Profiling B-Type Natriuretic Peptide Cleavage ... - ACS Publications

Sep 1, 2017 - and Jennifer E. Van Eyk*,†. †. Heart Institute, Cedars-Sinai Medical Center, 127 South San Vicente Boulevard, Los Angeles, Californi...
0 downloads 0 Views 901KB Size
Subscriber access provided by - Access paid by the | UCSB Libraries

Article

Profiling B-Type Natriuretic Peptide Cleavage Peptidoforms in Human Plasma by Capillary Electrophoresis with Electrospray Ionization-Mass Spectrometry. Shenyan Zhang, Koen Raedschelders, Marcia Santos, and Jennifer E. Van Eyk J. Proteome Res., Just Accepted Manuscript • DOI: 10.1021/acs.jproteome.7b00482 • Publication Date (Web): 01 Sep 2017 Downloaded from http://pubs.acs.org on September 3, 2017

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 free 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 accessible to all readers and 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.

Journal of Proteome Research 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 11

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

Journal of Proteome Research

Profiling B-Type Natriuretic Peptide Cleavage Peptidoforms in Human Plasma by Capillary Electrophoresis with Electrospray Ionization-Mass Spectrometry. Shenyan Zhang1, Koen Raedschelders1, Marcia Santos2, Jennifer E. Van Eyk1* 1

Heart Institute, Cedars-Sinai Medical Center. 127 S San Vicente Blvd. Los Angeles CA. 2Sciex Separations, 250 S Kraemer Blvd, Brea, CA KEYWORDS Capillary Electrophoresis, Mass Spectrometry, CESI 8000, Peptidoforms, Brain Natriuretic Peptide, Multisegment Injection, Neutral Capillary, proteolysis profile. ABSTRACT: B-type Natriuretic Peptide (BNP) is a biologically active circulating hormone. Plasma concentrations of BNP are routinely used in the diagnosis of heart failure and the intravenous infusion of recombinant BNP can be used for heart failure treatment. Like many bioactive polypeptides, multiple plasma enzymes are known to cleave circulating BNP, and as part of the CVDB/D-HPP mandate, we sought to develop a technique capable of profiling these catabolic processes in plasma. We used a neutralcoated Capillary Electrophoresis-Electrospray Ionization (CESI) separation system coupled with high-resolution mass spectrometry to profile the proteolysis of exogenous recombinant BNP1-32 in plasma. Our method utilizes electrokinetic injection of minimally processed plasma samples to simultaneously monitor the dynamic generation and breakdown of at least five BNP peptidoforms in plasma. By integrating multisegment injection, our method can produce a multi-point BNP proteolytic profile for one sample within an hour. We envision applying this method to assess the potential relation between plasma-based BNP proteolysis and heart failure, as well as a means of monitoring BNP bioavailability after therapeutic infusion.

Introduction Heart disease continues to persist as a major cause of worldwide mortality, underscoring the urgent need for improved diagnostic and risk stratification tools1. B-type Natriuretic Peptide (BNP) is a key biomarker whose quantitative analysis is used to clinically assess heart failure2-3. In fact, BNP is one of the most popular human proteins, a designation based on the number of citations, which was determined as part of the cardiovascular biology/disease-human proteome project (CVD B/DHPP) of the human proteome organization4. Bioactive BNP is a peptide hormone that is secreted from myocytes and is translated as a 132 amino acid prehormone and cleaved to the 108 amino acid proBNP glycopeptide. Furin/Corin-mediated proBNP cleavage produces two fragments, NT-proBNP1-76 and BNP1-32, which are secreted into the circulation in equimolar amounts5. BNP1-32 facilitates cardiovascular fluid homeostasis through the counter-regulation of the renin-angiotensin-aldosterone system, increasing cellular cyclic guanosine monophosphate, activating protein kinase G, resulting in smooth muscle cell relaxation and decreased cardiac load. This protective mechanism can be impaired in heart failure, and these patients can have greater than 4 times higher concentration of circulating BNP than healthy individuals, with persistently low ventricular ejection fraction6. Circulating BNP is also elevated in other pathologies including atrial fibrillation, acute myocardial infarction, hypertension, acute lung injury, and chronic renal failure. Ultimately, physiological levels of circulating BNP1-32 reflect a state of dynamic equilibrium in which secretion from cardiomyocytes is counterbalanced by active degradation in plasma by at least three known peptidases3, neutral endopeptidase (also called as neprilysin, NEP), dipeptydilpeptidase IV (DPPIV), and insulin degrading enzyme (IDE)7-8, alt-

hough it is possible that there could be additional plasma enzymes involved since it was reported that other peptidoforms of BNP were found in human plasma9. Recent attempts to treat heart failure by the activation of the protein kinase G pathways have included manipulating BNP1-32 levels using drugs that inhibit proteolysis10. One example is the combined formulation of valsartan and the NEP-inhibitor Sacubitril, which may be superior to Enalapril-mediated angiotensin-converting enzyme inhibition to reduce hospitalization and mortality for patients with chronic heart failure with reduced ejection fraction.11 Alternatively, intravenous infusion of recombinant BNP1-32 can be used to treat acute decompensated heart failure (ADHF)12, but this strategy may increase the risk of deteriorating renal function in patients13. These data solicit questions concerning the relative influence of different BNP pepidoforms in plasma on the effectiveness of heart failure treatment, and these peptidoforms may in turn reveal important diagnostic and prognostic information. Addressing these questions is challenging for several important reasons; 1) Routine clinical analysis of BNP is performed by immunoassays14 that do not necessarily distinguish between peptidoforms or account for cleaved or alternatively modified peptides; 2) Sampled plasma contains intact catabolic and processing steps for BNP, but it is devoid of any input of newly secreted BNP. Sampling therefore disrupts the homeostatic equilibrium, and the length and conditions of sample storage upstream of BNP analysis represent an important and often overlooked variable15. Like many bioactive polypeptides, multiple plasma enzymes are known to cleave circulating BNP, and as next part of the CVD-B/D-HPP mandate4, our goal isto develop a mass spectrometry (MS) based method capable of identifying and profiling the formation of BNP peptidoforms in

ACS Paragon Plus Environment

Journal of Proteome Research

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

plasma samples. Specifically, focusing on obtaining the degradation kinetics of intact BNP1-32. We hypothesize that the composite of endogenous plasma proteases involved in the processing of endogenous BNP is reflected by the proteolysis of exogenously added BNP1-32. The addition of recombinant BNP1-32 to plasma is straightforward, simplifies sample preparation and analysis, and overcomes sensitivity challenges inherent to plasma matrices. Furthermore, we propose that changes in the kinetics to one or more of the proteases may occur with disease and that these alterations will change the proteolytic profile of exogenously added BNP1-32. By developing a technique that can account for BNP peptidoforms, we aim to help clarify the extent to which BNP is degraded by endogenous factors in plasma with the ultimate goal of providing more accurate diagnostic and prognostic parameters for heart failure, as well as a method for monitoring BNP bioavailability and assessing overall BNP-degrading enzyme activity. We identified several requirements and impediments to the development of such a method. First, in order to avoid skewing the proteolytic profile by disfavoring or outright elimination of potential peptidoforms during sample preparation, the method must be able to preserve peptidoform diversity. In the specific case of BNP proteolysis, several enzymes are known to serially process multiple peptidoforms. This scenario, in which an analyte can serve as a substrate and a product that can be continuously generated and subsequently proteolyzed, can be analyzed by the well-established labeling techniques developed to characterize protein processing, such as COFRADIC16-17 and TAILS18. However, the processing time involved with these strategies may not be feasible for routine clinical analysis. Accordingly, we sought to build a technique with minimal sample preparation. Second, since clinically relevant BNP concentrations can be below 100pg/µL, we anticipate that the direct measurement of endogenous BNP cleavage peptidoforms would require analytical sensitivity into the low to sub-ng/µL range. Third, a method for endogenous BNP proteolytic profiling would have to achieve sufficient sensitivity while simultaneously overcoming the inherently complexity of plasma, which is often achieved through an enrichment step. Immunoprecipitation-based sample enrichment strategies represent the obvious approach9, 19, but these strategies can skew proteolytic profiles by disfavoring peptidoforms with potentially altered epitopes. Furthermore, enrichment strategies and similarly complex sample preparation steps lengthen protocols, increase variance, and are unlikely to ultimately be adopted for routine clinical analyses. In this study, we present an alternative approach for BNP proteolytic profiling based on using neutral-coated capillary electrophoresis with electrospray ionization in a single unit that is directly coupled to a mass spectrometry (CESI-MS), and that can measure individual BNP cleavage peptidoforms as they are generated over time in minimally processed plasma. In our approach, standard exogenous BNP1-32 is spiked into a plasma sample, where endogenous peptidases cleave BNP1-32 into its peptidoforms, which are then detected and profiled over time. This reaction can proceed in a CE sample vial that can be sampled for analysis at any desired time interval. By integrating multisegment injection (MSI), where multiple samples are sequentially introduced into a single capillary and separated by a buffer spacer for simultaneous analysis20, our method can allow for the paral-

Page 2 of 11

lel analysis of multiple plasma samples where successive CESI-MS runs providing a time course for BNP proteolytic profiling. Similarly, MSI can be used to produce a multipoint BNP proteolytic profile from one plasma sample across the protracted timeframe of a single CESI-MS run. Materials and Methods Recombinant human BNP1-32 was purchased from SigmaAldrich Cat#B5900, and dissolved in Optima grade water (Fisher Scientific W6500) at 2.5mg/mL. These BNP1-32 standards were stored at -80°C in 10µL aliquots prior to use. Artificial plasma consisted of 2.25g bovine serum albumin (Recho Ref#: 03117332001) dissolved in 50mL 1x PBS pH 7.4 (Quality Biological Cat#:119-069-131, Lot#: 720744) with 1 tablet of protease inhibitors (Thermo Scientific Cat#:88266). Human plasma was purchased from Bioreclamation, including human heparin plasma (Cat#: HMPLNAHP, Lot#: BRH181304) and human EDTA plasma (Cat#: HMPLEDTA, Lot#: BRH1120184). The 4 BNP peptidoform standards (BNP3-32, BNP3-29, BNP1-30 and BNP3-30) were synthesized by Synpel Chemical. Plasma sample preparation. All plasma samples were centrifuged through a 0.22µm spin filter (E&K scientific, Cat#: EK-680850) for 15 minutes at 16100g. the filtered plasma was stored at -80°C in 10µL aliquots. The plasma aliquots were thawed on ice immediately prior to a CE-MS experiment, and mixed with a designated BNP1-32 solutions to achieve a final BNP1-32 concentration of 250ng/µL (unless the concentration is otherwise stated). Capillary Electrophoresis and Mass Spectrometry. CE System: CE experiments were carried out using a CESI 8000 High Performance Separation-ESI Module (Sciex Separations, Brea, CA). The capillary temperatures were maintained at 25°C. The capillary used in this study was the OptiMS Neutral Surface Cartridge (Sciex Separations, Brea, CA). Prior to use, the capillary was first washed by 0.1M hydrochloric acid (Sigma-Aldrich, Cat#258148), then rinsed with background electrolyte (BGE) consisting of 10% acetic acid (Fisher Scientific, Cat#: A38-500), and finally rinsed with deionized water for 30 min at 100 psi, stored overnight filled with water. Before each run, the capillary was rinsed with 0.1M HCl and flushed with fresh BGE for 10 minutes at 100 psi. Unless otherwise stated, samples were injected by 10kV voltage for 5 secs and the BGE spacer was added between samples by hydrodynamic injection. A separation voltage of 30 kV was applied across the capillary with a supplemental forward pressure of 1.5 psi. Mass Spectrometry. CESI−MS experiments were performed using a Q Exactive+ mass spectrometer (Thermo Fisher Scientific, San Jose, USA). The electrospray voltage used was 1.8kV. Data were acquired with automatic gain control of 3x106 and a maximum injection time of 100msec. The scan range was set to 200-1200m/z. The MS resolution was set to 70K and 17.5K for the full MS1 and the DD-MS 2 scans, respectively, the default charge was 4 and NCE 30eV for fragmentation. Data Analysis. We performed peptide mapping analysis for the identification and confirmation of BNP peptidoforms using Biopharmfinder 1.0 SP1 software (Thermofisher Sci-

ACS Paragon Plus Environment

Page 3 of 11

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

Journal of Proteome Research

entific, San Jose, USA). Accurate MS1 quantitation of identified peptidoforms was accomplished using Tracefinder 3.1 (Thermofisher Scientific, San Jose, USA).

Results & Discussion Our goal for this study was to develop a mass spectrometry-based method capable of reproducibly analyzing a proteolysis profile of BNP1-32 derived in a plasma matrix, with an overarching longer-term vision of providing a deeper understanding of heart failure and treatment. To do this, we added exogenous BNP1-32 to plasma and observed the disappearance of the longer peptide and the formation of specific truncated peptides over time. In keeping with our goal, we identified the following constraints that our method would have to overcome. First, because a bottom-up proteomics approach21 could directly impact the diversity of the cleavage peptidoforms we seek to detect, our method necessitated analysis without trypsin digestion. Second, we estimated the concentration of endogenous BNP peptidoforms in plasma to be in the sub-pg/µL range22, which our initial screening experiments suggested was within the reach of nano-flow LCMS for the analysis of standard BNP1-32 reconstituted in a clean aqueous matrix (data not shown). Unfortunately, achieving a similar level of sensitivity with sufficient accuracy and reproducibility in plasma samples likely requires enrichment strategies with extensive sample cleanup and/or fractionation. Finally, while catabolic processes in plasma remain active, sources for secretion of newly synthesized BNP1-32 are absent once blood is drawn. Therefore, the accuracy and reproducibility for endogenous BNP peptidoform analysis may be secondary to the disruption inherent to sample extraction.

Table 1. Theoretical pI values for putative BNP peptidoforms and known plasma proteolytic enzymes with UniProtKB accession number. Protein name BNP1-32 (active part of P16860)

Theoretical pI 12.14

Neutral endopeptidase (P08473)

5.54

Dipeptydilpeptidase IV (P27487)

5.67

Insulin degrading enzyme (P14735)

6.16

of endogenous BNP peptidoforms, we can assume that plasma enzymes are saturated with BNP. We based our method on a CESI-MS platform for several key reasons: First, all putative BNP peptidoforms are small peptides that share a common central loop motif and a highly alkaline theoretical pI. Thus, BNP1-32 carries a positive charge across a wide pH range (Table 1). This alkaline pI that is inherent to BNP peptidoforms endows them with an electrophoretic mobility that exceeds a majority of other plasma proteins and helps minimize interfering signals and plasma matrix effects. Second, the low sample consumption of a given CE run provides an opportunity for multiple successive sampling at different time points from a single vial. Third, CE methods can be built to include MSI, where multiple sample injections separated by short background electrolyte spacers can be simultaneously run and analyzed. The increase in throughput afforded by MSI is an especially attractive feature where the analysis of multiple time points, potential enzyme kinetics, and/or larger clinical cohorts are concerned. Our experimental strategy involved the following approach: 1) Profile the formation of exogenous BNP peptidoforms in plasma by CESI-MS; 2) Analyze the effect on the BNP proteolysis profile in plasma with different plasma storage conditions 3) Establish and validate reproducibility and accuracy of MSI for creation a BNP standard curve; 4) Explore the application of MSI technique in one-hour BNP proteolytic kinetic profile (5 time points). Profile the generation of BNP peptidoforms in plasma by CESI-MS. Eleven putative BNP cleavage peptidoforms have previously been described in the literature9. These are derived from at least three known BNP-specific proteolytic enzymes: NEP, DPPIV and IDE8. Table 1 summarizes the theoretical isoelectric point of all putative BNP peptidoforms in contrast with those of the BNP proteolytic enzymes. Given that our goal was direct sampling from plasma, where the pH falls between the theoretical pI of BNP and that of the proteolytic enzymes, we built our CE method with electrokinetic sample injection. The use of electrokinetic injection in a neutral pH sample endows our method with the ability to selectively introduce high pI analytes, while excluding the lower pI catabolic enzymes responsible for BNP degradation. Therefore, enzymatic reactions can proceed uninterrupted in the sample vial and not the capillary, and selective sampling simultaneously serves as a cleanup step to decrease matrix effect and analysis complexity.

Our approach aims to simultaneously circumvent these constraints by spiking exogenous standard BNP1-32 in plasma, followed by direct sampling and analysis of the BNP peptidoforms that result from those intact catabolic processes. By spiking a high concentration of exogenous BNP1-32 into plasma that is several orders of magnitude above those

ACS Paragon Plus Environment

Journal of Proteome Research

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 4 of 11

Figure 1. Enzymatic proteolysis curves of recombinant BNP peptidoforms. Simultaneous profiling of five peptidoforms from 20 consecutive CESI-MS runs sampled after spiking in 250ng/µL of BNP1-32 into a plasma. Panel A depicts representative electropherograms of the total ion count (TIC), BNP1-32, BNP3-32, BNP3-29, BNP1-30 and BNP3-30. We extracted the accurate MS1 spectra of the most abundant form of each BNP peptidoforms: BNP1-32 with 5+ charge, BNP3-32 with 5+, BNP3-29 with 4+ charge, BNP1-30 with 5+ charge and BNP3-30 with 5+ charge. Panel B depicts the time course profile of each peptidoform as individual peak areas over a total of 14hrs post spiking in BNP (n=3). Note: Peaks below the quantitative threshold of 5e5 could not be accurately quantified and were excluded.

Capillary electrophoresis was performed using commercially available neutral CESI capillaries23-24, which circumvents the problems of adsorption and interaction of basic BNP peptidoforms and other intact cationic plasma proteins with the anionic silanol groups that constitute the inner surface of uncoated fused silica capillaries. After spiking recombinant BNP1-32 in EDTA plasma, the sample was placed in the CESI instrument auto-sampler at 25oC, and injected sequentially for CESI-MS analysis in 34 min intervals for 657 min. Thus producing a time-dependent profile of the number and quantity of each BNP peptidoform. We observed a clear time-dependent decrease of BNP1-32 and a reciprocal appearance and increase of secondary peaks. We used BioPharma Finder 1.0 Mass Informatics Software (ThermoFisher) to identify these new product peaks as BNP cleavage peptidoforms. Ultimately, successive sampling across 657min from an individual reaction vial enabled the quantitative profiling of five enzymatically generated BNP peptidoforms in plasma, BNP1-32, BNP3-32, BNP3-29, BNP1-30, and BNP3-30. Although these peptidoforms differ by only a few amino acids, the resolution of CE achieved baseline separation. To confirm the identities and assist in the further characterization of these five peptidoforms, we synthesized their standards to

generate reference spectra. We reconstituted each of these synthetic peptides in water and analyzed them with CESIMS, extracting the MS1 and MS2 spectra (Supporting Figure 1-5) to compare with the ones generated in plasma samples. Moreover, we chose the charge state of each BNP peptidoforms with the highest abundance for the following quantitation. The reproducibility of those peaks were evaluated, which showed that the CVs were all under 30% (n=4) (Supporting Table 1). Our results indicate that the proteolysis of exogenous BNP1-32 occurs rapidly, with a signal decaying to below our detection limit within 100min (Figure 1). The difference in the timing of the appearance of several peptidoforms suggests that products of one cleavage reaction serve as substrates for other reactions. For example, BNP3-32 is the primary product of DDPIV-mediated cleavage of BNP1-32, and its abundance increases quickly within 1 hour after BNP1-32 is spiked in. Prolonged incubation produces a gradually decreasing peak of BNP3-32, while another BNP peptidoform, BNP3-29 is reciprocally generated, presumably by IDEmediated cleavage of BNP3-32. Thus, we propose that the simultaneous measurement of multiple BNP peptidoforms over time best captures the proteolytic capacity of individual patients.

ACS Paragon Plus Environment

Page 5 of 11

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

Journal of Proteome Research

Figure 2: Comparison of three BNP proteolysis profile in identical EDTA plasma with three different storage conditions. Commercially purchased human plasma collected in EDTA tubes was aliquoted and stored in -80oC then thawed on ice to give the same baseline activity for 3 aliquots. One of them was analyzed directly after thawed, while the other two continued incubate at room temperature and 4oC for another 24 hours. Peak areas are all normalized by TIC to reduce CE-MS error.

Effect of storage conditions on the exogenously added BNP1-32 proteolysis profile. To evaluate the storage conditions on the BNP proteolysis profile, we compared three different temperatures; storage of the plasma at 4oC for 24 hours, at room temperature for 24 hours and direct analysis without storage. Human EDTA plasma was aliquoted and stored in -80oC to freeze the enzyme activities in plasma. After this baseline treatment, three sets of aliquots were thawed. The first set were incubated at room temperature for 5min to achieve room temperature, and was considered as a substitution to simulate fresh plasma. The other two sets of aliquots were incubated at room temperature and at 4oC for another 24 hours respectively (Figure 2). Plasma samples were subsequently spiked with the BNP1-32 solution as per our established protocol, and analyzed by CESI-MS. Our results indicate that storage at 4oC and direct analysis result in similar profiles, indicating that cleavage capacity is largely unaffected. Conversely, 24hr incubation at room temperature alters these profiles, indicating that prolonged roomtemperature storage adversely affects proteolysis (Figure 2). The altered kinetics we observed with prolonged room temperature sample storage also suggest the importance of storage conditions for other BNP-related analyses. Assessing reproducibility and accuracy of multisegment injection using a recombinant BNP1-32 standard. To increase the throughput of this CE-MS analysis, we applied a multisegment injection to our method. We estab-

lished baseline parameters by developing a method for CESI-MS analysis of standard BNP1-32 dissolved in water. This sample was electrokinetically25 injected at 5kV for 10 seconds into a capillary filled with a BGE of 10% acetic acid, and a separation voltage of 30kV supplemented with a forward pressure of 0.5psi was applied throughout the run. The buffer spacer between every sample injection is the same BGE with hydrodynamic injection for 3min. We tested the performance of our CESI-MS method by simultaneously analyzing peak areas (Table 2) and migration times (Table 3) for four successive runs consisting of five MSI segments each (Figure 3). These experiments resulted in an average percent coefficient of variance (%CV) between runs of 7% (n=4), and an average %CV between segments within a run of 21.6% (n=5). The overall average %CV for the combination of all segments and all runs was 20%. We found the migration time of the multiply injected peaks, the ESI spray, and the MS signal throughout the CE run to be stable and reproducible. We applied our method to analyze a dilution series of recombinant BNP1-32 and generated a linear 5-point standard curve in a single run with five MSI sample injections (Figure 4). The concentration range of BNP1-32 for this curve was between 50 ng/µL to 250 ng/µL, and all concentrations produced well resolved Gaussian peaks. All five concentration points resulted in CVs under 15% (n=3) and accuracies between 89%-108%, with the exception of the lowest concentration (50ng/µL) whose accuracy was 130% (Table 4).

ACS Paragon Plus Environment

Journal of Proteome Research

Figure 3. Reproducibility of 5 MSI runs of exogenous BNP proteolyzed in plasma. The overlaid total ion current electropherograms of four experiments, each consisting of 5 MSI segments from recombinant BNP1-32 protein in water (250ng/µL). Intra-run reproducibility was assessed by comparing individual MSI segments within a run, while Inter-run reproducibility was assessed by comparing the same peak across four successive runs. Segments were injected in 3 min intervals within each run, and successive runs were performed at 1hr intervals.

Table 2. Intra-run and inter-run peak area of BNP1-32 by CESI-MS with MSI Peak Area Run

MSI 1

MSI 2

MSI 3

MSI 4

MSI 5

%CV (intrarun)

1

3.4E+08

3.4E+08

3.3E+08

2.9E+08

2.9E+08

8%

2

3.7E+08

3.5E+08

3.6E+08

3.4E+08

3.1E+08

7%

3

3.1E+08

3.3E+08

3.3E+08

3.0E+08

2.8E+08

6%

4

2.2E+08

2.0E+08

2.2E+08

1.9E+08

1.8E+08

7%

%CV (inter-run)

22%

23%

20%

22%

21%

Table 3. Intra-run and inter-run migration time of BNP1-32 of CESI-MS with 5 MSI Migration time (min) Run

MSI 1

MSI 2

MSI 3

MSI 4

MSI 5

1

4.02

6.18

8.39

10.61

12.82

2

4.00

6.20

8.52

10.68

12.92

3

4.06

6.23

8.45

10.67

12.86

4

4.08

6.32

8.56

10.73

12.97

Std dev (inter-run)

0.04

0.06

0.08

0.05

0.07

%CV (all runs included)

20%

8.0×10 8

Peak area of BNP1-32

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 6 of 11

6.0×10 8 4.0×10 8 2.0×10 8 0 0

100

200

300

BNP concentration (ng/µ µl)

Figure 4. BNP dilution curve in plasma (n=3). Calibration curve produced from three separate CESI-MS runs with MSI, each consisting of five segments of increasing

ACS Paragon Plus Environment

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

Journal of Proteome Research

recombinant BNP1-32 concentrations separated by a background electrolyte spacer. Curve includes all data points (mean +/-Std Dev). Table 4. Reproducibility and Accuracy of five BNP1-32 calibration curves constructed as 5 MSI per curve (n=3) BNP1-32 Concentration (ng/µL)

%CV

% Accuracy*

50

12%

130%

100

13%

101%

150

15%

89%

200

10%

92%

250

13%

108%

*% accuracy is defined as the quantitative value of each peak from the calibration curve relative to its theoretical concentration, and is calculated using the average of replicates at the same concentration level.

Applying CESI-MS with MSI to determine the freezethaw stability of BNP proteolysis assay. Our results in table 2 indicate that a sample that is successively analyzed in separate CESI-MS runs has a greater amount of variance than the same sample simultaneously analyzed by five MSI segments. For this reason, we applied MSI to simultaneously evaluate the effect of up to 5 freeze-thaw cycles on the enzymatic cleavage capacity of plasma by quickly measuring and comparing the ratio of BNP3-32-to-BNP1-32 in an overnight sequence. Segments were electrokinetically injected into the capillary in order of decreasing freeze-thaw cycles. Aside from the initial plasma preparation, the incubation and successive overnight analyses were performed within and by the CE instrument as part of a sequence protocol. This experiment (Figure 5) demonstrated that three or less freeze-thaw cycles did not result in significant differences in BNP1-32 proteolysis profiles in plasma. However, when the plasma underwent more than 3 freeze-thaw cycles, the ratio of BNP3-32-to-BNP1-32 was increased in every timepoint, indicating that those enzymes using BNP3-32 as a substrate are more susceptible to freeze-thaw cycles than those using BNP1-32 as a substrate. Regardless of individual enzyme susceptibility, accurate proteolysis profiles appear to depend on minimizing the number of freeze-thaw cycles. MSI enabled us to simultaneously acquire BNP proteolysis profiles for 5 different samples. We envision that this type of throughput increase can similarly accelerate the analysis of large-scale studies.

Peak area ratio of BNP 3-32 to 1-32

Page 7 of 11

Ratio of BNP 3-32 to 1-32 8 4

1 cycle 2 cycles 3 cycles 4 cycles 5 cycles

2 1 0.5 0.25 0.125 0.0625 0.03125 0

5

10

15

20

Time (h)

Figure 5: Ratio changes of BNP3-32 and BNP1-32 in the plasma samples with different freeze-thaw cycles. Commercial pooled plasma was aliquoted and frozen at 80oC. Each freeze-thaw cycle, the aliquot was taken out and stored at room temperature for 30min then put back to -80oC for 30min. The 5 samples with 1 to 5 freeze-thaw cycle each were spiked in recombinant BNP simultaneously, then injected sequentially into capillary as one segment, with a BGE spacer in between. All the samples were injected and analyzed every hour by CE-MS. The peak areas of BNP3-32 and BNP1-32 of every run was measured for the comparison of their proteolysis profile.

One-hour exogenous BNP proteolytic profiling using CESI-MS with MSI. The prospective application of our proteolytic profiling technique for discovery and research cohorts, warrants the quantitative detection and profiling of as many peptidoforms as possible. The disadvantage inherent to a comprehensive characterization of five BNP cleavage peptidoforms is its requirement for an overnight protocol to allow sufficient reaction time for the generation of slower-forming peptidoforms. We propose that incorporating MSI into the method can increase throughput by enabling the simultaneous analysis of multiple different plasma samples within a single MS run, thereby achieving a faster turnaround. Conversely, we also applied MSI orthogonally to sample a single individual reaction vial every 3 minutes for a total of 5 closely spaced time points. This iteration of our method can provide an acute quantitative profile of primary BNP proteolysis products, which we depict as the ratio of BNP3-32:BNP1-32, from plasma in less than one hour, including sample preparation. Although this 1hr method under current conditions can detect additional peptidoforms, the short reaction time precludes a reliable signal with sufficient intensity for their quantitative analysis. Alternatively, adjusting the plasma dilution ratio may facilitate the quantitative analysis of slower-forming BNP peptidoforms within this 1hr timeframe.

ACS Paragon Plus Environment

Journal of Proteome Research

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 6. Proteolysis BNP profile is suppressed in Heparin verse EDTA plasma. Proteolytic profiles of exogenous BNP1-32 (250 ng/µL) spiked into 10x diluted EDTA, Heparin, and artificial plasma using neutral-coated CESIMS with MSI (n=5, each). Samples were incubated on the CE auto-sampler at 25oC for 18 min prior to injection of BNP peptidoforms using MSI with 3 min intervals (5 segments per run). Data points depict mean +/- Std Dev lines of best fit reflect quadratic nonlinear regression with 95 % CI. EDTA and heparin collection tubes are commonplace for plasma sampling in the hospital setting. Buckley et al26 suggested EDTA collection tubes be employed where BNP analysis from plasma is concerned when using traditional radioimmunoassay-based methods. While our results clearly indicate that the endogenous plasma proteases retain their ability to cleave BNP1-32 after plasma collection, our study was not designed to assess the extent to which protease activities alter the quantitative accuracy of immunoassay-based methods. Neither was this study designed to assess the suitability of various sampling tubes as they pertain to CESI-MS-based profiling of BNP proteolysis. Nevertheless, our assay provides a measure of the outcome of all plasma proteases that degrade BNP, and provides a biological readout of the underlying physiology or pathophysiology.

Furthermore, as an initial application for our 1-hr detection method, we performed a pilot side-by-side analysis to compare human plasma collected in either Heparin tubes or EDTA tubes (Figure 6). In one hour, our method clearly describes a difference in the profiles of BNP1-32:BNP3:32 between plasma collected in these two common tubes. Among the enzymes known to cleave BNP, both IDE and NEP are known to use Zn2+ as a cofactor27-29. IDE can cut the last 3 amino acids from the C-terminal of BNP while NEP can cleavage between the fourth and fifth amino acids from N-terminal in reactions required to produce the primary proteolysis products of BNP8. We have previously detected (but not quantified) BNP3-29 in plasma from heparin collection tubes within an hour, but not

Page 8 of 11

from EDTA tubes which require substantially longer incubation times (data not shown). We speculate that the activity of IDE may be partially inhibited by EDTA due to its chelation of Zinc ions. The absence of BNP1-32 degradation was the product of previously described enzymatic activities in plasma samples, we measured the time-dependent ratio of BNP1-32:BNP3-32 of the BNP1-32 spiked into a commercially available human plasma matrix along with the same ratio that is also spiked into an artificial plasma matrix consisting of albumin dissolved in PBS. In contrast with the human plasma matrix, a small amount of BNP3-32 was consistently detected in the artificial plasma group, but the BNP1-32:BNP3-32 ratio was stable (Figure 6). We did not detect any signals for BNP3-29, BNP1-30, or BNP3-30 in artificial plasma across the timespan of this experiment, and we speculate that the residual BNP3-32 peak represents a low-level degradation product. However, the data confirms that the degradation of BNP1-32 that we detect in human plasma is enzyme mediated.

Conclusion Even though BNP is one of the most studied cardiovascular disease-related protein, there are currently limited means to accurately measure its biological processing in blood. In this study, we present a novel approach in which exogenous BNP132 is added to plasma where it undergoes processing by endogenous proteases to form specific peptidoforms which are profiled using CESI-MS. In our approach, standard exogenous BNP1-32 is spiked into a plasma sample, where it is cleaved by endogenous plasma peptidases to produce various product peptidoforms. This reaction can proceed in a temperaturecontrolled sample vial, and sampled into a capillary using electrokinetic injection at any desired time interval. We present the first use of MSI with CESI-MS using a neutral coated capillary for proteolysis analysis in plasma. The use of MSI can endow this method with increased throughput via parallel analysis of multiple plasma samples. In this case, each parallel sample is represented by an individual sequentially injected segment, while successive CESI-MS runs provide a time course profile for the generation and degradation of BNP peptioforms across an entire CE sequence. Similarly, we applied MSI orthogonally to the same sample, producing a multi-point BNP peptidoform profile from an individual plasma sample across the protracted timeframe of a single CESI-MS run. This second iteration of our method produces a 5-point profile in under an hour, including all sample preparation steps. In the future, we envision that this technique could be used to study the proteolysis of other bioactive protein or peptide hormones (e.g. complement factor C330, endostatin31, and fibrillin 132), and to assess the bioavailability of other peptide drugs in plasma.

ACS Paragon Plus Environment

Page 9 of 11

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

Journal of Proteome Research

AUTHOR INFORMATION Corresponding Author *email: [email protected]

Author Contributions The manuscript was written through contributions of all authors. / All authors have given approval to the final version of the manuscript.

ACKNOWLEDGMENT The authors wish to acknowledge Sciex Separations for their support with all CESI methods, and neutral coated cartridges; ThermoFisher Scientific for QExactive plus and access to data analysis software support. This work was supported by Erika Glazer Endowed Chair for Women’s Heart Health, NIH-NHLBI (P01HL112730), Department of Defense (W81XWH-16-10592) and the AHA challenge grant (#15GPSGC24470098, subaward #566963).

REFERENCES 1. Murray, C. J.; Lopez, A. D., Alternative projections of mortality and disability by cause 1990-2020: Global Burden of Disease Study. Lancet 1997, 349 (9064), 1498-504. 2. Suzuki, T.; Lyon, A.; Saggar, R.; Heaney, L. M.; Aizawa, K.; Cittadini, A.; Mauro, C.; Citro, R.; Limongelli, G.; Ferrara, F.; Vriz, O.; Morley-Smith, A.; Calabro, P.; Bossone, E., Editor's Choice-Biomarkers of acute cardiovascular and pulmonary diseases. Eur Heart J Acute Cardiovasc Care 2016, 5 (5), 416-33. 3. Simons, J. E.; Don-Wauchope, A. C., Evaluation of natriuretic peptide recommendations in heart failure clinical practice guidelines. Clin Biochem 2016, 49 (1-2), 8-15. 4. Lam, M. P.; Venkatraman, V.; Xing, Y.; Lau, E.; Cao, Q.; Ng, D. C.; Su, A. I.; Ge, J.; Van Eyk, J. E.; Ping, P., Data-Driven Approach To Determine Popular Proteins for Targeted Proteomics Translation of Six Organ Systems. J Proteome Res 2016, 15 (11), 41264134. 5. Hall, C., Essential biochemistry and physiology of (NTpro)BNP. Eur J Heart Fail 2004, 6 (3), 257-60. 6. O'Donoghue, M.; Chen, A.; Baggish, A. L.; Anwaruddin, S.; Krauser, D. G.; Tung, R.; Januzzi, J. L., The effects of ejection fraction on N-terminal ProBNP and BNP levels in patients with acute CHF: analysis from the ProBNP Investigation of Dyspnea in the Emergency Department (PRIDE) study. J Card Fail 2005, 11 (5 Suppl), S9-14. 7. Norman, J. A.; Little, D.; Bolgar, M.; Di Donato, G., Degradation of brain natriuretic peptide by neutral endopeptidase: species specific sites of proteolysis determined by mass spectrometry. Biochem Biophys Res Commun 1991, 175 (1), 22-30. 8. Volpe, M.; Rubattu, S.; Burnett, J., Jr., Natriuretic peptides in cardiovascular diseases: current use and perspectives. Eur Heart J 2014, 35 (7), 419-25. 9. Niederkofler, E. E.; Kiernan, U. A.; O'Rear, J.; Menon, S.; Saghir, S.; Protter, A. A.; Nelson, R. W.; Schellenberger, U., Detection of endogenous B-type natriuretic peptide at very low concentrations in patients with heart failure. Circ Heart Fail 2008, 1 (4), 258-64. 10. McMurray, J. J.; Packer, M.; Desai, A. S.; Gong, J.; Lefkowitz, M. P.; Rizkala, A. R.; Rouleau, J. L.; Shi, V. C.; Solomon, S. D.; Swedberg, K.; Zile, M. R.; Investigators, P.-H.; Committees, Angiotensin-neprilysin inhibition versus enalapril in heart failure. N Engl J Med 2014, 371 (11), 993-1004. 11. Fala, L., Entresto (Sacubitril/Valsartan): First-in-Class Angiotensin Receptor Neprilysin Inhibitor FDA Approved for Patients with Heart Failure. Am Health Drug Benefits 2015, 8 (6), 330-4. 12. O'Connor, C. M.; Starling, R. C.; Hernandez, A. F.; Armstrong, P. W.; Dickstein, K.; Hasselblad, V.; Heizer, G. M.; Komajda, M.; Massie, B. M.; McMurray, J. J.; Nieminen, M. S.; Reist, C. J.; Rouleau, J. L.; Swedberg, K.; Adams, K. F., Jr.; Anker, S. D.; Atar, D.; Battler, A.; Botero, R.; Bohidar, N. R.; Butler, J.; Clausell, N.;

Corbalan, R.; Costanzo, M. R.; Dahlstrom, U.; Deckelbaum, L. I.; Diaz, R.; Dunlap, M. E.; Ezekowitz, J. A.; Feldman, D.; Felker, G. M.; Fonarow, G. C.; Gennevois, D.; Gottlieb, S. S.; Hill, J. A.; Hollander, J. E.; Howlett, J. G.; Hudson, M. P.; Kociol, R. D.; Krum, H.; Laucevicius, A.; Levy, W. C.; Mendez, G. F.; Metra, M.; Mittal, S.; Oh, B. H.; Pereira, N. L.; Ponikowski, P.; Tang, W. H.; Tanomsup, S.; Teerlink, J. R.; Triposkiadis, F.; Troughton, R. W.; Voors, A. A.; Whellan, D. J.; Zannad, F.; Califf, R. M., Effect of nesiritide in patients with acute decompensated heart failure. N Engl J Med 2011, 365 (1), 32-43. 13. Sackner-Bernstein, J. D.; Skopicki, H. A.; Aaronson, K. D., Risk of worsening renal function with nesiritide in patients with acutely decompensated heart failure. Circulation 2005, 111 (12), 1487-91. 14. Clerico, A.; Emdin, M., Diagnostic accuracy and prognostic relevance of the measurement of cardiac natriuretic peptides: a review. Clin Chem 2004, 50 (1), 33-50. 15. Buckley, M. G.; Marcus, N. J.; Yacoub, M. H., Cardiac peptide stability, aprotinin and room temperature: importance for assessing cardiac function in clinical practice. Clin Sci (Lond) 1999, 97 (6), 689-95. 16. Staes, A.; Van Damme, P.; Helsens, K.; Demol, H.; Vandekerckhove, J.; Gevaert, K., Improved recovery of proteomeinformative, protein N-terminal peptides by combined fractional diagonal chromatography (COFRADIC). Proteomics 2008, 8 (7), 136270. 17. Van Damme, P.; Van Damme, J.; Demol, H.; Staes, A.; Vandekerckhove, J.; Gevaert, K., A review of COFRADIC techniques targeting protein N-terminal acetylation. BMC Proc 2009, 3 Suppl 6, S6. 18. Kleifeld, O.; Doucet, A.; Prudova, A.; auf dem Keller, U.; Gioia, M.; Kizhakkedathu, J. N.; Overall, C. M., Identifying and quantifying proteolytic events and the natural N terminome by terminal amine isotopic labeling of substrates. Nat Protoc 2011, 6 (10), 1578611. 19. Chappell, D. L.; Lee, A. Y.; Bernstein, H. S.; Lassman, M. E.; Laterza, O. F., Development and validation of an IA-LC/MS method to quantitate active and total B-type natriuretic peptide in human plasma. Bioanalysis 2016. 20. Kuehnbaum, N. L.; Kormendi, A.; Britz-McKibbin, P., Multisegment injection-capillary electrophoresis-mass spectrometry: a high-throughput platform for metabolomics with high data fidelity. Anal Chem 2013, 85 (22), 10664-9. 21. Zhang, Y.; Fonslow, B. R.; Shan, B.; Baek, M. C.; Yates, J. R., 3rd, Protein analysis by shotgun/bottom-up proteomics. Chem Rev 2013, 113 (4), 2343-94. 22. Maisel, A., B-type natriuretic peptide levels: diagnostic and prognostic in congestive heart failure: what's next? Circulation 2002, 105 (20), 2328-31. 23. Sun, L.; Knierman, M. D.; Zhu, G.; Dovichi, N. J., Fast topdown intact protein characterization with capillary zone electrophoresiselectrospray ionization tandem mass spectrometry. Anal Chem 2013, 85 (12), 5989-95. 24. Neuberger, S.; Rafai, A.; Neususs, C., Screening of Small Intact Proteins by Capillary Electrophoresis Electrospray IonizationMass Spectrometry (CE-ESI-MS). Methods Mol Biol 2016, 1466, 43-56. 25. Hirokawa, T.; Okamoto, H.; Gas, B., High-sensitive capillary zone electrophoresis analysis by electrokinetic injection with transient isotachophoretic preconcentration: electrokinetic supercharging. Electrophoresis 2003, 24 (3), 498-504. 26. Belenky, A.; Smith, A.; Zhang, B.; Lin, S.; Despres, N.; Wu, A. H.; Bluestein, B. I., The effect of class-specific protease inhibitors on the stabilization of B-type natriuretic peptide in human plasma. Clin Chim Acta 2004, 340 (1-2), 163-72. 27. Shen, Y.; Joachimiak, A.; Rosner, M. R.; Tang, W. J., Structures of human insulin-degrading enzyme reveal a new substrate recognition mechanism. Nature 2006, 443 (7113), 870-4. 28. Oefner, C.; Roques, B. P.; Fournie-Zaluski, M. C.; Dale, G. E., Structural analysis of neprilysin with various specific and potent inhibitors. Acta Crystallogr D Biol Crystallogr 2004, 60 (Pt 2), 392-6. 29. Oefner, C.; Pierau, S.; Schulz, H.; Dale, G. E., Structural studies of a bifunctional inhibitor of neprilysin and DPP-IV. Acta Crystallogr D Biol Crystallogr 2007, 63 (Pt 9), 975-81. 30. Manning, M. L.; Williams, S. A.; Jelinek, C. A.; Kostova, M. B.; Denmeade, S. R., Proteolysis of complement factors iC3b and C5 by the serine protease prostate-specific antigen in prostatic fluid and seminal plasma. J Immunol 2013, 190 (6), 2567-74.

ACS Paragon Plus Environment

Journal of Proteome Research

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

31. Ferreras, M.; Felbor, U.; Lenhard, T.; Olsen, B. R.; Delaisse, J., Generation and degradation of human endostatin proteins by various proteinases. FEBS Lett 2000, 486 (3), 247-51. 32. Hindson, V. J.; Ashworth, J. L.; Rock, M. J.; Cunliffe, S.; Shuttleworth, C. A.; Kielty, C. M., Fibrillin degradation by matrix metalloproteinases: identification of amino- and carboxy-terminal cleavage sites. FEBS Lett 1999, 452 (3), 195-8.

ACS Paragon Plus Environment

Page 10 of 11

Page 11 of 11

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

Journal of Proteome Research

For table of contents only

ACS Paragon Plus Environment