Metabolite Identification in Peptide Drugs and Its Challenges

Peptide drugs are widely used in the treatment of a variety of diseases, including rare diseases, tumors, diabetes, gastrointestinal, orthopedic, immune, and cardiovascular diseases. Up to now, nearly 100 peptide drugs have been approved, playing a crucial role in treating diseases that are challenging to manage with small molecules or therapeutic antibodies ^[1]. Despite their significance, peptide drugs are prone to degradation by proteolytic enzymes, leading to low oral bioavailability, rapid metabolism, and short half-lives due to their chemical and physiological instability. Consequently, peptide structures are often optimized through cyclization, N- or C-terminal modifications, substituting L-type amino acids with D-type or unnatural amino acids, and increasing molecular weight to enhance metabolic stability ^[2].

The introduction of unnatural amino acids or organic linkers necessitates metabolite identification studies to assess major and potentially active metabolites in circulation ^[3]. The results of metabolite identification can assist in identifying the metabolic soft spots, guiding structural optimization, and allowing comparisons of metabolic differences across species, aiding in toxicological species selection ^[4]. Moreover, it can elucidate the clearance pathway of peptide drugs and support new drug filings. In this article, we will introduce the experimental system of peptide drug metabolite research, the challenges, and coping strategies of metabolite identification research. Drawing on the experience of WuXi AppTec DMPK in in vitro and in vivo metabolism study of the GLP-1 receptor agonist semaglutide, we will also illustrate the method of semaglutide metabolite identification and its correlation between in vitro and in vivo metabolites.

Experimental system for metabolite identification research of peptide drug

Peptide drugs are primarily hydrolyzed by peptidases, including exopeptidases and endopeptidases, which operate through distinct mechanisms. These enzymes are widely distributed in the body, including the liver, kidney, stomach, intestine, lung, blood, vascular endothelium, skin, and other tissues and organs. Initially, endopeptidases hydrolyze polypeptides into oligopeptides, which are subsequently broken down into amino acids by exopeptidases (aminopeptidases or carboxypeptidases) at the N- or C-terminus. Variations in the molecular weight and structure of peptide drugs can result in differences in their metabolic enzymes and mechanisms [5]. The kidneys and liver are the primary metabolic organs for peptide drugs, peptides are mainly excreted through urine and feces^[2].

Peptide drugs are commonly administered via subcutaneous injection, intravenous injection, oral routes, and nasal administration. Studying peptide metabolism in plasma or whole blood is essential regardless of the administration route. For peptide compounds being developed as oral or gut-restricted drugs, routine plasma or whole blood metabolism studies should be complemented with in vitro experiments using simulated intestinal fluid (SIF) ^[6]. Intestinal fluid appears to be the primary metabolic barrier of peptides, which may be due to the broad-spectrum specificity of the substrate of pancreatic peptidase, higher enzyme activity, etc. Additionally, in vitro studies in human intestinal epithelial cells indicate that peptides are vulnerable to hydrolases, reductases, and oxidases present in these cells. Given these factors, both simulated intestinal fluid and intestinal epithelial cell systems can be used to investigate the metabolism of peptide compounds in the gastrointestinal tract.

It has been reported that the kidney is the main metabolic site for peptides with molecular weights greater than 1000 Da ^[5]. Smaller peptides are hydrolyzed into amino acids by exopeptidases located on the brush border membrane of the proximal tubule following glomerular filtration. Larger peptides are filtered by the glomeruli and subsequently hydrolyzed into amino acids via endocytosis and lysosomal activity. Kidney S9 and kidney homogenate are one of the most appropriate in vitro experimental systems in the kidney system.

For peptides weighing less than 1000 Da, particularly cyclic peptides, the liver frequently serves as the primary metabolic site^[6]. Moreover, certain peptide drugs, like cyclosporine, depend on cytochrome P450 (CYP3A4) for metabolism and can produce over 30 metabolites through various reactions ^[7]. In the in vitro liver metabolism system, hepatocytes, which contain the highest concentration of metabolic enzymes, are often chosen. However, peptide metabolism in hepatocytes primarily occurs via carrier-mediated transport or endocytosis and pinocytosis, necessitating an extended incubation time. Furthermore, when certain peptides struggle to enter hepatocytes, liver S9 can be utilized as an incubation system to prevent restrictions on peptide entry. In this case, only phase I and phase II metabolic cofactors need to be added to simulate hepatocyte conditions ^[8].

In summary, commonly used peptide in vitro metabolism systems include plasma/whole blood, simulated intestinal fluid/simulated gastric fluid (SGF), intestinal epithelial cells, kidney homogenate/kidney S9, hepatocytes, liver S9, etc. In addition, for in vivo metabolism studies, plasma, urine, feces, and targeted tissue (liver or kidney, etc.) samples from animals can also be used to study the metabolism of peptide drugs. Selecting an appropriate experimental model for peptide drugs at various research stages or administration routes is crucial. Drawing from literature and WuXi AppTec DMPK’s extensive experience in peptide projects, we summarize the research systems for analyzing and identifying peptide drug metabolites, as illustrated in Figure 1.

Figure 1. Experimental system for the study of metabolites for conventional administration and oral administration

Challenges and strategies for peptide metabolite identification

Peptide drugs can generally be categorized into linear peptides made solely of natural amino acids and those containing unnatural amino acids, including both linear and cyclic forms. Due to their different cyclic bonds, cyclic peptides can be divided into homocyclic peptides with only amide bond formation and heterocyclic peptides with at least one non-amide bond formation (such as disulfide bond and thioether bond). Peptides are usually composed of 10-40 amino acids, with molecular weights between small molecules and biological macromolecules, and are prone to form multi-charged ions in mass, resulting in mass signal dispersion. Additionally, peptides generally exhibit poor UV absorption and can suffer from significant endogenous matrix interference in biological samples. Consequently, identifying metabolites of peptide drugs presents several challenges compared to conventional small molecule drugs.

#1. In the in vitro experimental system, the main metabolic pathway of polypeptides is relatively simple, mainly hydrolysis, but there are many hydrolysis sites of polypeptides and the metabolism is fast. In addition to hydrolysis, some polypeptides may also be metabolized by CYP enzymes to undergo phase I metabolic reactions. In in vivo studies, the metabolic pathways of peptide drugs are relatively complex, and there are often multiple metabolic enzymes that produce more metabolites, such as dozens of metabolites produced in the animal matrix after linaclotide administration ^[9]. This complexity makes it challenging to comprehensively and accurately identify all metabolites.

#2. Polypeptides exhibit high physiological activity, typically requiring small dosages, resulting in low concentrations of drugs and metabolites in the body. Additionally, most peptides have weak UV absorption, complicating analysis to the point where it often relies on comparison with a blank matrix for background subtraction in the absence of isotope labeling. Therefore, this is not only a challenge to the technical capabilities of metabolite identification, but also to the resolution and sensitivity of mass spectrometry.

#3. It is difficult for most data processing software to identify the metabolite structure of cyclic peptides with non-disulfide bonds and non-amide groups linked into rings, resulting in the inability to process data through software, and the data needs to be processed manually, which is labor-intensive and requires significant expertise. While some software can assist in discovering and identifying peptide metabolites, issues such as missing metabolites or errors in identifying multiple charges can lead to inaccuracies in metabolite discovery.

#4. Due to the weak absorption of UV characteristics and the low dosage of peptide compounds, it is difficult to use the UV peak area to perform semi-quantitative analysis (calculate the relative abundance) of parent drugs and metabolites, and in most cases, the mass peak area is selected for semi-quantitative analysis. Utilizing mass peak areas for analysis often leads to significant variation, primarily due to two reasons: (1) differences in ionization efficiency between the parent drug and its metabolites can create substantial discrepancies between the relative abundance of mass peak areas and their actual relative content; (2) Peptide compounds frequently generate multi-charged ions, complicating the selection of appropriate ions with varying valence states during integration. This raises questions about whether to choose the monoisotopic peak of multi-charged ions, the peak with the highest abundance, or the combined mass peak area of all ions across multiple valence states.

These factors make it necessary to consider the semi-quantitative analysis of metabolites and data interpretation of peptide drugs, to better provide assistance for peptide drug development based on the results of metabolite identification studies.

To identify the metabolites of peptides comprehensively and accurately, WuXi AppTec DMPK has established a liquid chromatography-high-resolution mass spectrometry (LC-HRMS) analysis platform for the discovery and identification of peptide metabolites. This platform integrates diverse data processing software (Compound Discoverer, BioPharma Finder, UNIFI^®, and Mass-MetaSite) and data processing techniques (background subtraction, mass loss filtering, and feature ion extraction). By employing both targeted and non-targeted analytical methods, along with a combination of software and manual approaches, it enables the rapid and thorough discovery of peptide metabolites in both in vitro and in vivo experimental systems.

Figure 2. High-resolution mass spectrometry methods and data processing for peptide metabolite identification

Case sharing

Semaglutide is the first oral GLP-1 receptor agonist approved by the FDA for the treatment of type 2 diabetes. Currently, there are no published studies on the metabolites of semaglutide in the in vitro kidney S9 system. Therefore, we will use in vitro metabolite identification in the kidney S9 of mice, rats, monkeys, and humans as a case study to illustrate the method of identifying peptide drug metabolites and discuss the correlation of these results with in vivo findings.

In this study, semaglutide was incubated for 4 hours and 24 hours in mouse, rat, monkey, and human kidney S9, terminated the reaction with the organic solvent, and the samples were centrifuged, nitrogen drying and concentration, and finally reconstituted with an appropriate solvent. The metabolites of semaglutide in kidney S9 were identified by the LC-UV-HRMS analysis method and data processing flow established above.

The metabolite analysis process in mouse S9 is illustrated below. The metabolite M18 and the parent drug semaglutide were identified by its LC-UV (Figure 3A) and LC-MS (Figure 3B) chromatogram after 24 hours of incubation in mouse kidney S9.

Figure 3. LC-UV (240-300 nm) (A) and LC-MS chromatogram (B) of samples after 24 h incubation of semaglutide in mouse kidney S9

Nine metabolites were detected in the incubated samples after background subtraction, and the LC-MS chromatograms are shown in Figure 4A. Extracted by characteristic ions (m/z 334.1501, N-terminal 3 amino acid (H-Aib-E) residue sheet ends), 10 metabolites were detected in the incubated sample, and the LC-MS extracted ion chromatograms are shown in Figure 4B. After processing by metabolite data processing software, 15 metabolites were detected and their LC-MS extracted ion chromatograms are shown in Figure 4C.

Figure 4. LC-MS chromatogram of the sample after 24 hours of incubation in mouse kidney S9, LC-MS chromatogram after background subtraction (A), chromatogram of characteristic fragment extraction (B), and LC-MS extracted ion chromatogram (C) of the metabolites calculated by data processing software

Each of these data processing methods has its strengths and weaknesses, making it challenging to identify all metabolites using a single approach. Therefore, employing multiple methods for screening and subsequently comparing and integrating data can lead to a more comprehensive and accurate identification of peptide drug metabolites.

A total of 23 metabolites of semaglutide were detected after 4 and 24 hours of incubation in mouse, rat, monkey, and human kidney S9 using targeted and untargeted analysis techniques, software, and manual combination. The abundance information of each metabolite is shown in Table 1.

Note：ND： Not detected

Table 1. Relative abundance of semaglutide and its metabolites after 4 h and 24 h incubation in kidney S9

The structures of 23 metabolites were identified according to the MS and MSMS data of each metabolite by comparing with that of parent drugs, and the possible metabolic pathways of semaglutide in kidney S9 are shown in Figure 5.

Figure 5. Possible metabolic pathways of semaglutide in kidney S9

According to the above experimental results, the 23 metabolites in human kidney S9 were also detected in three animal species (mouse, rat, and cynomolgus monkey), indicating that there was no significant species difference in the metabolite profile of semaglutide in kidney S9 of various genera. Among them, M18 was the main metabolite, and the metabolic pathway was hydrolysis between lysine at position 26 and alanine at position 25, and hydrolysis between glutamic acid at position 27. The relative abundance of M18 was increased along with the increased incubation time. For example, the relative abundance of M18 was approximately 15% in human kidney S9 after 4 hours of incubation, and increased to approximately 70% after 24 hours of incubation.

According to literature reports ^[10], semaglutide is mainly excreted through the hydrolysis of DPP-4 protein and neutral peptide endonuclease, and the oxidative metabolism of side chains (fatty acid diacid). The metabolic pathways of the main metabolites U6 and U7 in urine are the hydrolysis of endonucleases and the oxidation of fatty acid side chains, that is, the oxidation of fatty acid side chains further occurs based on metabolite M18. The main metabolite of semaglutide in human plasma, P3B, has also been reported to be detectable in the S9 system in vitro (M19). The results showed that the main metabolites or related metabolites of semaglutide in humans could be detected in the S9 system in vitro, and the metabolic results in the kidney S9 system in vitro were consistent with those in vivo.

Concluding remarks

While hydrolysis is the primary metabolic pathway for peptide drugs, analyzing and identifying peptide metabolites remains more challenging than for small molecules. Metabolite identification results of peptides can not only help optimize the structure of peptides and support the selection of toxicological species but also elucidate the clearance pathways of peptide drugs and support the potential safety assessment of metabolites. WuXi AppTec DMPK has established a peptide metabolism, and metabolite analysis and identification research platform based on UPLC-UV-HRMS technology, providing high-quality metabolic and metabolite analysis and identification study reports to help customers accelerate the rapid development of peptide drugs.

Authors: Rong Yang, Ruixing Li, Weiqun Cao 

Talk to a WuXi AppTec expert today to get the support you need to achieve your drug development goals.

Committed to accelerating drug discovery and development, we offer a full range of discovery screening, preclinical development, clinical drug metabolism, and pharmacokinetic (DMPK) platforms and services. With research facilities in the United States (New Jersey) and China (Shanghai, Suzhou, Nanjing, and Nantong), 1,000+ scientists, and over fifteen years of experience in Investigational New Drug (IND) application, our DMPK team at WuXi AppTec are serving 1,600+ global clients, and have successfully supported 1,500+ IND applications.