Panel discussion on large molecule quantification by LC–MS


The increasing importance of large molecule therapeutics has been coupled with advancements in LC–MS technology, opening up new opportunities. Typically, large molecules are analyzed using ligand-binding assays but with advanced progress in MS technology the community has seen increased success in overall sensitivity and selectivity. However, what does this mean for future analysis and what challenges remain for the use of LC–MS for large molecules.

In this free panel discussion, our experts will provide insights into their own research with large molecules including the challenges they have had to overcome, key trends they have seen and their future outlook of the development of this field. There is also an opportunity for your own questions to be answered in the live Q&A.

What will you learn?

  • Best practices of protein quantification and digestion
  • Considerations for top-down vs bottom-up approaches
  • Future developments (including enhancing assay sensitivity)

Q&A follow-up

QFor clarification: Reduction of sample preparation time when protein digestion is required?Cory: New reagents and buffers are being developed to accommodate for fast digestions.  In addition, there are now recombinant enzymes that react at much faster rates, at lower temperatures, than using conventional proteases.

QWhat are the dose levels in a preclinical study that can be handled by an LC-MS assay? Michael: Depends on the assay and the requirement of the project team?  Generally, at a preclinical stage, it should be possible to reach the desired LLOQs using modern top end LCMS, eg Xevo, 6500, Altis.

QDo the panelists have an opinion on how best to quantify intact proteins from raw LC-MS data?  Use one charge state?  Summation of charge states?  Deconvoluted signal? Michael: I have no direct experience, but at ASMS the consensus in a workshop appeared that deconvolution gave best sensitivity (though the software isn’t great for this).
QCan the panel ever imagine a single generic method for large molecule analysis, similar to protein crash for small molecules? Cory: There were some papers published by Davy Guillarme and Alain Beck on generic methods for size exclusion chromatography (SEC) and hydrophobic interaction chromatography (HIC) analyses.  I would recommend the interested reader to look those up.A universal approach can be applied for different proteins in animal matrices.

Michael: No, because there are different stages of molecule development which require different sensitivity and specificity levels, and therefore different kinds of approach. But at different stages there generally are common approaches. E.g. for say preclinical testing of human antibodies for example, approaches are now pretty standard. Also we have creative companies always looking for better approaches, leading to innovation. 

QAre any of you using MSIA for Immunoaffinity capture and purification as part of your sample prep workflows? Michael: Yes, we use MSIA for the analysis of intact insulins in biological matrix. We use these in combination with a Versette robot preparation system. Results are god and the technique is very robust.  
QWhen quantifying large multi-subunit protein analyte, how can an LC–MS method be sure the molecule is intact? Omnia: The immunocapture step itself is sometimes based on the intact 3D structure of the molecule, in addition to monitoring different surrogate peptides from different subunits.
QFor peptide quantitation, how essential is having a stable-label internal standard for your method? Omnia: It is very important to track variabilities during sample extraction and detection.
Michael: It is always nice, but not often possible. For large peptides, we use analogue large peptides, which work well if they have similar physicochemical properties. 
QWhat would you use to setup the calibration curve? Serial dilution of the protein standard, digested and use the peptide peak area/IS area to setup the cal curve, or use peptides to find the relationship with the protein concentration, then use peptides to setup the cal curve? So one is using protein standard and the other is using peptide standard, which would be more accurate and reliable? Rob: We construct calibration curves by spiking the protein reference material into matrix. This is considered the most accurate approach, normalising for digestion efficiency and any protein level purification.

QOne of the deficiencies of LBA assays is the inability to incorporate an internal standard.  So if you can include a suitable internal standard, then why not stay with chromatographic criteria (15%)? Michael: For hybrid assays we are using ligand binding as a part of the approach, so it is logical to use ligand binding acceptance levels. Also hybrid assays are more complex than simple small molecule approaches. I recommend AAPS (Rand Jenkins et al) and EBF publications (eg Schmidt and Knuttson) which discuss this in detail.

QWhat have you done to reduce the amount of time it takes for sample preparation?
Rob: We optimize reduction, alkylation and digestion times, establishing the shortest time needed to achieve the required assay sensitivity and robustness. We have also automated immunoprecipitation and SPE steps.
Michael: We use a robot preparation system for MSIA (Versette). For magnetic beads, the Kingfisher is a popular alternative. There are other auto prep systems available. 
QCan Rob talk a little more on his experience with automation in sample preparation? Rob: We have automated all immunoprecipitation steps using magnetic beads on the Kingfisher Flex platform. We have also recently implemented the A200, automated SPE platform. However, with a trap and back elute LC system we can inject and concentrate large volumes of dilute immunoprecipitated samples following digestion, negating the need for an offline SPE step.

QWhich kind of 2D-LC did you use? Heart-cut, multiple heart-cut or comprehensive? Rob: We used single heart cutting, although with modern LC systems that are capable of fast gradient delivery,  multiple heart-cutting is possible without the need for parking peaks in a loop or trap.

Q LC-MS instruments constantly get better (even ToF technology that is around for decades), when will the instruments be good enough in terms of sensitivity, acquisition speed and resolution? Or are we there already? Rob: As bioanalysts we are all too aware of the constant demands from our colleagues to achieve better assay sensitivity, with more potent and targeted drugs and a shift toward intact/sub-unit analysis I see the demand for more sensitivity remaining for some time. QqQ acquisition speed is now very fast and with scheduled MRM we can monitor hundreds of transitions in one run if we require, so I don’t feel there is a demand for more speed here. The speed of ToF detection has increased quite significantly recently, with some vendors now effectively decoupling sensitivity from acquisition speed. Although we have seen some good improvements over the years there still remains a trade-off for orbitrap technology however between acquisition speed and resolution. The major advantage to orbitrap technology is the level of resolution it can offer over the competitors, now offering up to 1 million. As mentioned earlier, with a trend toward intact or sub-unit analysis, we are going to require this resolving power and more in quantifying these molecules in the future.

Michael: Although I do not have direct experience, from presentations viewed recently at ASMS and other venues, I would suggest we are there already, apart from perhaps the most demanding clinical assays where ligand binding assays are still prevalent and probably preferred. 

QWhen quantifying large multi-subunit protein analyte, how can an LCMS method be sure the molecule is intact? Rob: The nature of a bottom up approach doesn’t owe itself to this task. However we can measure different surrogate peptides originating from the different sub-units and monitor ratio between them, which would be indicative of presence of sub-units but not necessarily infer the protein is intact. To ensure the molecule is intact we can use an immunoprecipitation. If we pull the protein down through capturing one of the sub-units, any sub-units not intact will be removed through subsequent washing. We then monitor the presence of the other sub-units to provide confidence that the protein is intact.

QFor labs suffering from high turnover, how do you justify developing the necessary experience to develop large molecule assays and hybrid approaches? Is it better to bring in the experience instead? Michael: Such a difficult question: The vendors e.g Thermo, Waters, are good at providing training now in the use of their kits, and how to optimize their methodologies. We are going to use them to help us develop our skills. ASMS for example do good value training courses in protein quan which can help with understanding. You have to convince management that this training will bring benefits. Large molecule analysis is mainstream now, so it is a necessary skillset for example, CROs. 

QAre there dis/advantages to using chemical extraction (protein crash) and sample clean-up (LLE) over IA approaches to sample prep prior to LC-MS quantitation? Michael: The big disadvantage is that the extracts (PPt or LLE) will contain peptides and proteins. These cause interferences. This may be OK for discovery or low sensitivity assays, but as you move up the development chain greater sensitivity is required and hybrid assays will generally give better sensitivity due to their more specific clean up.

Meet our panelists

MB x90
Michael Blackburn
Bioanalytical Scientist, Method Development
ARCINOVA (Alnwick, UK)

Michael completed postgraduate work at the University of Wales (Bangor, UK) and began work for the government in the UK, measuring the concentrations of pollutants in aquatic samples by gas and later liquid chromatography-mass spectrometry. He later joined the pharmaceutical industry with Sanofi (Guildford, UK), as a mass spectrometrist, doing bioanalytical and drug metabolism studies. Since then he has worked as a Principle Investigator and Head of Method Development at Covance (Alnwick, UK) and latterly for ARCINOVA as a method developer. The development of better insulin assays using hybrid MS approaches has been a major interest for about 10 years.

BronsemaKees x90
Kees Bronsema
Senior Scientist
Kees Bronsema, Senior Scientist in Bioanalytical Method Development LC–MS/MS at PRA (Groningen, Netherlands), has nearly 20 years of experience in the field. Initially, the focus was on small molecule quantitation but following the trend towards biopharmaceuticals, he completed postgraduate work on absolute quantitation of proteins and peptides with LC–MS/MS at the University of Groningen (Groningen, Netherlands) in 2015. His current role is to develop analytical methods for proteins and peptides with LC–MS/MS, which include both endogenous biomarkers and biopharmaceuticals.
Cory E. Muraco
Senior R&D Scientist
MilliporeSigma (MA, USA)

Cory is a Senior R&D Scientist in the Liquid Separations R&D group at MilliporeSigma, Bellefonte, PA. He completed his graduate studies at Youngstown State University in 2013, focusing on the analysis and characterization of oxidized proteins. Upon graduation, Cory joined MilliporeSigma in 2013, first joining the chemical standards R&D group, then transferring to the liquid separations R&D group. His current role at MilliporeSigma is to research, develop, and present on new particle technology for improved chromatographic separations of both small and large molecules as well as to develop new methodologies for characterizing biomacromolecules by several modes of chromatography.

Omnia A. Ismaiel
Senior Research Scientist

Dr Omnia Ismaiel is a Senior Research Scientist at Bioanalytical Labs, Pharmaceutical Product Development, USA. She is also an Associate Professor of Pharmaceutical Analytical Chemistry at Faculty of Pharmacy (Zagazig University, Egypt) and Affiliate Assistant Professor at School of Pharmacy (Virginia Commonwealth University, VA, USA). She was a Postdoctoral Fellow at Virginia Commonwealth University and a Postdoctoral Research Associate at University of Georgia, USA. She has been in the Bioanalysis field for more than 12 years. and has many years of teaching experience too. She is currently working on development and validation of Bioanalytical LC–MS methods for therapeutic peptides/proteins, glycan analysis and high-resolution MS.

Robert Wheller x90
Robert Wheller
Principal Scientist
LGC (Cambridge, UK)

Robert is a principal scientist at LGC where he leads of a group of scientists focused on providing a large molecule LC-MS bioanalytical service. He has expertise in the bioanalysis of small molecules, oligonucleotides, peptides and proteins utilizing LC-MS/MS and ligand binding assay technologies. He has developed a keen interest in the implementation of immunoaffinity, 2D-LC and HR-MS techniques for protein LC-MS quantification workflows.

Hongbin Yu x90

Hongbin Yu
Boehringer Ingelheim Pharmaceuticals (CT, USA)


Hongbin Yu is the Director of Bioanalytical Mass Spectrometry in the DMPK department at Boehringer Ingelheim Pharmaceuticals in Ridgefield (CT, USA). He obtained his Ph.D in organic chemistry from the University of Missouri-Columbia (MO, USA). He joined Boehringer Ingelheim in 2006 and has supported drug development for both small and large molecules. For small molecules, his responsibilities focused on biotransformation and bioanalysis utilizing LC–MS. For large molecules, his responsibilities included supervising immunoassay development/validation, critical reagent generation and pharmacokinetic analysis. He currently oversees regulated and non-regulated bioanalysis for small molecules by LC–MS and large molecules by hybrid LBA/LC–MS.

Sanjeev Bhardwaj

Senior Scientist
Janssen Biotherapeutics (PA, USA)

Sanjeev is working as a Senior Scientist in PK/PD Assay Development group of Biologics Discovery sciences at Janssen Biotherapeutics (PA, USA). He obtained his PhD in Protein Mass Spectrometry from the University of Cincinnati (OH, USA). He started his professional career with Applied Biosystems (now Sciex) as a scientist developing MS methods for proteins and peptides. His next job was at Merck (NJ, USA) developing assays for protein/peptide biomarkers of toxicity in animal models utilizing LC–MS and hybrid LC–MS approaches. In his current job at Janssen, he is leading the development of hybrid LBA/LC–MS assays for mAbs/proteins/peptides to support in vivo PK studies to support early drug development.

waters 150x80150x80 pra150x80 logo150x80150x80 lgc