In Vitro Anti-Oxidant, In Vivo Anti-Hyperglycemic, and Untargeted Metabolomics-Aided-In Silico Screening of Macroalgae Lipophilic Extracts for Anti-Diabetes Mellitus and Anti-COVID-19 Potential Metabolites

COVID-19 patients with comorbid DM face more severe outcomes, indicating that hyperglycemic conditions exacerbate SARS-CoV-2 infection. Negative side effects from existing hyperglycemia treatments have urged the need for safer compounds. Therefore, sourcing potential compounds from marine resources becomes a new potential approach. Algal lipids are known to possess beneficial activities for human health. However, due to limitations in analyzing large amounts of potential anti-hyperglycemic and anti-COVID-19-related marine metabolites, there is an increasing need for new approaches to reduce risks and costs. Therefore, the main aim of this study was to identify potential compounds in macroalgae Sargassum cristaefolium, Tricleocarpa cylindrica, and Ulva lactuca lipophilic extracts for treating DM and COVID-19 by an integrated approach utilizing in vitro anti-oxidant, in vivo anti-hyperglycemic, and metabolomic-integrated in silico approaches. Among them, S. cristaefolium and T. cylindrica showed potential anti-hyperglycemic activity, with S. cristaefolium showing the highest anti-oxidant activity. A GC-MS-based untargeted metabolomic analysis was used to profile the lipophilic compounds in the extracts followed by an in silico molecular docking analysis to examine the binding affinity of the compounds to anti-DM and anti-COVID-19 targets, e.g., α-amylase, α-glucosidase, ACE2, and TMPRSS2. Notably, this study reveals for the first time that steroid-derived compounds in the macroalgae T. cylindrica had higher binding activity than known ligands for all the targets mentioned. Studies on drug likeliness indicate that these compounds possess favorable drug properties. These findings suggest the potential for these compounds to be further developed to treat COVID-19 patients with comorbid DM. The information in this study would be a basis for further in vitro and in vivo analysis. It would also be useful for the development of these candidate compounds into drug formulations.


Introduction
Diabetes mellitus (DM) is still a global health problem.Individuals with DM are also prone to experience worse conditions when infected by other diseases, such as COVID-19 [1][2][3].Increased disease severity and mortality among COVID-19 patients when DM co-occurs was observed.Patients suffering from DM experience more severe COVID-19 symptoms than non-diabetic patients [4][5][6], because DM is a risk factor for upregulated angiotensin-converting enzyme-2 (ACE2) gene expression due to hyperglycemic conditions [7,8].Hyperglycemia in DM triggers glycosylation of the host ACE2 and viral S-protein [9], thereby increasing the binding between the virus and ACE2.This, in turn, facilitates the entry of the virus [10], thereby worsening the condition of COVID-19 patients [11].Another study highlighted that COVID-19 patients with DM had increased levels of interleukin-6 and C-reactive protein compared with non-DM patients [12].Some anti-DM drugs, such as α-glucosidase inhibitors, have been used against COVID-19 [13], and findings suggest that these compounds show neutral mortality rates in COVID-19 patients compared with other anti-DM drugs [14].However, conventional treatments for DM have drawbacks, including gastrointestinal side effects, such as abdominal pain and diarrhea [15,16].Further complications, including weight gain, cardiac conditions, anemia, headache, hypoglycemia, nausea, vomiting, fatigue, constipation, and dyspepsia, have also been associated with standard anti-DM drugs [17].This highlights the urgent need for new compounds that can effectively combat hyperglycemia and COVID-19 but have fewer side effects.
The search for novel drug compounds is increasingly turning to natural resources, with marine biota of particular interest.Macroalgae, abundant in tropical regions, have become a focal point in these efforts.Among them, Sargassum cristaefolium, Tricleocarpa cylindrica, and Ulva lactuca are the prominent macroalgae species found along the coast of Lombok in Indonesia.S. cristaefolium is a member of Phaeophyta (brown macroalgae), T. cylindrica is a member of Rhodophyta (red macroalgae), and U. lactuca is a member of Chlorophyta (green macroalgae).Previous studies have demonstrated the diverse pharmacological potential of macroalgae, including anti-oxidant, anti-aging, and antimelanoma activities [18,19].The lipophilic extract of the brown macroalga Sargassum polycystum was able to decrease hyperglycemia in alloxan-induced rats [20].Another study also showed that macroalgae lipophilic compounds have the ability to reduce inflammation and oxidation [21][22][23].Meanwhile, the Ulva fasciata lipid extract mainly containing palmitic acid has been shown to have an anti-oxidant and anti-genotoxic effect [24].Moreover, non-flavonoid compounds from brown macroalgae show higher bioactivities than the other classes [25].These non-flavonoid compounds include lipophilic substances, such as fatty acids, steroids, and terpenoids [25].
Fatty acids have demonstrated efficacy against metabolic diseases, including DM [26,27], through α-glucosidase [28][29][30][31] and α-amylase inhibitory activities [32,33].The consumption of Omega-3 fatty acids was able to reduce COVID-19 complications [34].Meanwhile, sterols are known for their α-amylase and α-glucosidase inhibitory [35][36][37] and anti-viral effects [38].More specifically, stigmasterol and β-sitosterol were able to improve glycemic regulation and stimulate pancreatic β-cells insulin secretion [39].Several studies have also observed the activities of macroalgae lipid compounds.The fatty acids of Antarctic macroalgae have potential cytotoxic activity [40], and sterols from macroalgae have been shown to have α-glucosidase and α-amylase inhibitor activities [41].Meanwhile, the brown macroalga Turbinaria ornata contains sterols that have cytotoxicity effects on cancer cells [42][43][44].Several other macroalgae, including Himanthalia elongate, Undaria pinnatifida, Phorphyra spp., Chondus crispus, Cystoseira spp., and Ulva spp., show anti-cholesterol activity by the reduction in the total and LDL-C levels [45] through their sterol compounds.Meanwhile, fucosterol isolated from the brown macroalga Pelvetia siliquosa was able to reduce hyperglycemia by 25-33% [46].Other brown macroalgae, Eisenia bicyclis and Ecklonia stolonifera, also contain fucosterols that are able to inhibit α-glucosidase [47].Meanwhile, other studies have shown that fucosterol from macroalgae has the ability to inhibit ACE in endothelial cells [48], which shows the potential anti-viral activity of this compound.Interestingly, COVID-19 patients with DM had a reduced risk of death after treatment with α-glucosidase inhibitors compared with other anti-DM approaches [49].Thus, further research is needed to explore the potential usage of lipophilic compounds from marine resources as a treatment for DM and COVID- 19.In this study, the anti-hyperglycemic efficacy of macroalgal lipophilic extracts using an alloxan-induced diabetic rat model and their anti-oxidant activity were assessed.A Gas Chromatography-Mass Spectrometry (GC-MS)-based untargeted metabolomics analysis on the lipophilic extracts of S. cristaefolium, T. cylindrica, and U. lactuca was also performed.This was followed by evaluating the activity of the identified compounds against anti-DM and anti-COVID-19 protein targets using molecular docking and bioinformatics analysis.

Sample Collection, Preparation, and Lipophilic Compounds Extraction
The aim of this study was to compare the differences in the activity of three classes of macroalgae found in an abundant amount on the Lombok, Indonesia, coast, namely, Phaeophyta (brown macroalgae), represented by S. cristaefolium; Rhodophyta (red macroalgae), represented by T. cylindrica; and Chlorophyta (green macroalgae), represented by U. lactuca.The macroalgae were sourced from the North Lombok, West Nusa Tenggara, Indonesia, coast.The collection of the macroalgae samples was performed when the tide was low and only healthy macroalgae were used in this study.The collected macroalgae samples were identified by referring to electronic algae databases [50] and other literature (S. cristaefolium based on the description in [51], T. cylindrica based on the description in [52], and U. lactuca based on the description in [53]).
The collected macroalgae samples were immediately cleaned using clean seawater.These samples were then air-dried for three days followed by dehydration at 30 • C for 24 h.Once dry, the samples were ground to achieve an optimal particle size of 40 µm.This particle size was recommended for macroalgae [54].The powdered samples underwent lipid extraction using chloroform:methanol:water (2:1:0.5).Initially, the powdered samples were soaked with the solvent mixture and mixed for 5 min at 4 • C.This was then followed by sonication for 15 min.After 15 min, the mixture was centrifugated at 5000 rpm, 2 • C for 30 min, leading to the formation of two phases.The organic phase, which contains lipophilic compounds was then collected and then filtered.This extraction was performed three times, as this was the most sufficient number of replicates required for the extraction method, and the extracts from each run were combined.The solvent was evaporated for approximately 3 h at below 30 • C on a water bath where the sample was placed on ceramic porcelain covered with pierced aluminum foil to avoid oxidation and light denaturation of the compounds.The obtained thick extract was then allocated for further analysis.

In Vivo Anti-Diabetes Analysis in an Alloxan-Induced Diabetic Mouse Model
Male Wistar rats (Rattus norvegicus) with an average weight ranging from 200 to 250 g were used as the model animals in this study; all the rats were housed in ventilated cages and had free access to rat food and clean water.The rats were obtained from the Animal Laboratory Facility, Faculty of Medicine, University of Mataram, Indonesia.This research received ethical permission from the Medical Research Ethics Committee, Faculty of Medicine, University of Mataram, Indonesia (Approval No:285/UN18.F7/ETIK/2022), and all the experiments on the animals were performed based on the guidelines of the proper care and use of laboratory animals.The rats were classified into six groups (n = 4 per group): (1) untreated rats; (2) diabetic rats treated with the vehicle (0.5% CMC-Na with NaCl solution); (3) diabetic rats treated with glibenclamide (diluted in 0.5% CMC-Na with NaCl solution); (4) diabetic rats treated with S. cristaefolium lipophilic extract; (5) diabetic rats treated with T. cylindrica lipophilic extract; and (6) diabetic rats treated with U. lactuca lipophilic extract.All the tested samples were diluted in 0.5% CMC-Na with NaCl solution.Before being treated, the rats underwent an acclimatization phase for one week, where they were given standard rat food and clean water.Next, hyperglycemia was induced in Groups 2 to 6 by administering two intraperitoneal injections of alloxan monohydrate (150 mg/kgBW) (Sigma Aldrich, Gillingham, UK) for two consecutive days.The fasting blood glucose levels were assessed before induction and 72 days post alloxan induction using the Easy-Touch-GCU device (Bioptik Technology, Inc., Jhonghua, Taiwan).Only the rats showing fasting blood glucose levels exceeding 126 mg/dL were categorized as diabetic [55,56] and were included in the next phase of the study.
Post confirmation of the hyperglycemia status, the treatment was performed as follows: the rats in Group 1 received only food and water; Group 2 was given a vehicle solution; Group 3 was given glibenclamide orally at a dose of 0.45 mg/kgBW; and the rats in Groups 4 to 6 were treated orally with lipophilic extracts at a daily dose of 100 mg/kgBW, respectively.The fasting blood glucose was measured on days 0, 5, 10, and 15 after treatment.The recorded data underwent a statistical analysis using a homogeneity test followed by a two-way ANOVA with Dunnett's multiple comparison test to confirm the significant differences in the fasting blood glucose levels between the treated and control diabetic rats.GraphPad Prism version 9 was used for the statistical analysis.

DPPH Assay
The DPPH assay was performed following the method described in [57], with some modifications.The lipophilic extract was subjected to serial dilution.For every 100 µL of the sample and blank (ethanol), 100 µL of the 200 µM DPPH solution (Sigma Aldrich, Gillingham, UK) was added.This procedure was carried out in triplicate.The anti-oxidant activity was determined by where the Scavenging effect is a percentage that shows how effective the sample is in neutralizing free radicals.Acontrol represents the control absorbance, which usually contains all reagents except the sample being tested.It provides basic absorption without any anti-oxidant activity.Asample is the absorbance of the tested sample when mixed with a DPPH solution.By subtracting the sample absorbance from the control absorbance, we obtain a measure of how much the sample has reduced the initial amount of DPPH radicals, indicating its anti-oxidant ability.

ABTS Assay
The scavenging activity of the lipophilic extracts against the ABTS radical cations was assessed following the method described in [58], with minor modifications.Stock solutions were prepared using 7 mM ABTS in water and 2.4 mM potassium persulfate.To prepare the working solution, both stock solutions in equal volumes were mixed and incubated for 16 h at room temperature in the dark.This solution was then diluted by combining 250 µL of the ABTS mixture with 12 mL of ethanol, aiming to achieve an absorbance of approximately 0.700 units at 734 nm as measured with a spectrophotometer.For the test, 1 mL of samples at various concentrations was combined with 1 mL of dilute ABTS solution.After incubation for 7 min at room temperature, the absorbance was measured at 734 nm.The ABTS scavenging capacity of the extract was determined using Equation (1).

Untargeted Gas Chromatography-Mass Spectrometry (GC-MS) Metabolomics Analysis
The lipophilic extracts from S. cristaefolium, T. cylindrica, and U. lactuca that were obtained by extraction (Section 2.1) were then analyzed using GC-MS.For the sample preparation, the extract was diluted in chloroform:methanol, mixed in a microtube, vortexed, and centrifuged at 9500 rpm for 3 min.The resulting supernatant was transferred to a GC vial and injected into a Shimadzu GCMS-QP2010S instrument (Shimadzu Corporation, Kyoto, Japan).A 30 m long Agilent HP-5MS UI column was used.The carrier gas was Ultra High Purity Helium (He), with the settings including an injector temperature of 250 • C, a split flow of 510 mL/min, a front inlet flow of 1.00 mL/min, an MS transfer line temperature of 300 • C, and an ion source temperature of 280 • C. The compound identities were analyzed by comparing the observed properties with the properties in the NIST MainLib Library.

In Silico Analysis of Anti-Diabetic Activity and Anti-COVID-19 Activity
Molecular docking studies were carried out in silico to test the impact of the lipophilic compounds identified from the macroalgae extracts on the protein target receptors associated with DM and COVID-19.These receptors include α-amylase, α-glucosidase, angiotensin-converting enzyme 2 (ACE2), and transmembrane serine protease 2 (TM-RPSS2).The respective native ligand for α-amylase and α-glucosidase was acarbose, for ACE2 it was captopril, and for TMRPSS2 it was nafamostat.The receptor preparation was carried out using DS BIOVIA Discovery Studio 2016 [59].The ligand preparation includes creating a 2D structure using ChemDraw Ultra [42] and converting it to a 3D structure using Chem3D Pro [60].During the re-docking phase, the grid box size was set using AutoDockTools [61], and the molecular docking simulations were performed with AutoDock Vina [62].An analysis of the receptor-ligand complex interactions was performed with DS BIOVIA Discovery Studio 2016 [59], and the visualization of the binding sites and interactions between the ligand and receptor was facilitated by PyMOL [63].The drug similarity scores and oral bioavailability, based on Lipinski's rule, were determined using SWISS ADME tool (http://www.swissadme.ch(accessed on 1 April 2023).Figure 1 outlines this workflow.
ured at 734 nm.The ABTS scavenging capacity of the extract was determined using Equation (1).

Untargeted Gas Chromatography-Mass Spectrometry (GC-MS) Metabolomics Analysis
The lipophilic extracts from S. cristaefolium, T. cylindrica, and U. lactuca that were obtained by extraction (Section 2.1) were then analyzed using GC-MS.For the sample preparation, the extract was diluted in chloroform:methanol, mixed in a microtube, vortexed, and centrifuged at 9500 rpm for 3 min.The resulting supernatant was transferred to a GC vial and injected into a Shimadzu GCMS-QP2010S instrument (Shimadzu Corporation, Kyoto, Japan).A 30 m long Agilent HP-5MS UI column was used.The carrier gas was Ultra High Purity Helium (He), with the settings including an injector temperature of 250 °C, a split flow of 510 mL/min, a front inlet flow of 1.00 mL/min, an MS transfer line temperature of 300 °C, and an ion source temperature of 280 °C.The compound identities were analyzed by comparing the observed properties with the properties in the NIST MainLib Library.

In Silico Analysis of Anti-Diabetic Activity and Anti-COVID-19 Activity
Molecular docking studies were carried out in silico to test the impact of the lipophilic compounds identified from the macroalgae extracts on the protein target receptors associated with DM and COVID-19.These receptors include α-amylase, α-glucosidase, angiotensin-converting enzyme 2 (ACE2), and transmembrane serine protease 2 (TMRPSS2).The respective native ligand for α-amylase and α-glucosidase was acarbose, for ACE2 it was captopril, and for TMRPSS2 it was nafamostat.The receptor preparation was carried out using DS BIOVIA Discovery Studio 2016 [59].The ligand preparation includes creating a 2D structure using ChemDraw Ultra [42] and converting it to a 3D structure using Chem3D Pro [60].During the re-docking phase, the grid box size was set using AutoDock-Tools [61], and the molecular docking simulations were performed with AutoDock Vina [62].An analysis of the receptor-ligand complex interactions was performed with DS BIO-VIA Discovery Studio 2016 [59], and the visualization of the binding sites and interactions between the ligand and receptor was facilitated by PyMOL [63].The drug similarity scores and oral bioavailability, based on Lipinski's rule, were determined using SWISS ADME tool (http://www.swissadme.ch).Figure 1 outlines this workflow.

In Vivo Anti-Hyperglycemic Activity of Sargassum Cristaefolium, Tricleocarpa cylindrica, and Ulva lactuca Lipophilic Extract
A preliminary in vivo assay was performed to evaluate the anti-hyperglycemic potential of the macroalgae lipophilic extracts, which is an important aspect in the management of DM.The differences in the activity of the three classes of macroalgae, namely, Phaeophyta, Rhodophyta, and Chlorophyta, were observed.We aimed to choose the species that were found abundantly on the Lombok, Indonesia, coast.S. cristaefolium was chosen to represent Phaeophyta, T. cylindrica to represent Rhodophyta, and U. lactuca to represent Chlorophyta.
The rats, after receiving two intraperitoneal doses of alloxan monohydrate (150 mg/kgBW) for two days, showed an increase in the fasting blood sugar levels ranging from approximately 300 to 350 mg/dL (Figure 2).This increase was significantly greater than that observed in the untreated rats.These hyperglycemic rats were then chosen as a DM model and given glibenclamide (0.45 mg/kgBW) and lipophilic macroalgae extract (100 mg/kgBW) every day orally for 15 days.
The rats, after receiving two intraperitoneal doses of alloxan monohydrate (150 mg/kgBW) for two days, showed an increase in the fasting blood sugar levels ranging from approximately 300 to 350 mg/dL (Figure 2).This increase was significantly greater than that observed in the untreated rats.These hyperglycemic rats were then chosen as a DM model and given glibenclamide (0.45 mg/kgBW) and lipophilic macroalgae extract (100 mg/kgBW) every day orally for 15 days.A decrease in the fasting blood glucose was clearly visible in the rats treated with glibenclamide on day 5 post-treatment and continued to decrease until day 15.In contrast, treatment with the macroalga S. cristaefolium lipophilic extract caused a significant reduction in the fasting blood sugar levels on day 10, and this trend was maintained until day 15 after treatment.Meanwhile, another macroalga, T. cylindrica, did reduce the blood glucose levels on day 15 in the alloxan-induced rat model; however, the effect was not as significant as S. cristaefolium.Meanwhile, U. lactuca showed lower anti-hyperglycemia activity compared to the other macroalgae, S. cristaefolium and T. cylindrica (Figure 2).These preliminary findings demonstrate the anti-hyperglycemic properties of the lipophilic extracts.

In Vitro Anti-Oxidant Activity of Sargassum cristaefolium, Tricleocarpa cylindrica, and Ulva lactuca Lipophilic Extract
The in vitro anti-oxidant studies revealed that the lipophilic extracts of S. cristaefolium, T. cylindrica, and U. lactuca exhibited anti-oxidant activity even though the activity was considered a low-potential anti-oxidant.Specifically, they showed IC50 values of 206.7 ± 0.11 µg/mL, 252 ± 0.10 µg/mL, and 308.6 ± 0.13 µg/mL, respectively, as determined by the DPPH assay.Meanwhile, the ABTS assay produced IC50 values of 200.47 ± 0.089 µg/mL, 353.22 ± 0.070 µg/mL, and 386.42 ± 0.050 µg/mL for the same extract.Ascorbic acid A decrease in the fasting blood glucose was clearly visible in the rats treated with glibenclamide on day 5 post-treatment and continued to decrease until day 15.In contrast, treatment with the macroalga S. cristaefolium lipophilic extract caused a significant reduction in the fasting blood sugar levels on day 10, and this trend was maintained until day 15 after treatment.Meanwhile, another macroalga, T. cylindrica, did reduce the blood glucose levels on day 15 in the alloxan-induced rat model; however, the effect was not as significant as S. cristaefolium.Meanwhile, U. lactuca showed lower anti-hyperglycemia activity compared to the other macroalgae, S. cristaefolium and T. cylindrica (Figure 2).These preliminary findings demonstrate the anti-hyperglycemic properties of the lipophilic extracts.

In Vitro Anti-Oxidant Activity of Sargassum cristaefolium, Tricleocarpa cylindrica, and Ulva lactuca Lipophilic Extract
The in vitro anti-oxidant studies revealed that the lipophilic extracts of S. cristaefolium, T. cylindrica, and U. lactuca exhibited anti-oxidant activity even though the activity was considered a low-potential anti-oxidant.Specifically, they showed IC 50 values of 206.7 ± 0.11 µg/mL, 252 ± 0.10 µg/mL, and 308.6 ± 0.13 µg/mL, respectively, as determined by the DPPH assay.Meanwhile, the ABTS assay produced IC 50 values of 200.47 ± 0.089 µg/mL, 353.22 ± 0.070 µg/mL, and 386.42 ± 0.050 µg/mL for the same extract.Ascorbic acid was used as a positive control for the assay and showed IC 50 values of 3.667 ± 0.02 µg/mL (DPPH assay) and 5.58 ± 0.02 µg/mL (ABTS assay) as summarized in Table 1, showing that the method performed well.The IC 50 values of the macroalgae lipophilic extracts obtained in this study are lower than ascorbic acid but are still considered acceptable and in the normal range for crude macroalgae extracts, as shown in other studies.The acetone extracts of the macroalga Cystoseira amantacea showed an IC 50 value of 408.81 µg/mL with the DPPH method [64].Meanwhile, another study has also shown that the ethyl acetate extract of Sargassum angustifolium had an anti-oxidant IC 50 value of 400 µg/mL by using the ABTS method [65].This is due to the fact that the macroalgae lipophilic extract is still a whole crude extract and not a single isolated compound.The lipophilic extract of S. cristaefolium had superior anti-oxidant activity compared to T. cylindrica and U. lactuca.This is in line with the anti-hyperglycemic data where the lipophilic extract of S. cristaefolium produced a more pronounced reduction in fasting blood glucose, followed by T. cylindrica, and then U. lactuca showed the least effective.These observations suggest that the anti-hyperglycemic action of S. cristaefolium lipophilic extract may be mediated through the reduction in ROS.Contemporary advances in metabolic research have identified that molecules with anti-oxidant capacity can influence blood glucose concentrations and monitor changes in biochemical indicators related to glucose metabolism [66].Anti-oxidants may improve insulin sensitivity and modulate carbohydrate metabolism [67].In addition, pioglitazone, a known type-2 anti-diabetic agent and peroxisome proliferator-activated receptor agonist, regulates glucose regulation through its anti-oxidant activity.Thus, entities with anti-oxidant attributes show promise in their potential as anti-diabetic agents.In addition, the solid anti-oxidant activity of S. cristaefolium may be due to the synergy of several bioactive compounds, not just the identified lipophilic molecules.It is important to understand that although these findings are significant, further isolation and testing of these compounds is necessary to decipher the exact mechanisms underlying their anti-oxidant properties.

Untargeted GC-MS Metabolomics Analysis of Macroalgae Lipophilic Extract
Performing a phytochemical analysis of the compounds in extracts helps identify new compounds with potential bioactivity.Therefore, in this study, after verifying the anti-hyperglycemic and anti-oxidant properties of the lipophilic macroalgae extracts, the metabolites in the extracts were analyzed.A GC-MS-based untargeted metabolomics approach was used to profile the compounds present in the lipophilic extracts of the three macroalgae: a total of 37 peaks were observed in S. cristaefolium, 41 peaks were observed in the T. cylindrica extract, and 37 peaks were observed in the U. lactuca extract.Figure 3 displays the GC-MS spectrum.Each peak was identified using the NIST MainLib database by referring to the retention time, molecular formula, and molecular weight.
Metabolites 2023, 13, x FOR PEER REVIEW 8 of 22 acids, terpenes, and steroids.For S. cristaefolium, the dominant compound is ethyl isoallocholate, a steroid.In addition, several fatty acids were identified.In T. cylindrica, the main compounds are steroids.Other major compounds in the T. cylindrica lipophilic extracts were fatty acids.Meanwhile, for U. lactuca, the main compound is cholest-5-en-3ol,24-propylidene-, (3β)-, a steroid.Other fatty acids were also detected.Detailed data on the compounds detected in the lipophilic extracts of these macroalgae are presented in Table 2. Table 2. Identified compounds from Gas Chromatography-Mass Spectrometry (GC-MS) analysis of macroalgae lipophilic extract.Based on NIST MainLib identification, 16 lipophilic compounds were detected in the S. cristaefolium extract, 20 in the T. cylindrica extract, and 14 in the U. lactuca extract.Most of the lipophilic compounds detected in the macroalgae belong to the family of fatty acids, terpenes, and steroids.For S. cristaefolium, the dominant compound is ethyl iso-allocholate, a steroid.In addition, several fatty acids were identified.In T. cylindrica, the main compounds are steroids.Other major compounds in the T. cylindrica lipophilic extracts were fatty acids.Meanwhile, for U. lactuca, the main compound is cholest-5-en-3-ol,24-propylidene-, (3β)-, a steroid.Other fatty acids were also detected.Detailed data on the compounds detected in the lipophilic extracts of these macroalgae are presented in Table 2.

In Silico Molecular Docking Analysis
Natural products including marine metabolites offer potential treatments for a variety of human ailments.However, challenges in analyzing large numbers of compounds in the laboratory have hampered progress in the development of medicinal compounds based on these natural products.As a result, computational analysis has become an invaluable tool in drug development, especially when working with natural resources rich in metabolites.Virtual screening methods are often used to reduce the costs and time associated with developing natural product-based drugs.Specifically, more negative values of the total net charge indicate a higher binding affinity [68].
An in silico molecular docking technique was used in this study to screen the lipophilic compounds of macroalgae (S. cristaefolium, T. cylindrica, and U. lactuca), which were identified via an untargeted GC-MS metabolomics analysis.These compounds were evaluated to determine their potential ligand binding affinity toward established anti-DM and COVID-19 targets, specifically α-glucosidase, α-amylase, ACE2, and TMPRSS2.Compounds that have the ability to inhibit α-glucosidase and α-amylase activity could cause an inhibitor of sugar or carbohydrate absorption and digestion from the intestine, which then causes the reduction in post-prandial blood glucose spikes.Consequently, inhibitors of this enzyme may have therapeutic potential in the management of DM [69].In the context of COVID-19, compounds that inhibit ACE2 have the potential to prevent the virus infection.This is because the binding of the virus SARS-CoV-2 S-protein to the ACE2 receptor initiates viral entry, which then could cause viral replication and viral spread [70].In addition to inhibiting ACE2, other characteristics of a potential anti-COVID-19 agent is the ability to inhibit ACE2 s co-receptor known as TMPRSS2, which has the role of facilitating viral entry into host cells [71][72][73][74].
Our findings indicate that several lipophilic compounds from S. cristaefolium, T. cylindrica, and U. lactuca have the ability to bind and inhibit proteins associated with either DM or COVID-19 and even both (Figure 4).Notably, its binding affinity exceeds that of the reference compound (native ligand or positive control).For example, ethyl iso-allocholate, present in the lipophilic extracts of S. cristaefolium and T. cylindrica, showed effective binding and inhibition of ACE2 and TMPRSS2.It also showed effective binding toward α-glucosidase and α-amylase, but the activity was slightly lower than the positive controls.This indicates that the anti-hyperglycemic activity of the S. cristaefolium, as shown in the in vivo hyperglycemic study, could be due to other mechanisms, possibly through the reduction in ROS, as the S. cristaefolium had the highest anti-oxidant activity compared to the other macroalgae (Table 1).

Analysis of Physiochemical Properties and ADME Studies of the Compounds
Three lipophilic compounds were identified due to their strong binding affinity to DM and COVID-19-related protein targets, Compound 1, -2, and -3.To measure their potential as a drug, its physicochemical properties were evaluated based on Lipinski's rule of five.The viability of a compound as a drug increases if it meets these criteria and no more than one physicochemical criterium is violated [86].Compound 1 fully complies

Analysis of Physiochemical Properties and ADME Studies of the Compounds
Three lipophilic compounds were identified due to their strong binding affinity to DM and COVID-19-related protein targets, Compound 1, -2, and -3.To measure their potential as a drug, its physicochemical properties were evaluated based on Lipinski's rule of five.The viability of a compound as a drug increases if it meets these criteria and no more than one physicochemical criterium is violated [86].Compound 1 fully complies with Lipinski's rule, as it does not violate any criteria.All properties, including the molecular weight, hydrogen bond donor, hydrogen bond acceptor, log P value, and molar refractivity, were within the accepted range.Compound 2, although displaying a log P value exceeding 5, indicating high lipophilicity, can still be considered for further drug discovery and development because it has only one violation of Lipinski's rule of five.Meanwhile, based on this analysis, Compound 3 violates two criteria: the log P value was higher than 5 and the molar refractivity exceeded 130. Therefore, Compound 3 would need to be developed into a certain drug delivery system in order to have higher oral bioavailability and to be used as a drug.Detailed data regarding the physicochemical properties of these compounds are tabulated in Table 3.The oral bioavailability of a compound is an important criterion for drug likeness and can be predicted using a bioavailability radar [87].The radar, depicted in Figure 10A-C, evaluates the suitability of compounds for oral consumption.The optimal range for each property, such as the lipophilicity, size, polarity, solubility, saturation, and flexibility, is represented by the pink area.The results of the following compounds show that the Compound 1 properties were all within the optimal pink area.Meanwhile, Compound 2 and Compound 3 were partly in the pink area.All these compounds qualify as druglike.However, Compound 2 and -3 require extra consideration during drug formulation due to their high lipophilicity, which was also evidenced by their log P value, which exceeded 5, summarized in Table 3.Many currently emerging drug candidates exhibit high lipophilicity.For example, novel tuberculosis drugs that were lipophilic performed better than less lipophilic drugs [88,89].Therefore, despite their lipophilicity, Compound 2 and -3 showed potent activity against target proteins, thus underlining their potential as drug candidates.The intricate relationship between the structure of the lipophilic compounds and their binding affinity to DM and COVID-19 protein targets offers promising avenues in drug design.Moreover, variations in Lipinski's rule of five violations among compounds underscore the complexity of translating molecular properties into drug-like behavior.These findings emphasize the need for a multidimensional approach when evaluating potential drug candidates, including considering traditional criteria such as Lipinski's rule of five and innovative methods, such as the bioavailability radar.

Discussion
The main aim of this study was to explore the potency of macroalgae lipophilic compounds as drug candidates for COVID-19 patients with DM comorbidity treatment, by using several approaches.Our findings demonstrate the importance of a multifaceted approach to drug discovery.Exploration of these compounds has significance considering the ongoing challenges faced by the global healthcare community in addressing this health problem, including limitations in performing laboratory experiments, such as analyzing a wide range of plant metabolites for COVID-19 studies where the SARS-CoV-2 virus is highly infectious to humans.The methods we used allow a rigorous and detailed The pharmacokinetic profile of candidate compounds can be assessed using a boiled-egg diagram, as shown in Figure 10D.This diagram evaluates passive gastrointestinal absorption (HIA) and brain penetration (BBB).The white part indicates a greater possibility of passive absorption in the gastrointestinal tract, while the yellow part indicates a greater possibility of penetration into the brain.Red dots indicate that the compounds are not affected by P-glycoprotein-mediated extrusion from the CNS.These findings revealed that all compounds were not substrates for P-gp.Compound 1 is located in the yellow area, indicating the possibility of being passively absorbed by the gastrointestinal tract and it can penetrate the BBB.In contrast, Compound 2 and -3 were located in the gray area, indicating that they may not be efficiently absorbed by the gastrointestinal tract or penetrate the BBB.These results may originate from its structural complexity or increased lipophilicity.Addressing these challenges may require customized delivery systems to target specific regions.
The intricate relationship between the structure of the lipophilic compounds and their binding affinity to DM and COVID-19 protein targets offers promising avenues in drug design.Moreover, variations in Lipinski's rule of five violations among compounds underscore the complexity of translating molecular properties into drug-like behavior.These findings emphasize the need for a multidimensional approach when evaluating potential drug candidates, including considering traditional criteria such as Lipinski's rule of five and innovative methods, such as the bioavailability radar.

Discussion
The main aim of this study was to explore the potency of macroalgae lipophilic compounds as drug candidates for COVID-19 patients with DM comorbidity treatment, by using several approaches.Our findings demonstrate the importance of a multifaceted approach to drug discovery.Exploration of these compounds has significance considering the ongoing challenges faced by the global healthcare community in addressing this health problem, including limitations in performing laboratory experiments, such as analyzing a wide range of plant metabolites for COVID-19 studies where the SARS-CoV-2 virus is highly infectious to humans.The methods we used allow a rigorous and detailed assessment of the potential of these compounds as drug candidates.In this study, we integrated in vitro, in vivo, and in silico approaches to screen the large number of lipophilic metabolites in macroalgae in order to find the potential drug candidates.Moreover, using Lipinski's rule of five criteria provides a comprehensive framework for understanding the pharmacokinetics and pharmacodynamics of these compounds.The application of these methods provides advantages that cannot be easily obtained from traditional techniques.
The analysis of the lipophilic extracts of macroalgae S. cristaefolium, T. cylindrica, and U. lactuca has shown that these extracts have anti-oxidant and anti-hyperglycemic properties.Recent opinions have emerged regarding the efficiency of anti-oxidant consumption and treatment of diseases in the human body, as reviewed in [90].It is mentioned that anti-oxidants might not have a direct correlation with disease progression and treatment.However, based on other studies, it is shown that anti-oxidant consumption is an important factor in disease treatment, especially diseases that are a result of high ROS production, such as DM.Anti-oxidants play a significant role in the management of DM and could reduce hyperglycemia conditions, even though not directly.For example, dietary supplementations of anti-oxidants such as N-acetylcysteine, ascorbic acid, and α-lipoic acids are able to reduce DM complications [91].The consumption of anti-oxidants could also reduce key biochemical changes found in hyperglycemia caused by superoxide radicals, including increased flux through the polyol pathway (glucose conversion to sorbitol) and the formation of AGE products [92,93].Some other studies have also shown that individuals with low concentrations of anti-oxidants have a higher risk of DM complications [94][95][96].Therefore, it could be assumed that anti-oxidants still play an important role in DM treatment.
Further, in silico molecular docking assessments have shown that three compounds detected in the T. cylindrica lipophilic extract, identified as steroid derivatives, have the ability to inhibit DM and COVID-19 protein targets, α-amylase, α-glucosidase, ACE2, and TMRPSS2, and their binding affinity was higher than their respective positive controls.Although all three compounds are promising (Compound 1, Compound 2, and Compound 3), they present varying challenges and benefits.Compound 1 fully complies with Lipinski's rule of five; however, the binding affinity toward all the protein targets was not as significant as Compound 2 and Compound 3.Meanwhile, Compound 2 and Compound 3 were highly lipophilic.Therefore, this needs to be addressed by further optimization and development of the right drug delivery system to increase their oral bioavailability.
This research presents a step forward in searching for effective drug candidates against DM and COVID-19.By integrating established rules, advanced visualization tools, and rigorous analysis, we were able to navigate the complexities of marine-based drug discovery and development.The journey from a potential metabolite to a market-ready drug is still long and full of challenges.Further research including the isolation of these single compounds from the macroalgae lipophilic extract, evaluation of compound activity and toxicity through in vitro and in vivo assays, and optimization of the drug formulation are all critical steps needed to bring these promising candidate metabolites to the hand of COVID-19 with DM comorbidities patients.

Figure 1 .
Figure 1.Workflow of in silico molecular docking analysis.Figure 1. Workflow of in silico molecular docking analysis.

Figure 1 .
Figure 1.Workflow of in silico molecular docking analysis.Figure 1. Workflow of in silico molecular docking analysis.
a shows statistical differences (p < 0.05) between samples and standard, b shows statistical differences (p < 0.05) between samples and S. cristaefolium.

Table 2 .
Identified compounds from Gas Chromatography-Mass Spectrometry (GC-MS) analysis of macroalgae lipophilic extract.

Table 3 .
The physicochemical and pharmacokinetic properties of the lipophilic compounds detected in macroalga T. cylindrica lipophilic extracts with a high binding affinity toward DM and COVID-19related protein targets.