19-X5
Immunoassays
or
LC-MS/MS?
A
Comparison
Revealing
the
Properties
of
Modern
Methods
for
Insulin,
Pro-insulin,
C-peptide
and
Glucagon
Quantification
Anna
Ivert
Nordén,
Temis
Martinez,
Casper
Tiger,
Viktor
Umenius,
Ruta
Upite,
Susanna
Wärmegård
Beställare:
Mercodia
AB
Beställarrepresentant:
Johan
Ledin
Handledare:
Lena
Henriksson
1MB332,
Självständigt
arbete
i
molekylär
bioteknik,
15
hp,
vt
2019
Civilingenjörsprogrammet
i
molekylär
bioteknik
Abstract
The purpose of this report is to compare seven different methods for biomarker detection and quantification based on previously published papers. The methods investigated are ELISA, LC-MS/MS, UPLC-MS/MS, LC-IM/MS, IA-LC-MS/MS, MSIA-HR/AM, HTRF and AlphaLISA® .
The focus lies on biomarkers relevant for diabetes, obesity and cardiovascular diseases. Namely insulin, proinsulin, glucagon and C-peptide. Particular significance is assigned to the comparison of the currently widest used method, ELISA, with various types of
LC-MS/MS.
The report concludes ELISA being superior to LC-MS/MS methods in terms of recovery and precision, while LC-MS/MS is superior in accuracy, multiplexing, specificity,
throughput and sample cost. This suggests that different types of LC-MS/MS has the potential to gain momentum in the field of biomarker quantification if they become more available.
Table of Contents
1. Vocabulary and Abbreviations 4
2. Background 6
2.1 Rationale for this project 6
2.2 Detection today 6
3. Workflow 8
4. Technology Descriptions 9
4.1 ELISA 9
4.2 Liquid Chromatography and Mass Spectroscopy Methods 10
4.2.1 LC-MS/MS 10
4.2.2 UPLC-MS/MS 12
4.2.3 LC-IM-MS 12
4.2.4 IA-LC-MS/MS 13
4.2.5 MSIA-HR-AM 15
4.3 Other Methods for Detection and Quantification 16
4.3.1 HTRF 16
4.3.2 AlphaLISA® 17
5. Explanation of parameters upon comparison 19
6. Summarised comparison of the methods 21
6.1 Insulin 21
6.2 Proinsulin 22
6.3 C-peptide 23
6.4 Glucagon 24
7. Remarks regarding comparison of the methods 25
7.1 Comparison Between ELISA and LC-MS Methods 25
7.1.1 Accuracy for LC-MS/MS 25
7.1.2 Automatisation and multiplexing 26 7.1.3 Sample preparation with respect to complex matrices 26
7.1.4 UPLC-MS/MS 27
7.1.5 LC-IM-MS 27
7.1.6 Availability of data about the biomarkers among the methods 27 7.2 Comparison Between ELISA and other immunochemical methods 28
7.2.1 AlphaLISA® 28
7.2.3 Hook effect in immunochemical methods 29
7.3 Economics 29
8. Conclusion 30
9. Ethical Discussion 31
9.1 Production of Antibodies 31
9.2 Diagnostics or Research Only 32
9.3 New Standard Technology for Detection and Quantification 32
9.4 Production and Transport 33
10. Acknowledgements 34 11. Statement of Contribution 34 12. References 36 12.1 Articles 36 12.2 Online sources 39 Appendix 1 41
Not Selected Technologies 41
1. Vocabulary and Abbreviations
2D-method: A method that uses two different liquid chromatography columns to separate the sample before being analyzed by two mass spectrometers in series.
Biomarker: A measurable biological indicator. In this report, biomarker is referred to as the molecules insulin, proinsulin, C-peptide and glucagon.
Cross-reactions: In this report - cross-reactivity of an antibody in the context of
immunology. Reliant on an antibody’s antigen-binding site several proteins/peptides might get bound if the target molecule shares similarity meaning that the same antibody will detect several molecules. Depending on the purpose of immuno-recognition this might lead to false positives upon detecting a specific target molecule.
CV: Coefficient of variation upon cross validation. Performed with analytes in different concentrations. Smaller variation between the used data sets indicate greater precision. Endogenous peptide: Peptide that is produced in the sampled organism.
Exogenous peptide: Peptide added to a sample/sampled organism.
FRET: Fluorescence Resonance Energy Transfer, upon close contact (<10nm) two molecules can transfer energy to one another without using radiation.
Hyphenated techniques: Hyphenation refers to coupling different techniques. Hyphenated methods in this report are combinations of different liquid chromatography and mass spectroscopic techniques as well as immunoaffinity based separation techniques.
Immunochemical methods: Methods using antibodies for the purpose of detecting specific proteins/peptides. ELISA and RIA are examples of immunochemical methods.
Immunosorbent: An insoluble support for an antigen or antibodies used to adsorb
antibodies or antigen from a mixture. The adsorbed material can then be rinsed out in pure form for analysis. Many different substances are used, including sepharose, glutaraldehyde, copolymers of anhydrides, polyacrylamides and others.
Matrix-effects: Matrix depending signals. Non-target molecules can affect quantification upon measurements. Interferences may enhance or suppress signals.
Multiplexing: Detecting multiple peptides simultaneously. In biomarker detection it is of interest to detect isoforms arising from, for example, posttranslational modifications as well as relevant exogenous analogs of the same peptide.
PTM: Posttranslational modifications
Recovery: A measure of accuracy, quantifying agreement between a known reference value (standard) and the found value.
RIA: Radioimmunoassay, quantifies analyte by detecting radioactivity of radiolabeled antigens that are released from a complex as the interesting antigen outcompetes the radioactive ones from an antibody-antigen complex.
Signal-to-noise ratio: Describes the ratio between the wanted signal and the unwanted background noise from the sample.
Targeted Mass Spectrometry (MRM/SRM): Multiple reaction monitoring (MRM, also known as Selective Reaction Monitoring – SRM) is a highly specific and sensitive mass spectrometry technique. It uses triple quadrupole MS where the molecule is fragmented in several steps and the target ions as well as their daughter ions are tracked. The technique is especially suitable for detection of bigger peptides with mass range outside of the measurement range of simpler MS instruments.
Top down and bottom up analysis: Top down analysis uses the whole protein in the MS/MS analysis. Bottom up analysis breaks down the compound into smaller components prior to the MS/MS analysis. Bottom up is more complex but more suitable for bigger molecules.
2. Background
In this report we aimed to explore available methods for biomarker quantification and detection aside from the currently most widely used method, Enzyme-linked
ImmunoSorbent Assay (ELISA) . In particular, various liquid chromatography tandem mass spectrometry (LC-MS/MS) methods were of interest. Other antibody utilising methods such as Homogeneous Time Resolved Fluorescence (HTRF) and AlphaLISA® are reviewed as
well. LC-MS/MS methods are of interest as they have seen an upswing in popularity during the last decade as they offer improvements of quantification that traditional methods like ELISA do not support.
2.1 Rationale for this project
This study was requested by Mercodia AB in order to gain a summarising view of what types of methods for biodetection that currently exist. The company develops ELISA kits for complex matrices such as blood serum and plasma. Therefore, the articles analysed in this study concerns primarily these matrices. Following biomarkers were researched: insulin, proinsulin, C-peptide and glucagon, as these are the most important biomarkers that Merodia AB develops kits for and are therefore of greatest interest.
Identifying and quantifying biomarkers is important for tracing and detecting a wide variety of diseases. Thus biomarker detection is a foundation in many areas of research in order to understand the biological processes causing disorders in living organisms.
Diabetes, obesity and cardiovascular diseases can for example be linked to the biomarkers we aim to look at in this study. Societal costs for these diseases, due to their effect on the quality of life, contributes to demand for reliable biomarker detection. The interest to detect biomarkers is also of great importance in other fields of research, such as common animal studies. Biomarkers in this study have multifaceted functions in organisms,
understanding of their function may lead to increased knowledge in other areas beyond the mentioned diseases.
2.2 Detection today
of more than 20 SEK (Sakamoto et al. 2018). Moreover, there is a risk for interference from other molecules, such as cross reactivity with similar compounds to the target antigen, which results in false positives. Also, ELISA has a relatively long total runtime, around 2-4 hours, with many steps involved during the procedure (Farino et al. 2016). Lastly, ELISA requires tailored and specific antibodies for every biomarker of interest.
These are the reasons for exploring potential new methods for biodetection. Furthermore, it is shown that consensus for concentrations among different ways of measuring
biomarker levels is not reached among methods used today. MS-based methods have been suggested to offer more coinciding results due to streamlined methodology that minimizes manual labour. However, many of the methods studied in this report solve some or most of these issues, but as we will see, they do come with some disadvantages of their own.
3. Workflow
A literary study was conducted to gain background knowledge about different ways of quantifying the biomarkers of interest from serum and plasma. Information about several technologies was gathered and stored. The different biomarkers had almost an equal number of articles, but insulin is overall the most studied. Comparative articles between different technologies and ELISA were collected as well.
Later, a list was made with all the twelve technologies found we deemed interesting: LC-MS/MS, IA-LC-MS/MS, LC-IM/MS, UPLC-MS/MS, HTRF, HRMS, MSIA-HR/AM, AlphaLISA® , ECLIA, CLIA, IRMA and IECMA. These were assigned among the group
members and each person investigated how two to three of these technologies functioned. Shortly thereafter, another round of method selection was carried out based on the
individual research.
After the initial investigation, we chose to research further about the following methods, all which will be described in the next section of the report: ELISA, LC-MS/MS,
UPLC-MS/MS, IA-LC-MS/MS, LC-IM/MS, MSIA-HR/AM, HTRF and AlphaLISA® . The
choice was based on the amount of information that was found and our judgement of the potential of the technologies in comparison to ELISA. Each member then searched for information regarding two of these technologies, in articles where they were applied on the four biomarkers of interest for this report. The methods that were not chosen from the initial investigation, namely ECLIA, CLIA, IRMA and IECMA, are shortly described in
Appendix 1 to show why these technologies were not selected.
Finally, summarising tables 1.1, 1.2, 2.1, 2.2, 3.1, 4.1, 4.2, 4.3, 4.4, 5.1, 6.1, 6.2, 6.3, 7.1, 7.2 and 7.3 were compiled to the result tables 1, 2, 3 and 4 for an easier overview.
4. Technology Descriptions
Below follows a description of how the seven chosen methods work. It also includes an explanation of ELISA. Their estimated prices are included. To simplify the comparison between technologies all the prices, as cost per sample and cost of instrumentation, were converted to SEK. The publication date of the article was taken in consideration for the conversion ratios that year.
4.1 ELISA
Enzyme linked immunosorbent assay (ELISA) is an immunochemical method, and the currently most widespread method for biomarker detection. There are many variations of ELISA, namely indirect ELISA, direct ELISA, sandwich ELISA and competitive ELISA. They all differ slightly from one another but share the core mechanics. Looking closer on one method, Sandwich ELISA (Figure 1), which works by connecting an overabundance of antibodies specific to the biomarker of interest to a chemically inert plate. The sample is then washed over the plate, resulting in only the biomarkers of interest binding to the antibodies and the rest being washed away. Then a second antibody with a connected enzyme is washed over the plate, binding only where antigen has already been bound with the first sets of antibodies. This ensures specificity. By adding a substrate catalysed by the enzyme, a detectable product that is directly proportional to the amount of bound antigen will be produced. Measuring these will give an estimation of the concentration of the target biomarker within the sample. (Sakamoto et al. 2018)
In contrast to sandwich ELISA, direct and indirect ELISA let everything in the sample first bind to the plate and then wash with antibodies for detection. Competitive ELISA uses different types of antibodies that competes for the antigen and then the concentration is calculated through the ratio of catalysed substrate compared to the number of antibodies added of each type. The typical cost per sample for ELISA is 20-50 SEK and no expensive machines are required other than for detection of the coloured product from the enzyme reaction. This equipment is standard lab equipment however and will not be a problem for most research teams. A typical ELISA kit might cost around 5000 SEK for an assay covering around 100 wells (50 SEK per sample, Thermo Fisher ScientificTM, 2019). (Sakamoto et al.
Figure 1. Description of Sandwich ELISA. 1) Sample is washed over plate with antibodies 2) Antigen of interest binds 3) A second wave of antibodies with a linked enzyme is washed over the plate 4) The second antibody binds to the antigen 5) Substrate is added which are catalysed by the enzyme into detectable product. Figure created with BioRender.com.
4.2 Liquid Chromatography and Mass Spectroscopy Methods
4.2.1 LC-MS/MS
LC-MS/MS is a liquid chromatography separation combined with tandem mass
spectrometry. The sample is first separated through a liquid chromatography as shown in
mass analyser, in a similar way as in the first mass analyser, and finally detected (see Figure
3). (Pitt 2009)
When analysing a complex matrix, one has to take into account the matrix effects. These are caused by all components in the sample that are not quantified and can affect the reproducibility of the experiment. Therefore complex samples often need a pre-treatment before they are injected into the LC-MS/MS machine. (Smeraglia et al. 2002)
Depending on the type of LC machine used, the price can range from 10 000- 400 000 SEK (Liquid Chromatographs (HPLC), Conquer Scientific. 2019). The same applies to the MS/MS machines, where second hand machines can range in price from 500 000-2 500 000 SEK (LabX. 2019).
Figure 2. LC workflow. 1) The mobile phase is pumped into the column. 2) The sample is added. 3) The sample is separated based on the affinity of the column. 4) The sample is detected. 5) Sample is analysed in a computer. 6) Output from the separated sample for further investigation. Figure created with BioRender.com.
Figure 3. MS/MS workflow 1) The inserted sample is ionised. 2) The ionised sample is separated in the first mass analyser. 3) The separated sample is fragmented 4) The fragmented ions are separated in the second mass analyser 5) Detection of the wanted compounds. Figure created with BioRender.com.
4.2.2 UPLC-MS/MS
UPLC-MS/MS stands for Ultra Performance/Pressure Liquid Chromatography – tandem mass spectrometry and is also known under the name UHPLC-MS/MS where the H stands for “high”. It is a method for separation of components that combines qualitative liquid chromatography with mass spectrometry. This liquid chromatography is better than the ordinary because of the special columns that are used. These columns have a particle size diameter of less than two micrometers which makes the system more efficient. In addition, the eluent is driven through the column at a higher pressure than with a HPLC. Altogether the properties of UPLC contributes to a time reduction. (Guillarme & Veuthey 2017)
The system needs a mobile phase, a pumping system, an autosampler, a special column, detectors in series and a data system (Chromatography & Mass Spectrometry Solutions). Almost everything in the system can be controlled by the data system. This type of
instrumentation could cost 2 000 000 SEK (Cross & Hornshaw 2016). However, according to the article by Howard et al. (2017) the price will be ameliorated after a couple of years due to the systems multiplexing ability.
4.2.3 LC-IM-MS
Liquid chromatography ion mobility mass spectrometry (LC-IM-MS or LC-IMS-MS) utilises three coupled separations to create one powerful method which separates molecules with regard to both mobility and mass/charge. The liquid chromatography and mass
All the instruments needed in this method have high prices, including the IMS which cost around 350 000 SEK (CBRNE Tech Index 2018). The most time-consuming step in this method is LC while IMS only takes a few seconds. In recent years, IMS has seen an increased use in the detection of biomolecules, such as peptides. As can be seen in an article by Lapthorn et al. (2013) the peak capacity for LC-IM-MS is 100 times higher than LC-MS and greater by a factor of 25 000 compared to UHPLC. Peak capacity is a
measurement of how good the separation is by calculating a number of maximum number of peaks that fit in a 2D method (Lapthorn et al. 2013).
4.2.4 IA-LC-MS/MS
Immunoaffinity liquid chromatography tandem mass spectrometry (IA-LC-MS/MS) uses antibodies to isolate and capture analytes of interest in a complex matrix and then runs a LC-MS/MS for analysis. The immunoaffinity can be done in two main ways, either by using an offline approach or by using an online approach. The online-column based method usually requires less sample preparations, since most steps are combined in a single online method (Dufield & Radabaugh 2012). For this, one uses affinity columns where the
chromatography media is covalently coupled with antibodies, as shown in Figure 4. The online immunoaffinity is directly coupled to an analytical technique for analysis. With a relatively high concentration of antibodies, this approach gives a rapid enrichment of the sample and can be regenerated and used for several runs. Disadvantages with using this method is that there is a risk for sample transfer and it has difficulties running large sample volumes.
The offline based approach of enrichment uses either magnetic particles coupled with antibodies (see Figure 5), beads coupled with antibodies or immobilised antibodies in pipet tips. The advantages with the offline method is that samples can be processed in parallel by using larger sample volumes. (Whiteaker & Paulovich 2011)
The price for this method, excluding the price for LC and MS equipment, depends on the immunoaffinity approach. A bead-based approach using Dynabeads™ (Thermo Fisher
ScientificTM, 2019) costs between 24 000 to 33 000 SEK for 10 mL. Using the columns
Figure 4. Online column based immunoaffinity which is directly coupled with an analytical technique. 1) Sample is added to an antibody coated column. 2) After rinsing, everything but the antigen bound to the antibodies will leave the column. 3) Only antigens remain from the sample, they are then eluated out and introduced into an
LC-MS/MS system. Figure created with BioRender.com.
Figure 5. Offline bead based immunoaffinity. 1) Fill tube with magnetic beads. 2) Added antibodies will bind to the beads. 3) After addition of sample, antigen will bind to the antibodies. 4) Using a magnet to hold the beads, everything else is removed. 5) Eluate out beads with antigen and wash off antigen, then transfer into an
4.2.5 MSIA-HR-AM
Mass spectrometric immunoassay - high resolution - accurate mass (MSIA-HR-AM) applies immunoaffinity based enrichment of biomarkers with high resolution LC-MS/MS for quantification of the peptides of interest. Use of MSIA microcolumns allows streamlined, multiplexed screening for biomarkers as the workflow is partially automatised. The monolithic microcolumns used for MSIA are activated with an antibody and fixed in a pipette tip. Immunoaffinity retrieval is obtained with 100-250 cycles of aspiration and dispensation on a pipettor (see Figure 6) (Oran et al. 2011, Peterman et al. 2014, Nedelkov et
al. 2018).
Enrichment with help of a polyclonal antibody serves as a precondition for further analysis in MS. Antibodies are polyclonal with respect to isoforms and PTMs as several forms of analysed peptides are recognized. The elute from the MSIA step is loaded in a LC-MS/MS system. To gain high-resolution data that fulfills HR-AM standards, a spectrometer with sufficient resolution is required. Data analysis is then performed using software such as Pinpoint™ Software (Thermo Fisher ScientificTM) in order to quantify biomarkers as well as
to perform a qualitative analysis of the obtained peaks (Peterman et al. 2014).
A MSIA step combined with LC-MS/MS offers a top down analysis. MSIA can be
hyphenated with selected reaction monitoring (SRM) - a bottom up approach that obtains an even higher resolution.
Throughput of the method is highly dependent on the pipettor used. Thermo Fisher ScientificTM offers the MSIA tips and several products applicable to MSIA protocols -
multichannel pipette, a liquid handler system and a fully automated system. Studies available and used in our comparison uses a multichannel pipette or a 96-channel liquid handling system. This system (VersetteTM system for MSIA application, Thermo Fisher
ScientificTM, 2019) costs 400 000 SEK. The custom MSIA D.A.R.T.’STM (Disposable
Figure 6. Immunoenrichment steps in MSIA tips. 1) Pipette tips are connected to the pipettor. 2) Closer look at the pipette tip. 3) The sample is filtered by pipetting the sample up and down multiple times through the antibody filled column within the pipette tip, antigens bind to the antibodies. 4) The antigens are eluted. Figure created with BioRender.com.
4.3 Other Methods for Detection and Quantification
4.3.1 HTRF
Homogeneous Time Resolved Fluorescence (HTRF) (Figure 7) is a method that utilises Fluorescence Resonance Energy Transfer (FRET) between two molecules attached to two separate antibodies. The first molecule is called the “donor” and this excites the second molecule, the “acceptor”, via FRET. When the antibodies are not connected to the antigen, the molecules attached to the antibodies are too far apart (>10nm) for FRET to occur. When both antibodies bind to the antigen on two different sites the distance is shortened so FRET can occur. When administering energy, the wavelength of both the excited donor and acceptor will be detected if FRET occurred, and only the donors wavelength if FRET did not occur. Since antibodies are added in excess, the amount of fluorescence from the acceptor is directly linked to the number of antigens in the sample.
Figure 7. 1) No FRET may occur because of distance 2) Distance is shortened after antibodies bind to antigen 3) FRET occurs, donor gives energy to acceptor, acceptor gives out light of a specific wavelength that is measured.
Figure created with BioRender.com.
4.3.2 AlphaLISA
®AlphaLISA® is a commercial method branded by PerkinElmer, that is claimed to be an
improved version of ELISA when working with difficult matrices or running several assays of the same sample (PerkinElmer. 2019. Alpha Reagents). As shown in Figure 8, it is a sandwich assay with two antibodies coupled to two separate beads. A Biotinylated Anti-analyte Antibody (antibody with a biotin bound to it) first binds to the analyte of interest and then an Anti-Analyte Antibody with an AlphaLISA® Acceptor Bead binds to the
analyte. A Streptavidin-coated Alpha Donor Bead binds to the Biotinylated Anti-analyte
Antibody and form a complex. This complex is transported to a laser reader, which excite the donor bead, causing it to release an oxygen molecule to the acceptor bead. As a result, the acceptor bead emits a sharp light peak at 615 nm which can be detected.
(PerkinElmer. 2011. TDS – Human Insulin.)
To implement an AlphaLISA® assay the kit itself as well as a plate reader with laser
excitation are needed. One plate reader branded by Perkin Elmer is EnVision Multimode Plate Reader. A refurbished instrument costs up to 170 000 SEK and new ones have significantly higher prices, giving this method a high starting cost. Apart from that, it is useful to have dark plates, something to cover the plates with and a dimmed room since the Donor beads are light-sensitive (PerkinElmer. 2011. TDS – Human Insulin.). AlphaLISA®
Figure 8. Schematic figure over the AlphaLISA® reaction. An oxygen molecule is released from the Streptavidin on
the donor bead when it is excited by the 680 nm wave emissions. This excites the molecule connected to the acceptor bead and this emits a 615 nm wave which is measured. This may only occur if they are in close proximity, prompted by the antibodies binding to the antigen. Figure created with BioRender.com.
5. Explanation of parameters upon comparison
To compare the different methods, fourteen parameters were considered. The parameters sample volume, sensitivity, measure range, throughput, sample cost, multiplexing,
specificity, required equipment were desired by Mercodia AB. Equipment requirements are described above in the technology description. Number of samples, accuracy, precision, LOD, LOQ and recovery were added since these were often given in the articles and thus used to obtain a better comparison. In Appendix 2, the parameter number of samples describes the number of individuals used in the study. These are included to ensure the reliability of the article. The sample volume is referred to as the volume of plasma or serum, in μL, that is required to conduct the experiment, or how much that is inserted in a well. In the latter case, it is explained in the table. Note: the volumes are generally the sample volumes from the patient e.g. plasma or serum, that is needed prior to adding buffers or similar for the analytical sample preparation. Accuracy is how close the results are to the true value. Values are presented in percent and ppm or with phrases as “good” according to the descriptions in the original article. Precision is how close the results are relative to each other, here the values are given in intra- and interassay percentage.
Sensitivity describes the smallest amount of the sample that can be detected by the system. The parameter is displayed as “good”, “high” or in g/mL or in mol/L depending on the article it is taken from. This is closely connected to LOD, which stands for Limit Of Detection and is given in mol/L or g/mL and LOQ, which represent the Limit Of
Quantification in mol/L or g/mL. Table 2.2, 4.4, 7.2 and 7.3 Appendix 2 show the value for Lowest Limit Of Quantification (LLOQ). Measure range is an interval between the lowest and highest amount of the biomarker that can be detected with reasonable accuracy, given in concentration per volume or mass per volume. Note: different units are compiled in the tables as they are taken directly from the corresponding articles. The throughput refers to the time (minutes or hours) it takes to do the analysis, excluding sample preparations. Sample cost is the cost in SEK for each sample without taking the equipment in consideration.
Multiplexing is a yes or no parameter that describes whether or not the method can effectively analyse several biomarkers in one run. Recovery describes the concentration of the target molecule that is reported recovered/detected by the method in contrast to a known concentration within a sample. The result agrees with a standard in a percentage that could be over 100. The closer to 100% the better.
correspondence to identify true m/z ratios. This ranking is biomarker specific, regards each table separately and must not be compared across different tables. The scale goes from A to D, where A is the highest score in each separate table.
Values for the parameters did not have the same units and we have avoided to convert the units in order not to tamper with interpretations, thus some parameters are given with several units.
All values presented in the tables are for biomarkers in complex matrices - serum or plasma.
6. Summarised comparison of the methods
Below, all the seven different methods and ELISA are compared to one another in respect to the four different biomarkers mentioned in the background of this report. The values are merged from the gathered data presented in Appendix 2. Values for the parameters did not have the same units and we have avoided to convert the units. This in order not to tamper with interpretations, thus some parameters are given in several units.
6.1 Insulin
Table 1. Values in the respective fields are assembled from the corresponding insulin tables for each method in Appendix 2. Upon different values in the same field there were no clear consensus among the same methods in the tables found in Appendix 2.
ELISA ELISA Ultra- sensitive insulin LC-MS/MS UPLC-MS/ MS
LC-IM/MS IA-LC-MS/MS MSIA-HR-A
M HTRF AlphaLISA® Sample volume 25 μL per well 25 μL per well 150 - 250 μL (<5 µL) 250 - 500 µL (<5 µL) 200 μL 100 - 500 μL 500 μL 20 μL per well 1 - 10 µL per well Accuracy - - 92.8-112 % Inter- assay: 87-96 % Intra- assay: 80-92 % - 92 - 99 % 85 -127 % 0.96 - 22 ppm Very good - Precision Inter- assay: 4.0-5.0 % Intra- assay: 2.8-4.0 % Inter- assay: 5.4-6.0 % Intra- assay: 4.2-5.3 % Inter- Assay: 3.0-13.0 % Intra- assay: 1.27-14.0% - 12 - 19 % 6-25% Inter- assay: 6.8-22.5% Intra- assay: 2.2-10.7% Inter- assay: <8% Inter- assay: 7.6% - 9.8% Intra- assay : 3.4-4.3% Inter- assay: 7.5-12.3 % Intra- assay: 2.9-8.5 %
Sensitivity - - ≤50pg/mL - Good Good High High -
Specificity D D A A B A A D D
Measure range 3 - 200 mU/L 0.15 - 20 mU/L 38.5 - 38500 pg/ mL - 0 - 4.0 ng/mL See table 4.1 1.5 - 1920 pmol/L - 0.8 - 3000 µlU/mL Through- put Incubation 60+15 min Incubation 60+30 min LC: <8 min per sample 25 samples/ day. 35 min/ high sensitivity assay 12 min/ sample IA: 1 - 2 h LC-MS/MS: 4 - 33 min High Medium through -put 3 h
Sample cost - - - - Low - - ≈ 2 SEK/
sample
(0.43 μg/L) μ mU/L (0.0065 μg/L) (10.8 pmol/L) ng/mL LOQ - - ≥ 38.46 pg/mL 200 pg/mol 0.3 ng/mL 0.5 ng/mL Alt. 0.78 pmol/L 1.5 - 15 pmol/L 0.345 ng/mL - Recovery Upon addition 94-113 % Upon addition 97-107 % 68-113 % - 64 % 33-44 % alt. 76-108 % - 104-126 % 94-112 %
6.2 Proinsulin
Table 2. Values in the respective fields are assembled from the corresponding proinsulin tables for each method in Appendix 2. Upon different values in the same field there were no clear consensus among the same methods in the tables found in Appendix 2.
ELISA LC-MS/MS UPLC-MS/MS LC-IM/MS IA-LC-MS/MS MSIA-HR-A
M HTRF AlphaLISA® Sample volume 50 μL per well - - - 100 μL - - - Accuracy - - - - 85 - 127 % - - - Precision Intraassay: 2.5 - 3.2 % Interassay: 5.0 - 6.1 % - - - 6 - 25 % - - - Sensitivity - - - - ~0.05 ng/mL - - - Specificity B - - - A - - - Measure range 3.3 - 130 pmol/L - - - - Through- put Incubation 60+60+15 min - - - IA: 1 - 2 h LC-MS /MS: ~33 min - - - Sample cost - - - - Multi- plexing No - - - Yes - - - LOD 1.7 pmol/L - - - ~0.05 ng/mL - - - LOQ - - - -
Recovery Upon dilution:
6.3 C-peptide
Table 3. Values in the respective fields are assembled from the corresponding C-peptide tables for each method in Appendix 2. Upon different values in the same field there were no clear consensus among the same methods in the tables found in Appendix 2.
ELISA ELISA Ultrasensitiv e C-peptide LC-MS/MS UPLC-MS/ MS LC-IM/MS IA-LC-MS/ MS MSIA-HR-A M HTRF AlphaLISA® Sample volume 25 μL per well 50 μL per well - - - 100-150 µL - - 1-10 µL
Accuracy - - - 85 - 127 % - Very good -
Precision Intra- assay: < 6.8 % Inter- assay: < 4.8 % Intra- assay: 3.9-6.2 % Inter- assay: 2.2-5.3 % - - - 4-25 % - Intra- assay: 3.5-5.0 % Inter- assay: 4.9-8.8 % Intra- assay: 3.9-4.8 % Inter- assay: 7.7-11.8 %
Sensitivity - - - Good - High Good
Specificity C C - - - A - C+ B Measure range 100-4000 pmol/L (0.3-12.0 ng/mL) 5-280 pmol/L (0.015-0.846 ng/mL) - - - 0.11 - 27.7 ng/ml - - 23.5 - 30 000 pg/mL Through- put Incu- bation 60+60+15 min Incu- bation 60+60+30 min - - - IA: 1 - 2 h LC-MS /MS: 4 - 33 min - - 3 h Sample cost - - - ~ 20 - 30 SEK/ sample Multi- plexing
No - - - - Yes - No No
6.4 Glucagon
Table 4. Values in the respective fields are assembled from the corresponding glucagon tables for each method in Appendix 2. Upon different values in the same field there were no clear consensus among the same methods in the tables found in Appendix 2.
ELISA LC-MS/MS UPLC-MS/MS LC-IM/MS IA-LC-MS/MS MSIA-HR-A
M HTRF AlphaLISA® Sample volume 25 μL per well 200-250 μL (<5 µL) 400 µL (<5 µL) - 500 µL - - 1 - 10 µL Accuracy - Interassay: <4.0 % Intraassay:: <12 % 68.6 - 87.4 % - - - - - Precision Intraassay: 3.3 - 5.1 % Interassay: 7.5 - 9.5 % Intraassay: <13 % Interassay: <14 % Intra- and interassay >20 % - Intraassay: 2.2 - 10.7 % Interassay: 6.7 - 22.5 % - - Intraassay: 12% Interassay: 13%
Sensitivity 1 pmol/L Good High - Good - - Good
Specificity C B A - A - - D
Measure range 1.5 - 120 pmol/L 0.5 - 100 pmol/L alt. 10 – 10 000 pg/mL 15 - 1000 pg/mL - Detection range: 2 - 750 mg/dL - - 20 - 30 000 pg/mL Throughput 18-22 h (overnight) + 15 min substrate incubation - 3.6 - 5 min/ sample
- IA: 1h & 15 min
LC-MS/MS: ~13 min
- - 3 h
Sample cost - - Low/very low - - - - ≈ 20 - 30 SEK/
sample Multi-
plexing
No - Yes - Yes - - No
7. Remarks regarding comparison of the methods
Here we discuss the results between the seven methods and ELISA in respect to the different parameters within Table 1-4. There is also a short economic discussion regarding the prices of the different methods.
The measuring ranges in our results were not consistently documented. Quantification ranges are more common than detection ranges and thus used for comparing measuring range. Also, LLOD (lowest limit of detection) is documented more frequently than ULOD (upper limit of detection) and serves as a guideline for sensitivity comparisons which are more relevant to the studied biomarkers as they occur in low concentrations.
The number of samples used in the different experiments is shown in tables in Appendix 2, where the information was available. This information is not compiled in the result tables since the number of samples are different between the articles and therefore difficult to compile into an interval.
7.1 Comparison Between ELISA and LC-MS Methods
From the various articles read across this project, some clear strengths and weaknesses for both ELISA and LC-MS/MS methods have been noted. As described in many of the articles, and elegantly summarised by Cross & Hornshaw (2016),
ELISAs main strength lies in the high amount of sensitivity that can be achieved. This without the need of any expensive equipment, while at the same time being fairly easy to use. Most of the articles use ELISA to verify the results of other methods.7.1.1 Accuracy for LC-MS/MS
As described in the result, accuracy is a parameter of great importance to the comparison since it defines reliability. The accuracy depends on both sensitivity and specificity as it includes aspects of detecting the right compound as well as the right amount. While LC-MS/MS methods can achieve a high specificity, one of the greater problems is the method sensitivity. This is due to poor fragmentation, formation of multiple charged ions, and matrix effects such as adsorption of analytes to the containers and instability of its compounds (Lapko et al. 2013). However, a high sensitivity can be achieved with
LC-MS/MS offers a more accurate quantification with less false positives than other immunochemical methods such as ELISA (Cross & Hornshaw 2016). This technology is quickly developing to be a powerful technique in the field of quantifying peptides, previously only analysed by immunochemical assays (Lapko et al. 2013).
However, in Table 1, 2, 3 and 4 it is shown that ELISA overall has better recovery rates indicating higher accuracy than LC-MS/MS based methods. The conflicting results can be partially explained by the fact that accuracy is not that well documented in most studies, and for some biomarkers not studied at all. In general, the recovery rates were either very much lower than for ELISA, or relatively close but broader and less precise.
For immunochemistry-based methods, specificity is depending on antibody-antigen interactions and thus different interaction conditions will affect specificity. MS-based methods may use immunoreactions but only for the purpose of enriching the sample, allowing a more concentrated sample to reach the MS-step. The immunoaffinity step is not specificity-determinant. LC-MS/MS methods have high specificity due to having a
detection mechanism where the mass/charge ratio is a very specific characteristic, it does require a spectral library in order to map the right peak to the right molecule.
7.1.2 Automatisation and multiplexing
Modern instruments allow LC-MS/MS workflow to be more automatised, which reduces the influence of the human error. Automatisation makes it possible to run many samples unsupervised which increases the throughput, lets the system perform with minor sample volumes and comes with high intra- and interassay reproducibility (Cross & Hornshaw 2016). In addition, the successful development of LC-MS/MS has given the method the quality of multiplexing, an advantage that is lacking in antibody-based methods.
Features discussed above are offered by LC-MS/MS methods due to the multiple
instrumentation that forms the system. The workflow might require supervision and the results gained are often complicated graphs that need training to be understood. In other words, these methods require qualified personnel with experience in the subject,
something that might be hard to find.
7.1.3 Sample preparation with respect to complex matrices
When it comes to complexity of a sample, it can affect the analytical correctness and reproducibility of an experiment when using a LC-MS/MS method. Without sufficient sample preparation, matrix effects will occur. Therefore, it is important with a thorough, sample preparation to get a good signal-to-noise ratio. This is of utter importance when working with complex matrices such as plasma and serum (Smeraglia et al. 2002).
LC-MS/MS. However, the reference material used for calibrating the detection curve with ELISA needs to be chosen with care to avoid misdetection among analogs or isoforms. The used antibody need to have corresponding specificity towards both the analyte and the standards used as reference (Lapko et al. 2013). The secondary structure might vary among different references due to minor changes in primary structure, which affects
immunochemistry-based methods but not MS-based ones, as quantification relies solely on the integrity of the analytes primary structure.
In immunoaffinity LC-MS/MS, antibodies are used in the purification step of the analysis. This combines the specificity of the LC-MS/MS method with the sensitivity of the
immunoaffinity, making it possible to match the sensitivity of current immunoassays (Taylor et al. 2016). A drawback with this method is the cost of antibodies and the incubation time of the sample preparation, which affects the throughput of the method negatively.
7.1.4 UPLC-MS/MS
Results referring to UPLC-MS/MS are shown in Table 1 and 4. It has the disadvantage of not covering C-peptide nor proinsulin. However, with the existing data for glucagon and insulin, it has the highest specificity score and a quite low cost per sample. In Table 4 we see that the method presents a high sensitivity and a LOQ of 15 pg/mL and a wide
measuring range of 15 - 1000 pg/mL. It is debatable whether an accuracy from 68.6 - 87.4 % is good because of the lack of information from ELISA. Nonetheless, it is an advantage to avoid using antibodies due to cross-reactivity and the false positives that they might
contribute to. ELISA in general had a poorer detection range than the LC-MS/MS methods. UPLC-MS/MS was shown to have the best overall detection range with very low LOQ values.
7.1.5 LC-IM-MS
During this project we were pleasantly surprised about the great potential of LC-IM-MS. Unfortunately, very little information was found about this method, specially for the
biomarkers of interest. The method showcases better precision than several other methods with the lowest LOQ in our results, see Table 1. LC-IM-MS has potential to give great
results in the future, tentatively in combination with bottom up approach, and therefore this method should be more investigated for biomarker quantification.
7.1.6 Availability of data about the biomarkers among the methods
for its absence in the studies. (Thomas et al. 2017)
Another difficult biomarker to quantify is C-peptide which showed to be difficult to analyse by the LC-MS/MS methods. Only IA-LC-MS/MS was able to detect it among the LC-MS based methods. This is most likely because IA-LC-MS/MS utilises an immunochemical separation step which the other LC methods do not. We do not otherwise know what factor makes it hard to find via LC-MS methods.
7.2 Comparison Between ELISA and other immunochemical
methods
ELISA has better recovery rate in comparison to the other methods, although AlphaLISA®
and ELISA performed almost identically when screening for insulin and had comparable results for C-peptide. AlphaLISA® had a slightly worse performance than ELISA when
analysing glucagon. HTRF was also similar but in general had larger intervals in recovery percentages than ELISA.
7.2.1 AlphaLISA
®When comparing AlphaLISA® and ELISA, the former has a lower sample volume and is
faster due to the washing-steps not being required. In addition, AlphaLISA® has lower
cross-reactivity to proinsulin than some commercially available ELISA kits when analysing C-peptide. This makes AlphaLISA® a method that potentially can outshine ELISA. Yet, there
does not exist any commercially available AlphaLISA® kits for proinsulin, while it does exist
for ELISA. ELISA is a broader method with the advantage that it can be used on biomarkers that are not covered by AlphaLISA® . AlphaLISA® also requires a special laser instrument for
detection, making the method more expensive to begin with.
When compared, Table 3 presents the need of 25 to 50 µL of the sample for ELISA while AlphaLISA® only needs 10 µL to quantify C-peptide. The latter method also has a better
specificity and faster throughput for the same biomarker. Table 4 shows that ELISA needs 25 µL in sample volume while AlphaLISA® still only needs 10 µL to quantify glucagon. In
addition, AlphaLISA® has faster throughput while ELISA has a better specificity.
7.2.2 HTRF
The major differences between ELISA and HTRF are mainly price, time of execution and sample preparation. HRTF is a faster analytic method. The whole process takes half the time compared to ELISA, it only has about 5 steps while ELISA has approximately 13.
We believe that every immunochemical method mentioned in this report should in theory work for all the same biomarkers as ELISA does, since they all operate based on antibodies. Although, sufficient data could not always be found, such as for HTRF on proinsulin and glucagon.
7.2.3 Hook effect in immunochemical methods
Articles studied during this project showed an inaccurate specificity for all the
immunoassay methods. This could be a consequence of cross-reactivity, which causes false positives or of the “Hook effect” which causes false negatives. Hook effects occurs when the concentration of antibodies is much greater than the concentration of analytes. All the antigens will bind to one pair of antibodies until there is no free antigens left. This high concentration of unbound antibodies will bind to the antigens in solution and will be washed away in the final washing step. This leads to underestimation of the true antigen concentration as some are removed from the sample. The Hook effect only occurs when the sandwich principle is used, and it is named after the point where the solution starts to be saturated, the so-called Hook point. (Einhorn & Krapfenbauer 2015)
7.3 Economics
When it comes to expenses, ELISA is the method least dependent of equipment and thus require no large investment in order to be used. This comes at the price of having the highest sample cost out of all the methods that were researched. In contrast, most of the LC-MS/MS methods have a very low sample cost but all require expensive equipment to be used. HTRF and AlphaLISA® can be considered the middle ground, where they require
some equipment but offer a lower sample cost than ELISA, in the case of HTRF even a significantly lower sample cost. An interesting note is that the articles we found stated a price range of 2 - 8.4 SEK. These ranges were for the material costs only, if one were to perform the method independently (Einhorn & Krapfenbauer 2015). In contrast, some ready to use HTRF kit products that were found online could have a cost per sample as high as 28 SEK and we found none cheaper than 11.5 SEK. To reach that price you had to buy a kit of 10 000 wells for 115 000 SEK, which is much too expensive for most studies.
8. Conclusion
To conclude, our results indicate that advantages of immunoassays include great recovery and precision, in particular when it comes to ELISA. The availability of these diverse
commercial kits strengthens the expanse of the methods which is manifested in our results. However, parameters as specificity and throughput are proved to be superior for LC-MS/MS methods.
In addition, accuracy, LOD and LOQ are also better for LC-MS/MS methods when compared to immunoassays, which makes the methods an alternative to ELISA.
Multiplexing is a great advantage of LC-MS/MS. The cost per sample is another asset, but this advantage comes at the expense of higher equipment prices. Furthermore, it requires more expertise to interpret the results from these methods.
In summary, LC-MS/MS methods are extensively supported by scientific articles reporting on biomarker detection and quantification. The methods demand streamlining and
standardisation for each analyte and ELISA will be used to evaluate reliability of LC-MS/MS for years to come.
So, the question remains: Immunoassay or LC-MS/MS? Our answer is that it depends on the purpose of the quantification and the requirements for each parameter.
9. Ethical Discussion
This is meant as an additional section to raise the readers awareness about potential ethical dilemmas that could occur within the fields of biomarker detection. It also touches upon some problems that Mercodia AB might have to deal with in the future in case there is a big shift into LC-MS/MS technologies within the research field.
9.1 Production of Antibodies
Antibodies are truly valuable tools in research such as detection and quantification of biomarkers in complex samples and treatment of, for example, immune deficiencies. There are many advantages of using antibodies for these applications due to their high sensitivity and specificity. But there are some disadvantages, especially during the production of the antibodies. To produce polyclonal antibodies, one need to inject a chosen substance into laboratory or farm animals such as mice, rabbits or horses. The immune systems of the animal then starts to produce antibodies directed against the injected substance. The issue of using animals to obtain antibodies is that the antibodies become species-specific,
leading to the need of being humanised before the antibodies can be used for human samples. The binding parts of the antibodies will keep its animal origin but the non-binding parts are exchanged for human types. (Fredholm. 2007)
To collect the antibodies the animal is first heavily anaesthetised, then a needle is inserted directly into the heart. When no more blood can be extracted the animal is put down with a lethal injection of pentobarbitone (Animalresearchinfo. 2014). Some may argue that this is a harsh treatment of animals and some may say that it is pain-free and necessary. One thing is certain: we are finding ourselves crossing into an animal rights discussion for research purposes. According to the Swedish Animal Protection Act a strict requirement for animal use is that the research can not be accomplished in any other satisfactory manner without the use of animals (Riksdagsförvaltningen. 2018).
From the production method described above we receive polyclonal antibodies. These antibodies come from different B-cells that can recognise multiple epitopes on the same antigen. Monoclonal antibodies represent, in contrast to polyclonal antibodies, antibodies that only bind to a specific epitope on an antigen which make them favorable for
survive the fusion process (Denovo Biotechnology. 2017). There is also some benefits with polyclonal antibodies because they are more tolerant of minor changes in the antigen.
9.2 Diagnostics or Research Only
The kits produced by companies such as Mercodia AB has the main purpose of being used for research and not for clinical diagnostic purposes. Nevertheless, the products are being used for diagnostic purposes in some places around the world and therefore reliability is essential to avoid misdiagnosed patients. In such cases, companies like Mercodia AB could be argued to have a responsibility to produce reliable products when it comes to clinical diagnostics as well, even though that is not their initial purpose. However, it is debatable whether the producer carries an ethical responsibility rather than judicial for how their products are applied by the clients. Since there is an interest for diagnostic methods, by expanding the area of use, the company would probably gain customers and also probably profit a lot from it.
In this report we have compared many different technologies and methods for detection and quantification of biomarkers. Some of them seem to be better than ELISA and could therefore be better suited for diagnostic purposes. However, some of them require
expensive equipment and a lot of experience which probably would not make it suitable for diagnostic purposes. The cost for reliable tests may not be a price everyone is able to pay which further divides the rich from the poor and increases the social gaps. So, what is a company, such as Mercodia AB, supposed to do? Stay with the same technology and try to make it more available, or go for the already more advanced technologies and join the future?
9.3 New Standard Technology for Detection and
Quantification
Let us say for the sake of discussion that we found out through our research that the best way to detect and quantify biomarkers in complex matrices is through a form of
LC-MS/MS method. These are methods that, once again, require expensive equipment that not all research facilities have access to nor can afford. An UHPLC (second hand) can cost up to 400 000 SEK (Liquid Chromatographs (HPLC), Conquer Scientific. 2019) and a mass spectrometer-LCMS system (second hand) can cost over 1 000 000 SEK (Mass
equipment regularly. Many small research facilities may not have the funds to do this and therefore will not be able to perform the same high-quality experiments.
If this became the new standard for all kits that companies like Mercodia AB would adapt their methods for, then it could potentially cripple small research facilities that cannot afford the required equipment. They would perhaps be forced to send their samples for analysis to a contract research organisation such as Sci Life in Sweden, and that may increase the cost and could affect the reliability of the result due to the changes of environment. This may also mean a loss of customers for companies that would start developing kits customised for LC-MS/MS methods instead of for example ELISA.
9.4 Production and Transport
10. Acknowledgements
The authors would like to thank:
Our supervisor, Lena Henriksson, for guidance throughout the project.
Lecturers Elisabeth Sköld and Lennart Köhler for instructions in project organization and report writing.
Representative of Mercodia AB, Johan Ledin for answering our questions and providing useful feedback.
Mercodia AB for the project and an interesting site visit.
Last but not least, a big thanks to our fellow students on the course for valuable responsive comments on our report.
11. Statement of Contribution
All authors have conducted the literature study together in which the data was gathered. All authors have been taking active parts in discussing all aspects of the report and given feedback to one another. All authors have had a hand in the formation of each section in this report.
C. Tiger was the main contributor to the background section of the report and wrote the technology sections on ELISA and HTRF. He created fig 1-3 and 7 and wrote figure texts for figures beside 2 and 3.
A. Ivert has contributed to the report by writing the IA-LC-MS/MS method description and the ethical discussion. She also designed the tables.
T. Martinez has written the description of the method UPLC-LC/MS and data collection of HTRF. She also contributed with the workflow, the explanation of parameters and the conclusion.
R. Upite has contributed with the MSIA-HR/AM method description, figure 6 and
collecting data for this method used in the comparisons. She also compiled the “Vocabulary and Abbreviations”.
V. Umenius has contributed to this report with the method description for LC-MS/MS and the figure texts for figure 2 and 3. He also contributed with the data collection of
LC-MS/MS.
S. Wärmegård contributed mostly to the method description and data collection for
LC-IM-MS and AlphaLISA® . She also created figure 8, contributed to the explanation of the
parameters and designed the tables.
12. References
12.1 Articles
Blackburn M. 2013. Advances in the quantitation of therapeutic insulin analogues by LC–MS/MS. Bioanalysis 5: 2933–2946.
Carslake HB, Pinchbeck GL, McGowan CM. 2017. Evaluation of a Chemiluminescent Immunoassay for Measurement of Equine Insulin. Journal of Veterinary Internal Medicine 31: 568–574.
Chambers EE, Fountain KJ, Smith N, Ashraf L, Karalliedde J, Cowan D, Legido-Quigley C. 2014. Multidimensional LC-MS/MS Enables Simultaneous Quantification of Intact Human Insulin and Five Recombinant Analogs in Human Plasma. Analytical
Chemistry 86: 694–702.
Chen Z, Caulfield MP, McPhaul MJ, Reitz RE, Taylor SW, Clarke NJ. 2013. Quantitative Insulin Analysis Using Liquid Chromatography-Tandem Mass Spectrometry in a High-Throughput Clinical Laboratory. Clinical Chemistry 59: 1349–1356.
Cross TG, Hornshaw MP. 2016. Can LC and LC-MS ever replace immunoassays? Journal of Applied Bioanalysis 2: 108–116.
Dong S, Gu Y, Wei G, Si D, Liu C. 2018. Direct comparison of LC–MS/MS and RIA methods for the pharmacokinetics assessment of human insulin in preclinical development. Biomedical Chromatography 32: e4323.
Dufield DR, Radabaugh MR. 2012. Online immunoaffinity LC/MS/MS. A general method to increase sensitivity and specificity: How do you do it and what do you need?
Methods 56: 236–245.
Einhorn L, Krapfenbauer K. 2015. HTRF: a technology tailored for biomarker
determination—novel analytical detection system suitable for detection of specific autoimmune antibodies as biomarkers in nanogram level in different body fluids. The EPMA Journal, doi 10.1186/s13167-015-0046-y.
Farino ZJ, Morgenstern TJ, Vallaghe J, Gregor N, Donthamsetti P, Harris PE, Pierre N, Freyberg R, Charrier-Savournin F, Javitch JA, Freyberg Z. 2016. Development of a Rapid Insulin Assay by Homogenous Time-Resolved Fluorescence. PLOS ONE 11: e0148684.
Fredholm L. 2007. Antikroppar som läkemedel. WWW-dokument 2007-12-01:
Guillarme D, Veuthey J-L. 2017. Chapter 1 - Theory and Practice of UHPLC and UHPLC–MS. I: Holčapek M, Byrdwell WmC (red.). Handbook of Advanced Chromatography/Mass Spectrometry Techniques, s. 1–38. AOCS Press,
Henry H, Lannoy D, Simon N, Seguy D, D’Herbomez M, Barthélémy C, Décaudin B, Dine T, Odou P. 2017. Immunoassay quantification of human insulin added to ternary parenteral nutrition containers: comparison of two methods. Analytical and Bioanalytical Chemistry 409: 3677–3684.
Heyduk E, Moxley MM, Salvatori A, Corbett JA, Heyduk T. 2010. Homogeneous Insulin and C-Peptide Sensors for Rapid Assessment of Insulin and C-Peptide Secretion by the Islets. Diabetes 59: 2360–2365.
Howard JW, Kay RG, Jones B, Cegla J, Tan T, Bloom S, Creaser CS. 2017. Development of a UHPLC–MS/MS (SRM) method for the quantitation of endogenous glucagon and dosed GLP-1 from human plasma. Bioanalysis 9: 733–751.
Howard JW, Kay RG, Tan T, Minnion J, Ghatei M, Bloom S, Creaser CS. 2014. Development of a high-throughput UHPLC–MS/MS (SRM) method for the quantitation of endogenous glucagon from human plasma. Bioanalysis 6: 3295–3309.
Kanu AB, Dwivedi P, Tam M, Matz L, Hill HH. 2008. Ion mobility–mass spectrometry. Journal of Mass Spectrometry 43: 1–22.
Kinumi T, Mizuno R, Takatsu A. 2014. Quantification of serum C-peptide by isotope-dilution liquid chromatography–tandem mass spectrometry: Enhanced detection using chemical modification and immunoaffinity purification. Journal of Chromatography B 953–954: 138–142.
Lapko VN, Miller PS, Brown GP, Islam R, Peters SK, Sukovaty RL, Ruhn PF, Kafonek CJ. 2013. Sensitive glucagon quantification by immunochemical and LC–MS/MS methods. Bioanalysis 5: 2957–2972.
Lapthorn C, Pullen F, Chowdhry BZ. 2013. Ion mobility spectrometry-mass spectrometry (IMS-MS) of small molecules: Separating and assigning structures to ions. Mass Spectrometry Reviews 32: 43–71.
Lee AYH, Chappell DL, Bak MJ, Judo M, Liang L, Churakova T, Ayanoglu G, Castro-Perez J, Zhou H, Previs S, Souza SC, Lassman ME, Laterza OF. 2016. Multiplexed
Quantification of Proglucagon-Derived Peptides by Immunoaffinity Enrichment and Tandem Mass Spectrometry after a Meal Tolerance Test. Clinical Chemistry 62: 227–235.
Legg KM, Labay LM, Aiken SS, Logan BK. 2019. Validation of a Fully Automated
double-antibody sandwich ELISA for rapid detection to C-peptide in human urine. Journal of Pharmaceutical and Biomedical Analysis 162: 179–184.
Miyachi A, Kobayashi M, Mieno E, Goto M, Furusawa K, Inagaki T, Kitamura T. 2017. Accurate analytical method for human plasma glucagon levels using liquid chromatography-high resolution mass spectrometry: comparison with
commercially available immunoassays. Analytical and Bioanalytical Chemistry 409: 5911–5918.
Moser AC, Hage DS. 2010. Immunoaffinity chromatography: an introduction to applications and recent developments. Bioanalysis 2: 769–790.
Nedelkov D, Niederkofler EE, Oran PE, Peterman S, Nelson RW. 2018. Top-down mass spectrometric immunoassay for human insulin and its therapeutic analogs. Journal of Proteomics 175: 27–33.
Oran PE, Jarvis JW, Borges CR, Sherma ND, Nelson RW. 2011. Mass spectrometric immunoassay of intact insulin and related variants for population proteomics studies. PROTEOMICS – Clinical Applications 5: 454–459.
Peterman S, Niederkofler EE, Phillips DA, Krastins B, Kiernan UA, Tubbs KA, Nedelkov D, Prakash A, Vogelsang MS, Schoeder T, Couchman L, Taylor DR, Moniz CF, Vadali G, Byram G, Lopez MF. 2014. An automated, high-throughput method for targeted quantification of intact insulin and its therapeutic analogs in human serum or plasma coupling mass spectrometric immunoassay with high resolution and accurate mass detection (MSIA-HR/AM). PROTEOMICS 14: 1445–1456.
Pitt JJ. 2009. Principles and Applications of Liquid Chromatography-Mass Spectrometry in Clinical Biochemistry. The Clinical Biochemist Reviews 30: 19–34.
Riksdagsförvaltningen. 2018. Djurskyddslag (2018:1192) Svensk författningssamling 2018:2018:1192 - Riksdagen. WWW document:
https://www.riksdagen.se/sv/dokument-lagar/dokument/svensk-forfattningssa mling/djurskyddslag-20181192_sfs-2018-1192. Accessed 4 June 2019.
Sakamoto S, Putalun W, Vimolmangkang S, Phoolcharoen W, Shoyama Y, Tanaka H, Morimoto S. 2018. Enzyme-linked immunosorbent assay for the
quantitative/qualitative analysis of plant secondary metabolites. Journal of Natural Medicines 72: 32–42.
Simmons E, Johnson D, Delhaye R, Colwell J, Clor E, Santa P, Bowsher RR. 2019.
Smeraglia J, Baldrey SF, Watson D. 2002. Matrix effects and selectivity issues in LC-MS-MS. Chromatographia 55: S95–S99.
Taylor SW, Clarke NJ, Chen Z, McPhaul MJ. 2016. A high-throughput mass spectrometry assay to simultaneously measure intact insulin and C-peptide. Clinica Chimica Acta 455: 202–208.
Thomas A, Brinkkötter PT, Schänzer W, Thevis M. 2017. Simultaneous determination of insulin, DesB30 insulin, proinsulin, and C-peptide in human plasma samples by liquid chromatography coupled to high resolution mass spectrometry. Forensic Toxicology 35: 106–113.
Thomas A, Schänzer W, Thevis M. 2014. Determination of human insulin and its analogues in human blood using liquid chromatography coupled to ion mobility mass
spectrometry (LC-IM-MS). Drug Testing and Analysis 6: 1125–1132.
Warnken T, Huber K, Feige K. 2016. Comparison of three different methods for the quantification of equine insulin. BMC Veterinary Research, doi
10.1186/s12917-016-0828-z.
Whiteaker JR, Paulovich AG. 2011. Peptide immunoaffinity enrichment coupled with mass spectrometry for peptide and protein quantification. Clinics in laboratory medicine 31: 385–396.
12.2 Online sources
Animalresearchinfo. 2014. Antibody production. WWW-document 11-11-2014:
http://www.animalresearch.info/en/designing-research/experimental-design-an alysis/antibody-production/. Accessed 28 May 2019.
CBRNE Tech Index. 2018. Ion Mobility Spectrometry (IMS). WWW-document:
http://www.cbrnetechindex.com/Chemical-Detection/Technology-CD/Spectrom etry-CD-T/Ion-Mobility-Spectrometry-CD-S. Accessed 28 May 2019.
Conquer Scientific. 2019. Liquid Chromatographs (HPLC). WWW-document:
https://conquerscientific.com/lab-equipment/liquid-chromatographs-hplc/. Accessed 4 June 2019.
Conquer Scientific. 2019. Mass Spectrometers-LCMS Systems. WWW-document:
https://conquerscientific.com/lab-equipment/mass-spectrometers-lcms-systems /. Accessed 4 June 2019.
Denovo Biotechnology 2017. What are the advantages and disadvantages of using Monoclonal antibodies Vs Plyclonal antibosies? WWW-document:
https://www.denovobiotech.com/support-center/view-article/2-what-are-the-a dvantages-and-disadvantages-of-using-monoclonal-antibodies-vs-polyclonal-antib odies. Accessed 28 May 2019.
Eurofins (1). 2019. HTRF kit 10 000 wells. WWW-document:
https://www.eurofinsdiscovery.com/33-036-pi-3-kinase-htrf-assay-10-000-wells Accessed 5 June 2019.
Eurofins (2). 2019. HTRF kits 384 wells. WWW-document:
https://www.eurofinsdiscovery.com/33-016-33-016-pi-3-kinase-htrf-assay-384-wells Accessed 5 June 2019.
LabX. 2019. Mass spectrometer Listings. WWW-document:
https://www.labx.com/mass-spectrometer. Accessed 28 May 2019.
Pacific Immunology. 2019. WWW-documents:
https://www.pacificimmunology.com/resources/antibody-introduction/polyclon al-vs-monoclonal-antibodies/. Accessed 7 June 2019.
PerkinElmer. 2011. TDS – Human Insulin. WWW-document:
https://www.perkinelmer.com/lab-solutions/resources/docs/TDS_AlphaLISA_A L204.pdf. Accessed 28 May 2019
PerkinElmer. 2012. TDS – Human Connecting Peptide (C-peptide). WWW-document: https://www.perkinelmer.com/lab-solutions/resources/docs/TDS_AlphaLISA_AL299.p
df. Accessed 28 May 2019.
PerkinElmer. 2019. Alpha Reagents. WWW-document:
https://www.perkinelmer.com/se/category/alpha-reagents. Accessed 28 May 2019.
PerkinElmer. 2012. TDS – Glucagon AlphaLISA Immunoassay Kit. WWW-document:
http://www.perkinelmer.com/lab-solutions/resources/docs/TDS_AlphaLISA_AL 312.pdf Accessed 28 May 2019.
Thermo Fisher ScientificTM. 2019. MSIA Microcolumns. WWW-document:
