Increasing Analytical Separation and Duty Cycle with ...faculty.fiu.edu/~fernandf/pubs/84_inc-an-sep.pdfIncreasing Analytical Separation and Duty Cycle with Nonlinear Analytical Mobility - [PDF Document] (2024)

Increasing Analytical Separation and Duty Cycle with NonlinearAnalytical Mobility Scan Functions in TIMS-FT-ICR MSPaolo Benigni,† Jacob Porter,† Mark. E. Ridgeway,‡ Melvin. A. Park,‡ and Francisco Fernandez-Lima*,†,§

†Department of Chemistry and Biochemistry, Florida International University, Miami, Florida 33199, United States‡Bruker Daltonics Inc., Billerica, Massachusetts 01821, United States§Biomolecular Sciences Institute, Miami, Florida 33199, United States

*S Supporting Information

ABSTRACT: In this work, nonlinear, stepping analyticalmobility scan functions are implemented to increase theanalytical separation and duty cycle during tandem TrappedIon Mobility Spectrometry and FT-ICR MS operation. Thedifferences between linear and stepping scan functions aredescribed based on length of analysis, mobility scan rate, signal-to-noise, and mobility resolving power. Results showed that forthe linear mobility scan function only a small fraction of thescan is sampled, resulting in the lowest duty cycle 0.5% andlongest experiment times. Implementing nonlinear targetedscan functions for analysis of known mobilities resulted inincreased duty cycle (0.85%) and resolving powers (R up to 300) with a 6-fold reduction in time from 30 to 5 min. For broadrange characterization, a nonlinear mobility stepping scan function provided the best sensitivity, resolving power, duty cycle(4%), and points per peak. The applicability of nonlinear mobility scan functions for the analysis of complex mixtures isillustrated for the case of a direct infusion of a MCF-7 breast cancer cell digest, where isobaric peptides (e.g., DFTPAELR andTTILQSTGK) were separated in the mobility domain (RIMS: 110) and identified based on their CCS, accurate mass (RMS: 550k),and tandem MS using IRMPD in the ICR cell.

A common technical challenge faced when couplingtraditional preseparation technique (e.g., gas and liquid

chromatography) with FT-ICR MS is that FT methods rely onmeasuring the free induction decay of ions over a period ofseconds per scan in order to achieve ultrahigh massresolution.1,2 The ultrahigh resolving power of FT-ICR MSinstrumentation has unique advantages in the analysis ofenvironmental,3 biological,4 and petroleum mixtures5 comparedto non-FT methods. However, optimal operation of the FT-ICR MS (e.g., sensitivity and resolution) requires a well-definedion population in order to minimize coalescence of the ionsignal.6−8 Several groups have shown the analytical advantagesof coupling postionization separation techniques like ionmobility spectrometry (IMS) to FT-ICR MS. For example,the coupling of scanning IMS techniques (e.g., FAIMS)9 to FT-ICR MS has been advantageous for the analysis of poly-(ethylene glycol),10 proteins,11−13 and glycans.14 The couplingof time-dispersive IMS techniques (e.g., drift tube IMS) hasalso been shown for the separation of phosphopeptides15 andto study ion−molecule reaction chemistry.16 Alternatively, wehave shown that time-independent IMS analyzers, such asTrapped Ion Mobility Spectrometry (TIMS),17−19 can beeffectively coupled to FT-ICR MS in selective accumulation,20

oversampling,21 and gated22 modes of operation. These modesof operation have enabled mobility resolving powers up to 300and mass resolution up to 1 million, giving the unique

advantages when coupled to FT-ICR MS for the analysis ofcomplex mixtures.23−26

In this work, for the first time, nonlinear, stepping analyticalmobility scan functions are implemented in gated TIMS toincrease the analytical separation and duty cycle during tandemTIMS and FT-ICR MS operation. Following the concept ofnonlinear mobility scan steps used in TIMS-TOF MS,27 thedifferences between linear and nonlinear stepping scanfunctions are described based on length of analysis, mobilityscan rate, signal-to-noise, and mobility resolving power. Threemodes of operations (i.e., linear, nonlinear targeted, andnonlinear stepping) are evaluated and their applicationillustrated for the mobility separation of a binary mixture ofisobaric peptides in a complex mixture, as well as the potentialfor mobility selected tandem MS experiments.

■ EXPERIMENTAL SECTIONMaterials and Reagents. A Tuning Mix calibration

standard (G24221A) was obtained from Agilent Technologies(Santa Clara, CA) and used as received. Peptides DFTPAELRand TTILQSTGK were added at 10 μM to an MCF-7 digest26

at 2.68 μg/μL and 10% of 0.1% methanol/formic acid.

Received: October 2, 2017Accepted: January 26, 2018Published: January 27, 2018

Technical Note

pubs.acs.org/acCite This: Anal. Chem. 2018, 90, 2446−2450

© 2018 American Chemical Society 2446 DOI: 10.1021/acs.analchem.7b04053Anal. Chem. 2018, 90, 2446−2450

TIMS and FT-ICR MS Analysis. TIMS separation wasperformed using nitrogen as a bath gas at about 300 K, P1 =2.2−2.47 and P2 = 0.9 mbar, and a constant Vout = 50 V and rf(840 kHz and 240−260 Vpp). The TIMS-FT-ICR MSexperiments were acquired in chromatography mode, whereeach MS scan was a single 1−8 Megaword (0.5−4 s) transient,processed using a sine-squared apodization followed by fast-Fourier transform (FFT) in magnitude mode with anexperimental resolving power of 50,000−550,000 at m/z 400.The individual scans are converted to mobility using thereported mobilities of hexakis fluorinated alkyl phosphazinesfound in the tuning mix mixture m/z 622 K0 = 1.008, m/z 922K0 = 0.826, m/z 1222 K0 = 0.711, m/z 1522 K0 = 0.632 cm2

V−1 s−1) more details can be found in the SupportingInformation.

■ RESULTS AND DISCUSSIONIn TIMS, ions are spatially resolved along the analyzer axis, in atime independent process where ions reach an equilibriumbetween the drag force caused by collisions with a moving gasand the increasing electric field as a function of theirmobilities.17−19 The electric field is established by a voltagegradient between the front (Vramp) and the end (Vout) of theanalyzer section, and stepwise reduction of Vramp causes the ionsto move toward the end of the analyzer in discrete mobilityresolved packets. The TIMS experiment, which typically lastsbetween 100 and 500 ms, is traditionally coupled with a Timeof Flight (ToF) mass analyzer, which has an experiment time of100−150 μs, and is capable of quickly sampling the mobilityseparated ions as they elute from the TIMS analyzer. Whencombined with FT-ICR MS, which has experiment times on theorder of seconds, the challenge becomes achieving ananalytically significant number of points across a mobilitypeak in a reasonable amount of time. When coupled to FT-ICRMS, the mobility range is sampled using discrete gating pulses,therefore, the number of points across the peak is dependenton the size of the elution steps. In a typical linear scan gatedTIMS experiment, the mobility range, mobility resolution,analysis time, and number of points across a mobility peak aredefined by scan rate, Vramp range and gate width (see Figure 1).For example, for a Vramp = 200 V, tramp = 100 ms, and a 1 msgate, the Tune Mix mobility peaks are characterized by just 3−4points, from baseline, (i.e., 3−4 FT-ICR MS scans), which istoo low for a true analytical characterization of an IMS peakprofile (i.e., quantitative analytical characterization of a peak istypically defined as 15−20 points per peak). By reducing thegate width by half (1.0−0.5 ms), the number of points acrossthe mobility peak increases from 3−4 to 6−8 peaks at the costof doubling the analysis time (6−12 min). Another strategy toincrease the number of points per peak, is to decrease the scanrate; for example, by doubling the ramp time from 100 to 200ms over Vramp = 200 V with a gate width of 1 ms, the number ofpoints across the mobility peak increases from 3−4 to 6−8peaks in 15 min, with slightly higher mobility resolution (Figure1e vs d). If the gate width is then decreased from 1.0 to 0.5 ms,10−12 points, from baseline, across the mobility are observedwith higher resolving power (R = 130−160). While reducingthe scan rate and gate width are effective methods for betterpeak profiles and higher mobility resolving power, they come ata cost of increasing the experiment time and reducing the dutycycle (in this work duty cycle is calculated as the ratio of gatewidth to the total IMS time). For example, in the describedTIMS-FT-ICR MS experiment a duty cycle of 0.5% is obtained

for the tramp = 200 ms, ΔVramp = 200 V, and 1.0 ms gate pulsewidth. That is, in this approach, the majority of the analysistime goes to scanning the Vramp voltage, with only a smallfraction of the scan steps allowed to be transmitted andaccumulated in the FT-ICR MS collision cell (defined by thegate width).An alternative to this approach is to utilize nonlinear targeted

mobility scan functions (nonlinear ΔVramp) where the scan rateis adjusted based on the mobility regions of interest usinganalytical (slow) and nonanalytical (fast) scan rates (see Figure2). This procedure is like a targeted quadrupole MS experimentwhere only specific masses are measured, and the others areskipped. For example, when performing a targeted analysis forions at specific mobilities, the slower analytical portion of theramp are centered on the targeted mobility, which is knownbeforehand. This approach not only reduces the analysis timebut has the potential for even slower scan rates potentiallyleading to higher resolving power. For example, using anonlinear scan rate function the Tune Mix analysis is reducedfrom 30 to 5 min using the same tramp and gate width andΔVramp range with similar or higher resolving power (see Figure2b,c). Note that the ratio between analytical and nonanalyticalcan be varied as a function of the number of mobility peaks ofinterest without increasing the total experiment time. Underthese conditions, the duty cycle is improved to 0.85%, butremains low since a narrow gate width is required to guaranteeenough points across the mobility peak of interest.

Figure 1. Typical linear scan TIMS profiles for Tuning Mix standard.(a) Scan over (ΔVramp = 200 V) and tramp of 100 and 200 ms. (b) Gateprofile per step as a function of the gate time, letting subsequentportions of the ramp into the FT-ICR MS. (c) IMS profiles for tramp =100 ms and 1 ms gate pulse, (d) IMS profiles tramp = 100 and 0.5 msgate pulse, (e) IMS profiles tramp = 200 and 1 ms gate pulse, and (f)IMS profiles tramp = 200 and 0.5 ms gate pulse. FT-ICR MS spectrawere collected at 1 MW with 0.5 s transient and RMS ∼ 50000 at m/z400.

Analytical Chemistry Technical Note

DOI: 10.1021/acs.analchem.7b04053Anal. Chem. 2018, 90, 2446−2450

2447

In order to better correlate the timing of the analytical step inthe TIMS profile with the gate width, a nonlinear stepping scanfunction can be designed such that only ions that are mobilityseparated are gated into the FT-ICR MS collision cell (seeFigure 3). That is, a profile composed of a single analytical step

can be used along with a gate at a fixed position. Subsequentmobilities are selected and separated by incrementing thevoltages of the analytical step, changing the mobility region thatundergoes high resolution separation (see Figure 3a). Thisapproach does not require prior knowledge of an ion’s mobility,as in the previous approach, but allows the entire mobilityrange to be separated with the highest resolving power (i.e.,using the slowest ramp speed across the analytical step),

without exponential increases in time. For example, thisnonlinear scan function with mobility stepping results in highresolving powers (up to 160), 26−30 points characterizing thepeaks in the IMS profile, and better duty cycle (∼4%; Figure3d). Another advantage of this approach is that a prioriknowledge of the scan step of the peaks of interest is notnecessary to define the analytical scan, since the full mobilityrange is acquired using slow scan rates (high mobility resolvingpower). A summary of the presented methodologies isprovided in the Supporting InformationThe proposed methodology was applied for the targeted

analysis of isobaric peptides from a MCF-7 breast cancer digestsample: DFTPAELR (m/z 948.47852) and TTILQSTGK (m/z948.53603). Direct infusion of the MCF-7 breast cancer digestsample results in a highly complex MS spectrum. Takingadvantage of the ultrahigh mass resolution of the FT-ICR MS,the signals corresponding to the DFTPAELR (m/z 948.47852)and TTILQSTGK (m/z 948.53603) can be easily separatedwith a resolving power of RMS ∼ 225000 at m/z 948. Moreover,the addition of the mobility dimension is particularly useful inanalyzing these peptides because it enables complementaryconfirmations based on (i) their collision cross section of thepeptides and (ii) tandem MS without the need of massselection of the parent ion in the mass domain since they areseparated in the mobility domain.26

The TIMS analysis using a linear scan function (tramp = 100ms and gate width of 1 ms) lasted 1 h and provided a detailedcharacterization of the MCF-7 breast cancer digest sample andaccurate CCS values of the two isobaric peptides (<2% error),but they were not resolved in the mobility domain (resolvingpower of 30). When nonlinear scan functions are utilized, theanalysis time is reduced to 10 min (Figure 4c) and 35 min (4d),and the two peptides were resolved in the mobility domain(resolving power exceeding 100). Because the nonlinear scanfunction with mobility stepping has a greater duty cycle thanthe linear or multisegments nonlinear functions, this mode canbe utilized to increase the analysis sensitivity with multiple fillsin the collision cells of mobility selected ion packets. TandemMS of mobility selected ion packets using infrared multiphoton

Figure 2. Comparison of the IMS profiles using (a) linear scan TIMS (ΔVramp = 200 V and tramp = 200 ms) and nonlinear scan TIMS with analyticalramps of (b) ΔVramp = 10 V and (c) ΔVramp = 4 V over 15 ms. For all experiments, a 0.5 ms gate pulse was used. (Top; scanning ramp profiles areshown in blue and bottom) IMS profiles are shown in CCS. FT-ICR MS spectra were collected at 1 MW with 0.5 s transient and RMS ∼ 50000 at m/z 400.

Figure 3. (a) Typical nonlinear stepping mobility scan TIMSexperiment. The portions of the ramp highlighted in red are notsampled, and with each subsequent step the voltages of the analyticalramp are incremented, as shown by the black arrows, maintaining thesame ΔVramp. Typical Tune Mix results are shown for an ΔVramp = 3 Vscanned over (b) 3, (c) 6, and (d) 9 ms. FT-ICR MS spectra werecollected at 1 MW with 0.5s transient and RMS ∼ 50000 at m/z 400.

Analytical Chemistry Technical Note

DOI: 10.1021/acs.analchem.7b04053Anal. Chem. 2018, 90, 2446−2450

2448

dissociation (IRMPD) in the ICR cell permitted furtherconfirmation of the identities of DFTPAELR andTTILQSTGK with 100% sequence coverage, respectively (seetandem MS spectra in Figure S1). This example illustrates thepotential of using nonlinear scan mobility functions forachieving high resolution mobility separation, better analyticalprofile (number of points across the mobility peak), accurateCCS measurements and increase duty cycle in tandem withultrahigh MS and tandem MS/MS separation. We anticipatethat this powerful and flexible instrumental platform will findwide application in the analysis of complex mixtures inbiomedical, environmental, and forensic applications.

■ CONCLUSIONS

Several modes of operation of a TIMS analyzer in tandem withFT-ICR MS were discussed. Differences in analysis time,analytical profiles (number of points across the mobility peaks)and resolving power were shown between linear and two formsof nonlinear mobility scan functions: targeted and stepping.Results showed differences in duty cycles from up to 1% withthe linear TIMS to 5% using nonlinear functions with variablestepping TIMS. Compared to linear TIMS, the nonlinearfunctions are able to achieve greater resolving powers bydecreasing the scan rate of the analytical step without increasingthe total analysis time. The advantage of the proposedtechnology was successfully evaluated in a complex MCF-7breast cancer digest sample for the separation and identificationof isobaric peptides with mobility resolving powers exceeding110, accurate CCS measurements (<1% error) and tandem MSusing IRMPD in the ICR cell for sequence identification.

■ ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acs.anal-chem.7b04053.

Additional experimental information, Table S1 summa-rizes the experimental results, and Figure S1 shows theIRMPD spectra of peptides DFTPAELR andTTILQSTGK (PDF).

■ AUTHOR INFORMATIONCorresponding Author*Phone: 305-348-2037. Fax: 305-348-3772. E-mail: [emailprotected] Fernandez-Lima: 0000-0002-1283-4390NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThis work was supported by the National Science FoundationDivision of Chemistry, under CAREER Award CHE-1654274,with cofunding from the Division of Molecular and CellularBiosciences to FFL. P.B. acknowledges the fellowship providedby the National Science Foundation Award (HRD-1547798) toFlorida International University as part of the Centers forResearch Excellence in Science and Technology (CREST)Program. This is contribution number 857 from the SoutheastEnvironmental Research Center in the Institute of Water &Environment at Florida International University. The authorswould like to thank Dr. Steven Van Orden and Dr. MichaelEasterling from Bruker and the Advance Mass SpectrometryFacility (AMSF-FIU).

■ REFERENCES(1) Nikolaev, E. N.; Boldin, I. A.; Jertz, R.; Baykut, G. J. Am. Soc. MassSpectrom. 2011, 22, 1125−1133.(2) Benigni, P.; DeBord, J. D.; Thompson, C. J.; Gardinali, P.;Fernandez-Lima, F. Energy Fuels 2016, 30, 196−203.(3) Hertkorn, N.; Harir, M.; Cawley, K.; Schmitt-Kopplin, P.; Jaffe,́ R.Biogeosciences 2016, 13, 2257−2277.(4) Xie, Y.; Zhang, J.; Yin, S.; Loo, J. A. J. Am. Chem. Soc. 2006, 128,14432−14433.(5) Klein, G. C.; Kim, S.; Rodgers, R. P.; Marshall, A. G.; Yen, A.Energy Fuels 2006, 20, 1973−1979.(6) Comisarow, M. B.; Marshall, A. G. Chem. Phys. Lett. 1974, 25,282−283.(7) Marshall, A. G.; Hendrickson, C. L. Int. J. Mass Spectrom. 2002,215, 59−75.(8) Qi, Y.; O’Connor, P. B. Mass Spectrom. Rev. 2014, 33, 333−352.(9) May, J. C.; McLean, J. A. Anal. Chem. 2015, 87, 1422−1436.(10) Robinson, E. W.; Garcia, D. E.; Leib, R. D.; Williams, E. R. Anal.Chem. 2006, 78, 2190−2198.(11) Robinson, E. W.; Leib, R. D.; Williams, E. R. J. Am. Soc. MassSpectrom. 2006, 17, 1469−1479.(12) Robinson, E. W.; Sellon, R. E.; Williams, E. R. Int. J. MassSpectrom. 2007, 259, 87−95.(13) Robinson, E. W.; Williams, E. R. J. Am. Soc. Mass Spectrom.2005, 16, 1427−1437.(14) Kailemia, M. J.; Park, M.; Kaplan, D. A.; Venot, A.; Boons, G. J.;Li, L.; Linhardt, R. J.; Amster, I. J. J. Am. Soc. Mass Spectrom. 2014, 25,258−268.(15) Tang, X.; Bruce, J. E.; Hill, H. H., Jr. Rapid Commun. MassSpectrom. 2007, 21, 1115−1122.

Figure 4. Analysis of isobaric peptides from a MCF-7 breast cancerdigest sample. (a) 2D-TIMS-FT-ICR MS contour plot acquired usinga nonlinear stepping mobility scan function. (b) Typical MS spectrumfor the MCF-7 digest with inset for the targeted peptide molecularions. Typical mobility profiles for DFTPAELR (black squares and bluefit) and TTILQSTGK (red circles and yellow fit) using nonlineartargeted (c) and stepping (d) mobility scan functions. FT-ICR MSspectra were collected at 8 MW with 4s transient and RMS ∼ 550000 atm/z 400.

Analytical Chemistry Technical Note

DOI: 10.1021/acs.analchem.7b04053Anal. Chem. 2018, 90, 2446−2450

2449

(16) Bluhm, B. K.; Gillig, K. J.; Russell, D. H. Rev. Sci. Instrum. 2000,71, 4078.(17) Fernandez-Lima, F. A.; Kaplan, D. A.; Park, M. A. Rev. Sci.Instrum. 2011, 82, 126106.(18) Fernandez-Lima, F.; Kaplan, D. A.; Suetering, J.; Park, M. A. Int.J. Ion Mobility Spectrom. 2011, 14, 93−98.(19) Hernandez, D. R.; Debord, J. D.; Ridgeway, M. E.; Kaplan, D.A.; Park, M. A.; Fernandez-Lima, F. Analyst 2014, 139, 1913−1921.(20) Benigni, P.; Thompson, C. J.; Ridgeway, M. E.; Park, M. A.;Fernandez-Lima, F. Anal. Chem. 2015, 87, 4321−4325.(21) Benigni, P.; Fernandez-Lima, F. Anal. Chem. 2016, 88, 7404−7412.(22) Ridgeway, M. E.; Wolff, J. J.; Silveira, J. A.; Lin, C.; Costello, C.E.; Park, M. A. Int. J. Ion Mobility Spectrom. 2016, 19, 77−85.(23) Benigni, P.; Sandoval, K.; Thompson, C. J.; Ridgeway, M. E.;Park, M. A.; Gardinali, P.; Fernandez-Lima, F. Environ. Sci. Technol.2017, 51, 5978−5988.(24) Benigni, P.; Marin, R.; Sandoval, K.; Gardinali, P.; Fernandez-Lima, F. J. Visualized Exp. 2017, 121, e55352.(25) Benigni, P.; Bravo, C.; Quirke, J. M. E.; DeBord, J. D.; Mebel, A.M.; Fernandez-Lima, F. Energy Fuels 2016, 30, 10341−10347.(26) Garabedian, A.; Benigni, P.; Ramirez, C. E.; Baker, E. S.; Liu, T.;Smith, R. D.; Fernandez-Lima, F. J. Am. Soc. Mass Spectrom. 2017, na.(27) Silveira, J. A.; Danielson, W.; Ridgeway, M. E.; Park, M. A. Int. J.Ion Mobility Spectrom. 2016, 19, 87−94.

Analytical Chemistry Technical Note

DOI: 10.1021/acs.analchem.7b04053Anal. Chem. 2018, 90, 2446−2450

2450