Studying Microwave-Assisted Hydrolytic Degradation of Colchicine using RP-Chromatographic Methods: In Silico ADME/Tox Profile and Molecular Docking

Colchicine, is a natural amide containing anti-gout treatment with versatile applications. Microwave assisted hydrolytic degradation is a newly alternative method thought to be more promising than traditional procedures of heating. It is an ecofriendly method that has more reproducible results due to the control of parameters. From this point, carrying on hydrolytic degradation of colchicine was tested for the first time under acidic conditions with the aided of microwave. The drug was hydrolyzed with the formation of deacylated analogue. Isolation of the resulted degradate was carried out using flash chromatography, the isolated one was elucidated based on 1 H NMR data. Moreover, the results of human pharmacokinetic predictions conducted from in silico data showed that colchicine had higher blood brain barrier (BBB), plasma protein binding, and oral absorption than its deacylated derivative. The study was also extended to forecast the binding of colchicine and its degradate to the target protein. Furthermore, two stability indicating chromatographic methods were developed for quantification of the drug and its degradation product with high sensitivity. The first method was RP-TLC densitometric method that based on using a solvent mixture of water: methanol: diethylamine (70: 30: 15, by volume). The second one was RP-HPLC at which a mixture of water (containing 0.02% diethyl amine): methanol: acetonitrile (50: 20: 30, by volume) was the used mobile phase. Validation parameters were calculated according to ICH recommendations and all were within the acceptable limits. These methods were used for determination of colchicine in its available tablets. They are the first developed stability indicating methods for analysis of colchicine and its degradation product. following the instructions of each method. Mean percentage recovery of each analyte was calculated and found to be 99.38 and 100.59% (for RP-TLC densitometric method) and 99.92 and 100.01% (for RP-HPLC method) for the drug and the degradate, respectively. All the results are near the acceptable value (100%) ensuring accuracy of the two

Despite the fact that colchicine is a terminal amide containing drug which is susceptible to hydrolytic degradation with the production of the deacylated colchicine, none of the published methods studied the hydrolytic degradation of this drug. Stability study forms an essential part of the registration of a new drug product or substance which ensures the compliance of the newly discovered drug with the international regulations. It is also imperative to reflect the effect of various environmental conditions (like temperature, humidity, and light) on the quality of the drug, to ensure its safety and efficacy regardless the place in the world it will be shipped, and to identify potential degradation products [27][28][29]. Microwave heating assisted degradation is a 'cold' in-situ process at which heating takes place only when the molecule or the solution of interest absorbs microwave energy. It is used as an alternative tool to the traditional forced degradation method which minimizes the waste chemicals and hence reduces the hazardous environmental effect. Similarly, it reduces dramatically the time required for generation of the degradation products without the production of intermediates, increases product yield, and enhances the purity of the resulted product. Recently, it is widely used to study the forced degradation of several pharmaceuticals [30][31][32].
Flash chromatography is a fast and economic approach substitution to the elderly used preparative column chromatography. It is widely used for isolation and purification of different molecules in highly pure form which is essential for performing microanalysis like MS and NMR analyses [33].
The pharmacokinetic profile of a drug molecule relies on ADME/Tox properties of a compound that deals with its absorption, distribution, metabolism, excretion, and toxicity throughout the human body. The vital goal of in silico ADME/Tox is to predict behavior of compounds in the whole body by collecting all kinetic processes in one global model. Prediction of molecule ADME/Tox in the early stages of drug development process are widely recognized. Fast evaluation of ADME/Tox properties will save both time and expense. Therefore, a large number of in-silico ADME/Tox soft wares have been recently used [34]. Molecular modeling, reveals the best fit orientation of a ligand that binds to a certain protein of interest. It affords useful information about the interaction and the binding orientation between the candidates and their protein targets [35].
Owing to the pharmaceutical importance of colchicine and lacking of methods published for studying its stability, the work in this manuscript aimed to study the hydrolytic degradation of colchicine using microwave aided heating through developing of two selective and sensitive chromatographic methods depending on using reversed stationary phases. Furthermore, isolation of the resulted degradate in highly pure form was done by using flash chromatography and then its structure was elucidated by 1 H-NMR analysis. Likewise, the in silco data was used for the first time to expect ADME/Tox profile of the drug as well as its degradation product which is very important for expecting the toxicity resulted from degradation of dosage forms containing colchicine.
Also, molecular docking was used to study the interaction of the drug and the degradate separately with the target protein. This work is the first developed stability indicating one that can be easily applied in QC laboratories. In order to prepare the solution of the tested tablets (200 µg/mL), the content of ten tablets were accurately weighed, grinded, and mixed well. A quantity equivalent to 2 mg of colchicine was accurately transferred to 10-mL volumetric flask and then dissolved in 7 mL methanol, ultrasonicated for about 10 min. The extracted tablets were then left to cool and the volume was finally adjusted with methanol. The prepared solution was then filtered and used to prepare different dilutions within the linearity range of each method either by using methanol (for RP-TLC densitometry) or by the mobile phase mixture (for RP-HPLC method).

1. Preparation of the degradation product
0.1 g of colchicine in 1 mL methanol and 2 mL of different concentrations of HCl (1N and 2N) were subjected to microwave radiation for 2 minutes. Methanol was distilled off under vacuum using rotaevaporator and the formed precipitate was then dissolved in 3 mL distilled water and then neutralized by aqueous NaHCO 3 till pH 7-7.5. The deacylated degradate was obtained by extraction with dichloromethane (2X 10 mL) and then further purification was carried out using flash chromatography.

2. Purification of the degradation product
The resulted degradation product was subjected to purification using Reveleris ® X2 Flash Chromatography system on C18 Column (50µm, 12g) RP-silica. Isolation was carried out in gradient elution manner starting with 10% methanol in water (containing 0.02% diethyl amine): till reaching 100% methanol (containing 0.02% diethyl amine). The collected fractions were combined together, evaporated under vacuum using rotavaporator and then tested for purity using RP-TLC. The isolated degradate was then subjected to IR and 1 HNMR analyses for structural elucidation.

3. Linearity and construction of calibration curves
For RP-TLC densitometric method, linearity was achieved in the concentration ranges of 0.2-4 and 0.1-4 µg/ band for colchicine and its degradate, respectively. Different concentrations of the drug and the degradate (10-400 µg/ mL) were separately prepared in 10 mL methanol from their respective stock solutions (1 mg/mL). Accurate 10 µL of each sample was applied in triplicates as band to RP-TLC aluminum plates (20 x 10 cm). The width of the band was adjusted to be 3 mm and the bands  Dilutions were done by the mobile phase mixture of water (containing 0.02% diethyl amine): methanol: acetonitrile (50: 20: 30, by volume). 40 µL of each flask was then injected in triplicates to HPLC system. The mobile phase used was pumped at 2mL/ min, the run time was adjusted at 4 min and UV scanning was performed to 254 nm. Peaks area of the eluted analytes was recorded and used for plotting its corresponding calibration graph and computing the regression equation.

4. Analysis of marketed samples
For RP-TLC densitometric method, accurate 7.5 µL was applied to RP-TLC aluminum plates from the previously prepared tablets solution (200 µg/ mL) (n= 6). Chromatographic conditions following during construction of calibration curves for RP-TLC densitometric method was then followed and the computed regression equation for colchicine was used to calculate its recovered conecentration.
For RP-HPLC method, a sample equivalent to 15 µg/ mL colchicine were prepared in 10 mL mobile phase mixture from sample solution previously prepared (200 µg/ mL) (n= 6). 40 µL of each dilution was injected to HPLC system and then instructions under linearity were followed. Concentrations of the drug were then obtained by substitution in its corresponding regression equation. Moreover, standard addition technique for each method was performed on three different levels for both methods.

Molecular docking
MOE software was used to protonate the crystal structure of tubulin [39]. The following parameters are adjusted to carry out the docking of colchicine and its degradate into the active site of tubulin. (1) The triangle matcher algorithm was selected to dock the targeted compounds to the active site; (2) London dG scoring was specified to rank the poses; (3) The Force field refinement was specified to release the poses, and (4) Affinity ΔG scoring was specified to classify the poses using the refinement step. Van der Waals, ionic, and hydrogenated interactions that were recognized between the protein and the ligand contribute in calculating the binding free energy score. After the docking process was completed, ligand interaction 2D & 3D diagrams for each compound was created based on the analysis of the binding sites

Results And Discussions
Regarding to the recommendations of International Conference on Harmonization (ICH) [40], it is a must to study the drug degradation profile as well as to characterize the produced degradate and its expected toxic effect. Inclusive searching in the published methods for determination of colchicine, no stability indicating assay method has been reported for studying the hydrolytic degradation of colchicine and the toxicity of its hydrolytic degradation product which is essential to minimize the risk of toxicity due to degradation of drug substance or product [41,42]. Additionally, performing the hydrolytic degradation under microwave controlled heating has advantages of significantly reducing time of degradation from several hours to few minutes, giving reproducible results, and almost pure degradation product. In this manuscript, the hydrolytic degradation of colchicine was studied for the first time by using two concentrations of HCl (1 and 2 N) under microwave controlled heating. The work was then protracted to isolate the resulted degradate in highly pure by using flash chromatography. Likewise, molecular modeling of the resulted degradate on the target protein was performed. The drug and its degradate were also quantified with high sensitivity even though their structural similarity through developing of two new chromatographic methods, RP-TLC densitometric and RP-HPLC methods. 3.

Degradation behavior of colchicine
It was observed that the drug was sensitive for hydrolytic degradation and its degradation rate affected by concentration of HCl used. Upon degradation with 1N HCl for two minutes, 65% degradation was observed while 85 % degradation was resulted on using 2N HCl for the same time.
After degradation, the produced degradate was isolated by extraction of the neutralized solution with methylene chloride (3 X 10 mL) and then purified with flash chromatography by using mixture of protons. FT-IR spectra revealed the disappearance of amidic carbonyl group of colchicine with the presence of tropolone carbonyl group at 1769 cm -1 and forked signal of NH 2 group around 3299-3276 cm -1 . The pathways adopted for the degradation of colchicine is depicted in Fig. 1 3.

1. RP-TLC densitometric method
Trials for TLC densitometric method began with using normal phase TLC plates using different developing systems, unfortunately none of the tested systems succeeded in elution of the produced degradate with symmetric non tailed peak. After that, RP-TLC plates were tested using different solvent mixtures such as water: methanol, water: acetone, methanol: ethylacetate, and water: acetonitrile. All these systems were tried in different ratios but the most acceptable separation between the drug and the degradate was resulted on using a mixture of water: methanol (70: 30, v/v) but with highly tailed peak for the degradate. Using methanol ˃ 30%, resulted in decreasing the resolution between the two components but decreasing its ratio increased the time required for chromatographic development dramatically. In a trial to enhance the shape and the symmetry of the peak of the degradate, different ratios of acetic acid (1-5%, v/v) and diethyl amine (1-20%, v/v) were separately added to the solvent mixture. Diethyl amine was reported to act as silanol blocker as well as ion pair reagent [43], hence it was found that addition of 15% diethyl amine was necessary to elute the degradate from the stationery phase with nearly symmetric non tailed peak and hence the optimum chromatographic efficiency. The

RP-HPLC method
Depending on the previous trials of the developed RP-TLC densitometric method, diethyl amine was necessary to elute the degradation product from reversed stationary phase with acceptable peak  , Fig. 3.

Validation of the developed methods
Recommendations reported in ICH [44] guidelines for method validation was followed.
Linearity of the proposed RP-TLC densitometric method was tested and was verified in the ranges of 0.2-4 and 0.1-4 µg/ band for colchicine and the degradate, respectively using polynomial regression which resulted in lower LOD and LOQ values than linear regression. On the other hand, linearity of RP-HPLC method was ascertained in the range of 2-40 µg/ mL for both analytes using linear regression (for colchicine) and polynomial regression (for the degradate). Parameters of the computed regression equations are given in Table 1, values of the resulted correlation coefficients revealed that the developed methods are linear within the tested ranges.
Besides, accuracy of these methods was tested by applying them to analyze different pure samples (n= 9) each of the drug and its degradate following the instructions of each method. Mean percentage recovery of each analyte was calculated and found to be 99.38 and 100.59% (for RP-TLC densitometric method) and 99.92 and 100.01% (for RP-HPLC method) for the drug and the degradate, respectively. All the results are near the acceptable value (100%) ensuring accuracy of the two methods. Likewise, intraday and interday precision, represented as SD, were tested by analysis of different three concentrations of each compound either on the same day (for intraday precision) or on three consecutive days (for interday precision). Low SD values (˂ 2%) were resulted confirming that the developed methods are precise, Table 1. Specificity of the methods was assured by the good separation of the parent drug and its deacylated analogue Figs. 2 and 3. Additionally, no interference from tablets additives were found on analysis of the available tablets dosage form which was proved from the results given in Table 2. Limits of detection (LOD) and quantitation (LOQ) were calculated using the slope method (LOD= 3.3*SD/ slope, LOQ= 10*SD/ slope) where low values were resulted, Table 1 confirming that the methods possess the sensitivity required for detection of low concentrations of the resulted degradate. In addition, robustness of the method determined the ability of the method to resist small intended changes in parameters and it was expressed as %RSD.
Robustness was checked by changing amount of diethyl amine (± 2%), methanol (± 1%) [for RP-TLC densitometric method], mobile phase flow rate (± 0.05 mL/ min), and detection wavelength (± 2 nm) [for RP-HPLC method] and measuring the effect of these changes on values of R f (for RP-TLC densitometric method) or t R (for RP-HPLC method). All the obtained %RSD values were within the acceptable limits ensuring that the methods did not affected by the studied changes, Table 1. Finally, system suitability testing parameters such as peak symmetry, capacity, selectivity, and resolution factors as well as number of theoretical plates and height equivalent to theoretical plates were calculated. All the calculated parameters ensured that the methods have good specificity and efficiency, Table 3.
After validation of the proposed chromatographic methods, they were successfully used to measure concentrations of colchicine in Colchicine ® tablets and the recovered concentration of the active drug (105.38 and 107.16 % for RP-TLC densitometric and RP-HPLC, respectively) ensured the validity of the method, Table 2. Standard addition technique was also done and the results obtained indicated that there was no interference from tablets excipients confirming accuracy of these methods, Table 2.
As well, statistical comparison between the results of the developed methods and the reference one [17] was carried out by using student's t and F test for variance and the results showed that the difference between the two methods and the reference one is not significant, Table 1.

In silico molecular and ADME properties
In most drug discovery projects sorting out the ADME/ Tox issues is the most challenging part of the project as pharmacokinetic (ADME) and pharmacodynamic (e.g., toxicological) properties (druglikeness) are of great importance during pre-clinical evaluation to optimize a lead compound into a successful drug candidate, and to minimize attrition rates during clinical trials. Some active compounds pass through Blood-brain barrier (BBB), thus it is important in pharmaceutical sphere to study it. Forecasting human intestinal absorption (HIA %) of drugs is considered the key physicochemical parameter for identifying potential drug candidate [45].  Table   4. From the results, colchicine showed relatively higher BBB value than its deacylated analogue and this is contributed to the hydrophilic nature of the amino group of the degradate. Additionally, colchicine showed higher plasma protein binding, oral drug absorption, G protein-coupled receptors ligand, ion channel modulation, kinase inhibition, and nuclear receptor ligand.
A drug-like molecule (DLM) possesses the physicochemical properties that might enable it to become a drug. Generally speaking, an orally available molecule that satisfies Lipinski's rule (Lipinskicompliant) and shows a balance between lipophilicity and hydrophilicity would qualify it to be a druglike molecule [47].
An orally active compound should obey both Lipinski's and Veber's rules. As shown from Table 5, both compounds followed both Lipinski's and Veber's conditions.
Toxicity screening results of preADMET for colchicine and its deacylated product showed mutagenicity against AMES mutagenicity with positive TA100_NA stain and negative for the other strains used. Moreover, both compounds showed no potential rodent carcinogenicity except against rat. Also, hERG_inhibition is of low risk for both the drug and its degradate. Yet the degradate has a slight increase in the toxicity value against algae_at, daphnia_at, medaka_at & minnow_at, Table 6.
The results of in silico screening highlights the necessity to conduct lead optimization for colchicine containing pharmaceutical dosage forms to minimize its degradation to more toxic analogue.

Molecular docking
In this present work, the automated molecular docking was used to determine the orientation of colchicine and its deacylated analogue bound in the active site of the tubulin (PDB: 4LZR). In order to achieve different binding conformations, the structures of both the drug and its degradate as well as the enzyme were kept flexible and the best-docked complex obtained was recognized and analyzed in detail. Validation of molecular docking was carried out by extracting the native specific ligand of tubulin from the binding site and then redocked the conformation of the ligand to the binding site to guarantee the reliability and reproducibility of the docking procedure.
The docking of target molecules with tubulin reveals that they were exhibiting the bonding with two amino acids in the active site which is showed in Fig. 4. From the resulting docking procedures, it is clear that the compounds attached to the active binding site of the target enzyme. Colchicine showed high energy docking binding score (-13.87 kcal/mol) with hydrophobic binding interaction with Ile146 amino acid, Fig. 4A. Additionally, the deacylated analogue showed relatively the same docking score (-13.86 kcal/ mol) linked to the active site of tubulin through hydrogen bond interaction between amino group and ASN140, Fig. 4B.

Conclusion
Forced degradation study of colchicine has been conducted with assistance of microwave which resulted in the formation of a product that was not reported before. The produced degradate was identified as deacylated colchicine by 1 HNMR and IR spectra. Quantitation of the drug and its deacylated analogue was achieved through developing of two chromatographic methods, RP-TLC densitometric and RP-HPLC methods. Analysis time required for the two methods was short and thus minimum amounts of hazardous solvents was needed. The study expanded knowledge space concerning ADME/Tox profile of the drug and the formed degradate. Additionally, this work demonstrated that the resulted degradate possess lower penetration to BBB and slightly higher algae_at, daphnia_at, medaka_at, and minnow_at values than the parent drug. Also, docking study revealed the ability of the degradate to bind in the same active site of the target tubulin.

Compliance with ethical standards
Disclosure of potential conflicts of interest: Authors declare that they have no conflict of interest.
Research involving human participants and/ or animals: no human volunteers or animals were used in this work.     Table 4. The values of ADME properties of the studied compounds obtained using in-silico method.   Figure 1 Schematic diagram for the hydrolytic degradation pathway of colchicine.

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download.