Counting Needles in a Haystack: Improving Sensitivity and Quantitation of Low-Level Tryptic Peptides
Carmen Fernández-Metzler, president of PharmaCadence Analytical Services, believes that quantitation should not be the limiting factor in biological studies. She took this mantra to heart when determining the basal levels of the membrane-bound isoforms of UDP-glucuronsyl-transferase (UGT), a major enzyme in the phase II-elimination of over 200 xenobiotic drugs and endogenous metabolites. Having a good understanding of UGT's role in drug metabolism, by correlating both the activity and absolute protein levels, provides a handle on effective dosages for clinical trials. But, these calculations require exact protein quantitation at very low cellular concentrations-around 2-100 pmol/mg of microsomal tissue. When tackling this issue with the UGT family of enzymes, Dr. Fernández-Metzler was presented with a challenging situation. "We didn't have a pure UGT protein standard at the time, so the difficulty was in quantitation of overexpressed recombinant protein. Additionally, peptide concentrations did not always agree with each other due to variable digestion efficiency and recovery," explained Dr. Fernández-Metzler. To address these problems, a strategic workflow was devised: isotopic dilution of signature peptides, tightly-controlled tryptic digestions, and analysis using the sensitive SCIEX QTRAP® 6500 System and highly-reproducible chromatographic separations using the Eksigent microLC System to achieve quantitation of six endogenous UGT isoforms in a complex matrix (Figure 1).
Figure 1: Overview of UGT tryptic peptide quantitation using the MIDAS™ Workflow1
After in silico digestion of the target protein, Skyline Software computed ideal representative (or signature) peptides based on charge sites, MS/MS fragment ions, and resistance to post-translational modification.
MRM lists are computed in Skyline Software for each UGT isoform using prevalidated peptides prior to transfer of the MRM lists to Analyst® Software.
Predicted signature peptides sequences were verified using microflow separation followed by analysis using the MIDAS™ workflow. Eluting peaks corresponding to MRM transitions of signature peptides triggered full-scan MS/MS. All fragments were captured and then scanned out of the trap on a UHPLC time scale, providing additional in-depth peptide structural and quantitative information, as well as selectivity.
Standard curves generated using rUGT microsomes in the presence of rat liver microsomes and stable isotope-labeled surrogate peptides were used to quantify UGT levels in human liver samples using the MIDAS workflow. A batch analysis of various UGT isoforms was enabled, capturing multi-MRM transitions for each sample.
The resulting data for each isoform was imported into MultiQuant™ Software for data processing and results reporting.
For peaks with background interference, SelexION™ technology or MRM3 assays provided an additional filtering step.
Consistent peptide release and quantitation
The first hurdle in the quantitative strategy was to develop optimal tryptic digest conditions for the protein standard (recombinant human UGT (rUGT) expressed in insect cells and prepared as microsomes), an essential step towards consistent peptide release and quantitation. The signature peptide method makes a key assumption-that the tryptic digest of the parent protein will go to completion, liberating one peptide from every instance of that protein-a uniformity that was not consistently maintained between peptides from the same isoforms. Optimizing tightly-controlled tryptic digest conditions, including the timing and concentration of all components, made the assay come together, producing effective standard curves. As Dr. Fernández-Metzler explained, "When you have reproducibility in the digest, you will have reproducibility in the mass spectrometry data." With digestion conditions optimized, Dr. Fernández-Metzler then further refined peptide separation and quantitation techniques, working closely with SCIEX application specialists. Faced with measuring numerous human liver samples, PharmaCadence needed a high-throughput application for separating low-abundance peptides, methods traditionally handled by more sedate nanoflow techniques. Combining divergent chromatographic conditions from small molecule and proteomics studies, Dr. Fernández-Metzler devised a microflow-based separation using an Eksigent microLC 200 System, which generated higher resolution data by using wider columns and faster flow rates than traditional nanoflow regimens. Sensitivity and reproducibility were not compromised under these conditions, and coupling microflow LC with the improved detector dynamic range of the QTRAP® 6500 system yielded a 3-9-fold elevation in raw signal and 2-5-fold improvement in S/N ratios of UGT peptides compared to microflow conditions on the QTRAP 5500 system (Figure 2).
Quantitative and qualitative assessment at once
Even with reproducible digestions and chromatographic separations, distinguishing low-abundance UGT peptides from the multitudes of other tryptic peptides remained a nuanced process; very small amounts of analyte needed to be selectively plucked from the sample milieu while still retaining a meaningful signal. Dr. Fernández-Metzler's team quantitated the UGT signature peptides of interest using multiple reaction monitoring (MRM) methods on the QTRAP 6500 system, screening tryptic digest peaks through two mass filters. Multiple peaks for the same MRM signal are the norm, not the exception, necessitating an additional discovery step-enhanced product ion (EPI) analysis. In this scan type, precursor ions are fragmented by true collision-induced dissociation, and then the fragments are collected, concentrated and scanned from the linear ion trap at speeds much more rapid than are possible using traditional triple quadrupole instruments. This process is called an MRM information-dependent acquisition (IDA)-based method (MIDAS™ workflow), enabling both quantitative and qualitative assessment of peaks in the same run. Confirmation of a precursor peptide's identification can then be derived from these information-rich product ions.
Figure 2: Sensitivity improvements on the 6500 versus the 5500 QTRAP® system for MRM detection of UGT tryptic peptides following microflow separation.2
An UGT tryptic peptide corresponding to MRM transition 554.3/893.5 showed 3.2-fold raw signal and 2-fold S/N improvements for fragments detected on the 6500 (blue trace) versus the 5500 QTRAP system (red trace). Similar data were obtained for three other MRM transitions:
MRM 523.3/589.3 showed a 5-fold raw signal and 3-fold S/N improvements.
MRM 523.3/718.4 showed a 9-fold raw signal and 5-fold S/N improvements.
MRM 523.3/817.4 showed a 9-fold raw signal and 4-fold S/N improvements. Each sample was acquired n=4 times on two different QTRAP 6500 systems (3.3-7.5% CV) to show reproducibility.
Figure 3: Elimination of tryptic background interferences from UGT signature peptides using MRM3.3
For y1 tryptic peptide, DIVEVLSDR, the MRM-basedchromatogram for the signature transition shows background interference peaks at 2.97 min and at 4.31 min.
The MRM3 chromatogram for DIVEVLSDR (Figure A) completely eliminated the interference peak and improved peak integration for a better %CV.
For y6 tryptic peptide, YIPCDLDFK, the MRM-based chromatogram shows an interfering signal at 4.29 min, which was 40% of the area of the parent peak (4.50 min).
The MRM3 chromatogram for YIPCDLDFK (Figure C)
The SCIEX QTRAP® 6500 System's 20-fold improvement in detector dynamic range provided the necessary sensitivity for detection of low-abundance UGT peptides; but, most importantly, the QTRAP 6500 system's fast linear ion trap scan speeds allowed rapid MS/MS analysis while still providing 10 data points across the peak for optimal quantitation. Even with the selectivity of the MIDAS™ workflow, coeluting contaminants and closely-overlapping isobaric peaks constrained optimal peak integration conditions for a number of the UGT peptides. To remove interferences from the peptide spectra without introducing additional chromatography or sample preparation steps, the QTRAP system offered an additional advantage-the MRM3 scan. During the EPI scan, when the fragments of the precursor peptide are assembled in the trap, a selected ion is further isolated and fragmented. This secondary fragmentation process produced additional ions for further structural analysis and high-resolution quantitation of UGT signature peptides, removing background interferences (Figure 3). "After MRM3, only one peak remained in the chromatogram, and it was really easy to process the data, as the automated integration routines worked more reliably with the MRM3 data," noted Dr. Fernández-Metzler, who added that the "MRM3 method is very clean, very selective, but requires a lot more work to set up."
Reduced matrix interferences
If method development time is limited or if the second generation fragments are either not specific enough or are too low, differential ion mobility separation (DMS) based on SelexION™ technology can provide an additional degree of selectivity. This technique exploits an ion's mobility through a set of plates with high and low energy fields applied to quickly resolve isobaric species and single and multiple charge state interferences on a timescale compatible with UHPLC and MRM acquisition. For certain UGT signature peptides, these background interferences from overlapping peaks were problematic, complicating peak integration. To acquire a clean spectrum, interfering ions were essentially tuned out of the instrument using DMS, significantly improving the MS/MS spectrum for UGT-specific peptides that were previously muddled by overlapping peaks (Figure 4). Furthermore, SelexION technology reduced the matrix interferences, effectively boosting the signal to noise and sensitivity of the UGT assays.
Figure 4: Elimination of co-eluting multiply-charged interferences from UGT signature peptides using SelexION™ Technology.4
An enhanced product ion (EPI) scan of UGT-2B7 tryptic peptide, IEIYPTSLTK (fragments labeled in pink) was captured on a QTRAP 6500 system, which enhanced sensitivity without loss of qualitative sequence information.
An EPI scan with Q0 trapping of IEIYPTSLTK shows that a co-eluting, interference (fragments circled in yellow) produced a chimeric spectrum.
The interfering peaks were removed by DMS filtering using SelexION technology, yielding a clean MS/MS spectrum for quantitation.
Overlays of the tryptic peptide chromatograms for UGT-peptide, IEIYPTSLTK (pink) and the interference (blue) are shown.
Figure 5: Concentrations of UGT isoforms in 10 individual human livers. Quantitation of each human isoform in liver samples was conducted as described1 using rUGTinfused rat liver microsomes and stable-label signature peptides to create standard curves for the assessment of UGT isoform levels in human liver samples.
The range of concentrations obtained for four UGT isoforms 1A1, 1A6, 1A9 and 2B7 from ten separate human liver samples are graphically displayed here, visually representing the quantitative variability.
A table summary of mean concentrations and ranges of all six UGT isoforms in 50 individual human liver microsomes.
After perfecting the experimental process, Dr. Fernández-Metzler obtained highly accurate quantitative data on the basal levels of UGT isoforms by creating high quality calibration curves under a stringent assay optimization process where tightly-controlled digestions gave rise to extremely reproducible conditions1 (Table 1). When constructing the standard curves, occasionally the concentrations of different peptides from the same proteins did not correspond well because of poor signal to noise. After processing the results in MultiQuant™ Software, even peptides such as UGT 1A1 with high background produced good quality standard curves with accuracies within 15% CV and showed precision within 15-20% CV, an appropriate range for a discovery assay. Other UGT isoforms produced three closely corresponding peptide concentrations (such as those from isoform 2B7) yielding calibration curves with a % CV less than 10%. In the end, was all this work worth the effort? "Yes!" confirmed Dr. Fernández-Metzler. "This study will help design a clinical trial that will hopefully lead to a better medicine one day."
Fernández-Metzler C. (August 2013) "Peptide Quantification on the QTRAP® Mass Spectrometers with MicroflowLC: Bridging the Best of Small Molecule and Proteomic Analysis." SCIEX Mass Spec Webinar Series.
UGT Family of Enzymes: Quantification of Tryptic Peptides. Part 1 of 3: The QTRAP® 6500 Platform and MicroLC Provide the Combination of Sensitivity, Specificity and Robustness for the Quantitation of UGT Enzymes." (White Paper) SCIEX. Accessed November 2013.
UGT Family of Enzymes: Quantification of Tryptic Peptides. Part 2 of 3: Accelerating MRM3 Workflows on QTRAP® 6500 System for Enhanced Selectivity in Complex Matrices like Tryptic Digests." (White Paper) SCIEX. Accessed November 2013
UGT Family of Enzymes: Quantification of Tryptic Peptides. Part 3 of 3: Using SelexION™ Technology for Additional Selectivity by Separating Multiple Charge State Ions in Tryptic Digests." (White Paper) SCIEX. Accessed November 2013.