Automated quantitative microarray assay platforms

A microarray-based assay platform has been developed for high test throughput, with a fully automated protocol.

Figure 1. SQiDworks diagnostics platform.

Protein microarrays have become powerful tools for examining various aspects of the immune system. Investigators can now test for the presence of multiple specific (multiplexed) antibody reactivity in patient samples with one reaction by using antigen microarrays spotted on planar substrates.1 For example, such microarrays can identify clinical subtypes of rheumatoid arthritis.2 However, unique challenges exist for developing clinically useful microarrays that reproducibly provide multiple simultaneous and quantitative patient test results. These challenges mandate technical advances in such areas as capture analyte printing and probing, in-assay calibration and multiple marker signal normalization, and complex software algorithms to ensure consistent measurable results.3

In addition, when target proteins are expressed in the normal and disease states across a wide range of concentrations, the simultaneous measurement of both low- and high-abundance proteins in the same reaction needs special consideration. Such challenges are worth overcoming because of the enhanced clinical utility of simultaneous quantitative protein measurement in biomarker-based diagnostic, prognostic, and drug efficacy assays.4,5

Current enzyme-linked immunosorbent assays (ELISAs) affix an unknown amount of an antigen to a surface, and a specific antibody is washed over the surface to bind to the antigen. Enzyme-linked antigen/antibody complexes emit a fluorescent signal with intensity proportional to the amount of antigen in the sample. For most autoimmune diseases, no single biomarker provides a high enough clinical sensitivity and specificity to indicate and confirm the disease process.

For rheumatoid arthritis (RA), many laboratories offer a test menu with an RA panel, in which a patient sample is examined using four separate tests. For each rheumatoid factor (RF)—IgM, IgG, IgA, and anti-CCP IgG—ELISA protocols are used to detect the presence of an antibody or antigen in a sample. However, despite the recognized potential benefits of multiplex testing, no FDA-approved, commercial multiplex assay for simultaneous testing of multiple antigens and classes or isotypes of antibodies is available as a laboratory test.

The QuantiSpot Rheumatoid Arthritis Assay by SQI Diagnostics Inc. (Toronto), which is licensed for sale in Canada, and the IgX Plex Rheumatoid Arthritis Qualitative Assay, which has been submitted to FDA for 510(k) Premarket Notification Clearance, are sensitive CCP IgG assays multiplexed with three additional analytes: rheumatoid factors IgA, IgG and IgM. These multiplex assays simultaneously test these three RF biomarkers and anti-CCP IgG in one sample reaction, and obtain quantitative or qualitative multiplex results, respectively, for diagnosing rheumatoid arthritis. The quality control (QC) in the sample reaction provides built-in confidence in the test results that is not possible with a traditional manual ELISA. The assay also reduces lab technician input and the potential for operator error since this test is a fully automated, internally calibrated and integrated device.

The SQiDworks Diagnostics Platform is an automated 96-well microtiter plate robotic platform that integrates and automates the manual aspects of the assay protocol. The platform was designed to be configurable for all future QuantiSpot and IgX Plex assays developed. Moreover, the design of the QuantiSpot RA Assay provides the basis for future multiplex, quantitative assays that are calibrated to a recognized and traceable international standard. The flexible architecture of the platform and microarray characteristics that are built with SQI Diagnostics' ISO 13485-compliant manufacturing processes provides the framework for the design, development, and manufacture of a menu of QuantiSpot multiplex assays, and may allow for the rapid development of custom assays to support research or clinical needs.

Automated Microarray Multiplex Diagnostic Platform

The SQiDworks platform fully automates an immunoassay protocol from end-to-end, including sample pipetting, serum dilution, incubation, washing, and drying (see Figure 1). The instrument integrates an automated pipetting station, a fluorescent scanner, washing and drying stations, and ancillary hardware components under dedicated instrument control. The software controls perform scheduling, task verification, data acquisition, data management, algorithms analysis, and results reporting.

The automated robotic processing of QuantiSpot multiplex diagnostic assays reduces costs and the time required to process patient tests due to the following: simple-to-use platform, automated sample identification and consumable barcode tracking, up-front verification of all loaded materials for load-and-go performance, advanced liquid handling and robotic automation of the assay protocol, and simultaneous multiple analyte detection and quantification for each patient sample in each test well.

In addition, the automation of the multiplex test arrays translates into calibrated patient test results with a high degree of integrity owing to the following: automated scanning of assayed plate and analysis of fluorescence using internal quality controls and comparative standards; in-well normalization providing intra-lot, lot-to-lot, site-to-site, and operator-to-operator consistency; data invalidation algorithms providing enhanced internal quality controls; and integrated data output.

High-Throughput Microarray Assay Plate

A QuantiSpot assay uses a standard, Society for Biomolecular Screening–sanctioned 96-well microtiter plate format (see Figure 2). Depending on the assay, at least 73 wells are available for testing individual samples (i.e., one well per patient). When four analytes are tested, such as with the QuantiSpot RA assay, the total number of results per plate is 292. A single plate's processing time of 192 minutes results in a throughput of 0.38 patient samples, or 1.52 test results, per minute. The SQiDworks platform is designed to run up to three individual plates in one batch, allowing processing and analysis of 219 patient samples, or 876 test results. With an estimated runtime of 360 minutes, this process would have a throughput of 2.43 results per minute.

[1]

Figure 2. (click to enlarge [1]) QuantiSpot microarray plate.

The QuantiSpot RA microarray plate has protein and antibody replicate spots covalently bound to the surface of coated glass in each well of a 96-well assay plate (see Figure 2). One well is enlarged in Figure 2, which shows the array layout.

Figure 3 shows the magnified microarray spots as printed in a single QuantiSpot RA well. Each well has the identical configuration of microarray elements, which include capture, normalization, and control subarrays.

Microarray Normalization and In-Well Quality Control

Signal and data normalization represents a challenge for microarray-based technologies, and becomes even more accentuated when applied to multiplex quantitative arrays. This problem has been mitigated in the QuantiSpot RA Assay and other assays in development by utilizing GMP control protocols and integrating consistent solid-phase coating, plate uniformity in manufacturing, and built-in normalizing controls in the assay.

[2]

Figure 3. (click to enlarge [2]) QuantiSpot microarray well layout (patent pending).

The assays are calibrated using external standards traceable to recognized reference standards and an internal normalization process. In every test well, the assay contains a series of normalization microarray spots. This normalization series minimizes well-to-well and plate-to-plate variations. When variations occur, algorithms automatically normalize the capture signal to ensure a comparable signal response across equivalent test spots in different wells and on different plates. The normalized values are then translated to a standardized result. In the case of RF analytes, the assay standards calibrate to the international reference standard (WHO, British First standard 64/2). Since there is no equivalent standard recognized for the anti-CCP antibody analyte, it is standardized against a commercially available IVD predicate test.

Current ELISA tests have no in-well QC mechanisms. At the very least, laboratories routinely perform duplicate tests to detect sample test failure. While this practice is effective, testing samples in replicate significantly affects time and testing costs. The multiplexed QuantiSpot assays have integrated in-well QC that mitigates errors such as no sample added, improper testing operations, and incorrect reagent addition—any of which may lead to false negative or false positive results. Such controls ensure that invalid test results at the analyte, well, or plate level are not reported. In-well QC of microarray assays offers additional assurance for obtaining correct patient test results the first time, in addition to significant labor, time, and cost reductions.

Platform Use and Assay Operation

Following a ten-minute startup initiation routine of the SQiDworks platform, an operator loads the bar-coded reagents for the specific assay (serum diluent and reporter antibody), the patient samples in standard bar-coded serum vials, and the QuantiSpot microarray device into the fluidics station. The automated system ensures that only samples with valid barcode and reagents are loaded before the assay is performed. The operator responds to user prompt screens to confirm the type of assay being run, the number of patient samples loaded, and that the reagents being loaded are correct. The prompt screens also provide an option to rescan any unregistered bar codes.

The SQiDworks platform automatically completes analyses of the entire immunoassay protocol. For the RA assay, the process includes the following steps: sample dilution, incubation, a wash series, reporter tagging incubation, a second wash, and a final plate drying. Upon completing these operations, the fluorescence scanner reads the plate. The test results are analyzed according to specific assay algorithms. The set of final test results is formatted and reported for each sample, based on its identifying barcode. The status bar on the viewing interface tracks and monitors every step of the assay process to completion. The user is continuously informed of the status of the test run (see Figure 4).

[3]

Figure 4. (click to enlarge [3]) Status window indicating the current stage of the assay.

The QuantiSpot RA Assay is the first test developed for the SQiDworks platform. The platform accepts various patient sample tube sizes, the smallest being a 1.5 mL tube requiring a minimum of 60 µL of serum sample. The QuantiSpot RA Assay uses immunochemistry similar to predicate ELISA tests and targets the same analytes that are currently used to diagnose RA. QuantiSpot assay immunochemistry is based on a sandwich design for analyte capture and signal detection through a reporter antibody, using a sensitive fluorescent microarray scanner.

QuantiSpot and IgX Plex assays differ from other multiplex assays in that they can multiplex the different varieties of antibodies known as isotypes. For example, the QuantiSpot RA assay detects and measures the IgG, IgA, and IgM isotypes for each antigen. This ability to multiplex isotypes is incremental to the ability to multiplex capture responses. These assays provide time and labor savings advantages compared with traditional assay methods, and offer increased utility of test results at the antibody class level.

[4]

Figure 5. (click to enlarge [4]) IgX-Plex assay principle.

With the IgX Plex assay principle, two capture antigens, an IgG protein and CCP-based capture peptides, along with the control proteins, are printed in each test well (see Figure 5). Analytes in the samples are captured during the sample incubation step. Excess analytes are removed by a plate washer. The antihuman antibodies in the reporter mix tagged with a fluorescent dye are then added. After a second incubation step, the excess reporter reagents are removed by washing the wells in the integrated plate washer. The wash buffers are completely aspirated, and the wells are dried to allow for proper reading of fluorescent signals. The microarray signal for each spot is read separately by using three optical channels. The software completes the analysis with statistical methods to ensure proper representation of the signal for each type of spot obtained. The software also normalizes and standardizes data to produce and report the final results for the tested samples. The whole process is illustrated in Figure 6.

Study Results

The QuantiSpot RA assay claims were derived from blinded validation studies that consisted of the following three main components: clinical sensitivity and specificity, method correlation, and multisite reproducibility.

[5]

Figure 6. (click to enlarge [5]) The QuantiSpot rheumatoid arthritis assay flowchart.

Clinical Sensitivity and Specificity. Sensitivity and specificity were independently calculated for RA patients versus each disease classification and against normal controls. Non-RA diseases with rheumatic conditions, such as systemic lupus erythematosus (SLE) which is known to contain RF but not cyclic citrullinated peptide (CCP) autoantibody, were tested. Selected infectious disease samples of Epstein-Barr virus and syphilis were compared as they contain cross-reactive antibodies to CCP.6

[6]

Table I. (click to enlarge [6]) Clinical sensitivity and specificity.

Validation studies showed that when analyzed on the automated SQiDworks platform, multiplexed QuantiSpot RA assays have demonstrated reliable performance for clinical sensitivity and specificity (see Table I).

Method Correlation. Validation studies compared QuantiSpot RA test results run on the SQiDworks platform to manual predicate IVD methods (see Table II). Using 350 samples, the results of this method-correlation study demonstrated a positive correlation in diagnostic outcome to predicate devices currently available on the market. FDA-cleared predicate methods that were used for this study included the Quanta Lite ELISA assays by Inova Diagnostics Inc. (San Diego) for all four analytes.

Multisite Reproducibility. A blinded multisite study demonstrated the reproducibility of the QuantiSpot RA test on the SQiDworks platform when performed by different operators at different sites. Table III illustrates that the assays run on the platforms at each of the sites consistently and accurately determined the same diagnostic outcome.

[7]

Table II. (click to enlarge [7]) Method correlation including confidence interval (CI).

The study results indicated that the site-to-site reproducibility of the QuantiSpot RA Assay is commensurate with other methods using multiplexed microarrays. More importantly, the concordance of the final results among samples that were tested in multiple runs and analyzed for between-lot, between-operator, and between-site performance was good.

The SQiDworks platform also enables and sustains high sensitivity. One study concluded that microarray spots would have enhanced sensitivity due to increased signal density and low analyte depletion, which has been confirmed by SQI Diagnostics' research.2,7 The results for detection limits are summarized in Table IV.

Pending Multi-Assay Panels

[8]

Table III. (click to enlarge [8]) Site-to-site agreement.

The launch of the RA assay panel (RF-IgA, IgG, IgM, anti-CCP IgG) is scheduled for late 2009 in Canada and early 2010 in the European Union. The U.S. launch date is to be determined, pending regulatory clearance. SQI Diagnostics' QuantiSpot 9-plex thrombosis assay is currently being developed and will have the potential to report the analysis of 2,052 discrete patient biomarker results in one three-plate run.

SQI Diagnostics also has a strategic plan to complete a number of critical test panels to aid clinicians in diagnosing various autoimmune diseases such as antiphospholipid syndrome, lupus, celiac disease, and irritable bowel syndrome.

Conclusion

In one study, researchers printed microarrays containing immobilized antigens at different concentrations to determine simultaneously titers of different antibodies in a single immunoassay experiment.8 The results confirmed that microarray immunoassays can be very sensitive and highly specific with little or no cross-reactivity for nonspecific proteins. The study also proposed to adapt fluorescence-labeled probes and include standardized antibodies labeled with a second fluorophore that are similar to the internal standards used in DNA chip technology. Including an internal standard for each antigen in one assay could allow quantitative analysis.

[9]

Table IV. (click to enlarge [9]) Summarized limit of detection (LOD) and limit of quantitation.

Solid phase-bound sandwich immunoassays are a major achievement in the field of protein quantitation, especially when combined with highly sensitive multispot and multianalyte immunoassay image scanning detectors. Miniaturized assay systems have to demonstrate precision, sensitivity, and reliability that is suitable for rapid analysis and amenable to automation.9 The simultaneous incubation of an assay with two or more reporters in the same detection mixture allows multiplex quantitation of target antigen and antibody classes including IgA, IgG, and IgM.2

In summary, up to 30% of patients with RA may not have RF and CCP autoantibodies, especially early in the disease process. In addition, the presence of RF, especially at a low level, is not uncommon in patients who do not have and may not develop RA. The anti-CCP antibody is often associated with RA, and if it is elevated without typical symptoms of RA, they may develop RA. Early detection of anti-CCP antibodies would benefit patients.

Currently, differential RA diagnosis also needs additional blood tests including ESR and CRP. These serum tests confirm inflammation and are typically elevated in patients with active RA. While normal levels do not rule out RA, such patients may be less likely to develop damage in their joints than others with high levels of inflammation, especially elevated CRP. Although detection of antinuclear antibody (ANA) signifies SLE, low levels of ANA are also common in RA. High levels of ANA may also indicate lupus, especially if the CCP and RF are negative. In subsequent clinic visits, RF and CCP tests are not typically reordered if previously found positive. However, ESR and CRP tests are frequently ordered as they can help to confirm whether the arthritis is active or in remission.

In the future, to increase the clinical utility of the QuantiSpot RA assay to aid in diagnosing and treating rheumatoid arthritis, second-generation multiwell microarrays could incorporate additional currently clinically relevant serum tests such as anti-CCP IgA to confirm diagnosis, disease treatment, and control, and improve patient outcomes.

References

1. MM Ling, C Ricks, and P Lea, “Multiplexing Molecular Diagnostics and Immunoassays Using Emerging Microarray Technologies,” 7, no. 1 (2007): 87-98.

2. P Biessels, et al., “A Novel Anti-Citrullinated Peptide Antibody Assay Using a Four-Analyte Multiplexed QuantiSpot Rheumatoid Arthritis Test System,” poster presentation at the American Association for Clinical Chemistry 40th Oak Ridge Conference, April 17-18, 2008.

3. JE Larkin, et al., “Independence and Reproducibility Across Microarray Platforms,” 2 (2005): 337- 44.

4. M Hartmann, et al., “Expanding Assay Dynamics: A Combined Competitive and Direct Assay System for the Quantitation of Proteins in Multiplexed Immunoassays,” 54 (2008): 956-63

5. P Lea and MM Ling MM, “New Molecular Assays for Cancer Diagnosis and Targeted Therapy,” 10, no. 3 (2008): 251-259.

6. C Anzilotti, et al., “Antibodies to Viral Citrullinated Peptide in Rheumatoid Arthritis,” 33 (2006): 647-51.

7. RP Ekins, “Ligand Assays: From Electrophoresis to Miniaturized Microarrays,” 44 (1998): 2015-2030.

8. D Stoll, et al., “Protein Array Technology,” 7 (2002): 13-32.

9. MF Templin, et al., “Protein Microarrays and Multiplexed Sandwich Immunoassays: What Beats the Beads?,” (2004): 223-229.

Peter Lea, PhD, is the founder and chief science officer of SQI Diagnostics Inc. (Toronto). He can be reached at plea@sqidiagnostics.com [10]

Thomas E. T. O'Connor is senior vice president, sales and marketing at SQI Diagnostics Inc. (Toronto). He can be reached at toconnor@sqidiagnostics.com [11]

Kate Smith is vice president, technology at SQI Diagnostics Inc. (Toronto). She can be reached at ksmith@sqidiagnostics.com [12]

Claude Ricks is president and chief executive officer at SQI Diagnostics Inc. (Toronto). He can be reached at cricks@sqidiagnostics.com [13]

Author:

Peter Lea, Thomas E. T. O'Connor, Kate Smith, and Claude Ricks