Bioassay Development for Ultrasensitive Detection of Influenza A

Aug 4, 2016 - This antibody was implemented on an in-house developed digital ELISA platform for ultrasensitive detection of recombinant nucleoprotein,...
0 downloads 10 Views 1008KB Size
Subscriber access provided by University of Florida | Smathers Libraries

Article

Bioassay development for ultrasensitive detection of influenza A nucleoprotein using digital ELISA Karen Leirs, Phalguni Tewari Kumar, Deborah Decrop, Elena Pérez Ruiz, Pelin Leblebici, Bram Van Kelst, Griet Compernolle, Hanne Meeuws, Liesbeth Van Wesenbeeck, Ole Lagatie, Lieven Stuyver, Ann Gils, Jeroen Lammertyn, and Dragana Spasic Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.6b00502 • Publication Date (Web): 04 Aug 2016 Downloaded from http://pubs.acs.org on August 4, 2016

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 9

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

Bioassay development for ultrasensitive detection of influenza A nucleoprotein using digital ELISA Karen Leirsa, Phalguni Tewari Kumara, Deborah Decropa, Elena Pérez-Ruiza, Pelin Leblebicia, Bram Van Kelsta, Griet Compernolleb, Hanne Meeuwsc, Liesbeth Van Wesenbeeckc, Ole Lagatiec, Lieven Stuyverc, Ann Gilsb, Jeroen Lammertyna* and Dragana Spasica a

BIOSYST-MeBioS, KU Leuven, Belgium Laboratory for Therapeutic and Diagnostic Antibodies, University of Leuven (KU Leuven), Belgium c Janssen Diagnostics, Beerse, Belgium b

ABSTRACT: Flu is caused by the influenza virus that, due to mutations, keeps our body vulnerable for infections, making early diagnosis essential. Although immuno-based diagnostic tests are available, they have low sensitivity and reproducibility. In this paper, the prospect of detecting influenza A virus using digital ELISA has been studied. To appropriately select bioreceptors for this bioassay, seven commercial antibodies against influenza A nucleoprotein were methodically tested for their reactivity and binding affinity. The study has been performed on two markedly different platforms, being an enzyme-linked immunosorbent assay and a surface plasmon resonance system. The selected antibodies displayed completely different behavior on the two platforms and in various assay configurations. Surprisingly, the antibodies that showed overall good reactivity on both platforms had the highest dissociation constant among the tested antibodies, suggesting that, although important, binding affinity is not the only parameter to be considered when selecting antibodies. Moreover, only one antibody had the capacity to capture nucleoprotein directly in lysis buffer used for releasing this viral protein, which might pose huge advantage when developing assays with fast time-to-result. This antibody was implemented on an in-house developed digital ELISA platform for ultrasensitive detection of recombinant nucleoprotein, reaching a detection limit of 4 ± 1 fM in buffer and 10 ± 2 fM in 10 fold diluted nasopharyngeal swabs which is comparable to currently available fast molecular detection techniques. These results point to a great potential for ultrasensitive immuno-based influenza detection.

1. Introduction Flu epidemics, occurring yearly during winter period, result in high morbidity, economic loss and even mortality in elderly, children or people with a compromised immune system1. Flu is caused by the influenza virus, an RNA virus of the Orthomyxoviridae family that infects the respiratory tract. Among the three types of influenza virus (A, B and C), type A and B cause the most severe symptoms in humans2. The different influenza virus types are differentiated based on the nucleoprotein located inside the envelope of the viral particle. In contrast to type B and C, influenza A viruses are further classified into subtypes based on their neuraminidase and hemagglutinin proteins that are found on the surface of the virus3. Although flu vaccination prevents influenza infections to a great extent, mutations in the viral genome lower its permanent effect. Moreover, the commercial influenza vaccines are produced based on the estimation of the three influenza strains that might be dominant during the upcoming winter season, thereby leaving the possibility of getting infected by another (sub)type of influenza virus even if vaccinated4. Also, different viral and bacterial infections are inducing similar symptoms as the influenza virus. Therefore, appropriate diagnosis and therapy is essential not only for the patient but also for the society to control local outbreaks5,6. Whereas antibiotic treatment is useful for bacterial infections, viral infections can be treated with antiviral drugs, which are, however, only effective if administered within 48 hours after the onset of symp-

toms7. Moreover, there is a huge variation in viral load between different people and sample specimens, which complicates diagnosis. Therefore, fast and sensitive detection of viruses is crucial, but it depends on the availability of the right diagnostic tools8. The same tests can also be essential in the case of pandemic outbreaks, caused by new types of influenza viruses generated through interspecies re-assortment of influenza genes. For many years, viral culture was considered a gold standard for influenza diagnosis that allows typing and subtyping of the virus with high sensitivity. However, the time-to-result is usually long (2-3 days), which is an obstacle in making decisions for immediate treatment. Moreover, some viruses are difficult or even impossible to cultivate3,9. Therefore, quantitative polymerase chain reaction (qPCR) has become the method of choice for influenza virus detection. Although conventional qPCR is particularly sensitive, it still has a relatively long detection time (2 h), high cost and requires trained personnel10,11. To meet the need for fast diagnostics of influenza viruses, numerous immuno-based easy-to-use rapid diagnostic tests have been developed in the past years, such as Directigen EZ Flu A and B (Becton Dickinson, USA), Binax Now Influenza A/B antigen kit (Binax, USA) and Quidel Quick Vue Test (Quidel, USA)7. Most of these tests have a turnaround time between 15 and 30 min and some even allow discrimination between influenza A and B. However, the main draw-

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

backs of these tests are the overall low sensitivity and reproducibility due to the variability of different sample specimens and variations in viral load within patients6,10,12. Recent advancements in molecular detection techniques have led to the development of fast PCR-based point-of-care tests for influenza diagnosis (e.g. Alere i (Alere), Cobas Liat (Roche) and Idylla (Biocartis)). These tests are easy-to-use, have a sampleto-result time of less than 50 min and a hands-on time of only 2 minutes9,11. However, the use of RNA as the target for detection has some drawbacks. The stability of the reagents needed for the amplification can result in a false negative outcome. Moreover, the high mutation rate of the virus and occurrence of new viruses require constant update of the primers used in the test13. Recently, the sensitivity of immuno-based assays has been improved by the introduction of digital ELISA. Here, similar to conventional ELISA, the amount of captured target is determined by the enzymatic conversion of substrate. However, by performing the reaction in femtoliter-sized microwells, the accumulated product is detectable even for very low concentrations of the target, thereby achieving single molecule detection14. The immunocomplexes are built on magnetic microparticles that are individually seeded into the microwells. When detecting low target concentrations, each microparticle will either capture one or no target molecule, following Poisson distribution. Consequently, it is possible to count the amount of captured target molecules by simply counting the amount of microwells showing enzymatic activity15. As mentioned above, the use of immunoassays for diagnostic purposes has some distinct advantages, such as their ease-ofuse and low complexity. However, their sensitivity is often substandard. Therefore, the aim of this paper is to study the potential of digital ELISA for the ultrasensitive detection of influenza A virus. Here, the major emphasis is put on the development of the bioassay, with the choice of the bioreceptor being one of the most crucial aspects for efficient target detection and thereby requiring methodical research. The selection depends on the kinetic properties as well as the specificity of the bioreceptors in complex samples. The ideal bioreceptors display fast association with the target molecule, whereas dissociation should be slow. Furthermore, high specificity should be conserved in the clinical specimens16. Therefore, in this work the performance of seven commercially available antibodies against influenza A nucleoprotein, selected based on a high variety of possible applications previously described17,18, is profoundly evaluated. Nucleoprotein is selected as target since it is conserved within all influenza A strains, which allows universal detection of influenza A virus with a single test without the risk of missing an infection due to mutations of the virus19. The performance of the antibodies is studied in relation to their potential of being used as bioreceptors in the ultrasensitive bioassay. Two platforms are used for this in depth study of the antibody behavior, being an enzyme-linked immunosorbent assay (ELISA) and a surface plasmon resonance (SPR) platform, thereby providing both endpoint and kinetic measurements, respectively. Although the antibody characterization is mainly performed using recombinant nucleoprotein, some selected antibodies are also tested for their binding performance towards nucleoprotein obtained from real influenza A virus. The most suitable antibody pair is

Page 2 of 9

finally used for ultrasensitive detection of recombinant nucleoprotein on a digital ELISA platform. 2. Materials and Methods 2.1. Reagents

Reagents were purchased from Sigma-Aldrich (Belgium) unless stated otherwise. Seven commercially available antibodies were selected for testing their binding affinity towards recombinant influenza A nucleoprotein: one monoclonal (11675-MM03, referred to as SinoM) and one polyclonal (11675-RP01, referred to as SinoP) antibody were purchased from SinoBiological Inc. (China); four monoclonal antibodies (ab110661, ab128193, ab66191, ab67428, in this paper referred to as Ab11, Ab12, Ab66 and Ab67, respectively) were purchased from Abcam plc. (United Kingdom); a polyclonal antibody (IMG-5134A, referred to as ImP) was purchased from Imgenex (United States). The selection of antibodies was based on a high variety of possible applications stated in the datasheet (ELISA, western blot, flow cytometry, …). All monoclonal antibodies were raised against nucleoprotein isolated from real virus. Custom made recombinant influenza A nucleoprotein was purchased from Imgenex (United States). Fluorinert FC-40 was obtained from 3M (United States) and Teflon-AF was purchased from Dupont (United States). Parylene-C dimer and Silane A174 were purchased from Plasma Parylene Coating Services (Germany). Reagents for photolithography were obtained from Rohm and Haas (United States). The fluoroalkylsilane, Dynasylan F8263, was a kind gift from Evonik Degussa International AG (Essen, Germany). Dynabeads M280 Tosylactivated superparamagnetic beads were purchased from Life Technologies (United States). Superblock buffer was purchased from Thermo Scientific (United States). PlusOne Drystrip Coverfluid oil was obtained from GE Healthcare (The Netherlands). Beta-galactosidase labeled goat-anti-rabbit antibody was obtained from Abcam (ab130804, United Kingdom). 2.2. Antibody screening using ELISA

Antibody screening using ELISA was first performed with nucleoprotein immobilized directly on the microtiter plate surface (Costar, Corning Inc., United States), as described in supporting information. Next, the antibodies with the highest reactivity were further tested by immobilizing them onto the microtiter plate surface and detecting nucleoprotein free in solution. Based on previously described protocols20,21, 200 µL of 4 µg/mL antibody diluted in PBS buffer, was added to each well and incubated for 72 h at 4 °C. Following incubation, the plate was blocked with 200 µL of 1 % BSA solution in PBS for 2 h at room temperature (RT). Next, the plate was washed six times with PBS containing 0.002 % v/v Tween 80 using an ELx405 washer (BioTek Instruments Inc., United States), followed by the introduction of 200 µL of nucleoprotein in serial five-fold dilutions (from 80 to 0.128 ng/mL) and incubated overnight at 4 °C or 1 h at RT. 180 µL of secondary antibody (3 µg/mL) and 170 µL of HRP-labeled detection antibody were applied sequentially, each with a 2 h incubation time at RT. To maximize the signal-to-noise ratio, different secondary antibodies (ImP and SinoP) and detection antibodies (referred to as GAR1 (Bio-Rad Laboratories Inc., United States) and GAR2 (Life Technologies, Norway)) were tested. According to manufacturer guidelines, 1/2000 dilution for GAR1 and 1/8000 dilution for GAR2 were made in PTA buffer. After every incubation step, the plate was washed as mentioned before. The signal was generated using 160 µL of

ACS Paragon Plus Environment

Page 3 of 9

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

0.4 mg/mL o-phenylenediamine with H2O2 in 0.1 M sodium citrate and 0.2 M di-sodium phosphate solution, pH 5. The reaction was stopped after 10 min by adding 50 µL of 4 M H2SO4 and the absorbance was measured at 492 nm with an ELx808 absorbance microplate reader (BioTek Instruments Inc., United States). The measured absorbance was corrected for the background signal generated due to aspecific interactions of the secondary and detection antibodies with the antibody coated on the microtiter plate surface. 2.3. Antibody screening and affinity constant measurement using SPR

SPR analysis was performed using a Biacore 3000 analytical system (GE Healthcare Bio-Sciences AB, Sweden) and CM5 sensor chips, consisting of a gold-coated glass slide modified with a carboxymethyl dextran layer. The sensor was functionalized by EDC/NHS chemistry using the recommended amine coupling kit (GE Healthcare Bio-Sciences AB, Sweden). The chip surface was activated using 35 µL of a solution containing 37.5 mg/mL EDC and 5.75 mg/mL NHS. After activation, a 5 µg/mL protein (either nucleoprotein or antibody) solution, diluted in 10 mM sodium acetate buffer, pH 4.5 (GE Healthcare Bio-Sciences AB, Sweden), was injected until 1200 RU (response units) were reached22,23. Remaining active COOH-groups were deactivated with 35 µL of 1 M ethanolamine-HCl, pH 8.5 and the surface was washed with 10 µL glycine, pH 1.5 to remove loosely bound protein. 500 nM of analyte solution (either antibody or nucleoprotein) was prepared in HBS-EP buffer (0.01 M HEPES pH 7.4, 0.15 M NaCl, 3 mM EDTA, 0.005 % v/v surfactant P20) and 180 µL of this solution was injected onto the CM5 chip surface with a flow rate of 30 µL/min. After delivery of the analyte, 180 µL of HBS-EP buffer was injected with the same flow rate, to allow dissociation. The glycine solution was also used for regeneration of the surface between consecutive analyses. A sensorgram was obtained for each condition and the response of the antibodies was determined by the amount of signal (in RU) generated when binding of antibody and nucleoprotein occurred. The antibodies with a response > 15 RU, which is recommended by the manufacturer for kinetic analysis, were investigated further to determine association and dissociation constants. Antibody immobilization on CM5 chips was performed as described above. Subsequently, the chip surface was incubated with different concentrations of recombinant nucleoprotein (from 0 nM to 500 nM, which corresponds to 0 µg/mL – 28 µg/mL) and association/dissociation constants were determined by fitting the obtained curves using Biacore Evaluation Software version 4.1 (GE Healthcare Bio-Sciences AB, Sweden). The curves were fitted locally using a Langmuir model (with the assumption of 1:1 binding) or a bivalent analyte model. 2.4. Nucleoprotein detection in lysis buffer

N-lauroylsarcosine sodium salt (NLS) was tested as buffer for lysis of influenza viruses by using a sandwich ELISA. The assay was performed with Ab11 or Ab12 as capture antibody, SinoP as secondary antibody and GAR2 as detection antibody, using the same protocol as described in section 2.2. Concentrations of 0.5 % and 0.1 %24,25 of NLS, diluted in PTA, were tested on purified A/Puerto Rico/8/1934 H1N1 Influenza A virus (Zeptometrix Corporation, United States). Ten-fold serial dilutions (109 to 107 viral particles/mL) of the purified virus were prepared in lysis buffer. The virus solutions were added onto antibody coated wells and incubated for 15 and 30 min.

Next, secondary and detection antibody were applied sequentially with 2 h incubation time each. Signal was generated by applying 100 µL TMB (3,3',5,5'-tetramethylbenzidine) substrate (Thermo Fisher Scientific Inc., United States) and after 10 min the reaction was stopped with 100 µL 1 N HCl. Absorbance at 450 nm was measured with an Infinite M1000 absorbance reader (Tecan Group Ltd., Switzerland). 2.5. Functionalization of magnetic beads with antibodies

Monoclonal antibody (Ab11) against influenza A nucleoprotein was immobilized on 2.8 µm tosylactivated superparamagnetic beads using a functionalization protocol recommended by the supplier. First, 83 µL of the beads stock solution (2 x 109 beads/mL) was washed once in 500 µL of 0.1 M sodium phosphate buffer. Next, 50 µg of antibody, diluted in PBS, was added together with extra sodium phosphate buffer to reach a total volume of 75 µL. 50 µL of 3 M ammonium sulfate was also added and the mixture was incubated overnight at 37 °C and 1400 rpm using a thermoshaker. Next, the supernatant was removed and 500 µL PBS + 0.2 % BSA was added and incubated for 2 h at the same conditions in order to block free tosyl-groups by the BSA. The beads were washed 3 times in PBS, resuspended in 500 µL Superblock buffer and incubated for 1 h at RT while rotating. This step was performed to stabilize the beads and minimize aspecific interactions between the beads and the reagents used in further experiments. Finally, the beads were washed 2 times in Superblock buffer with 0.05 % Tween 20, resuspended in 83 µL of the same buffer and stored at 4 °C until further use. 2.6. Fabrication of hydrophilic-in-hydrophobic microwells

To allow trapping of single beads and preventing them from escaping, hydrophilic-in-hydrophobic (HIH) femtoliter-sized microwells were fabricated by patterning a Teflon-AF layer deposited on glass slides, using a dry lift-off method26. In short, glass slides (51 x 76 x 1 mm) were coated with a 3 µm thick Teflon-AF layer of which the adhesion was improved by first applying fluoroalkylsilane. Next, a thin layer of ParyleneC was deposited to maintain the hydrophobicity of the Teflon during photolithography. Then, an aluminium hard mask (100 nm thickness) was applied. This aluminium mask was patterned by standard photolithography and wet chemical etching. After patterning the aluminium hard mask, the exposed underlying Parylene-C and Teflon-AF layers were etched with oxygen plasma to reveal the array of hydrophilic patches on the glass surface. Finally, the Parylene-C mask was peeled off together with the aluminium hard mask using tweezers. By using this dry lift-off method, the hydrophobicity of the Teflon-AF layer remained intact, while standard photolithographic processes were used to transfer the pattern to the Teflon-AF layer. This fabrication process resulted in 62.500 femtoliter-sized microwells of 4 µm diameter and 3 µm depth, arranged in a hexagonal pattern in a square patch of 2 x 2 mm2. The volume of each femtoliter well was calculated to be 38 fL. 2.7. Detection of recombinant influenza A nucleoprotein using digital ELISA

For the capture of recombinant nucleoprotein on the superparamagnetic beads, 2 µL of the functionalized bead solution was washed once in 100 µL Superblock buffer and resuspended in 100 µL of the same buffer. 3.6 µL (144 000 beads) of this solution was added to 100 µL of sample. This was either re-

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

combinant nucleoprotein spiked in incubation buffer (Superblock + 0.05 % Tween 20) or in nasopharyngeal swab eluted in 3 mL universal transport medium (Copan, Italy) and diluted 10 times in incubation buffer. The mixture was incubated for 1 h at RT, while rotating. Next, the sample was washed twice in 100 µL washing buffer (5x PBS + 0.1 % Tween 20). 100 µL of 1 nM polyclonal detection antibody (SinoP), diluted in incubation buffer, was added and incubated for 45 min at 23 °C and 500 rpm using a thermoshaker. Finally, 100 µL of 7.2 pM beta-galactosidase-labeled labeling antibody, diluted in incubation buffer was added and incubated under the same conditions as the detection antibody. The beads were washed 8 times in washing buffer and resuspended in 10 µL PBS + 1 % BSA + 0.1 % Tween 20. To determine the amount of bound enzyme, which corresponds to the amount of captured nucleoprotein, the beads were seeded into the HIH microwells. This was done by transporting a droplet of 5 µL back and forth over the array (15 seeding cycles) that was placed above a permanent magnet (NdFeB, 6 mm diameter, 12.7 N, Supermagnete, Germany). The sample droplet was then removed and 40 µL of 152 µM fluorescein di(beta-D-galactopyranoside) substrate, diluted in PBS, was added. This droplet was covered with 180 µL of PlusOne Drystrip Coverfluid oil to seal the substratecontaining microwells with oil. The presence of reaction product was determined using an inverted fluorescence microscope (IX71, Olympus Corporation, Japan) with a mercury lamp, 20x objective and EM-CCD camera (Hamamatsu Photonics K.K., Japan). First, a background image was taken before the start of the enzymatic reaction to correct for the autofluorescence of the beads. After 20 min, a new image was taken to visualize the microwells with enzymatic activity (i.e. active beads). 1/6 of the total array was analyzed. To select the correct excitation and emission wavelength, a WIBA filter (excitation filter BP460-495, emission filter BA510-550 and dichromatic filter DM505) was used for imaging. Neutral density filters were used to decrease the light intensity to 3 %. Finally, a third image was taken where the total number of seeded beads was visualized, to calculate the percentage of active beads. Here, a WIGA filter (excitation filter BP530-550, emission filter BA575-625 and dichromatic filter DM570) was used in combination with a light intensity of 25 %. The amount of active beads were counted using ImageJ software by determining the amount of fluorescent spots using the ‘find maxima’ command.

3.

Page 4 of 9

Results and discussion

A thorough study on seven commercially available antibodies against influenza A nucleoprotein was conducted in this paper using ELISA and SPR, to select the best performing antibodies for the ultrasensitive detection of influenza A nucleoprotein using digital ELISA. 3.1. Antibody screening using ELISA

In a first screening experiment antibodies were identified that showed a good interaction with nucleoprotein immobilized at the surface of the microtiter plate (supporting information, Figure S1). Three antibodies (Ab11, Ab12 and Ab66) were selected for which the reactivity was further studied in a typical sandwich ELISA setting in order to detect nucleoprotein in solution. Hereto, the surface of the microtiter plate was coated with Ab11, Ab12 or Ab66 as capture antibody and incubated with recombinant nucleoprotein, followed by the incubation of a secondary and detection antibody. The most optimal pair of secondary and detection antibody (GAR2 and SinoP, respectively), exhibiting a high signal-to-noise ratio, was established as described in supporting information (Table S1 and S2). The experiment was performed with different concentrations of recombinant nucleoprotein (0.128 – 80 ng/mL, five-fold dilutions) to mimic the typical range of viral load in patients27,28. The target was incubated with the capture antibody either for 1 h or overnight (Figure 1). For overnight incubation, all three antibodies showed good binding performance with Ab11 giving the highest reactivity, followed by Ab12 and Ab66. Whereas shorter incubation time had a minimal influence on Ab11 and Ab12, with the latter showing almost no drop in signal, Ab66 exhibited a 4 times lower signal for 1 h incubation compared to overnight. Collectively, based on the results of the ELISA screening, it was concluded that Ab11 and Ab12 hold great potential for the development of bioassays for nucleoprotein detection due to their high reactivity. 3.2. Antibody screening and affinity constant measurement using SPR

Although endpoint measurements provide useful information on the binding performance of the antibodies, kinetic information is essential to determine the association and dissociation constants. Moreover, for developing both sensitive and fast assays, the capacity of antibodies to rapidly recognize and bind their target needs to be investigated. Therefore, all the antibodies were challenged using an SPR system, i.e. Biacore 3000. Screening was performed following two approaches. First, recombinant nucleoprotein was coupled to the CM5 chip

Figure 1: Reactivity of the capture antibodies Ab11 (A), Ab12 (B) and Ab66 (C) towards recombinant influenza A nucleoprotein during overnight (ON) and 1 h incubation. Ab11, Ab66 or Ab12 was immobilized on the microtiter plate and incubated with recombinant nucleoprotein. SinoP was used as secondary antibody and GAR2 as detection antibody. Absorbance (492 nm) was corrected for background signal. For each antibody, several concentrations of nucleoprotein were tested (0.128 – 80 ng/mL), of which the complete calibration curves are shown in supporting information (Figure S2). Here, the linear dynamic range for a nucleoprotein concentration between 0 and 16 ng/mL is shown, with the corresponding linear fitting. Error bars are standard deviations based on three repetitions.

ACS Paragon Plus Environment

Page 5 of 9

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

surface and 500 nM of each antibody was used as analyte. Second, each antibody was coupled to the surface and 500 nM recombinant nucleoprotein was used as analyte. The interaction time between the nucleoprotein and the antibody was 6 min for both configurations. The selected concentration of 500 nM is commonly used as analyte concentration in Biacore 3000 experiments resulting in high response signal22,23. The results of these interactions are shown in supporting information (Table S3). SinoP, Ab66, Ab11 and Ab12, showing a good interaction with the recombinant nucleoprotein in solution, were further used for determining their affinity constants. Here, the antibody was immobilized on the sensor chip surface and nucleoprotein was injected using different concentrations (ranging from 0 to 500 nM, corresponding to 0 µg/mL – 28 µg/mL, for SinoP, Ab11 and Ab12 and from 0 to 200 nM for Ab66 because of its strong interaction with the nucleoprotein). Both association and dissociation rates were determined by fitting the response curves locally using a 1:1 Langmuir model for Ab66, SinoP and Ab12 and a bivalent analyte model for Ab11 (supporting information, Figure S3). The need for a bivalent model for Ab11 can be explained by the possibility of multimeric nucleoprotein binding to two different antibodies. This is more pronounced for antibodies with a higher dissociation rate or if the epitopes are located further away from the binding sites of two nucleoprotein monomers29. The obtained affinity constants are shown in Table 1. The highest KD value (2.57 µM) was obtained for Ab11, which is due to the highest dissociation rate observed among the tested antibodies. The noticed differences in performance of the antibodies on ELISA and SPR can be attributed to the different dynamics between both platforms. For ELISA, the interactions are limited by diffusion, while due to the constant flow of liquid in the SPR system, the dissociation rate has a major influence on the binding performance. From the affinity measurements, it was observed that Ab66 has the lowest dissociation constant, which explains its superior performance on the SPR platform compared to lower reactivity on ELISA. The same observation can be made for Ab11, which showed the highest dissociation constant and therefore a lower performance on the SPR platform compared to ELISA. Even though high variations in binding performances were observed for some of the antibodies on ELISA and the SPR platform, two monoclonal antibodies (Ab11 and Ab12) showed good binding performance on both platforms. This indicates their applicability under different conditions, which is desirable for bioassay development. 3.3. Nucleoprotein detection in lysis buffer

Due to the location of the nucleoprotein, i.e. inside the viral particle, viral lysis is required. To reduce the complexity of the bioassay, it is important to enable immediate capture of the

released nucleoprotein in lysis buffer, thereby omitting washing steps and also minimizing the sample preparation time. Assuming that the detergent present in the lysis buffer will reduce the interactions between the nucleoprotein and the antibodies, the performance of the two most promising antibodies (Ab11 and Ab12) was further tested in the presence of NLS detergent. The NLS was selected as it has been previously described for its capacity to lyse viral particles and thereby liberate nucleoprotein24,25. With this test the binding performance of the antibodies towards nucleoprotein obtained from real influenza virus was investigated, which has been reported so far only for Ab1217. The lysis capacity of the NLS was first tested for two concentrations of NLS (0.1 and 0.5 %), which both showed similar lysis capability (supporting information, Figure S4). 0.1 % was selected for further experiments, in order to minimize the negative effect of the detergent on the antibody-target interaction. The impact of NLS on the antibody performance was tested in a sandwich ELISA assays with 15 and 30 min incubation time between nucleoprotein and capture antibody. The purified viral solution was incubated together with the lysis buffer on the ELISA plate to allow immediate interaction of the released nucleoprotein with the capture antibody (Ab11 or Ab12 coated on the plate; secondary and detection antibody used were SinoP and GAR2, respectively). In order to exclude the possibility of signal interference from free nucleoprotein in the viral solution, a control sample was included in the ELISA test being the viral solution without prior treatment with the lysis buffer. This control measurement revealed that a certain amount of free nucleoprotein is present in the purified viral solution, which should be subtracted from the signal obtained from the samples treated with the lysis buffer. Ab12 clearly showed binding to the viral nucleoprotein already present in the viral solution, indicating its binding capability to nucleoprotein obtained from real virus (data not shown). However, in the presence of the lysis buffer, it showed low binding capability for both incubation times and the three tested viral concentrations (107, 108 and 109 viral particles/mL), indicating the influence of the detergent on its binding performance. In contrast to Ab12, Ab11 still showed reactivity with the released nucleoprotein in the presence of 0.1 % NLS. The results for the different incubation times and different viral dilutions for Ab11 are given in Figure 2. Based on the previously published conceptions that each influenza A viral particle contains approximately 1000 nucleoproteins30, it appears that the ELISA signals obtained for the different virus concentrations correspond to the values obtained from the calibration curve made with recombinant nucleoprotein when using the same antibody, namely Ab11 (Figure 2A). This demonstrates that NLS has minimal influence on the performance of the Ab11 antibody. For 108 and 107 viral particles/mL, the difference between both incubation

Table 1: Affinity constants of Ab66, SinoP, Ab12 (1:1 Langmuir fitting) and Ab11 (bivalent analyte fitting). Errors are standard deviations based on three repetitions for each concentration tested. Average Ab66 SinoP Ab12 Ab11 kon (1/Ms) koff (1/s) KD (M) kon2 (1/RUs) koff2 (1/s) KD2 (RU)

(1.89 ± 1.79) x 104 (8.81 ± 1.83) x 10-5 (9.68 ± 6.99) x 10-9 / / /

(8.62 ± 8.42) x 104 (1.38 ± 1.16) x 10-4 (5.86 ± 8.86) x 10-9 / / /

(1.57 ± 1.51) x 104 (1.10 ± 0.41) x 10-3 (1.17 ± 0.73) x 10-7 / / /

ACS Paragon Plus Environment

(2.52 ± 3.66) x 104 (1.13 ± 0.55) x 10-2 (2.57 ± 3.17) x 10-6 (5.31 ± 17.7) x 10-4 (1.85 ± 5.61) x 10-3 11.6 ± 12.7

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

times was not significant. However, for 109 viral particles/mL, 15 min incubation time gave the highest signal. This can be explained by possible multimerization of the nucleoprotein in the solution. In the viral particle, multiple nucleoproteins form the ribonucleoprotein to protect the viral RNA31. A higher nucleoprotein concentration can enhance the degree of multimerization, which becomes more pronounced for a longer incubation time. In conclusion, both antibodies (Ab11 and Ab12) can be used for the detection of nucleoprotein obtained from real virus. However, only Ab11 has the ability of capturing nucleoprotein directly in lysis buffer (0.1 % NLS) and this even in significantly short time (15 min) thus minimizing the assay time.

Figure 2: Detection of nucleoprotein from influenza A virus using 0.1 % NLS as lysis buffer. Sandwich ELISA was performed with Ab11 as capture antibody, SinoP as secondary antibody and GAR2 as detection antibody. Two different incubation times (15 and 30 min) were tested for multiple virus dilutions. Absorbance at 450 nm is given, corrected for the background signal. Error bars are standard deviations based on three repetitions. 3.4. Detection of nucleoprotein using digital ELISA

Based on the described study and the performance of the tested antibodies, Ab11 and SinoP were selected for developing a nucleoprotein-specific digital ELISA. In this assay, the nucleoprotein was first captured using superparamagnetic beads functionalized with Ab11. The antibody-functionalized magnetic beads were incubated with different concentrations of nucleoprotein (ranging between 2.5 and 100 fM) for 1 h at RT to compose a calibration curve in incubation buffer. Additionally, the bioassay was performed in a clinically relevant sample matrix, namely nasopharyngeal swabs, which was selected among different types of clinical specimens that are used for

Page 6 of 9

influenza detection as it provides the most reliable outcome32. Nasopharyngeal swab samples from healthy donors were first diluted 10 times in incubation buffer in order to reduce the aspecific interaction from the matrix, spiked with different concentrations of nucleoprotein (ranging between 10 and 100 fM) and incubated for 1 h at RT. A sandwich complex was built by introducing a detection antibody (SinoP), followed by a beta-galactosidase-labeled antibody to allow visualization of the bound complexes. The obtained immunocomplexes were seeded into the HIH femtoliter-sized microwells for determining the amount of captured nucleoprotein. After adding the enzyme substrate, the reaction product was visualized using a microscope. A schematic of the procedure is shown in supporting information (Figure S5). An average seeding efficiency (i.e. the amount of wells containing a single bead divided by the total number of wells) of 70 ± 14 % was obtained. Figure 3 shows an example of the obtained microscopy images for the digital ELISA performed in incubation buffer. The percentage of microwells with enzymatic activity (i.e. percentage of active beads) was calculated by dividing the number of fluorescent wells by the total number of seeded beads26. Based on the obtained microscopy images, calibration curves were generated for both sample matrices, as shown in Figure 4. The limit of detection was calculated by interpolating, from the linear curves, the concentration corresponding to a signal equal to 3 times the standard deviation above the mean background signal value. For the capture in buffer, a detection limit of 4 ± 1 fM was obtained, which corresponds well with the experimental results. Although the capture efficiency in 10 times diluted nasopharyngeal swab samples was lower compared to buffer, an excellent detection limit of 10 ± 2 fM was achieved. It was observed that the nasopharyngeal swab matrix is causing suppressing aspecific interactions which are reducing the sensitivity of the assay and can induce false negative results33. The influence of the swab matrix on the capture efficiency was observed in standard ELISA as well (data not shown) and needs to be further investigated in order to reduce this effect, for example by diluting the matrix further or by adding extra blocking agents. Taking into account the estimation of one viral particle containing 1000 nucleoproteins30, the detection limit reached when using the presented digital ELISA concept was estimated as 3.35 log10 and 4.79 log10 viral particles/mL for spiked buffer samples and nasopharyngeal swabs, respectively. Therefore, the obtained detec-

Figure 3: Representative fluorescent images of the digital quantification of nucleoprotein capture (top) and amount of seeded beads (bottom, 2x zoom). Antibody-functionalized magnetic beads were incubated for 1 h at RT with the nucleoprotein spiked in incubation buffer, followed by the binding of detection and labeling antibody. The obtained immunocomplexes were seeded into HIH microwells and substrate was added to generate signal. The beta-galactosidase enzyme attached to the labeling antibody allowed conversion of the substrate into fluorescent product that was read with a fluorescent microscope.

ACS Paragon Plus Environment

Page 7 of 9

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

tion limit in clinical sample is comparable to the Alere i molecular test (4.86 log10 viral particles/mL, Influenza A&B Package Insert). Although this detection limit is already an enormous advancement compared to other influenza-specific immuno-based assays, additional improvement can be made by using superior antibodies. It has been shown theoretically that for efficient detection, the preferred characteristics of the applied antibodies are low KD (< 10-9 M) and high on-rate (> 105 M-1 s-1)34. Here, confined by the use of commercial antibodies, the choice has been made for a capture antibody with inferior affinity but performing on both ELISA and SPR screening platforms and capable of binding nucleoprotein in lysis buffer, the latter being crucial when aiming to reduce sample-to-result time and minimize washing steps. Taking into consideration previously published work14,15,26,34 and the obtained detection limit while being restricted by the use of commercial antibodies, digital ELISA shows great potential for ultrasensitive immuno-based target detection. Even though the intrinsic detection limit of an assay is always dependent on the aspecific interactions that occur and Poisson noise increases for the low target concentrations that are measured with digital ELISA33, subfemtomolar detection limits have been shown already for the detection of prostate specific antigen when using different digital ELISA platforms35,36. This indicates the huge potential of digital ELISA for developing extremely sensitive diagnostics tests. 4.

Conclusions

The performance of seven commercially available antibodies against influenza A nucleoprotein was investigated and their potential to be used as bioreceptors for developing a nucleoprotein-specific digital ELISA was tested. Since nucleoprotein is conserved within all influenza A strains, the risk of missing an infection due to mutations of the virus is avoided. The antibodies were selected for this study based on a high variety of possible applications previously described17,18. They were screened for their reactivity and binding affinity towards recombinant influenza A nucleoprotein using two platforms, providing endpoint (ELISA) and kinetic (SPR) measurements. Different assay configurations were studied, including immobilized target versus immobilized antibody and short versus longer incubation times. Whereas this comprehensive study showed that most of the antibodies exhibited different behav-

ior under different conditions, two monoclonal antibodies (Ab11 and Ab12) showed overall good reactivity with the target on both the ELISA and SPR platform. This suggests their applicability under several conditions including short assay time and low concentration of target, which are mostly desirable aspects for bioassay development. Surprisingly, neither Ab11 nor Ab12 had the lowest KD value among the tested antibodies, indicating that, although important, the dissociation constant is not the only parameter to be considered when selecting an antibody for bioassay development. Moreover, the association and dissociation rate of antibodies can determine the preferred platform for the interaction with the target. It was observed that the antibodies with the lowest dissociation constant had superior performance on the SPR platform compared to a lower reactivity on ELISA. Similarly, antibodies showing the highest dissociation constant had a lower performance on the SPR platform compared to ELISA. The performance of Ab11 and Ab12 was also tested in the lysis buffer (0.1 % NLS) needed for releasing nucleoprotein located inside the viral particle. This revealed that both antibodies were able to bind nucleoprotein from real virus, whereas only Ab11 showed reactivity directly in NLS. Therefore, Ab11 represents an apparent choice when developing a fast detection assay because it allows immediate capture of nucleoprotein in lysis buffer thus substantially shortening the sample preparation step, which often presents the major hurdle in every bioassay design. Finally, Ab11 was implemented on an in-house developed digital ELISA platform for ultrasensitive detection of recombinant nucleoprotein. Nucleoprotein was captured in buffer and 10 times diluted nasopharyngeal swab samples. A detection limit of 4 ± 1 fM and 10 ± 2 fM was obtained for capture in buffer and diluted swabs, respectively. Based on the estimation that each influenza viral particle contains 1000 nucleoproteins30, this corresponds to 3.35 log10 and 4.79 log10 viral particles/mL for buffer and nasopharyngeal swabs, respectively. The obtained detection limit in clinical sample is therefore comparable to the currently available fast molecular tests (Alere i, 4.86 log10 viral particles/mL). However, it has been theoretically shown that by using superior antibodies, the detection limit can be further improved34, indicating the enormous potential of using digital ELISA for ultrasensitive detection.

Figure 4: Calibration curves for A) capture of nucleoprotein in buffer (0, 2.5, 5, 10, 25, 50 and 100 fM) and B) nucleoprotein capture in 10 times diluted nasopharyngeal swab samples (0, 10, 25, 50 and 100 fM). The percentage of active beads was calculated based on the microscopy images obtained after 20 minutes of signal generation. Linear weighted fitting (weight function: ωi = 1/σi2, with σi the error bar size) was applied. Error bars are standard deviations based on three repetitions for the capture in buffer (A) and 2 repetitions for the capture in diluted swab samples (B). The limit of detection (LOD) was calculated by interpolating from the linear calibration curves the concentration corresponding to a signal equal to 3 times the standard deviation above the mean background signal value, being 4 ± 1 fM and 10 ± 2 fM for capture in buffer and diluted swab samples, respectively.

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

ASSOCIATED CONTENT Supporting Information Additional information as reported in text. This material is available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION Corresponding Author * E-mail: [email protected]. Fax: +32 16 32 29 55.

Author Contributions All authors have given approval to the final version of the manuscript. The authors declare no competing financial interest.

ACKNOWLEDGMENT The authors would like to thank the European Union (IMI RAPPID, Norosensor FP7-NMP-2013-SMALL-7604244 and Horizon 2020 research and innovation program under the Marie Skłodowska-Curie grant agreement No 675412), KU Leuven (IOF/KP/12/002 NanoDiag, IOF/KP/12/009 Atheromix, OT 13/058 and C3 2/15/005), Research Foundation - Flanders (FWO G086114N) and De Vlaamse Liga tegen Kanker for financial support. The nasopharyngeal swabs were obtained from the Laboratory of Medical Microbiology of the Vaccine and Infectious Disease Institute of the University of Antwerp.

REFERENCES 1. Mahony, J.; Chong, S.; Bulir, D.; Ruyter, A.; Mwawasi, K.; Waltho, D. 2013. J. Clin. Virol. 58, 127-131. 2. Centers for Disease Control and Prevention, www.cdc.gov, (assessed April 2015). 3. Krejcova, L.; Hynek, D.; Adam, V.; Hubalek, J.; Kizek, R. 2012. Int. J. Electrochem. Sci. 7, 10779-10801. 4. Kreijtz, J. H.; Osterhaus, A. D.; Rimmelzwaan, G. F. 2009. Hum. Vaccin. 5, 126-135. 5. Call, S. A.; Vollenweider, M. A.; Hornung, C. A.; Simel, D. L.; McKinney, W. P. 2005. JAMA. 293, 987-997. 6. Hazelton, B.; Gray, T.; Ho, J.; Ratnamohan, V. M.; Dwyer, D. E.; Kok, J. 2015. Influenza Other Respir. Viruses. 9, 151-154. 7. van der Vries, E.; Anber, J.; van der Linden, A.; Wu, Y.; Maaskant, J.; Stadhouders, R.; van Beek, R.; Rimmelzwaan, G.; Osterhaus, A.; Boucher C.; Schutten, M. 2013. J. Mol. Diagn. 15, 347-354. 8. Parida, M.; Shukla, J.; Sharma, S.; Santhosh, S.R.; Ravi, V.; Mani, R.; Thomas, M.; Khare, S.; Rai, A.; Ratho, R.K.; Pujari, S.; Mishra, B.; Rao, P.V.L.; Vijayaraghavan, R. 2011. J. Mol. Diagn. 13, 100-107. 9. Bell, J.; Bonner, A.; Cohen, D. M.; Birkhahn, R.; Yogev, R.; Triner, W.; Cohen, J.; Palavecino, E.; Selvarangan, R. 2014. J. Clin. Virol. 61, 81-86. 10. Cho, C. H.; Woo, M. K.; Kima, J. Y.; Cheong, S.; Lee, C.-K.; An, S. A.; Lim, C. S.; Kim, W. J. 2013. J. Virol. Methods. 187, 51-56. 11. Chen, L.; Tian, Y.; Chen, S.; Liesenfield, O. 2015. Eur. J. Microbiol. Immunol. 5, 236-245. 12. Sutter, D. E.; Worthy, S. A.; Hensley, D. M.; Maranich, A. M.; Dolan, D. M.; Fischer, G. W.; Daum, L. T. 2012. J. Med. Virol. 84, 1699-1702. 13. Peaper, D. R.; Landry, M. L. In Handbook of Clinical Neurology, Vol 123 Neurovirology; Tselis, A.; Booss, J.; Elsevier, 2014, ch 5; 123-147. 14. Rissin, D.; Kan, C. W.; Campbell, T. G.; Howes, S. C.; Fournier, D. R.; Song, L.; Piech, T.; Patel, P. P.; Chang, L.; Rivnak, A. J.;

Page 8 of 9

Ferrel, E. P.; Randall, J. D.; Provuncher, G. K.; Walt, D. R.; Duffy, D. C. 2010. Nat. Biotechnol. 28, 595-599. 15. Rissin, D. M.; Fournier, D. R.; Piech, T.; Kan, C. W.; Campbell, T. G.; Song, L.; Chang, L.; Rivnak, A. J.; Patel, P. P.; Provuncher, G. K.; Ferrel, E. P.; Howes, S. C.; Pink, B. A.; Minneham, K. A.; Wilson, D. H.; Duffy, D. C. 2011. Anal. Chem. 83, 2279-2285. 16. Zichi, D.; Eaton, B.; Singer, B.; Gold, L. 2008. Curr. Opin. Chem. Biol. 12, 78-85. 17. Smith, S. E.; Gibson, M. S.; Wash, R. S.; Ferrara, F.; Wright, E.; Temperton, N.; Kellam, P.; Fife, M. 2013. J. Virol. 87, 1295712966. 18. Thompson, C. I.; Barclay, W. S.; Zambon, M. C.; Pickles, R. J. 2006. J. Virol. 80, 8060-8068. 19. Portela, A.; Digard, P. 2002. J. Gen. Virol. 83, 723-734. 20. Van Stappen, T.; Brouwers, E.; Tops, S.; Geukens, N.; Vermeire, S.; Declerck P. J.; Gils, A. 2015. Ther. Drug Monit. 37, 479-485. 21. Gils, A.; Alessi, M-C.; Brouwers, E.; Peeters, M.; Marx, P.; Leurs, J.; Bouma, B.; Hendriks, D.; Juhan-Vague, I.; Declerck, P. J. 2003. Arterioscler. Thromb. Vasc. Biol. 23, 1122-1127. 22. Gils, A.; Ceresa, E.; Macovei, A. M.; Marx, P. F.; Peeters, M.; Compernolle, G.; Declerck, P. J. 2005. J. Thromb. Haemost. 3, 2745-2753. 23. de Mol, N.J. In Surface Plasmon Resonance, Methods in Molecular Biology; de Mol, N. J.; Fischer, M. J. E.; Humana Press: London, 2010, ch. 6; 101-111. 24. Yang, K.; Wills, E. G.; Baines, J. D. 2011. J. Virol. 85, 11972-80. 25. Bilello, J. P.; Manrique, J. M.; Shin, Y. C.; Lauer, W.; Li, W.; Lifson, J. D.; Mansfield, K.G.; Johnson, R. P.; Desrosiers, R. C. 2011. J. Virol. 85, 12708-12720. 26. Witters, D.; Knez, K.; Ceyssens, F.; Puers, R.; Lammertyn, J. 2013. Lab Chip. 13, 2047. 27. Gavin, P. J.; Thomson, R. B. Jr. 2003. Clin. Appl. Immunol. Rev. 4, 151-172. 28. Lee, N.; Chan, P. K. S.; Hui, D. S. C.; Rainer, T. H.; Wong, E.; Choi, K-W.; Lui, G. C. Y.; Wong, B. C. K.; Wong, R. Y. K.; Lam, W-Y.; Chu, I. M. T.; Lai, R. W. M.; Cockram, C. S.; Sung, J. J. Y. 2009. J. Infect. Dis. 200, 492-500. 29. Pabbisetty, K. B.; Yue, X.; Li, C.; Himanen, J-P.; Zhou, R.; Nikolov, D.B.; Hu, L. 2007. Prot. Sci. 16, 355-361. 30. Lamb, R. A.; Krug, R. M. In Orthomyxoviridae: the viruses and their replication; Knipe, D. M.; Howley, P. M.; Griffin, D. E. et al.; Fields Virology, 4th edn. Lippincott Williams & Wilkins, 2001; 1487–1531. 31. Chenavas, S.; Estrozi, L. F.; Slama-Schwok, A.; Delmas, B.; Di Primo, C.; Baudin, F.; Li, X.; Crépin, T.; Ruigrok, R. W. H. 2013. PLoS Pathog. 9, 1-10. 32. Forman, M.; Valsamakis, A. In Specimen Collection, Transport, and Processing: Virology; Versalovic, J.; Carroll, K.; Funke, G.; Jorgensen, J.; Landry, M.; Warnock, D.; Manual of Clinical Microbiology, 10th edn. ASM Press, 2011; 1276-1288. 33. Davies, C. In Immunoassay performance measures; Wild, D.; The Immunoassay Handbook: Theory and applications of ligand binding, ELISA and related techniques, 4th edn. Elsevier, 2013; 11-26. 34. Chang, L.; Rissin, D. M.; Fournier, D. R.; Piech, T.; Patel, P. P.; Wilson, D. H.; Duffy, D. C. 2012. J. Immunol. Methods. 378, 102-115. 35. Rissin, D. M.; Fournier, D. R.; Piech, T.; Kan, C. W.; Campbell, T. G.; Song, L.; Chang, L.; Rivnak, A. J.; Patel, P. P.; Provuncher, G. K., Ferrell, E. P.; Howes, S. C.; Pink, B. A.; Minnehan, K. A.; Wilson, D. H.; Duff, D. C. 2011. Anal. Chem. 83, 2276-2285. 36. Kim, S. H.; Iwai, S.; Araki, S.; Sakakihara, S.; Iino, R.; Noji, H. 2012. Lab Chip.12, 4986-4991.

ACS Paragon Plus Environment

Page 9 of 9

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

Analytical Chemistry For TOC only

ACS Paragon Plus Environment

9