In silico Pharmacokinetic and Toxicity Prediction of Compounds from Andrographis Paniculata ( Burm.F .) Nees.

: Many compounds have been isolated from Andrographis paniculata (Burm. f.) Nees (AP). In drug discovery and development, plant secondary metabolites are popular as resources for drug candidates. A high-quality drug candidate should not only be effective against the therapeutic target, but it should also be safe and have good pharmacokinetic features. This study aimed to predict the pharmacokinetic features and toxicity potencies of 46 compounds from AP using the pKCSM online tool. According to pKCSM prediction, among the forty-six compounds from AP, compound 1 (14-Deoxy-11,12-didehydroandrographolide), compound 2 (14-Deoxyandrographolide), and compound 39 ((-)-beta-Sitosterol) have good pharmacokinetic features and do not have potencies to be mutagenic and hepatotoxic agents. The lethal dosage values (LD50) of compounds 1, 2, and 39, are 1935, 2053, and 2424 (mol/kg), respectively. However, further research is still needed to confirm these predictions.

For decades, plant secondary metabolites and their structural analogs have remained popular drug candidates in drug discovery and development [12,13].However, despite the fact that some compounds have been shown to have certain bioactivities, they cannot be developed and have failed in clinical trials due to pharmacokinetic issues.Thus, pharmacokinetic screening is needed [14,15].The evaluation of the pharmacokinetic properties, including absorption, distribution, metabolism, and excretion (ADME) and the toxicity potencies (T) of compounds can be done using in vitro and in vivo methods, however, these experiments are expensive, especially when testing a large number of compounds.Many in silico models are being developed to predict the ADMET features of compounds.This approach has decreased experimental drug trials and increased success rates, making it a helpful method [16].
In this study, AP compounds are collected from the Knapsack, a complete database of plants and their metabolites [17].The pharmacokinetic properties and the toxicity potencies of AP compounds were evaluated using pKCSM online tool.pKCSM is a free web server that uses graphbased signatures to create predictive models of ADMET features [18].The computational prediction limitations of this study require future investigation to verify the results.

MATERIALS AND METHODS
The compounds of AP were collected from the Knapsack (http://www.knapsackfamily.com/)[17].The SMILES (simplified molecular input line entry system) strings of each compound were collected from PubChem (https://pubchem.ncbi.nlm.nih.gov/) and were then introduced into pKCSM (http://biosig.unimelb.edu.au) to evaluate the physicochemical properties, pharmacokinetic characteristics, and toxicological potencies of the AP compound.Both Knapsack and pKCSM were accessed in March 2022.

Collection of AP compounds
Forty-six compounds of AP were downloaded from the Knapsack database (see Table 1), including twenty-two compounds from the diterpene lactone group, sixteen flavonoids, a compound from the sterol group, four phenolic acids, and three compounds from the sesquiterpene class.The pKCSM online tool [18] was then used to predict the physicochemical properties, pharmacokinetic characteristics, and toxicity potencies of compounds.

Pharmacokinetics characteristics and toxicity potencies of AP compounds
When a drug is taken orally, it travels from the stomach to the small intestine, where most of it is absorbed [19].To evaluate drug absorption, pKCSM predicts the percentage of the absorbed compound in human intestinal (% HIA) and intestinal mucosa permeability (Caco-2 permeability).
In the pkCSM predictive model, a compound with an absorption value >80% is well-absorbed, while <30% is poorly-absorbed.Moreover, a compound is predicted to have high intestinal mucosa permeability if it is predicted to have Caco-2 permeability values >0.90 [18].According to pKCSM predictive models (see The volume of distribution (VDss) is an important indicator for estimating the proportion of a drug's total amount in the body versus its plasma concentration at a given time [20].pKCSM developed the human VDss predictive model.The low VDss if log VDss < -0.15 and the high VDss if log VDss > 0.45 [18].Drugs can also be distributed to the brain.However, the blood-brain barrier (BBB) prevents drugs from entering the brain.To predict whether a drug would cross the BBB and cause effects on the central nervous system, pKCSM provides BBB and CNS permeability predictions.
A compound with logBB >0.3 is assumed to penetrate the BBB easily, whereas a compound with logBB < -1 is assumed to be poorly distributed to the brain.Moreover, a compound with logPS >-2 is considered to penetrate the central nervous system (CNS), while those with logPS <-3 are considered unable to penetrate the CNS [18].According to the pKCSM prediction (see have logPS value (D3) >-3, which means that they can penetrate the CNS.There have been safety concerns in the development of drugs that can easily cross the BBB and penetrate the CNS due to unforeseen neurotoxicity [21].
In the liver, Cytochrome P450 (CYP) deactivates some drugs, and it can also activate several drugs [22].A CYP inhibitor is a molecule that inhibits the detoxification activity of CYP.The CYP inhibition activity of the molecule likely mediated many drug interactions.Therefore, it becomes essential to assess the CYP substrates and inhibitors of drug candidates.pKCSM provides predictive models of five CYP isoforms that are responsible for drug metabolisms, such as CYP1A2, CYP2C19, CYP2C9, CYP2D6, and CYP3A4 [18].In metabolism features, compounds 6, 23-25, 27, and 29-30, were predicted as CYP1A2 inhibitors (M1).Compounds 23-30 were predicted as CYP2C19 inhibitors (M2).
Organic cation transporter 2 (OCT2) is a transporter in the human kidney that controls drug reuptake from the blood.It plays a key role in the disposition and renal clearance of drugs [23].Thus, assessing a potential molecule candidate to be re-uptaken by OCT2 (OCT2 substrates) provides useful information regarding not only its clearance (excretion) but also potential contraindications [18].
Toxicity evaluation is important to assure that drug candidates are safe.pKCSM also predicted the potential mutagenicity (based on AMES toxicity) and hepatotoxicity of AP compounds.
In addition, the lethal dosage value (LD50) is a standard measurement to assess the acute toxicity of compounds.LD50 is the amount of a compound given all at once that causes the death of 50% of a group of test animals [18].As a result, compounds 3, 28, and 44-46 were predicted to have the potential to be mutagenic agents (T1).Compounds 11, 17, and 43 were predicted to have potencies to be hepatotoxic agents (T2).The predicted lethal dosage values (LD50) of 46 AP compounds range from 241 to 3204 (mol/kg).
Based on the results above, compounds 1, 2, and 39 were predicted to have good absorption and distribution features.None of these compounds were predicted to be CYP1A2 inhibitors, CYP2C19 inhibitors, CYP2C9 inhibitors, CYP2D6 inhibitors, CYP3A4 inhibitors, or OCT2 substrates.
The therapeutic effects of AP are attributed to four major active diterpenoids, including andrographolide, neoandrographolide, 14-deoxy-11,12-didehydroandrographolide, and 14deoxyandrographolide.The highest content of 14-deoxyandrographolide was found in leaves at the transfer stage (between the seedling and vegetative stages).Meanwhile, 14-deoxy-11,12didehydroandrographolide was at its highest level during the vegetative stage [24].Compound (-)beta-sitosterol is a phytosterol that is widely distributed in the plant kingdom and possesses many bioactivities [25].In 2011, Xu et al. isolated (-)-beta-sitosterol along with 27 other compounds from the roots of AP [26].

CONCLUSION
This study predicted the pharmacokinetic properties and toxicity of 46 AP compounds, including compounds from the diterpene lactone group, flavonoid group, phenolic group, and sterol group, using the pKCSM online server.According to the pKCSM prediction, compounds 1 (14-Deoxy-11,12-didehydroandrographolide), 2 (14-Deoxyandrographolide), and 39 ((-)-beta-Sitosterol) have good pharmacokinetic features, and non-toxic.This computational method is a helpful approach to testing a large number of compounds.However, further research is still needed to confirm these predictions.