A novel anti-hyperglycemic sulfated pyruvylated polysaccharide from marine macroalga Hydropuntia edulis

Abstract Dipeptidyl peptidase is a crucial enzyme that regulates glucose metabolism by degrading incretins, such as glucagon-like-peptide-1, thereby reducing insulin secretion from the pancreatic β-cells. Consequently, dipeptidyl peptidase-IV inhibitors are an important remedial approach to moderate the hyperglycemic pathophysiology. A pyruvylated polysaccharide characterized as [→3)-4,6-O-(1-carboxyethylidene)-β-D-galp-(2SO3 −)-(1→4)-3,6-α-L-AnGalp-(2OMe)-(1→], isolated from the marine macroalga Hydropuntia edulis, showed attenuation potential against dipeptidyl peptidase-IV (IC50 4.44 μM). The structure was elucidated using mass and one/two-dimensional nuclear magnetic resonance spectroscopic analyses of hydrolyzed polysaccharide besides glycosidic linkages obtained from partially methylated alditol acetate derivative. The isolated polysaccharide also revealed potential anti-carbolytic properties against α-amylase/α-glucosidase (IC50 45–47 μM). The results proved the candidacy of pyruvylated polysaccharide isolated from H. edulis as a potential therapeutic lead against hyperglycemia. Graphical Abstract


Introduction
Diabetes mellitus constitutes a cluster of metabolic diseases described by chronic hyperglycemia either because of insulin deficiency (diabetes type I) or insulin resistance (diabetes type II). Among numerous different ways for the treatment of type-2 diabetes, such as glucagon receptor antagonist, insulin-releasing glucokinase, and dipeptidyl peptidase-IV inhibitor (DPP-IV inhibitor), the latter was found to be safer attributable to lesser side effects. DPP-IV is an enzyme targeted for inhibiting the rapid destruction of incretin hormones, which further prevents post-prandial hyperglycemia. One of the important treatment strategies to prevent diabetes is to inhibit DPP-IV thereby reducing the degradation of incretin hormones and increasing insulin production (Pozharitskaya et al. 2020). Earlier reports found that low molecular weight polysaccharides in marine macroalgae and other marine organisms stimulate the discharge of insulin from the beta-cell of pancreas (Zhang et al. 2008) and exhibited attenuation potential against DPP-IV and carbolytic enzymes (Krishnan et al. 2021). Besides this, post-prandial hyperglycemia can also be treated by inhibiting carbohydrate hydrolyzing enzymes, namely a-amylase and a-glucosidase (Gunathilaka et al. 2020). Hence, the attenuation of these enzymes has emerged as a subject of great importance to treat hyperglycemia. Naturally originated polysaccharides were reported to inhibit the carbolytic enzymes and development of advanced glycation end products that result in hyperglycemic conditions .

Results and discussion
2.1. Bioactive potential of crude polysaccharides from red marine macroalgae The yield of crude polysaccharides obtained via alkaline extraction of red macroalgae P. hornemanii, G. corticata, H. edulis, and A. spicifera were between 1.2% and 5.1% (Table S1). However, the yield of Gracilaria species was found to be lesser than that of prior reports (Barros et al. 2013).
One of the available means to control postprandial hyperglycemia is to interrupt the carbohydrate absorption by cells after food consumption. Carbohydrate hydrolyzing enzyme a-amylase breaks long-chain carbohydrates, while a-glucosidase is responsible for converting starch and disaccharides to glucose by cleavage of 1,4-glycosidic bonds. H. edulis (IC 50 0.26-0.31 mg/mL) and G. corticata (IC 50 0.32-0.47 mg/mL) displayed significantly greater anti-hyperglycemic activities when compared to acarbose (IC 50 0.81 mg/mL) as well as other studied species. Similarly, H. edulis registered a significant DPP-IV inhibition activity (IC 50 0.21 mg/mL) followed by those of G. corticata (IC 50 0.31 mg/mL), A. spicifera (IC 50 0.62 mg/mL), and P. hornemanii (IC 50 0.71 mg/mL), in descending order. The most interesting was that isolated crude polysaccharides have demonstrated greater anti-hyperglycemic activity than displayed by synthetic inhibitor acarbose, and that further paves the way for developing new therapeutic drugs based on polysaccharides of red macroalgal origin.
Natural polysaccharides always co-exist with proteins, nucleic acids, lipids and amino acid residues, which are linked by either hydrophobic interactions or hydrogen bonding, cavities or crevasses. They never occur as singly in the system. Hence, antioxidant activity of polysaccharides could be partly because of the presence of these co-existing partners besides molecular weight, presence of uronic acids, sulfate content, substitution groups/degree of substitutions, and other chemical components Xie et al. 2016). A linear correlation between molecular weight and uronic acids against anti-oxidant capacity is well recognized . Hence, the greater anti-oxidant activity of crude polysaccharide of H. edulis might suggest the presence of a higher number of hydroxyl groups and different substitution patterns compared to other studied algal polysaccharides.
TBARS assay measures lipid peroxidised products like malondialdehyde (MDA) as a quantitative method to detect free radical generation. The crude polysaccharide extract of H. edulis (0.72 mM MDAEQ kg À1 ) has also displayed significant inhibition of lipid peroxidation than those exhibited by other studied macroalgae (! 1 mM MDAEQ kg À1 ) ( Table S1). The addition of crude polysaccharides has been evidenced to reduce lipid peroxidation as previously reported (Maneesh and Chakraborty 2018).
Anti-inflammatory potential of selected red macroalgae were analyzed by assessing their attenuation potential against pro-inflammatory 5-lipoygenase (5-LOX) enzyme wherein H. edulis (IC 50 0.52 mg/mL) presented significantly higher activity followed by G. corticata (IC 50 0.63 mg/mL), A. spicifera (IC 50 0.71 mg/mL), and P. hornemanii (IC 50 1.79 mg/mL), in descending order. Inflammatory responses analyzed via cyclooxygenase-2 isoform (COX-2) exhibited a similar pattern, wherein H. edulis (IC 50 0.61 mg/mL) showed significantly higher inhibition activity than the rest of the extracts screened. The results of this study are consistent with that of prior reports focused on antiinflammatory activity of G. textorii (Okada et al. 1995), G. verrucosa (Yoshizawa et al. 1996), G. salicornia (Chakraborty et al. 2019a;Antony and Chakraborty 2020), and G. cornea (Coura et al. 2012), and it is also in agreement with findings showing antiinflammatory activity of aqueous extract compared to methanolic extract of H. edulis (Vijayalakshmi 2015). Likewise, polysaccharides from H. edulis showed inhibition activities against ACE-I and HMGCR enzymes (IC 50 0.75-0.87 mg/mL).

Fractionation of crude polysaccharide of H. edulis and chemical analysis
Based on preliminary biological activities, the crude polysaccharide of H. edulis was deproteinated and further purified on a DEAE-cellulose anion exchanger to yield four fractions, namely SP-He-1, SP-He-2, SP-He-3 and SP-He-4 using subsequent elution with distilled water, 0.1 N, 0.2 N, and 0.3 N of NaCl, respectively ( Figure S1). The yields of collected fractions ranged between 14.3% and 32.5% (Table S2).
The crude polysaccharide SP-He of H. edulis was comprised of 55.2% carbohydrate, 5.36% protein, 8.5% sulfate, 2% uronic acid, and 9.07% pyruvic acid, while the monosaccharide analysis detected the presence of galactose, anhydrogalactose, and xylose in 62.5, 35 and 2 mol %, respectively (Table S2). The purified fractions demonstrated carbohydrate contents between 65.6-19.1%, which are consistent within the range of carbohydrate content reported in other Gracilaria species from India (Sudharsan et al. 2015). The content of protein was noticed in trace amounts for SP-He-2 (1.38%) and SP-He-3 (0.04%) after deproteination. Fraction SP-He-1 (9.43%) registered higher amount of uronic acid in comparison with other fractions, SP-He-2 (0.97%), SP-He-3 (2.19%), and SP-He-4 (1.97%). These fractions exhibited sulfate contents ranging from 2.36% to 0.19% of which SP-He-2 being the highest. Purified fractions displayed pyruvic acid content (16-0.07%), which also accords with earlier observations of Chiovitti et al. (1996) demonstrating that red algal polysaccharides from different sources are rich in pyruvate acetal or methyl ether substitution. Monosaccharide analysis of purified fractions using HPLC-ELSD detected galactose and anhydrogalactose as dominant sugars with minor quantities of xylose and glucose ( Figure S2). This study produced the results which corroborate the findings of Pomin and Mourão (2008) who mentioned the presence of sulfated galactans in red algae. The purified fraction designated as SP-He-2 displayed higher carbohydrate content exemplified by the presence of galactose (58%), anhydrogalactose (40%), pyruvic acid (15.4%), sulfate (2.36%), and a trace amount of uronic acid (0.97%). Hence, fraction SP-He-2 was considered for extensive study using spectroscopic techniques.

Spectroscopic description of purified SP-He-2 from H. edulis
The molecular mass of total polysaccharide was calculated via HRESIMS negative mode and found m/z at 513.4296 ( Figure S3). UV-visible spectrum of SP-He-2 ( Figure  S4) in the region 200-800 nm showed sharp peaks below 250 nm, and a broad shoulder between 290 and 310 nm as described earlier by Srivastava and Kumar (2013). These peaks could be attributed to carbonyl moiety of pyruvic acid ketal in SP-He-2.
The purified polysaccharide of H. edulis was analyzed using IR spectroscopy in order to identify specific absorption bands (Table S3). The polysaccharide displayed IR band ( Figure S5) at 3462 cm À1 signifying hydroxyl stretching frequencies while 2930 cm À1 designated -CH stretching vibrations (Sudharsan et al. 2015). The bands around the region 1340-1260 cm À1 were due to sulfate esters, wherein that at 830 cm À1 was assigned to secondary equatorial sulfate at C2 of the 1,3-linked D-galactose (Zablackis and Perez 1990). As these bands exhibited a low intensity, it was considered that degree of substitution of sulfate esters is low in the expected structure (Barros et al. 2013). Signals stroked at 1261 and 1143 cm À1 showed C-C, C-O stretching in pyranosyl ring as well as C-O-C stretching of glycosidic bonds, whereas those at 1077, 746 and 687 cm À1 were ascribed to skeleton bending of galactose ring (Bedoux et al. 2017, Sudharsan et al. 2015. The presence of 3,6-anhydro-a-L-galactose was confirmed by C-O-C bridge band vibration at 930 cm À1 (Bedoux et al. 2017) and absence of bands at 825 cm À1 indicated hydroxyl group than sulfate at C6 of galactose residue (Alencar et al. 2019). Further, sulfated polysaccharide from H. edulis did not produce any bands at 805 and 845 cm À1 thereby demonstrating the absence of sulfate esters at C2 of 3,6-anhydro-L-galactose and C4 of D-galactose residue (Alencar et al. 2019). Apart from this, IR spectrum demonstrated an unusual band at 1620 cm À1 , which provided evidence for existence of pyruvic acid ketal substitution or acetate residues. IR data perceptibly showed the presence of sulfate esters, which is reported to have a dynamic part in the bioactivity of red algal polysaccharides (Sudharsan et al. 2015). Thus, outcomes of IR spectrum identified the structure of polysaccharide to be consisting of pyruvated C2-sulfated 1,3-linked-D-galactose and 3,6-anhydro-a-L-galactose with no sulfation.

In vitro anti-hyperglycemic potential of SP-He-2
DPP-IV is a multifunctional transmembrane glycoprotein involved in enzymatic incretin degradation by inhibiting glucose-dependent insulinotropic peptide and glucagon-like peptide-1, which enhance post-prandial insulin secretion from b-cells of pancreas. DPP-IV inhibitors act by extending the life of incretin hormones, thus enhancing insulin secretion, thereby lowering blood glucose level (Patil et al. 2015). Sulfated polysaccharides are natural DPP-IV inhibitors, and have been proven as effective in developing anti-hyperglycemic agents (Zhang et al. 2008, Maneesh and Chakraborty 2018, Antony et al. 2022. The sulfated pyruvylated polysaccharide SP-He-2 obtained from H. edulis displayed attenuation potential against DPP-IV (IC 50 4.44 lM) (Table S2), and was comparable to that exhibited by diprotin-A (IC 50 4.21 lM). Noticeably, the polar/electrostatic interaction between DPP-IV and negatively charged sulfate group of SP-He-2 of H. edulis might restrain the activities of the proteolytic enzymes. Conspicuously, inhibition of enzyme catalyzing the breakdown of complex carbohydrates is another therapeutic approach that could retard post prandial hyperglycemia (Nasab et al. 2020). SP-He-2 showed greater inhibition potential against carbolytic enzymes a-amylase (IC 50 45.25 lM) and a-glucosidase (IC 50 47.02 lM) compared to acarbose (IC 50 50.56 and 50.15 lM, respectively) ( Table S2). The correlation between SP-He-2 and anti-hyperglecemic activity could also be associated with the formation of a complex between sulfated pyruvylated polysaccharide and carbolytic enzyme thereby resulting in a slower diffusion of glucose, and its disrupted absorption in the cellular system (Cho et al. 2011). Thus, the attenuation potential of SP-He-2 against DPP-IV and carbolytics could be important to attribute its promising anti-hyperglycemic property.

Kinetics of DPP-IV inhibition and molecular docking analysis
The isolated polysaccharide SP-He-2 was deduced to inhibit DPP-IV in a non-competitive manner as determined by the Lineweaver-Burk plot ( Figure S13). This inferred that SP-He-2 was bound at a different site from that of the substrate, and consequently, decreases the efficacy of enzyme catalysis (Chakraborty et al. 2019b). The SP-He-2 was subjected to molecular docking against DPP-IV (Table S5) resulting in lesser binding and intermolecular energies of À1.76 and À3.24 kcal mol À1 , respectively in addition to form five hydrogen bonds (three with Ser 239.A , Ser 239.B and residual two with Tyr 241.B , and Phe 713.A ) with the binding site of DPP-IV (Table S5, Figure S14), which substantiated its promising inhibitory activity against the adenosine deaminase complexing glycoprotein (IC 50 4.44 lM). It was reported that potent inhibitors of DPP-IV, such as saxagliptin interacts with serine residue at the enzyme active site to form covalent bonds, thus reducing blood glucose level (Wang et al. 2019).

Chemical and instrumentation
All chemicals, reagents and solvents utilized in the present study were procured from Sigma-Aldrich (St. Louis, MO, USA), E-Merck (Darmstadt, Germany), HiMedia (West Chester, PA, USA) and Sisco Research Laboratories (Mumbai, India), and were of analytical grade. Details of instrumentation are given in the supporting information (supplementary information S1). Nuclear magnetic resonance (NMR) spectra (one and two dimensional) of purified polysaccharide were recorded at Bruker AVANCE III spectrometer operating at 500 MHz (AV 500) (Bruker, Karlsruhe, Germany) using deuterated water (D 2 O) as solvent and tetramethylsilane (TMS) as the internal standard. 1 H, 13 C and 135 DEPT NMR spectra were recorded at 512, 12,000 and 6000 scans, respectively, while 1 H-1 H-COSY and HSQC experiments were performed at 8 and 16 scans, respectively. Obtained results were processed using MestReNova (version 7.1.1-9649, Mestrelab Research S.L) software. . The macroalgal extracts were prepared through the alkaline treatment method (Kravchenko et al., 2020 ) with suitable modifications to obtain crude polysaccharides. Briefly, macroalgae were treated for depigmentation using acetone:methanol mixture (3 Â 1 L) to remove organic soluble fractions. The depigmented macrolgae were then soaked in 5% KOH (1 L) for 3-4 h and heated at 80-85 C for alkali treatment followed by filtering and washing with distilled water to eliminate excess KOH salt. The filtrate was subjected to neutralization using 3 M HCl (1 L). After filtration, polysaccharide was extracted using NaHCO 3 (pH 8-9, 1:50 w/v) for 3 h at elevated temperature (60-80 C) before being added with KCl (1%, 500 mL). The content was centrifuged (8000 rpm, 4 C) to obtain crude polysaccharide as residue, which was lyophilized.

Purification of sulfated pyruvylated polysaccharide
The polysaccharide from H. edulis (SP-He) was subjected to deproteination as reported elsewhere (Staub 1965; supplementary information S3), and yield was calculated. The crude polysaccharide was subjected to anion exchange chromatography using diethylaminoethyl (DEAE) cellulose on a glass column (25 Â 4 cm) (Maneesh and Chakraborty 2018;Chakraborty et al. 2020a). Initially, slurry was prepared with 9 g of DEAE-cellulose in 30 mL tris buffer (50 mM, pH -7.4) and kept it for swelling for 1 h followed by packing the column with same. The crude polysaccharide (2 g) dissolved in water (15 mL) was applied to the packed anion exchanger column, and was washed with several column-volumes of tris buffer followed by different concentrations of NaCl (0.1-0.3 N). The collected fractions ( Figure S1) were tested for carbohydrate (Supplementary information S3) after lyophilization.

Chemical methods, analysis of monosaccharide composition and linkage analysis
Samples were analyzed for the presence of uronic acids, whereas sulfate content was determined by turbidimetric procedure. Protein quantification was done by Lowry's method, and pyruvic acid by utilizing pyruvic acid as the standard (supplementary information S3). The polysaccharide fractions ($5 mg) were subjected to mild acid hydrolysis using trifluoroacetic acid (TFA, 3 M; 60 C, 20 mg in 4 mL) for monosaccharide compositional analysis  before being concentrated to eliminate the traces of trifluoroacetic acid (Antony et al. 2022). The analysis of monosaccharide composition was performed by following the method reported by Maneesh and Chakraborty (2018) using the HPLC-ELSD (high-performance liquid chromatography-elusive light scattering detector) (Chakraborty et al. 2020b) for hydrolyzed polysaccharide sample of H. edulis in which galactose (Gal), anhydrogalactose (AnGal), glucose (Glc), mannose (Mann) and xylose (Xyl) were used as the standards. To carry out linkage analysis, it is imperative to desulfate and depyruvylate the polysaccharide. Briefly, an aliquot of polysaccharide (15 mL) in water was passed through a column loaded with an Amberlite CG-120 resin (200-400 mesh) to collect the eluate, which was neutralized using pyridine (1 mL) before being lyophilized. This was dissolved in dimethyl sulfoxide (DMSO) (20 mL) and pyridine (1 mL), incubated at 100 C for 3 h and was dialyzed using distilled water before being lyophilized to obtain desulfated polysaccharides (Ermakova et al., 2016). Pyruvic acid residues from desulfated polysaccharide were removed (Bilan et al. 2007), wherein the sample (50 mg) was treated with acetic acid (1% v/v, 10 mL) under heating ($100 C) for 4 h. This was followed by neutralization using NaHCO 3 , dialyzation and lyophilization to obtain depyruvylated desulfated polysaccharide. Desulfated and depyruvylated polysaccharide (2 mg) was methylated using reductive cleavage (Kiwitt-Haschemie et al. 1996). Initially, polysaccharide was dissolved in DMSO (3 mL) and stirred at 80 C until it dissolves. This was followed by addition of NaOH (100 mg) and continued stirring at room temperature for 3 h. Methyl iodide (MeI, 0.2 mL) was added to the mixture, and later quenched with distilled water (1 mL) after 2.5 h. The methylated sample was extracted using dichloromethane, and excess MeI was removed under reduced pressure, passed through anhydrous sodium sulfate, and dried over a nitrogen atmosphere. Reductive depolymerisation and acetylation of the methylated sample was performed in the presence of trimethylsilyl methanesulfonate (TMS-O-mesylate) (5 equivalent/glycosidic bond) and boron trifluoride etherate (BF 3 À etherate) (1 equivalent/glycosidic bond). Acetylation was carried out in situ with a mixture of acetic anhydride and trifluoroacetic acid (TFA).

Kinetics of DPP-IV inhibition
Kinetic analyses by Lineweaver-Burk plot were used to assess the mechanism to inhibit DPP-IV by the purified polysaccharide SP-He-2. The DPP-IV inhibition was analyzed at different concentrations (0.0625, 0.125, 0.25 and 0.5 mM) of the substrate (Gly-Pro-pnitroanilide), in the presence and absence of the polysaccharide, in Lineweaver-Burk double reciprocal plot. The polysaccharide was taken in different concentrations, and the inhibitory activities were expressed in mM.

In silico molecular docking analysis
Molecular docking simulation study of polysaccharide unit was analyzed using Autodock 4 software to realize the binding of ligand to enzyme. The structure of polysaccharide motif was built up by Chemsketch 2.5, and Open Babel 2.4.1 (3 D conformation) was utilized for docking experiments along with crystal structure of human DPP-IV (downloaded from protein data base) (Antony et al. 2022). USCF Chimera 1.11.2 was utilized for visualizing the docking results.

Statistical analysis
Statistical Program for Social Sciences 13.0 (SPSS Inc., Chicago, USA, ver. 13.0) was used for performing statistical studies. In order to evaluate the significance level (p < 0.05), variance analysis (ANOVA) of triplicate results was conducted.