each compound. The results showed that the solutions were remarkably stable after 24 h.

Table 9 Analytical solution stability analysis.

Compound	Precision
	Samples freshly prepared	Samples after 24 h
	Mean ± RSD
P42	99.34 ± 0.86	99.92 ± 1.48
SACL4	101.29 ± 1.57	99.55 ± 1.61
PACL7	100.59 ± 0.69	100.35 ± 0.69
MACL6	99.96 ± 0.61	99.67 ± 1.58

where P42-catechin, SACL4-syringic acid, PACL7-p-coumaric acid and MACL6-morin

3.3.8

3.3.8 Range

The linearity test determined the assay's upper and lower limits. The 40 to 160 % range chosen for this analysis validated the method and confirmed that the compounds can be precisely and accurately analyzed using this method.

3.3.9

3.3.9 Quantification of antidiabetic compounds in extract

The quantity of the polyphenolic compounds isolated from the extracts of leaves of Tradescantia pallida are expressed in Table 10. The findings indicate that morin possesses the highest concentration, while catechin has the lowest concentration.

Table 10 Quantification of compounds from Tradescantia pallida extract.

Compound	Amount in extract
	mg/mL ± RSD
P42	0.072 ± 1.27
SACL4	0.135 ± 2.00
PACL7	0.152 ± 0.91
MACL6	0.160 ± 1.12

where P42-catechin, SACL4-syringic acid, PACL7-p-coumaric acid and MACL6-morin

4 Discussion

Phenolic compounds derived from natural sources, predominantly plants, exhibit a myriad of health benefits (El Gizawy, 2021; Pandey and Rizvi, 2009). Diets rich in polyphenols have been associated with significantly positive impacts on human health, offering protection against various serious ailments such as diabetes (Imtiaz, 2023), cardiovascular disease, and cancer (Arts and Hollman, 2005; Tungmunnithum, 2018). Polyphenols play an important role in diabetes (Imtiaz, 2022) treatment by activating β-cells and regulating the digestive enzymes (Imtiaz, 2023) responsible for carbohydrate metabolism (Zhang, 2022; Zheng, 2020). This study is novel in isolation of phenolic compounds from Tradescantia pallida leaves. Notably, there is a lack of a HPLC method for the simultaneous identification of these compounds. Thus, the present investigation involves the isolation of these compounds, exploration of their antidiabetic potential, development of an HPLC method, and the subsequent validation of the assay.

Compound SACL4 was identified as syringic acid through spectroscopic methods, including NMR (Maity, 2011; Mogana, 2014. 2014.; Mogana, et al., 2020). Similarly, p-coumaric acid was confirmed as compound PACL7 (An, 2008; Karthikeyan et al., 2015. 2015.); morin as compound MACL6 (Bunyapraphatsara, 2000; Hussain, 2014) and catechin as compound P42 (Annamalai et al., 2021; Zhou, 2020). The identification process aligns with established studies, providing consistency and validation to our findings.

In this current study, phenolic compounds exhibited significant inhibitory effects on the activity of the digestive enzyme α-amylase and effectively prevented the glycosylation of hemoglobin. P42 and MACL6 emerged as particularly potent in their antienzymatic and anti-non-enzymatic (glycosylation of hemoglobin) activities. While literature has previously reported on the potential of catechin and its derivatives in combating diabetes through the inhibition of digestive enzymes (Zhou, 2020; Bhandari, 2008; Justino, 2018), there is a notable absence of recognized research focused on their impact on glycated hemoglobin. Moreover, the antidiabetic efficacy of morin has been demonstrated in a streptozotocin-induced rat model and through the monitoring of insulin signals in HepG2 cells (Jiang, 2020; Razavi et al., 2019). SACL4 displayed a noteworthy 61 % inhibition of α-amylase and a 57 % glycosylation inhibition of the hemoglobin protein in our study. Previous research has supported the potential of syringic acid in diabetes treatment by enhancing insulin production, revitalizing pancreatic β-cells, and impeding advanced glycation end products (AGEs) (Bhattacherjee and Datta,2015;Jan, 2022;Muthukumaran,2013). Our observations align with these findings, underscoring the potential therapeutic impact of syringic acid. Similarly, PACL7 demonstrated a substantial 45 % inhibition of α-amylase and a 52 % inhibition of non-enzymatic processes. Consistent with previous research, the antidiabetic properties of p-coumaric acid were attributed to its ability to modulate glucose levels in the pancreas by stimulating the glucose transporter 2 (Amalan, 2016), and 5′-adenosine monophosphate-activated protein (AMPK) release (Yoon, 2013). The results from our in vitro antidiabetic assessment are in concordance with previously published data.

Molecular docking, a contemporary computer-based methodology (Dissanayake, 2022), offers insights into potential compounds and streamlines the drug discovery process by revealing intricate receptor-ligand interactions, assessing energetic compatibilities, and elucidating binding modes (Guedes et al., 2014). Our investigation underscores that MACL6 and P42 exhibit robust binding affinities with α-amylase and hemoglobin proteins. The α-amylase binding pockets are categorized into −1 to N subpockets, where −1 represents the innermost subpocket, followed by + 1, +2, +3, and so forth. The glycosidic linkage targeted for cleavage resides in the −1 to + 1 subpockets (Proença, 2019). The catalytic pocket in the α-amylase enzyme comprises triad residues ASP197, ASP300, and GLU233, positioned in the −1 subpocket (Zhang, 2022). Our findings indicate that all isolated polyphenols interacted within the same binding pocket, with key residues including HIS201, GLH233, ASP197, ASP300, THR163, LYS200, TYR151, LYS200, and HIP305, suggesting a deep pocket within the −1 subpocket. This interaction is likely facilitated by the formation of hydrogen bonds. Similarly, in molecular docking with 2DN1, MACL6 exhibited interactions with HIS87, TYR42, and HIE58 as the primary amino acid residues, displaying an energy range of −44.113 Kcal/mol. The hemoglobin protein encompasses ten binding pockets, with four larger than the others and positioned in each hemoglobin chain. Our investigation unveiled that the binding pocket for the isolated polyphenols in the 2DN1 protein is located in helix 6 and helix 7. This aligns with prior studies on molecular docking, supporting the potential of phenolic compounds as potent agents against diabetes (Diker and Kutluay, 2021; Gancar, 2020; Nazir, et al., 2018).

The Prime MM-GBSA method is a rigorous and widely accepted approach for validating docked complexes by computing binding free energies (Genheden and Ryde, 2015). In biomolecular studies, the precise computation of binding free energies is a crucial goal, as these energies govern all processes at the molecular level (Rastelli, 2010). Our data indicated that MACL6 and P42 exhibited the most negative energy, underscoring the stability of the optimized complexes. Our study is innovative in conducting Prime MM-GBSA calculations on MACL6 and P42 against the 5U3A and 2DN1 targets.

MD simulation studies provide insight into the dynamic and thermodynamic characteristics of biological systems under specific physiological conditions (Azam, 2019; Azam, 2018). The study aimed to validate and assess the stability of the P42 and MACL6 docked complexes with 5U3A and 2DN1. The root mean square deviation (RMSD) trajectories of both ligand–protein complexes and the Apo proteins initially followed similar patterns. Specifically, the RMSD of the P42-5U3A complex and the 5U3A Apo form exhibited somewhat parallel trajectories, with convergence points occurring around 40–50 ns. However, there were instances where trajectories did not converge, indicating potential loose connections between the ligand and the protein. Nevertheless, points of convergence demonstrated strong and stable interactions. In contrast, the trajectories of the MACL6-2DN1 complex showed nearly complete convergence throughout the maximum simulation duration, indicating a stable ligand–protein complex. Only a small portion displayed non-convergence between 70 and 80 ns, while the majority of the trajectories converging indicated the formation of a stable and robust complex.

The interaction fraction diagram of P42-5U3A indicated that hydrogen bonds were the predominant type of interaction. ASP197 and ASP356 exhibited the highest interaction fraction values, signifying significant hydrogen bond interactions, followed by HIS299, THR163, and ASP300. Although hydrophobic interactions were relatively scarce in this complex, water bridges played a prominent role, involving several amino acid residues not observed in molecular docking studies. ASP356, for instance, formed multiple contacts with the ligand's hydroxyl residue, with one hydrogen bond directly and another interaction mediated by a water bridge. Both interactions persisted significantly throughout the simulation duration. The second most notable interaction involved ASP197 and the hydroxyl residues of catechin, with interaction fractions of 98 % and 82 % during the simulation, suggesting the formation of a stable complex between P42 and 5U3A. The interaction fraction analysis of MACL6-2DN1 revealed several amino acid residues with substantial interactions with ligand atoms, some forming multiple contacts not observable in molecular docking results due to their static nature. PHE98 exhibited the most significant interaction fraction, surpassing 1.0, indicating more than one contact point with more than one ligand atom. The primary interaction force with PHE98 was hydrophobic in nature. Additionally, hydrophobic interactions involving LEU101, VAL132, VAL93, and LEU136 displayed noteworthy interaction fraction values. Hydrogen bond interactions, crucial for stable complex formation, were observed, with SER102 demonstrating the most significant hydrogen bond interaction, boasting an interaction fraction exceeding 0.8, indicating substantial binding during 80 % of the simulation time. Other hydrogen bonds involved HIS87, HIS58, and ASN97, each with noteworthy interaction fraction values. Minor concentrations of water bridges were also observed. Furthermore, SER102, HIS87, and PHE108 contributed most to ligand binding. SER102 displayed binding for almost 97 % of the simulation duration, while HIS87 and PHE108 exhibited binding durations of 67 % and 66 %, respectively. Intriguingly, the ligand's intramolecular hydrogen bond interactions with certain atoms may have limited their availability for binding, potentially strengthening interactions if not present.

All the findings collectively indicate the establishment of a robust and stable ligand–protein complex between MACL6 and 2DN1. Notably, the complex formed by MACL6 and 2DN1 exhibited greater stability compared to the complex formed by P42 and 5U3A. The observed flexibility of the compounds aligns with existing literature on polyphenols, confirming their dynamic nature (Ghosh, 2021).

The optimization of the HPLC method for the simultaneous estimation of antidiabetic compounds from Tradescantia pallida involved the utilization of a gradient system analysis. The mobile phase consisted of 0.1 % formic acid and acetonitrile. Notably, this specific mobile phase composition has not been previously employed for the simultaneous estimation of the isolated polyphenolic compounds. Therefore, this method stands as an innovative approach for the identification and quantification of catechin, syringic acid, morin, and p-coumaric acid.

All the compounds designated for simultaneous analysis demonstrated complete solubility in the diluent acetonitrile. To ensure the appropriateness and effectiveness of the chromatographic setup for subsequent analyses, various suitability parameters of the system were evaluated. These parameters encompassed the separation factor, capacity factor, resolution, tailing factor, and the number of theoretical plates. The tailing factor, specifically, serves as an indicator of the peak's symmetry upon elution from the column (T) and increases proportionally with the elongation of the peak's tail (Tesoro, 2022). The number of theoretical plates (N) serves as an indicator of the efficiency of the column's stationary phase. A low number of theoretical plates suggests poor separation effectiveness of the column (Li, 2020) therefore, to ensure effective separation, the number of theoretical plates should ideally exceed 2000 (Mangla, 2020). In accordance with these criteria, our findings demonstrated that each compound achieved more than 2000 theoretical plates, coupled with a tailing factor of ≥ 1. Resolution, a measure of the difference in peak heights between two distinct components eluting from the column at different retention times, was also considered in our evaluation (Pérez-Cova et al., 2021). Additionally, the capacity factor, illustrating the interaction duration between the active substance and the stationary phase, revealed favorable conditions for efficient separation. The separation factor, crucial for peak resolution, underscored the distinct migration rates of the compounds (Freitag, 2020). Overall, the analysis adhered to stringent system suitability parameters, ensuring robust and reliable chromatographic performance for the simultaneous estimation of the identified polyphenolic compounds.

The correlation coefficients obtained in this study ranged from 0.9909 to 0.9997 for all analytes, suggesting excellent linearity within the measured range. To determine the limits of detection (LOD) and quantification (LOQ), signal-to-noise ratios of 3.3 and 10 were employed, respectively. The calculated LOD values varied from 3.37 to 4.98 µg/L, while the LOQ values ranged from 10.23 to 15.10 µg/L, as determined from the calibration curves. These results affirm the sensitivity and reliability of the developed HPLC method for the simultaneous estimation of the identified polyphenolic compounds in Tradescantia pallida. The analytical method's precision offers insight into random errors, indicating the agreement among measurements obtained from multiple samplings of a homogenous sample under specified conditions. This concept includes both repeatability (intra-day precision) and intermediate precision (inter-day precision) (ICH, Ich, 2017). The intra-day variation was assessed through the analytical procedure within the same laboratory, employing the same analyst, equipment, and day. For inter-day precision, the procedure was replicated over three consecutive days (n = 6). The precision was determined by analyzing the retention times and peak areas of reference standard compounds and calculating the percent relative standard deviation (% RSD). The data demonstrated a favorable concordance among individual test results. The six-sample percentage assay's relative standard deviation (RSD) should not exceed 2.0 (Vikas, 2020) and our results followed this benchmark. Throughout the testing process in this study, every run exhibited a relative standard deviation (RSD) of less than 2 %, indicating excellent precision of the developed HPLC method. Accuracy was assessed by spiking known quantities of standard substances with the samples, and the percentage recoveries consistently exceeded 98 %. The quantification of catechin, syringic acid, p-coumaric acid, and morin in the extract of Tradescantia pallida leaves was determined, marking the first study to reveal the amounts of these compounds in these leaves.

In the context of our investigation on Tradescantia pallia leaves, the significance of utilizing HPLC lies in its ability to accurately identify individual compounds amidst the intricate amalgamation of phenolic substances. This method not only facilitated precise quantification of each identified compound but also contributed to the standardization of Tradescantia pallia leaves extract. By characterizing and quantifying key phenolic components, HPLC ensured the consistency and reproducibility of results, addressing a crucial aspect of quality control for subsequent studies or industrial applications. Moreover, HPLC enabled the comparative analysis of different samples, allowing for the assessment of variations in phenolic composition. This aspect is particularly pertinent for maintaining uniformity in herbal extract preparations and ensuring product quality. Additionally, the correlation of HPLC data with the biological activities of the isolated compounds established connections between the presence or concentration of specific phenolic compounds and their antidiabetic potential. Thus, the inclusion of HPLC profiling in our analytical approach enhanced our understanding of the therapeutic potential of Tradescantia pallia leaves and provided valuable support for future research endeavors in antidiabetic drug development.

5 Conclusion

In conclusion, our investigation unveiled the presence of syringic acid, p-coumaric acid, morin, and catechin in Tradescantia pallia leaves. These compounds exhibited significant antidiabetic potential by inhibiting α-amylase enzyme and non-enzymatic glycosylation of hemoglobin. The results were further validated using computer-aided drug design technology, molecular docking, Prime MM-GBSA calculations, and MD simulations, elucidating the intricate molecular interactions underlying these inhibitory effects. Furthermore, the developed HPLC method demonstrated simplicity, repeatability, and precision, suggesting its potential for further exploration in clinical and industrial applications. Our results suggest that these isolated compounds could serve as valuable markers for Tradescantia pallida leaves and hold promise for future developments in antidiabetic drug design and optimization.

Funding.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

CRediT authorship contribution statement

Fariha Imtiaz: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Data curation, Writing – original draft, Writing – review & editing, Supervision. Muhammad Islam: Conceptualization, Methodology, Formal analysis, Investigation, Data curation, Writing – original draft, Writing – review & editing. Hamid Saeed: Conceptualization, Methodology, Investigation, Writing – original draft, Supervision. Muhammad Ishaq: Validation, Software, Formal analysis, Data curation, Writing – review & editing. Usman Shareef: Methodology, Software, Formal analysis. Muhammad Naeem Qaisar: Methodology. Kalim Ullah: Methodology. Sibghat Mansoor Rana: Software. Anam Yasmeen: Methodology. Aneeqa Saleem: Formal analysis. Romia Javaid Saddiqui: Writing – original draft. All authors have read and agreed to the published version of the manuscript.