Insulin is a peptide hormone produced by the beta cells of the pancreas. It is secreted into the plasma to promote efficient glucose utilization and to maintain glucose homeostasis. Measuring the levels of insulin can be challenging as it is present in very small quantities (sub ng/ml) in a complex biological matrix, so must be extracted or enriched before measurement. Although commercial assays are available, results can differ markedly even between laboratories using the same assay. The existence of different insulin analogues further increases the challenge, as an assay needs to clearly detect, differentiate and measure the different entities for clinical purposes, forensics or sports doping.

Scientists at Thermo Fisher Scientific, led by Scott Peterman,1 have developed a new analytical method to detect and quantify intact insulin and its analogues from human plasma. The new method combines insulin-specific extraction with a targeted quantification by high-resolution LC-MS.  This high-throughput semi-automatic method is sensitive and selective.

Specific analyte extraction, simple sample handling

To overcome the challenge of measuring the very low levels of insulin in plasma, Peterman and his team used anti-human insulin antibodies to extract insulin and its analogues from the complex biological matrix. This pan-insulin antibody enables one efficient extraction method to be used for all samples and uses only small plasma volumes.

The antibody was coupled to MSIA D.A.R.Ts (Disposable Automated Research Tips). An automated liquid handler was used for the many aspiration and dispense cycles required to bind the antibody, affinity extract the insulin, rinse and elute the insulin into a microplate. The sample was dried down and re-suspended in an LC appropriate buffer.

Straightforward full scan MS

The samples were separated on a ProSwift monolithic column using a linear gradient and full scan MS data were acquired using a mass range of 800-2000 m/z on a Q Exactive mass spectrometer. The MS was operated in data dependent/dynamic exclusion mode.

Data analysis

The team analyzed the data using Pinpoint software. A straightforward full scan HR/AM MS data acquisition method was used for quantification. Additional qualitative analysis was conducted using the three most abundant precursor charge states (+4, +5 and +6) and the six most abundant isotopes per charge state of each insulin analogue.

The scientists tested the method with four different human insulin analogues (Humulin, Lantus, Apidra, Novorapid), bovine insulin and porcine insulin which was used as an internal standard. They verified that the extraction, detection and quantitation in buffer and plasma was consistent for each analogue, and that detection efficiency was consistent across a biologically relevant concentration range (1.5 to 960 pM in 0.5 mL of plasma). Finally, as the scientists used the same stock human plasma for all insulin analogue spiked samples, they used their assay to determine that the amount of human endogenous insulin detected was consistent between samples, with a variance of 4.26%.

To conclude, Peterson and colleagues have successfully created a new high-throughput, semi-automated assay for insulin immune-extraction, LC separation and quantification with HR/AM MS. The method is very sensitive and detects all isoforms with one assay, consumes very little sample volume and utilizes equipment that is commonly available. In addition, the method can be modified to include other insulin variants as they emerge in the future.

  1. Peterman S, et al. An automated, high-throughput method for targeted quantification of intact insulin and its therapeutic analogs in human serum or plasma coupling mass spectrometric immunoassay with high resolution and accurate mass detection (MSIA-HR/AM) (2014) Proteomics, 14 (1445-1456)