Method for Identification of Threonine Isoforms in Peptides by

patterns, which are independent of the residues adjacent to threonine or iso-threonine in a peptide sequence. .... Such sequences were used to simulat...
1 downloads 0 Views 451KB Size
Subscriber access provided by UNIV OF LOUISIANA

Article

Method for Identification of Threonine Isoforms in Peptides by Ultraviolet Photofragmentation of Cold Ions Elizaveta M. Solovyeva, Vladimir N. Kopysov, Aleksandr Y Pereverzev, Anna A. Lobas, Sergei A. Moshkovskii, Mikhail V. Gorshkov, and Oleg V. Boyarkin Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.9b00770 • Publication Date (Web): 01 May 2019 Downloaded from http://pubs.acs.org on May 1, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Method for Identification of Threonine Isoforms in Peptides by Ultraviolet Photofragmentation of Cold Ions Elizaveta M. Solovyeva§,†,‡ , Vladimir N. Kopysov§, Aleksandr Y. Pereverzev§, Anna A. Lobas‡, Sergei A. Moshkovskii#, Mikhail V. Gorshkov‡, and Oleg V. Boyarkin§* §

Laboratoire de Chimie Physique Moléculaire, École Polytechnique Fédérale de Lausanne, Station-6, 1015 Lausanne, Switzerland †

Moscow Institute of Physics and Technology (State University), 9 Institutskiy per., Dolgoprudny, Moscow Region, 141701, Russia ‡

V.L. Talrose Institute for Energy Problems of Chemical Physics, Russian Academy of Sciences, 38 Leninsky Pr., Bld.2 Moscow, 119334, Russia # Institute

of Biomedical Chemistry, 10 Pogodinskaya str., Moscow, 119121, Russia

ABSTRACT Identification of isomeric amino acid residues in peptides and proteins is challenging but often highly desired in proteomics. One of the practically important cases that require isomeric assignments is that associated with single-nucleotide polymorphism substitutions of Met residues by Thr in cancer-related proteins. These genetically encoded substitutions can yet be confused with the chemical modifications, arising from protein alkylation by iodoacetamide, which is commonly used in the standard procedure of sample preparation for proteomic analysis. Similar to the genetically encoded mutations, the alkylation also induces a conversion of methionine residues, but to iso-threonine form. Recognition of the mutations therefore requires isoform-sensitive detection techniques. Herein, we demonstrate an analytical method for reliable identification of isoforms of threonine residues in tryptic peptides. It is based on ultraviolet photodissociation mass spectrometry of cryogenically cooled ions and a machinelearning algorithm. The measured photodissociation mass spectra exhibit isoform-specific patterns, which are independent of the residues adjacent to threonine or iso-threonine in a peptide sequence. A comprehensive metric-based evaluation demonstrates that, being calibrated with a set of model peptides, the method allows for isomeric identification of threonine residues in peptides of arbitrary sequence.

1 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 23

Mass spectrometry-based shotgun proteomics is currently a method of choice for identification of sample proteins in biomedical and clinical applications.1–3 These identifications are based on a search against the unique reference protein sequence database. Because this search can return only the proteins present in the database, the information about alterations, such as alternative splicing, RNA editing, or single nucleotide polymorphisms (SNPs), etc. will be missing. Single amino acid variants (SAVs) are the products of nonsynonymous SNPs, considered as the main source of the genome variations. According to the recent sequencing data, there are 3 to 5 million SAVs in a human individual.4-8 A change in a single amino acid residue may substantially affect protein conformation, stability, or enzymatic activity, thus resulting in a change in the protein function.5 For instance, up to 60% of Mendelian diseases are caused by amino acid substitutions alone.6 Accounting for this information and identification of the sample-specific protein sequences is in the core of the proteogenomic concept introduced recently.7 Within this concept, the mass spectra are searched against a customized database generated using genomic and/or transcriptomic data obtained for particular sample, thus revealing the DNA or RNA alterations at the protein level.8 Regarding the complexity and the importance of the potential applications, the identification of SAVs must be performed with a high confidence. In shotgun proteogenomics, the identification of SAVs is commonly performed using mass spectrometry (MS),9–11 which measures a change in the mass of the residues that were substituted. The MS identifications are severely complicated by the circumstance that certain chemical modifications of amino acids (occurring both in vivo and in vitro) result in the same mass shifts for the corresponding residues in proteins. These modifications can be confused with the genetically induced SAVs, leading to false positive identifications of the variant peptides. An example of such a modification is methionine, Met, to iso-threonine (homoserine) chemical conversion, arising from protein alkylation by iodoacetamide (Figure S1). This reagent is typically and widely used in the standard procedure of preparation of protein samples for MS analysis.

12,13

The product

of this chemical conversion, isoThr, has to be distinguished from a genetically encoded point mutation of Met to (exclusively) Thr residue. In addition, Thr to Met substitution may be converted back by the same chemical modification, thus, resulting in an isoThr residue. When the two isoforms cannot be distinguished, the mutation is disregarded and lost as a SAV identification. The Met/Thr and Thr/Met substitutions are not a rare case and, on average, account for more than 2% of all SAVs, which is much higher than 0.3% median of the 2 ACS Paragon Plus Environment

Page 3 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

distribution for all 20 amino acids (Figure S2; generated from the data of the NCI-60 panel analysis14,15). Disregarding Met/Thr/Met SAVs can therefore noticeably reduce the number of identified genetic mutations, any of which might be the only origin of a severe disease. This high importance of the Met/Thr mutations in the cell functioning and disorders diagnostic motivates further developments of fast and reliable methods for identifications of Thr isoforms in protein sequences. Reversed phase high performance liquid chromatography (RP-HPLC), which is a widely employed method for separating isoforms of peptides,16 provides only a marginal separation of tryptic peptides that differ only by isoforms of Thr residues (typically, less than 1.5 min for the 50-minute gradient17). Moreover, the assignment of chromatographic peaks to the particular isoform is not always straightforward due to high sensitivity of retention times to the experimental conditions, which may result in inversion of the elution order for closely retained peptides.18,19 Occasionally, the induced fragmentation of parent ions analyzed by tandem mass spectrometry (MS/MS) reveals the presence of isoform-specific fragment ions (or fragmentation patterns), which allows for isoform assignment of the parent. For example, hot electron capture dissociation,20 collision-induced dissociation (CID),21 electron transfer dissociation (ETD)22, and fast atom bombardment23 of peptides with Leu/Ile residues produce isomer-specific fragment ions. The difference in higher-energy dissociation (HCD) fragment mass spectra was reported also for peptides with Thr/isoThr residues,17 although the generality of this effect (in particular for peptides with different residues adjacent to Thr) has yet to be assessed. The known high sensitivity of HCD mass spectra to experimental conditions makes it however challenging to reproduce these HCD spectra in different laboratories and with different instrumentation. Ultraviolet photodissociation (UVPD) is, perhaps, one of the most structure-specific fragmentation approaches. Different from all types of collision-induced fragmentations, photoexcitation by pulsed lasers of nanosecond (ns) duration is virtually an instantaneous collision-free process of photon absorption with well-defined energy, which can be easily reproduced with any appropriate laser light source. These factors drastically narrow the distribution of internal energy in ions, as compared with collisional excitations, making the subsequent fragmentation of the ions more sensitive to their structural details. Cryogenic cooling of protonated peptides further narrows the internal energy distribution enabling additional selectivity in the fragmentation. Here, we demonstrate the use of UVPD MS of 3 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

cryogenically cooled ions for highly confident and fast identification of Thr and isoThr residues in tryptic peptides of arbitrary sequence. The observed isomeric specificity of UVPD MS was first systematically verified using a set of 40 model synthetic peptides containing all 20 genetically encoded amino acids as the neighbors of the Thr/isoThr residues. The developed approach was then validated using a set of synthetic “human” tryptic peptides, allowing for development of an algorithm for unambiguous identification of the isomeric forms of threonine in peptides. EXPERIMENTAL SECTION Peptides. Two sets of peptides were synthesized in-house using a Hamilton Microlab Star® (Bonaduz, Switzerland) automated liquid handling workstation equipped with SynPhas Lanterns Mimotopes (Victoria, Australia) with a trityl linker and polypropylene 96-well plate Arctic White LLC, PE Frit 25 μm (Bethlehem, PA). Peptides were dried in an Eppendorf concentrator Plus (Hamburg, Germany), dissolved at 5*10-5 M concentration in 50/50 water/acetonitrile solutions with 1% of formic acid for the further analyses. The set of model peptides (Set 1) contained 20 isomeric pairs with general sequences of PXXTXXGSLGSLK and PXXisoTXXGSLGSLK, where “X” residue corresponded to one of the 20 natural amino acids. Such sequences were used to simulate typical tryptic peptide features, including Cterminal residue, sequence length, charge state and hydrophobicity. The second set (Set 2) contained 5 isomeric pairs of different peptides (Table S1 in SI), which were randomly selected from the in-silico digest of the human proteome. Each pair contained one isomer with Thr and another one with isoThr residue(s). UVPD of cold ions. Our versatile experimental setup (Figure S3 for 2D UV-MS spectroscopy of cold biomolecules has been described in detail elsewhere.24,25 Briefly, the protonated peptides are produced from solution using a nano-electrospray ionization, the ions of interest are selected by a quadrupole mass filter and transferred to a cryogenic octupole ion trap.26 This trap is cooled to T=6K by a closed cycle He refrigerator. He buffer gas is pulsed into the trap and quickly cooled in collisions with the walls of the trap. The arriving ions are collisionally cooled down to T=10-15 K. Once cooled and trapped, the ions are irradiated by a laser pulse from a VUV/UV optical parametric oscillator (OPO; 5 ns pulse duration at 10 Hz repetition rate, 0.5-5 mJ puse energy; NT342C model, EKSPLA, Lithuania). After a delay of 10-40 ms the parent ions and the generated photofragments are released from the octupole trap 4 ACS Paragon Plus Environment

Page 4 of 23

Page 5 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

and transferred into the high resolution Orbitrap mass-analyzer (Exactive, Thermo Scientific, Bremen, Germany). The 2D UV-MS spectra of Thr-containing peptides were recorded by scanning the wavelength of the OPO while measuring fragmentation MS at each wavelength. In UVPD measurements at the fixed wavelength of 193 nm, we performed 25 analytical scans, averaging 40 miscroscans per each scan. Each single microscan analyzed the parent and fragment ions accumulated in the C-trap of the Exactive MS during five consecutive cycles (trapping, fragmentation, release of ions) of the cold trap (one OPO pulse per cycle). Collisional fragmentation. In the HCD experiments, the OPO beam was blocked and the parent ions were transferred through the cold trap further to the collisional cell of the analyzing mass spectrometer. The isolation windows were chosen manually in the range from 2 to 3 Th depending on the isotopic distribution width. The collision gas was dry nitrogen, and the normalized collision energy (not specified by the instrument maker in absolute units) was set to 30% with the maximum injection time of 100 ms for precursor ion accumulation. For every studied peptide, 25 analytical spectra with 8 microscans in each scan were recorded and then averaged. Data analysis. 2D UV-MS data arrays were pre-processed using the Peak by PeakTM software (Spectroswiss Sarl, Lausanne, Switzerland). MS spectra were converted to ms127 and commonly used Mascot Generic Format (mgf) using MSConvert (ProteoWizard release: 2.1.2575)28 and analyzed by Python scripts written in-house using Pyteomics library.29 RESULTS AND DISCUSSION Probe peptides. Figure 1a shows an example of the UV photofragmentation spectra for a pair of

doubly

protonated

cold

isomeric

“probe”

peptides

DNIQGITKPAIR

and

DNIQGIisoTKPAIR randomly chosen from Set 2. Despite the cooling, the spectra do not appear vibrationally resolved. Nevertheless, suppression of the inhomogeneous thermal broadening makes the two spectra significantly different in the 200-240 nm UV spectral range, although, they become almost identical in the 200-192 nm VUV region, where the dissociation yield reaches the maximum around 193 nm.

Regarding the previous experience in

identification and quantification of isomeric peptides with the method of 2D UV-MS fingerprinting, the observed difference should allow for an identification and relative quantification of the two peptides with a few percent accuracy.30,31 Such identifications require, 5 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

however, a preliminary measurement of 2D UV-MS library for all isomeric peptides that are expected to appear in the analysis. Because the bottom-up proteomics deals with thousands of peptides, such a library may be unavailable or incomplete for all desirable peptides. This circumstance, eventually, prohibits a use of the 2D UV-MS library-based method for an arbitrary peptide. A close look at our data revealed that at any wavelength in the spectra in Fig. 1a, the relative abundances of certain photofragments appear to be specific to isoforms of Thr. Figures 1b and 1c show UVPD mass spectra of the two isomeric peptides, for which the UV photofragmentation spectra are shown in Fig. 1a. UVPD was performed at 193 nm because (i) the fragmentation yield is maximum (within the tuning range of our OPO) at this wavelength; (ii) this is the wavelength of ArF exсimer lasers, which are commonly used in UVPD MS2 studies.32–35 The two ratios of abundances of the photofragments, b6+/b7+ and y6+/y5+, are distinctly higher for the peptide that contain Thr residue than for the isomeric peptide with isoThr. These fragments correspond to the cleavage of the peptide bonds between the Thr, or isoThr in the 6th position of the sequence and the adjacent residues. These observations suggest that, perhaps, the peptide bond to the isoThr on the C-terminal side is weaker than the same bond to Thr. If the observed change in the relative abundances of the respective fragments is an intrinsic property of the isomeric substitution, it can be used for analytical identification of Thr isoform in any peptide regardless of the physical nature of this effect. Similar isomeric preferences in the respective fragment ion intensities were observed for all studied herein pairs of isomeric probe peptide with a Thr residue. The scale of the effect varies in a wide range for the peptides of different sequence, reflecting the differences in the bond energies between Thr/isoThr and different adjacent residues in the peptides of different sequences. To assess the versatility of this effect, we performed a systematic study of UVPD fragmentation MS spectra of cold peptide ions in which Thr/isoThr residue was surrounded by pairs of all natural amino acids. Model peptides. Set 1 of model peptides contains 20 isomeric pairs with the sequences PXXTXXGSLGSLK and PXXisoTXXGSLGSLK, where X residue is one of the 20 natural amino acids. The set resembles typical tryptic peptides and allows for studying the influence of a particular amino acid residue adjacent to Thr/isoThr on the fragmentation pattern. Figures 2a and 2b summarize the results of UVPD MS analysis for the model peptides. The figures show 6 ACS Paragon Plus Environment

Page 6 of 23

Page 7 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

the ratios of intensities of the fragment ions, which arise from the cleavage of the peptide bonds between Thr/isoThr (b3+/b4+ and y9+/y10+) and the adjacent residues X. There is clear evidence that, regardless of the identity of the residues X, all b3+/b4+ and y10+/y9+ ratios are systematically higher for the peptides with Thr residue than for the peptides with isoThr. Moreover, with a single noticeable exception, the b3+/b4+ ratios for the former and for the latter are separated by a threshold line of ~0.75. The only exception is the pair of peptides, in which the adjacent to threonine residues X are prolines. This residue is known to inhibit the dissociation of peptides with the formation of respective b-type fragment ions,36,37 which, indeed, could not be reliably detected due to their low abundance for this pair of peptides. Consistently, even after averaging of 25 analytical scans, the signal-to-noise ratio for b3 fragment is equal to one for both peptides. On the contrary, both y10 and y9 fragment ions have been detected in UV photodissociation mass spectra for all peptides. Based on the data shown in Figure 2, we tentatively propose a metric, Rexp, for the identification of the isoform of Thr residue in peptides. This metric considers the sum of the intensities of b- and y-fragment ions that originate from the cleavages of the peptide bonds adjacent to Thr/isoThr residue and is calculated using the following equation:

Rexp 

bN  yN , bC  yC

(1)

bN, yN and bC, yC designate the intensity of fragments, resulting from the cleavage of peptide bonds to Thr/isoThr residue on the side of N- and C-terminus , respectively. For instance, in the model peptides, bN, yN, bC and yC are the b3, y10, b4 and y9 ions, respectively. The intensities of the b- and y- fragment ions of interest were normalized by the total intensity of the selected group of fragment ions, y2 to y5, common for the all model peptides. The other common fragment ions, including y1, y6, and y7 were not used for normalization. The first fragment has low mass close to the lower detection mass range. The other two fragments were excluded because of their proximity to the variable part of peptides. Such correction should largely account for all potential differences in the ionization process, ion transport efficiency, differences in absorption cross-sections, etc. The normalized intensities of b- and yfragment ions that correspond to the cleavage of the peptide bonds adjacent to Thr/isoThr in the sequences are given for all natural amino acids based on the peptides from Set 1 in 7 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 23

Supplementary Tables S2 and S3. Once generated, this library of normalized fragment intensities can be further used for predicting the ratios of intensities for photofragments originated from cleavage of peptide bonds adjacent to Thr/isoThr residue in cold peptides of arbitrary sequence (see below). Human-like peptides. The proposed method for distinguishing peptides with threonine isoforms was further evaluated using Set 2 containing five pairs of synthetic peptides with sequences corresponding to the tryptic peptides of human proteins; each peptide in the test set contained either Thr or isoThr. In general, the pattern of a photofragmentation MS at a fixed UV wavelength is governed by the strength of the peptide bonds and by the proton affinities of different residues in the sequence. In contrast to the studied above model peptides (Set 1), the residues before and after threonine are different in the Set 2 peptides. In order to account for this circumstance, the metric Rexp (Eq.1) for the model peptides is to be modified as follows:

RTheor

bNX  yNX  Y Y bC  yC

(2)

where X and Y designate amino acid residues adjacent to Thr/isoThr from the left and right sides, respectively (i.e.,…XT(isoT)Y…). The normalized fragment ion intensities for these adjacent residues are taken from the experiments with model peptides (Table S2). Since the first step in the mechanism of the peptide ion fragmentation is a proton transfer and some particular residue such as glutamic and aspartic acids, proline, histidine, lysine, arginine and others may lead to distinctive peptide bond cleavage, the sequence specificity of the fragmentation pattern is not limited to the residues adjacent to the cleaved bond.38 Moreover, the probability of mobile proton to stay on the one or other part of a peptide formatting b- or yfragments also depend on more than just two neighboring residues. A comprehensive metric should, therefore, include a normalization that reflects the relative proton affinities of all residues in a peptide. A direct theoretical calculation of the proton affinity effect is, however, challenging. Instead, a number of machine learning algorithms were proposed to predict fragment ion intensities for any required peptide sequence.39–41 In this work we employed the MS2PIP Web Server (https://iomics.ugent.be/ms2pip/) to predict the intensities of fragment ions for the peptides containing Thr.39,42 The predicted intensities were used in Eq. 1 for calculating a metric, RMS2PIP, for the corresponding b- and y-ions. This empirical metric accounts for the proton affinity effects, but, because it is based on experimental data obtained 8 ACS Paragon Plus Environment

Page 9 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

for Thr, it predicts well the fragmentation pattern only for the peptides with this threonine isoform. We denote the ratio of RTheor calculated for peptide with Thr to the predicted value RMS2PIP for the same peptide as α. With this notation, for each pair of the isomeric human peptides of the same sequence, the normalized values of RTheor, obtained from Eq.2 can be Norm T Norm isoT ]   [RTheor ]T and [RTheor ]   [RTheor ]isoT for peptides with Thr and expressed as [RTheor

isoThr, respectively. A comparison of these predicted ratios with Rexp allows distinguishing between the peptides with Thr and isoThr residues. A peptide has that isoform of Thr, for Norm which the respective RTheor is closer to Rexp. The described procedure of metric calculation is

shown schematically in Figure 3. For the five pairs of synthetic human peptides from Set 2, containing six Thr/isoThr Norm residues, the isoform metrics RTheor calculated according to the described procedure are shown

in Figure 4. An isoform is considered as identified correctly if the experimentally observed ratio of fragment intensities (blue and red dots in Fig. 4 for peptides with Thr and isoThr, respectively) is closer to the ratio calculated for the same isoform (blue and red asterisks in Fig. 4 for peptides with Thr and isoThr, respectively). Graphically, this means that for the peptide in Fig. 4, a dot should be closer to the asterisk of the same color. The probability for a particular Thr isoform to be in a peptide can be accessed from the ratio between the distances from a dot (the experimental ratio) to the two respective asterisks. For instance, if a dot is twice closer to the asterisk corresponding to isoThr than to the Thr, the probability for the peptide to contain isoThr is 66%. If the distance ratio is close to 1, both options are equally possible. The application of the derived metric for analysis of the experiments with peptides of Set 2 correctly identified threonine isoforms in all but G4 peptide (TGHSLLHTLYGR), in which Thr follows His. The exception may relate to the known fact that the presence of His in a peptide often makes its fragmentation selective, involving an atypical b-ion structure, which complicates the prediction of fragment intensities.43 To test this assumption, the fragmentation HCD mass spectra were measured for all peptides of Set 2. The derived from the HCD data HCD metric, RExp , differs indeed for G4 by a factor of more than 2 from the value predicted by

MS2PIP (RMS2PIP), while for all other peptides the deviations are less than 20% (Figure S4). Regarding this exceptionalness, the experimental HCD intensities were further used instead of the MS2PIP prediction to calculate the theoretical ratios for G4 (denoted as color-coded triangles in Fig. 4). Such rectified procedure results in a correct identification of the Thr 9 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

isoform in the G4 pair of human peptides. This example demonstrated that the development of more precise algorithms for calculating HCD fragmentation pattern could further improve the predictability of the proposed method. Origin of isomeric specificity. The main experimental observation in this work is the significantly higher fragmentation efficiency for the peptide bond from the C-terminal side of the isoThr compared with the one of the Thr residue. We may speculate that this effect originates from the presence and specific location of the hydroxyl group on the side chain of the threonine residue. The hydroxyl group of the Thr side chain is closer to the peptide bond than in the case of isoThr. Because the nitrogen of the peptide bond has a lone electron pair, it may form an H-bond with the OH group of the Thr side chain. According to the currently accepted mobile proton model, the fragmentation process in an internally excited peptide begins with protonation of the amide group in a peptide bond.44–46 The involvement of the amide nitrogen in the H-bond obstructs the proton transfer by reducing the proton affinity of the site. Alternatively to the H-bond formation, the appearance of the additional electronegative group close to the peptide bond (OH of Thr side chain) may decrease the probability for the mobile proton to appear close to the nitrogen atom for initiating the fragmentation (Figure S5. Elucidation of the gas-phase geometry of a pair of the Thr/isoThr-containing peptides should shed light on the origin of the observed effect, which we employ for the identification of Thr isoforms. CONCLUSIONS Overall, we have developed and demonstrated an approach for distinguishing peptides with two different isoforms of threonine residues. It employs UV photofragmentation mass spectrometry of cryogenically cold peptide ions and monitors the intensities of the fragment ions that originate from the cleavage of the two peptide bonds adjacent to Thr/isoThr in the sequence. The systematic substitution of the two residues adjacent to Thr/isoThr on each side with all 20 natural amino acids reveals that the relative abundance of these fragments remains specific to the isomeric forms of Thr for any adjacent residue. This observation allows us to propose an algorithm for identification of Thr isoforms in any typical tryptic peptides. The algorithm compares the experimentally measured and the empirically predicted intensities of the specific fragments. It was successfully tested with isomeric pairs of synthetic human peptides. The experimental data obtained in this work enable a straightforward way of using the 10 ACS Paragon Plus Environment

Page 10 of 23

Page 11 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

proposed algorithm in other laboratories and with different instrumentation. In particular, this approach can be used in proteogenomic studies to distinguish on a timescale of tens of seconds the genetically encoded substitutions of Met to Thr from the in vitro chemical conversion of Met into isoThr residue. The employed combination of UVPD of cold ions with predictive machine learning algorithms, can, potentially be applied for identification of peptides with other than Thr isomeric residues. ASSOCIATED CONTENT Supporting information. Reaction of methionine to iso-threonine (homoserine) chemical conversion (Figure S1), Amino acid substitution frequencies (Figure S2); layout of the instrument (Figure S3); The HCD prediction errors (Figure S4); Schematic of mobile proton accommodation in peptides with Thr and isoThr residues (Figure S5); Sequences of synthetic peptides from Set 2 (Table S1); Normalized intensities of the selected fragments of peptides from Set 1 (Table S2).

AUTHOR INFORMATION Corresponding Author *E-mail: [email protected] Notes The authors declare no competing financial interest. ACKNOWLEDGMENTS This work was supported by the Swiss National Science Foundation (STCPSR program, grant IZLRZ2_163865; grant 206021_164101), EPFL and the Russian Foundation for Basic Research (grant 16-54-21006). The work of S.A.M. and M.V.G. was also funded by the Program for Basic Research of State Academies of Sciences for 2013–2020. The authors thank Dr. Yury. O. Tsybin for fruitful discussions, Ksenia Kuznetsova for assistance with the sample preparation, as well as Lev I. Levitsky and Dr. Mark V. Ivanov for help with data processing.

11 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

REFERENCES (1)

Yates, J. R. The Revolution and Evolution of Shotgun Proteomics for Large-Scale Proteome Analysis. J. Am. Chem. Soc. 2013, 135 (5), 1629–1640. DOI: 10.1021/ja3094313.

(2)

Geyer, P. E.; Kulak, N. A.; Pichler, G.; Holdt, L. M.; Teupser, D.; Mann, M. Plasma Proteome Profiling to Assess Human Health and Disease. Cell Syst. 2016, 2 (3), 185– 195. DOI: 10.1016/j.cels.2016.02.015.

(3)

Theis, J. D.; Dasari, S.; Vrana, J. A.; Kurtin, P. J.; Dogan, A. Shotgun-Proteomics-Based Clinical Testing for Diagnosis and Classification of Amyloidosis. J. Mass Spectrom. 2013, 48 (10), 1067–1077. DOI: 10.1002/jms.3264.

(4)

Ahn, S.-M.; Kim, T.-H.; Lee, S.; Kim, D.; Ghang, H.; Kim, D.-S.; Kim, B.-C.; Kim, S.Y.; Kim, W.-Y.; Kim, C.; et al. The First Korean Genome Sequence and Analysis: Full Genome Sequencing for a Socio-Ethnic Group. Genome Res. 2009, 19 (9), 1622–1629. DOI: 10.1101/gr.092197.109.

(5)

Yutani, K.; Ogasahara, K.; Sugino, Y.; Matsushiro, A. Effect of a Single Amino Acid Substitution on Stability of Conformation of a Protein. Nature 1977, 267, 274–275. DOI: 10.1038/267274a0.

(6)

Botstein, D.; Risch, N. Discovering Genotypes Underlying Human Phenotypes: Past Successes for Mendelian Disease, Future Approaches for Complex Disease. Nat. Genet. 2003, 33 (3s), 228–237. DOI: 10.1038/ng1090.

(7)

Bunger, M. K.; Cargile, B. J.; Sevinsky, J. R.; Deyanova, E.; Yates, N. A.; Hendrickson, R. C.; Stephenson, J. L. Detection and Validation of Non-Synonymous Coding SNPs from Orthogonal Analysis of Shotgun Proteomics Data. J. Proteome Res. 2007, 6 (6), 2331–2340. DOI: 10.1021/pr0700908.

(8)

Nesvizhskii, A. I. Proteogenomics: Concepts, Applications and Computational Strategies. Nat. Methods 2014, 11 (11), 1114–1125. DOI: 10.1038/nmeth.3144.

(9)

Sheynkman, G. M.; Shortreed, M. R.; Frey, B. L.; Scalf, M.; Smith, L. M. Large-Scale Mass Spectrometric Detection of Variant Peptides Resulting from Nonsynonymous 12 ACS Paragon Plus Environment

Page 12 of 23

Page 13 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Nucleotide Differences. J. Proteome Res. 2014, 13 (1), 228–240. DOI: 10.1021/pr4009207. (10)

Su, Z.-D.; Sun, L.; Yu, D.-X.; Li, R.-X.; Li, H.-X.; Yu, Z.-J.; Sheng, Q.-H.; Lin, X.; Zeng, R.; Wu, J.-R. Quantitative Detection of Single Amino Acid Polymorphisms by Targeted Proteomics. J. Mol. Cell Biol. 2011, 3 (5), 309–315. DOI: 10.1093/jmcb/mjr024.

(11)

Tanaka, K.; Takenaka, S.; Tsuyama, S.; Wada, Y. Determination of Unique Amino Acid Substitutions in Protein Variants by Peptide Mass Mapping with FT-ICR MS. J. Am. Soc. Mass Spectrom. 2006, 17 (4), 508–513. DOI: 10.1016/j.jasms.2005.12.006.

(12)

Gundlach, H. G.; Moore, S.; Stein, W. H. The Reaction of Iodoacetate with Methionine. J. Biol. Chem. 1959, 2657 (6), 1761–1765.

(13)

Chernobrovkin, A. L.; Kopylov, A. T.; Zgoda, V. G.; Moysa, A. A.; Pyatnitskiy, M. A.; Kuznetsova, K. G.; Ilina, I. Y.; Karpova, M. A.; Karpov, D. S.; Veselovsky, A. V.; et al. Methionine to Isothreonine Conversion as a Source of False Discovery Identifications of Genetically Encoded Variants in Proteogenomics. J. Proteomics 2015, 120, 169–178. DOI: 10.1016/j.jprot.2015.03.003.

(14)

Gholami, A. M.; Hahne, H.; Wu, Z.; Auer, F. J.; Meng, C.; Wilhelm, M.; Kuster, B. Global Proteome Analysis of the NCI-60 Cell Line Panel. Cell Rep. 2013, 4 (3), 609– 620. DOI: 10.1016/j.celrep.2013.07.018.

(15)

Abaan, O. D.; Polley, E. C.; Davis, S. R.; Zhu, Y. J.; Bilke, S.; Walker, R. L.; Pineda, M.; Gindin, Y.; Jiang, Y.; Reinhold, W. C.; et al. The Exomes of the NCI-60 Panel: A Genomic Resource for Cancer Biology and Systems Pharmacology. Cancer Res. 2013, 73 (14), 4372–4382. DOI: 10.1158/0008-5472.CAN-12-3342.

(16)

Winter, D.; Pipkorn, R.; Lehmann, W. D. Separation of Peptide Isomers and Conformers by Ultra Performance Liquid Chromatography. J. Sep. Sci. 2009, 32 (8), 1111–1119. DOI: 10.1002/jssc.200800691.

(17)

Kuznetsova, K. G.; Trufanov, P. V; Moysa, A. A.; Pyatnitskiy, M. A.; Zgoda, V. G.; Gorshkov, M. V; Moshkovskii, S. A. Threonine versus Isothreonine in Synthetic Peptides Analyzed by High-Resolution Liquid Chromatography/tandem Mass 13 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Spectrometry. Rapid Commun. Mass Spectrom. 2016, 30 (11), 1323–1331. DOI: 10.1002/rcm.7566. (18)

Tarasova, I. A.; Perlova, T. Y.; Pridatchenko, M. L.; Goloborod’ko, A. A.; Levitsky, L. I.; Evreinov, V. V.; Guryca, V.; Masselon, C. D.; Gorshkov, A. V.; Gorshkov, M. V. Inversion of Chromatographic Elution Orders of Peptides and Its Importance for Proteomics. J. Anal. Chem. 2012, 67 (13), 1014–1025. DOI: 10.1134/S1061934812130102.

(19)

Ogata, K.; Krokhin, O. V.; Ishihama, Y. Retention Order Reversal of Phosphorylated and Unphosphorylated Peptides in Reversed-Phase LC/MS. Anal. Sci. 2018, 34 (9), 1037–1041. DOI: 10.2116/analsci.18SCP11.

(20)

Kjeldsen, F.; Haselmann, K. F.; Sørensen, E. S.; Zubarev, R. A. Distinguishing of Ile/Leu Amino Acid Residues in the PP3 Protein by (Hot) Electron Capture Dissociation in Fourier Transform Ion Cyclotron Resonance Mass Spectrometry. Anal. Chem. 2003, 75 (6), 1267–1274. DOI: 10.1021/ac020422m.

(21)

Armirotti, A.; Millo, E.; Damonte, G. How to Discriminate between Leucine and Isoleucine by Low Energy ESI-TRAP MSn. J. Am. Soc. Mass Spectrom. 2007, 18 (1), 57–63. DOI: 10.1016/j.jasms.2006.08.011.

(22)

Lebedev, A. T.; Damoc, E.; Makarov, A. A.; Samgina, T. Y. Discrimination of Leucine and Isoleucine in Peptides Sequencing with Orbitrap Fusion Mass Spectrometer. Anal. Chem. 2014, 86 (14), 7017–7022. DOI: 10.1021/ac501200h.

(23)

Takemichi Nakamura, Hidemi Nagaki, Yasuko Ohki, T. K. Differentiation of Leucine and Isoleucine Residues in Peptides by Consecutive Reaction Mass Spectrometry. Anal. Chem. 1990, No. 14, 311–313.

(24)

Kopysov, V.; Makarov, A.; Boyarkin, O. V. Colors for Molecular Masses: Fusion of Spectroscopy and Mass Spectrometry for Identification of Biomolecules. Anal. Chem. 2015, 87 (9), 4607–4611. DOI: 10.1021/acs.analchem.5b00822.

(25)

Boyarkin, O. V. Cold Ion Spectroscopy for Structural Identifications of Biomolecules. Int. Rev. Phys. Chem. 2018, 37 (3–4), 559–606. DOI: 10.1080/0144235X.2018.1547453. 14 ACS Paragon Plus Environment

Page 14 of 23

Page 15 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

(26)

Boyarkin, O. V.; Kopysov, V. Cryogenically Cooled Octupole Ion Trap for Spectroscopy of Biomolecular Ions. Rev. Sci. Instrum. 2014, 85 (3), 33105. DOI: 10.1063/1.4868178.

(27)

McDonald, W. H.; Tabb, D. L.; Sadygov, R. G.; MacCoss, M. J.; Venable, J.; Graumann, J.; Johnson, J. R.; Cociorva, D.; Yates, J. R. MS1, MS2, and SQT—three Unified, Compact, and Easily Parsed File Formats for the Storage of Shotgun Proteomic Spectra and Identifications. Rapid Commun. Mass Spectrom. 2004, 18 (18), 2162–2168. DOI: 10.1002/rcm.1603.

(28)

Kessner, D.; Chambers, M.; Burke, R.; Agus, D.; Mallick, P. ProteoWizard: Open Source Software for Rapid Proteomics Tools Development. Bioinformatics 2008, 24 (21), 2534–2536. DOI: 10.1093/bioinformatics/btn323.

(29)

Levitsky, L. I.; Klein, J. A.; Ivanov, M. V.; Gorshkov, M. V. Pyteomics 4.0: Five Years of Development of a Python Proteomics Framework. J. Proteome Res. 2019, 18 (2), 709–714. DOI: 10.1021/acs.jproteome.8b00717.

(30)

Kopysov, V.; Gorshkov, M. V.; Boyarkin, O. V. Identification of Isoforms of Aspartic Acid Residues in Peptides by 2D UV-MS Fingerprinting of Cold Ions. Analyst 2018, 143 (4), 833–836. DOI: 10.1039/C7AN02044A.

(31)

Kopysov, V.; Makarov, A.; Boyarkin, O. V. Identification of Isomeric Ephedrines by Cold Ion UV Spectroscopy: Toward Practical Implementation. Anal. Chem. 2017, 89 (1), 544–547. DOI: 10.1021/acs.analchem.6b04182.

(32)

Barbacci, D. C.; Russell, D. H. Sequence and Side-Chain Specific Photofragment (193 Nm) Ions from Protonated Substance P by Matrix-Assisted Laser Desorption Ionization Time-of-Flight Mass Spectrometry. J. Am. Soc. Mass Spectrom. 1999, 10 (10), 1038– 1040. DOI: 10.1016/S1044-0305(99)00077-X.

(33)

Madsen, J. A.; Boutz, D. R.; Brodbelt, J. S. Ultrafast Ultraviolet Photodissociation at 193 Nm and Its Applicability to Proteomic Workflows. J. Proteome Res. 2010, 9 (8), 4205– 4214. DOI: 10.1021/pr100515x.

(34)

Choi, K. M.; Yoon, S. H.; Sun, M.; Oh, J. Y.; Moon, J. H.; Kim, M. S. Characteristics of Photodissociation at 193 Nm of Singly Protonated Peptides Generated by Matrix15 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Assisted Laser Desorption Ionization (MALDI). J. Am. Soc. Mass Spectrom. 2006, 17 (12), 1643–1653. DOI: 10.1016/j.jasms.2006.07.021. (35)

Morgan, J. W.; Russell, D. H. Comparative Studies of 193-Nm Photodissociation and TOF-TOFMS Analysis of Bradykinin Analogues: The Effects of Charge Site(s) and Fragmentation Timescales. J. Am. Soc. Mass Spectrom. 2006, 17 (5), 721–729. DOI: 10.1016/j.jasms.2006.02.004.

(36)

Raulfs, M. D. M.; Breci, L.; Bernier, M.; Hamdy, O. M.; Janiga, A.; Wysocki, V.; Poutsma, J. C. Investigations of the Mechanism of the “Proline Effect” in Tandem Mass Spectrometry Experiments: The “Pipecolic Acid Effect.” J. Am. Soc. Mass Spectrom. 2014, 25 (10), 1705–1715. DOI: 10.1007/s13361-014-0953-5.

(37)

Schwartz, B. L.; Bursey, M. M. Some Proline Substituent Effects in the Tandem Mass Spectrum of Protonated Pentaalanine. Biol. Mass Spectrom. 1992, 21 (2), 92–96. DOI: 10.1002/bms.1200210206.

(38)

Paizs, B.; Suhai, S. Fragmentation Pathways of Protonated Peptides. Mass Spectrom. Rev. 2005, 24 (4), 508–548. DOI: 10.1002/mas.20024.

(39)

Degroeve, S.; Martens, L. MS2PIP: A Tool for MS/MS Peak Intensity Prediction. Bioinformatics 2013, 29 (24), 3199–3203. DOI: 10.1093/bioinformatics/btt544.

(40)

Elias, J. E.; Gibbons, F. D.; King, O. D.; Roth, F. P.; Gygi, S. P. Intensity-Based Protein Identification by Machine Learning from a Library of Tandem Mass Spectra. Nat. Biotechnol. 2004, 22 (2), 214–219. DOI: 10.1038/nbt930.

(41)

Arnold, R. J.; Jayasankar, N.; Aggarwal, D.; Tang, H.; Radivojac, P. A Machine Learning Approach to Predicting Peptide Fragmentation Spectra. Pacific Symp. Biocomput. 2006, 230, 219–230.

(42)

Degroeve, S.; Maddelein, D.; Martens, L. MS 2 PIP Prediction Server: Compute and Visualize MS 2 Peak Intensity Predictions for CID and HCD Fragmentation. Nucleic Acids Res. 2015, 43 (W1), W326–W330. DOI: 10.1093/nar/gkv542.

(43)

Tsaprailis, G.; Nair, H.; Zhong, W.; Kuppannan, K.; Futrell, J. H.; Wysocki, V. H. A Mechanistic Investigation of the Enhanced Cleavage at Histidine in the Gas-Phase 16 ACS Paragon Plus Environment

Page 16 of 23

Page 17 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Dissociation of Protonated Peptides. Anal. Chem. 2004, 76 (7), 2083–2094. DOI: 10.1021/ac034971j. (44)

Wysocki, V. H.; Tsaprailis, G.; Smith, L. L.; Breci, L. A. Mobile and Localized Protons: A Framework for Understanding Peptide Dissociation. J. Mass Spectrom. 2000, 35 (12), 1399–1406. DOI: 10.1002/1096-9888(200012)35:123.0.CO;2-R.

(45)

Jones, J. L.; Dongré, A. R.; Somogyi, A.; Wysocki, V. H. Sequence Dependence of Peptide Fragmentation Efficiency Curves Determined by Electrospray Ionization/Surface-Induced Dissociation Mass Spectrometry. J. Am. Chem. Soc. 1994, 116 (18), 8368–8369. DOI: 10.1021/ja00097a055.

(46)

Tsaprailis, G.; Nair, H.; Somogyi, Á.; Wysocki, V. H.; Zhong, W.; Futrell, J. H.; Summerfield, S. G.; Gaskell, S. J. Influence of Secondary Structure on the Fragmentation of Protonated Peptides. J. Am. Chem. Soc. 1999, 121 (22), 5142–5154. DOI: 10.1021/ja982980h.

FIGURE CAPTIONS Figure 1. (a) Normalized photofragmentation UV spectra, and (b,c) fragmentation mass spectra (measured at OPO wavelength of 193 nm) of doubly protonated cold peptides DNIQGITKPAIR and DNIQGIisoTKPAIR. The peaks of b- and y- fragment ions labeled on the mass spectra correspond to the cleavages of the bonds adjacent to Thr/isoThr residue. The inserts in (b) and (c) depict the chemical structures of the Thr and isoThr residues, respectively. Figure 2. Ratios of intensities of singly charged (a) b-fragments and (b) y-fragments, originating from the cleavage of the two peptide bonds to Thr/isoThr residues for different adjacent to Thr residues. The intensities were measured in photofragmentation mass spectra of model peptide pair, PXXTXXGSLGSLK and PXXisoTXXGSLGSLK, where X is one of the 20 natural amino acids. Figure 3. The workflow for calculating the metric used for the identification of Thr isomers in an arbitrary peptide using photofragmentation MS of cold ions and pre-built library for all essential amino acids (see text for details). Figure 4. The experimental (color-coded dots) and normalized predicted (color-coded asterisks) ratios of fragments for five pairs of synthetic human-like tryptic peptides (A1-6= AVANQT/isoTSATFLR; A1-9= AVANQTSAT/isoTFLR; C1= DNIQGI(T/isoT)KPAIR; G3= 17 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

MSSPEDDSD(T/isoT)KR; G4=TGHSLLH(T/isoT)LYGR, and A10= TGNFQV(T/isoT)ELGR). The color-coded triangles show the predicted fragments ratios that are based on experimental HCD spectra for the G4 pair. An isoform is considered as identified correctly, if a dot (experimental ratio) is closer to the asterisk (theoretical ratio) of the same color than to the differently colored asterisk. Blue and red colors code the peptides with Thr and isoThr, respectively.

18 ACS Paragon Plus Environment

Page 18 of 23

Page 19 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Figure 1.

19 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 2

20 ACS Paragon Plus Environment

Page 20 of 23

Page 21 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Figure 3

21 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 4

22 ACS Paragon Plus Environment

Page 22 of 23

Page 23 of 23 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

TOC figure

23 ACS Paragon Plus Environment