TO DOWNLOAD A COPY OF THIS POSTER, VISIT WWW.WATERS.COM/POSTERS ©2017 Waters Corporation INTRODUCTION Recombinant human insulin and its analogs are well known as the standard of care for insulin dependent diabetes. With the increasing incidence of type 2 diabetes and limited alternative treatments, numerous insulin biosimilars are being developed. LC-MS has emerged as a key platform for quantification across the drug discovery and development continuum, showing advantages over ligand binding assays (LBAs) which are challenged with specificity issues and multiplexing capabilities. Previous work has highlighted the ability of microflow LC-MS using the ionKey/MS system as an effective approach to obtaining ultimate analytical sensitivity for insulin and its analogs with low sample volume. 1 Microflow LC-MS as an alternative to analytical scale is well known to address high analytical sensitivity requirements, yet often with the compromise of longer cycle times. Recent work has shown that the ionKey/MS system can achieve 3 minute cycle times with up to a 6- fold increase in analytical sensitivity when moving from analytical scale to microflow scale. 2 This work described herein, demonstrates that analytically sensitive, accurate, robust and high throughput bioanalytical analysis can be achieved with microflow regimes using the ionKey/MS system and 300 µm I.D. iKey HT Separation Device for insulin and five of its analogs extracted from human plasma for clinical research. INCREASING THROUGHPUT FOR HIGH ANALYTICAL SENSITIVITY BIOANALYSIS OF HUMAN INSULIN AND ITS BIOTHERAPEUTIC ANALOGS USING MICROFLOW LC-MS/MS FOR CLINICAL RESEARCH Anthony Marcello, Michael Donegan, James Murphy, and Erin E. Chambers Waters Corporation 34 Maple Street Milford, MA 01757, USA METHODS Sample Preparation Samples were prepared according to the method outlined by Chambers et. al 3 . Human plasma was fortified with human insulin and five insulin analogs, Levemir (insulin detemir), Apidra (insulin glulisine), Humalog (insulin lispro), Lantus (insulin glargine) and Novolog (insulin aspart), at concentrations ranging from 25 - 10,000 pg/ mL. Bovine insulin was used as the internal standard (IS). Samples were extracted from plasma by first performing protein precipitation where 250 µL of sample was added to 25 µL of IS and 250 µL of 1:1 (v:v) ACN:CH 3 OH containing 1% acetic acid. Samples were vortex mixed and centrifuged at 13,000 rcf for 10 minutes. The supernatant was removed and added to 900 µL of 5% NH 4 OH in water. Following the protein precipitation step, samples were further refined and concentrated by solid phase extraction (SPE). An Oasis MAX µElution plate was first conditioned with 200 µL of methanol, followed by equilibration with 200 µL of water. The diluted supernatant from the protein precipitation step was added in two steps and slowly pulled through the SPE plate. Samples were then washed with 200 µL of 5% NH 4 OH in water, followed by 200 µL of 5% methanol containing 1% acetic acid in water. The insulins were eluted with 2 x 25 µL of 60/30/10 methanol/water/acetic acid and diluted with 50 µL of water for analysis. LC-MS Conditions LC-MS quantification of insulin and its analogs was performed using a Waters ionKey/MS ™ System. Chromatographic separation was performed with a Waters® BEH C 18 iKey (300 µm x 50 mm, 1.7 µm). Mobile phase A and B consisted of water and acetonitrile, respectively, each containing 0.1% formic acid. To maximize analytical sensitivity, 15 µL of sample was injected using a trap and elute workflow. Trap conditions involved loading the sample with a 85:15 mobile phase A:B ratio onto a 300 µm x 50 mm Symmetry C 18 trap column for 2 minutes at a flow rate of 25 µL/min. The flow was then reversed into the analytical column for analysis. The insulins were eluted using a linear gradient (15-55% B) over 2 min. with re-equilibration to 5 minutes and a flow rate of 6 µL/min. LC-MS quantification of all analytes was performed using a Waters Xevo TQ-XS triple quadrupole. MS conditions are summarized in Table 1. DISCUSSION In a previous application 1 , ultra high analytical sensitivity for insulin and its analogs was demonstrated by applying microflow LC/MS. That application utilized the 150 µm I.D. iKey, which offered unmatched analytical sensitivity. In general, microflow LC/MS has been shown to offer significant analytical sensitivity gains over traditional 2.1mm x 50 mm UPLC approaches. 4,5 However, one of the downsides of using a microfluidic approach is that lower flow rates can also translate into longer run times. This challenge has led to the development of the iKey HT. The use of a larger microfluidic channel enables higher pressures that allows for higher flow rates and shorter run times. The LC gradient in this study utilized a 5-minute cycle time, thereby providing faster results while still maintaining very high analytical sensitivity. Figure 3 shows a comparison of analytical sensitivity and run times for insulin analogs among conventional UPLC, 150 µm iKey, and 300 µm iKey HT. The iKey HT had the fastest run time among all three methods with the analytical sensitivity comparable to the levels acquired with the 150 µm iKey. It should be noted that there are some differences among the methods--the UPLC method extracted 250 µL and injected 30 µL. The ionKey HT method extracted the same volume, but only injected 15 µL. The ionKey 150 µm method extracted only 100 µL of sample and injected just 10 µL. When considering the injection volume and extraction volume differences between the methods, the overall analytical sensitivity gains observed with microflow in both 150 µm and 300 µm scale becomes even more pronounced when compared to conventional scale. CONCLUSION This study highlights the novel and highly efficient ionKey/MS system, combined with the robust and high throughput 300um I.D. iKey HT separation device for the analytically sensitive and accurate quantification of therapeutic analytical insulin and five analogs using tandem quadrupole LC-MS. For ultra analytical sensitivity, the iKey 150 µm I.D. columns achieved LLOQ of a 25 pg/mL for most analogs while using the least amount of sample, but with 1.7X run time compared to the 2.1 mm analytical scale method. The iKey HT (300 µm x 50 mm) delivered a minimal compromise on analytical sensitivity with a LLOQ of 50 pg/ml and a 2- fold improvement in run time compared to the 150 µm microflow method, which is more appropriate for routine bioanalysis labs. The fully integrated ionKey/MS system enables bioanalysts the flexibility to modulate a microflow LC-MS system between ultra analytical sensitivity analysis and higher throughput by simply switching between the iKey columns for clinical research. RESULTS Human plasma extracts yielded LLOQ’s in the range of 25-100 pg/ml for both the insulin analogs, and endogenous insulin. Standard curves were established over 3 orders of magnitude with R 2 values > 0.99 and mean accuracy values > 93% for all analytes. Figure 1 illustrates the area response for the lowest 3 standards of Apidra (glulisine) as compared to the human plasma blank. The standard curve for Apidra is shown in Figure 2. Figure 2: Calibration curve from the insulin analog Apidra (glulisine) extracted from plasma and analyzed using the ionKey HT Table 1. MS conditions for human insulin, insulin analogs and the internal standard bovine insulin References Chambers, E.E., et al., Reducing Sample Volume and Increasing Sensitivity for the Quantification of Human Insulin and 5 Ana- logs in Human Plasma using ionKey/MS, Waters Application Note, p/n 72000519EN (2016) Michael Donegan et al., High Throughput Microflow LC-MS: Sensitivity Gains on a Practical Timescale. App Note, Waters Corp., (2016). Chambers, E.E., et al., Multidimensional LC-MS/MS Enables Simultaneous Quantification of Intact Human Insulin and 5 Re- combinant Analogs. Analytical Chemistry, 86(1), 694-702 (2014). Y.W. Alelyunas, G. Roman, J. Johnson, C. Doneanu and M. Wrona, High throughput analysis at microscale:performance of ionKey/MS with Xevo G2-XS QTof under rapid gradient conditions. J. Appl. Bioanal. 1(4), 128-135 (2015) P.D. Rainvile, J. Langridge, M. Wrona, I. Wilson and R. Plumb. Integration of microfluidic LC with HRMS for the analysis of an- alytes in biofluids:past, present and future. Bioanalysis, 7(11), 1397-1411 (2015). Specific Insulin MRM Transition Cone Voltage (V) Collision Energy (eV) Human insulin 1162 -> 226 50 40 Glargine 1011->1179 60 30 Detemir 1184-> 454.4 60 35 Aspart 971.8 -> 660.8 50 22 Lispro 1162.2 -> 217 80 45 Glulisine 1165 -> 1370 60 20 Bovine (IS) 956.6 -> 1121.2 60 18 Figure 1: Chromatographic performance observed from the lowest 3 standards of Apidra (glulisine) extracted from plasma and analyzed using the ionKey HT Figure 3: Insulin performance comparison using conventional UPLC (2.1 x 50 mm), micro- flow LC iKey HT (300µm x 50 mm) and microflow LC with standard iKey (150 µm x 50 mm) Std. Curve Range (pg/mL) Analyte 2.1 x 50 mm 300 µm x 50 mm 150 µm x 50 mm Lispro 50-10,000 100-10,000 25-10,000 Glargine 50-10,000 50-10,000 25-10,000 Detemir 200-10,000 100-5,000 50-10,000 Glulisine 50-10,000 25-10,000 25-10,000 Aspart 100-10,000 50-10,000 25-10,000 Sample Vol 250 µL 250 µL 100 µL Injecon Vol 30 µL 15 µL 10 µL Run Time 8 min 7 min 13.5 min For Research Use Only, Not for use in Diagnostic Procedures