Online Preconcentration in Capillaries by Multiple Large-Volume

Subscriber access provided by READING UNIV

Article

On-line preconcentration in capillaries by multiple large-volume sample stacking: an alternative to immunoassays for quantification of amyloid beta peptides biomarkers in cerebrospinal fluid Cedric Crosnier de Lassichere, Thanh Duc Mai, Markus Otto, and Myriam Taverna Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b03843 • Publication Date (Web): 15 Jan 2018 Downloaded from http://pubs.acs.org on January 15, 2018

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.

Analytical Chemistry 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 38 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

On-line preconcentration in capillaries by multiple large-volume sample stacking: an alternative to immunoassays for quantification of amyloid beta peptides biomarkers in cerebrospinal fluid

Cédric Crosnier de Lassichère†, Thanh Duc Mai†, Markus Otto‡ and Myriam Taverna*† †

Institut Galien Paris Sud, UMR 8612, Protein and Nanotechnology in Analytical Science (PNAS),

CNRS, Univ. Paris-Sud, Univ. Paris-Saclay, 5 rue Jean Baptiste Clément, 92290 ChâtenayMalabry, France ‡

University of Ulm, Department of Neurology, Oberer Eselsberg 45, 89081 Ulm, Germany

Correspondence: E-mail: [email protected];

Fax: +33-1-46-83-54-62

Keywords: capillary electrophoresis, Alzheimer’s disease, stacking, preconcentration, Amyloid β peptides, biomarkers, CSF

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 38

ABSTRACT

A novel electrokinetic preconcentration approach, so-called multiple pressure-assisted large-volume sample stacking with an electroosmotic flow pump (M-PA-LVSEP), was developed to allow incapillary enrichment and separation of analytes from unlimited sample volumes. With this approach, the inherent limitation of in-capillary electrokinetic preconcentrations to the separation capillary volume can be overcome. The M-PA-LVSEP protocol relies on repeated cycles of pressure-assisted electroosmotic pumping and injection of extremely large sample volumes for analyte stacking and sample matrix removal. This technique was developed to address the challenge of sensitive and simultaneous determination of several amyloid β (Aβ) peptides, which are biomarkers for the molecular diagnosis of Alzheimer’s disease (AD). For the first time, reliable quantification of different species of fluorescently derivatized Aβ peptides, i.e. Aβ 1-42, Aβ 1-40 and Aβ 1-38 down to sub nM ranges in cerebrospinal fluids (CSF) from AD and AD non-demented patients

(healthy

controls)

was

made

possible

without

recourse

to

immunoassay,

immunoprecipitation or mass spectrometry approaches. Based on the stacking from a sample plug representing up to 400% of the total capillary volume, sensitive enhancement factors up to 800 could be achieved with this ‘antibody free’ approach. Quantification limits for these Aβ peptides down to 0.05 nM with capillary electrophoresis coupled with laser induced fluorescent detection could be obtained. Excellent agreement between results from M-PA-LVSEP and the gold standard ELISA method was achieved for measurements of Aβ 1-42 in CSF, with a determination correlation (r2) better than 0.993.


Page 3 of 38 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


INTRODUCTION

To overcome the inherent drawback of capillary electrophoresis (CE) in unsatisfied limits of detection, there has been a strong interest in the development of analyte enrichment strategies prior to CE operation. Efforts have been devoted to in-capillary, on-line and automated electrokinetic preconcentration techniques. An overview of such techniques can be found in some recent reviews14

and in a systematic guide.5 Their preconcentration performance is nevertheless limited by the fact

that technically it is not readily possible to inject more than one capillary volume of sample. Recently, with the goal of overcoming this limitation, our group has laid the groundwork for multiple cycles of electrokinetic preconcentration with repeated hydrodynamic injection of unlimited sample volumes.6 The concept was demonstrated for isotachophoresis (ITP), allowing an enrichment of a sample plug representing up to 300 % of the total capillary length after 9 ITP cycles. This approach, though very efficient, did not allow quantification of different preconcentrated analytes from biological samples as no separation was implemented after this step. Another electrokinetic technique with the advantage of combining both preconcentration and species separation is the large-volume sample stacking with an electroosmotic flow pump (LVSEP) coupled with capillary zone electrophoresis (CZE).7 Fundamentals of this technique and its recent applications can be found in a review

7

and some recent publications.8,9 The reported protocols

always stop after 1 cycle of LVSEP, the application of multiple cycles of LVSEP, remains so far unexplored. For diagnosis of Alzheimer’s disease (AD), the 42-amino acid long amyloid β (Aβ 1-42) peptide in cerebrospinal fluid (CSF) has been used as an established biomarker

10

as the level of Aβ 1-42

peptide is decreased in the CSF of AD patients.11 Nevertheless, its quantification in CSF is only one



Page 4 of 38

hallmark used in the molecular diagnosis of AD.12 To improve the differential diagnostic power between AD and healthy subjects or to discriminate better AD from other neurodegenerative diseases, several groups have proposed the use of Aβ 1-42 / Aβ 1-40 ratios or other combinations of various Aβ isoforms.13-17 From a practical aspect, the employment of Aβ 1-42 / Aβ 1-40 ratios help minimizing bias (adsorption of the peptides on tubes or tips; high inter-individual variation of global Aβ amyloid peptide concentration) thus contributing to increase diagnostic accuracy.18 Among the methods for tracing Aβ peptides in CSF, immunoassays, notably ELISA19,20, single molecule array (SiMoA)21,22 and multi-analyte profiling assay (Luminex xMAP)23,24, have been up to now the most practiced ones in clinical routine. Some considerations (if not drawbacks) should be nevertheless considered. First, cross reactions of different Aβ peptides with the antibodies employed may occur. In addition, while antibodies specific for Aβ 1-42 and Aβ 1-40 are readily available, this is not the case when expanding the immunoassays to other Aβ peptides found in biological fluids. Second, unsatisfactory inter-laboratory reproducibility of these tests is a major issue. While state of the art and costly ELISA instruments have been introduced for automated and / or routine analyses25, conventional ELISA is still widely practised, especially outside the hospital context. The fact that several operations are required for the immunocapture, washing and detection steps can lead to error accumulation.26,27 Last, accurate quantification is hindered by the difficulty to maintain the standard peptides used for calibration under their monomeric forms and avoid their in-vitro loss or oligomerisation / aggregation.28 Efforts to replace these antibody-based methods with electrokinetic approaches have been communicated for both capillary29-33 and microchip-based configurations.34-37 However, sample treatment and enrichment by immunocapture were still required in these cases to reach the detectable levels of Aβ peptides in CSF. This again raises the issue of antibody specificities /


Page 5 of 38 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


availability and capture efficiency. Recourse to mass spectrometry (MS) for determining these Aβ peptides is an attractive alternative that has been explored with a high detection performance.38-43. Immuno-precipitation employing antibodies is again very often used to reduce the matrix complexity prior to MS.41 The little accessibility (if not unavailability) for isotopically-labelled Aβ peptides used as internal standards for quantification, the need for intense sample pre-treatment(s) (immuno-precipitation in most cases) of biological samples to ensure the MS performance, together with high cost and requirement of deep instrumental and operational expertise are points to consider when employing this technique for AD diagnosis. Following our pioneering study on multiple ITP preconcentration6 which was an extension of our precedent work on simple ITP for Aβ peptide detection30, it is reported herein and for the first time the development of repeated cycles of LVSEP coupled with CZE for in-capillary enrichment and separation of several Aβ peptides from unlimited sample volumes. Our aim is to reach the picomolar concentration and to circumvent the use of antibodies. For this purpose, electrokinetic preconcentration has been realised for the first time with fluorescently labelled Aβ peptides (Aβ 138, Aβ 1-40, Aβ 1-42, Aβ 2-40 and Aβ 5-40). Accordingly, a new CZE method to analyse these derivatized Aβ peptides in a neutral capillary was developed to eliminate adsorption of these species onto the inner capillary wall. Pressurisation was employed to compensate for the decreased EOF during single and multiple LVSEP. Efficiency of the novel approach followed by CZE was demonstrated for quantification of different species of N-truncated fluorescently-labelled Aβ peptides, i.e. Aβ 1-42, Aβ 1-40 and Aβ 1-38, down to sub nM ranges in CSF from AD and nondemented patients (healthy controls). The new developed method was quantitatively compared to the classical ELISA one. The originality of our work relies on two aspects: i) the development of a novel multi-electrokinetic enrichment concept that allows to preconcentrate the analytes from an



Page 6 of 38

unlimited sample volume, which was not possible so far with any developed single-electrokinetic preconcentration approach; and ii) the ultra-sensitive detection of bioanalytes without recourse to any immunocapture or immunoaffinity steps.


Page 7 of 38 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


EXPERIMENTAL SECTION

Chemicals, reagents and samples Amyloid beta peptides Aβ 1-38, Aβ 1-40, Aβ 1-42, Aβ 2-40 and Aβ 5-40 were purchased from Eurogentec (Seraing, Belgium). Central nervous system (CNS) perfusion fluid used for the preparation of standard solutions was obtained from CMA/Microdialysis AB (Kista, Sweden). Sodium hydroxide (1 M) was obtained from VWR (Fontenay-sous Bois, France). Tricine (≥ 99 %), formic acid, acetic acid, 2-(N-morpholino)ethanesulfonic acid MES), phosphoric acid, 3-(Nmorpholino)propanesulfonic acid (MOPS), 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), 28.1 % ammonium hydroxide (m / V) and dimethyl sulfoxide (DMSO, 99.9 %) were all provided by Sigma (St. Louis, MO, United States). All buffers were prepared with deionized water and were filtered through a 0.22 µm membrane prior to use. The Fluoprobe 488 NHS ester was obtained from Interchim (Montluçon, France) and was dissolved in DMSO to obtain aliquots of 10 mg mL−1 which were then stored at −20 °C in the darkness.

Apparatus and Material Uncoated capillaries were purchased from Phymep (Paris, France), whereas polyacrylamide (PAA) and polyvinylalcool (PVA) capillaries were obtained from Sciex Separation (Brea, CA) and Agilent Technologies (Santa Clara, CA) respectively. The studies have been performed using a Beckman Coulter PA 800 plus system (Sciex Separation, Brea, CA) equipped with either a photodiode array detection (DAD) or coupled with a solid-state laser induced fluorescence detector (λexcitation: 488 nm, λemission: 520 nm). Data acquisition and instrument control were carried out using Karat 9.1 software (Sciex Separation, Brea, CA). Deionized water used in all experiments was purified using



Page 8 of 38

a Direct-Q3 UV purification system (Millipore, Milford, MA, USA). Conductivity and pH values of buffer solutions and samples were acquired with a SevenCompact pH meter (Mettler Toledo, Schwerzenbach, Switzerland). Selection of background electrolyte (BGE) and buffer ionic strength (IS) calculations were based on simulations with the computer program PhoeBus (Analis, Suarlée, Belgium)

Methods Dissolution and storage of peptides All amyloid beta peptides were dissolved at 2 mg.mL-1, either in pure DMSO or 0.10 - 0.16 % ammonium hydroxide (m / v) in order to prevent in vitro aggregation of the peptide.31,44 For peptides prepared in ammonium hydroxide, aliquot solutions (5 µL) of individual peptides were lyophilized to remove all traces of ammonia and finally stored at - 20°C. For peptides dissolved in DMSO, aliquot solutions of 5 µL were stored immediately after reconstitution at - 32 °C.

Fluorescent Labelling of peptides and sample preparation Each aliquot of amyloid peptide was diluted in CNS perfusion fluid for preparation of standard solutions. These solutions were then diluted in a sodium borate buffer (pH 10.5, IS 40 mM) containing Fluoprobe 488 NHS to obtain the desired concentration with a molar ratio of 200:1 (Fluoprobe / peptide). This ratio led to the formation of mainly di-labelled species for each peptide. The reaction was carried out at room temperature. About two minutes after addition of the fluorescent dye into the sample, the mixture was filtered without the need of reaction quenching.

Design of experiments for the filtration process


Page 9 of 38 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


A two-level-three-factors (32) full factorial experiment was designed to optimize the filtration protocol used for the sample preparation using Minitab statistical software 17. The three factors considered were (i) centrifugation time (low 4 min and high 8 min), (ii) number of filtration (2 to 4) and (iii) replacement of membranes for successive filtrations (0-1). Two responses were evaluated: the recovery of Aβ amyloid peptides and the removal efficiency of the two main peaks arising from the fluorophore (both estimated using the corresponding peak areas of CE-LIF analyses). A Pareto chart was used to determine the optimal conditions for removing most of the fluorescent dye without losing analytes of interest.

Sample preparation To remove the excess of fluorophore, before analysis and preconcentration steps, each sample was first filtered through a 10 kDa membrane (Amicon Ultra-15 Centrifugal Filter Unit, Millipore (UK) Limited, Hertfordshire, UK). The filtrate was then filtered again on a 3 kDa membrane (Amicon Ultra-15 Centrifugal Filter Unit, Millipore (UK) Limited, Hertfordshire, UK). To recover the peptides, 100 µL of water was added to the filter unit and another filtration of this solution was performed on a new 3 kDa membrane. Each centrifugation was performed at 14000 g for 4 min at room temperature. Samples (retentates) were then recovered into the appropriate medium and at the desired concentration using water, BGE or diluted BGE. The peptide recovery after the 3 filtrations was estimated by CZE-LIF, with the optimized CZE method, by analysing 8 different standard peptide solutions and three CSF samples before and after these filtrations by centrifugation, and using the ratio of the peak area of each peptide peak before and after filtration

Separation of peptides by capillary zone electrophoresis (CZE)



Page 10 of 38

The capillary (PVA capillary; I.D. of 50 µm, total and effective lengths of 90 cm and 80 cm respectively) was pre-conditioned with the BGE-1 (120 mM Tricine-NaOH, pH 8.2) during 10 min. Before each injection, the capillary was rinsed with 10 mM phosphoric acid for 10 min then with water for 5 min and then with the BGE for 5 min.. Then, a sample plug accounting for 1 % of the capillary volume was injected at the inlet by pressure (10 sec, at 3.4 kPa). The separation was carried out at -30 kV and 25 °C. Samples were stored at 4 °C in the instrument. .

PA-LVSEP-CZE and M-PA-LVSEP-CZE in a PVA capillary For PA-LVSEP, before each run, the PVA capillary was pre-conditioned and equilibrated as described above. Samples were injected for 75 s at 138 kPa (whole capillary injection, ∼1.8 µL). The applied voltage was - 30 kV for 60 min with BGE vials at inlet and outlet. An additional backward pressure was applied at 700 Pa during 6 min from the beginning of the process. Between each run the capillary was rinsed as stated before. For M-PA-LVSEP, we performed the same process for capillary preconditioning, equilibrating and the first injection. The applied voltage was kept the same, but the duration of pressure application was reduced to 180 s. Then a second sample injection was carried out at 138 kPa for 27 s (40 % of the capillary volume). For the second LVSEP cycle, the same voltage (- 30 kV) and pressure (700 Pa) were applied for 180 s. The process was repeated up to ten times for multiple LVSEP cycles. At the end of the M-PA-LVSEP, CZE separation of the preconcentrated species was triggered by application of the voltage (- 30 kV) without pressure assistance for 60 min. To guarantee the highest performances the PVA capillary was replaced after 120 analyses. The corrected migration time was calculated by subtracting from the observed migration time the duration required for the current to drop to a stable and minimal value upon application of the voltage. This corrected


Page 11 of 38 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


migration time is applied to correct for slight difference in the time required to remove all SM between analyses (when the current reaches its minimal value). The corrected peak areas were calculated by dividing the obtained peak areas to the corrected migration times.

Measurement of electroosmotic flow (EOF) in PVA and PAA capillaries To determine the residual EOF velocity in the PVA coated capillaries, DMSO (0.1 %) was injected (10 sec, at 3.4 kPa) at the outlet of the preconditioned capillary. The analysis was performed in triplicate at - 30 kV for 30 min with the tricine BGE.

CSF sample collection All CSF samples were taken by the department of Neurology, university of Ulm (Ulm, Germany), aliquoted and stored at - 20 °C until use. Samples from five AD patients and three non-demented persons (used as controls) were tested.. Their collection and analysis were approved by the Ethics Committee at the University of Ulm. Additionally, Aβ 1-42 was measured with a commercially available assay (Fujirebio, Hanover, Germany) according to the manufacturers’ instructions. Mean coefficients of variation of the assays were below 20 %.



Page 12 of 38

RESULTS AND DISCUSSION

Neutral-capillary based CZE of fluorescence-labelled Aβ peptides CZE of Aβ peptides is generally carried out under alkaline conditions to enhance their solubility and minimize their aggregation.29,32 Under such conditions, it is not trivial to maintain very reproducible migration times over several analyses due either to fluctuation of the elevated EOF under high pHs and/or to possible adsorption of peptides onto the capillary wall. Dynamic coatings used by Verpillot et al.33 to solve this problem are not compatible with efficient electrokinetic preconcentrations. We developed therefore a new CZE method employing neutral capillaries such as PAA or PVA coated ones. Preliminary results on PAA capillary showed some co-migration between Aβ peptides and peaks of the fluorophore used to derivatize them. After measuring the residual EOF inside PAA coated capillaries, we concluded that it was still too high to achieve a good separation between peptides that vary only by a few amino acids. In contrast, PVA coated capillaries and a BGE-1 composed of 120 mM tricine - NaOH at pH 8.2, high resolutions of different amyloid peptides (i.e. Aβ 1-38, Aβ 1-40, Aβ 1-42, Aβ 2-40 and Aβ 5-40) with minimal adsorption were obtained. Better performances in terms of resolution were even achieved compared to the previous dynamic coating-based method. To achieve sufficient detection sensitivity to determine different amyloid peptides from biological fluids, the electrokinetic preconcentration was developed, in this study, for fluorescence - labelled Aβ peptides allowing there subsequent detection by LIF. Optimization of fluorescent labelling was implemented for several Aβ peptides (i.e. Aβ 1-38, Aβ 1-40, Aβ 1-42, Aβ 2-40 and Aβ 5-40) and for CSF samples using the Fluoprobe F488 according to our previous investigation.32 Although three amino groups are available in each Aβ peptide for covalent bonding, only mono and di-tagged


Page 13 of 38 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


species could be obtained after derivatization due likely to steric constrains and electrostatic repulsions. This led to at least 3 different peaks for each Aβ peptide in CZE. To improve the detection sensitivity and facilitate peak identification, only one peak in CE (di-tagged peptide form) is desirable. To achieve this purpose, we therefore optimized the molar ratio between the Fluoprobe 488 NHS ester and the peptide from 50 to 200. The formation of tagged species for each peptide after the labelling was confirmed with CE-QTOF-MS (see more details in the supplementary electronic information ESI). It was found that the ratio of 200 allowed obtaining only one derivatized species per peptide (see Figure S1). The main form obtained corresponded to a ditagged peptide. No unlabelled peptide could be detected by CE-MS indicating a derivatization yield close to 100 %. However, a large amount of free fluorescent dye that could disturb the electrokinetic preconcentration was present in our sample. Moreover, the derivatization was performed in a borate buffer that is not compatible with the PVA coating capillary. Several approaches were therefore tested to remove the excess of fluorophore and borate from the samples, including: fraction collection using CZE, dialysis and filtrations through membranes (see Figure S2). The filtration method offered the best results and straightforward operations. We thus further optimized this step, by studying 3 factors (centrifugation time, number of filtration and changing or not the membrane between filtrations) using a design experiment at 2 levels (see Figure S3). Two centrifugal filtrations on 2 different 3 kDa membranes during 4 min gave the best recovery of the peptides (between 87 and 92 %) and the highest removal of Fluoprobe (more than 99 %). To remove large species (like proteins) from CSF samples, one pre-filtration of the derivatized CSF on a 10 kDa membrane was carried out prior to the two 3kDa filtrations.. The CZE profiles obtained for fluorescence-labelled Aβ peptides with and without filtrations are compared in Figure 1. The slight loss of peptides occurring during these filtrations was considered not critical as it was always



Page 14 of 38

the same whatever the solution / sample investigated and could therefore be easily corrected through a constant factor or using a calibration curve.

Concept of M-PA-LVSEP-CZE Recently our group developed a multiple - ITP method to increase detection sensitivity for amyloid peptides in CE.6 We extended this concept of multiple cycle electrokinetic approach to PA-LVSEP, as shown in Figure 2, using a mixture of the five standard peptides. First, the capillary was totally filled with the sample mixture (Aβ 1-38, Aβ 1-40, Aβ 1-42, Aβ 2-40 and Aβ 5-40, or CSF samples (after derivatization and filtration). Then the process of PA-LVSEP was applied with the BGE-1 (Tricine - NaOH pH 8.2 IS 120 mM) at the inlet and outlet for 6 min. After this preconcentration step another plug (40 % of the capillary volume) was injected by pressure followed by the same process of preconcentration. The cycles were repeated several times and finally ended by one separation step. As the performance of M-PA-LVSEP is concurrently dependent on that of LVSEP and the number of repeated hydrodynamic injections, optimization of these two processes was implemented. LVSEP of fluorescence labelled Aβ peptides prepared in deionized water rather than (diluted) BGE was found to give the best performance (i.e. high and sharp peaks, data not shown), thanks to the stacking effect promoted under the very low conductivity of the water matrix. For removal of the SM after sample stacking, instead of using only the electroosmotic pumping as in conventional LVSEP, pressurization was employed to create a hydrodynamic flow to compensate for the weak EOF generated in the PVA capillary especially when there is less than 20 % of SM. Based on the theoretical model proposed by Kawai et al.45 and the Poiseuille Law, we developed a new equation to estimate the time (t) necessary to totally remove the SM plug from the capillary (see equation 1)


Page 15 of 38 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


when both electroosmotic mobility (µEOF) and hydrodynamic pressure (∆P) are contributing to this SM removal.

t=

8η L ∆Pr 2

x max

∫ x

dx +

γµEOF , SM Vx (1) {(γ − 1) x + L}L

Where γ is the ratio of the conductivities (σ) of the SM and BGE (γ = σ BGE /σ SM), L is the total length of the capillary; r its radius, x is the length of the SM plug, η is the viscosity of the BGE, µep the electrophoretic mobility of the anionic analyte and µEOF, Sdx 2.3). The linearity of the detection was evaluated by triplicate analyses of solutions containing the mixture of the ﬁve peptides at concentrations ranging from 0.3 to 50 nM. The determination coeﬃcients were from 0.979 to 0.991 showing a good linearity of response (Table 1).



Page 16 of 38

After single PA-LVSEP optimisation, the method was extended to multi - cycle operation to further improve the detection sensibility. In the preliminary tests, baseline separation of Aβ 1-42, Aβ 1-40 and Aβ 1-38 could not be achieved after three successive cycles. This was attributed to the electrophoretic mobilities of the peptide becoming higher than the µEOF, so that the preconcentration of the peptides was not maintained. We investigated this problem by monitoring the EOF magnitudes in the PVA capillary filled with different volume ratios of SM / BGE. It was found that the electrophoretic mobilities of the peptides are stronger than the µEOF when SM occupied less than 40 % of the capillary volume. By reapplying the equation (1), it was estimated that 180 s are necessary to keep 40 % of the SM in the capillary without losing any preconcentrated analyte. This allowed us to estimate when we had to stop the voltage and the pressure to keep enough matrix in the capillary to maintain a global µEOF higher than the µeps. This allowed us to increase the cycle number. We performed a sample injection, filling 100 % of the capillary at the first cycle, applied the process of PA-LVSEP for 180 s (to keep more than 40 % of the SM in the capillary). Then a second injection of 40 % was applied and the same process of preconcentration was done. The cycle was repeated several times. This way we could maintain the resolution during the CZE separation performed even after 8 cycles while increasing the peak intensities, as shown in Figure 3A. The total sample plug concentrated after 8 cycles was equal to 380 % of the capillary volume. A linear relationship between peak area and injection/cycle number was demonstrated (see Figure 3B) showing that there is no loss of sample provoked by the repetition of cycles, nor decrease in the concentrating performances. This linearity was demonstrated at two different concentration levels (0.5 nM and 5 nM) allowing us to decrease the previous LOQ obtained with single PA-LVSEP to 0.06 nM, compatible with the expected concentrations of 0.1 nM for Aβ 1-42 in CSF. The limits of detection (LODs, related to peak heights and calculated as 3 × signal-to-noise


Page 17 of 38 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


ratios) were determined to be 0.2 nM for Aβ 1-42 and 0.4 nM for the other Aβ peptides after one single PA-LVSEP. Signal improvements were 3 - 4 folds after 8 cycles of M-PA-LVSEP. In the particular case of M-PA-LVSEP, for evaluation of signal improvements, the LOQs obtained from the calibration curve made with peak area measurements (using the linear regression method) were used because peak areas were found to respond in a more linear way than peak heights during the multiple cycles of electrokinetic preconcentration 6. For a better demonstration of the M-PALVSEP performance, a comparison between normal CZE, single LVSEP and multi-LVSEP for detection of Aβ 1-42 was made. LOQs for Aβ 1-42 achieved with CZE-LIF, single LVSEP-LIF and multi-LVSEP-LIF were 10 nM

29

, 0.6 nM and 0.06 nM, respectively. Accordingly, a detection

improvement of around 170 folds was achieved with our multi-LVSEP approach. Application to CSF samples The optimized conditions for the derivatization, filtration and M-PA-LVSEP of standards Aβ peptides were then applied to real CSF samples. Eight cycles were performed to allow the detection and quantitation of Aβ 1-42, Aβ 1-40, Aβ 1-38 directly from the derivatized and filtrated CSF samples (see Figure 4). Other peaks in the electropherogram correspond to species that can be derivatized in real CSF. To exclude the possibility of analytes’ co-migration, we checked the separation with another BGE composed of MOPS-NaOH buffer (IS 100mM, pH 7.5) (BGE-2) that produced a shift of migration over 4 min for Aβ (data not shown). It was confirmed from this experiment that there was no co-migration of other species in the CSF with the amyloid peptides. It can be noted that the shape of Aβ1-38’s peak was a bit distorted. This phenomenon, which was more evidently seen in the zooming mode (see the insert in Fig. 4), was already described elsewhere46 and was explained by the distortion of a small peak adjacent to a much larger one behind. The two additional peaks at 45 and 48 min could correspond to those of Aβ 2-40 and Aβ 5-



Page 18 of 38

40 according to the migration times. However, the detected signals did not correspond to the expected concentrations of these species in CSF samples. We evaluated the repeatability of our approach on real biological samples by pooling several CSF samples (n = 5) from AD patients and performing 4 analyses (entire process with 8 cycles). RSD values of peak areas obtained from analyses performed within the same day were between 5.5 % and 6.6 % for Aβ 1-38, Aβ 1-40 and Aβ 1-42. The intermediate precision estimated by analyzing the same sample over 3 days. (table 2) was below 6.9 % for the corrected peak area and less than 3 % for the corrected migration time. The specificity of the method was further checked by comparing the corrected peak areas between BGE-2 and BGE-1 from 5 analyses. We observed a random variation of about 5% (either positive or negative) for Aβ 1-42 and Aβ 1-38, proving that no species co - migrated with those 2 peptides. For Aβ 1-40, the variation of 5% was systematically negative indicating a probable slight overestimation of that peptide with our optimized conditions (BGE-1). However, as our method has a repeatability of 6.6% for Aβ 1-40 in CSF, we considered this variation not critical. We then checked the accuracy of the method by spiking a pool of 5 CSF samples from 5 AD patients with either 200 pM or 600 pM using a concentrated Aβ 1-42 solution. M-PA-LVSEP with 8 cycles was performed and estimated concentrations of Aβ 1-42 were compared to the expected concentration based on the basal level already measured by M- LVSEP in these samples and adding the amount of Aβ 1-42 spiked. It was observed that the peak resolutions started to deteriorate over 8 cycles. To guarantee the highest performance of M-PA-LVSEP, further cycles were not implemented. Table 3 shows that accuracy of the method was between 92.3 and 96.3% according to the amount of Aβ spiked to the sample. For the concentration range expected in real samples, the accuracy above 95% is considered as very satisfactory. Finally, we compared our method to the classical ELISA test (Fujirebio assay) for 8 different CSF samples, 3 non-demented patients and 5 from AD patients.


Page 19 of 38 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


Among those samples, four were measured under a double blinded approach. Figure 5 shows that a very nice correlation was found between the two methods with a r² of 0.992. However, we noticed that the concentrations of Aβ 1-42 estimated with M-PA-LVSEP were systematically lower (by 1013%) than those obtained by ELISA. This can be attributed to several parameters. In ELISA, the use of antibodies does not preclude that in addition to Aβ 1-42, some other species (like other structurally close Aβ peptides) can be captured by the antibodies too. The CSF concentration of this peptide evaluated using different ELISA kits has been recently investigated by Bjerke et al.25 and variations between concentrations obtained using different ELISA suppliers were noticed (from 5 to 20 %)25. The slight discrepancy between ELISA and our “antibody free” approach could also be explained by the fact that ELISA tests could measure not only the monomeric Aβ 1-42 but also to some extent oligomers present in the samples which would be still recognized by the antibody. This would not occur in the M-PA-LVSEP considering that the CZE separation based on species mobilities, precludes this. It is worthy to note that our approach requires much shorter time (i.e. 1h30 for M-PA-LVSEP preconcentration and CZE separation, excluding 15 min needed for offline sample filtration and derivatization) than an ELISA protocol (at least 3 hours for detection of Aβ 142), while allowing simultaneous detection of several Aβ peptides in a single run. With the use of immunocapture, result bias could occur due to cross-reactivity of antibodies towards different species in the sample. This problem on the other hand could be solved with our antibody free approach. We then measured the ratio of Aβ peptide 1-42 and 1-40 instead of Aβ 1-42 alone as recently suggested by several studies.13,15 Based on the 8 different CSFs from AD patients or healthy controls, we defined a cut-off of 0.05 for the discrimination between AD and healthy patients (see



Page 20 of 38

Figure 6) reaching a sensitivity and specificity of 100%. By plotting only, the concentration of Aβ 1-42, the sensitivity was lowered to 80 % and the specificity maintained to 100%.


Page 21 of 38 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


CONCLUSION We developed here a new multi cycles LVSEP based approach to preconcentrate efficiently Aβ peptides from standard solutions with a good repeatability close to 1 % and 5 % for the corrected migration times and the peak areas, respectively. Together with our recently published work on multi-ITP 6, the present work on multi-LVSEP opens the door to a completely new concept of multi-electrokinetic preconcentration that solves the problem of limited performance and preconcentration factors of conventional single-electrokinetic approaches. With this novel approach, the ultra-sensitive detection of Aβ peptide based biomarkers of AD in CSF without recourse to any immunocapture or immunoaffinity steps was made possible, allowing to expand this method to the assay of all other peptides in the amyloid peptide family (being truncated or not) even those for which specific antibodies do not exist. The method was successfully automated thanks to the development of the mathematical model allowing to estimate precisely the time necessary to remove the matrix from the capillary during the process, and also the amount of matrix that has to be kept in the capillary to allow the preconcentration to be operated efficiently. This method was successfully applied to human CSF samples with a minimal and simple sample pretreatment which takes less than 15 minutes. Comparison of this new automated and antibody-free approach to conventional ELISA method showed a very good correlation, although slightly lower values (by 10 %) were systematically obtained for the measurement of Aβ 1-42 in the case of MPA-LVSEP. This feature disserves further investigation to understand whether this arises from an overestimation of ELISA-based methods or an underestimation of M-PA-LVSEP or simply an artefact that stands within the RSD of the method itself.

Acknowledgement



Page 22 of 38

This work has been financially supported by the ‘Institut Universitaire de France’ (Senior Grant IUF 2017 to Prof. Myriam Taverna), by the DIGIDIAG project (n° ANR 10 −NANO 0207), the JPND network PreFrontAls (01ED1512), the German Federal Ministry of Education and Research (FTLDc O1GI1007A), the EU (FAIR-PARK II, the foundation of the state Baden-Württemberg (D.3830), Boehringer Ingelheim Ulm University BioCenter and the Thierry Latran Foundation. We would like to thank Dr. Thuy Tran-Maignan (PNAS, Université Paris Sud) for valuable discussions and all study participants.

References: (1) Breadmore, M. C.; Tubaon, R. M.; Shallan, A. I.; Phung, S. C.; Abdul Keyon, A. S.; Gstoettenmayr, D.; Prapatpong, P.; Alhusban, A. A.; Ranjbar, L.; See, H. H.; Dawod, M.; Quirino, J. P. Electrophoresis 2015, 36, 36-61. (2) Mala, Z.; Slampova, A.; Krivankova, L.; Gebauer, P.; Bocek, P. Electrophoresis 2015, 36, 1535. (3) Breadmore, M. C.; Wuethrich, A.; Li, F.; Phung, S. C.; Kalsoom, U.; Cabot, J. M.; Tehranirokh, M.; Shallan, A. I.; Abdul Keyon, A. S.; See, H. H.; Dawod, M.; Quirino, J. P. Electrophoresis 2017, 38, 33-59. (4) Šlampová, A.; Malá, Z.; Gebauer, P.; Boček, P. Electrophoresis 2017, 38, 20-32. (5) Breadmore, M. C.; de Griend, C. E. S. V.; Majors, R. E. Lc Gc North America 2014, 32, 174-+. (6) Mai, T. D.; Oukacine, F.; Taverna, M. J Chromatogr A 2016, 1453, 116-123. (7) Kitagawa, F.; Kawai, T.; Otsuka, K. Anal Sci 2013, 29, 1129-1139.


Page 23 of 38 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


(8) Kawai, T.; Ueda, M.; Fukushima, Y.; Sueyoshi, K.; Kitagawa, F.; Otsuka, K. Electrophoresis 2013, 34, 2303-2310. (9) Kitagawa, F.; Kinami, S.; Takegawa, Y.; Nukatsuka, I.; Sueyoshi, K.; Kawai, T.; Otsuka, K. Electrophoresis 2017, 38, 380-386. (10) Somers, C.; Goossens, J.; Engelborghs, S.; Bjerke, M. Biomark Med 2017, 11, 169-178. (11) Querfurth, H. W.; LaFerla, F. M. N Engl J Med 2010, 362, 329-344. (12) Olsson, B.; Lautner, R.; Andreasson, U.; Ohrfelt, A.; Portelius, E.; Bjerke, M.; Holtta, M.; Rosen, C.; Olsson, C.; Strobel, G.; Wu, E.; Dakin, K.; Petzold, M.; Blennow, K.; Zetterberg, H. Lancet Neurol 2016, 15, 673-684. (13) Lewczuk, P.; Esselmann, H.; Otto, M.; Maler, J. M.; Henkel, A. W.; Henkel, M. K.; Eikenberg, O.; Antz, C.; Krause, W. R.; Reulbach, U.; Kornhuber, J.; Wiltfang, J. Neurobiol Aging 2004, 25, 273-281. (14) Graff-Radford, N. R.; Crook, J. E.; Lucas, J.; Boeve, B. F.; Knopman, D. S.; Ivnik, R. J.; Smith, G. E.; Younkin, L. H.; Petersen, R. C.; Younkin, S. G. Arch Neurol 2007, 64, 354-362. (15) Lewczuk, P.; Lelental, N.; Spitzer, P.; Maler, J. M.; Kornhuber, J. J Alzheimers Dis 2015, 43, 183-191. (16) Janelidze, S.; Zetterberg, H.; Mattsson, N.; Palmqvist, S.; Vanderstichele, H.; Lindberg, O.; van Westen, D.; Stomrud, E.; Minthon, L.; Blennow, K.; Swedish Bio, F. s. g.; Hansson, O. Ann Clin Transl Neurol 2016, 3, 154-165. (17) Wiltfang, J.; Esselmann, H.; Smirnov, A.; Bibl, M.; Cepek, L.; Steinacker, P.; Mollenhauer, B.; Buerger, K.; Hampel, H.; Paul, S.; Neumann, M.; Maler, M.; Zerr, I.; Kornhuber, J.; Kretzschmar, H. A.; Poser, S.; Otto, M. Ann Neurol 2003, 54, 263-267.



Page 24 of 38

(18) Willemse, E.; van Uffelen, K.; Brix, B.; Engelborghs, S.; Vanderstichele, H.; Teunissen, C. Alzheimer's & Dementia 2017. (19) Kang, J. H.; Korecka, M.; Toledo, J. B.; Trojanowski, J. Q.; Shaw, L. M. Clin Chem 2013, 59, 903-916. (20) Lista, S.; Zetterberg, H.; Dubois, B.; Blennow, K.; Hampel, H. J Neurol 2014, 261, 1234-1243. (21) Song, L.; Lachno, D. R.; Hanlon, D.; Shepro, A.; Jeromin, A.; Gemani, D.; Talbot, J. A.; Racke, M. M.; Dage, J. L.; Dean, R. A. Alzheimers Res Ther 2016, 8, 58. (22) Wilson, D. H.; Rissin, D. M.; Kan, C. W.; Fournier, D. R.; Piech, T.; Campbell, T. G.; Meyer, R. E.; Fishburn, M. W.; Cabrera, C.; Patel, P. P.; Frew, E.; Chen, Y.; Chang, L.; Ferrell, E. P.; von Einem, V.; McGuigan, W.; Reinhardt, M.; Sayer, H.; Vielsack, C.; Duffy, D. C. J Lab Autom 2016, 21, 533-547. (23) Olsson, A.; Vanderstichele, H.; Andreasen, N.; De Meyer, G.; Wallin, A.; Holmberg, B.; Rosengren, L.; Vanmechelen, E.; Blennow, K. Clin Chem 2005, 51, 336-345. (24) Kang, J. H.; Vanderstichele, H.; Trojanowski, J. Q.; Shaw, L. M. Methods 2012, 56, 484-493. (25) Bjerke, M.; Andreasson, U.; Kuhlmann, J.; Portelius, E.; Pannee, J.; Lewczuk, P.; Umek, R. M.; Vanmechelen, E.; Vanderstichele, H.; Stoops, E.; Lewis, J.; Vandijck, M.; Kostanjevecki, V.; Jeromin, A.; Salamone, S. J.; Schmidt, O.; Matzen, A.; Madin, K.; Eichenlaub, U.; Bittner, T.; Shaw, L. M.; Zegers, I.; Zetterberg, H.; Blennow, K. Clin Chem Lab Med 2016, 54, 11771191. (26) Mattsson, N.; Andreasson, U.; Persson, S.; Carrillo, M. C.; Collins, S.; Chalbot, S.; Cutler, N.; Dufour-Rainfray, D.; Fagan, A. M.; Heegaard, N. H.; Robin Hsiung, G. Y.; Hyman, B.; Iqbal, K.; Kaeser, S. A.; Lachno, D. R.; Lleo, A.; Lewczuk, P.; Molinuevo, J. L.; Parchi, P.; Regeniter, A.; Rissman, R. A.; Rosenmann, H.; Sancesario, G.; Schroder, J.; Shaw, L. M.;


Page 25 of 38 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


Teunissen, C. E.; Trojanowski, J. Q.; Vanderstichele, H.; Vandijck, M.; Verbeek, M. M.; Zetterberg, H.; Blennow, K.; Alzheimer's Association, Q. C. P. W. G. Alzheimers Dement 2013, 9, 251-261. (27) Le Bastard, N.; De Deyn, P. P.; Engelborghs, S. Clin Chem 2015, 61, 734-743. (28) Brinet, D.; Kaffy, J.; Oukacine, F.; Glumm, S.; Ongeri, S.; Taverna, M. Electrophoresis 2014, 35, 3302-3309. (29) Mai, T. D.; Pereiro, I.; Hiraoui, M.; Viovy, J. L.; Descroix, S.; Taverna, M.; Smadja, C. Analyst 2015, 140, 5891-5900. (30) Oukacine, F.; Taverna, M. Anal Chem 2014, 86, 3317-3322. (31) Haussmann, U.; Jahn, O.; Linning, P.; Janssen, C.; Liepold, T.; Portelius, E.; Zetterberg, H.; Bauer, C.; Schuchhardt, J.; Knolker, H. J.; Klafki, H.; Wiltfang, J. Anal Chem 2013, 85, 81428149. (32) Verpillot, R.; Esselmann, H.; Mohamadi, M. R.; Klafki, H.; Poirier, F.; Lehnert, S.; Otto, M.; Wiltfang, J.; Jean-Louis, V.; Taverna, M. Anal Chem 2011, 83, 1696-1703. (33) Verpillot, R.; Otto, M.; Klafki, H.; Taverna, M. J Chromatogr A 2008, 1214, 157-164. (34) Mesbah, K.; Verpillot, R.; Chiari, M.; Pallandre, A.; Taverna, M. Analyst 2014, 139, 65476555. (35) Svobodova, Z.; Reza Mohamadi, M.; Jankovicova, B.; Esselmann, H.; Verpillot, R.; Otto, M.; Taverna, M.; Wiltfang, J.; Viovy, J. L.; Bilkova, Z. Biomicrofluidics 2012, 6, 24126-2412612. (36) Mohamadi, M. R.; Svobodova, Z.; Verpillot, R.; Esselmann, H.; Wiltfang, J.; Otto, M.; Taverna, M.; Bilkova, Z.; Viovy, J. L. Anal Chem 2010, 82, 7611-7617. (37) Mohamadi, M. R.; Verpillot, R.; Taverna, M.; Otto, M.; Viovy, J. L. Methods Mol Biol 2012, 869, 173-184.



Page 26 of 38

(38) Mikkonen, S.; Jacksen, J.; Roeraade, J.; Thormann, W.; Emmer, A. Anal Chem 2016, 88, 10044-10051. (39) Mikkonen, S. Electrophoretic focusing in microchannels combined with mass spectrometry: Applications on amyloid beta peptides. KTH Royal Institute of Technology2016. (40) Grasso, G. Mass Spectrom Rev 2011, 30, 347-365. (41) Bros, P.; Delatour, V.; Vialaret, J.; Lalere, B.; Barthelemy, N.; Gabelle, A.; Lehmann, S.; Hirtz, C. In Clinical Chemistry and Laboratory Medicine (CCLM), 2015, p 1483. (42) Varesio, E.; Rudaz, S.; Krause, K. H.; Veuthey, J. L. J Chromatogr A 2002, 974, 135-142. (43) Lame, M. E.; Chambers, E. E.; Blatnik, M. Analytical biochemistry 2011, 419, 133-139. (44) Shen, C. L.; Murphy, R. M. Biophys J 1995, 69, 640-651. (45) Kawai, T.; Sueyoshi, K.; Kitagawa, F.; Otsuka, K. Anal Chem 2010, 82, 6504-6511. (46) Kuban, P.; Hauser, P. C. Electrophoresis 2009, 30, 176-188.


Page 27 of 38 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 captions Fig. 1. CZE of derivatized Aβ 1-38, Aβ 1-40, Aβ 1-42, Aβ 2-40 and Aβ 5-40 at 1 µM before and after the whole filtration protocol. CE conditions: PVA capillary (I.D. of 50µm, total length of 90 cm), BGE 120 mM Tricine-NaOH, pH 8.2, Voltage - 30 kV, temperature 25 °C, LIF detection: λex 488 nm, λem 520 nm. Fig. 2. Principle of the developed M-PA-LVSEP-CZE. 1 - First hydrodynamic injection of the sample, 2 - LVSEP process by exploiting conductivity differences between sample and the separation buffer, 3 - Second hydrodynamic injection, 4 - Second process of LVSEP, 5 Separation and detection of the peptides. Steps 3 and 4 can be repeated several times (cycles) for further preconcentration of analytes newly injected at each cycle. Fig. 3. (A) M-PALVSEP of the 5 peptides (5 nM) after derivatization and filtrations for different number of cycles from 1 to 8. For the first injection, the capillary is totally filled with the sample. For the subsequent cycles 40 % of sample is injected. Between each injection the voltage was set at - 30 kV with an additional pressure of 700 Pa for 180 s. Other conditions as in Figure 1. (B) Peak area of the 5 analysed peptides against the injection/cycle number over 8 cycles of M-PALVSEP at a concentration of 5 nM Fig. 4. M-PA-LVSEP of one CSF sample obtained from a healthy patient after 1 or 8 cycles. In the rectangle, after 1 cycle, only Aβ 1-40 can be quantified but after 8 cycles Aβ 1-40, Aβ 1-42 and Aβ 1-38 can be quantified. Fig. 5. Correlation between ELISA and M-PA-LVSEP for the quantitation of Aβ 1-42 in CSF of 4 patients with known concentration (●) and 4 patients in double blinded measurements (○). The first in the brackets indicates the concentration found by ELISA method and the second with M-PA-LVSEP 27 ACS Paragon Plus Environment


Page 28 of 38

Fig. 6. Plot of the concentration of Aβ 1-42 alone (left) and the ratio of Aβ 1-42 / Aβ 1-40 (right) in CSF of 8 patients as measured by the M-PA-LVSEP method.


Page 29 of 38 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


Determination

RSD of corrected Peptides

RSD of peak area (%) coefficient (r²)

migration time (%) 1-38

0.7

2.5

0.991

1-40

0.9

3.6

0.981

1-42

0.8

3.2

0.987

2-40

1.2

4.1

0.985

5-40

1.5

5.8

0.979

Table 1. Repeatability of migration times and peak areas of PA-LVSEP and linearity of the response for five amyloid peptides



RSD corrected

Intermediate

Intermediate

precision corrected

precision corrected

migration time (%)

peak area (%)

RSD corrected

Peptides migration time (%)

Page 30 of 38

peak area (%)

1-38

1.3

5.5

2.1

5.9

1-40

1.4

6.6

2.9

6.9

1-42

1.4

6.2

2.5

6.5

Table 2. Validation of the intermediate precision of M-PA-LVSEP for the analysis of three amyloid peptides in one pooled CSF sample.


Page 31 of 38 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


Overspike concentration (pM) 0 (n=4) 200 (n=3) 600 (n=3)

Expected concentration (pM) 100.3 300.3 700.3

Mean calculated concentration (pM) 96.6 285.9 646.4

mean accuracy (%) 96.3 95.2 92.3

Table 3. Accuracy of Aβ 1-42 measured by M-PA-LVSEP at 2 overspiked concentrations of a pool of 5 CSF samples from AD patients.



Graphical abstract

ACS Paragon Plus Environment

Page 32 of 38

Page 33 of 38

Fluorophore

150

RFU

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


100 A 1-38 A 1-40 A 1-42 A 2-40 A 5-40 Before

50 Fluorophore

After

0 20

30 40 Migration time (min)

50

Figure 1.



Figure 2.


Page 34 of 38

Page 35 of 38

60

A

1500000

B

A 1-38

Cycle 8

A 1-40

40 Cycle 7 RFU

Cycle 6 Cycle 5

20

Cycle 4

Peak area (A.U.)

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


A 1-42

1000000

A 2-40 A 5-40

500000

Cycle 3 Cycle 2 Cycle 1 0 30

35

40

45

Migration time (min)

50

55

0 1

2

3 4 5 Number of injections

Figure 3.


6

7

8


150

AE 1-40

100 AE 1-38

RFU

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

AE 1-42

50

AE 1-40

Cycle 8

0 20

Cycle 1

30 40 Corrected migration time (min)

50

Figure 4.


Page 36 of 38

Page 37 of 38

0.4

(0.351,0.293)

0.3 M-PA-LVSEP A 1-42 (nM)

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


0.2

(0.210,0.183) (0.190,0.160)

0.1

(0.161,0.149)

(0.104,0.095)

(0.104,0.094) (0.093,0.078)

(0.067,0.060)

0.0 0.0

0.1

0.2 ELISA A 1-42 (nM)

0.3

0.4

Figure 5.



Figure 6.


Page 38 of 38

Online Preconcentration in Capillaries by Multiple Large-Volume

Recommend Documents