Best practice for protein quantitation is use of full-length protein stable isotope-labeled internal (protein-SIL) standards to correct for variations in sample preparation, processing (including digestion), matrix effects, and instrument performance. At times protein-SIL internal standards may not be feasible for use, especially during test development due to the lack of commercially available products and the expense of protein-SIL production. The use of winged-SIL and peptide-SIL may be a viable option, but must be carefully evaluated to ensure accurate patient results.
To evaluate the effectiveness of different internal standard approaches using a validated test method and a test method currently in development.
Winged-SIL (TNGSPRLLIKYASESMSGIP*SR*FSGSG), and peptide-SIL (YASESMSGIP*SR) were compared to our current protein-SIL (infliximab 13C6, 14N¬4, Sigma) utilized as the internal standard in our test method for infliximab quantitation by peptide LC-MS/MS. Briefly, sample and IS were added to a filter plate and enriched using Melon™ Gel (Thermo Fisher). The enrichment was digested and analyzed by LC-MS/MS (Sciex API 5000).
Winged-SIL (SVNGKEDAIWNL*LRQAQ and DKSPKFQLFGSPSGQK*DLLFK) and peptide-SIL (EDAIWNL*LR and FQLFGSPSGQK*DL) were utilized during the development of a new test method for lactoferrin. Stool samples (1 g) were extracted with a solution of water:isopropyl alcohol:formic acid (50:33:17) and vigorously vortexed for 10 minutes to lyse cells and form a homogenous solution. After addition of IS, reduction, alkylation, and trypsin digestion were performed. The samples were injected and analyzed by HPLC-MS/MS (Sciex API 6500+).
For infliximab, within-run precision and bias (N=24, 10 mg/mL expected) were acceptable with the protein-SIL (10.0 mg/mL, 3.2% error) but showed more variation for both the winged-SIL (11.8 mg/mL, 21% error) and peptide-SIL (16.2 mg/mL, 31% error). Bias was also evaluated for 24 residual patient comparisons. The protein-SIL showed comparable results with reported; -3.9% difference, range 0.1 to -6.7%. The use of the winged-SIL also gave an acceptable bias; 0.7% difference, range 10% to -11%. The use of peptide-SIL gave unacceptable bias; 32% difference, range 7.9% to 48%.
For lactoferrin, 30 replicates of 4 stool samples were measured, to compare the precision of winged-SIL and peptide-SIL internal standards. We found that winged-SIL were generally advantageous compared to peptide-SIL, but the results were not as conclusive as those outlined above for infliximab. For example, the %CVs for the IS corrected intensity were generally 3X lower for the FQLFGSPSGQKDL target peptide when using winged-SIL compared to peptide-SIL. However, one sample exhibited atypically high %CV when using winged-SIL. For the EDAIWNLLR target peptide the difference in %CV between winged-SIL and peptide-SIL was much closer, ranging from 20% worse to 50% improvement.
The infliximab study confirms that the use of protein-SIL internal standards is best practice for a quantitative protein or peptide mass spectrometry test method, with winged-SIL peptides providing significant benefits over peptide-SIL. Although more exploration is needed to form a definitive conclusion and achieve desired reproducibility criteria, our lactoferrin results suggest that winged-SIL are a much better surrogate for protein-SIL than peptide-SIL. When obtaining protein-SIL is not feasible, winged-SIL peptides provide advantages relative to confirmation of peak picking/retention time and digestion during assay development. The totality of these data demonstrates the advantages of using protein-SIL when possible. The extent to which winged-SIL is viable as an alternative to protein-SIL must be carefully assessed analytically and in the context of the clinical implications for each testing scenario.