Absolute Quantitation of Tryptophan-Containing Peptides and Amyloid Beta Peptide Fragments by Coulometric Mass Spectrometry

Abstract

Isotope labeled internal standards are routinely used for mass spectrometry (MS)-based absolute quantitation. However, syntheses of isotope-labeled peptides are time-consuming and costly. To tackle this issue, we recently developed a coulometric mass spectrometric (CMS) approach for absolute quantitation without the use of standards, based on the electrochemical oxidation of cysteine or tyrosine-containing peptides followed with mass spectrometric measurement of the oxidation yield. To further expand the utility of this method, herein, we present the CMS method for absolute quantitation of peptides based on tryptophan electrochemical oxidation. Several tryptophan-containing peptides, such as WGG, WQPPRARI, WAGGDASGE, RTRPLWVRME, and KVPRNQDWL were successfully quantified with a quantification error ranging from −4.5 to +4.3%. Furthermore, this quantitation approach is also applicable to protein, in which protein can be digested and a surrogate peptide can be selected for quantification to reflect the amount of the parent protein, as exemplified by CMS analysis of peptide GITWK from cytochrome c. The CMS result agreed well with the traditional isotope dilution method, with only a small difference of 3.5%. In addition, CMS was used to successfully quantify amyloid beta (Aβ) peptide fragments (up to 28 amino acid residues) based on tyrosine oxidation. The validity of CMS method for peptide and protein absolute quantitation without using isotope-labeled peptide standards would greatly facilitate proteomics research.

Graphical Abstract

Introduction

Quantitative mass spectrometry reporting the relative or absolute quantities of target peptides and proteins is very important for biological research.^1–9 Although relative quantification is now routinely used to provide valuable information on alteration of protein abundance in a proteome-wide scale,^10–14 the absolute quantitation determines the absolute concentration of target peptides and proteins within a sample, providing a far more precise description of molecular events in the biological processes.¹⁵ Moreover, absolute quantitation is crucial for the evaluation of clinical biomarker candidates and enables the data integration and comparison among different laboratories.^16–18 For the MS-based absolute quantitation of peptides and proteins, the isotope-labeled internal standards and the calibration curve are often needed.¹⁶ However, there are some challenges for employing those isotope-labeling absolute quantitation strategies. First, the most suitable standard needs to be selected and synthesized in advance, which is time consuming, expensive and sometimes unavailable. Second, using this costly technique, the concentration of only a few peptides of interest can be determined and it unsuits large-scale proteomic studies.¹ To address this, we recently developed a novel MS-based absolute quantitation method, named as coulometric mass spectrometry (CMS), which combines electrochemistry (EC) with MS for quantitation without the need of using any standards or calibration curve.^19–23

As illustrated in Scheme 1a, the basic workflow of CMS involves two steps. First, an electrochemically active substance that elutes from the LC column undergo an electrochemical reaction (oxidation or reduction). Electrons are transferred resulting in an electrical current peak, which then can be integrated over time to obtain the total charge Q. According to the Faraday’s Law, the total electric charge (Q) involved in the electrochemical reaction is proportional to the quantity of the oxidized/reduced substance: Q = nzF, where n denotes the moles of the oxidized/reduced analyte, z is the number of electrons transferred per molecule during the redox reaction, and F is the Faraday constant taken to be 9.65 × 10⁴ C/mol. Second, from the acquired MS spectra before and after electrolysis, the redox conversion yield Δi can be determined by measuring the relative change of the target analyte peak area in the extracted ion chromatogram (EIC). Thus, the total amount of the analyte can be calculated as the quotient of the amount of the oxidized analyte n and the oxidation yield Δi (i.e., Q/(zFΔi)).

Scheme 1.

a) CMS workflow and b) schematic illustration of our LC/EC/MS apparatus for CMS quantitation.

Previously, we have demonstrated that CMS can be used for accurately quantifying tyrosine- or cysteine-containing peptides^{19, 21} as well as proteins.²⁰ However, mainly relatively short peptides were examined and long peptides were not investigated. Additionally, peptide quantitation based on tryptophan oxidation were not studied. Among the 20 naturally occurring amino acids, tryptophan (W) is another oxidizable amino acid with a relatively low oxidation potential (1.01 V vs. NHE) compared to other oxidizable residues (others are cysteine (C), tyrosine (Y), and methionine (M)).²¹ In our trials, we also found that tryptophan-containing peptides generally exhibits a higher electric current signal than tyrosine-containing peptides, indicating that CMS would have a better sensitivity, if it could be built on tryptophan oxidation. Therefore, in this study, we attempted to explore the feasibility of CMS quantitation based on tryptophan residue oxidation in the purpose of expanding CMS applications for absolute peptide/protein quantitation. We successfully quantified a set of tryptophan-containing peptides as well as one surrogate peptide GITWK digested from cytochrome c with good quantitation accuracy. In addition, we successfully quantified two long tyrosine-containing peptides, Aβ peptide fragment Aβ_1-16 and Aβ_1-28, using CMS based on tyrosine oxidation.

EXPERIMENTAL SECTION

Materials.

[Glu¹¹]-amyloid β 1-16 (Aβ_1-16, DAEFRHDSGYEVHHQK, HPLC grade), [Gln¹¹]-amyloid β 1-28 human (Aβ_1-28, DAEFRHDSGYQVHHQKLVFFAEDVGSNK, HPLC grade), fibronectin adhesion-promoting peptide (WQPPRARI, HPLC grade), delta sleep inducing peptide (WAGGDASGE, HPLC grade) and cytochrome c from equine heart (99% purity) were bought from Sigma-Aldrich (St. Louis, MO). Trp-Gly-Gly-OH (WGG, HPLC grade) were obtained from Chem-Impex (Wood Dale, IL). BDC2.5 Mimotope (RTRPLWVRME, HPLC grade) and gp 100 (25-33) human (KVPRNQDWL, HPLC grade) were purchased from AnaSpec (Fremont, CA). Stable isotope-labeled peptide standards GITWK^ (labeled at lysine, ¹³C6, ¹⁵N2) was purchased from Vivitide (Gardner, MA). Trypsin/Lys-C mix (mass spec grade) was purchased from Promega (Madison, WI). Acetonitrile (ACN, HPLC grade) and Formic acid (HPLC grade) were bought from Fisher Scientific (Fair Lawn, NJ). Ammonium bicarbonate (bioultra grade) were bought from Sigma-Aldrich (St. Louis, MO). A Millipore Direct-Q5 purification system (Burlington, MA) was used to obtain purified water for sample preparation.

Instrumentation.

As shown in Scheme 1b, the experimental setup used for CMS consisted of a ultra-performance liquid chromatography (UPLC, Waters, Milford, MA) coupled with an electrochemical thin-layer flow cell (BASi, West Lafayette, IN; cell dead volume: ca. 1 μL) and a high-resolution Orbitrap Q Exactive mass spectrometer (Thermo Scientific, San Jose, CA). The electrochemical cell equipped with a glassy carbon disc (i.d., 3 mm & 6 mm, catalog# MF-1095 & MF-1015, respectively) as the working electrode (WE). An Ag/AgCl (3 M NaCl) electrode was used as the reference electrode (RE, catalog# MW-2021) and the cell stainless steel body served as a counter electrode (CE, catalog# MF1092). The working electrode was cleaned by polishing with alumina slurry before and after use.

A reversed phase column (BEH C18, 2.1 mm × 50 mm, 1.7 μm) was used for separation. A potential of +1.00 V or +1.05 V (vs Ag/AgCl) was applied to WE to trigger the oxidation of peptides. An Epsilon Eclipse™ potentiostat (BASi, West Lafayette, IN) was used to monitor and record the oxidation current. OriginPro 2018b was used to import and integrate the electric current peak to calculate the total electric charge of Q. The peptide flowing out of electrochemical cell was online analyzed using the Orbitrap mass spectrometer equipped with a heated electrospray ionization (HESI) source with the following parameters: sheath gas flow rate, 35; auxiliary gas flow rate, 10; spray voltage, 4.0kV; sweep gas flow rate, 0; capillary temperature, 300 °C; auxiliary gas heater temperature, 100 °C; S-lens RF level, 50. Mass spectra were acquired by Thermo Xcalibur (3.0.63).

Proteolytic Digestion.

Cytochrome c (100 μg) was dissolved in 50 mM ammonium bicarbonate (NH₄HCO₃, pH 8.0) followed by adding 50 μL of 0.2 μg/μL trypsin/lys-c mix solution. The protein sample was incubated at 37 °C for overnight. The digested cytochrome c sample was then diluted to the final concentration of 10 μM by adding mobile phase A (water with 0.1% formic acid).

LC/EC/MS analysis.

For LC/EC/MS analysis of Aβ_1-16 (DAEFRHDSGYEVHHQK), the mobile phase flow rate was 200 μL/min. In an 8-min linear gradient elution, the mobile phase B (acetonitrile with 0.1% formic acid) increased from 5% to 10% in 3 min, and reached 40% in 1 min. Then, mobile phase B was kept at 40% for 1 min before returned to 5%. The concentration of Aβ_1-16 used was 20.02 μM. The injection volume was 3 μL per analysis.

For LC/EC/MS analysis of Aβ_1-28 (DAEFRHDSGYQVHHQKLVFFAEDVGSNK), fibronectin adhesion-promoting peptide (WQPPRARI) and Delta sleep inducing peptide (WAGGDASGE), the mobile phase flow rate was 200 μL/min. In an 8-min linear gradient elution, the mobile phase B (acetonitrile with 0.1% formic acid) increased from 5% to 20% in 3 min, and reached 40% in 1 min. Then, mobile phase B was kept at 40% for 1 min before returned to 5%. The concentrations of Aβ_1-28, WQPPRARI and WAGGDASGE were 20.87 μM, 20.07 μM and 20.13 μM, respectively. The injection volume was 3 μL per analysis.

For LC/EC/MS analysis of WGG, the mobile phase flow rate was set as 200 μL/min. An isocratic elution program using 90% A (mobile phase A: water with 0.1% formic acid and mobile phase B: acetonitrile with 0.1% formic acid) for 6 min was used. The concentration of WGG used was 29.65 μM and the injection volume was 3 μL per analysis.

For LC/EC/MS analysis of BDC2.5 Mimotope (RTRPLWVRME) and gp 100 (25-33) human (KVPRNQDWL), the mobile phase flow rate was 200 μL/min. In an 8-min gradient elution, the mobile phase B (acetonitrile with 0.1% formic acid) increased from 5% to 30% in 3 min, and reached 50% in 1 min. Then, mobile phase B was kept at 50% for 1 min before returned to 5%. The concentrations of RTRPLWVRME and KVPRNQDWL were 20.03 μM and 20.05 μM, respectively. The injection volume was 3 μL per analysis.

For LC/EC/MS setup of digested cytochrome c, the mobile phase flow rate was 250 μL/min. The mobile phase B (acetonitrile with 0.1% formic acid) increased from 5% to 10% in 1 min, and then reached to 14% from 1-16 min. Then the mobile phase B climbed to 70% in 1 min and remained at 70% for 3 min before returned to 5%. The injection concentration of digested cytochrome c was 10 μM. The injection volume was 6 μL per analysis.

RESULTS AND DISCUSSION

CMS quantitation of Amyloid beta peptide fragments

Alzheimer’s disease (AD) is becoming a major public health problem worldwide.^{24, 25} Amyloid beta (Aβ) peptide and their peptide fragments, derived from the proteolytic processing of amyloid precursor protein (APP),^{26, 27} have been considered as the diagnostic biomarker and therapeutic target of Alzheimer’s disease (AD).^28–30 Therefore, quantification of those different Aβ isoforms and their fragments is significant for early diagnosis of AD^{29, 31} as well as facilitating the investigation of disease mechanism and thus benefiting drug discovery. MS is widely used and represents an important tool in the field of AD due to their capability of providing both qualitative and quantitative information on the Aβ involved in AD.^{32, 33} However, for traditional MS-based absolute quantitation of Aβ peptides,^{34, 35} isotopically labeled Aβ peptides are often needed as internal standards, which are time-consuming and expensive. In this study, we first conducted quantifications of two long Aβ peptide fragments Aβ_1-16 and Aβ_1-28 using CMS based on the electrochemical oxidation of tyrosine residue without adding isotope labeling standards.

Aβ peptide fragments, Aβ_1-16, was first chosen for CMS test, which is a Tyr-containing peptide (sequence: DAEFRHDSGYEVHHQK). Based on our previous studies, tyrosine residue can be oxidized into semi-quinone by losing two electrons and two protons.^{20, 21} A blank sample (solvent only) was first injected and no electric current peak was observed when the oxidation potential +1.05 V was applied (Figure 1e). In contrast, after the injection of Aβ_1-16 sample, a sharp electric current peak was observed under the same oxidation potential of +1.05 V (Figure 1f), indicating that the current peak observed in Figure 1f was derived from electrochemical oxidation of Aβ_1-16. Indeed, it was further confirmed by the corresponding pair of MS spectra recorded upon the electrochemical oxidation. Compared to the mass spectrum (Figure 1a) without electrolysis in which the +4 ion of DAEFRHDSGYEVHHQK was observed at m/z 489.47, a new peak at m/z 488.97 arose (Figure 1b), corresponding to +4 ion of the oxidized peptide product, when +1.05 V was applied to the cell for oxidation. The integrated EIC peak area of m/z 489.47 shown in Figure 1d was reduced by 11.3% compared to the same peak in Figure 1c, suggesting that the oxidation yield for DAEFRHDSGYEVHHQK was 11.3% (see detailed data in SI, Table S1). The amount of the oxidized Aβ_1-16 was calculated to be 6.6 pmol, based on the integration of the current peak area shown in Figure 1f and the Faraday’s Law. Therefore, the measured amount of Aβ_1-16 was 58.3 pmol (SI, Table S1). A triplicate measurement gave the average amount of Aβ_1-16 to be 58.3 pmol, which turned out to be very close to the theoretical amount of 60.1 pmol, with the measurement error of −3.0% (SI, Table S1). Aβ_1-28, another long amyloid peptide that contains Tyr (sequence: DAEFRHDSGYQVHHQKLVFFAEDVGSNK), was also analyzed by CMS. The average amount of Aβ_1-28 was measured at 60.6 pmol with the quantitation error of −3.2% (SI, Table S2).

Figure 1.

MS spectra of Aβ_1-16 (a) when the cell was off and (b) when the cell was turned on (applied potential: + 1.05 V). The oxidation product of Aβ_1-16 was detected at m/z 488.97. EICs of Aβ_1-16 were acquired (c) when the cell was off and (d) when the cell was turned on (applied potential: + 1.05 V). Electric current diagrams were collected from (e) blank solvent and (f) the oxidation of Aβ_1-16.

CMS quantitation of tryptophan containing peptides

The result above reveals that our CMS method is applicable to absolute quantitation of quite long tyrosine-containing peptides, besides short peptides that we demonstrated before. To further extend the CMS application, we investigated the possibility of using it for absolute quantitation of tryptophan-containing peptides in this study. Tryptophan is known to be electroactive and has a relatively low oxidation potential (+1.01 V, vs NHE²¹). The major oxidation products and pathways of Trp-containing peptides have been reported in literature using combined EC/MS systems.^{36, 37} As illustrated in Scheme 2, in general, the major oxidation products are observed as the singly hydroxylated M+16 (2e⁻ oxidation product) and the ketone product M+14 (4e⁻ oxidation product) and minor products include the doubly hydroxylated M+32 (4e⁻ oxidation product) and cleavage product R₁W+14 (4e⁻ oxidation product, cleavage products were observed in oxidation of some peptides such as KVPRNQDWL and GITWK in this study). The various oxidation products can be readily monitored by our online LC/EC/MS system. Based on the similar structures of those oxidation products, we assume that the intensity ratios of the different mass peaks of the oxidized products are approximately equal to ratios of moles of the different oxidized products, as shown in eq. 1–3. For calculation, we assume that moles of products M+16, M+14, M+32 and R₁W+14 are n₁, n₂, n₃ and n₄, respectively, and the total mole of the oxidized peptide is n=n₁+n₂+n₃+n₄. Based on the Faraday’s Law, the electricity Q is contributed from all the four oxidation pathways (eq. 4). From eq. 1–4, the total amount of oxidized peptide n can be calculated as denoted in eq. 5, where the Q is experimentally obtained by integrating the Faradaic current over time and F is the Faraday constant (9.65 × 10⁴ C/mol). Once we get n and the oxidation yield, the total amount of the peptide could be calculated as the quotient of the amount of the oxidized peptide n and the oxidation yield Δi (i.e., n/Δi).

Scheme 2.

Equation showing electrochemical oxidation of a Trp-containing peptide. R₁ and R₂ denote the peptide N-terminal and C-terminal moieties, respectively.

$$\frac{{\text{n}}_{\text{2}}}{{\text{n}}_{\text{1}}}\text{=}\frac{\left[\text{M+14}\right]}{\left[\text{M+16}\right]}$$ (1)

$$\frac{{\text{n}}_{\text{3}}}{{\text{n}}_{\text{1}}}\text{=}\frac{\left[\text{M+32}\right]}{\left[\text{M+16}\right]}$$ (2)

$$\frac{{\text{n}}_{\text{4}}}{{\text{n}}_{\text{1}}}\text{=}\frac{\left[{\text{R}}_1\text{W+14}\right]}{\left[\text{M+16}\right]}$$ (3)

$${\text{Q=2n}}_{\text{1}}{\text{F+4n}}_{\text{2}}{\text{F+4n}}_{\text{3}}{\text{F+4n}}_{\text{4}}\text{F}$$ (4)

$${\text{n=n}}_1{\text{+n}}_{\text{2}}{\text{+n}}_{\text{3}}{\text{+n}}_{\text{4}}=\frac{\text{Q}}{\text{2F}}\times \frac{1+\left(\left[\text{M+14}\right]\text{+}\left[\text{M+32}\right]\text{+}\left[{\text{R}}_{\text{1}}\text{W+14}\right]\right)/\left[\text{M+16}\right]}{1+2\times \left(\left[\text{M+14}\right]\text{+}\left[\text{M+32}\right]\text{+}\left[{\text{R}}_{\text{1}}\text{W+14}\right]\right)/\left[\text{M+16}\right]}$$ (5)

To explore the feasibility of CMS peptide quantitation based on tryptophan oxidation, a tripeptide WGG was first tested. A blank solvent sample was first injected and no electric current peak was observed when the oxidation potential +1.00 V was applied (Figure 2e). In contrast, after the injection of WGG sample, a sharp electric current peak was observed under the same oxidation potential of +1.00 V (Figure 2f), indicating that the current peak observed in Figure 2f originated from electrochemical oxidation of WGG. Before electrolysis (Figure 2a), the protonated WGG was detected at m/z 319.14. After electrolysis (Figure 2b), the oxidation products WGG+14, WGG+16, and WGG+32 were observed at m/z 333.12, m/z 335.13 and m/z 351.13 respectively. In particular, the ion intensities of M+14 and M+16 were much higher than the M+32 (the ion intensities and intensity ratios are shown in SI Table S3). The integrated EIC peak area of m/z 319.14 shown in Figure 2d was smaller by 4.4% compared to the same peak in Figure 2c, suggesting that the oxidation yield for WGG was 4.4% (see detailed data in SI Table S3). Based on eq. 5 using the integration of the current peak area (i.e., total electric charge Q, Figure 2f) as well as the intensity ratios of M+14/M+16 and M+32/M+16, we calculate the oxidized amount of WGG to be 3.8 pmol. In combination of the oxidation yield 4.4%, the measured amount of WGG was 87.1 pmol. The average amount of a triplicate measurement was to be 87.1 pmol, which is quite close to the injection amount of 89.0 pmol with the measurement error of – 2.1%.

Figure 2.

MS spectra of WGG (a) when the cell was off and (b) when the cell was turned on (applied potential: + 1.00 V). The oxidation products of WGG were detected at m/z 333.12, 335.13 and 351.13. EICs of WGG were acquired (c) when the cell was off and (d) when the cell was turned on (applied potential: + 1.00 V). Electric current diagrams were collected from (e) blank solvent and (f) the oxidation of WGG

To confirm that the oxidized products M+14, M+16 and M+32 were truly produced from electrochemical oxidation of tryptophan in the WGG as shown in Scheme 2, MS/MS analysis of m/z 319.14, m/z 333.12, m/z 335.13, and m/z 351.13 were conducted. In this experiment, the WGG sample was directly injected into the electrochemical cell by a syringe pump for oxidation at + 1.00 V potential. The resulted oxidation products and unoxidized peptide were analyzed using online EC/MS/MS. Collision-induced dissociation (CID) spectra of m/z 319.14, m/z 333.12, m/z 335.13, and m/z 351.13 were recorded. As shown in Figure 3, upon CID, the unoxidized peptide ion (m/z 319.14) gave rise to fragment ion b₂ (m/z 244.11,Figure 3a), while b₂+14 (m/z 258.09, Figure 3b), b₂+16 (m/z 260.10, Figure 3c), and b₂+32 (m/z 276.10, Figure 3d) were observed for oxidation product WGG+14, WGG+16 and WGG+32 respectively, in agreement with the fact that the observed m/z 333.12, m/z 335.13, and m/z 351.13 resulted from the oxidation of tryptophan residue in the peptide. Moreover, the b₁ ion readily loses CO to form a₁ ion (m/z 159.09) which was observed upon CID of the unoxidized peptide ion (m/z 319.14, Figure 3a). As shown in Figures 3b–d, the corresponding a₁ ions were all observed from CID of WGG+14, WGG+16 and WGG+32 ions with mass shift of +14, +16 and +32 Da, respectively, confirming that the oxidation occurred to the first tryptophan residue of WGG.

Figure 3.

CID MS/MS spectra of (a) the unoxidized peptide ion [WGG+H]⁺ (m/z 319.14), (b) the oxidized peptide ion [WGG+14+H]⁺ (m/z 333.12), (c) [WGG+16+H]⁺ (m/z 335.13) and (d) [WGG+32+H]⁺ (m/z 351.13).

WAGGDASGE, another Trp-containing peptide with 9 amino acid residues was also analyzed by the same approach. Without oxidation (Figure 4a), +1 ion of WAGGDASGE was observed at m/z 849.34. As shown in Figure 4b, after electrolysis, the oxidation products at m/z 863.32 (M+14 ion), m/z 865.33 (M+16 ion) and m/z 881.33 (M+32 ion) were observed. Based on EIC peak change form/z 849.34 upon oxidation (Figure 4c and d), the oxidation yield for WAGGDASGE was measured as 19.0% (see detailed data in SI Table S4). Figure 4f shows the electric current peak from oxidation of WAGGDASGE (as a control, no oxidation current peak was observed in Figure 4e from solvent blank under the same oxidation potential). The amount of oxidized WAGGDASGE was calculated to be 11.1 pmol, based on eq. 5 using the integration of the current peak area (total electric charge Q) and the intensity ratios of m/z 863.32 and m/z 881.33 to m/z 865.33. Therefore, the measured amount of WAGGDASGE was 58.4 pmol (SI Table S4). In a triplicate measurement, the averaged quantity of this peptide measured by CMS was 57.7 pmol which is close to the initial amount of 60.4 pmol injected with a measurement error −4.5% (SI Table S4). WQPPRARI and RTRPLWVRME, another two peptides longer than WGG containing tryptophan either as the first residue or in the middle of the sequence, were also quantified using CMS. The measurement error were 4.3% (SI Table S5 and Figure S2) and −2.3% (SI Table S6 and Figure S3) respectively.

Figure 4.

MS spectra of WAGGDASGE (a) when the cell was off and (b) when the cell was turned on (applied potential: + 1.00 V). The oxidation product of WAGGDASGE was detected at m/z 863.32, m/z 865.33 and m/z 881.33. EICs of WAGGDASGE were acquired (c) when the cell was off and (d) when the cell was turned on (applied potential: + 1.00 V). Electric current diagrams were collected from (e) blank solvent and (f) the oxidation of WAGGDASGE.

In particular, upon electrochemical oxidation of another Trp-containing peptide KVPRNQDWL, besides M+14, M+16 and M+32 products, +3 ion of the cleavage product (KVPRNQDW+14, SI Figure S4b) with low intensity was observed at m/z 352.84. The oxidized amount of KVPRNQDWL was calculated as 5.9 pmol by eq. 5 based on the integration of the current area (SI Figure S4f), and the intensity ratio of the oxidation products observed (SI Table S7). The integrated EIC peak area of unoxidized peptide ion at m/z 385.88 decreased by 9.6% after the cell was turned on, indicating the oxidation yield was 9.6%. Therefore, the total amount of KVPRNQDWL was measured to be 61.0 pmol. Triplicate measurements gave an averaged amount of KVPRNQDWL of 61.0 pmol, which is close to the injected amount of 60.1 pmol with a small quantitation error 1.5% (SI Table S7).

CMS quantitation of cytochrome c

With the success in quantifying a series of Trp-containing peptides, we took a step further to use our CMS method for the absolute quantitation of proteins. Cytochrome c from equine heart (104 amino acids, sequence shown in Figure 5a) was chosen as a test sample. A surrogate peptide GITWK digested from cytochrome c was identified and quantified using CMS. Blank solvent was first injected and no current peak was observed when +1.00 V was applied (Figure 5f). In contrast, an electric current peak from oxidation of GITWK was observed under the same potential (Figure 5g). Before oxidation, +2 ion of GITWK was observed at m/z 302.68. As shown in Figure 5b, after the cell was turned on, ions of the M+14, M+16 and M+32 products as well as small amount of cleavage product GITW+14 were observed at m/z 309.67, m/z 310.67, m/z 318.67 and m/z 490.23, respectively (Figure 5c). The oxidation yield for GITWK was measured to be 9.8% (SI Table S8), by comparing the peak area of m/z 302.68 before and after oxidation. Based on eq. 5 using the integration of electric current peak area from peptide oxidation and product ion intensity ratios, the oxidized amount of GITWK was calculated as 5.4 pmol. Further, according to the oxidation yield, the total amount of GITWK was measured to be 53.4 pmol from a triplicate experiment (SI Table S8). Based on the assumption that stoichiometrically one protein molecule produces one peptide molecule in theory, the measured protein amount is therefore 53.4 pmol, which is also close to the amount of protein of 60 pmol (−11.0% error, SI Table S8) for generating the 6 μL of the protein digest that was injected. The −11.0% measurement error might be from the sample loss caused by the tryptic digestion of proteins to peptides.²⁰ To further check our measurement accuracy, the same diluted cytochrome c digest sample was also measured using traditional isotope dilution method with an isotope labeled peptide GITWK^ (labeled at lysine, ¹³C6, ¹⁵N2). The concentration of the diluted digest sample was quantified to be 8.6 μM (injection volume 6 μL, total amount of 51.6 pmol) based on the known amount of the isotope labeled peptide GITWK^ added and the intensity ratio of both peptides (the calibration curve obtained was shown in SI Figure S5). The difference of the measured amount of surrogate peptide GITWK from cytochrome c digest between CMS (53.4 pmol) and isotope dilution method (51.6 pmol) was only 3.5%, validating the feasibility of using CMS for protein absolute quantitation based on the oxidation of tryptophan residue.

Figure 5.

(a) Sequence of cytochrome c (the chosen surrogate peptide GITWK for CMS is highlighted in bold). MS spectra of GITWK from cytochrome c (b) when the cell was off and (c) when the cell was turned on (applied potential: + 1.00 V). The oxidation product of GITWK was detected at m/z 309.67, m/z 310.67, m/z 318.67 and m/z 490.23. EICs of GITWK were acquired (d) when the cell was off and (e) when the cell was turned on (applied potential: + 1.00 V). Electric current diagrams were collected from (f) blank solvent and (g) the oxidation of GITWK.

In addition, in such an experiment of CMS quantitation of proteins, more than one surrogate peptide in protein digest could be quantified to obtain quantitation redundancy.²⁰ For instance, in a separate trial, both GITWK and another tyrosine containing surrogate peptide TGQAPGFTYTDANK from the same cytochrome c digest were quantified together by CMS. Those two peptides were separated and quantified in the same LC/EC/MS run. Results showed that the GITWK and TGQAPGFTYTDANK were quantified to be 53.5 and 52.7 pmol by CMS (theoretical amount was 60.5 pmol based on the initial amount of cytochrome c used for digestion; data shown in SI Figure S6 and Table S9). The CMS quantitation results for two peptides agreed with each other well, with only 1.5% difference. This result demonstrates the feasibility of quantifying multi-peptides in a single run and the reliability of CMS quantitation for a target protein based on different surrogate peptides.

From the results shown above, one can see that the quantitation error between our measured and theoretical values is only −4.5%− +4.3% (Table 1). It clearly shows that CMS can be applicable for quantifying Trp-containing peptides and our assumption used for the calculation (eq. 1–3) is valid, although the oxidation undergoes multiple pathways (Scheme 2). Indeed, M+14 and M+32 are structurally similar to M+16 and the only difference among these products are from the substituents in the tryptophan side chains. The cleavage product is shorter than the product M+16 and their ionization efficiency could have some differences so that the calculation of n₄ used intensity ratios could have some errors. However, the detectable cleavage product were only observed in the oxidation of two peptides in our experiments and the cleavage product ion intensity were quite small in comparison to the sum of all other products (the ratio of the former to the latter is less than 3.4%). Thus, the measurement error of the oxidized peptide amount n from the estimated calculation of n₄ should be very small. This explains why all the Trp-containing peptides were measured by CMS with good accuracy.

Table 1.

List of Trp-containing peptides quantified by CMS in this work

#	Name	Peptide sequence	Molecular weight (Da)	Measured amount (pmol)	Theoretical amount (pmol)	Quantitation measurement error
1	L-Tryptophylglycylglycine	WGG	318.3	87.1	89.0	−2.1%
2	Fibronectin adhesion-promoting peptide	WQPPRARI	1023.2	62.8	60.2	4.3%
3	gp 100 (25-33) human	KVPRNQDWL	1155.3	61.0	60.1	1.5%
4	BDC2.5 Mimotope	RTRPLWVRME	1343.6	58.7	60.1	−2.3%
5	Delta sleep inducing peptide	WAGGDASGE	848.8	57.7	60.4	−4.5%

The sensitivity of CMS based on the tryptophan oxidation was also evaluated by testing a series of low concentration of peptide WGG ranging from 0.025 μM, 0.05 μM, 0.1 μM (3 μL injection volume). The quantitation results were shown in SI Table S10. 3 μL of 0.025 μM WGG (injection amount 75 fmol) was injected into LC/EC/MS for quantitation and the measured peptide quantity was 74.1 fmol on average (−1.2% quantitation error, see detailed results in Table S10 in Supporting Information), suggesting a high quantitation sensitivity and accuracy. Indeed, in comparison to the CMS quantitation of Y-containing peptides where 300 fmol of DRVY was successfully quantified,²¹ the quantitation sensitivity of using W-containing peptide is improved, under a similar experimental condition. This is in line with our expectation mentioned before, based on our observation that W-containing peptide provides a relatively high oxidation current upon oxidation. Besides, the peak width at half height of EIC peaks and their corresponding EC peaks for peptides studied are compared (SI Table S11). A slight peak broadening was observed EIC peaks, which might be caused by the extra dead volumes of the electrochemical flow cell and the connection tubing between EC and MS.

CONCLUSIONS

In this study, we demonstrated the viability of using CMS for the absolute quantitation of very long peptides such as Aβ peptide fragments (Aβ_1-16 and Aβ_1-28) based on the tyrosine-oxidation. Moreover, to further extend the application of our method in peptides and protein quantitation, we also extended the CMS absolute quantitation to tryptophan-containing peptides. Good quantitation accuracy and sensitivity were achieved. Furthermore, absolute quantitation of protein (using cytochrome c as an example) was also demonstrated and confirmed with traditional isotope dilution method. One striking strength of our method is that there is no need for isotope-labeled internal standard and calibration curve which is routinely used in MS-based absolute quantitation, showing the advantage of cost and time efficiency for CMS. Since CMS has been shown to be capable of quantifying either peptides or proteins containing either tryptophan, tyrosine, or cysteine residues, it would have a good application potential in quantitative proteomics. To realize the application of CMS in the real proteomics world, reducing the current setup from micro scale to nanoscale (i.e., nanoLC and nanoESI) would be necessary. Such experiment is underway.