Flow-Induced Dispersion Analysis for Probing Anti-dsDNA Antibody

Subscriber access provided by CORNELL UNIVERSITY LIBRARY

Article

Flow-Induced Dispersion Analysis for Probing Anti-dsDNA Antibody Binding Heterogeneity in Systemic Lupus Erythematosus Patients: Towards a New Approach for Rapid Diagnosis and Patient Stratification Nicklas N. Poulsen, Morten E. Pedersen, Jesper Ostergaard, Nickolaj Jacob Petersen, Christoffer Tandrup Nielsen, Niels H. H. Heegaard, and Henrik Jensen Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.6b01741 • Publication Date (Web): 29 Aug 2016 Downloaded from http://pubs.acs.org on September 1, 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 7

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

Flow-Induced Dispersion Analysis for Probing Anti-dsDNA Antibody Binding Heterogeneity in Systemic Lupus Erythematosus Patients: Towards a New Approach for Rapid Diagnosis and Patient Stratification Nicklas N. Poulsen1, Morten E. Pedersen1, Jesper Østergaard1, Nickolaj J. Petersen1, Christoffer T. Nielsen2, Niels H. H. Heegaard2,3 and Henrik Jensen1,* 1

Department of Pharmacy, University of Copenhagen, Universitetsparken 2, 2100 Copenhagen, Denmark; 2Department of Autoimmunology & Biomarkers, Statens Serum Institut, Artillerivej 5, 2300 Copenhagen, Denmark; 3Department of Clinicial Biochemistry, Odense University Hospital, University of Southern Denmark, Sdr. Boulevard 29, 5000 Odense C. *Corresponding author: [email protected] ABSTRACT: Detection of immune responses is important in the diagnosis of many diseases. For example, the detection of circulating autoantibodies against double-stranded DNA (dsDNA) is used in the diagnosis of Systemic Lupus Erythematosus (SLE). It is, however, difficult to reach satisfactory sensitivity, specificity and accuracy with established assays. Also, existing methodologies for quantification of autoantibodies are challenging to transfer to a point-of-care setting. Here we present the use of Flow-Induced Dispersion Analysis (FIDA) for rapid (minutes) measurement of autoantibodies against dsDNA. The assay is based on Taylor Dispersion Analysis (TDA) and is fully automated with the use of standard Capillary Electrophoresis (CE) based equipment employing fluorescence detection. It is robust towards matrix effects as demonstrated by the direct analysis of samples composed of up to 85% plasma derived from human blood samples, and it allows for flexible exchange of the DNA sequences used to probe for the autoantibodies. Plasma samples from SLE positive patients were analyzed using the new FIDA methodology as well as by standard indirect immunofluorescence and solid-phase immunoassays. Interestingly, the patient antibodies bound DNA sequences with different affinities, suggesting pronounced heterogeneity amongst autoantibodies produced in SLE. The FIDA based methodology is a new approach for autoantibody detection, and holds promise for being used for patient stratification and monitoring of disease activity.

Systemic Lupus Erythematosus (SLE) is an autoimmune disease that can affect many organs. In the diagnosis of the disease, detection of autoantibodies against double stranded DNA (dsDNA) is used as an important criterion1. However, substantial discrepancies are observed between assays and results from different laboratories2,3. Autoantibodies against dsDNA represent a heterogeneous population in SLE4-6 and some recent evidence suggest that not all antibodies against dsDNA in SLE patients are associated with the disease and indeed that some types of anti-DNA antibodies may be found in perfectly healthy individuals2,7-9. Consequently, diagnosis is based on both clinical criteria and serological manifestations10. Quantification of anti-dsDNA antibodies in blood samples are often carried out using the Crithidia luciliae Immuno Fluoresence Test (CLIFT)8,11 and/or an Enzyme-Linked ImmunoSorbent Assay (ELISA)2,7,11. These assays are based on specific interactions and involve multiple washing steps and secondary detection reagents. Often, non-specific binding and surface adsorption has been found to adversely influence the results, especially from ELISA12. Also, the multistep ELISA procedure presents challenges in relation to automation; ELISA is therefore often performed as a manual procedure taking hours to complete13,14. Furthermore, the complexity of ELISA assays presents challenges in inter- and intra-laboratory precision as well as in assay optimization. While the currently used assays are primarily used for diagnostic purposes there is a

growing need for additional biomarkers and technologies suitable for testing new treatments and/or for evaluating efficacy in patient sub-populations15. Thus, a new methodology where the heterogeneous anti-dsDNA antibody population can be probed in greater detail would be of high value. Recently, Flow-Induced Dispersion Analysis (FIDA) was presented as a new homogeneous technique for protein quantification in complex sample matrices such as plasma16. In the FIDA assay a small ligand with high affinity to the target protein is used as an indicator. Because the indicator is smaller than the protein analyte the apparent size of the indicator is observed to increase upon binding. The apparent size of the indicator is conveniently measured by quantifying the apparent indicator diffusivity by Taylor Dispersion Analysis (TDA)17. TDA is based on analysing the dispersion of the indicator in a hydrodynamic flow system. TDA is a calibration free methodology enabling a wide size range to be studied18,19. Further, as the radial diffusion is measured inside a narrow capillary (typically 50-75 µm inner diameter) the analysis is completed within minutes. Thanks to the simplicity of the method it is easy to automate, it has high reproducibility, and new assays are straightforward to develop. In the present work FIDA was developed for quantification of autoantibodies against double stranded DNA. The FIDA assay was optimized in a model system using a monoclonal mouse antibody against

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 7

dsDNA as analyte, and subsequently used for assessing antidsDNA autoantibodies in samples from SLE patients. Three 32 bp indicator DNA sequences were used and showed a strong sequence dependency of anti-dsDNA autoantibody binding. The optimisation strategy in the present work is expected to be applicable for other plasma or serum-based FIDA assays.

ultra-pure water and filtered using a 13 mm 0.2 µm nylon syringe filter. The peak variance was determined by fitting the indicator peak to a Gaussian function. The apparent diffusivities (Dapp) were calculated from equation 117:

EXPERIMENTAL SECTION

where r is the inner radius of the capillary, tR is the peak appearance time and σ2 is the peak variance. The experimental conditions were chosen in order for equation 1 to be valid.16,22 As further described in the supporting information this involves restriction to the flow rates and capillary dimensions employed. For the fused silica capillaries of inner diameter 75 µm used in the present work the conditions are easily fulfilled. Protocol I was used for FIDA experiments performed in phosphate buffer alone. Initially the capillary was rinsed with 48% v/v ethanol and subsequently buffer both at 1 bar for 5 min. Then an analyte zone was injected at 50 mbar for 30 s followed by the indicator zone injected at 50 mbar for 5 s. Following the indicator zone a new zone of analyte was injected for 30 s at 50 mbar. The indicator and analyte zones were eluted with buffer using a pressure gradient of 50 mbar for 14 min. The binding curve in buffer was obtained using protocol I. In a series of experiments, 50 nM DNA-I was used as indicator and the monoclonal antibody against dsDNA was used as analyte in the following concentrations: 0, 50, 100, 300, 500 and 1000 nM. In another series of experiments the indicator was mixed with the analyte prior to analysis. In this mixture 50 nM DNA-I in the target analyte concentration was used as indicator. In this series of experiments the following analyte (the anti-dsDNA monoclonal antibody) concentrations were used: 0, 50, 100, 500 and 1000 nM. Assay selectivity was tested by measuring binding of DNA to the SLE irrelevant monoclonal antibody anti-beta-2microglobulin of the same isotype. DNA-I in a concentration of 50 nM was used as indicator and an antibody against beta2-microglobulin was used as analyte in the following concentrations: 0, 10, 30, 100, 300, 1000 nM. Protocol II was employed for FIDA experiments performed in 20% (v/v) human plasma. Initially the capillary was rinsed with 48% v/v ethanol and subsequently buffer both at 1 bar for 5 min. Then an analyte zone was injected at 50 mbar for 5 min followed by the indicator zone injected at 50 mbar for 5 s. Following the indicator injection the capillary inlet was dipped in buffer. The indicator zone was eluted with the analyte at a pressure gradient of 50 mbar. The binding curve in 20% plasma (diluted in phosphate buffer) was obtained as described in protocol II. In a series of experiments, 50 nM DNA-I prepared in phosphate buffer was used as indicator and the antibody against dsDNA was used as analyte in the following concentrations: 0, 10, 30, 50, 100, 300, 500 and 1000 nM. A second series of experiments was conducted using a bare fused silica capillary instead of a PEG-coated capillary. In this set of experiments the following antibody concentrations were used as analyte: 0, 50, 100, 500, and 1000 nM. Protocol III was employed for FIDA assays performed in 85% (v/v) plasma. With the purpose of minimizing the consumed volume of patient sample as well as minimizing the risk

Equipment: A fused silica capillary with inner:outer diameter of 75:360 µm, 75 cm total length and 65 cm to the detection window was used. Capillaries were obtained from Polymicro Technologies (Phoenix, AZ, USA). Unless otherwise stated the inner surface of the capillary was covalently coated with poly(ethylene glycol) (PEG) brushes as described elsewhere20. An Agilent 3DCE instrument (Agilent Technologies, Waldbronn, Germany) employing laser induced fluorescence detection (ZETALIF Evolution featuring a 488 nm Melles Griot Diode laser, Picometrics, Labege, France) was used to conduct the experiments. The capillary was thermostated inside the Agilent 3DCE instrument, except for the small fraction used to connect to the ZETALIF Evolution detector. Materials and chemicals: Ultrapure water (18 MΩ cm) was obtained from a Millipore (Billerica, MA, USA) MilliQ water purification system. The antibody against beta-2 microglobulin (HYB290-03) and the antibody against dsDNA (HYB331-01) were acquired from Statens Serum Institut (SSI, Copenhagen, Denmark). TITAN nylon syringe filters (pore size 0.2 µm) were obtained from Sun-Sri (Rockwood, TN, USA). Human plasma was obtained from pooled plasma from healthy blood donors. The six anonymous SLE patient samples were obtained from the biobank at SSI21. The three dsDNA sequences labelled with the ATTO-dye were purchased from DNA Technology A/S (Aarhus, Denmark): DNA-I: ATTO-488 – 5-AACAGCTATGACCATGCCGACTGCATATCCGT-3, DNA-II: ATTO-488 – 5-GACGTTGACGTTGACGTTGACGTTGACGTT-3 and DNA-III: ATTO-488 – 5-GCAGTACGTGTACTACGTCAT TTGCAGTATCT-3. Procedures: The FIDA methodology was carried out using automated capillary electrophoresis (CE)-based equipment, but without using the high voltage power supply. The detailed procedures are described in protocol I-III (vide infra) and the overall principle is illustrated in figure 1.

Figure 1. The automated FIDA protocol is performed using commercial equipment for capillary electrophoresis. The inlet vials contain sample, indicator/sample solution and washing solutions. A hydrodynamic flow is employed giving rise to a parabolic flow profile (insert) which results in a Gaussian signal shape at the detection point.

Throughout, the measurements were performed using a 100 mM phosphate buffer pH 7.9. The buffer was prepared using


 



  

ACS Paragon Plus Environment

(1)

Page 3 of 7

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

of adsorption of the target proteins to the filter23, a centrifugation procedure was used in order to eliminate precipitates from the plasma: 250 µL plasma was centrifuged at 8000 g for 5 min and 175 µL of the supernatant was carefully withdrawn immediately. The patient plasma samples were diluted to 85% with buffer. Initially the capillary was flushed with buffer for 10 min at 1 bar. Subsequently the 85% patient plasma sample was injected for 3 min followed by the indicator (100 nM in 85% patient plasma) for 5 s both at 50 mbar. Then the capillary inlet was placed in buffer (to clean inlet) before the elution with 85% patient plasma for 15 min at 50 mbar. The binding curve in 85% plasma was obtained as described in protocol III. In a series of experiments, 100 nM DNA-I prepared in phosphate buffer was used as indicator and the antibody against dsDNA was used as analyte in the following concentrations: 0, 10, 30, 50, 100, 300, 500 and 1000 nM. Agilent Chemstation method files are available upon request. ELISA & CLIFT: The model monoclonal antibody against dsDNA (HYB331-01) was studied using a commercial antidsDNA VARELISA kit for comparison with the FIDA method. The model antibody was diluted in the assay buffer and used as analyte in the following concentrations: 6.7 nM, 1.7 nM, 0.4 nM, 0.11 nM, 27 pM, 6.7 pM, and 1.7 pM. The procedure was performed in accordance with the manufacturer recommendations with the following exceptions: The secondary anti-human IgG from the kit was exchanged with alkaline phosphatase-conjugated goat anti-mouse IgG (Sigma A3688) in a final concentration of 1.45 µg/mL. Phosphatase substrate (Sigma S0942) was used instead of tetramethylbenzidine (TMB) from the kit. The plates were read 30 min after the substrate was added at 405 nm with 690 nm as reference. The indirect immunofluorescence test with the haemoflagellate Crithidia luciliae (CLIFT) was performed according to the instructions of the manufacturer (ImmunoConcepts, Sacramento, CA, USA) and samples were scored negative or weakly, medium or strongly positive based on microscopic examination of the fluorescence intensity.

Figure 2. Representative Taylorgrams of 29 nL 50 nM DNA-I injected in blank phosphate buffer (full line) and in the presence of 500 nM model monoclonal antibody in phosphate buffer (broken line) obtained using protocol I. A small offset has been applied to the broken curve in order to facilitate visual comparison.

A standard curve for the DNA-I mAb interaction obtained in buffer is shown in Figure 3. It can be observed that increasing the antibody concentrations leads to a decrease in the apparent indicator diffusivity. The binding curve was fitted to the following binding isotherm:


 

  ∙




(2)

where DI is the indicator diffusivity, DIA is the diffusivity of the complex, K is the equilibrium dissociation constant and [A] is the analyte concentration16,22. A non-linear fitting procedure resulted in estimates of DI, DIA and K (supporting information and figure 3). The data obtained in 85% plasma was corrected for differences in viscosities as described in the supporting information. A dissociation constant of 4.24 · 106 M-1 was obtained in buffer, which is close to the value of 1.11 · 107 M-1 for the corresponding biotinylated dsDNA found by affinity capillary electrophoresis (ACE) under similar conditions24.

RESULTS AND DISCUSSION For optimization of the assay conditions, a mouse monoclonal antibody (mAb) against dsDNA, originally developed from a lupus mouse, was selected as model analyte. A fluorescently labeled 32 bp dsDNA sequence previously found to bind the antibody24 was used as indicator and denominated DNA-I. An indicator concentration of 50 nM was used, which is sufficiently high to provide a suitable fluorescent signal and well below the dissociation constant previously found for the analyte-indicator couple24, which ensures maximum sensitivity towards the analyte16. Sample Taylorgrams of DNA-I in the presence and absence of anti-dsDNA obtained using protocol I are shown in Figure 2. The broadening of the indicator peak in the presence of the mAb shows that the apparent diffusivity of the dsDNA probe decreases indicating binding to the mAb. It may be noted that the peak shape and symmetry also can be used as a measure of quality control of the assay. For example a pronounced peak tailing would be indicative of adsorption of the indicator and/or indicator/analyte complex.

Figure 3. Binding curves for the interaction between DNA-I and the monoclonal antibody against dsDNA in 100 mM phosphate buffer, pH 7.9, (open circles, protocol I), 20% (v/v) human plasma (crosses, protocol II) and 85% (v/v) human plasma (black diamonds, protocol III) . The curves were obtained in a PEG coated capillary with ID 75 µm and a length to the detector of 65 cm (75 cm total length) at a mobilization pressure of 50 mbar. The fitting procedure resulted in affinity constants of K = 4.24 · 106 M-1 (0 % plasma, solid line), K = 3.60 · 106 M-1 (20 % plasma, broken line) and K = 2.76 · 106 M-1 (85 % plasma, dotted line, corrected for viscosity differences as described in the supporting information), respectively. Additional data are provided in the supporting information.

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

If mixing/binding is slow compared to the time scale of the experiment, the binding constants may be expected to be higher than suggested from the fitted binding isotherm. It might therefore be expected that mixing analyte and indicator before introduction into the capillary would result in a more pronounced change in dispersion as the binding equilibrium would be established prior to the FIDA experiment. At antibody concentrations lower than 500 nM, the protocol involving capillary mixing (protocol I) and the pre-mixing procedure gave similar results, indicating that the reported affinity constants are not influenced by non-equilibrium conditions. However, at higher concentrations of the antibody (1000 nM), the pre-mixing experiment resulted in lower apparent indicator diffusivity (see supporting information). This is similar to the high dose hook effect seen in other immunoassays25 where the signal intensity decreases at high analyte concentrations. This has been reported to be caused by saturation26, similar protein variants27 and fibrillation28. In the present study we interpret the decrease in apparent indicator diffusivity to be caused by immuno-precipitation as a decrease in indicator peak area is simultaneously observed. The shorter reaction time when the binding process occurs inside the capillary during the hydrodynamic flow prevents precipitation of the dsDNA – antibody complex. The pre-mix experiments demonstrate an important aspect of the assay as it shows that indicator peak area, in addition to peak shape, can be used as a convenient quality control of assay performance. The matrix tolerance of the FIDA assay was first tested by measuring the binding isotherm in 20% normal control plasma without autoantibodies as shown in figure 3. Due to autofluorescence of plasma, the injection procedure was slightly modified as described in protocol II. In Figure 3 it can be seen that the binding and diffusivity was not influenced by the presence of 20% plasma, and, after correction for viscosity differences (see supporting information), only to a minor extent in 85% plasma. A standard fused silica capillary may also be used for plasma content up to 20%, but in general the results obtained using the uncoated capillary had a significantly larger standard deviation (see supporting information). The binding curve may be used as a standard curve for obtaining the concentration of the monoclonal antibody against dsDNA in unknown samples. The sensitivity in terms of limit of detection is then determined by the strength of the binding and the standard deviation of the measured diffusivities. A thorough statistical analysis (t-test, P < 0.05) of the data results in a limit of detection of the monoclonal antibody of 9.3 nM in a sample matrix composed of 20 % plasma (see supporting information). To further investigate specificity, a standard curve was prepared using another IgG antibody, but against beta-2microglobulin, as the analyte. In this case, the indicator diffusivity and peak area did not change with increasing antibody concentrations (data not shown). It may thus be concluded that the assays, as expected, is not probing an antibody isotype. Two additional DNA sequences (DNA-II and DNA-III) were used for the characterization of patient samples without the need for any further assay development (see supporting information). In accordance with recent findings4,6,11, the model monoclonal antibody showed pronounced sequence specificity as it was found not to interact with DNA-II and III (data not shown). As the anti–dsDNA antibody was found to bind dsDNA with high sequence specificity it seemed plausible that

Page 4 of 7

sequence specific binding might also be observed for human SLE anti-dsDNA autoantibodies. Any increase in dispersion reflects a binding event which is governed by affinity as well as concentration; the human SLE anti-dsDNA autoantibodies can therefore not be quantified using the standard curve displayed in figure 2. Still, human SLE anti-dsDNA antibodies were unequivocally detected via a decrease in the apparent indicator diffusivity. It is thus the anti-dsDNA binding capacity of the SLE patient samples which is probed, and which reflects a combination of human SLE anti-dsDNA antibody concentration as well as binding affinity, respectively. Six plasma samples from patients diagnosed with SLE based on clinical symptoms in a prior study (see supporting information)21,29,30 were analyzed applying the three DNA probes. Sample 1-3 were found to be positive for human SLE anti-dsDNA antibodies using CLIFT and ELISA (Figure 4a), whereas sample 4-6 were CLIFT and ELISA negative and diagnosed with SLE solely on the basis of clinical symptoms (Figure 4b). To ease the comparison between the three indicators, the relative rather than absolute diffusivity is given in Figure 4. All diffusivities were corrected for differences in plasma viscosity (see supporting information). In order to increase the assay sensitivity, the plasma concentration was increased from 20 to 85%, as it was found that the higher plasma content did not affect assay performance (figure 3). Even at the highest plasma content of the samples, peak area and shape indicated optimal performance of the assay. Figure 4 shows that the human SLE anti-dsDNA antibodies found in the plasma samples from SLE patients exhibit sequence specific binding, as was also the case for the monoclonal anti-dsDNA antibody used in the assay optimization. This finding may partially explain the discrepancies found between different SLE assays where a plethora of different DNA sources are used3,4,11,14,31,32. A t-test is used to descriminate if patient indicator diffusivity is statistically different from that of the healthy control. Patient 2 and 3 are positive towards one or more of the applied DNA-sequences (p < 0.05). Patient 3 suffered from nephritis as the only patient and had a higher Systemic Lupus Erythematosus Disease Activity Index (SLEDAI) score at the time of the blood sampling (see supporting information). Notably, patient 3 exhibits a somewhat different binding pattern than the other patients which may suggest that the FIDA assay could be used for patient stratification and monitoring of SLE disease activity, although this should be confirmed in a more extensive study. In the FIDA method the viscosity of the sample matrix can be obtained as part of the procedure as changes in viscosity cause a proportional change in the observed peak appearance time. The plasma viscosity contains in itself useful information, as an increased viscosity is a non-specific marker of inflammation33,34, but a more detailed analysis of this parameter was outside the scope of the present study.

ACS Paragon Plus Environment

Page 5 of 7

Analytical Chemistry classes.

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

Figure 4. Plasma samples from six SLE patients analysed with three different DNA indicators (DNA I-III). The indicator diffusivity is given relative to the diffusivity in plasma from healthy donors. Plasma samples where anti-dsDNA have been found with the standard assays CLIFT and ELISA are shown in A and plasma samples where no anti-dsDNA have been found with the standard assays CLIFT and ELISA are shown in B. An asterisk (*) indicates the measurements where there is a significant change in the indicator diffusivity compared with blank plasma (p < 0.05). All plasma samples have been investigated in triplicate. The relative diffusivities were corrected for differences in viscosity as described in the supporting information. Patient 6 has only been investigated with two out of three sequences due to limited availability of the human plasma sample.

A direct comparison between the FIDA method and an ELISA assay is challenging as it would require surface immobilization and optimization of the three DNA probes. In Figure 5 is shown the results of a modified ELISA assay employing a large dsDNA strand which at least offers an opportunity to compare the dynamic range of the two assays. Both assays show a dynamic range of two orders of magnitude. As seen in Figure 3 the FIDA response does not change when the plasma concentration in the buffer is increased from 0 to 20%. ELISA of antibodies against dsDNA is normally performed at 1% plasma concentration. To give a better comparison of the assays ability to detect antibodies in plasma the antibody concentration in ELISA is corrected for the lower plasma concentration by multiplication by 20. In ELISA, a large dsDNA strand is employed which undoubtedly will be able to bind a larger number of anti-dsDNA antibodies, resulting in better sensitivity. On the other hand, the FIDA assays provides a facile approach for assessing the sequence specificity of the human SLE anti-dsDNA binding which may have potential for addressing disease activity as well as for patient stratification. It should also be noted that the DNA sequences employed in the FIDA setup was not systematically optimized for their binding affinities. An optimized binding would of course also result in a higher sensitivity. In most ELISA assays a secondary antibody is used which binds to the human anti-dsDNA antibody. Often the secondary antibody is specific for a certain class (e.g. IgG) of human SLE anti-dsDNA antibodies, excluding detection of other

Figure 5. ELISA and FIDA signals pertaining to binding to different dsDNA. The black ELISA curves have been prepared in phosphate buffer while the red FIDA results have been prepared in 20% plasma. Error bar from triplicate measurements are shown for the FIDA results. The concentration of the monoclonal model antibody in the ELISA assay is corrected for the extra dilution needed in ELISA (normally performed in 1% plasma) by multiplying by 20.

Furthermore, the washing steps employed in ELISA assays may remove human SLE anti-dsDNA autoantibodies with low affinity for dsDNA35. This drawback is not present in the FIDA method, as antibodies of low affinity will contribute to the binding of the indicator and will thus be detected if present in sufficient concentration.

CONCLUSIONS FIDA presents a new approach in the detection of human circulating anti-dsDNA antibodies. Among the important qualities are the short analysis time (minutes) combined with a modest sample volume consumption of a few hundred nanoliters. The FIDA methodology was developed to quantify a model monoclonal antibody against dsDNA in both buffer and plasma as well as to detect autoantibodies against dsDNA in six plasma samples from Lupus patients. The ability to provide a more sequence specific antibody characterization by using short and well defined dsDNA sequences is an important feature of the FIDA method. The dsDNA sequences can even be interchanged without the need for assay optimization. The added value of using multiple dsDNA sequences is clearly apparent as human SLE anti-dsDNA antibodies are found in a higher number of patients by combining the results from assaying different dsDNA sequences. Further, the assay is very robust and can be performed even in 85% plasma. The potential for patient stratification and an increased specificity in diagnosis may provide a useful approach in the treatment of SLE patients as well as in the development and test of new drugs against the disease. Finally it should be stressed that the simplicity of the FIDA methodology holds promise for transferring the methodology into a miniaturized assay for e.g. point-of-care testing of autoantibodies as well as other disease related biomarkers36. The realization of these perspectives requires extended validation and characterization in larger, well-characterized cohorts of patients suffering from various autoimmune conditions.

ASSOCIATED CONTENT

ACS Paragon Plus Environment

Analytical Chemistry Supporting Information

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

The Supporting Information is available free of charge on the ACS Publications website. The supporting information contains additional experimental details and results.

AUTHOR INFORMATION Corresponding Author *E-mail: [email protected]

Author Contributions All authors have given approval to the final version of the manuscript.

Notes Henrik Jensen and Jesper Østergaard are co-founders of FIDATech ApS aiming at commercializing the FIDA technology.

ACKNOWLEDGMENT Financial support from the Danish Council for Independent Research, Technology and Production Science (grant: 11-106647) is gratefully acknowledged.

REFERENCES (1) Petri, M.; Orbai, A. M.; Alarcon, G. S.; Gordon, C.; Merrill, J. T.; Fortin, P. R.; Bruce, I. N.; Isenberg, D.; Wallace, D. J.; Nived, O.; Sturfelt, G.; Ramsey-Goldman, R.; Bae, S. C.; Hanly, J. G.; SanchezGuerrero, J.; Clarke, A.; Aranow, C.; Manzi, S.; Urowitz, M.; Gladman, D.; Kalunian, K.; Costner, M.; Werth, V. P.; Zoma, A.; Bernatsky, S.; Ruiz-Irastorza, G.; Khamashta, M. A.; Jacobsen, S.; Buyon, J. P.; Maddison, P.; Dooley, M. A.; van Vollenhoven, R. F.; Ginzler, E.; Stoll, T.; Peschken, C.; Jorizzo, J. L.; Callen, J. P.; Lim, S. S.; Fessler, B. J.; Inanc, M.; Kamen, D. L.; Rahman, A.; Steinsson, K.; Franks, A. G.; Sigler, L.; Hameed, S.; Fang, H.; Pham, N.; Brey, R.; Weisman, M. H.; McGwin, G.; Magder, L. S. Arthritis Rheum. 2012, 64, 2677-2686. (2) Compagno, M.; Rekvig, O. P.; Bengtsson, A. A.; Sturfelt, G.; Heegaard, N. H.; Jönsen, A.; Jacobsen, R. S.; Eilertsen, G. Ø.; Fenton, C. G.; Truedsson, L.; others. Lupus science & medicine 2014, 1, e000007. (3) Hernando, M.; González, C.; Sánchez, A.; Guevara, P.; Navajo, J. A.; Papisch, W.; González-Buitrago, J. M. Clin. Chem. Lab. Med. 2002, 40, 1056–1060. (4) Buhl, A.; Page, S.; Heegaard, N. H. H.; von Landenberg, P.; Luppa, P. B. Biosens. Bioelectron. 2009, 25, 198-203. (5) Cerutti, M. L. J. Biol. Chem. 2001, 276, 12769-12773. (6) Uccellini, M. B.; Busto, P.; Debatis, M.; Marshak-Rothstein, A.; Viglianti, G. A. Immunol. Lett. 2012, 143, 85-91. (7) Compagno, M.; Jacobsen, S.; Rekvig, O.; Truedsson, L.; Heegaard, N.; Nossent, J.; Jönsen, A.; Jacobsen, R.; Eilertsen, G.; Sturfelt, G.; Bengtsson, A. Scand. J. Rheumatol. 2013, 42, 311-316. (8) Haugbro, K.; Nossent, J. C.; Winkler, T.; Figenschau, Y.; Rekvig, O. P. Ann. Rheum. Dis. 2004, 63, 386-394. (9) Merbl, Y.; Zucker-Toledano, M.; Quintana, F. J.; Cohen, I. R. J. Clin. Invest. 2007, 117, 712-718. (10) Hochberg, M. C. Arthritis Rheum. 1997, 40, 1725-1725. (11) Antico, A.; Platzgummer, S.; Bassetti, D.; Bizzaro, N.; Tozzoli, R.; Villalta, D.; (SIMeL), S. G. o. A. D. o. t. I. S. o. L. M. Lupus 2010, 19, 906-912. (12) Dieker, J. W.; Sun, Y.-J.; Jacobs, C. W.; Putterman, C.; Monestier, M.; Muller, S.; van der Vlag, J.; Berden, J. H. J. Immunol. Methods 2005, 296, 83-93. (13) Comley, J. Drug Discov. 2012, 23. (14) Žigon, P.; Lakota, K.; Čučnik, S.; Švec, T.; Ambrožič, A.; Sodin-Šemrl, S.; Kveder, T. Croat. Med. J. 2011, 52, 694-702. (15) European Medicines Agency, Committee for Medicinal Products for Human use, 2015.

Page 6 of 7

(16) Poulsen, N. N.; Andersen, N. Z.; Østergaard, J.; Zhuang, G.; Petersen, N.; Jensen, H. Analyst 2015, 140, 4365-4369. (17) Taylor, G. P. Roy. Soc. A-Math. Phy. 1953, 219, 186-203. (18) Cottet, H.; Biron, J.-P.; Cipelletti, L.; Matmour, R.; Martin, M. Anal. Chem. 2010, 82, 1793-1802. (19) Cottet, H.; Biron, J.-P.; Martin, M. Anal. Chem. 2007, 79, 9066-9073. (20) Poulsen, N. N.; Ostergaard, J.; Petersen, N. J.; Daasbjerg, K.; Iruthayaraj, J.; Dedinaite, A.; Makuska, R.; Jensen, H. J. Sep. Sci. 2016, Submitted. (21) Nielsen, C. T.; Ostergaard, O.; Johnsen, C.; Jacobsen, S.; Heegaard, N. H. H. Arthritis and Rheum. 2011, 63, 3067-3077. (22) Jensen, H.; Østergaard, J. J. Am. Chem. Soc. 2010, 132, 40704071. (23) Mahler, H.-C.; Huber, F.; Kishore, R. S. K.; Reindl, J.; Rückert, P.; Müller, R. J. Pharm. Sci. 2010, 99, 2620-2627. (24) Heegaard, N. H. H.; Olsen, D. T.; Larsen, K.-L. P. J. Chromatogr. A 1996, 744, 285-294. (25) Leboeuf, R.; Langlois, M. F.; Martin, M.; Ahnadi, C. E.; Fink, G. D. J. Clin. Endocr. Metab. 2006, 91, 361-364. (26) Erickson, J. A.; Grenache, D. G. Clin. Chim. Acta 2016, 452, 120-123. (27) Nerenz, R. D.; Song, H. W.; Gronowski, A. M. Clin. Chem. 2014, 60, 667-674. (28) Lu, C.-H.; Kalamar, B.; Malaspina, A.; Greensmith, L.; Petzold, A. J. Neurosci. Methods 2011, 195, 143-150. (29) Nielsen, C. T.; Ostergaard, O.; Stener, L.; Iversen, L. V.; Truedsson, L.; Gullstrand, B.; Jacobsen, S.; Heegaard, N. H. H. Arthritis Rheum. 2012, 64, 1227-1236. (30) Ostergaard, O.; Nielsen, C. T.; Iversen, L. V.; Tanassi, J. T.; Knudsen, S.; Jacobsen, S.; Heegaard, N. H. H. Arthritis Rheum. 2013, 65, 2680-2690. (31) Slater, N. G.; Cameron, J. S.; Lessof, M. H. Clin. Exp. immunol. 1976, 25, 480. (32) Wold, R. T.; Young, F. E.; Tan, E. M.; Farr, R. S. Science 1968, 161, 806-807. (33) Harkness, J.; Whittington, R. B. Rheol. Acta 1971, 10, 55–60. (34) Késmárky, G.; Kenyeres, P.; Rábai, M.; Tóth, K. Clin. Hemorheol. Micro. 2008, 39, 243–246. (35) Bobrovnik, S. A.; Demchenko, M.; Komisarenko, S.; Stevens, F. J. Mol. Recognit. 2010, 23, 448-456. (36) Hatch, A.; Kamholz, A. E.; Hawkins, K. R.; Munson, M. S.; Schilling, E. A.; Weigl, B. H.; Yager, P. Nat. Biotechnol. 2001, 19, 461-465.

ACS Paragon Plus Environment

Page 7 of 7

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

Insert Table of Contents artwork here

ACS Paragon Plus Environment