Metabolism and Chemical Degradation of New Antidiabetic Drugs (Part II): A Review of Analytical Approaches for Analysis of Gliptins

This paper is part II of the review on metabolism and chemical degradation of new antidiabetic drugs from glutides, gliflozins and gliptins. It is well known that metabolism data can be helpful for deriving safe levels of degradation impurities and their qualifying as far as toxicological aspects are concerned. As a result, it could link the quality of respective pharmaceutical products to clinical practice and patients. Some overlapping pathways of transformations of these important drugs of different chemical structures and different mechanisms of action were discussed. At the same time, the paper summarized interesting analytical tools for conducting modern drug metabolism as well as drug degradation experiments. The methods described here include liquid chromatography (LC) and liquid chromatography coupled with mass spectrometry (LC-MS or LC-MS/MS), which are widely used for detection and quantitative measurements of the drugs, their metabolites and degradants, as well as radiometric methods that are suitable for pharmacokinetic experiments. Special attention was paid to dedicated types of packing in chromatographic columns, as well as to special solutions in the LC-MS procedures. The present part addresses the analytical approaches elaborated for examining the metabolism and degradation pathways of gliptins that are dipeptidyl peptidase 4 (DPP-4) inhibitors.


Introduction Drug Metabolism and Drug Degradation Overlapping
Information on absorption, metabolism and excretion of drugs is necessary to support the studies on their pharmacokinetics and potential drug-drug interactions. Moreover, the knowledge on drug metabolism is one of the crucial factors used to assess their pharmacokinetic profile in patients with some dysfunctions. It is especially important in diabetic patients with higher incidence of chronic liver and kidney problems [1]. All of this creates the need to elaborate modern analytical methods suitable for determination of drugs and their metabolites, especially the drugs used for the treatment of diabetic patients [1,2]. On the other hand, drug stability studies are carried out to select the storage conditions and estimate the shelf life of pharmaceuticals. In addition, such studies allowed us to elaborate on the stability indicating methods to determine the drugs in the presence of their degradation products. Finally, such degradation studies are essential to inspect the changes in active pharmaceutical ingredients (APIs) and different drug products over the storage period and to assess the effect of environmental factors like pH, light and oxidative conditions on their chemical stability. Stability experiments in the stressed conditions aim to generate the maximum of degradation products in a short time in comparison with general stability studies (i.e., long term and accelerated stability studies). It is well known that some degradants may possess different biological properties than parent drugs, including toxic and immunity properties. Therefore, it is obligatory to estimate their levels in APIs or the marketed formulations, to know their structures, the way they are formed and finally, their potential amounts or concentrations [3].
The drug metabolism and drug degradation pathways may overlap, resulting in the formation of identical constituents. Therefore, the metabolism data can be helpful for deriving safe levels of degradation impurities and developing patient-centric therapy. Qualification of the degradation products or related substances is the process of acquiring and evaluating data to establish the biological safety of an individual degradation product, impurity or metabolite at the specified levels [4]. For all these purposes, information on metabolism of drugs from animal studies could be used to derive the no observed adverse effect level (NOAEL). Then, it could be converted to a human equivalent that was determined taking into account the safety and efficacy for patients. Additionally, such knowledge could be used to better understand the pharmaceutical product behavior on the market and more strongly connect the quality of this pharmaceutical product to clinical practice [5].
Generally, both drug metabolism in the body and drug degradation during processing and/or storage can undergo similar chemical reactions. Consequently, many impurities generated during degradation are also metabolites. It could be the important question when the process of qualification of impurities is concerned. It is obvious that simple comparison of an impurity with a detected metabolite in humans or animals is not adequate to qualify an impurity because the detected compound could be a metabolite or an impurity. Therefore, it is obligatory to establish that an exposure is equal to or greater than the level that might result from the real exposure to a particular impurity, following administration of some individual substance or product [5]. Thus, the ICH documents on respective thresholds at which impurities in new drug substances and products must be reported, identified and qualified must be taken into account [4].
Bearing in mind possible chemical connections between drug metabolism in the body and drug degradation under different environmental conditions, the goal of the present study was to collect the most interesting papers concerning three groups of important antidiabetics, i.e., gliptins (dipeptidyl peptidase 4 (DPP-4) inhibitors), glutides (glucagonlike peptide 1 (GLP-1) receptor agonists, GLP-1 RAs) and gliflozins (sodium glucose co-transporter 2 inhibitors, SGLT2 inhibitors), and to compare their metabolites and degradation products. We intended to find as many of the above connections as possible, as well as similar products of both metabolism and degradation, to express the idea of projecting the studies connecting these two essential aspects of antidiabetic therapy. The second goal of the present paper was to collect the most interesting analytical tools used to analyze these important drugs in terms of their metabolism and chemical degradation. Between them, liquid chromatography (LC) and liquid chromatography coupled with mass spectrometry (LC-MS), as well as high-resolution mass spectrometry (HRMS) were emphasized.
In Part I of our review (sent for publication in Biomedicines) we presented the data from the literature on metabolism, degradation and respective analytical tools for antidiabetic drugs from glutides (GLP-1 receptor agonists) and gliflozins (SGLT2 inhibitors) while the present Part II is devoted to gliptins (DPP-4 inhibitors).

Gliptins (DPP-4 Inhibitors)
Gliptins have been developed to prevent degradation of endogenously released incretins, glucose-dependent insulinotropic polypeptide (GIP) and glucagon-like peptide 1 (GLP-1), by DPP-4 enzyme. Thus, they act as DPP-4 inhibitors and prolong respective actions of these endogenous incretins. They do not pass the blood-brain barrier, have no direct effect on satiety, and in contrast with GLP-1 receptor agonists (GLP-1 RAs), did not alter gastric emptying. However, when compared with GLP-1 RAs, the DPP-4 inhibitors offer several advantages such as oral administration, absence of gastrointestinal adverse effects and lower costs. What is more, some clinical data suggest that gliptins could exert positive cardiovascular effects. Because of their safety profile, especially their very low risk of hypoglycemia, gliptins could be indicated in elderly patients [6,7].
Five of these DPP-4 inhibitors, i.e., anagliptin ( [8]. Gliptins can be classified into peptidomimetic (i.e., ANA, OMA, SAXA, SITA, TENE and VILDA) and non-peptidomimetic (i.e., ALO, EVO, LINA and TRELA) subtypes. Due to their specificity to the substrate site of the enzyme, some of them have substituted pyrrolidines or thiazolidines as a proline mimetic moiety. Although their chemical structures differ from each other, all DPP-4 inhibitors are substrate-competitive active site binders and have common interactions with the key residues of the target protein [7]. Based on the half-life and time of dissociation from the DPP-4 enzyme, they are prescribed twice a day (e.g., ANA and VILDA), once a day (e.g., ALO, EVO, LINA, SITA, SAXA) or once a week (e.g., OMA, TRELA) [6]. The structures of the mentioned DPP-4 inhibitors are presented in Figure 1.
iomedicines 2023, 11, x. https://doi.org/10.3390/xxxxx www.mdpi.com/journal/biomedicines direct effect on satiety, and in contrast with GLP-1 receptor agonists (GLP-1 RAs), did not alter gastric emptying. However, when compared with GLP-1 RAs, the DPP-4 inhibitors offer several advantages such as oral administration, absence of gastrointestinal adverse effects and lower costs. What is more, some clinical data suggest that gliptins could exert positive cardiovascular effects. Because of their safety profile, especially their very low risk of hypoglycemia, gliptins could be indicated in elderly patients [6,7].  [8].
Gliptins can be classified into peptidomimetic (i.e., ANA, OMA, SAXA, SITA, TENE and VILDA) and non-peptidomimetic (i.e., ALO, EVO, LINA and TRELA) subtypes. Due to their specificity to the substrate site of the enzyme, some of them have substituted pyrrolidines or thiazolidines as a proline mimetic moiety. Although their chemical structures differ from each other, all DPP-4 inhibitors are substrate-competitive active site binders and have common interactions with the key residues of the target protein [7]. Based on the half-life and time of dissociation from the DPP-4 enzyme, they are prescribed twice a day (e.g., ANA and VILDA), once a day (e.g., ALO, EVO, LINA, SITA, SAXA) or once a week (e.g., OMA, TRELA) [6]. The structures of the mentioned DPP-4 inhibitors are presented in Figure 1.

Metabolism of Gliptins
Alogliptin (ALO) is a highly potent and selective inhibitor of DPP-4 that was developed using the technology of structure-based drug design (SbDD). In the paper from the literature [9], it was shown that ca. 10% of ALO is metabolized, while 60-70% of the dose is excreted as unchanged drug in the urine. Two minor metabolites were detected following oral administration of [14C]-ALO, i.e., N-demethylated ALO (ALO-M1) (<1% of the dose) ( Table 1) that inhibits DPP-4 similar to the parent molecule, and inactive N-acetylated ALO (<6% of the dose). It was also speculated that ALO is mainly metabolized by CYP2D6 and CYP3A4.
Because anagliptin (ANA) is not frequently used in other countries than Japan, only a small number of reports exists investigating its biological properties and pharmacokinetics. The major metabolic pathway of ANA was proposed as the cyano group hydrolysis to generate the carboxylic acid metabolite ANA-M1 (Table 2), which accounted for 29.2% of the dose. The parent ANA and ANA-M1 were eliminated mainly with urine where the mechanisms of the active transport were probably involved [10].
More information on the possible metabolic pathways of evogliptin (EVO) in humans was reported. Metabolism of EVO was proposed as the phase I reactions where the drug is metabolized to 4-oxo-EVO (EVO-M1), 4(S)-hydroxy-EVO (EVO-M2) and 4(R)-hydroxy-EVO (EVO-M3) (Table 1), and as the II phase reactions forming 4(S)-hydroxy-EVO glucuronides and EVO-N-sulfates. At the same time, formation of EVO-M2 and EVO-M3 was inhibited by the CYP3A4 antibody, suggesting that CYP3A4 played a major role in the metabolism of EVO. It was also shown that EVO-M2 could be further metabolized to respective glucuronides by UDP-glucuronosyltransferases UGT2B4 and UGT2B7 [11].
The above data on EVO were confirmed in the next study from the literature [12]. The metabolism of EVO was also shown by the phase I reactions, i.e., hydroxylation and oxidation, as well as by opening of piperazine ring and oxidation, forming EVO-M4, EVO-  Alogliptin (ALO) is a highly potent and selective inhibitor of DPP-4 that was developed using the technology of structure-based drug design (SbDD). In the paper from the literature [9], it was shown that ca. 10% of ALO is metabolized, while 60-70% of the dose is excreted as unchanged drug in the urine. Two minor metabolites were detected following oral administration of [14C]-ALO, i.e., N-demethylated ALO (ALO-M1) (<1% of the dose) ( Table 1) that inhibits DPP-4 similar to the parent molecule, and inactive N-acetylated ALO (<6% of the dose). It was also speculated that ALO is mainly metabolized by CYP2D6 and CYP3A4.
Because anagliptin (ANA) is not frequently used in other countries than Japan, only a small number of reports exists investigating its biological properties and pharmacokinetics. The major metabolic pathway of ANA was proposed as the cyano group hydrolysis to generate the carboxylic acid metabolite ANA-M1 (Table 2), which accounted for 29.2% of the dose. The parent ANA and ANA-M1 were eliminated mainly with urine where the mechanisms of the active transport were probably involved [10].
More information on the possible metabolic pathways of evogliptin (EVO) in humans was reported. Metabolism of EVO was proposed as the phase I reactions where the drug is metabolized to 4-oxo-EVO (EVO-M1), 4(S)-hydroxy-EVO (EVO-M2) and 4(R)hydroxy-EVO (EVO-M3) (Table 1), and as the II phase reactions forming 4(S)-hydroxy-EVO glucuronides and EVO-N-sulfates. At the same time, formation of EVO-M2 and EVO-M3 was inhibited by the CYP3A4 antibody, suggesting that CYP3A4 played a major role in the metabolism of EVO. It was also shown that EVO-M2 could be further metabolized to respective glucuronides by UDP-glucuronosyltransferases UGT2B4 and UGT2B7 [11].
The above data on EVO were confirmed in the next study from the literature [12]. The metabolism of EVO was also shown by the phase I reactions, i.e., hydroxylation and oxidation, as well as by opening of piperazine ring and oxidation, forming EVO-M4, EVO-M5 and EVO-M6 (Table 1). In addition, the phase II reactions were shown to be involved, i.e., glucuronidation, sulfation, and conjugation with glycine-cysteine moieties.
The chemical structure of linagliptin (LINA) is based on xanthine moiety that distinguishes this drug from other gliptins. It may offer some differences in pharmacokinetic and pharmacodynamic properties of LINA, e.g., its low dissociation from the enzyme and greater potency than other gliptins. It is known that LINA is predominantly eliminated unchanged after both oral and intravenous administration [7]. However, the inactive metabolite LINA-M1 was identified in plasma as a major metabolite after oral administration. A two-step mechanism was proposed for its formation, i.e., CYP3A4-dependent conversion of the secondary amine of the parent drug to the corresponding ketone via oxidative deamination, followed by reduction. In excreta, the main metabolite after oral and intravenous administration was LINA-M2 formed by hydroxylation of the methyl group of the butinyl side chain. Next, minor metabolites could be formed by combinations of the following reactions: oxidation of the butinyl side chain and the piperidine moiety followed by oxidative degradation of the piperidine moiety, and finally N-acetylation and glucuronidation. The oxidation of the methyl group at position 4 of the quinazoline moiety was also proposed, and resulted in the corresponding carboxylic acid metabolite LINA-M3 (Table 1). A cysteine adduct and its sulfate conjugate were additionally observed in urine after intravenous administration of LINA [13].
In the study of Xu et al. [14], absorption, metabolism and excretion of omarigliptin (OMA) were evaluated in healthy male subjects after a single oral dose of 25 mg of [14C]-OMA. Radioactivity levels in plasma and excreta were determined via accelerator mass spectrometry (AMS). As a result, minimal metabolism of OMA was observed, as indicated by the fact that the parent drug accounted for ca. 89% of the radioactivity in urine. However, some oxidative metabolites were detected in plasma, with each comprising less than 3% of the total radioactivity.  (Table 1). In addition, the phase II reactions were shown to be involved, i.e., glucuronidation, sulfation, and conjugation with glycine-cysteine moieties. The chemical structure of linagliptin (LINA) is based on xanthine moiety that distinguishes this drug from other gliptins. It may offer some differences in pharmacokinetic and pharmacodynamic properties of LINA, e.g., its low dissociation from the enzyme and greater potency than other gliptins. It is known that LINA is predominantly eliminated unchanged after both oral and intravenous administration [7]. However, the inactive metabolite LINA-M1 was identified in plasma as a major metabolite after oral administration. A two-step mechanism was proposed for its formation, i.e., CYP3A4-dependent conversion of the secondary amine of the parent drug to the corresponding ketone via oxidative deamination, followed by reduction. In excreta, the main metabolite after oral and intravenous administration was LINA-M2 formed by hydroxylation of the methyl group of the butinyl side chain. Next, minor metabolites could be formed by combinations of the following reactions: oxidation of the butinyl side chain and the piperidine moiety followed by oxidative degradation of the piperidine moiety, and finally N-acetylation and glucuronidation. The oxidation of the methyl group at position 4 of the quinazoline moiety was also proposed, and resulted in the corresponding carboxylic acid metabolite LINA-M3 (Table 1). A cysteine adduct and its sulfate conjugate were additionally observed in urine after intravenous administration of LINA [13].
In the study of Xu et al. [14], absorption, metabolism and excretion of omarigliptin (OMA) were evaluated in healthy male subjects after a single oral dose of 25 mg of [14C]-OMA. Radioactivity levels in plasma and excreta were determined via accelerator mass spectrometry (AMS). As a result, minimal metabolism of OMA was observed, as indicated by the fact that the parent drug accounted for ca. 89% of the radioactivity in urine. However, some oxidative metabolites were detected in plasma, with each comprising less than 3% of the total radioactivity.  n.a. [9] ANA-M1  (Table 1). In addition, the phase II reactions were shown to be involved, i.e., glucuronidation, sulfation, and conjugation with glycine-cysteine moieties. The chemical structure of linagliptin (LINA) is based on xanthine moiety that distinguishes this drug from other gliptins. It may offer some differences in pharmacokinetic and pharmacodynamic properties of LINA, e.g., its low dissociation from the enzyme and greater potency than other gliptins. It is known that LINA is predominantly eliminated unchanged after both oral and intravenous administration [7]. However, the inactive metabolite LINA-M1 was identified in plasma as a major metabolite after oral administration. A two-step mechanism was proposed for its formation, i.e., CYP3A4-dependent conversion of the secondary amine of the parent drug to the corresponding ketone via oxidative deamination, followed by reduction. In excreta, the main metabolite after oral and intravenous administration was LINA-M2 formed by hydroxylation of the methyl group of the butinyl side chain. Next, minor metabolites could be formed by combinations of the following reactions: oxidation of the butinyl side chain and the piperidine moiety followed by oxidative degradation of the piperidine moiety, and finally N-acetylation and glucuronidation. The oxidation of the methyl group at position 4 of the quinazoline moiety was also proposed, and resulted in the corresponding carboxylic acid metabolite LINA-M3 (Table 1). A cysteine adduct and its sulfate conjugate were additionally observed in urine after intravenous administration of LINA [13].

Metabolite
In the study of Xu et al. [14], absorption, metabolism and excretion of omarigliptin (OMA) were evaluated in healthy male subjects after a single oral dose of 25 mg of [14C]-OMA. Radioactivity levels in plasma and excreta were determined via accelerator mass spectrometry (AMS). As a result, minimal metabolism of OMA was observed, as indicated by the fact that the parent drug accounted for ca. 89% of the radioactivity in urine. However, some oxidative metabolites were detected in plasma, with each comprising less than 3% of the total radioactivity.   (Table 1). In addition, the phase II reactions were shown to be involved, i.e., glucuronidation, sulfation, and conjugation with glycine-cysteine moieties. The chemical structure of linagliptin (LINA) is based on xanthine moiety that distinguishes this drug from other gliptins. It may offer some differences in pharmacokinetic and pharmacodynamic properties of LINA, e.g., its low dissociation from the enzyme and greater potency than other gliptins. It is known that LINA is predominantly eliminated unchanged after both oral and intravenous administration [7]. However, the inactive metabolite LINA-M1 was identified in plasma as a major metabolite after oral administration. A two-step mechanism was proposed for its formation, i.e., CYP3A4-dependent conversion of the secondary amine of the parent drug to the corresponding ketone via oxidative deamination, followed by reduction. In excreta, the main metabolite after oral and intravenous administration was LINA-M2 formed by hydroxylation of the methyl group of the butinyl side chain. Next, minor metabolites could be formed by combinations of the following reactions: oxidation of the butinyl side chain and the piperidine moiety followed by oxidative degradation of the piperidine moiety, and finally N-acetylation and glucuronidation. The oxidation of the methyl group at position 4 of the quinazoline moiety was also proposed, and resulted in the corresponding carboxylic acid metabolite LINA-M3 (Table 1). A cysteine adduct and its sulfate conjugate were additionally observed in urine after intravenous administration of LINA [13].
A few metabolites of sitagliptin (SITA) were detected, e.g., diketone metabolite SITA-M1. In addition, hydroxylation of both amine group and aromatic ring followed by formation of glucuronide metabolites, as well as oxidation of NH2 and hydroxylation followed by loss of HF to form SITA-M2, was proposed (Table 1). On the other hand, it was clearly shown that only 3.1% of the parent drug was metabolized over 2 h incubation [19].
n.a. [7] n.a.-not available in the literature; *-low-resolution LC-MS. A few metabolites of sitagliptin (SITA) were detected, e.g., diketone metabolite SITA-M1. In addition, hydroxylation of both amine group and aromatic ring followed by formation of glucuronide metabolites, as well as oxidation of NH2 and hydroxylation followed by loss of HF to form SITA-M2, was proposed (Table 1). On the other hand, it was clearly shown that only 3.1% of the parent drug was metabolized over 2 h incubation [19].
In the next paper concerning SITA, Vincent et al. [19] reported the phase I and phase II reactions of [14C]-SITA and detected its six metabolites in human plasma, urine and feces, using radiometric detection and LC-MS/MS, using the TurboIonSpray ® interface operated in positive ion mode. The metabolites were shown to be the products of hydroxylation and N-sulfation of SITA. In addition, oxidation of piperazine ring followed by cyclization was proposed to form SITA-M3 (Table 1). Glucuronidation was also included in SITA metabolism, but it was observed that glucuronides were further transformed and were not detected in feces.
As far as the next drug from gliptins, i.e., vildagliptin (VILDA) is concerned, its metabolism accounts for near 70% of the dose and is the dominant elimination pathway in humans. The major metabolite VILDA-M1 (Table 1) is pharmacologically inactive and is the product of hydrolysis of the cyano moiety. Next, identified metabolites of VILDA were formed by glucuronidation. At the same time, it was shown that the kidneys may be one of the major organs contributing to the hydrolysis of VILDA to VILDA-M1 [7].
In the next two studies from the literature [22,23], it was investigated whether different gliptins, i.e., ANA, ALO, LINA, SITA and VILDA, have the potential to covalently bind to macromolecules and, as a consequence, to initiate immune-mediated hepatotoxicity. As a result, the possibility of covalently binding to macromolecules was shown for ANA and VILDA. It was also found that VILDA and ANA, which both contain a cyanopyrrolidine moiety, rapidly reacted with L-cysteine in no enzymatic manner. Structural analysis of these adducts revealed that the nitrile moieties of both drugs were irreversibly converted to thiazoline acid. Thus, it was found that VILDA and ANA have the potential to covalently bind to a thiol residue of L-cysteine in proteins. Such binding may lead to unpredictable immune responses in humans.

Analytical Methods for Elucidating the Metabolism of Gliptins
Structural characterization of ANA and its metabolites in humans was performed with radiometric and chromatographic methods ( Table 2). Quantitative measurements in respective fractions of mobile phase were taken using a radio flow detector (for urine and feces) or liquid scintillation counter (LSC) (for plasma). Comparing the LC retention times and MS spectra with respective synthetic standards allowed identification of some of the detected metabolites [10].
To identify EVO and its metabolites in human liver microsomes, an orbitrap mass spectrometer coupled with the UPLC system was used. Separation was performed on a C8 column using gradient elution. A higher-energy collision dissociation (HCD) system was employed to investigate the fragmentation pattern of EVO and its metabolites ( Table 2). As a result, five EVO metabolites were detected and identified via comparison with the retention times and MS/MS spectra of the corresponding standards [11]. In addition, a HPLC system coupled with a radioactivity detector (RAD) and an ion trap mass spectrometer (Table 2) was used in order to study of metabolism of [14C]-EVO after oral administration to male rats and dogs [12].
In the study of Xu et al. [14], absorption, metabolism and excretion of [14C]-OMA were evaluated in healthy male subjects via the AMS method. It is worth noting that AMS shows a special power to separate a rare isotope from an abundant neighboring mass [24]. In the mentioned study, the metabolite profiling was conducted using HPLC fractionation followed by AMS to generate radiochromatograms, and finally, metabolites were identified by LC-MS/MS with a hybrid system that consisted of a quadrupole and orthogonal acceleration TOF tandem mass spectrometer (Table 2).
A specific and rapid LC-MS/MS method [15] was proposed for simultaneous determination of SAXA and its active metabolite, 5-hydroxy-SAXA (SAXA-M1) in human plasma. Sample preparation was accomplished via solid phase extraction (SPE) using SDS as an ion-pair reagent. In addition, next, the LC-MS/MS method for quantitative determination of SAXA and SAXA-M1 was described in the literature [16] (Table 2).
In the study of Peng et al. [17], hydrophilic interaction liquid chromatography (HILIC) coupled with mass spectrometry (HILIC-MS) was applied to determine the plasma concentrations of SAXA and SAXA-M1 in the presence of metformin. Chromatographic separation was achieved on a HILIC Chrom Matrix HP amide column that was the optimal choice because of the high differences in hydrophilic properties of the mentioned analytes. Mass spectrometry was performed using a quadrupole tandem mass spectrometer in an MRM mode under positive ESI (Table 2).
A similar analytical strategy was applied in the next study of Khreit et al. [18], where the metabolism of SITA was investigated via incubation of the drug with rat hepatocytes. For LC method, a zwitter ionic hydrophilic interaction ZIC-HILIC column with gradient elution and an orbitrap mass spectrometer were used (Table 2).
Additionally, experiments on TENE metabolism were based on the use of AMS technology [20]. Metabolite profiling in plasma, urine and feces was performed via offline measurement of radioactivity using a HPLC system. Radioactivity was quantified using a LSC analyzer, with automatic quench correction by an external standard method, while identification of TENE and its metabolites was performed via ESI using a orbitrap hybrid Fourier Transform (FT) mass spectrometer ( Table 2). Such hybrid mass spectrometer combines a linear ion trap MS and the orbitrap mass analyzer. Ions generated by a compound are collected and followed by axial ejection to the C-shaped storage trap which is used to store and cool ions before injection into the orbital trap. Signals from each of the orbital trap outer electrodes are amplified and transformed into a frequency spectrum by fast FT, which is finally converted into a mass spectrum [25].
Next, a rapid and sensitive LC-MS/MS method for simultaneous quantification of TENE and its active metabolite, TENE-M1, in human plasma was elaborated by Park et al. [21], using deuterated TENE as an internal standard. This method was proposed as an alternative to the AMS method described above [20], in order to avoid its disadvantages, including its complicated procedure and the expense of designing the radioactive isotope (Table 2).  SAXA HILIC Chrom Matrix HP amide column (3.0 × 100 mm, 5 µm), ACN and 5 mM ammonium formate buffer containing 0.1% HCOOH. Ion spray voltage 5500 V, ion spray temperature 550 • C, ion source gas 1: 50 Arb, ion source gas 2: 55 Arb, curtain gas (N 2 ) 30 Arb, collision gas (N 2 ) 10 Arb, entrance potential 10 V, collision cell exit potential 12 V. [20] TENE CAPCELL Pak C18 column (2.0 × 2150 mm, 5 µm) and isocratic elution with ACN, MeOH and H 2 O, 025 mL/min; temperature of the column and autosampler 50 • C and 10 • C. MS: collision gas 5 psi, curtain gas 10 psi, ion source gas (nebulizer) 30 psi, ion spray voltage 5500 V, and collision energy of 37 eV for TENE; declustering potential, entrance potential, and collision exit potential were 106 V; ESI positive ion mode using MRM.

Chemical Degradation of Gliptins
Using HPLC method with two detectors, i.e., UV and charged aerosol detector (CAD), it was shown that alogliptin (ALO) degraded about 60% in different stress conditions. After acidic degradation, two additive peaks were observed with both the CAD and UV detectors. One peak was observed early in the run and the other after the peak of the parent ALO. At the same time, the retention times observed with the CAD and the UV detector were not the same, suggesting the presence of two different compounds. Under thermal conditions, one degradation peak was observed in both detectors after the peak of parent ALO, with a retention time similar to the peak observed in the UV detector in acidic conditions. Under the photolytic conditions, one degradation peak before ALO using the CAD detector and one peak after ALO using the UV detector were observed, similarly to the peaks detected after acidic degradation [26].
Based on this gem-dimethyl effect, formation of ANA-D5 can be explained by hydrolysis of nitrile to amide followed by cyclization between the imine carbon and secondary amine group with loss of the water molecule. Similarly, ANA-D7 could be formed via hydroxylation of methylpyrazolopyrimidine ring followed by cyclization with an elimination of water molecule. As far as ANA-D6 was concerned, it was formed due to N-hydroxylation of the secondary amine (attached to tetrahedral carbon) under oxidative stress conditions.  (Table 3) [28,29]. In the study of Pandya et al. [31], two degradation products were detected, i.e., ANA-D4 that was described above [28,29] and one new degradant ANA-D8 (Table 3). This oxidative degradant was characterized as N- 5-a]pyrimidine-N-oxido-6-carboxamide [31].
What is more, it was emphasized that this impurity contains the N-acylated amino aryl groups, which are well-known DNA reactive substances [33].
Forced degradation studies were also conducted for tablets containing LINA and metformin. As a result, three new possible degradants were identified, i.e., LINA-D13, LINA-D14 and LINA-D15 (Table 3). The LINA-D13 was formed due to the presence of reducing sugar, while the LINA-D14 was formed due to the presence of HCOOH (because of their presence as tablet excipients). As far as LINA-D15 was concerned, it was formed in the presence of metformin in these two component tablets. However, the authors also proposed a general mechanism to form amide in the presence of amine and acid that could also be adequate for other similar APIs [34].
It is also worth noting that some studies from the literature concerning LINA and its process-related impurities showed possible risk of toxicity of these molecules, according to different computational programs, e.g., pKCSM, Osiris and LAZAR [35]. It was concluded that some impurities of LINA presented prediction of mutagenicity with the use of all software used. Moreover, some suggestions of carcinogenic effect with the use of Osiris and LAZAR were discussed. What is more, an experimental Comet assay was performed, in which LINA and its process related impurities showed some DNA damage at concentrations of 10 × Cmax, which was in accordance with the suggestions for in silico studies mentioned above.
The work of Tantawy et al. [36] was elaborated for profiling the chemical stability of omarigliptin (OMA). The formed degradation products were then identified by means of IR and MS. It was found that OMA showed stability under photolytic and thermal conditions, but was sensitive to acidic and alkaline conditions, where complete degradation was achieved, and similar degradants were observed. This assumption was confirmed by LC-MS analysis, where their positive mass scan determinations showed the same molecular ion peaks. OMR was also susceptible to oxidative degradation and was completely oxidized when stored at room temperature for 50 h. The formation of the oxidative derivative product OMA-D1 (Table 3) was confirmed via the respective GC-MS method.
What is more, the next two degradants, i.e., SITA-D4 and SITA-D5 were identified after oxidative degradation of SITA [39]. Their exact masses indicated that they were formed by dehydration of 3-aminobutan-1-one. The results showed that they could be positional isomers with respect to the double bond, where both molecules had the same m/z (Table 3).
In the next study from the literature [40], SITA-D1, SITA-D2 and SITA-D3 degradants were identified after degradation in acidic and alkaline conditions. In addition, two new products with the same m/z, i.e., SITA-D7 and SITA-D8 were detected after decomposition of the parent drug with the loss of ammonia from an abundant ion under basic and oxidative stress (Table 3).
Forced degradation of VILDA was performed at extreme pH values and in oxidative conditions [46]. UHPLC-DAD-MS experiments allowed us to separate and identify four degradants of VILDA, from VILDA-D1 to VILDA-D4 (Table 3). VILDA-D1 was detected when the drug was stressed in acidic and basic conditions. It was supposed that the pyrrolidine-2-carbonitrile motif was left from the structure of the parent drug to produce the corresponding carboxylate. Thus, it could be described as [(3-hydroxytricyclo[3.3.1.13,7]decan-1-yl)amino]acetic acid. As far as VILDA-D2 is concerned, it was detected in basic, oxidative and acidic conditions. It was probably formed via hydrolysis of the cyano group of VILDA into the amide one. It could be identified as 1-{[(3-hydroxytricyclo[3.3.1.13,7]decan-1-yl)amino]acetyl}pyrrolidine-2-carboxamide. Thus, the results from the literature pointed to hydrolysis of the cyano group in basic, oxidative, as well as acidic conditions. Further hydrolysis of the amide group of VILDA-D2 afforded the corresponding carboxylic acid, leading to VILDA-D3 (Table 3).

Methods for Elucidating Degradation Pathways of Gliptins
Stability-indicating HPLC methods using an UV detector and a charged aerosol detector (CAD) were elaborated for the assessment of chemical stability of ALO. The analysis was performed on a C8 column (4.6 × 250 mm, 5 µm) at a flow rate of the mobile phase of 0.8 mL/min, using ACN-10 mM acetate buffer of pH 3.5 (90:10, v/v) and UV detection at 241 * [48] n.a.-not available in the literature; *-low-resolution LC-MS; **-high-resolution LC-MS.

Methods for Elucidating Degradation Pathways of Gliptins
Stability-indicating HPLC methods using an UV detector and a charged aerosol detector (CAD) were elaborated for the assessment of chemical stability of ALO. The analysis was performed on a C8 column (4.6 × 250 mm, 5 µm) at a flow rate of the mobile phase of 0.8 mL/min, using ACN-10 mM acetate buffer of pH 3.5 (90:10, v/v) and UV detection at 275 nm [26]. The above method is worth mentioning because of its use of the CAD detector, which may offer entirely new possibilities by exploiting a large range of analytes. This is frequently impossible when using other detection methods. Other benefits of using CAD include an accurate response for all nonvolatile compounds independently of the analyte structure and high sensitivity [49].
Other work from the literature [27] showed significant degradation of ALO leading to five degradants. Highly specific and efficient LC-MS-Q-TOF was used for their separation. The eluent was split and introduced into the ESI source of the Q-TOF unit, operating in positive ionization mode ( Table 4).
The next few studies from the literature were assessed to examine the degradation behavior of ANA under different stress conditions followed by elucidation of the structures of respective degradants. All the stressed samples were analyzed using UPLC/PDA and LC-MS/MS methods (Table 4) [28,29]. In the study of Pandya et al. [31], preparative HPLC separation of ANA and its degradants was performed with a chromatographic system with a PDA detector and a Daisogel-SP-100-10-ODS-P column (20 × 250 mm, 10 µm). The flow rate of the mobile phase was kept at 5 mL/min. A HRMS system consisting of an UHPLC, NANO Chipcube ® interface and iFunnel Q-TOFs was used ( Table 4).
Separation of LINA and its 12 stress degradation products was achieved using HPLC-PDA and LC-MS/MS experiments [32] (Table 4). Separation of LINA and its degradation products was also achieved on a Zorbax Eclipse XDB C18 column (250 × 50 mm, 10 µm) using mobile phases consisting of phosphate buffer (20 mM) of pH 2.5) and ACN with the flow rate of 0.6 mL/min. The gradient program, the column temperature 30 • C and UV detection at 225 nm were used [33].
Forced degradation studies were also conducted for tablets containing LINA and metformin [34]. Separation of the drugs and degradants was achieved using LC-MS and UPLC-TOF/MS methods with a Zorbax SB-Aq column, which is a RP HPLC phase which is designed to retain hydrophilic and other compounds when using highly aqueous mobile phases, even including 100% water ( Table 4).
The work of Tantawy et al. [36] was elaborated for profiling of chemical stability of OMA under various stress conditions. The formed degradation products were then identified by means of IR and MS. In addition, thin-layer chromatography (TLC) with mobile phase of MeOH:ethyl acetate:33% NH4OH (2:8:1, v/v/v) was used for separation. What is more, formation of the oxidative degradant of OMA (OMA-D1) was confirmed by the respective GC-MS method.
In the study of Sridhar et al. [37], comprehensive estimation of chemical stability of SAXA was performed using LC-PDA, LC-MS, and LC-MS/MS methods. In the next study [40], a new LC-MS analysis of SITA and its degradants was performed using an accurate mass Q-TOF spectrometer equipped with a dual ESI-Jet stream ion source (Table 4).
To develop a suitable RP-HPLC method for identifying TENE and its degradation products, a system with a variable wavelength UV detector was used [41]. TENE was stressed under different conditions and the obtained samples were subjected to HPLC separation. For LC-MS experiments, the products were ionized in ESI mode to obtain their mass data (Table 4).
UPLC-MS/MS, UPLC-UV, HPLC-UV and UHPLC-UV methods were developed for multifaceted comparative analysis of TRELA stability (Table 4) [43]. A validated HPLC-UV method was selected for degradation kinetic study due to better separation for the degradation products and lower cost than the LC-MS/MS method. In the study of Luo et al. [44], rapid, sensitive and accurate LC-UV and LC-MS methods were developed for separation of TRELA and its eight potential process-related impurities. Between the methods, HPLC-UV using a XSelect CSH C18 column was proposed (Table 4).
It is worth noting that charged surface hybrid (CSH) technology can help overcome some problems such as peak shape asymmetry and poor loading for basic compounds, column bleed and slow mobile phase reequilibration. It was also found that the XSelect CSH C18 column yielded the best chromatographic separation between the critical pairs of peaks, while the other columns caused some peaks either to be co-eluted or to have comparatively low resolution. In addition, the XSelect CSH C18 column can give superior peak shape for basic compounds due to their improved loading capacity under low-ionic strength conditions and has excellent stability at low pH [50].
Forced degradation of VILDA was performed in extreme pH values and oxidative conditions [46]. Then, selective LC-UV was used for quantitative determination of a nondegraded drug in the presence of its degradation products. Finally, identification of degradation products of VILDA was performed using an UHPLC-DAD-MS with positive ESI ( Table 4). The main objective of the next chromatographic method [47] was to achieve the separation of VILDA from VILDA-D4 and VILDA-D5 (Table 4). Additionally, a simple RP-HPLC-UV method was developed for separation of VILDA and its new degradation products, e.g., VILDA-D6, VILDA-D7, VILDA-D8 and VILDA-D9. Acetate buffer of pH 7.5 and MeOH with an Athena C18-WP (250 mm) column was used [48]. Athena C18-WP packing has great stability due to the high purity of spherical silica and can be used with 100% water as mobile phase. As a consequence, it can be used for the separation of many compounds with different acid-base properties. In the present study, phosphate or acetate buffers were mixed with MeOH in the ratio 90:10 (v/v).

Compound Conditions
Ref.

Conclusions
All gliptins except for LINA predominantly undergo renal excretion, with ca. 70-80% of the dose eliminated as unchanged parent compound in urine. In contrast, LINA is excreted mostly unchanged in feces, and therefore appears to be safe in diabetic patients with renal disturbances. What is more, in patients with liver impairments, the dose adjustment seems unnecessary. However, the dose reduction of the gliptins with predominantly renal excretion (e.g., ALO, SITA, SAXA, but not VILDA) is needed in the case of such renal complications [7].
As far as metabolic transformations of gliptins are concerned, they undergo CYPdependent and CYP-independent metabolism, and after that some of them have active metabolites. Bearing in mind all of the above results, it could be summarized that the dominant metabolic pathways for gliptins are hydroxylation and/or oxidation, affecting the butinyl side chain and particular heterocyclic moieties (Figures 2 and 3).
As far as the reactions of II phase of the metabolism are concerned, N-acetylation (ALO, LINA), glucuronidation (EVO, SITA, VILDA) as well as sulfation (EVO, SITA) were reported, although these products, especially glucuronides, were further transformed and were not typically detected in excreta. After oral administration of VILDA, the cysteine-and glutathione-containing adducts were also detected in rat plasma, urine, feces and bile. What is more, formation of these metabolites was confirmed by in vitro incubation experiments. In addition, the possibility of forming the adducts with cysteine was also shown for ANA, EVO and LINA. It is worth noting that such reactivity may lead to unpredictable immune responses in humans [23].
Biomedicines 2023, 11, x. https://doi.org/10.3390/xxxxx www.mdpi.com/journal/biomedicine with renal disturbances. What is more, in patients with liver impairments, the dose ad justment seems unnecessary. However, the dose reduction of the gliptins with predomi nantly renal excretion (e.g., ALO, SITA, SAXA, but not VILDA) is needed in the case o such renal complications [7]. As far as metabolic transformations of gliptins are concerned, they undergo CYP dependent and CYP-independent metabolism, and after that some of them have active metabolites. Bearing in mind all of the above results, it could be summarized that the dominant metabolic pathways for gliptins are hydroxylation and/or oxidation, affecting the butinyl side chain and particular heterocyclic moieties (Figures 2 and 3).  As far as the reactions of II phase of the metabolism are concerned, N-acetylation (ALO, LINA), glucuronidation (EVO, SITA, VILDA) as well as sulfation (EVO, SITA) were reported, although these products, especially glucuronides, were further transformed and were not typically detected in excreta. After oral administration of VILDA, the cysteineand glutathione-containing adducts were also detected in rat plasma, urine, feces and bile What is more, formation of these metabolites was confirmed by in vitro incubation experiments. In addition, the possibility of forming the adducts with cysteine was also shown for ANA, EVO and LINA. It is worth noting that such reactivity may lead to unpredictable immune responses in humans [23].
It was interesting to observe that metabolism and drug degradation pathways may It was interesting to observe that metabolism and drug degradation pathways may overlap for many gliptins (ANA, ALO, SAXA, TRELA and VILDA), resulting in the formation of similar constituents. The major transformation of the above drugs, during metabolism as well as degradation, was proposed as the cyano group hydrolysis to generate the carboxylic acid or amide, resulting in such products as ANA-M1 and VILDA-M1 ( Figure 4). It was interesting to observe that metabolism and drug degradation pathways may overlap for many gliptins (ANA, ALO, SAXA, TRELA and VILDA), resulting in the formation of similar constituents. The major transformation of the above drugs, during metabolism as well as degradation, was proposed as the cyano group hydrolysis to generate the carboxylic acid or amide, resulting in such products as ANA-M1 and VILDA-M1 (Figure 4). The presented results pointed to hydrolysis of the cyano group in basic, oxidative, as well as acidic conditions. Further hydrolysis of the amide group afforded the corresponding carboxylic acid.
Metabolism data could be helpful for deriving the safe levels of the related compounds by using it to qualify these impurities to reduce the cost, animal use and time. The presented results pointed to hydrolysis of the cyano group in basic, oxidative, as well as acidic conditions. Further hydrolysis of the amide group afforded the corresponding carboxylic acid.
Metabolism data could be helpful for deriving the safe levels of the related compounds by using it to qualify these impurities to reduce the cost, animal use and time. Thus, our considerations about gliptins could be leveraged to enhance their behavior in pharmaceutical products and link the quality of respective pharmaceutical products to clinical practice. As was also shown, the main analytical tools for such modern analysis of gliptins are chromatographic methods like LC-MS or HRMS. Such hyphenated techniques present the power of separation and quantification for more in-depth analysis of complex samples from metabolism and degradation experiments, in turn allowing us to solve these complex problems. In addition, radiometric methods using the stable isotopes, as well as chromatographic methods with the deuterated standards, could be used for determination of the mentioned drugs in biological fluids or excreta. As was described above, [14C]-technology was widely applied to track the gliptins and to provide their metabolic profiles.