Complex protein structures pose analytical difficulties that can be addressed by advanced mass spectrometry technologies and workflows, which can be used to comprehensively describe them.
Advances in protein engineering have empowered pharmaceutical developers to optimize and exploit the therapeutic potential of proteins much more quickly and cost effectively than usual, while maintaining or even enhancing their safety profile (1). Innovative recombinant protein therapeutics, such as monoclonal antibodies (mAbs), fusion proteins, and antibody-drug conjugates, are now being used to treat an expanding range of conditions, including cancers, inflammatory, autoimmune, and genetic diseases. Structurally larger in size than small-molecule drugs and functionally dependent on post-translational modifications (PTMs), these classes of therapeutics bring added complexity to the protein characterization process.
With greater protein complexity comes the need for robust analytical methods. Current analytical technologies, workflows, and data processing methods need to accommodate the requirements for biotherapeutic protein characterization (i.e., confirming identity and detecting PTM status for each residue, and measuring abundance of both major and minor intact protein isoforms to ensure the quality and consistency of these products). Such a thorough characterization effort would not only help protect patients for whom the drug will be developed, but also reduce the time required to bring these protein-based therapeutics to market. A robust, reproducible characterization protocol also functions as a quality control step to measure batch-to-batch variability of the drugs. As with all high-throughput fields, giving researchers the ability to automate time-consuming steps will make the characterization process even more productive.
This article considers the challenges associated with determining the structure of complex protein biotherapeutics, and how researchers can employ the latest mass spectrometry (MS) technologies and workflows to confidently characterize these increasingly important drugs in an efficient and comprehensive manner.
Recombinant protein therapeutics are intricate molecules, typically orders of magnitude larger in size than traditional small-molecule drugs. The production of these biotherapeutics relies on living cells or organisms that are extremely sensitive to a range of factors, such as species origin and culture conditions (1). As therapeutic efficacy often requires preservation of precise secondary or tertiary molecular structures, comprehensive characterization is required during development and manufacture. Gaining such structural insights can be difficult when synthesis is complicated-often requiring over 5000 critical process steps to generate a recombinant biotherapeutic (2).
Given the need to assure the safety and quality of these therapeutic products, there is a growing requirement for reliable analytical methods capable of quickly characterizing the structure of these molecules. There is also a need for these methods to reduce the number of test strategies required and the potential for operator error. Moreover, with the growing number of hybrid therapeutic products under development, such as antibodies fused to highly glycosylated or otherwise analytically challenging molecules, the technologies and workflows used by researchers to confirm the structure of these complex molecules must keep pace with the evolving biopharmaceutical landscape.
Comprehensive protein characterization typically involves a combination of intact mass analysis and peptide mapping to confirm both the total mass of the species and elucidate the substructure in fine detail. Ongoing improvements to the capabilities of high-resolution accurate mass (HRAM) analyzers and complementary dissociation techniques, as well as increasingly reliable and efficient protein digestion solutions, have enabled highly effective peptide mapping workflows that can probe protein structure at the individual amino acid level. Additionally, recent advances in MS technology are now enabling powerful non-denaturing liquid chromatography–mass spectrometry (LC–MS) methods for intact mass analysis, reducing the reliance on orthogonal, non-LC–MS-based approaches. These so-called “native” LC–MS strategies are enabling researchers to support the characterization of microheterogeneous isoform mixtures of covalently assembled molecules. The combined analysis directly correlates identified features to each proteoform, removing the need for manual inferences.
Traditionally, denatured, intact mass analysis generates convoluted MS spectra containing a mix of intact mass isoforms and wide charge state envelopes with much greater charge state overlap, which result in the production of complex mass spectra. Although spectral deconvolution algorithms can be used to interpret this complexity, these computational workarounds have their limitations. Performing MS in non-denaturing conditions, a characteristic feature of native MS, can help mitigate these issues. By spraying the intact protein sample in physiologically approximate buffers consisting of volatile salts, native MS enables the protein to retain its native structural characteristics and, in effect, shield internal basic residues from becoming protonated in a typical positive mode MS analysis. Folded protein isoforms have decreased charge states and increased spectral separation between each of the isoforms in a mixture, resulting in relatively simplified mass spectra. With native MS, it is often possible to achieve the resolution of intact protein isoforms, such as mAb glycoforms, without the need for sample pre-treatment.
Coupling native MS with separation techniques that are compatible with “native” mobile phases, such as size exclusion chromatography (SEC), provides a convenient and automated analytical workflow that can effectively resolve the complex protein architecture. As a result, native SEC–MS intact mass analysis can deliver cleaner spectra and provides reliable characterization data in a high-throughput fashion.
To characterize complex proteins in their intact state, in addition to optimizing the native MS methods, significant efforts are made towards charge reduction techniques with the goal of obtaining improved resolution. In recent years, advances in SEC–MS intact mass analysis have been driven by the advent of solution-phase charge reduction additives, such as triethylammonium acetate (TEAA). Due to the high pKa values associated with these additives, these reagents enable protons to be efficiently abstracted from the protein, reducing the overall charge (z) on the analyte. The lower value of zimparted by the additives shifts the mass-to-charge ratio (m/z) distribution, resulting in higher m/zratios and better spectral separation. Higher concentrations of additive can facilitate enhanced charge reduction and improve the spectral separation of intact protein isoforms. However, a frequent challenge encountered when working with solution phase additives is the need to regularly clean heated capillaries because the build-up of material can quickly have a detrimental impact on analytical performance.
Proton transfer charge reduction (PTCR), an alternative technology using gas phase ion source reagents, can help overcome the challenges posed by solution-phase additives. The ion–ion proton transfer reactions serve as an effective method for gas-phase charge reduction, yielding cleaner spectra while minimizing maintenance requirements. Recently, PTCR has become commercially available, providing researchers with wider access to this technology in a vendor-supported format. For example, an advanced ion source and modified dual-pressure linear ion trap (e.g., such as that only available in the latest Thermo Scientific Orbitrap Tribrid mass spectrometer) provides significantly increased ion–ion reaction efficiency, which provides support for enhanced PTCR capabilities. Through the application of other proprietary innovative ion management technologies, including hardware that more precisely manages electrical fields and reduces noise, these designs can maximize ion transmission from injection to detection, delivering robust qualitative and quantitative performance. Such improvements can eliminate electrospray ionization (ESI) source contamination and simplify the MSn spectra of intact proteins and complexes, ultimately increasing confidence with intact and top-down protein sequencing.
MS instrumentation. The use of PTCR as a powerful strategy for intact mass analysis has been further facilitated by the extended high-mass-range functionality offered by some modern instruments. The latest Orbitrap Tribrid system, for example, is now capable of performing high mass measurement and isolation up to 8000 m/z, enabling the higher-order analysis of large-protein complexes and their components. Moreover, this next-generation system is capable of supporting both peptide mapping and intact mass workflows in a single platform. Fragmentation techniques with higher energy collisional dissociation (HCD) and electron transfer dissociation (ETD) facilitate more detailed peptide mapping, while automated PTCR-enabled charge reduction applied to native MS yields improved resolution for intact mass analysis. Using ion trap isolation to further isolate smaller m/zwindows helps significantly improve the signal-to-noise ratio, allowing for more control in the experimental workflow to characterize complex biotherapeutics. Native MS coupled with sequential enzymatic dissection of individual subunits in highly glycosylated proteins, such as etanercept, can offer additional information on glycan heterogeneity compared to bottom-up peptide mapping, thereby acting as a fingerprinting tool to asses batch-to-batch variability of drugs (3).
The features of this advanced MS instrument help scientists obtain the high-quality protein characterization data necessary to drive the right decisions along with the ability to expand their experimental capabilities in the future.
Data acquisition. Further advances in MS technologies and workflows are helping researchers confidently characterize protein therapeutics more quickly. The latest intelligence-driven data acquisition strategies, such as charge state directed dissociation, are delivering improved analytical specificity and making data collection significantly faster. By enabling spectra to be processed using automated precursor determination, precursor charge state analysis can be performed in real-time, enabling the intelligent selection of the dissociation techniques to be employed as well as the optimal parameter settings for high-quality MSn acquisition. For example, higher charge state precursors may fragment significantly better using a combination of ETD and higher-energy collisional dissociation (EThcD), requiring both the duration of ion–ion reactions and relevant collisional activation energy to be imparted. Intelligent MS methods have been further extended utilizing data-independent acquisition PTCR to be implemented following ETD spectral acquisition. These workflows can boost efficiency and accelerate the generation of high-quality results through more complete MSn spectral acquisition and confident, automated data processing.
Data processing. Newly developed data processing tools, used in conjunction with state-of-the art MS systems, enable accurate interpretation of results. For example, the Sliding Window algorithm (Thermo Scientific BioPharma Finder software) used for the deconvolution of intact proteins “slides” along a chromatogram and acts as an overlay to generate time-integrated results (4). Traditional intact protein deconvolution involved averaging all the spectra corresponding to an arbitrarily selected chromatographic time. However, with numerous LC peaks eluting over time, averaging all the spectra in a selected time doesn’t accurately represent the protein isoforms present in separation. By interrogating smaller windows within a larger chromatographic range, modern software algorithms perform deconvolution multiple times in succession, resulting in a more accurate understanding of how the spectra behave and change over time. As the set window width moves along the LC time axis, different isoforms in the chromatographic separations can be detected multiple times. Redundant detections improve the confidence of reported isoforms, in terms of both mass and abundance. Using the minimal percentage offset enables the greatest number of redundant detections, thus allowing the detection of even minor components that may elute for shorter times. Applying these approaches to characterize biotherapeutic proteins allows researchers to analyze mass spectra of different protein isoforms with varied elution profiles. Performing deconvolution using the Sliding Window algorithm can, for example, provide more accurate drug-to-antibody (DAR) ratios, taking even the lower abundance species into account when analyzing antibody-drug conjugates.
A few years ago, analyzing an intact complex protein sample would have yielded a mass spectrum that would pose significant deconvolution challenges, making it difficult to be analyzed or interpreted. The latest MS systems, with the capacity for native analysis and charge reduction with PTCR, have opened up the possibility of comprehensive and automatable characterization of highly complex proteins in a single platform. However, with the ability to isolate every portion of the spectrum and closely examine the charge state distributions of the isoforms present, researchers must take care not to introduce implicit bias to their analysis. The application of the latest technologies can eliminate the risk of a skewed isoform distribution, especially when characterizing therapeutic proteins, relieving the possibility of introducing any bias.
The nature of protein therapeutics will continue to evolve as the field of protein engineering accelerates. Simultaneous advances in MS technology will ensure researchers have access to analytical tools that can keep up with the increasing complexity of modern biotherapeutic products.
1. H.A.D. Lagassé, et al., F1000Res. 6 (F1000 Faculty Rev): 113 (2017).
2. H. Schellekens, NDT Plus. 2, i27 (2009).
3. T. Wohlschlager, et al., Nature Communications 9 (1713) (2018).
4. A.O. Bailey et al., mAbs 10 (8) 1214–1225 (2018).
Aaron O. Bailey, PhD, is associate director of Mass Spectrometry Core Research at the University of Texas Medical Branch, Galveston, TX, and has worked in the Life Science Mass Spectrometry group at Thermo Fisher Scientific.
Vol. 44, No. 2
When referring to this article, please cite it as A.O. Bailey, “How Advanced Mass Spectrometry Technologies and Workflows are Delivering Comprehensive Protein Characterization,” Pharmaceutical Technology 44 (2) 43–45 (2020).