Mobility Enhanced Top-down Sequencing of Proteins on a TimsTOF
Loading...
Authors
Graham, Katherine
Issue Date
2024
Type
Dissertation
Language
en_US
Keywords
Intact Protein Ions , Ion Mobility , Mass Spectrometry , Top-down Proteomics , Trapped Ion Mobility Spectrometry
Alternative Title
Abstract
Proteins are the driving force behind many biological processes, and thus, elucidating protein expression, regulation, and interactions is vital to understanding the mechanisms behind the function and disfunction of organismal systems. Mass spectrometry (MS)-based proteomics is one of the predominate methods for quantitative and qualitative analyses of these protein systems. In the most utilized MS-based proteomic workflows, “bottom-up", proteins undergo proteolysis by an enzyme such as trypsin into peptides as most modern-liquid chromatography separations and MS technologies were developed to analyze small peptides. This technique is sensitive, easily implemented, and supported by robust bioinformatic platforms. However, highly acidic or hydrophobic regions are under sampled and inaccurately quantified due to instrumental limitations. Also, any post translational modification (PTM) cooperativity not found within a single peptide will be erased. These limitations have driven the field to increasingly focus on gas-phase sequencing of intact protein ions. While this “top-down" approach solves previously described limitations, gas-phase sequencing of intact proteins is limited by inefficient fragmentation efficiency and formation of complex convoluted mass spectra. Site-specific localization of PTMs requires robust gas-phase sequencing of protein ions via tandem MS (MS/MS). However, many of these techniques were developed for low mass peptides and have poor efficiency when applied to protein ions. This low fragmentation efficiency decreases the spatial resolution of PTM assignments and prohibits the differentiation of isomeric protein species with adjacent PTM sites. Fragmentation and sequencing of protein ions can also result in the formation of product ions with overlapping isotopic distributions, generating convoluted and difficult to analyze mass spectra. Ion mobility spectrometry (IMS) has shown promise in deconvoluting these spectra. However, few instrumental platforms are capable of ion activation prior to mobility analysis. To overcome these challenges, we developed a peptide and protein activation technique that is performed prior to mobility separation on a Bruker trapped ion mobility spectrometry time-of-flight (timsTOF). This collision-induced dissociation trapped ion mobility spectrometry (CIDtims), is achieved by decreasing the tunnel-in pressure and increasing the DC electric field �"6. First, this method is developed with peptides, melittin, and ubiquitin. As peptide or protein ions are transferred, they undergo collisional activation resulting in the robust formation of a-, b-, and y-type ions. The generated fragment ions are then dispersed along the mobility dimension providing a 2-dimensional data display of generated fragment ions. This has helped deconvolute mass spectra by allowing fragment ions with near isobaric m/z values to detangled into mobility-correlated mass spectra where they are no longer overlapped. We also demonstrate the ability to increase sequence coverage of melittin and ubiquitin by using the downstream quadrupole and collision cell for a pseudo-MS3 workflow to increase sequence coverage by isolating and fragmenting a CIDtims generated fragment ion. This workflow results in ~50% increase in sequence coverage compared to fragmentation within a single MS2 analysis. Next, this method is further developed and tuned for the top-down sequencing of protein ions ranging in size from 8.7 kDa to 66 kDa. Tunnel-in pressure is optimized, and ion populations are controlled to limit space-charge effects in the TIMS device. The sequence coverage of CIDtims is compared to the two other modes of collisional activation on the Bruker timsTOF, in-source collision-induced dissociation (isCID) and collision-induced dissociation (CID). After optimizing the tunnel-in pressure, we significantly improve CIDtims sequence coverage by minimizing space-charge effects in the TIMS. Decreasing space-charge effects results in ~24.1% increase in sequence coverage for proteins. We show that CIDtims is capable of generating comparable sequence coverage compared to in-source CID and CID. This results in 84%, 52%, 24.6%, 36%, and 6.1% sequence coverage for ubiquitin, cytochrome C, β-lactoglobulin, carbonic anhydrase, and BSA, respectively. We also demonstrate the ability to use the additional mobility dimension to separate overlapping product ions to increase sequence coverage compared to in-source CID. Following the characterization of this method for intact proteins, CIDtims is improved to allow for annotation of internal fragment ions. Internal fragment ions occur when an amide backbone is fragmented in more than one location, resulting in a fragment ion that is not anchored to the n-terminus or c-terminus. First, the method is improved to handle dynamic ion populations and the bioinformatics used to analyze the mobility dimension are optimized. Next, we transfer CIDtims to a timsTOF pro2 which improves the mass resolving power allowing for a more stringent mass tolerance for fragment ion assignments. Internal fragment ions are then assigned with a 5 ppm mass tolerance cutoff and an isotopic distribution score cutoff is set at 95%. We confirm our scoring metrics by using the downstream quadrupole and collision cell to sequence three internal ions. Finally, we show that IMS provides increased signal-to-noise ratios to improve isotopic fit scores to improve the assignment of internal ions. Including internal ions results in 89.3%, 91.3%, 91.4%, 72.2%, 69%, and 43.4% sequence coverage for ubiquitin, cytochrome C, lysozyme, β-lactoglobulin, carbonic anhydrase, and enolase, respectively. Overall, this provides a 41% average increase in sequence coverage. Collision-based fragmentation of protein ions typically yield poor sequence coverage as energy will be shuttled towards the most kinetically labile bonds which reduces the formation of sequence informative product ions. In chapter four, we overcome this limitation by developing a mobility enhanced pseudo-MS3 method. This improves sequence coverage isolating and fragmenting CIDtims generated fragment ions to increase the number of sequence informative fragment ions at less kinetically labile bonds. We explore this method using both auto-MS/MS on the timsTOF and parallel accumulation serial elution fragmentation (PASEF) on the timsTOF pro2. This generates 97.33%, 98.06%, 92.19%, 63.35%, 54.65%, and 36.2% sequence coverage for ubiquitin, cytochrome C, lysozyme, β-lactoglobulin, carbonic anhydrase, and enolase, respectively. We then compare our results to the alternative top-down fragmentation methods AI-ETD and UVPD. CIDtims produces 94.8% of sequence coverage generated by UVPD and 92% of the sequence coverage generated by UVDP. While these methods still outperform CIDtims, the use of a two-step collisional activation reduces false discovery rate that is commonly increased by the mixed fragmentation pathways that result from alternative methods. Furthermore, CIDtims can be performed at timescales amenable to coupling with liquid chromatographic separations.