Sensitive signature peptide quantification in a complex matrix using accurate mass spectrometry

Methods

Sample preparation: Plasma proteins were precipitated with cold methanol. After centrifugation, the supernatant was discarded while the pellet was solubilized in 200 mM ammonium bicarbonate in 10:90 (v/v) methanol/water. Digestion was performed using trypsin. After 1 hour at 60°C, the solution was acidified by adding formic acid.⁵ The digested plasma was diluted by 200x using 5:1:94 (v/v/v) acetonitrile/formic acid/water. Synthesized peptides (Table 2) were spiked into the digested plasma solution and followed by serial dilution in the matrix. The final injection volume was 10 µL.

Chromatography: The separation was performed at a flow rate of 0.4 mL/min using an ExionLC system. A Phenomenex bioZen Peptide XB-C18 column (2.1 x 50 mm, 2.6 µm, 100 Å) was used for separation. The column oven temperature was set to 40°C. The mobile phase A consisted of 0.1% formic acid in water, while the mobile phase B was composed of 0.1% formic acid in acetonitrile. Gradient conditions are summarized in Table 3. A volume of 10 µL was injected for analysis. All samples were analyzed in triplicate.

Mass spectrometry: Data were acquired in positive mode using Zeno MRM^HR on a ZenoTOF 7600 system. Collision energy (CE), source and MS parameters were optimized for all the signature peptides. A summary of the source and MS parameters and the Zeno trap settings is displayed in Table 4. Additionally, the MRM^HR parameters and fragments used for quantification for each of the signature peptides are summarized in Table 5. Unit Q1 resolution was used for the analysis.

Data processing: Data were processed using the Analytics module in SCIEX OS software with the MQ4 integration algorithm. A 1/x² weighting was used for quantification.

Zeno trap provides greater sensitivity

In traditional TOF MS/MS acquisition, ions are lost between TOF MS spectra acquisitions, resulting in a much lower MS/MS sampling efficiency and sensitivity than MRM on a triple quadrupole mass spectrometer. The Zeno trap increases ion transmission by providing control of the ion beam from the collision cell into the TOF accelerator (Figure 1).

Ions are gated then released based on potential energy. Generally, higher m/z ions are released first then followed by lower m/z ions. A wide range of ions now arrive in the accelerator to be pushed during the same pulse resulting in a 10-fold increase in MS/MS sampling efficiency.⁶

Overview of the signature peptide workflow

FNWYVDGVEVHNAK, AGLIVAEGVTK* and LGLDFDSFR* were used as model peptides to evaluate the quantification of signature peptides on the ZenoTOF 7600 system. The peptides were spiked into processed rat plasma at concentrations ranging from 0.025 fmol/µL to 2000 fmol/µL.

Quantification was performed using the following strategies: 1) using the most sensitive fragment ion and 2) using the sum of multiple highly abundant fragment ions. For the former method, fragment ions y6, y8 and y6 were used for the quantification of peptides AGLIVAEGVTK*, LGLDFDSFR* and FNWYVDGVEVHNAK, respectively. For the latter strategy, the XIC responses from fragment ions y5, y6 and y8 were added for the quantification of peptide LGLDFDSFR*. For the quantification of peptide FNWYVDGVEVHNAK, the XIC responses from fragment ions b2, y4 and y6 were added.

All calibration points were measured in triplicate. For the determination of the LLOQ, a %CV value of less than 20% and accuracy between 80% and 120% was required. For all other concentrations, a %CV value of less than 15% and accuracy between 85% and 115% of the nominal concentration was required. This criteria was applied to both quantification strategies.

Quantification using a single fragment ion

In this workflow, quantification of signature peptides was performed using the Zeno MRM^HR workflow. For the following discussion, the most abundant fragment ion was used for the quantification of the signature peptides.

LLOQs of 0.025 fmol/µL, 0.05 fmol/µL and 0.1 fmol/µL were achieved for peptides AGLIVAEGVTK*, LGLDFDSFR* and FNWYVDGVEVHNAK, respectively (Figure 2). No interferences were observed in the matrix blank.

The linear range for the single fragment ion approach covered concentrations from 0.025 fmol/µL to 2000 fmol/µL (Figure 3). An LDR greater than 4 orders of magnitude was observed for each of the calibration curves.

Calculated concentrations for each calibration point were within ±15% of the nominal value (Table 6). The overall accuracy of the LLOQ was within 8% of the nominal concentration, indicating a highly accurate quantification platform for peptides at low concentrations. As shown in Table 6, the %CV value was less than 15%, demonstrating high reproducibility.

Increasing sensitivity using the summation of multiple fragment ions

The accessibility of TOF MS/MS data can be highly advantageous, as fragment ions can be selected based on overall selectivity and sensitivity for quantification. When multiple highly abundant fragment ions are summed from the target peptide, the assay sensitivity can be further enhanced.³

The linear range for the multiple fragment ion approach covered concentrations from 0.025 fmol/µL to 2000 fmol/µL (Figure 4). Strong linearity was observed for each of the calibration curves with an LDR of greater than 4.3 orders of magnitude.

A 2-fold improvement in LLOQ was achieved when multiple highly abundant fragment ions were summed for quantification (Figure 5). LLOQs of 0.025 fmol/µL and 0.05 fmol/µL were achieved for peptides LGLDFDSFR* and FNWYVDGVEVHNAK, respectively. No interferences were observed in the matrix blank.

Calculated concentrations for each calibration point were within ±15% of the nominal value (Table 7). The overall accuracy was within 8% of the nominal concentration at the level of the LLOQ, indicating a highly accurate platform for low-level quantification. As shown in Table 7, the overall %CV value for the acquired data, including the LLOQ, was less than 20%.

Conclusions

• A highly sensitive signature peptide quantification workflow was developed using the ZenoTOF 7600 system
• Low-amol/µL levels of quantification for signature peptides were reached using a Zeno MRM^HR workflow
• Summation of multiple fragment ions enhances the LLOQ up to 2-fold with the availability of TOF MS/MS data and improvements in MS/MS sensitivity using the Zeno trap
• GLP-level accuracy and precision for signature peptide quantification was achieved with greater than 4.3 orders of magnitude in LDR.

Acknowledgment

We would like to thank Fang Xie and Zhe Wang from Amgen for their consultation and guidance throughout the development of the assay.

References

1. Mike-Qingtao Huang, Zhongping (John) Lin, Naidong Weng (2013). Applications of high-resolution MS in bioanalysis. Bioanalysis 5(10):1269-1276.
2. Yuan-Qing Xia, Jim Lau, Timothy Olah, Mohammed Jemal (2011). Targeted quantitative bioanalysis in plasma using liquid chromatography/high‐resolution accurate mass spectrometry: an evaluation of global selectivity as a function of mass resolving power and extraction window, with comparison of centroid and profile modes. Rapid Communications in Mass Spectrometry 25(19):2863-2878.
3. Enhanced sensitivity for peptide quantification in a complex matrix using high-resolution LC-MS/MS. SCIEX technical note, RUO-MKT-02-13324-A.
4. High sensitivity MRM workflow for signature peptide quantification. SCIEX technical note, RUO-MKT-02-11882-A.
5. Enabling new levels of quantification. SCIEX technical note, RUO-MKT-02-11886-A.
6. Qualitative flexibility combined with quantitative power. SCIEX technical note, RUO-MKT-02-13053-A.