Development of a High-throughput Mass Spectrometry-based SARS-Cov-2 Immunoassay

Abstract

The serious impact of the Covid-19 pandemic underscores the need for rapid, reliable, and high-throughput diagnosis methods for infection. Current analytical methods, either point-of-care or centralized detection, are not able to satisfy the requirements of patient-friendly testing, high demand, and reliability of results. Here, we propose a two-point separation on-demand diagnostic strategy that uses Laser Desorption/Ionization Time of Flight Mass Spectrometry (LDI-TOF MS) and adopts a stable yet cleavable ionic probe as mass reporter. The use of this reporter enables ultra-sensitive, interruptible, storable, restorable, and high-throughput on-demand detection. We describe a demonstration of the concept whereby we (i) design and synthesize a laser-cleavable reporter (DTPA), (ii) conjugate the reporter onto an antibody and verify the function of conjugate, (iii) detect with good turnaround and high sensitivity the conjugated reporter, (iv) analyze quantitatively by using a laser-cleavable internal standard, and (v) identify negative and positive samples containing the spike protein. The protocol has excellent sensitivity (amol for the SARS-CoV-2 Spike S1 subunit antibody) without any amplification. This strategy is also applicable for detection of other disease antigens besides SARS-CoV-2.

Graphical Abstract

The Covid-19 pandemic has caused significant morbidity and mortality worldwide since the outbreak at 2019^1,2. The severity of the Covid-19 pandemic has also stimulated development and expedited application in many scientific fields, including vaccine technology, drug development, and testing. Diagnostic tests by real-time reverse transcription-polymerase chain reaction (real-time RT-PCR) were quickly established after the viral genome sequence was released^3,4. Although RT-PCR remains the most dependable approach to date for Covid detection, it has several constraints. They include: (i) the lack of information on patients who recovered from Covid, owing to a cleared viral load after recovery; (ii) a limited number of tests that can be performed per day; (iii) the need for additional tests to cover a shortage of several instruments and reagents, severely limiting applicability for large-scale screening^5,6. In addition, epidemic prevention, control, and clinical diagnosis demand several analytical systems. For prevention and control, rapid, frequent, and large-scale screening of the population is needed. In this regard, nucleic acid amplification-based approaches are neither practical nor convenient⁶, motivating the development of new analytical methods. All of these factors drive this research to develop alternative diagnostic resolutions^7–9.

Point-of-care (POC) tests are the most appealing for advancing large-scale population screening¹⁰. Currently, hundreds of initiatives are on track to convey serologic and antigen detection platforms. Serologic immunoassays targeting IgA, IgM, or IgG provide historic information about viral exposure, but no current information on virus shedding. By contrast, antigen tests directly detect the presence of shed virus, but their sensitivity, especially in the early or acute phases of an infection, is not well developed¹¹. The results require abundant antigen for visual identification and depend on the observers’ visual acuity.

An on-demand, two-point separation diagnostic method may enable an on-site assay and off-site centralized detection and provide reliability and accuracy in POC detection. This approach will: (i) enable patient-friendly testing, (ii) avoid degradation of samples, and (iii) allow detection at a later time in a centralized fashion to improve reliability and archiving of results. Here we describe combining an immunoassay with Mass spectrometry (MS) to achieve this goal.

MS has advantages of multiplexing, low detection limits, large dynamic range, and high spectral resolving power¹². MS has become an important tool in clinical laboratories and is the current gold standard for several clinical applications¹³. Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-MS) is now often used to identify microbial species in clinical samples¹⁴, establishing its wide availability compatible with the demands of a pandemic.

There are two rising MS-based methodologies for SARS-CoV-2 testing: MALDI-TOF MS and liquid chromatography-tandem MS (LC-MS/MS). One example of MALDI-MS-based SARS-CoV-2 detection depends on spectral characteristics picked by machine learning, but the chemical identities of the detected markers remain unknown^15–17. Another example uses LC-MS/MS-based targeted proteomics^18,19.

Here, we describe steps toward a new MS immunoassay analysis platform that employs cleavable ionic reporter and enables on-demand analysis of biomarkers (Figure 1a). This should be a sensitive and high-throughput analytical platform for which 2,5-dioxopyrrolidin-1-yl-3-(tritylthio)propanoate (DTPA) was synthesized as a mass reporter. The reaction of the NHS ester (the conjugation unit) in DTPA is favorable given the numerous amino groups in the antibody arms (Figure 1b). Under laser irradiation in a conventional MALDI-TOF-MS source, the weaker carbon-sulfur bond breaks, releasing a tertiary carbocation for detection by TOF-MS. LDI-TOF-MS needs only 0.5 μL of sample and has a high-throughput because many samples can be spotted on a single plate for introduction to the spectrometer. Here, we report a demonstration of the feasibility of this design and its ultrahigh sensitivity without any signal amplification. This stable reporter makes it possible to interrupt, store the sample, and make repeated detections of antigens, after sample loading and immunoreaction, thereby allowing quantitative analysis now and/or later.

Figure 1.

(a) Schematic illustration of high-throughput detection of the Anti-SARS-Cov-2 by LDI-TOF-MS. (b) Structure of DTPA, its bioconjugation with Anti-SARS-Cov-2, and the release of the ionic reporter by LDI.

We synthesized DTPA, a simple and efficient laser-cleavable probe, by a one-step amidation reaction (Scheme S1). The synthesized DTPA was characterized by ¹H NMR (Figure S1) and ESI MS (Figure S2). To ensure applicability, we tested trityl cation release under LDI-MS (Figure S3) to see in a mass spectrum a strong signal for the relatively stable trityl cations of m/z 243 in the positive-ion mode. The signal intensity is high even under low laser intensity, indicating efficient release of the trityl cations under laser (355 nm) irradiation.

The MS-based immunoassay workflow consists of four steps: (i) fabrication of the capture antibody-reporter conjugation, (ii) removal of the excess DTPA, (iii) immunoreaction to capture the targeted antigen, and (iv) MS analysis. To ensure the success of the detection, we conjugated DTPA to the antibody. We chose myoglobin as a model and tested the reactivity of DTPA with the protein (Figures 2(a) and (b) show the mass spectra of denatured apo-myoglobin before and after conjugation, respectively). After deconvolution, we clearly see conjugation of up to seven reporters per myoglobin (MW 16951 Da). We also confirmed the conjugation by MALDI-TOF MS (Figure S4). The conjugation was also verified on β-lactoglobulin (Figure S5).

Figure 2.

Denaturing ESI mass spectra of (a) aMB and (b) the reporter-aMB conjugate. (c) & (d) deconvolved mass spectra from (a) and (b), respectively.

Several MS-based analysis methods use a reporter for biomarkers.^20–24 None of them, however, gives convincing evidence that their reporters are linked firmly with the capture carrier. In fact, most reporters interact with the capture carrier via noncovalent interactions (e.g., electrostatic, van der Waals), introducing unstable interactions. Thus, those designs are vulnerable to false positive or negative results. Although there are reports of high specificity, special care is needed for those workflows. One of our aims is to reduce the rate of false positives by ensuring that every reporter in the system is firmly conjugated (covalently) to the capture carrier to avoid observation of itinerant reporter ions causing false positives after washing.

After demonstrating that DTPA is conjugated to the protein, we optimized the conjugation of DTPA to SARS-CoV-2 Spike S1 subunit antibody. Owing to limited amounts of Spike S1 antibody, we first optimized the conjugation condition for a homolog, human IgG1 (humanized monoclonal antibody). We tested two different ratios of (lysines in protein):DTPA of 1:1 and 1:0.5 and found high conjugation levels of up to 7 reporters per one human IgG1 even at the lysine:reagent ratio of 1:0.5 (Figure 3). We then used native MS to confirm the successful conjugation of DTPA onto the antibody and to provide evidence that the high-order structure of the antibody is maintained after conjugation to (a) human IgG1 (Figure 3), (b) deglycosylated human IgG1, and (c) reporter-bound human IgG1 conjugation followed by deglycosylation. Because glycosylation leads to a complex mixture and makes reporter human IgG1 characterization difficult (Figure 3a), we submitted the sample to deglycosylation (Figure 3b, c). We did not observe any changes of charge state after conjugation, suggesting that the native structure of human IgG1 is retained upon conjugation, a needed outcome for later antibody-antigen recognition. After optimization of conjugation conditions for human IgG1, we repeated the reaction on SARS-CoV-2 Spike S1 subunit antibody and found similar high conjugation efficiency (Figure S6).

Figure 3.

Native MS of (a) control antibody, (b) deglycosylated antibody, and (c) reporter-antibody conjugate. Deconvoluted mass spectra of (d) control antibody, (e) deglycosylated antibody, and (f) reporter-antibody conjugate.

We found that a positive charged reporter, like the trityl group, can be easily released from the reporter-antibody conjugate during LDI MS detection (no matrix) in the residence time in the ion source (μs-ms). Besides the trityl ion of m/z 243, we see another finger print ion of m/z 165, representing the 9-flurenyl cation formed by losing C₆H₆ from the trityl cation (Figure S7), as reported previously²⁵. Fortunately, the reporter is stable under neutral conditions, even after months of storage. Both signals can be used for quantification because their intensities are directly proportional to concentration. Here, we chose the signal for the ion of m/z 243 because it has higher intensity than that of the m/z 165 ion. Good linearity was obtained (Figure S7b, Figure 4) over a dynamic range of concentration of greater than four orders of magnitude, with a limit of detection (LOD) of 18 pM (9 amol) for 1:1 conjugation and 72 pM (36 amol) for 1:0.5 conjugation ratio. High sensitivity is expected for MS detection of a charged reporter that is released upon irradiation.

Figure 4.

Relative quantification for reporter-antibody conjugation. y = 5.58x + 1.56, R² = 0.96, n = 3.

Acquisition of quantitative information is crucial to diagnosis and therapeutic purposes, but quantitation is still a challenge to date. We designed an internal standard (Figure 5, Figure S8), which has similar structure and ionization performance to the reporter (structure in Figure S7), but with each aryl group substituted with methyl groups. The reporter and internal standard generate signals at m/z 243 and m/z 285, respectively (Figure 5a). Good quantitative linearity was obtained (Figure 5b, c) from 800 amol to 7.6 pmol. Therefore, our design achieved absolute quantification with high sensitivity and reproducibility.

Figure 5.

(a) LDI-TOF MS spectra of reporter-antibody conjugate monitored with the addition of an internal standard. X axis: m/z, Y axis: concentration (μM), Z axis: intensity. (b) Calibration curve, linear range from 800 amol to 6 fmol, y = 2.56x + 0.013, R² = 0.96, n = 3. (c) Calibration curve, linear range from 6 fmol to 7.6 pmol, y = 0.0047x - 1.34, R² = 0.93, n = 3.

To validate that the Antibody conjugate maintained its structure and function and to demonstrate that the functionalized antibody still binds the antigen, we designed a sandwich-structure ELISA experiment (details in SI). The ELISA assay of SARS-Cov-2 spike protein and reporter-antibody was performed on a 96-well microplate. First, we absorbed the reporter-anti-SARS-Cov-2 as the capture antibody. Different concentrations of Sars-Cov-2 spike protein were added to evaluate the binding capability of the functionalized antibody. HRP-Anti-SARS-Cov-2 spike protein antibody was added to report the absorbance signal. The result (Figure 6) shows that from 4.5 nM (4.5 fmol) to 0.070 nM (70 amol), the absorbance has a linear response with the concentration of spike protein. At concentrations higher than 4.5 nM, the absorbance saturated. At concentration lower than 0.070 nM, the absorbance signal was background and too low to differentiate concentrations. This result also indicates that the sensitivity of the proposed MS method is comparable to that recorded by us with enzyme-amplified ELISA, although no amplification occurs in the MS method.

Figure 6.

The binding ability of reporter-antibody conjugation proved by ELISA.

To prove the effectiveness of the MS-based immunoassay, we prepared “healthy samples” as nasal swab solutions and SARS-Cov-2 “positive samples” (by spiking the SARS-Cov-2 Spike protein into a nasal swab solution). We submitted the samples to the protocol for antigen capture and detection, as described in SI. The results (Figure 7(a) and (b)) show nasal swab solutions from three healthy persons and three “positive” (spiked) samples. There is a significant difference between the “healthy” and the “positive samples” as seen by comparing the MS peak intensities. Here, there is no background signals for “negative” samples unlike in fluorescence. The difference is substantiated by a t-test, and the positive sample was statistically different, based on a p-value threshold of < 0.05 (Figure 7c). The whole MS detection of the six samples after sample preparation took less than one minute. We can imagine that when applied to large-group screening, 400 samples can be detected in an automated fashion within 1 h on one target plate.

Figure 7.

Detection results for SARS-Cov-2. (a) Mass spectra of “healthy samples”. (b) Mass spectra of three “positive samples”. (c) Comparison of the healthy and positive groups.

Although fluorescence detection in any immunoassay has advantages of sensitivity, throughput, and on-site testing, MS offers additional benefits. First is flexibility through centralized analysis, improving reliability and sample processing efficiency. These holds promise for early-stage and large-group screening at the outbreak of a pandemic when the population is unaware of the symptoms. Second, the MS-based strategy greatly simplifies the conjugation process by using a single reaction once the capture antibody is immobilized compared with other immunoassays. Third, MS ensures greater accuracy by providing specific signals based on the reporter m/z, thereby reducing the background signal compared to ELISAs by 10-100x while also reducing false positives. Fourth, multiplexing and sensitivity are also enhanced with MS, allowing fast detection of a large number of samples at high sensitivity, as demonstrated by our low limit of detection. Fifth, MS enhances selectivity through precise m/z measurement of the incorporated cleavable ionic reporter, enabling quantitative analysis. Unlike the limited “channel” for fluorescence-based detection, the detection channel for MS is nearly unlimited because it detects m/z and even accurate m/z. Thus, many specific reporters like trityl cation deployed in this work can be used with the same protocol (Fig. S9) for future multichannel detection. These advantages collectively make MS a compelling choice for any immunoassay strategy.

In summary, we demonstrated a novel LDI-TOF MS-based enzyme-free platform for future MS immunoassay approach, establishing the concept and rationally designing a reporter for a high-throughput LDI-TOF MS-based immunoassay platform. The design of cleavable ionic reporters to replace enzymatic reactions in an immunoassay enable on-demand quantitative analysis. This approach is suitable not only for SARS antigen but also for other disease antigens that have identified antibodies (e.g., middle east respiratory syndrome, cancer antigens, and carcinoembryonic antigens). The protocol provides a promising option to traditional in-hospital or point-of-care testing by creating opportunities for self-testing, followed by analysis and diagnosis after sending the sample to a central facility.