(M1330-02-09) GLP-1 Analog: Accelerating Method Development and Manufacturing with LC-UV/MS

Purpose: Glucagon-like-peptide-1 (GLP-1) receptor agonists have recently gained significant attention as a metabolic regulator for treating type 2 diabetes and obesity resulting in unprecedented demand. In response, manufacturers are actively reassessing their production processes to enhance efficiency and meet growing market needs. Solutions that can utilize information more efficiently to expedite decisions in a timely manner are required to ensure that manufacturers can meet regulatory standards and drug product safety while improving productivity. The purpose of this study was to evaluate the performance of a single quadrupole mass spectrometer in a reversed phase liquid chromatography mass spectrometry (RPLC-MS) workflow in terms of sensitivity and specificity in detection and putative identification of GLP-1 analogs, process-related impurities, and stability indicating impurities under stressed conditions.

Methods: Exenatide from two vendors were dissolved in acetate buffer at pH 4.5. Peptide mapping studies were completed using trypsin digestion of 30 minutes at 37 °C. Chemical stress studies were performed with hydrogen peroxide. Thermal stress studies were performed at 50 °C. The samples were analyzed via a UV detector, a single quadrupole MS, and a quadrupole time-of-flight MS.

Results: A single quadrupole mass detector offers orthogonal screening for impurity detection in exenatide. Exenatide samples from two vendors were tested for purity. UV detection (Fig. 1A) and base peak intensity (BPI) chromatogram from the mass detector (Fig. 1B) indicated high purity for both samples. However, the spectral intensity for vendor 2 was observed to be half as intense compared to vendor 1 despite being prepared at the same concentration. Spectral data (Fig. 1C), indicates an impurity in vendor 2 with the presence of an additional ion series (857.8 m/z and 1071.8 m/z). These ions, which were determined to be [M+5H]⁺⁵ and [M+4H]⁺⁴ charged species, represent a mass difference of +97.1 Da when compared to the expected mass of exenatide. This impurity, which accounts for up to 40% of the sample based on spectral abundance, was overlooked using LC-UV workflow alone. To further investigate this out of specification event, both samples were analyzed at a peptide level using trypsin digestion. Exenatide generated four unique peptide fragments (Fig. 2A) that are readily separated using RPLC (Fig 2B) and can be identified by comparing the observed masses with expected peptide mases obtained by in silico digestion (Table, Fig. 2A). Exenatide from vendor 2 generated a fifth fragment with observed mass of 1,119.5 Da (Fig. 2C). As indicated in the table, the mass difference when compared to the four expected fragment masses could represent an insertion event of a proline, glutamine, lysine, or glutamic acid within ± 1.0 Da accuracy for peak 1 and 2. The close agreement between the observed intact mass difference (+97.1 Da) and fragment T4 mass difference (+96.9 Da) (Table, Fig. 2A), putatively suggested the observed impurity was related to a proline insertion in fragment T4 which was confirmed with a high resolution mass spectrometry. As part of development and manufacturing, drug substances undergo forced degradation and stability testing. LC-UV/MS workflows can be utilized to quickly identify degradants and potential stability indicating attributes for routine monitoring. Exenatide was placed under thermal stress for 30 days. Then several new peaks were observed in the UV chromatogram (Fig. 3A) with 7 peaks exhibiting a significantly elevated peak area. The mass detector is readily capable of detecting all observed peaks and enables the use of orthogonal mass data to putatively identify degradants. Spectra data of peak 2 was determined to correlate to a mass of 3,254.1 Da. Using this information, a quick examination of the amino acid sequence indicated this mass potentially represents a single cleavage event (Fig. 3B). Extracted ion chromatograms (XICs) (Fig. 3C) demonstrated a matching pair of peaks related to the peptide fragments, potentially indicating a site of thermal instability that may need further development in terms of sequence and activity.

Conclusion: The proposed RPLC-single quadrupole MS workflow significantly enhances the analytical capabilities of supporting laboratories by providing complementary, orthogonal mass data. This enables faster, more confident data-driven decision-making and improves overall lab productivity without compromising safety or regulatory compliance. This workflow with an in-line mass detector has successfully identified an exenatide impurity that would likely be overlooked in a UV-only workflow, demonstrating the value of orthogonal mass detection in confirming sample identity and screening for impurities in manufacturing settings. When paired with peptide mapping, it efficiently identifies an amino acid insertion event without the need to isolate or enrich impurities. Moreover, it has positively identified a potential degradation site under elevation temperatures that may require further sequence optimization to improve drug stability. Collectively, these findings with exenatide demonstrate the utility of the proposed QC-friendly MS platform as a method to identify and monitor stability indicating attributes of GLP-1 analogs to ensure drug safety and efficacy, all while meeting regulatory requirements with confidence.


Fig. 1 Exenatide from two vendors were compared for purity via a LC-UV/MS workflow. (A) UV chromatograph. (B) Base peak intensity chromatograph. (C) Spectra of the main species peak.



Fig. 2 (A) Trypsin digestion of exenatide produced four unique peptide fragments. (B) Vendor 1 sample generated four peaks matching the expected masses. (C) Vendor 2 sample generated a fifth peak suggesting the insertion of an amino acid.



Fig. 3 (A) An Exenatide control sample (top panel) was exposed to thermal stress for a 30-day period (middle panel) with 7 “new” peaks observed in the stressed sample. B) As an example, peak 2 was determined to have an observed mass of 3,254.1 Da. which correlates to a non-enzymatic cleavage and was confirmed with retention time and mass using extracted ion chromatograms.