Multiplex proteomics allows researchers to analyse proteins in two or more different sample preparations simultaneously. When used quantitatively, multiplex proteomics can detect and quantify protein biomarkers in biological samples such as blood, to aid disease screening and diagnostics and guide timely treatment decisions. Multiplex proteomics can also provide deep insights into biological and disease processes, while eliminating the assay-to-assay variation that occurs when samples are analysed in parallel.
In this article, we provide an overview on the potential of quantitative multiplex proteomics within diagnostics and disease monitoring and look at some of the methodologies in use today. But before we dig into all of that, let’s start with a quick recap on what a typical proteomics workflow looks like.
Mass spectrometry-based proteomics
The predominant approach to proteomics is ’bottom-up ’or ’shotgun proteomics’, which is a multistep process that involves enzymatic digestion of proteins into peptides, separation of those peptides via liquid chromatography, and identification of the separated peptides via mass spectrometry. The combination of liquid chromatography (LC) and mass spectrometry (MS) is referred to as LC-MS.
Although less frequently used, top-down proteomics workflows have also been developed; these involve the analysis of intact proteins by mass spectrometry. Top-down methods have been reviewed elsewhere and will not be discussed further in this article (1, and references therein).
A typical shotgun proteomics workflow usually includes the following steps:
1. Protein extraction. Here, samples (e.g. cells or tissues) are lysed using one or more of the following: mechanical disruption, liquid homogenisation, sonication, freeze/thaw cycles, and grinding with a mortar and pestle. The resulting lysates may be fractionated by ultracentrifugation, e.g., to separate the nuclear and cytosolic fractions, and this may be followed by enrichment and/or isolation of particular proteins of interest (affinity purification), or removal of interfering or contaminating substances.
2. Enzymatic digestion. The serine protease trypsin is the most widely used enzyme for digesting proteins into peptides for proteomics applications. Trypsin cleaves peptide bonds at the C-termini of arginine and lysine residues and yields peptides with an average size of 700-1500 daltons, which is the ideal range for mass spectrometry.
3. High-performance liquid chromatography (HPLC) separation. Here, the mixture of peptides are separated according to their polarity, based on their relative affinities for a separation column.
4. Mass spectrometry (MS). The separated peptides from Step 3 are further resolved based on mass and charge in the presence of electric and magnetic fields in a mass spectrometer. In practice, Steps 3 and 4 often happen in a single LC-MS instrument. Since it is not possible to identify peptides solely based on their intact masses, the peptides within the mixture are often deliberately fragmented within the mass spectrometer after an initial round of MS, to yield fragment ions that can be identified to amino acid level including post-translational modifications. This is known as tandem mass spectrometry (tandem MS or MS/MS).
5. Database searching or software-based protein quantification and identification using the LC-MS data generated in step 3 and 4.
A multiplex MS-based proteomics workflow involves the same overall steps as described above, with the critical difference that chemical tagging of digested peptides allows multiple samples to be analysed simultaneously.
The major benefit of sample multiplexing within proteomics, regardless of the method used, is improved statistical robustness due to the ability to compare and analyse high numbers of biological samples together while also including an internal standard or quality control for normalisation. On a practical level, multiplexing allows more data to be generated from small volume samples, increases throughput and may be more cost- and time-effective than running monoplex assays in parallel.
Quantitative multiplex proteomics in diagnostics
The goal of most clinical diagnostic assays in use today is to measure the concentration of certain proteins in biological samples such as blood, urine and sputum. Monoplex immunoassays such as conventional ELISA are still considered to be the gold standard within diagnostics and monitoring response to therapy, in particular for diseases caused by viruses, parasites and insect-borne pathogens.
Despite the proven utility of monoplex diagnostic assays, the complexity of diseases – e.g., the involvement of many proteins and pathways and the progression of disease over time – means that making conclusions based on the presence of a single analyte or biomarker may increase the risk of misdiagnosis and inappropriate treatment, especially for disease screening or during early stages of disease where symptoms may not be present.
Furthermore, while monoplex proteomics may aid in diagnosis, information on a single biomarker does not provide insight into the mechanism of disease. As one example, this limitation is illuminated in the post-pandemic era by the many open questions about the potential long-term health implications of SARS-CoV-2 infection, especially for those experiencing long-COVID, and those who experienced mild symptoms and/or did not seek medical care.
By allowing the quantification of large numbers of proteins simultaneously, multiplex proteomics can aid our understanding of diseases with rare or unknown causes, and provide deeper insights about how disease affects individuals during early, intermediate and late stages.
The possibility to understand how an individual’s immune response responds to disease and treatment will also help to guide clinicians in devising personalised treatment plans, and help to identify novel biomarkers for serious disease risk. Moreover, the absolute quantification data that results from quantitative multiplex proteomics assays can be readily exchanged between researchers in clinical and research laboratories to facilitate data integration and intra-lab comparisons.
Relative quantification is possible with MS-based proteomics
During LC-MS, peptides are eluted from a thin opening in the separation column (Step 4 above) directly in front of the mass spectrometer. During this process, peptides become ionised or charged as a result of the voltage applied between the opening of the separation column and the inlet of the mass spectrometer. Many factors affect the efficiency of the ionisation process, which means that peptides will not be uniformly charged. Thus, the number of ions present inside the mass spectrometer is never a direct indicator of the number of peptides present in the original sample.
It is possible to compare ions representing the same peptide from different samples, assuming similar ionisation efficiencies. Here, samples are processed and analysed separately as described in the steps above. Following LC-MS, the peak intensities from each run are compared to calculate relative changes in protein abundance. This approach, known as label-free proteomics, is relatively cost-effective, easily scalable albeit with low throughput, and may be useful in revealing changes in protein abundance on a proteome-wide scale between two or more samples, e.g., between healthy and diseased tissues, or in response to drug exposure or varying culture conditions. The drawbacks of label-free proteomics are the need for multiple runs, large variability among low-abundance proteins, and the fact that a significant number of peptides may be missed in every run.
Another approach to relative quantification involves the use of stable isotope labels to discriminate between samples. Here, the samples to be analysed are differentially labeled with a stable isotope either by addition of a label to cells in culture or in vivo, or following trypinisation. The labelled samples are then combined and simultaneously processed by LC-MS. The ratio of peak intensity between the ions of an isotope pair gives the relative difference in abundance of the protein from which the peptide is derived. While this approach is inherently less error-prone than label-free proteomics, its capacity is limited to two or three multiplexed samples per MS run, it does not provide the absolute quantification needed for diagnostic applications, and it may not capture low-abundance proteins.
So, how is quantitative, multiplex proteomics carried out?
Several approaches to quantitative multiplex proteomics have been developed, and these can roughly be divided into 1) mass spectrometry-based methods and 2) affinity proteomics methods, which involve the use of antibodies or other binding reagents as protein-specific detection probes. In the remaining sections of this article, we will summarise the predominant multiplex proteomics methods in use today.
Multiplex proteomics with mass spectrometry
As mentioned earlier, the addition of chemical tags to enzymatically digested peptides has made it possible to run multiplex MS-based proteomics experiments. Specifically, the development of isobaric mass tags about 20 years ago allowed researchers to address many of the limitations of label-free proteomics and heavy isotope labelling.
An isobaric tag is a chemical reagent that can be used to covalently modify a peptide. These tags can be obtained commercially in sets, whereby all tagging reagents in each set have the same nominal mass (i.e., they are isobaric) and chemical structure. Each tag is composed of a reactive group, a balance group (also known as a spacer arm or mass normaliser), and a mass reporter. The isobaric nature of these tags allows simultaneous identification and quantification of proteins in different samples using MS/MS. Various isobaric tags are commercially available and on average, current technologies allow multiplexing of between two and ten samples.
A typical workflow for a multiplex experiment using isobaric mass tags is as follows:
- Each trypsin-digested sample is labelled with one of a set of isobaric labels.
- The labelled samples are mixed and subjected to MS/MS.
- In the first round of MS, the same peptide from different samples will be identical in molecular weight and will appear as a single peak in the mass spectrum.
- Samples of the same peak are collected to perform the second round of MS (ionisation). Here, the bond between the balance group and the reactive group is broken and the balance group is lost. Now, the same peptide with different isotopic labels yields different masses of reporter ions, and the reporter ions exhibit different peaks, allowing for the discrimination of peptides originating from different samples.
- Software and databases are used to extract quantitative information about the peptides originating from all samples.
Affinity proteomics
While a complete overview of all affinity proteomics methods is beyond the scope of this article, we summarise the multiplex proteomics technologies developed by Olink Proteomics, Somalogic and Alamer Biosciences as well as bead-based immunoassays, such as those developed by Luminex.
Proximity extension assay
Olink’s proteomics technology is based upon the principles of the proximity extension assay (PEA). Here, matched pairs of antibodies that carry unique oligonucleotide tags bind to target proteins in a pair-wise manner. When a pair of antibodies binds the same target protein, the matched oligonucleotides on those antibodies are brought into proximity, where they hybridise and form a DNA reporter template for real-time quantitative PCR.
PCR and subsequent data analysis allows each reporter template to be amplified, identified and quantified, providing quantitative data about the abundance of all target proteins in all samples. The requirement for dual antibody binding and hybrdisation of the oligonucleotides before any signal is generated ensures specificity and allows for multiplexing of many samples.
Highly specific protein capture DNA hybridisation probes followed by real-time PCR
Somalogic’s SomaScan technology leverages diverse libraries of (SOMAmer) protein-capture reagents, which are chemically-modified nucleotides with defined 3D structures that both bind to native proteins with high affinity and harbour unique DNA tags that are recognisable by specific DNA hybridisation probes. The technology can measure thousands of native (but not unfolded and denatured) proteins by transforming each individual protein concentration into a corresponding SOMAmer reagent concentration, which is then quantified by real-time quantitative PCR or other quantitative DNA analysis methods.
A major advantage of this technology is its ability to capture previous undetected biomarkers.
Sequential immunocomplex capture and release followed by next-gen sequencing
Alamar’s NULISA technology combines a proprietary sequential immunocomplex capture and release mechanism and next-generation DNA sequencing for sensitive and scalable proteomics workflows. Here, two highly specific antibodies that recognise the same target are conjugated to a pair of oligonucleotides; one antibody is conjugated to a oligonucleotide containing a polyA tail while the other is conjugated to an oligonucleotide that contains a biotinylated tail.
When the antibodies bind their protein target in a sample, an immunocomplex is formed. This is detected and quantified through sequential binding with oligodT beads, which bind the polyA tail originating from one of the oligonucleotides, and streptavidin, which binds the biotin tail on the other oligonucleotide. The result is a hybridised oligonucleotide reporter probe that can be amplified, quantified and identified using next-generation sequencing.
Multiplex bead-bead immunoassays
Where a monoplex immunoassay, e.g., ELISA, uses antibodies to detect and quantify a single analyte per sample, a multiplex immunoassay can simultaneously detect many protein targets in a single well of an assay plate. In a standard sandwich-based ELISA, a target protein is captured between two antibodies and quantified, typically via an antibody-enzyme conjugate that recognises and binds a primary antibody that has already bound the target. The conjugate contains an enzyme that reacts with an exogenously added substrate to produce a detectable signal that is typically measured on a spectrophotometer, fluorometer or luminometer.
Multiplexing is achieved by the addition of populations of magnetic, colour-coded beads that are dyed into spectrally distinct sets, which allows them to be individually identified by a suitable instrument. One prominent example is the MagPlex® bead-based multiplex assay from Luminex. Samples. Here, sets of magnetic MagPlex® beads are fluorescently coded with a unique ratio of two fluorophores, where each spectrally distinct set is known as a bead region. Target analyte-specific capture antibodies are bound to beads of a specific region, then unbound sites of the bead are blocked. The beads are then incubated with the test sample, washed, and incubated with a biotinylated secondary antibody. The unbound secondary antibody is washed away and the added streptavidin-phycoerythrin binds to the biotin of the “sandwich” immunoassay. The Luminex® instrument detects individual beads based on their unique fluorophore ratio (region) region as well as the signal that results from the reaction between biotin and streptavidin-phycoerythrin.
Stay tuned!
That was it for our introduction to multiplex proteomics! Stay tuned for our next article and learn how researchers in Sweden assayed dried blood samples collected via Capitainer cards using quantitative multiplex immunoassays to profile the proteomes of individuals previously infected with the SARS-CoV-2 virus.