Targeted proteomics analysis of plasma proteins using recombinant protein standards for addition only workflows

Quantitative proteomics using mass spectrometry provides sensitive and robust assays when quantifying proteins in complex biological backgrounds [1]. Targeted proteomics typically focuses on predefined sets of peptides measured with high reproducibility across many samples. This approach is particularly suitable for clinical profiling [2–4], and it is therefore an appealing alternative for precision medicine efforts [5,6]. We have earlier reported on the use of isotopic standards, either through metabolic or chemical labeling of the sample or by the addition of stable isotope standard (SIS) peptides [7], proteins [8] or combinations thereof. Multipeptide sequences can account for enzyme kinetics biases, which is introduced between different cleavage sites [9]. This allows for the standards to be digested together with the target protein and the release of prototypic peptides can thus be more tightly controlled. Several variants of this strategy have been described and concatemer [10,11] and stable isotope labeled protein epitope signature tags (PrESTs) [12] are two examples thereof. PrESTs represent a stretch of amino acids unique for the target protein of interest (Figure 1A) and the protein fragment can be isotopically labeled [12] and used for protein quantification by quantitative proteomics [13–15].

To optimize targeted proteomics approaches for highly quantitative analysis, there is a need for a simple procedure adapted for routine clinical assays and precision medicine. Here, we describe a novel approach in which many different isotope-labeled protein fragments are mixed in ratios representing the blood concentrations of the corresponding target proteins. The mixture is dried down into a pellet to decrease the assay volume and to facilitate storage and distribution of these standards in routine settings. We show that this vacuum-dried pellet can be stored at room temperature for at least 1 month and subsequently be used for targeted proteomics analysis in the presence of a chaotropic reagent. The effect of repeated cycling between -20°C and room temperature shows that the recombinant protein standards stored in this salt pellet are surprisingly stable (coefficient of variation <10%) over repeated freeze-–thaw cycles, and for at least 6 months. The targeted proteomics analysis is performed by dissolving the pellet with patient plasma followed by a standard protocol involving enzyme digestion and analysis using mass spectrometry. The concentration of the target proteins in the plasma sample is thus determined by comparing the ratio between peptides originating from the isotope (heavy) standard and the plasma-derived (light) sample. Here, we show that this simple procedure can be used for simultaneous and accurate, multiplex analysis of 100 protein targets in plasma.

Materials & methods

Preparation of samples for in solution freeze–thaw cycle stability

Ninety-six stable isotope standard (SIS)-PrESTs were randomly selected from a library of in-house produced SIS-PrESTs (Supplementary Table 1) and individually aliquoted to a 96-well plate. Subsequently, light protein-tag sequence (PTag)-light, absolutely quantified by amino acid analysis, was diluted to a final concentration of 10 μM in 1x phosphate-buffered saline (PBS), 1 M urea and aliquoted to another 96-well plate. Both plates were stored at -20°C. The two plates were thawed before the quantification once a week for the first 3 weeks, and thereafter once a month for the subsequent four analyses. For every analysis, 5 μl from the thawed plate with SIS-PrESTs were mixed with 5 μl from the thawed plate of the PTag-light and digested (as described subsequently) for quantification by LC-MS/MS. The two plates were then stored at -20°C until subsequent analysis.

Digestion of samples for in solution freeze–thaw cycle stability

Fifty microliters of 100 mM ammonium bicarbonate (ABC) solution was added to the wells of a 96-well plate. Subsequently, 5 μl quantification tag and 5 μl SIS-PrEST were added to each well and the samples were thoroughly mixed before adding 200 ng trypsin to each sample. The plates were incubated at 37°C overnight before the digestion was quenched by the addition of formic acid (FA) to a final concentration of 0.6%. The samples were then directly analyzed by LC-MS/MS operating in SRM mode (LC-SRM).

Quantification of SIS-PrESTs using LC-SRM

A previously established LC-SRM/MS [16] analysis was done using an Ultimate 3000 equipped with a trap cartridge (catalog no. 160438, Thermo Fisher Scientific, MA, USA) and a 15-cm-long analytical EASY-Spray column (catalog no. ES806A, Thermo Fisher Scientific) and a mobile phase made up of solvent A (3% acetonitrile [ACN], 0.1% FA) and solvent B (95% ACN, 0.1% FA). The LC was coupled online to a TSQ Altis (Thermo Fisher Scientific) mass spectrometer. Approximately 0.6 pmol of SIS-PrEST digest and quantification tag was loaded onto the system and the peptides were separated over a 7-min run with a flow rate of 3.0 μl/min. The gradient used for the peptide separation was as follows: 1% solvent B for 0.75 min, 1–30% solvent B in 1.25 min, 95% solvent B for 3.5 min followed by a reequilibration at 1% solvent B for 1.5 min.

The TSQ Altis was operated in SRM mode monitoring the transitions specified in Supplementary Table 2 in standard mode with a cycle time of 0.1 s for the first 5 min of the LC-SRM/MS method.

Data analysis

All raw files were analyzed using Skyline (v. 20.1.0.76) [17] and quantitative data were obtained by calculating the ratio to standard between the heavy labeled SIS-PrEST peptides and their corresponding quantification tag peptides. The total amount of SIS-PrEST present in each sample was thereafter determined by calculating the median molar concentration of four tryptic peptides DLQAQVVESAK, DLQAQVVESAKK, YGVSDYHK and ISEATDGLSDFLK (Supplementary Table 2), previously selected due to their quantitative performance (peak shape and signal-to-noise ratio).

Freeze–thaw cycle stability of SIS-PrESTs for quantification of apolipoproteins in blood plasma

SIS-PrESTs covering seven apolipoprotein sequences were pooled and kept as a single aliquot at -80°C. Subsequently, a pool of deidentified human blood plasma was diluted 20 times using 50 mM ABC, aliquoted into 21 tubes and kept at -80°C. Every month, three tubes of plasma aliquots were thawed on ice together with the tube of pooled SIS-PrESTs and mixed. The tube of remaining SIS-PrESTs was thereafter stored at -80°C. The procedure was repeated six times. During the third preparation, an aliquot of the SIS-PrEST pool was prepared and stored at room temperature for 7 days before addition into freshly thawed plasma.

The plasma-SIS-PrEST mixture was treated in 1% (w/w) sodium deoxycholate (SDC) and 10 mM DTT at 96°C for 10 min; 2-chloroacetamide (CAA) was added to the final concentration of 50 mM and samples incubated at room temperature (RT) in the dark for 20 min. The digestion was performed with LysC (enzyme:substrate ratio 1:100) for 3 h at 37°C and 500 rpm followed by trypsin (enzyme:substrate ratio 1:50) overnight at 37°C and 500 rpm. The digestion was quenched with 0.5% (v/v) trifluoroacetic acid (TFA). Samples were centrifuged at 13,200 rcf for 5 min, and the supernatant desalted on three layer C18 StageTips prepared according to Rappsilber et al. [18]. In brief, StageTips were activated with 50 μl of 100% ACN and equilibrated with 50 μl 0.1% TFA followed by the addition of the digested sample corresponding to 15 μg of proteins in raw plasma. The C18 matrix was washed twice with 0.1% TFA and peptides eluted in two steps with 80% ACN, 0.1% TFA. Eluted peptides were vacuum dried at 42°C and stored in -20°C before mass spectrometry (MS) analysis.

MS analysis was done using a Dionex Ultimate 3000 (Thermo Fisher Scientific) equipped with a trap cartridge (CN: 160438, Thermo Fisher Scientific) and a 15-cm-long analytical EASY-Spray column (catalog no. ES806A, Thermo Fisher Scientific). The LC was coupled online to a TSQ Altis (Thermo Fisher Scientific). Approximately 5 μg of the digested plasma sample was loaded onto the system and the peptides were separated over a 35 min long run with a flow rate of 3.0 μl/min. The gradient used for the peptide separation was as follows: 1% solvent B for 0.75 min, 1–30% solvent B in 29.25 min, 95% solvent B for 3.5 min followed by a reequilibration at 1% solvent B for 1.5 min. The MS was operated in SRM mode monitoring the transitions listed in Supplementary Table 3 in a nonscheduled mode with a cycle time of 0.6 s for the first 35 min of the LC-SRM/MS method. All raw files were analyzed using Skyline (v. 20.1.0.76) [17] and quantitative data were obtained by calculating the ratio between the peptides of the isotopically labeled internal standard proteins and their corresponding target proteins.

Sample preparation for stability assessment of vacuum-dried SIS-PrESTs

A 96-well plate of 96 SIS-PrESTs and a 96-well plate of PTag-light were prepared for standard quantification as described by Edfors et al. [15]. The SIS-PrESTs plate was distributed into eight new plates so every well contained 5 μl (∼50 pmol) of each SIS-PrEST. Five SIS-PrESTs plates were vacuum dried at 42°C, 5 mbar for 3 h. As a first step, three vacuum dried plates were prepared together with aliquoted SIS-PrESTs that were kept as a solution. Prior addition of 5 μl of 10 μM PTag-light protein to every well of each plate, 55 μl of 90.9 mM ABC was added to the vacuum dried SIS-PrESTs and 50 μl of 100 mM ABC was added to the SIS-PrESTs that were kept in solution. Plates were sonicated for 180 s and digested with 200 ng of trypsin overnight at 37°C. Digestion was quenched with FA added to the final concentration of 0.6% (v/v) and samples analyzed using LC-MS/MS operating in PRM mode (LC-PRM). As a second step, the other two vacuum-dried SIS-PrESTs plates were kept at room temperature for 7 and 31 days, respectively, before addition of Tag-light, digestion and LC-PRM analysis as described above.

Quantification of SIS-PrESTs using LC-PRM

The quantification was performed using an online system of Ultimate 3000 LC connected to Q Exactive HF mass spectrometer. 2.5 pmol of each SIS-PrEST was loaded onto an Acclaim PepMap 100 trap column (CN: 164535, Thermo Fisher Scientific), washed 3 min at 8.5 μl/min with solvent A and separated by a PepMap RSLC C18 column (catalog no. ES802A, Thermo Fisher Scientific). A linear 3-min gradient was used for eluting the peptides ranging from 3 to 20% solvent B at 0.6 μl/min. The analytical column was then washed for 3 min at 1 μl/min with 99% solvent B and equilibrated with 3% B at 0.6 μl/min for 4 min. The MS operated in PRM mode with each cycle comprising of one full MS scan performed at 15,000 resolution (AGC target 2e5, mass range 350 to 1600 m/z and injection time 55 ms) followed by 20 PRM MS/MS scans at 15,000 resolution (AGC target 1e6, NCE 27, isolation window 1.5 m/z and injection time 105 ms) defined by a scheduled (0.4 min windows) isolation list. Resulting raw files were loaded into Skyline (v. 20.1.0.76) [17] and ratios between areas under heavy and light curves of peptide DLQAQVVESAK (m/z (light) = 594.31696, m/z (heavy) = 598.32406) exported.

Designing protein recombinant isotope standards

A new set of recombinant protein standards were designated to include approximately 100 amino acids from the target protein sequence, not including the quantification sequence present in each recombinant protein standard. A list of experimentally verified proteotypic peptides was used as a template for the redesign and formed the foundation when selecting amino acid sequences for a new set of standards called protein recombinant isotope standards (PRecIS). This list consists of data acquired by the screening of in-house generated full-length secreted protein sequences produced in CHO-cells [19], analysis of human plasma samples and basic reverse phase fractionation of plasma pools. The final list of peptides was supplemented with a list of peptide identifications from proteins with experimental evidence in blood (943 genes). These assay coordinates were obtained from public resources [20] including the SRM atlas repository [21] and the CPTAC assay library [22]. The sequence stretch including most tryptic peptides identified by MS were selected and transferred into an expression vector and produced as previously described. These standards are fundamentally different from the PrEST sequences, which are optimized for antibody responses when generating polyclonal antibodies.

Plasma digestion with vacuum dried PRecIS

A pool of 100 PRecIS was aliquoted and vacuum dried for 3 h at 42°C, 5 mbar. SDC was diluted so after addition into vacuum dried PRecIS the final concentrations were 1% SDC, 1 M urea, 1x PBS. A pool of plasma was diluted 10 times with 1x PBS and amount corresponding to 0.5 μl of raw plasma and added into the PRecIS pool. Samples were treated in 10 mM DTT at 37°C for 1 h and 50 mM CAA for 30 min at RT in the dark. SDC was diluted to a final concentration of 0.25% (w/w) with 1x PBS before addition of trypsin (enzyme:substrate ratio 1:50). Digestion was performed at 37°C and quenched with 0.5% (v/v) TFA after 1, 2, 3, 4 and 16 h. Samples were centrifuged and supernatants desalted as described earlier. Desalted samples were dissolved in solvent A and amount corresponding to 4 μg of raw plasma subjected to LC-MS/MS analysis operation in DIA mode (LC-DIA).

LC-DIA quantification of plasma proteins using vacuum dried PRecIS

The LC-DIA analysis was performed using an online system of Ultimate 3000 LC connected to Q Exactive HF mass spectrometer. First, the amount corresponding to 4 μg of raw plasma was loaded onto a trap column (catalog no. 160438, Thermo Fisher Scientific) and washed for 1 min at 15 μl/min with solvent A. Peptides were then separated by a 15-cm analytical column (CN: ES806A, Thermo Fisher Scientific). A linear 50-min gradient was used for eluting the peptides ranging from 1 to 32% solvent B at 3.6 μl/min. The analytical column was washed with 99% B for 30 s followed by two seesaw gradients from 1% to 99% solvent B. Column was then equilibrated for 1 min with 1% solvent B. The MS operated in DIA mode with each cycle comprising of one full MS scan performed at 60,000 resolution (AGC target 3e6, mass range 300–1200 m/z and injection time 105 ms) followed by 30 DIA MS/MS scans at 30,000 resolution (AGC target 1e6, NCE 26, isolation window 12 m/z, injection time 55 ms).

Peptide identification

The resulting raw files were loaded into Skyline (v. 20.1.0.76) [17] file containing a spectral library. The library was generated using a deep learning network termed Prosit, which is integrated into ProteomicsDB [23]. As an input to Prosit, sequences from a set of 2000 in silico digested proteins that were previously detected in blood plasma by LC-MS/MS was used [20]. The resulting MSP file was converted into a BLIB file using tools built in EncyclopeDIA [24] and used as a library in Skyline. A whole human proteome (Homo Sapiens UniProt ID: #UP000009606, 20,394 entries, accessed 2018-06-22) was used as a background proteome to match the peptide sequences from library to protein IDs and to control for peptide uniqueness between proteins. The peaks were manually inspected and ratios between areas under heavy and light curves exported. The final list of quantified proteins and peptides can be seen in Supplementary Table 4.

Investigation of possible deamidation

A subset of 10 proteins that were the most abundant in plasma in the digestion course experiment was selected and their sequences imported into Skyline (v.20.2.0.286) [17]. Only peptides containing amino acids asparagine (N) or glutamine (Q) were kept. These peptides were duplicated and all the duplicates set to be modified with deamidation. The raw files acquired from the digestion course experiment were imported and the extracted chromatograms manually inspected.

Results & discussion

Storage conditions of internal standards

The quantitative performance of SIS-PrESTs exposed to repeated freeze–thaw cycles was investigated to evaluate their overall stability and the effect on quantitative accuracy and precision. A set of 96 SIS-PrEST standards (Supplementary Table 1) was individually aliquoted into a microtiter plate, frozen and stored at -80°C (Figure 1B). The selection criteria of the recombinant standards was to harbor at least one peptide with a molecular mass suitable for mass spectrometry shown to be detectable in our earlier analysis of the recombinant protein fragments [9]. The PTag used for quantification of each SIS-PrEST was also prepared and stored at -80°C. Both the SIS-PrEST and the PTag were repeatedly thawed and frozen again for a total of seven cycles. An equal amount of SIS-PrEST and PTag was spiked together at each thaw cycle and thereafter digested by trypsin the same day. Trypsin digests were later analyzed by LC-MS/MS and quantified using the ratio between heavy and light peptides. Technical reproducibility of the complete analysis workflow was performed during the fourth thawing cycle (week 3) as three replicates were prepared of each recombinant protein standard. A median CV of 4.8% was observed across the set of 96 SIS-PrEST replicate digestions (Figure 1D). Additionally, one of the plates was subjected to triplicate LC-MS/MS injections to assess the technical reproducibility of the mass spectrometer itself, which resulted in a median CV of 1.4% across repeated injections of 96 digestions.

The quantitative precision and accuracy for 94 SIS-PrESTs after normalization of the tag sequence can be seen in Figure 1E (and details in Supplementary Table 1). Four clusters of peptides were identified by hierarchical clustering based on the quantitative output across all seven freeze–thaw cycles. The largest group, consisting of 82 SIS-PrESTs (purple), shows stable profiles across all seven-time points. The second cluster consists of seven SIS-PrESTs (blue) shows some minor affect upon freeze-thawing, particularly after repeated cycles. The third (n = 4, yellow) and fourth (n = 1, green) clusters contain SIS-PrEST sequences that are heavily affected by repeated freeze-thawing. The five SIS-PrESTs in clusters 3 and 4 were therefore considered not suitable for the targeted proteomics protocol described here, while the majority of the SIS-PrESTs analyzed (89 out of 94) were considered suitable for MS-based quantification. Of note, the recombinant protein that showed the highest loss after repeated freeze-thawing was the PTag itself, which showed consistent decay over all replicates (Supplementary Figure 1).

Quantitative performance during repeated freeze–thaw cycles over 6 months

The schematic workflow outlined in Figure 2A was performed to address the stability of the recombinant protein standards during repeated freeze–thaw cycles independently of the PTag-sequence. A set of seven SIS-PrESTs directed towards a set of apolipoproteins was selected and mixed in ratios to represent the endogenous levels of their corresponding proteins present in a standardized pool of human plasma (Supplementary Table 2). This ensures that the standard is spiked into the same approximate level as the endogenous protein, thereby ensuring nonbiased comparison of intensities and limiting technical biases from large off-ratios. The SIS-PrEST mixture was frozen and stored at -80°C together with a number of plasma pool replicates that had been aliquoted into individual tubes. The SIS-PrEST tube was repeatedly thawed and frozen on a monthly basis and three plasma tubes were thawed alongside the SIS-PrEST tube. Once thawed, the SIS-PrEST standard mixture was spiked into 1 μl of plasma in triplicates and digested. The SIS-PrEST mixture was repeatedly thawed and frozen over 6 months. The plasma tubes were only thawed once and then disposed of after each experiment to ensure that all plasma samples in this experiment were synchronized and had been subjected to the same number of freeze–thaw cycles.

The result from the LC-SRM/MS assay is shown in Figure 2B. All PrEST peptides show stable quantitative performance over time even though they had been subjected to up to six freeze–thaw cycles. In a separate experiment, the same SIS-PrEST mixture was aliquoted into a separate tube and stored at room temperature (22°C) for 1 week before thawing triplicate plasma tubes and spiking them with the SIS-PrESTs. The results showed that the quantitative accuracy is lost and suggest an aggregation or degradation of the standards over time if stored in suspension at room temperature.

Quantitative performance of the vacuum-dried formulation

We next decided to investigate the possibility to vacuum-dry the protein standards containing urea to enable a simple addition-only protocol with a salt pellet containing the mixture of SIS-PrESTs for multiplex analysis. Many proteomics sample preparation protocols include buffers that contain urea due to its ability to break up the tertiary structure of proteins by destabilizing internal, non-covalent bonds. Vacuum centrifugation or lyophilization is an attractive route for reducing sample volume prior to downstream workflow and it can enable larger numbers of SIS-PrESTs to be pooled into one single vial before spiking them into the sample of interest. We first decided to study the quantitative accuracy of the targeted proteomics assays based on adding plasma samples to the dried down standards (Figure 3A). Ninety-six SIS-PrESTs were aliquoted into eight replicate-plates using robotic pipetting systems and five of the plates were vacuum dried for 3 h and stored at room temperature. Three replicates of the in-solution plate and three replicates of the vacuum-dried plate were subjected to the same sample processing as an aliquot of PTag was added following trypsin digestion and LC-MS/MS analysis. Each SIS-PrEST was quantified using the PTag and the results are shown in Figure 3B demonstrating no major difference between the two approaches. Interestingly, plates stored at room temperature over extended periods of time in the vacuum dried format (7 days and 4 weeks) showed low variation, as shown in Figure 3C. Moreover, the quantification data points of vacuum-dried SIS-PrESTs that were stored at room temperature for 1 week and 1 month lay within the standard deviation of the triplicate quantification of SIS-PrESTs that were quantified right after they were vacuum-dried (Figure 3D). This level of accuracy measured repeatedly over 4 weeks confirms the robustness of the quantification even after extended storage of vacuum-dried SIS-PrESTs at room temperature.

The downside of using urea includes the increased risk of chemically modifying the protein target of interest in the process prior to any addition of internal standards. Here, we address this by adding the protein standards already before denaturation and the study presented here demonstrates that the level of deamidation of the analyzed peptides is undetectable by the DIA assay under the current sample preparation protocol. Thus, the pre-aliquotation of standards into the final sample well allows for such a robust assay, in which the biological sample of interest is spiked directly to the pre-mixed standards. In summary, the approach involving the addition of plasma to the salt pellet provides a streamlined protocol with performance suitable for quantitative studies (inter-day CV <10%). The approach allows for short digestion protocols, which might be an advantage for clinical routine applications.

Improved proteotypic peptide repertoire

The original recombinant protein standards were not designed for targeted proteomics, but produced to generate antibodies towards the recombinant target [25]. We therefore decided to re-design some of the PrEST sequences with regards to the inclusion of suitable proteotypic peptides and bottom-up proteomics workflows. Therefore, a new iteration of recombinant protein standards was designed to improve the quantitative performance. A total of 100 PrEST standards were designed and produced as SIS-PrESTs and subsequently used to quantify the corresponding target proteins in a digestion course experiment. Replicate samples were prepared using the vacuum dried approach and trypsin digestion was quenched at different time points (1, 2, 3, 4 and 16 h) (Figure 4A). Four sets of peptide groups can be identified (Figure 4B). The two main clusters 2 and 4 show similar digestion kinetics as the endogenous human protein, which is in line with previous work [9]. Cluster 1 shows that there is higher efficiency in the digestion of the internal standard protein than of the endogenous protein for its few members and on the other hand for the few members of cluster 3 the efficiency of digestion is higher for the endogenous protein than of the isotopically labeled protein standard during the time course. Most importantly, regardless of where the quantified peptide clusters and regardless of the digestion time, the precision of quantification remains stable and high for all clusters, with median CVs ranging between 4.6% and 6.1% (Figure 4C). This allows for short digestion times and rapid sample preparation protocols with great precision in quantification. The set of 100 target proteins quantified was ranging four orders of magnitude in concentration in human plasma (Figure 4D). Eighty-five of these proteins are classified as secreted by the Human Protein Atlas [19] and 33 proteins are reported as FDA-approved biomarkers [26]. This quantification strategy using internal standard dilution schemes performed by mass spectrometry are attractive for quantitative measurements of the clinically relevant proteome, especially for many FDA-approved protein targets. The proposed workflow allows for an addition-only design where these standards are introduced directly to any sample of interest upfront enzymatic digestion and downstream processing. This results in a robust assay format with a technical CV <10%, which is suitable for quantitative exploration of a potentially large fraction of the human proteome. The assay volume is kept to a minimum thanks to the vacuum centrifugation step, reducing the sample volume and furthermore allowing the standards to be stored for extensive periods of time without introducing quantitation biases. This is crucial to consider when working with protein standards, which are much more prone do aggregate, precipitate or affect the quantification if stored under improper conditions as compared to AQUA peptides. Thus, the protocol describe here is suitable for clinical applications in which a stream-lined protocol suitable for automation is combined with high accuracy and low technical variability.

Conclusion

Both accuracy and precision are highly relevant for larger studies as all SIS peptides or proteins are pre-quantified prior to any experiment. Stability is an even more important issue for recombinant proteins, due to the inherent instability caused by proteolysis. The presented “reverse-addition” of samples directly into a mixture of SIS recombinant proteins is shown to be robust as multiple plates can be prepared in bulk and stored for months. The results of this study show that vacuum dried isotope-labeled recombinant protein standards stored in urea can be used for precise quantification using an assay with short standardized digestion protocols. We also present a workflow concept for targeted proteomics suitable for automated analysis of clinical samples. The addition of recombinant standards prior to the enzymatic digestion allows for differences in digestion kinetics of individual peptides, thus making the procedure more robust for simultaneous analysis involving hundreds of protein targets. We show that the protocol can be used for multiplex analysis of hundred plasma proteins spanning four orders of magnitude in target protein concentrations, but the target number can be expanded to encompass many hundreds of targets of interest for clinical studies.

Future perspective

This new analytical platform described is based on isotope-labeled recombinant standards stored in a chaotropic agent and subsequently dried down to a pellet allowing storage for extended time periods at ambient temperature. This enables a straightforward protocol suitable for robotic workstations in which plasma sample to be analyzed is simply added to the dried pellet followed by enzymatic treatment and MS analysis. This can unlock robust LC-MS/MS-based profiling of sample cohorts in a future clinical setting, as LC-MS/MS vendors move towards in vitro diagnostic (IVD) certification of their equipment. The absolute concentration of the target proteins can subsequently be determined using the ratio of heavy (standard) and light (target) peptides and can potentially replace traditional serology using only one microliter of sample material. The re-design of the protein standard repertoire will help improve the protein coverage and introduce new potential not covered by the original PrEST repertoire. The targeted proteomics protocol presented here is therefore suitable for precision medicine efforts where protein quantification must be reproducible over long time periods.