Determination of Collision Cross-Sections of Protein Ions in an Orbitrap Mass Analyzer.

We demonstrate a method for determining the collision cross-sections (CCSs) of protein ions based on the decay rate of the time-domain transient signal from an Orbitrap mass analyzer. Multiply charged ions of ubiquitin, cytochrome c, and myoglobin were generated by electrospray ionization of both denaturing solutions and ones with high salt content to preserve native-like structures. A linear relationship between the pressure in the Orbitrap analyzer and the transient decay rate was established and used to demonstrate that the signal decay is primarily due to ion-neutral collisions for protein ions across the entire working pressure range of the instrument. The CCSs measured in this study were compared with previously published CCS values measured by ion mobility mass spectrometry (IMS), and results from the two methods were found to differ by less than 7% for all charge states known to adopt single gas-phase conformations.

Introduction decays in ICR mass analyzers has long been recognized, 34-37 and several methods have recently 1 capitalized on this relationship. 38 For example, CCS measurements of crown ethers and amino 2 acids have been made based on the frequency domain spectral linewidths obtained under elevated 3 pressure conditions; a method dubbed "CRAFTI" (cross-sectional areas by Fourier transform ion 4 cyclotron resonance). 39-42 A similar technique, performed using a custom-built FT-ICR featuring 5 a more powerful magnet (9.4 T instead of 4.7 T), was used to measure CCS of biomolecules up 6 to the size of ubiquitin (8.5 kDa), which to our knowledge is the largest analyte measured by this 7 type of method so far. 43 This approach directly measured the decay rate of the time-domain 8 transient signal by performing a series of digital low-pass filtering and down-sampling steps 9 followed by fitting an exponential decay function to the resulting decay profile. 43,44 10 More recently, other mass spectrometers have been utilized for CCS measurements. For 11 instance, CCSs of several tetraalkylammonium cations and small peptides were measured from 12 spectral linewidths obtained on a home-built FT electrostatic linear ion trap (FT-ELIT) with 13 impressive accuracy. 45 In another study, a charge-detection mass spectrometer (CDMS) was used 14 to measure CCSs of proteins as large as bovine serum albumin. 46 The transient signal decay rate 15 in the Orbitrap mass analyzer was shown to be related to ion CCS, although without the ability to 16 directly measure the pressure of background gas in the analyzer chamber, only relative 17 measurements could be made. 47 Here, we describe a method inspired by this previous study to

30
The proteins were used as supplied for the denaturing solutions, whereas the proteins were purified 1 using P-6 Bio-Spin columns (Bio-Rad Laboratories, Hercules, CA) for the non-denaturing 2 solutions.

4
All experiments were performed on a Thermo Fisher Scientific™ Orbitrap Elite™ mass 5 spectrometer (Bremen, Germany) which was modified to incorporate an electronically variable 6 pressure regulator allowing precise control over the pressure of nitrogen collision gas in the HCD 7 ion trapping cell 48 (this feature is now standard in most Orbitrap mass spectrometers). Since N2 8 gas leaking from the HCD cell is the primary source of background gas in the Orbitrap analyzer  To ensure that a consistent number of isotope peaks with high signal to noise were selected 28 across the full range of masses and charge states that were analyzed, a window was defined for 29 each charge state that encompassed all FT bins that contained the seven most intense isotope peaks 1 in the isolation spectrum, and these peaks were shifted to near zero frequency. An array of complex 2 zeros was appended to the selected peaks so that the data contained a total of 2 13 FT bins. While 3 this method relies on access to pre-FT time domain data that requires a special license, it does not 4 require any major instrument modifications aside from the ability to regulate the flow of nitrogen   The background gas in the ultra-high vacuum chamber that houses the Orbitrap mass 9 analyzer is primarily composed of N2 flowing from the HCD cell. While this chamber is equipped 10 with a cold ion gauge to monitor pressure, the pressure inside the Orbitrap analyzer is expected to 11 be somewhat higher owing to the dynamics of collision gas flow from the HCD cell to the Orbitrap 12 chamber as illustrated in Scheme S1. 47 The pressure gauge readings are also biased by a smaller

17
To obtain a more accurate estimate of pressure for calculation of CCS, the 9+ charge state 18 of ubiquitin, which is reported to adopt a single gas-phase conformation, 28,29 was selected as the 19 calibration standard. The decay rate of this ion was measured each day that experiments were 20 performed and each time the pressure settings were changed, and was used to calculate a value for 21 the gas molecule density n that was used for all subsequent CCS calculations. The relationship 22 between the gauge reading and the estimated pressure in the Orbitrap analyzer was consistent day-23 to-day and across the full range of gas in-flow rates used in the experiments. A plot of this 24 relationship is shown in Figure S1. The calibration coefficients were found to be A = 2.15 and B 25 = 0.11 × 10 -10 torr.  varied over the range of 0.3 × 10 -10 to 0.6 × 10 -10 torr in the Orbitrap chamber (as measured by the 2 cold ion gauge), which roughly corresponds to a calculated pressure range of 0.5 × 10 -10 to 1.2 × 3 10 -10 torr in the Orbitrap analyzer (see the experimental section, Figure S1 and Scheme S1 for 4 pressure calibration details). If collisions with neutrals are the primary source of signal decay, a 5 linear relationship between pressure and decay rate that trends toward zero decay at zero pressure 6 would be expected. Any deviation from the linear trend at lower pressures would suggest that 7 other factors such as space charge or field imperfections contribute to non-collisional loss of ions 8 or ion coherence. Figure 1 shows the relationship between gas pressure and the experimentally- summarized as a function of charge state in Figure 3 and Table S1. As expected, CCS consistently 4 increases with increasing charge state for all three proteins. Interestingly, while the relationship 5 between charge state and CCS is nearly linear in the higher charge state range, the slope of each 6 data series appears to be steeper at lower charge states, particularly for those generated from native-7 like solution conditions. While the effect is subtle, it is consistent with other reports that suggest 8 that many elements of protein tertiary structure are preserved in the gas phase for very low charge  were available, they were adjusted as described in Figure S2 to account for the consistently larger 20 obtained from measuring CCS in N2 as opposed to He.
21 Figure 4 and Table 1 provide a comparison of CCSs measured using the FT transient indicate that the assumption that a single collision is always sufficient to remove an ion from the 30 coherently orbiting packet may not be uniformly valid for larger proteins that possess higher 31 kinetic energies and have more pathways to disperse internal energy acquired through a collision.

1
The upper mass limit of this method and the effects of using heavier or lighter collision gasses, 2 which will change the center of mass collision energy and potentially affect this limit, will be 3 investigated in a future study.

5
The coefficient of determination (R 2 ) for each fitted trendline is > 0.99, and the y-intercepts are <  Table 1). Each point 4 represents the average of 5 replicate measurements, and error bars represent ± 1 standard 5 deviation.