Conflict of interest: Dirk Roggenbuck is a shareholder of GA Generic Assays GmbH and Medipan GmbH. Both companies are diagnostic manufacturers.
Multiplex assessment of non-organ-specific autoantibodies with a novel microbead-based immunoassay†
Article first published online: 3 JAN 2011
Copyright © 2011 International Society for Advancement of Cytometry
Cytometry Part A
Volume 79A, Issue 2, pages 118–125, February 2011
How to Cite
Grossmann, K., Roggenbuck, D., Schröder, C., Conrad, K., Schierack, P. and Sack, U. (2011), Multiplex assessment of non-organ-specific autoantibodies with a novel microbead-based immunoassay. Cytometry, 79A: 118–125. doi: 10.1002/cyto.a.21009
- Issue published online: 24 JAN 2011
- Article first published online: 3 JAN 2011
- Manuscript Accepted: 19 NOV 2010
- Manuscript Revised: 17 NOV 2010
- Manuscript Received: 18 OCT 2010
- German Federal Ministry of Education and Research (BMBF, PtJ-Bio). Grant Number: 0313909
- BMBF Innovation Initiative “Entrepreneurial Regions,” InnoProfile-Projekt. Grant Number: 03 IP 611
- Brandenburg Ministry of Economics and European Union. Grant Number: 80133708
- HEp-2 cell;
- microbead-based immunoassay;
Advances in immunofluorescence assay development paved the way for the simultaneous detection of several antibodies in one sample, for the serological diagnosis of systemic rheumatic diseases. Standardized automated screening of such antibodies can be achieved by HEp-2 cell-based indirect immunofluorescence (IIF) using a multicolor fluorescence imaging technical platform. To create a common platform for both screening and specific antibody assessment, multiplex measurement of antibodies using fluorescence-coded immobilized microbeads was employed on the same platform. The multicolor fluorescence detection system VideoScan (AKLIDES®) was used for the fluorescence analysis of a multiplex microbead-based immunoassay (MIA). First, immunoglobulin G (IgG) was covalently coupled to one microbead population in duplicate and in three independent experiments. The coupled IgG was detected by a Cy™5-conjugated secondary antibody. Thus, intra- and interassay coefficients of variation (CV) were obtained. Second, a multiplex determination of antinuclear autoantibodies (ANA) to Scl-70, Sm, dsDNA, SS-A (Ro60), CENP-B, and La/SS-B by solid-phase MIA was investigated, using 72 sera from patients with autoimmune diseases such as systemic lupus erythematosus and systemic sclerosis (SS). The reproducibility study revealed intra-assay CVs ranging from 3.2% to 9.9%, and interassay CVs ranging from 9.6% to 14.7%. The detection of Scl-70-, Sm-, CENP-B-, and La/SS-B-ANA with MIA showed very good agreement with the ELISA results (kappa = 1.0). The resulting relative sensitivities and specificities for Scl-70-, Sm-, CENP-B-, dsDNA-, and La/SS-B-ANA were 100%, respectively, with the exception of dsDNA (specificity 97%). Multiplex detection by immobilized fluorescence-coded microbeads using multicolor fluorescence is a reliable method for the assessment of rheumatic-disease-specific antibodies. Multicolor fluorescence analyses with pattern detection algorithms provide a common platform technique for both the screening of ANA by cell-based IIF and specific antibody assessment by multiplex detection. © 2011 International Society for Advancement of Cytometry
Antinuclear antibodies (ANA) to extractable nuclear antigens (ENA) are hallmarks in the serological diagnosis of systemic rheumatic diseases (1). The detection of ANA by indirect immunofluorescence (IIF) employing HEp-2 cells as a multiple antigen source has been established as the “gold standard” in routine diagnostics (2). However, advances in assay development and recombinant technology have paved the way for the detection of ANA to individual ENA, such as double-stranded DNA (dsDNA), improving the diagnostic power of ANA testing. The growing variety of ANA found in different systemic rheumatic diseases has generated the need for innovative techniques to overcome the shortcomings of single ENA detection, with regard to cost and time taken to obtain results (3). Hence, multiplexed platforms have been developed recently to meet the demand for simultaneous assessment of several ANA in one sample. Multiplex assays based on fluorescent microbead flow cytometry and microarray systems have proven to be powerful tools, supporting greater throughput analysis and more comprehensive testing of multiple patient samples (4–11). The novel detection technology presented in this study employs fluorescence-coded microbeads detected by a multicolor fluorescence image-capture based system with novel pattern recognition algorithms for multiplex testing (12). The assay characteristics and clinical value of the determination of ANA to six defined ENA are investigated and discussed, as an aid in the serological diagnosis of systemic rheumatic diseases. The interpretation system can also be used for the automated reading of HEp-2 cell-based IIF for the detection of ANA (13) and would be the first combined platform for both ANA screening and differentiation. © 2010 International Society for Advancement of Cytometry
Materials and Methods
Overall, 72 sera (female n = 63, age from 17 to 80 years, mean age 46 years and male n = 9, age from 12 to 78 years, mean age 43 years) with positive ANA reactivities from patients with systemic lupus erythematosus (SLE; n = 34; thereof 9 with suspicion of SLE), systemic sclerosis (SS; n = 30; thereof n = 7 with suspicion of SS), Sjögren's syndrome (n = 2), mixed connective tissue disease (MCTD; n = 3), vasculitis (n = 1), arthritis (n = 1), and SLE SS overlap syndrome (n = 1) were obtained from a routine laboratory (Institute of Immunology, Medical Faculty of the Technical University, Dresden, Germany). The study was approved by the local ethics committee (EK 226112006).
In total, seven carboxylated polymethylmethacrylate microbead populations, containing two coding fluorescent dyes (Rhodamine, Coumarin) (emission wavelengths: 502 nm (dye1), 555 nm (dye2)) (PolyAn GmbH, Berlin, Germany), were employed. Different ratios of dye1 and dye2 and variation of microbead size from 10 to 20 μm were used for individual identification of each population (Pop ID) and offered a high multiplex grade of microbead populations (Fig. 1).
Components of the VideoScan (AKLIDES®) Platform
An inverse fluorescence microscope (IX81, Olympus GmbH, Hamburg, Germany) with a motorized scan stage (IM 120 × 100, Märzhäuser GmbH & Co. KG, Wetzlar, Germany) was used. To excite the fluorescent dyes, a Xenon arc lamp (75 W) was employed as a light source. The images were detected by a grayscale camera (DX2HC-FW, Sony CCD-chip ICX 285 AL, Kappa optoelectronics GmbH, Gleichen, Germany). For the correct allocation of each microbead population after image focusing, microbead classification was conducted by PopID and microbead size detection (Fig. 2), using novel pattern recognition algorithms (14).
Intra-assay and Interassay Precision Study with Microbead-Based Immunoassay
One microbead population was coated with IgG (Sigma-Aldrich GmbH, Hamburg, Germany) according to Nustad et al. and Wood and Gadow (15, 16), in duplicate and in three independent experiments. In summary, six groups of IgG-coated microbead populations were obtained. Subsequently, eight wells of a 96-well microtiter plate (Greiner Bio-One GmbH, Frickenhausen, Germany) were immobilized with each group of IgG-coated microbead population using 0.05 M PBS pH 7.4, respectively. Next, IgG was detected by incubation with 100 μl of 1:100 diluted anti-human IgG-Fcγ Cy™5-conjugate (Dianova GmbH, Hamburg, Germany) in 0.1 M PBS/Tween (PBS/T) for 1 h at room temperature (RT) on a reciprocal shaker (650 rpm; Eppendorf GmbH, Hamburg, Germany). Unbound components were removed by washing each well three times with 100 μl of 0.1 M PBS/T for 5 min at RT while shaking. Fluorescence of processed microbeads was read with VideoScan. Intra-assay and interassay precision was calculated as the mean CV of the six groups of IgG-coated microbead populations, within eight wells, and in three independent experiments. Interpopulation CV determined the reproducibility of IgG coating of the six groups in 3 days.
Multiplex ANA Assessment with Microbead-Based Immunoassay
Five bead populations were coated with one of the following recombinant (r) or native (n) polypeptide antigens: rLa/SS-B, nCENP-B, rSS-A (Ro60, Diarect AG, Freiburg, Germany), or nScl-70 (Arotec Diagnostics Limited, Wellington, New Zealand) according to Nustad et al. and Wood and Gadow (15, 16). A further bead population was coated with herring testes DNA (Clonetech Laboratories Inc., Mountain View, CA). One bead population was coated with IgG, as a positive control for secondary antibody reactivity. Antigen and IgG-coated microbeads were mixed in equal numbers and immobilized onto the solid phase of Greiner 96-well microtiter plates (Greiner Bio-One GmbH, Frickenhausen, Germany), using 0.05 M PBS.
The principle of ANA detection with microbead-based immunoassay (MIA) is shown in Figure 3. Each well was incubated with 100 μl of 1:100 diluted serum samples of SLE and SS patients, in 0.1 M PBS/T 0.1%, for 1 h at RT on a reciprocal shaker (650 rpm; Eppendorf GmbH, Hamburg, Germany). Unbound components were removed by washing each well three times, with 100 μl of 0.1 M PBS/T 0.1%, for 5 min at RT while shaking. Specific ANA were detected by incubation with 100 μl of 1:100 diluted anti-human IgG-Fcγ Cy5 conjugate, in 0.1 M PBS/T 0.1%, for 1 h at RT while shaking. After another wash cycle, as described earlier, the last wash solution was removed and each well was filled with 100 μl of 0.1 M PBS. After that, the microbeads within the 96-well microtiter plate were analyzed by VideoScan. The fluorescence of each microbead was read automatically by the system, (i) to classify each microbead population and (ii) to detect the fluorescence of Cy5-secondary antibody. The latter reflects the specific ANA reactivity of the serum sample and is expressed as the mean fluorescence intensity (MFI). The fluorescence analysis was based on the investigation of at least 100 microbeads of each population for both classification of microbead populations and analysis of Cy5-secondary antibody.
The Kolmogorov-Smirnov test was used to test the data for normal distribution. The Mann-Whitney test (independent samples) and Wilcoxon test (paired samples) was performed to test for statistically significant difference. Cohen's kappa test was applied for group comparison. P values of less than 0.05 were considered to be significant. Calculations and receiver operating characteristics (ROC) curve analysis were performed using MedCalc® statistical software (MedCalc, Mariakerke, Belgium).
Intra- and Interassay Precision Study
The intra-assay and interassay CVs of six groups of IgG-coated microbead populations are shown in Table 1. The intra-assay CV at day 1 ranged from 3.5% to 9.9%, at day 2 from 3.2% to 7.9%, and at day 3 from 4.8% to 8.9%. Subsequently, the following interassay CVs were obtained for each of six microbead populations, respectively: 8.1%, 5.7%, 15.1%, 11.5%, 6.1%, and 1.6%. The interpopulation CVs of IgG-coating to six microbead populations at 3 days were as follows: day 1: 14.7%, day 2: 11.8%, and day 3: 9.6%.
|Charge of IgG-coated microbead population|
|Intra-assay CV (%), n = 8|
|Interassay CV (%), n = 3|
|Interpopulation CV (%), n = 6|
Multiplex Assessment of ANA
Seven microbead populations coated with Scl-70, Sm, dsDNA, Ro60, CENP-B, La/SS-B, and IgG as a positive control, were used for multiplex assessment of ANA in 72 patient sera. After MIA and fluorescence analysis with VideoScan, the obtained MFIs were subjected to ROC curve analysis to obtain cut-off values for each ANA to its corresponding antigen (Fig. 4). The following cut-offs were calculated for the discrimination of positive and negative ANA to CENP-B, La/SS-B, Ro60, dsDNA, Sm, and Scl-70: 0.14; 1.25; 1.23; 0.28; 0.46; 0.22, respectively. Based on these cut offs, results from MIA were compared with ELISA data.
Patients with SS
Scl-70- and CENP-B-ANA are typical marker antibodies for the classification of early SS (17). In sera of the 23 patients with SS, the following ANA frequencies were obtained: anti-Scl-70 = 11, anti-Sm = 0, anti-dsDNA = 2, anti-Ro60 = 2, anti-CENP-B = 11, and anti-La/SS-B = 2. The median MFI of 11 Scl-70-ANA positive patient sera was 1.67 (95% CI = 1.28 to 2.05) and of 11 CENP-B-ANA positive patient sera 4.50 (95% CI = 3.69 to 4.73). In contrast, the median MFIs of the 12 Scl-70-ANA negative and 12 CENP-B-ANA negative patient sera were significantly lower according to the Mann-Whitney test (0.16; 95% CI = 0.12 to 0.22; P < 0.0001 and 0.11; 95% CI = 0.08 to 0.14; P < 0.0001, respectively). The group of dsDNA- and Ro60-ANA was too small for statistical analysis.
Patients with SLE
The diagnosis of SLE was classified by the criteria of the American College of Rheumatology. Marker antibodies are ANA to dsDNA, Sm, and ANA with high titers (18). In sera of the 25 SLE patients, the following ANA frequencies were found: anti-Scl-70 = 2, anti-Sm = 3, anti-dsDNA = 17, anti-Ro60 = 12, anti-CENP-B = 4, and anti-La/SS-B = 5. The median MFI of 17 dsDNA-ANA positive patient sera was 0.93 (95% CI = 0.70 to 1.22) and of 12 Ro60-ANA positive patient sera 2.96 (95% CI = 1.38 to 4.67). In contrast, the median MFIs of seven dsDNA-ANA and 11 Ro60-ANA negative patient sera were significantly lower according to the Mann-Whitney test (0.10; 95% CI = 0.07 to 0.14; P < 0.0001 and 0.36; 95% CI = 0.26 to 0.46; P < 0.0001, respectively). The group of Scl-70-, Sm-, CENP-B-, and La/SS-B-ANA positive sera was too small for statistical analysis.
Patients with Sjögren's Syndrome, Vasculitis, Arthritis, SLE SS Overlap Syndrome, and MCTD
The international consensus criteria for the diagnosis of Sjögren's syndrome contain as marker antibodies ANA to Ro/SS-A (Ro60) and La/SS-B (19). The two tested Sjögren's syndrome patient sera were positive for Ro60- (MFI = 6.35 and 5.22) and La/SS-B-ANA (MFI = 4.87 and 4.66).
Ro60-ANA (MFI = 4.87) were found in one serum from a patient with vasculitis. Scl-70-ANA (MFI = 0.98) and Ro60-ANA (MFI = 1.26) were detected in one serum of a patient with SLE SS overlap syndrome. The serum of the patient with arthritis demonstrated no ANA.
For the diagnosis of MCTD, marker antibodies to U1-RNP were described (20). The detection of U1-RNP-ANA was not possible with the antigen panel of the developed MIA. In fact, no Scl-70-, Sm-, dsDNA-, Ro60-, CENP-B-, and La/SS-B-ANA were found in the three sera of the patients with MCTD.
Comparison of MIA Results to Prediagnostic Findings by ELISA
The detection of Scl-70-, Sm-, dsDNA-, Ro60-, CENP-B-, and La/SS-B-ANA with MIA was compared with prediagnostic findings by ELISA. The determination of positive and negative results with MIA was based on the cut offs determined according to ROC curve analysis. The prediagnostic findings of ELISA were obtained from a routine diagnostic laboratory (Institute of Immunology, Medical Faculty of the Technical University, Dresden, Germany).
The inter-rater agreement (Cohen's kappa [κ]) is shown in Table 2. The detection of ANA to Scl-70, Sm, CENP-B, and La/SS-B with MIA is in perfect agreement with the ELISA results (κ = 1.0). For ANA to dsDNA, we found a very good agreement (κ = 0.961), and for ANA to Ro60, a good agreement (κ = 0.783) compared with ELISA results. The resulting relative sensitivities and specificities shown in Table 3 demonstrated perfect consistency for ANA detection to Scl-70, Sm, CENP-B, and La/SS-B (sensitivity and specificity 100%, respectively), a very good consistency for anti-dsDNA (sensitivity 100%, specificity 97%), and an acceptable consistency for anti-Ro60 (sensitivity 86%, specificity 94%) (Fig. 5).
|95% CI||80% to 100%||54% to 100%||85% to 100%||64% to 97%||78% to 100%||69% to 100%|
|95% CI||54% to 100%||80% to 100%||83% to 100%||70% to 100%||40% to 100%||77% to 100%|
Finally, the data obtained by both ELISA and MIA were compared by multiple line graphs using ELISA (IU/ml) and MIA (MFI/Cy5)-results of SLE and SS patients. For comparison, the results were scaled to 100% for both tests. ANA negative patient sera showed lower values with MIA in sera of patients with SS tested for anti-CENP-B (Fig. 6A) and in sera of patients with SLE tested for anti-Sm (Fig. 6C, Table 4). The latter difference was statistically significant according to the Wilcoxon test (P < 0.0038). Patient sera tested positive for CENP-B-, Ro60-, and Scl-70-ANA demonstrated higher values by MIA in sera of patients with SS (Fig. 6A, Table 4) and SLE (Figs. 6B and 6D). ANA to CENP-B were significantly higher in sera of patients with SS (P < 0.0026).
|Patient group||ANA||Number||IU/ml||MFI||IU/ml (%)||MFI (%)|
The aim of our study was to assess a novel MIA for the multiplex detection of ANA in patients with autoimmune diseases. The novel imaging-platform technology VideoScan (AKLIDES®), which has recently been shown to be useful for the automated assessment of cell-based IIF assays in the serology of rheumatic diseases, was employed (13, 21, 22). There is a continuously growing need for cost efficient autoantibody testing technology that may combine both screening by IIF and multiplex evaluation of specific antibodies, on one platform. Current approaches for multiplex testing such as Luminex® or microarray technologies do not appear to be applicable in this context since they do not provide pattern recognition (13).
In this study, microbeads coded by fluorescent dyes and size were immobilized onto the surface of 96-well microtiter plates, for the detection of disease-specific autoantibodies by fluorescent imaging microscopy. In contrast to the Luminex® technology, bound-free separation was achieved by simple plate washing, not requiring vacuum filtration or centrifugation. Novel pattern recognition algorithms were employed to read the specific reactivity, given in MFI, to six autoantigenic targets coated onto the microbeads. Thus, fluorescent images of immobilized microbeads were taken for microbead determination, classification (PopID), and analysis of Cy5-conjugated secondary antibody used. Hence, immobilized microbeads omitting the need for sophisticated microfluid handling by fluorescence activated cell sorting (FACS) technology.
The first assessment of solid-phase MIA was a reproducibility study, with six IgG-coated microbead populations, and the corresponding detection of IgG in three independent experiments. The intra-assay CVs, from 3.2% to 9.9%, and interassay CVs, from 1.6% to 15.1%, were within the guidelines for this type of assay of <10% and <20%, respectively (23). Reasons for variance in intra-assay and interassay CVs could be the common variability of microbead size and carboxylation present in microbead manufacturing. These effects can lead to the unbalanced coating of IgG and accordingly also of other proteins or of DNA, onto the microbeads. This in turn leads to higher variability of IgG or autoantibody detection.
Other studies with the basic principle of MIA reported intra- and interassay CVs for the detection of Ro52-, Ro60-, and La/SS-B-antibodies from 2.8% to 12.5%, and from 6.5% to 14.5%, respectively (24). In comparison to MIAs, the well-established ELISA method for the detection of Ro52-, Ro60-, and La/SS-B-antibodies offers an intra-assay CV ≤ 10.2% and interassay CV ≤ 12% (25).
The second part of this study comprised the solid-phase multiplex detection of ANA employing Scl-70-, Sm-, dsDNA-, Ro60-, CENP-B, and La/SS-B-coated microbeads immobilized in 96-well microtiter plates. Sera from patients with systemic rheumatic diseases such as SS and SLE were investigated. In the case of SS patients, we found high frequencies of ANA to Scl-70 and CENP-B as expected according to the criteria of LeRoy and Medsger, 2001 (17). In the investigated sera of patients with SLE, ANA to dsDNA, and Sm were demonstrated as expected according to the criteria of Tan et al., 1982 (18).
The detection of ANA to Scl-70, Sm, dsDNA, Ro60, CENP-B, and La/SS-B with MIA was compared with findings of an established routine test system (ELISA). For comparison, Cohen's kappa was determined and showed very good agreement for ANA to Scl-70, Sm, CENP-B, La/SS-B, and dsDNA. The relative sensitivities and specificities of MIA, compared with ELISA, for ANA to Scl-70, Sm, CENP-B, dsDNA, and La/SS-B was in each case 100%, with the exception of the lower but very good specificity of anti-dsDNA (97%). For some ANA reactivities, significant lower scaled values for negative samples and, in contrast, higher values for positive samples demonstrated a better diagnostic performance of MIA in comparison with ELISA.
In conclusion, the VideoScan platform (AKLIDES) provides a reliable and robust method for the multiplex detection of autoantibodies, for the differential diagnosis of systemic rheumatic diseases. The newly developed MIA enables the detection of multiple ANA as separate entities in one sample at the same time. Multiplex ANA testing is more efficient than the conventional ELISA technique. It is less time consuming compared with singleplex ANA assessment and can work with small sample volumes. This is of great importance particularly regarding pediatric patients. In contrast to other multiplex technologies, novel pattern recognition allows the assessment of cell-based IIF in addition to multiplex testing. Further studies are warranted to prove the cost effectiveness and reliability of this unique combination for screening and multiplex testing of autoantibodies.
The authors want to thank Ingo Berger for his excellent graphics.
- 12Technical solutions and advances for microbead-based immunoassays. In: ConradK,ChanEKL,FritzlerMJ,HumbelRL,von LandenbergP,ShoenfeldY, editors. From Pathogenesis to Therapy of Autoimmune Diseases. Lengerich/Berlin/Bremen/Miami/Riga/Viernheim/Wien/Zagreb: Pabst Science Publishers; 2009. pp 234–255., , , , .
- 14A new focus algorithm for automated microscopy. Cell Prolif 2005; 38: 177., , , , , .
- 15Covalent coupling of proteins to monodisperse particles: Preparation of solid phase second antibody. Agents Actions Suppl 1982; 9: 207–212., , , , , .