Advancing Antibody Analytics Through Mass Photometry

1 www.refeyn.com A Modern Tool for Next-Gen Therapeutics Handbook of Mass Photometry Applications in Antibody Analytics 2 www.refeyn.com Contents Introducing mass photometry for antibody analytics Antibody discovery and development steps Discovery Mass photometry: The perfect team player for antibody analytics Quantifying protein binding affinities using mass photometry Rapid analysis of bispecific antibody stability and target binding by mass photometry Candidate characterization Rapidly assessing reagent quality before binding studies with mass photometry Comparing mass photometry and SEC for antibody aggregation assessment Assessing protein samples by mass photometry and size-exclusion chromatography Measurement of NISTmAb aggregation with mass photometry AUC, and DLS Process development Characterization of forced antibody degradation with automated mass photometry Quantifying antibody aggregation at nano- and micromolar sample concentrations with the mass photometry antibody stability module Mass photometry end-to-end solutions 3 6 7 9 12 19 21 29 35 40 42 45 3 www.refeyn.com Introducing mass photometry for antibody analytics In this e-book, we present case studies demonstrating the capabilities of MP for antibody aggregation (a critical quality attribute, CQA) and binding assessment. The data showcase the strengths of MP for antibody characterization, and how it compares to frequently used techniques, like size-exclusion chromatography (SEC). MP offers key advantages that make it an invaluable tool to derisk antibody development and production: Measurements take just a few minutes and require only nanograms of sample. Antibody-based products are powerful therapeutic agents, but their discovery, development and production require careful, repeated bioanalytical characterization. Mass photometry (MP) is a powerful tool for biomolecular analysis that provides information about the mass distribution of a sample at the single-particle level, in solution and requiring little time and sample. With speed, resolution and ease of use superior to most other techniques, MP offers transformative capabilities for antibody characterization. 1 High-quality information with low resource investment: MP provides data comparable to gold-standard techniques, but with minimal acquisition time (1 minute) and sample consumption (15-30 ng). Each measurement costs <5$, and the instrument has a small footprint. 2 Measurements in solution at functionally relevant concentrations: MP measurements are performed in solution – with no column or surface interactions – at nM concentrations, in line with the usual physiological and functional concentrations of antibodies. If needed, it is possible to measure samples at μM concentration by using the MassFluidix™ HC microfluidics add-on. 3 User-friendly and modality-agnostic: MP instruments are easy to use, and sample preparation is straightforward. A simple dilution is enough to measure antigens and antibodies of any modality separately or in complex samples – saving processing time and minimizing optimization work. 4 www.refeyn.com Introducing MP MP works by illuminating 10–20 microliters of sample on a glass surface with a laser, and then measuring the interference of light scattered by molecules in a sample with light reflected by the glass surface (Fig. 2). The resulting signal, known as the ‘contrast’, is proportional to the mass of each molecule. MP reports the mass distribution of all the particles it detects in a sample, providing an overview of all the components present, with singlemolecule resolution. With over 1,300 peer-reviewed studies now citing MP, it is becoming an established technique for applications that benefit from rapid, label-free, singlemolecule mass distribution analysis, including the measurement of antibody aggregation and binding, adeno-associated virus (AAV) empty/full ratios, protein-protein interactions, and multi-attribute mRNA analysis. For a deeper understanding of how MP works, Read Refeyn’s handbook Understanding mass photometry Watch the webinar Mass photometry 101: A Deep Dive into Principles and Applications. As a single-particle technique, MP detects all the populations in a sample within its mass range, even those present in small quantities, allowing it to readily detect aggregation. Thanks to its quick measurements and low sample consumption, MP can be used to repeatedly test antibody samples for aggregation at-line – enabling timely, informed decisions. To support the analysis of large sets of samples – for example those resulting from forced degradation studies – Refeyn offers an automated instrument. Additionally, a custom software module streamlines the analysis of data from stability studies. A MP measurement detects, in a single measurement, free antigens and antibodies as well as their complexes, and quantifies their abundance. This information can be used to determine antibodies’ preferred stoichiometries and binding affinities. MP can also be applied to more complex samples, including bispecifics and multispecifics, with no additional preparation or method development. This makes MP ideal for work with large antibody libraries and new modalities. Aggregation detection Binding analysis i 5 www.refeyn.com Figure 2. MP measures the molecular mass distribution of samples at the single-particle level. It can be applied to a variety of biomolecule types, including including antibodies and other types of proteins. 4 How does mass photometry work? The particles which interferes with the light at the interface 2 Mass (kDa) Contrast The resulting interference contrast scales linearly with the particles’ mass 3 A mass histogram is generated from the single-particle measurements Particles, such as proteins, nucleic acids and viral vectors, on a glass surface are illuminated by a laser Incident light Particles in solution 4 1 Mass (kDa) Counts Contrast Scattered light Reected light Particles in solution 1 3 Particles in a sample on a glass surface are illuminated by a laser. The particles scatter light, which interferes with the light reected at the interface Mass (kDa) Contrast The resulting interference contrast scales linearly with the particles’ mass 3 A mass histogram is generated from the single-particle measurements Particles, such as proteins, nucleic acids and viral vectors, on a glass surface are illuminated by a laser Incident light Particles in solution 4 1 Mass (kDa) Counts Contrast Scattered light Reected light Particles in solution The resulting interference contrast scales linearly with the particles’ mass. 2 The particles scatter light, which interferes with the light reected at the interface 2 Mass (kDa) Contrast The resulting interference contrast scales linearly with the particles’ mass 3 A mass histogram is generated from the single-particle measurements Particles, such as proteins, nucleic acids and viral vectors, on a glass surface are illuminated by a laser Incident light Particles in solution 4 1 Mass (kDa) Counts Contrast Scattered light Reected light Particles in solution The particles scatter light, which interferes with the light reflected at the interface. 4 The particles scatter light, which interferes with the light reected at the interface 2 Mass (kDa) Contrast The resulting interference contrast scales linearly with the particles’ mass 3 A mass histogram is generated from the single-particle measurements Particles, such as proteins, nucleic acids and viral vectors, on glass surface are illuminated by a laser Incident light Particles in solution 4 1 Mass (kDa) Counts Contrast Scattered light Reected light Particles in solution MassFerence® P1 calibrant A mass histogram is generated from the single-particle measurements. TM 6 www.refeyn.com Antibody discovery and development steps DISCOVERY HIT GENERATION CANDIDATE SELECTION STABILITY TESTING What can mass photometry do...  Binding  Antigen quality check  Binding  Aggregation  Aggregation PROCESS DEVELOPMENT CANDIDATE CHARACTERIZATION 7 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Mass photometry: The perfect team player for antibody analytics Antibodies are powerful therapeutics used to treat a range of diseases, which is being expanded thanks to the development of new engineered formats like bispecifics. However, developing new antibodies comes with significant analytical challenges: • Antigens used for antibody development need to be characterized to ensure they are pure and have the desired oligomeric state • Antibody samples need to be analyzed to determine the functionality of their binding sites, their binding affinities and the stoichiometries in complex with their target antigen Current widely used techniques struggle with these analytical challenges. Techniques like SEC or AUC are informative, but they consume large amounts of sample that – particularly in the early stages of antibody development – can be precious. In addition, they work at high (μM) sample concentrations, far from the physiological conditions in which antibodies are meant to work. Other techniques like SPR and BLI are helpful for binding analyses, but they are bulk techniques that measure the average binding affinities of all species present in the sample. Often relying on the assumption that the antibody-antigen binding stoichiometry is 1:1, they struggle to characterize multicomponent systems like bispecific antibodies. Mass photometry (MP) is a powerful technique that provides an overview of all the components in a sample at the singleparticle level, in a matter of minutes and using only nanograms of sample. Moreover, MP measures take place in solution and at nanomolar concentration, closer to physiological conditions. These strengths make MP ideal to screen valuable antibody and antigen samples for purity, oligomerization behavior and complex formation. MP can also characterize antibody binding, provide an overview of antibody-antigen assemblies with multiple stoichiometries and quantify separate binding affinities for each complex in multicomponent systems. In short, the fast, informative measurements of MP can help you assess the quality of your samples and inform the results of other analytical techniques while saving time, sample and operating costs. Below, you can see two use cases for the technique: 1. MP evaluates the purity and oligomeric state of two antigen samples. 2. MP characterizes the complex binding of a bispecific antibody. 8 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Case 1: Quickly check your reagents In the SDS-PAGE data from their certificates of analysis, these two samples of VEGF165 antigen look similar and don’t show oligomerization. Case 2: Confirm 1:1 binding MP shows this bispecific antibody in mostly monomeric, with a few higher-order complexes. Supplier X Supplier Y When measured with MP, the VEGF165 from supplier X looks pure and stable, with a single peak at the expected mass. However, MP shows aggregation in supplier Y’s VEGF165 sample. Check your storage protocols or change suppliers. But when adding VEGF165, we see that binding isn’t 1:1. There are multiple complex stoichiometries, even before adding the second antigen. 9 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Quantifying protein binding affinities using mass photometry Protein interactions play pivotal roles in a wide range of biological processes. However, they are often highly complex, involving the binding of multivalent ligands to multiple binding sites. As a result, it can be challenging to identify the various complexes formed and assess the strength of their interactions.1 MP is a fast, label-free technology that measures the molecular mass of individual biomolecules in solution, and is ideal for studying protein-protein interactions. By providing the mass distribution of biomolecules in a sample, it gives a detailed overview of the binding partners present, the complexes they form, the relative abundance of each species and the strength of their interactions.1,2,3 Here, we apply MP to characterize interactions between Immunoglobulin G (IgG) antibodies of different origin species (human and bovine) and protein A. We quantify the relative abundance of each protein and the complexes they form in solution. We use the automated pipetting feature of Refeyn’s TwoMP Auto mass photometer to generate highly reproducible data. From these measurements, we calculate the equilibrium dissociation constant (KD) for each interaction. A MP measurement detects, in a single measurement, free antigens and antibodies as well as their complexes, and quantifies their abundance. This information can be used to determine antibodies’ preferred stoichiometries and binding affinities. MP can also be applied to more complex samples, including bispecifics and multispecifics, with no additional preparation or method development. This makes MP ideal for work with large antibody libraries and new modalities. Fig. 1 MP measurements of human and bovine IgG antibodies, and protein A. Both the human (blue) and bovine (orange) IgGs had a molecular mass of 150 kDa, while protein A (grey) had a mass of 42 kDa. MP measures sample homogeneity MP can be used to assess the purity and quality of control samples, ensuring that aggregation or protein oligomerization can be identified and taken into account when interpreting results. Here, MP measurements confirmed the expected mass of the two IgG antibodies, from human and bovine origin (~150 kDa for each), and protein A (~42 kDa)4 (Fig. 1). Mass photometry (MP) is a label-free bioanalytical tool that can assess the dynamics and strength of protein-protein interactions by directly measuring the mass and the relative abundance of individual proteins as well as the complexes they form in solution. 10 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT MP resolves complex equilibria MP can also be used to investigate binding and analyze complex formation. When the IgGs of differing origin were combined with protein A, MP revealed differences between the two IgGs in the formation of complexes and their relative abundance (Fig. 2). The IgGs were mixed with protein A in a 1:1 molar ratio and measured at concentrations of 5, 10 and 20 nM. At all concentrations measured, both IgGs formed 1:1 complexes with protein A, but differences emerged for higher-order complexes. Protein A interacted strongly with human IgG, as demonstrated by the formation of higher-order protein A: IgG heterocomplexes and the low percentage of free human IgG (33%). In contrast, protein A bound less readily to bovine IgG. In the case of bovine IgG, higher-order protein A: IgG complexes were not observed, and the majority of bovine IgG remained unbound (82%). The results show that MP can resolve and quantify the relative abundance of the multiple species that co-exist in complex equilibrium reactions. Fig. 2 MP reveals differences in complex formation between protein A and IgG antibodies of differing origin. IgGs of either human or bovine origin were mixed with protein A in a 1:1 molar ratio at 20 nM concentration. When human IgG was mixed with protein A (blue line), the following species could be resolved: Protein A (42 kDa); IgG (150 kDa); and protein A: IgG complexes of stoichiometry 1:1 (192 kDa), 1:2 (342 kDa), 2:3 (534 kDa) and 2:4 (684 kDa). When bovine IgG was mixed with protein A (orange line), the resolvable species were: Protein A (42 kDa), IgG (150 kDa), 1:1 protein A: IgG complex (192 kDa) and IgG dimers (300 kDa). MP measures interaction strength We observed above that protein A forms more complexes and higher-order complexes with the human IgG antibody as compared to the bovine IgG. This indicates a difference in affinity between the two interactions To quantify this difference, we calculated the KD for each interaction (Fig. 3) from MP measurements at 20 nM concentration (the data shown in Fig. 2 and two repeats). The calculations confirmed that protein A binds to the bovine IgG with lower affinity (KD = 69.4 ± 9.0 nM) than it binds to the human IgG (KD = 5.73 ± 2.6 nM). In addition, for the human IgG – protein A interaction, slightly lower KD values were observed for larger complexes, suggesting possible cooperativity effects involving IgG and/or protein A. While we repeated the experiment to improve the accuracy of the measurement, a single MP measurement can provide the data needed to calculate KD, even for complex equilibria. MP can be used to measure KD values for any interactions where both bound and unbound species are observable at the concentrations used in MP measurements. Fig. 3 Multiple KD values can be calculated from a single MP measurement. Schematics show the complex formation depicted in Fig. 2. KD values for interactions involving the IgG of human origin (blue) and of bovine origin (orange) are both shown. Since IgGs of bovine vs. human origin did not form all the same complexes, the KD calculation was not applicable for certain interactions (depicted as N/A). The KD values were calculated for each of three repeated MP measurements, and the mean ± standard deviation is given. 11 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Conclusion MP is a versatile, label-free bioanalytical tool that can be used to investigate complex equilibria at physiologically relevant concentrations. It can rapidly measure the mass of individual biomolecules and the complexes they form in solution, using very little sample. A single MP measurement can generate the data needed to calculate KD values, even for complex interactions that generate multiple species, as shown above. Refeyn’s TwoMP and TwoMP Auto mass photometers are both optimal for studying protein-protein interactions. In this experiment, data was generated using the TwoMP Auto, which offers automated pipetting for enhanced reproducibility. Methodological details • Protein A, bovine & human IgG antibodies were purchased from Sigma-Aldrich • Mixtures of IgG and Protein A (1:1 molar ratio) were incubated for 1 hour at RT prior to measurement & transferred to a 96-well plate, which was loaded onto the TwoMP Auto mass photometer • Measurements were performed within 40 min using PBS as the buffer for droplet dilution find focus • The KD values were calculated as previously described3 References 1 Wu and Piszczek, Anal Biochem 2020. https://doi.org/10.1016/j.ab.2020.113575 2 Wu and Piszczek, JoVE 2021. https://dx.doi.org/10.3791/61784 3 Soltermann et al., Angew Chem Int Ed 2020. https://doi.org/10.1002/anie.202001578 4 Yang, Biswas and Chen, Biophys J 2003. https://doi.org/10.1016/S0006-3495(03)74870-X 12 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Rapid analysis of bispecific antibody stability and target binding by mass photometry Analyzing bispecific antibodies can be challenging, as commonly used techniques struggle to provide information on their different binding sites. In this application note, we show how MP efficiently characterizes the purity, stability, and binding of multiple bispecific antibody candidates. Bispecific antibodies (bsAbs) are an emerging form of targeted immunotherapy that shows promise for treating a range of conditions. However, bsAb characterization presents significant challenges, as they are complex, artificially designed molecules that can be very variable in terms of structure, potency, immunogenicity, and the presence of aggregates or impurities. Among the main quality attributes that need to be measured are molecule size, fragmentation, aggregation and binding affinity. Characterizing bsAb binding affinities is especially challenging, as they have two different binding sites with their own binding affinities. These sites can cooperate or interfere with one another, and lead to the formation of higher-order assemblies. Conventionally used techniques for determining binding affinities include surface plasmon resonance (SPR) and biolayer interferometry (BLI). However, they are bulk techniques that do not accurately capture the contributions of all components in the sample. Also, because they tend to assume 1:1 binding, they struggle to fully characterize the complexity of bsAb interactions. MP is an emerging analytical technique that measures the mass of single particles in solution, in minutes and without labels. As it detects and quantifies all components in a sample and their masses, MP can characterize the following key attributes of bsAb samples with a single, quick measurement: • Presence and quantity of fragmentation, impurities and aggregates • Molecular mass of the bsAb – to compare with the expected mass • Higher-order interactions with target antigens (and their stoichiometry) • Binding affinities for each of the target antigens • Cooperativity or interference between bsAb components Table 1. Theoretical mass, mass measured with MP and binding sites of the bsAbs characterized in this application note. The agreement between the expected mass and the mass measured by MP (see also Fig. 1) shows that the bsAbs are properly assembled and that MP accurately measures their molecular masses. bsAB Theoretical mass (kDa) Measured mass (kDa) Binding sites HER2 75 – 105 83 n/a CD3 75 – 85 85 n/a OKT-3 150 154 2 CD3 Trastuzumab 150 153 2 HER2 bsAb-I 125 125 1 CD3, 1 HER2 bsAb-H 173 172 1 CD3, 2 HER2 bsAb-F 172 173 1 CD3, 2 HER2 bsAb-R 54 59 1 CD3, 1 HER2 bsAb-A 199 201 2 CD3, 2 HER2 To demonstrate the strengths and limitations of MP for bsAb analysis, we used it to analyze a panel of different bsAbs presenting HER2 and CD3 binding sites (Table 1). We show that MP readily informs on bsAb sample purity and binding functionality. In addition, we show that it can easily characterize high-affinity binding, and can also measure low-affinity binding with a previous crosslinking step or by using Refeyn’s MassFluidixTM HC microfluidics add-on. Analyzing bispecific antibodies can be challenging, as commonly used techniques struggle to provide information on their different binding sites. In this application note, we show how mass photometry (MP) efficiently characterizes the purity, stability, and binding of multiple bispecific antibody candidates. This application note was created in collaboration with: 13 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Evaluating mass, purity and aggregation Using MP, we measured a panel of bsAbs with different numbers of HER2 and CD3 binding sites (Fig. 1E-I). We also measured isolated HER2 and CD3 antigens to confirm their molecular masses (Figs. 1A and 1B) and two monoclonal antibodies as controls and (OKT-3, Fig. 1C; trastuzumab, Fig. 1D). Our MP measurements showed a single, narrow peak close to the expected molecular mass of each analyzed species (bsAbs, control mAbs and antigens, Table 1) – confirming the purity and stability of the samples. Fig.1 MP accurately measures the mass of monoclonal and bispecific antibodies and antigens. A) HER2 antigen. B) CD3 antigen. C) OKT-3 mAb. D) Trastuzumab mAb. E) bsAb-I. F) bsAb-H. G) bsAb-F. H) bsAb-R. I) bsAb-A. 14 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Quantifying high-affinity interactions MP operates at nanomolar concentrations, and it can readily detect and quantify high-affinity antigen binding. Here, we took the interaction with HER2 – with a dissociation constant (KD) of 0.5 nM for trastuzumab according to SPR1– as an example of a high-affinity interaction and measured its binding affinity to our antibody panel using MP. First, we measured bsAb-A (5 nM) mixed with HER2 at different concentrations (0.0, 2.5, 5.0, 10 and 20 nM) after reaching equilibrium (Fig. 2). As MP measures the mass of each molecule in a sample and counts the number of molecules with a given mass, it is possible to separately quantify the proportion of bsAbs bound to one or two HER2 antigens at each concentration and calculate the KD for each interaction (Table 2), as described before.2 We performed similar MP analysis with the rest of our antibody panel (Fig. 3) and quantified their binding affinities to HER2 (Table 2). For these measurements, both the antibody and HER2 antigen were present at a concentration of 5 nM. A Welch’s t-test found significant differences in the affinity for HER2 between trastuzumab and bsAb-A, bsAb-I and bsAb-R. There were no significant differences in the affinity for HER2 between trastuzumab and bsAb-F or bsAb-H. These measurements show that MP can readily determine site functionality and binding affinity for bsAbs. For bsAbs with one HER2 binding site, only 1:1 HER2-bsAb complexes were present, while in the case of bsAbs with two HER2 binding sites, only 1:1 and 2:1 complexes were present (Fig. 3). This suggests that there was no non-specific binding occurring at the CD3 binding sites of these antibodies. Finally, MP quickly distinguished and quantified single and double HER2 binding, revealing small differences in the binding affinities that may be helpful for optimizing protein design. KD ± SD (nM) Antibody Ab:HER2 (1:1) Welch’s t- test Ab:HER2 (1:2) Trastuzumab 1.00 ± 0.90 n/a 1.21 ± 0.57 bsAb-A 2.06 ± 0.67 0.020* 3.66 ± 1.31 bsAb-F 0.92 ± 0.16 0.824 2.09 ± 0.91 bsAb-H 0.40 ± 0.08 0.126 0.85 ± 0.34 bsAb-I 0.25 ± 0.21 0.004* n/a bsAb-R 0.29 ± 0.13 0.040* n/a Table 2. Binding affinities of HER2 to the different bsAbs. Dissociation constant values were calculated from equilibrium measurements with MP (Figs. 2-3). Because MP can resolve both 1:1 and 2:1 binding, the KD for each interaction can be calculated separately. Both the antibody and HER2 antigen were present at a concentration of 5 nM. The bsAb-I and bsAb-R antibodies have a single HER2 binding site, so the KD for the 2xHER2 complex cannot be calculated. The middle row shows the p value of a Welch’s t-test for the binding affinity of bsAb:HER2 (1:1) against Trastuzumab:HER2 (1:1). An asterisk indicates a satistically significant difference. Fig. 2 MP resolves complex bsAb-antigen interactions. The concentration of bsAb-A was kept constant at 5 nM, while the HER2 concentration was varied (0.0, 2.5, 5.0, 10 and 20 nM). MP histograms show peaks and corresponding counts for each individual species as well as 1:1 HER2-bsAb complexes and 2:1 HER2-bsAb complexes. As the HER2 concentration increases, the peaks corresponding to free antigen and the 2:1 complex become more prominent. 15 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Fig. 3 MP characterizes HER2 binding for bsAbs with different numbers of HER2 binding sites. A) Trastuzumab mAb + HER2. B) bsAb-I + HER2. C) bsAb-H + HER2. D) bsAb-F + HER2. E) bsAb-R + HER2. F) bsAb-A + HER2. For all Abs and HER2, concentration was 5 nM. Quantifying high-affinity interactions Standard MP measures samples at nanomolar concentration, which makes it challenging to study low-affinity interactions. This can be seen when we measure the interaction between the OKT-3 monoclonal antibody and its CD3 antigen. First, we incubated both species at μM concentrations to establish equilibrium, and then diluted the sample to the nM concentration required by MP. In these conditions, OKT-3 and CD3 were dissociated (Fig. 4A). To measure low-affinity interactions with MP, Refeyn offers MassFluidix HC – a microfluidics add-on to the TwoMP mass photometer. This system can perform a rapid dilution in only 37 milliseconds, so the MP measurement is performed before the reaction has shifted significantly from its higher-concentration equilibrium state. We tested this approach to measure CD3:OKT-3 interactions by incubating the sample at μM concentration until equilibrium was reached, then measuring it after rapid dilution with MassFluidix HC (Fig. 4B). Under these conditions, CD3:OKT-3 complexes were visible, with 1:1 complexes representing 36.3% of counts and 2:1 complexes representing 11.7%. Next, we compared the MassFluidix HC measurement results against a more traditional crosslinking approach by using a rapid crosslinking protocol.3 OKT-3 and CD3 were incubated at micromolar (μM) concentration for 10 minutes, at a 1:4 ratio (CD3:OKT-3). We then crosslinked the sample with disuccinimidyl dibutyric urea (DBSU), incubated the reaction for 45 minutes and proceeded with dilution to nanomolar concentration. The MP measurement showed that 50.8% of the counts corresponded to 1:1 CD3:OKT-3 complexes and 23.6% to 2:1 complexes (Fig. 4C), indicating that crosslinking preserved the weak interactions between OKT-3 and CD3 after dilution. A small mass increase could be seen on all components of the crosslinked sample when compared to the MassFluidix HC measurement due to the presence of the crosslinker. Fig. 4 MP can be used to characterize low-affinity interactions. A) Measurement of a mix of OKT-3 (1 μM) and CD3 (5 μM) diluted to nM concentration. B) Mix of OKT-3 (1 μM) and CD3 (4 μM) measured after a fast 2000x dilution with MassFluidix HC. C) Measurement of the same OKT- 3 and CD3 sample diluted to nM concentration after a crosslinking step. 16 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Fig. 5 MP captures low-affinity Ab-CD3 interactions with MassFluidix HC. For each antibody, the top histogram shows a standard MP measurement, the middle histogram shows a measurement taken with MassFluidix HC and the bottom one shows a measurement of a crosslinked sample. A) bsAb-I + CD3. B) bsAb-H + CD3. C) bsAb-F + CD3. D) bsAb-R + CD3. E) bsAb-A + CD3. Inset: Zooming in on the measurement of the bsAb-A + CD3 crosslinked sample shows that both 1:1 and 2:1 complexes are visible. 17 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT To test CD3 binding in our panel of antibodies, we applied the MassFluidix HC approach to measure mixtures of each with CD3 (Fig. 5). No CD3:Ab complexes were visible with standard MP alone, while the MassFluidix HC MP measurements showed the expected CD3:Ab complexes. As expected, measurements of antibodies with two binding sites (OTK3 and bsAb-A) showed 1:1 and 2:1 antigen:antibody complexes. For antibodies with only a single binding site, only 1:1 complexes were detected. These results show that MP – in combination with either MassFluidix HC or crosslinking – can characterize low-affinity interactions between antibodies and their antigens. With this type of analysis, binding site functionality can be confirmed. However, with neither approach can the sample be assumed to be at equilibrium, so they should not be used to calculate KD values. Observing double binding in bsAbs Next, we explore whether MP and crosslinking can be used to characterize bispecific antibodies with two targets of different affinities. One example is bsAb-I, which has one HER2 binding site (with higher affinity) and one CD3 binding site (lower affinity). We used MP to analyze a sample of bsAb-I (1 μM), HER2 (2 μM) and CD3 (4 μM), with and without crosslinking – following the procedure described in the previous section. Without crosslinking, only HER2:bsAb-I complexes were visible after diluting the sample to the nM concentrations required for MP. However, when a crosslinking step was performed before dilution, the bsAb- I:HER2:CD3 complex was also detected (Fig. 6). Conclusions MP is an ideal, all-in-one technique for bispecific antibody characterization. In a matter of minutes and consuming very little sample, it can detect fragmentation, assess purity and determine if the measured antibody has the expected mass – providing straightforward information into the success of protein design and purification processes. Furthermore, MP can be used for binding studies, quickly determining binding functionality, complex stoichiometry and the KD for high-affinity bsAb-antigen interactions. Lower-affinity interactions can be detected by using the MassFluidix HC rapid dilution add-on or a crosslinking step. As an additional benefit, the single-molecule nature and high mass resolution of MP make it possible to observe and measure the binding affinities and stoichiometries of antibodies with multiple targets. Widely used techniques like biolayer interferometry (BLI) or surface plasmon resonance (SPR) are powerful when studying 1:1 antibody-antigen interactions but, as they measure interactions in terms of the sample average, do not give the full picture when dealing with the more complex binding behavior of bsAbs. MP, on the other hand, gives an overview of all the different species and complexes in a sample and their relative abundances, characterizing the binding affinities of multiple binding sites with a single measurement.4 In addition, MP measures samples in minutes, consumes only nanograms of sample and is very user-friendly – needing only a simple dilution step as sample preparation. These practical advantages make it an ideal technique for repeated, in-house analysis of bsAb binding as well as other critical quality attributes like aggregation.5 Fig. 6 Complex formation of a bispecific antibody with two targets of varying affinities can be detected through MP and crosslinking. Top: Control MP measurement without crosslinking. Here, only the high-affinity HER2:bsAb-I complex is observed. Bottom: MP measurement of the same sample after crosslinking, showing the HER2:CD3:bsAb-I ternary complex. References 1 Bostrom et al., PLoS one. 2011. https://doi.org/10.1371/journal.pone.0017887 2 Soltermann et al., Angew. Chem. Int. Ed. 2020. https://doi.org/10.1002/anie.202001578 3 Gizardin-Fredon et al., Nat. Comm. 2024. https://doi.org/10.1038/s41467-024-47732-4 4 Wu and Piszczek, Anal. Biochem. 2020. https://doi.org/10.1016/j.ab.2020.113575 5 Rapid, reliable antibody aggregation analysis with MP https://refeyn.com/mass-photometry-antibody-aggregation 18 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Materials and methods Reagents • DMSO buffer, cat. No: 855190 ThermoFisher Scientific, LC-MS grade • DSBU (disuccinimidyl dibutyric urea, BuUrBu), ThermoFisher Scientific, cat. no. A33549 • 1M Tris HCl pH8.0, ThermoFisher Scientific, J22638.AE • MassGlassTM UC slides, Refeyn Ltd. Antibodies • Anti-CD3e, OKT-3, 1 mg/mL, mouse IgG2a, Ab00122-2.0, T1528802 • Anti-erbB2, Trastuzumab, 1 mg/mL, human IgG1, Ab00103-10.0, T1818A54 • HER2 from AcroBiosystems, Human Hert2/ErbB2 protein, His-Tag, cat. no. HE2-H5225 • Human CD3 epsilon&CD3 delta Heterodimer Protein, Fc Tag&Fc Tag, cat. no. CDD-H5225 • HER2-hOKT3 bsAb-A, 25 - IgG-scFv (HC C-term), cat. no. bAb0515 • HER2-hOKT3 bsAb-F, 95 - Heterodimeric IgG-scFv (type 4), cat. no. bAb0520 • HER2-hOKT3 bsAb-H, 167 - Heterodimeric Fab-scFv-Fc (type 3), cat. no. bAb0522 • HER2-hOKT3 bsAb-I, 41 - Heterodimeric Fab/scFv-Fc, cat. no. bAb0523 • HER2-hOKT3 bsAb-R, 44 - Tandem scFv, cat. no. bAb0532 All samples were provided by Absolute Antibody. Any requests for samples can be directed to [email protected] Crosslinking methodology For analyzing bsAb-HER2 complexes, bsAb and HER2 were incubated at room temperature for 45 min at nM concentrations. Chemical crosslinking and MP measurements were performed according to Gizardin-Fredon et al.3 This rapid crosslinking protocol enables detection of the low-affinity binding complex under just one hour. For bsAb or OKT-3 and CD3, samples were incubated at micromolar (μM) concentration, at 1:4 or 1:5 molar ratio, respectively. The antibody and ligand mix was incubated 10 min at room temperature. Samples were then crosslinked with disuccinimidyl dibutyric urea (DBSU) in 400x molar excess. The optimal crosslinking time is 45 min at room temperature. The reaction was quenched by adding 25 mM Tris, for 5 min. Finally, samples were diluted to nM concentration for MP measurements. 19 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Rapidly assessing reagent quality before binding studies with mass photometry Binding studies require careful assessment of reagent quality. In the case of multispecifics, binding stoichiometry needs to be determined to interpret the results of BLI or SPR analyses. Mass photometry (MP) provides fast, single-particle insights on antigen purity and stability, making it an ideal orthogonal technique for binding studies of a broad range of antibody modalities. Antibodies are powerful therapeutics used to treat a range of diseases, and that range is growing thanks to the development of new engineered formats such as bispecifics. However, antibody development comes with significant analytical challenges: • Antigens used for antibody development need to be characterized to ensure they are pure and have the desired oligomeric state. • Antibody samples need to be analyzed to determine the functionality of their binding sites, their binding affinities, and stoichiometries in complex with target antigen(s). Current widely used techniques struggle with these analytical challenges. Size-exclusion chromatography (SEC) and analytical ultracentrifugation (AUC) are informative but consume large amounts of sample that – particularly in the early stages of antibody development – can be precious. In addition, they work at high (μM) sample concentrations, far from physiological conditions. Other techniques, such as surface plasmon resonance (SPR) and biolayer interferometry (BLI), are standard tools for binding analyses, but they are bulk techniques that measure the average binding affinities of all species present in the sample. These techniques typically assume 1:1 binding for model fitting when studying monovalent antibodies. Analyzing multispecific antibodies or oligomeric antigens requires a different model to fit the data, and model selection requires knowledge of the oligomeric state and stoichiometry of the species in the sample. MP provides an overview of all the components in a sample at the single-particle level, in a matter of minutes and using only nanograms of sample. Moreover, MP measurements take place in solution and at nanomolar concentration, closer to physiological conditions. These strengths make MP ideal to screen valuable antibody and antigen samples for purity, oligomerization behavior and complex formation. MP can also characterize antibody binding, provide an overview of antibody-antigen assemblies with multiple stoichiometries, and quantify separate binding affinities for each complex in multicomponent systems. Here, we show two case studies that illustrate how MP reveals key information about antibody and antigen samples before dedicating time and sample to other techniques. 20 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Fig. 1 Evaluating antigen purity and stability with MP. Panels A and B show analysis data from two different suppliers. Insets: SDS-PAGE results from the antigens’ certificates of analysis. The left column shows a mass reference, while the right column shows monodisperse, denatured VEGF165 monomers. Histograms: MP measurements of the two VEGF antigens. Measured on the TwoMP in PBS. Case 2: Binding assessment In this experiment, we used MP to measure a bispecific antibody by itself and mixed with VEGF165 – one of its targets (Fig. 2). The measurements of the isolated antibody showed a mostly monomeric distribution with a few higher-order oligomers Fig. 2 MP reveals multiple binding stoichiometries. MP histograms corresponding to MP measurements of (A) an isolated bispecific antibody and (B) the same bispecific antibody mixed with VEGF165. The x-axis displays the mass of the counted particles in kDa, while the Y-axis displays the total counts for each mass bin. Measured on the TwoMP. Case 1: Antigen characterization As MP quickly detects all populations in a sample using little sample, it is ideal to make sure antigens are pure and stable before further analysis. In this case study, we analyzed two VEGF165 antigen samples from two suppliers (Fig. 1). In the data from their certificates of analysis, these two samples of VEGF165 antigen from different suppliers look similar and do not show oligomerization (Fig1 A and B, insets). When measured with MP in PBS, the VEGF165 from supplier A (Fig. 1A) looks pure and stable, as the MP measurement shows a single population of VEGF165 homodimers – the natural form of the antigen. In contrast, the sample from supplier B (Fig. 1B) shows aggregates of different sizes, indicating that the sample may have degraded during storage or delivery. (Fig. 2A). When measured in conjunction with VEGF165 (Fig. 2B), the results reveal multiple antigen-antibody binding stoichiometries. Knowing that the bispecific antibody binds its target in multiple stoichiometries helps interpret further SPR or BLI analyses. 21 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Comparing mass photometry and SEC for antibody aggregation assessment MP is an emerging tool for assessing multiple antibody attributes, including aggregation. Here, a comparison to size-exclusion chromatography (SEC) shows the two techniques agree, but MP has superior resolution, particularly for larger modalities. MP requires no optimization beyond sample dilution and readily characterized samples that could not initially be measured with SEC. In addition, MP requires >100x less sample and is 20x faster than SEC, and provides an informative mass readout. Antibody characterization is essential in biopharmaceutical development for ensuring the safety, efficacy and consistency of the therapeutic product. One important critical quality attribute (CQA) investigated is aggregation. Aggregation can affect the final product in three ways: 1. Increased immunogenicity risk, as aggregated antibodies are more likely to be seen as ‘foreign’ by the immune system. 2. Reduced efficacy, as aggregates may be inactive or interfere with the function of the monomeric antibody. 3. Impaired stability, as aggregation can indicate or lead to protein instability over time, affecting shelf life and formulation robustness. A powerful and widely accepted method for detecting and monitoring aggregation during development and release testing is size exclusion chromatography on a high-performance liquid chromatography setup (SEC-HPLC) coupled to UV detection (simply denoted as SEC in this article). SEC can resolve monomers, dimers and higher-order aggregates of antibodies in a single run. However, it frequently requires time-consuming column optimization, as some antibodies may stick to a column matrix or interact in unexpected ways, increasing the retention time and complicating the interpretation of traces. In addition, SEC is biased towards higher-mass species, as those with a greater mass generate a larger signal. MP is an emerging tool for characterizing multiple antibody CQAs, including aggregation. As a single-molecule technique, MP can detect and quantify low abundance populations. Its high resolution (25 kDa for a 66 kDa protein) enables population differentiation. It can detect antibody fragments, monomers, dimers and higher-order aggregates – and its mass readout makes it straightforward to identify the different populations detected. MP can also analyze antibody binding to targets. Measurements take just minutes and require only nanograms of sample. Here, we provide a detailed comparison of the techniques in the context of antibody aggregation analysis. We characterize a series of antibodies with SEC and MP. Samples were only selected for inclusion in this study if SDSPAGE analysis showed the intact antibody under non-reducing conditions, with disassembly into heavy and light chains under reducing conditions. SEC and MP comparison experiments were carried out side-by-side using the same samples and illustrate different cases for how SEC and MP compare. We show that: 1. MP and SEC results generally agree, although they are not directly comparable due to differences in what they measure. MP provides a mass readout. 2. MP, unlike SEC, does not require prior optimization and cannot be affected by column interactions. 3. For larger modalities, MP has superior resolution to SEC and provides much more detailed insights about sample composition. 4. Standard MP is run at a lower concentration than SEC, which could affect concentration-dependent multimers. This can be overcome by combining MP with the rapiddilution MassFluidix HC microfluidics system. This application note was created in collaboration with: Mass photometry (MP) is an emerging tool for assessing multiple antibody attributes, including aggregation. Here, a comparison to size-exclusion chromatography (SEC) shows the two techniques agree, but MP has superior resolution, particularly for larger modalities. MP requires no optimization beyond sample dilution and readily characterized samples that could not initially be measured with SEC. In addition, MP requires >100x less sample and is 20x faster than SEC, and provides an informative mass readout. 22 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Case 1: SEC and MP agree We first measured a monoclonal IgG antibody, IgG1 Human 1. The results from SEC generally aligned with MP, but the superior resolution of MP enabled detection of a fragment population. The SEC results showed a main peak at around 7.5 min, as expected, as well as higher-order structures that eluted prior to the main peak (Fig. 1, Table 1). Smaller populations also eluted around 12–18 minutes, but the peaks were too small to be fitted. This result agreed with SDS-PAGE, which showed one major band and a very faint second band at higher molecular weight under nonreducing conditions (Fig. 1). MP confirmed that the main species was a ~150 kDa molecule, as expected for a monoclonal antibody. MP further identified populations of probable dimers, with mass ~300 kDa, and a very small population of trimers (Table 1). Though the dimers and trimers had very low abundance relative to the monomer, they were detectable. The dimer and trimer peaks were more visible with SEC, in line with the mass weighting that occurs with this technique (see Box 1). Overall, quantification showed the two techniques were in general agreement about the relative sizes of the monomer, dimer and trimer populations (Table 1). However, MP also clearly resolved an additional population around 40 kDa that SEC did not detect. This population is likely fragmentation, based on the mass. It is a striking difference between the two techniques that this population, which may have considerable importance for understanding the sample’s stability, was only detected by MP and not SEC. Table 1. Proportions of species, as determined by SEC and MP for the monoclonal IgG antibody IgG1 Human 1 in Fig.1. Oligomeric state SEC measurement (%) Mass photometry measurement (%) IgG Human 1 Fragment 0.0 5.4 Monomer 86.4 89.8 Dimer 10.8 4.1 Trimer 2.8 0.7 Fig. 1 SEC and MP results agree for analysis of two monoclonal IgG antibodies. For antibody IgG1 Human 1, results are shown for SDS-PAGE under nonreducing (NR) and reducing (R) conditions (left), SEC analysis (middle) and MP (right). For SEC and MP, the colored peak regions correspond to antibody fragments (purple), monomers (blue), dimers (yellow) and trimers (green). 23 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Case 2: SEC ambiguous, MP clear and consistent Next, a different IgG antibody, IgG Rabbit 1, was analyzed by SEC and MP. In this case, in the SEC analysis, the retention time for this antibody was ~9 min (Fig. 2), whereas the expected retention time was ~7.5 min under the column conditions used. The slower elution time would suggest that the proteins in the sample were in fact smaller than full IgGs (e.g. fragments). The MP result, on the other hand, produced a main peak around 150 kDa, consistent with the sample containing mainly full, monomeric IgG antibodies, as expected. MP further identified a population of dimers, with mass ~300 kDa and even a very small trimer population (Table 2). The unexpected SEC result could have been caused by temporary retention of the molecules in the column. Without additional information from an orthogonal technique, it would be unclear from the SEC data whether the protein was the expected one (but slightly delayed due to column interactions), or in fact had fragmented into a lower-mass species. This example illustrates the advantage that MP offers by eliminating the need for a column. It also demonstrates the value of analysis by orthogonal techniques. Table 2. Proportions of species, as determined by SEC and MP for the monoclonal IgG antibody IgG Rabbit 1 in Fig. 2. Oligomeric state SEC measurement (%) Mass photometry measurement (%) IgG Rabbit 1 Monomer 99.1 97.4 Dimer 0.9 2.2 Trimer 0.0 0.2 Fig. 2 SEC suggested possible fragmentation or column interactions, while MP confirmed the presence of a monomeric full IgG population. For an IgG rabbit antibody, results are shown for SDS-PAGE under non-reducing (NR) and reducing (R) conditions (left), SEC analysis (middle) and MP (right). For SEC and MP, the colored peak regions correspond to monomers (blue) and dimers (light orange). 24 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Case 3: SEC fails, while MP works We next analyzed another human antibody, IgG1 Human 2 (Fig. 3A). In this case, the SDS-PAGE results clearly indicated the presence of an antibody, but the SEC trace provided negligible information – despite being run under identical conditions to the antibodies shown in Cases 1 and 2. This result is likely due to the antibody having been retained within the column matrix – a common issue in SEC that can take considerable time to resolve. Conversely, MP provided a clear readout of the sample’s aggregation state, and with no optimization required. A mouse fab fragment (Fab Mouse 2) also produced inconclusive results in the SEC analysis, but again with a clear MP readout (Fig. 3B). The MP data showed a main peak at ~80 kDa as well as a multitude of higher-order species present with low abundance, in agreement with SDS-PAGE, which showed a smear in nonreducing conditions and many different sized bands under reducing conditions. This could explain the poor SEC results, as the higherorder species could have clogged the column, resulting in the antibody’s retention. Fig. 3 MP, but not SEC, returned clear aggregation readouts for an IgG antibody and a mouse fragment. For antibodies (A) IgG1 Human 2 and (B) Fab Mouse 2, results are shown for SDS-PAGE run under non-reducing (NR) and reducing (R) conditions (left), SEC analysis (middle) and MP (right). For MP, the colored peak regions correspond to monomers (blue) and higher-order species (light orange). 25 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Case 4: Only MP could resolve higher-order IgM oligomers Finally, we analyzed samples of larger antibodies, IgMs, which feature a more complex structure than IgG antibodies, forming pentamers or hexamers. Hexameric IgM samples, IgM Human 1 and IgM Human 2, were obtained by expressing IgM without the J-chain. In both cases, SDS-PAGE under reducing conditions showed the expected heavy and light chains (Fig. 4). In both cases, in the SEC trace, three peaks were distinguishable in the expected range (Fig. 4AB, mid left). More peaks were visible from the MP analysis, and the mass information from MP makes it possible to identify the species. In addition to the fully assembled hexameric IgM, MP revealed significant populations of pentamers, tetramers, dimers and monomers (Fig. 4AB, mid right). The MP results suggest that the major eluting peaks from the SEC traces are associated with at least three species: Tetramers, pentamers and hexamers. The observation that SEC does not always resolve all oligomeric states in IgM aggregation has been made before, such as in comparison to data from analytical ultracentrifugation.1 Overall, it was only possible with MP to identify and quantify the range of oligomeric species in each sample (Fig. 4). MP requires that samples be at low (nanomolar) concentrations for measurements. The low concentrations can affect the formation of multimers, such as the IgM assemblies here, which form reversibly. Because SEC and MP require different concentrations for measurement, they may report different oligomeric distributions of the IgM species. We can overcome MP’s low-concentration requirement by using a rapid-dilution microfluidic system (Refeyn’s MassFluidix HC) coupled to the mass photometer. The MassFluidix approach allows samples to be diluted up to 10,000x immediately prior to measurement, before significant dissociation can occur, so the observed distribution of species reflects what was present in the sample at the higher concentration. To assess the effect of concentration on the oligomeric distribution, we tested the two IgM samples using the MassFluidix MP approach (Fig. 4AB, far right). The MassFluidix measurements verified that the additional species observed (monomers-pentamers) were also present at higher concentrations, so they were not due to the low concentration used for standard MP. However, the higherorder oligomers were present in higher proportions than in the standard MP measurement, indicating that some IgM hexamer disassembly occurs when the sample is diluted. Fig. 4 IgM pentamers and hexamers were only resolved by MP. For IgM antibodies (A) IgM Human 1 and (B) IgM Human 2, results are shown for SDS-PAGE run under reducing (R) conditions (far left), SEC analysis (mid left), standard MP (mid right) and MP with MassFluidix HC (far right). The colored peak regions correspond to monomers (blue), dimers (light orange) and higher-order species (green). In (B), an additional population eluted around 14 min, likely due to buffer contaminants. It was not detected by MP, suggesting the species had mass below the 30 kDa minimum for MP analysis. 26 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Discussion Here, we have shown that MP is an effective tool for antibody aggregation analysis that outperforms SEC due to having superior resolution, providing an informative mass readout, and eliminating both the need for optimization and the risk of column interactions. MP also has the advantages of being fast (with a measurement taking only 1 min, vs. 20–25 min for SEC) and using >100x less sample. A dedicated Antibody Stability software module2 further facilitates MP analysis of antibody aggregation. Differences between the methods are detailed in Box 1 and summarized in Table 3. Another important difference between SEC and MP is the concentration at which samples are measured. MP requires a low (nanomolar) concentration and small volume (10–20 μL). While MP’s low sample consumption is beneficial, the low concentration can cause the dissociation of higher-order species. Here, we have shown that this issue can be overcome by using the MassFluidix HC rapid dilution approach with MP. This rapid dilution approach could also be used to assess whether measured aggregates are dissociable or not. In practice, both SEC and MP are valuable tools for antibody aggregation analysis and can be used in complementary ways (Table 3). SEC is ideal for high-throughput screening and quality control (QC) in GMP-regulated environments (as there is not yet a 21CFR11-compliant MP solution for antibody aggregation analysis). MP, meanwhile, is ideal for rapid screening and situations where sample is limited, and its superior resolution is valuable for analysing larger modalities. The availability of the MassFluidix HC rapid dilution system means that MP is not limited to lowconcentration samples. References 1 Chouquet et al., Front. Bioeng. Biotechnol. 2022. https://doi.org/10.3389/fbioe.2022.816275 2 Refeyn product datasheet: Antibody Stability Module https://info.refeyn.com/antibody-stability-module 3 Refeyn tech note: Assessing proteins samples by MP & SEC https://refeyn.com/mass-photometry-vs-sec-in-protein-analysis Samples All samples were provided by Absolute Antibody. Any requests for samples can be directed to [email protected]. MP All MP measurements were carried out on a Refeyn TwoMP mass photometer. All measurements were acquired using AcquireMP 2024.1.1 and analyzed with DiscoverMP 2024.1.0. All measurements were carried out using the large field of view setting and each movie was recorded for 60 seconds. Calibrations, to enable conversion from the measured contrast to mass, were carried out using Refeyn’s MassFerenceTM P1 calibrant. Prior to measurement, samples were diluted from 1 mg/mL (approximately 6.6 μM for a 150 kDa protein) to 10 nM using PBS. MP with MassFluidix HC A Refeyn MassFluidix HC system was used with a Refeyn TwoMP mass photometer. IgM samples were at 1 mg/mL (~0.92 μM, for hexameric IgM). Prior to loading, samples were diluted 2x to 0.55μM, followed by rapid 1,000x dilution in the microfluidic chip. SEC-HPLC with UV detection Size-exclusion chromatography HPLC with UV detection was carried out using an Agilent 1100 series HPLC with a Superdex 200 Increase 5/150 GL column equilibrated in PBS, pH 7.2, at a flow rate of 0.2 mL/min. Proteins were analysed following concentration to 1 mg/mL by injecting 5 μL of sample. Absorbance was measured at 280 nm and proteins assessed for percentage purity (peak area) and size (peak retention time). SDS-PAGE Qualitative protein analysis by SDS-PAGE was performed on samples concentrated to 1 mg/mL. Samples were loaded to pre-cast gels (10% Bis-Tris) from Invitrogen™. Gels were run in MES at 180 V for 45 minutes before staining with Invitrogen SimplyBlue™ SafeStain and destained for 16 hours in deionized water. Gel images were captured using a Bio-Rad GS-900 densitometer. Materials and methods 27 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT SEC and MP operate in very different ways and these differences must be taken into consideration when comparing data from the two methods. We explain differences in key areas below and present an overall comparison (Table 3), highlighting differences related to sample requirements, resolution and other aspects. The measurement process SEC separates molecules based on their size by passing them through a column packed with porous beads. Large molecules are excluded from entering the pores and elute first, while smaller molecules enter the pores and take longer to pass through, eluting later. UV absorbance (at 280 nm) is then measured by a detector as the sample elutes. MP, meanwhile, measures the mass of single proteins in solution by measuring the interference between light scattered by individual proteins and light reflected at the measurement surface. The resulting contrast scales linear with the proteins’ mass. Data presentation SEC data are typically represented in a graph showing UV absorbance vs. elution time (or elution volume). As larger molecules elute first, the first peak in the graph represents the largest molecule and the last peak the smallest. This is opposite to MP, where data are represented in a mass histogram (counts vs. mass) with the first peak corresponding to the lowest-mass species. An advantage of MP is that, provided a mass-contrast calibration step has been done, the molecular mass of the species in each peak is immediately apparent. This makes it possible to identify which species is most likely present in each peak (e.g. the dimer peak would be the one with a mean around the expected mass value of a dimer). Data interpretation Another important difference relates to what peak area and height represent. In MP, the area under the peak corresponds to the relative number of proteins in that population (with that mass value). For example, a population with a peak around twice as large as another would have about twice the concentration. Box 1: How do SEC and MP compare? Generally, in MP, the proportions of species represented in the histogram counts are the same as their concentration in solution, so relative concentrations can be assessed directly from peak sizes. In SEC, by contrast, the amount of UV light absorbed by a protein depends on its extinction coefficient – which is a function of the number of aromatic (particularly tryptophan) residues in its sequence. A dimeric molecule, for example, would contain twice as many aromatic residues and so absorb roughly double the UV light, giving twice the signal of a monomer. Therefore, in SEC, the signal from a given species depends on its mass as well as its amino acid composition. The need to factor in extinction coefficients when interpreting SEC data is particularly important in the analysis of samples such as antibody-drug conjugates (ADCs), where the variability in the extinction coefficients of different species may be significant. SEC and MP data are not directly comparable Due to the differences between the techniques, MP and SEC data are not directly comparable. To compare them quantitatively, it is necessary to use the extinction coefficients of the species present in a SEC profile to convert the data so that the signal is purely based on concentration, as it is in MP. For a demonstration of the conversion, see Refeyn’s recent technical note, Assessing protein samples by MP and SEC.3 However, when SEC peaks contain multiple different species (which may occur due to SEC’s limited resolution for larger species), normalization by extinction coefficients may not be possible. Additional considerations Unlike SEC, which separates based on size and may dilute or disturb equilibrium, MP measures the native distribution of species in near-physiological conditions. Other advantages of MP are its versatility with respect to sample type, conditions and buffer, low sample consumption and rapid measurement time. For antibody aggregation studies using MP, samples only need to be diluted to appropriate concentrations (nanomolar) and then can be measured in just one minute, giving a clear read-out on aggregation state. This makes MP particularly useful for rapid aggregation profiling, and for stability studies in early and late-stage biopharma development. 28 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Mass photometry Size-exclusion chromatography Separation mode None Size-based separation by a column Readout Molecule counts over molecular mass UV protein absorbance over retention time Sample requirement ~38 ng (10 μL of 10–55 nM) 10 μg for measurements with MassFluidix HC ~5 μg (5 μL of 1 mg/mL) Labeling Label-free Label-free Resolution Detects fragments, monomers and all higherorder multimers with baseline resolution Depends on column, with molecular weight range and operational limitations Run time 1 min 20–25 min Operational costs <$5 per sample <$5 per sample Ease of data interpretation Straightforward, direct mass readout with particle counts Cumbersome, requiring conversion based on extinction coefficients Limitations Non-preparative, Mass detection range ~30 kDa to ~5 MDa, Sample concentration limited to <50 μM Non-specific column interactions, Column exclusion limits for larger aggregates (typically ~0.6 MDa), Sensitive to extinction coefficients Ideal use cases Rapid, native-state profiling, Measurement in physiological conditions (i.e. native buffer, nM concentration) Fractionation with quantitation, High-throughput analysis, GMP-regulated environments 20 Testimonials “Mass photometry provides a fast screening tool to investigate mRNA integrity and size.” De Vos et al. (2024), J Chromatogr A “The data confirm the great potential of [mass photometry] technology... as a fast and simple orthogonal method that provides insights into the homogeneity and stability of mRNA samples.” Camperi et al. (2024), Anal Chem Unit 9, Trade City, Sandy Lane West, Oxford OX4 6FF, United Kingdom ©2024 Refeyn Ltd information on products, demos and ordering, write to [email protected] Samux and Refeyn are registered trademarks of Refeyn Ltd. refeyn.com @refeynit Refeyn Refeyn About Refeyn Refeyn pioneers analytical instruments that put molecular mass measurement capabilities within easy reach for scientists. Refeyn’s unique products measure the mass of individual proteins, nucleic acids, complexes and viruses directly in solution – providing vital insights for scientific discovery, R&D and therapeutics production. Our instruments feature mass photometry technology, which uses light to quantify the mass of single particles in solution without labels, and macro mass photometry technology, which uses light to characterize large viral vectors. Providing intuitive data in minutes, mass photometry technologies help scientists solve their research questions, optimize R&D processes and focus on innovation. 20 Testimonials “Mass photometry provides a fast screening tool to investigate mRNA integrity and size.” De Vos et al. (2024), J Chromatogr A “The data confirm the great potential of [mass photometry] technology... as a fast and simple orthogonal method that provides insights into the homogeneity and stability of mRNA samples.” Camperi et al. (2024), Anal Chem Unit 9, Trade City, Sandy Lane West, Oxford OX4 6FF, United Kingdom ©2024 Refeyn Ltd information on products, demos and ordering, write to [email protected] Samux and Refeyn are registered trademarks of Refeyn Ltd. refeyn.com @refeynit Refeyn Refeyn About Refeyn Refeyn pioneers analytical instruments that put molecular mass measurement capabilities within easy reach for scientists. Refeyn’s unique products measure the mass of individual proteins, nucleic acids, complexes and viruses directly in solution – providing vital insights for scientific discovery, R&D and therapeutics production. Our instruments feature mass photometry technology, which uses light to quantify the mass of single particles in solution without labels, and macro mass photometry technology, which uses light to characterize large viral vectors. Providing intuitive data in minutes, mass photometry technologies help scientists solve their research questions, optimize R&D processes and focus on innovation. Table 3. A comparison of SEC vs. MP 29 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Assessing protein samples by mass photometry and size-exclusion chromatography MP and size exclusion chromatography (SEC) are complementary analytical tools that can provide users with a wealth of information regarding the biochemical and biophysical properties of biomolecules within a sample. However, when comparing datasets captured by the two techniques, there are some important considerations to be aware of. In this technical note, these considerations are explored. MP and SEC data are compared in two case studies. The first case study explores how the two techniques can be used to quantify the relative abundance of each protein in a sample mixture, across a wide mass range, while the second focuses on how they can characterize antibody aggregation. Case study 1: Assessing protein abundance within a sample mixture Differences between SEC and MP are illustrated by analyses of a sample mixture consisting of four different proteins (Fig. 1). The analyses, which used SEC-UV and MP, were done by the laboratory of Professor Justin Benesch (University of Oxford). Although the same sample mix was used for both the SECUV and MP measurements, there is a clear disagreement as to which proteins are the most abundant within the mixture: SEC-UV suggests thyroglobulin and ferritin are the most abundant, whereas MP suggests it is conalbumin and aldolase. To understand this apparent discrepancy, one must first consider the fundamental principles underlying how each technique works. Fundamental principles of MP vs. SEC MP, as a single molecule technique, provides a particle count versus mass, i.e., it detects and counts the number of particles of a given mass. Consequently, the intensity of a MP peak (the area under the peak) corresponds simply to the absolute number of molecules with the given mass that were detected during the measurement, which is proportional to the molecular concentration. By contrast, SEC-UV analysis measures UV absorbance vs. column elution time. This data can be converted to absorbance vs. mass using the species’ molecular weights (if known) and the fact that the species of greatest hydrodynamic volume usually elutes first. However, the absorbance data is not only a function of the molecule’s concentration; it also depends on its UV-absorbing properties. The specifications of the UV detector itself, e.g., its sensitivity at a particular UV wavelength, also influence the data. However, when the molar extinction coefficient of each molecule is known, the concentration of each molecule within a mixed sample can be determined from the absorbance data using the Beer-Lambert law. MP agrees with normalized SEC-UV analysis Applying the above approach to each protein in the sample gives rise to a normalized SEC-UV data profile that visually appears to match the MP data (Fig. 1). Furthermore, quantification of the relative abundance of each protein confirms that the results of SEC-UV and MP are in very close agreement (Fig. 2). Mass photometry (MP) is an analytical tool that enables the accurate mass measurement of single molecules in solution, in their native state and without the need for labels. In this technical note, MP is compared to the industry’s gold standard, size exclusion chromatography (SEC), for the analysis of protein abundance and antibody aggregation. 30 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Fig. 1 Analysis of the same sample by MP and SEC-UV illustrates fundamental differences between the techniques. The sample analyzed contained a mixture of four proteins: conalbumin, aldolase, ferritin and thyroglobulin. The molecular weights of each protein within the mixture were known and used to convert the SEC-UV profile from absorbance vs. elution time to absorbance vs. mass. The molar extinction coefficients of each protein were used to normalize the absorbance data. Fig. 2 Abundance (%) of each protein within a sample mixture as determined by MP and SEC-UV. For MP, the number of counts for each molecule is expressed as a percentage of the total number of counts. For SEC-UV, the relative abundances of each molecule are shown, before (SECUV) and after (SEC-UV normalized) normalizing for the molar extinction coefficient. Experimental details • The SEC measurements were performed on an Agilent 1260 Infinity II with a Superdex 200 increase 3.2/300 column, operated in accordance with the manufacturer’s recommended guidelines • A 10 μL volume of the sample mixture was loaded on to the SEC column (the concentration of each protein in the mixture was 14 μM except for ferritin at 1.4 μM) • The same sample mixture was used for MP measurements, but the mix was diluted 1000-fold prior to measurement, to ensure that the concentration was within the appropriate range for this technique 31 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Case study 2: Monitoring aggregation levels of monoclonal antibodies Aggregation is an important quality attribute of many proteinbased biopharmaceuticals, including monoclonal antibodies (mAbs) and multi-specific antibodies, which can influence product efficacy and safety. Monitoring of aggregation levels is therefore essential. SEC is widely considered to be the gold-standard analytical tool for assessing nanometer-sized aggregates and is often combined with multi-angle light scattering (MALS) to enable the determination of molecular weight and size. However, the technique can be complicated by several factors, including column and mobile phase optimization. In contrast, MP detects light scattered by single molecules, enabling the measurement of the molecular mass of biomolecules, in solution, and this technique can analyze small (μL) sample volumes at low concentrations (100 pM up to 100 nM) under native conditions within a few minutes. In the presented analysis, carried out by RIC biologics (Kortrijk, Belgium), both SEC and MP were used to measure aggregation levels of trastuzumab, a monoclonal antibody, and several trastuzumab biosimilars. Trastuzumab is a humanized IgG1 monoclonal antibody that can inhibit HER2 signalling pathways as well as activate antibodydependent cell-mediated cytotoxicity, helping to facilitate the treatment of cancers that overexpress the HER2 cell surface receptor. In recent years, trastuzumab biosimilars have been developed and chromatography has been at the forefront of assessing critical quality attributes, including aggregation, during biosimilar development. The results of this case study demonstrate that SEC and MP are complementary, highlighting the usefulness of MP as an orthogonal technique for monitoring aggregation of biopharmaceuticals such as mAbs. Experimental methods SEC and SEC-MALS measurements were performed on an Agilent Technologies 1260 Bio-inert HPLC Infinity II system equipped with a diode-array detector (DAD) and coupled to a refractive index (RI) detector and a Wyatt miniDAWN multi-angle light scattering detector. Sample compounds were first separated according to their hydrodynamic radius under native conditions using a SEC column and detected using a DAD, MALS detector, and RI detector consecutively. By using the DAD or RI signals as a concentration source, the light scattering data from multiple detector angles can be used to determine the molecular weight (MW) of analyte peaks. The sample load was increased to 315 μg to allow for accurate MW determination of the protein aggregates. MP data was acquired using a TwoMP system, measuring the interference between the scattered light coming from individual sample molecules and the reflected light of the glass slide measurement surface. The resulting signal (interferometric contrast) is directly correlated with molecular mass and can thus be easily converted using protein standards of known MW. A detailed overview of the instrumentation and experimental conditions are provided in Tables 1 and 2. SEC-MALS System Agilent Technologies 1260 Bio-inert HPLC Infinity II with RI detector and Wyatt miniDAWN MALS detector Column Waters XBridge Protein BEH SEC Column 200Å (7.8 x 300 mm x 3.5 μm) Temperature 22°C Mobile phase 0.2 M sodium phosphate, pH 7.0 Flow rate 0.8 mL/min Run time (Elution time) 24 min Injection 10 μg (SEC), 315 μg (SEC-MALS) DAD detection Wavelength 280 nm (band width 4 nm, no reference wavelength) Peak width > 0.2 min (1.25 Hz) R1 Detection Temperature 35°C Peak width > 0.025 min (18.5 Hz) Data processing Software OpenLAB CDS ChemStation and ASTRA V8 Table 1. Experimental conditions for SEC-MALS 32 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Table 2. Experimental conditions for MP MP System TwoMP Temperature Room temperature (21°C) Dilution solvent PBS Sample concentration (and mass of antibody per sample measurement) 20 nM (30 ng) Sample carrier slides Cleaned using water and 2-propanol Run time (Elution time) 1 min Data processing Software DiscoverMP Interpretation of SEC data Trastuzumab-producing Chinese hamster ovary cell (CHO) clone supernatants samples were purified using Protein A affinity chromatography and analysed using SEC. Fig. 1 shows the SEC chromatograms of the trastuzumab originator (Herceptin®, i.e., the ‘parent molecule’ or reference product) and four selected CHO clones. Distinct differences in both the high MW area and the low MW area, respectively left and right of the main monomer peak, can be distinguished. The main high MW peak is associated with a dimer of trastuzumab, which is particularly pronounced in clones 3 and 10. The relative peak areas for both monomer and dimer species (obtained from the SEC-UV chromatogram), which are the main molecules of interest, are provided in Table 3. To derive an estimation of the MW, a SEC-MALS experiment was run for Herceptin® and CHO clone 10 (Fig. 2). The latter showed a MW of 146.5 kDa for the monomer and 295.5 kDa for the dimer, which agrees well with the MW of the concurring peaks of the originator. Denaturing SEC-mass spectrometry analysis showed that noncovalent dimers are present in CHO clone 10, whereas covalently bound dimers are found in the originator product1. This also explains the retention time difference between dimer peaks observed in the clones versus the originator. Fig. 1 SEC-UV chromatograms of trastuzumab originator (Herceptin®) and trastuzumab-producing CHO clones (UV 280 nm). Fig. 2 Molecular weight determination by SEC-MALS of monomer and dimer peak for trastuzumab originator (Herceptin®) and trastuzumabproducing CHO clone 10 (dRI: differential refractive index, LS: light scattering, MW: molecular weight). 33 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Trastuzumab sample analysis by MP Each of the samples were also analyzed by MP and measured at a final concentration of 10 nM (Fig. 3). By using albumin, α-mannosidase and thyroglobulin as calibrants, the MW could be directly interpolated from the data. The MW of the trastuzumab monomer was measured as 155 kDa and for the dimer, 300 kDa, both of which are consistent with the values determined by SEC-MALS. For the analysis of all MP data sets, and to ensure consistency, monomer and dimer peaks are defined as the 120 –190 kDa and 280 – 350 kDa mass intervals, respectively. Percentage abundance of each species is calculated by determining the number of counts (i.e., those within the relevant defined mass interval) as a proportion of the total count number. Fig. 3 MP mass histogram of trastuzumab originator (Herceptin®) and trastuzumab-producing CHO clones. Monomer peaks are highlighted in blue, dimer peaks in orange. Normalizing SEC-UV data When comparing % abundance for both monomer and dimer from SEC-UV (Table 3) and MP (Table 5), we can see that the data is broadly in agreement, with clones 3 and 10 containing the greatest abundance of dimers relative to the predominant monomeric species. However, for a true quantitative comparison, a correction step is necessary. For biotherapeutic proteins such as rastuzumab, the DAD detector response at 280 nm is predominantly related to the number of tyrosine and tryptophan residues it contains. When dimers are formed, a higher number of absorbing residues are present per molecule with respect to the monomer, which affects the peak height (detector response) and consequently also the peak area. The Beer-Lambert law was used to convert the measured peak heights in the SEC chromatogram to analyte concentration [g/L], also applying a correction factor for the UV flow cell. The concentration was subsequently converted into molar concentration [mol/L] by using the MW of the monomer and dimer species, respectively. An overview of the monomer and dimer abundance for all samples, after normalizing absorbance for molar extinction coefficient is reported in Table 4. Table 3. Monomer and dimer abundance from SEC-UV chromatograms. To calculate percentage abundance of both monomer and dimer, the individual peak area for each was calculated and expressed as a percentage of the total/combined peak area. Sample trastuzumab originator Clone 3 Clone 8 Clone 9 Clone 10 Abundance (%) Monomer 99.6 94.9 99.0 97.7 94.7 Dimer 0.4 5.1 1.0 2.3 5.3 Sample trastuzumab originator Clone 3 Clone 8 Clone 9 Clone 10 Abundance (%) Monomer 99.9 97.7 99.8 99.2 97.6 Dimer 0.1 2.3 0.2 0.8 2.4 Sample trastuzumab originator Clone 3 Clone 8 Clone 9 Clone 10 Abundance (%) Monomer 99.3 96.8 98.8 98.7 97.1 Dimer 0.7 3.2 1.2 1.3 2.9 Table 5. Monomer and dimer abundance determined by MP. The 120 –190 kDa mass interval defines the monomer peak and the 280 – 350 kDa mass interval, the dimer peak. Percentage abundance of each species is expressed as a proportion of the total number of counts. Table 4. Monomer and dimer abundance after normalizing SEC-UV absorbance data. 34 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Conclusion MP is an analytical method that can analyze samples under native conditions within a one-minute analysis run time by interpreting the light scattering data from individual analyte molecules approaching the glass slide interface. The underlying detection principle of each of these analytical methods is fundamentally different and there are considerations to be made when comparing data from both, particularly when quantitatively assessing the relative abundance of different molecules in a sample mixture. As outlined in this technical note, to account for this, the measured peak area absorbance values should be normalized by converting to molar concentration, thereby allowing an unbiased comparison of samples analyzed by SEC and MP. This was illustrated with several trastuzumab biosimilars. SEC and MP provided comparable results in both cases; any remaining differences in the data captured by both techniques likely relate to the parameters/criteria used for peak integration. Where the trastuzumab biosimilars are concerned, if the monomer-dimer ratio is concentration dependent, the differences in the concentration of antibody sample used for SEC versus MP would also have an impact upon the relative abundance of monomer versus dimer. Biopharmaceuticals such as monoclonal antibodies are complex and the requirement to ensure efficacy and safety often necessitates the use of several analytical tools to provide a comprehensive view. As shown in this technical note, MP and SEC data – for a simple protein mixture, as well as several trastuzumab biosimilars – are in agreement, confirming the validity and utility of MP as an orthogonal analytical technique. 35 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Measurement of NISTmAb aggregation with mass photometry, AUC, and DLS Abstract This tech note presents analyses of aggregation of the antibody standard NISTmAb using three column-free techniques: mass photometry (MP), analytical ultracentrifugation (AUC), and dynamic light scattering (DLS). Samples of NISTmAb, an antibody reference standard from the US National Institute of Standards and Technology, were assessed before and after exposure to heat stress. The results showed that MP quickly analyses antibody aggregation, with accuracy comparable to AUC and resolution superior to DLS. MP is also faster and easier to use than AUC, and uses much less sample than AUC or DLS. Antibody aggregation is an important critical quality attribute to track, as antibody aggregates pose safety concerns and can affect therapeutic dosage. The most widely used technique for antibody aggregation assessment is sizeexclusion chromatography (SEC), but a significant drawback is that samples can have unforeseen interactions with the chromatographic columns. For this reason, it is important to apply a column-free orthogonal method to confirm SEC results. Column-free orthogonal methods in current use include analytical ultracentrifugation (AUC), dynamic light cattering (DLS), and MP (MP). MP can assess the aggregation levels in an antibody sample and, because it is a single-molecule technique, it readily distinguishes different species present (i.e. monomers, dimers, trimers) and reports the mass and relative abundances of each. Its further advantages are that it is fast, easy to use, and requires very little sample (Table 1). MP does require that samples be at a low concentration (nanomolar), but samples with concentrations up to the tens of micromolar can be measured with use of a rapid dilution microfluidic system (MassFluidix HC).1 Here, we assess how MP compares to AUC and DLS for antibody aggregation analysis. We apply all three methods to analyze the NISTmAb reference standard. NISTmAb (reference material 8671), a humanized IgG1k monoclonal antibody, was chosen as it is widely used for evaluating the performance of different methods and instruments for antibody characterization.2 These measurements were used to benchmark the ability of MP to detect and resolve different types of aggregates. At a glance • Compared to AUC and DLS, MP requires the least sample (10,000x less than AUC; 100x less than DLS). • MP takes only one minute and is straightforward to use. • MP can resolve and accurately quantify more different aggregate species than DLS or AUC. • MP results agree with those from AUC, where comparable. • MP provides rapid, accurate antibody aggregation analysis, enabling frequent, in-house analysis of antibody samples under different conditions. Table 1. Overview of typical requirements for running MP, AUC, and DLS analyses. Of the three methods, MP requires the least sample. Both MP and DLS have a low run time and require little expertise. Method requirements MP AUC DLS Sample (per run) 15 ng 250 μg 65 μg Run time 1 min 6 hr 1 min Expert needed No Yes Yes 36 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT MP resolves subspecies in forced degradation studies Analyses of heat-stressed and control NISTmAb by MP, AUC, and DLS show that all three methods detected a main peak corresponding to monomeric NISTmAb in both the control and heat-stressed samples (Fig. 1). They all also detected an increase in aggregate species in the heat-stressed sample (Fig. 1). However, while peaks for NISTmAb dimers, trimers and tetramers were all clearly visible in the mass histogram (Fig. 1A), AUC only clearly resolved the monomer, dimer and trimer populations (not tetramers) with baseline resolution (Fig. 1B), and DLS did not resolve the different aggregate species (Fig. 1C). These differences occur due to differences in how the three techniques work. Both AUC and DLS are bulk measurements. DLS is biased by larger particles and shape variability can influence the readout. Because MP detects and measures each molecule independently, and irrespective of their shape, it does not have that limitation. In addition, because MP provides a direct mass readout, it is straightforward to identify the species in each mass peak. AUC quantifies antibody aggregation by resolving molecular species according to their sedimentation rates in solution under centrifugal force, yielding size distributions of monomeric and aggregated forms based on their different mass and shape. DLS, meanwhile, measures the average hydrodynamic radius of the molecules in a sample based on the fluctuations in light scattering caused by Brownian motion. Both AUC and DLS are sensitive to molecular shape because they probe hydrodynamic behavior in solution (rather than the molecule’s intrinsic geometric dimensions). Elongated or asymmetric antibody aggregates sediment more slowly in AUC (yielding smaller sedimentation coefficients) and diffuse more slowly in DLS (yielding larger apparent hydrodynamic radii) as compared to spherical particles of the same size. Because larger aggregates tend to have more variable shapes, both techniques can be less accurate in characterizing their size and distribution. MP AUC DLS Figure 1. MP resolved more aggregate NISTmAb species than AUC or DLS. Samples of heat-stressed NISTmAb (20 minutes at 80°C; orange traces) and control NISTmAb (no heat; blue traces) were measured by (A) MP, inset zooms in on aggregate peaks; (B) AUC, inset zooms in on aggregates; and (C) DLS. In each case, a single measurement is shown. 37 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT We next sought to compare whether each of three methods reported the same level of aggregation in the heat-stressed samples. We could not compare the DLS results to the other methods because the different types of aggregates (dimers, trimers etc.) could not be resolved, but we were able to compare the results from MP and AUC for monomers up to trimers. It is important to note that MP is a particle-counting technique, whereas AUC measures UV absorbance, and this difference has implications for how results from the methods should be interpreted (Box 1). To compare the MP results to those from AUC, we converted the AUC results to molar concentration (using our knowledge of the different species’ relative mass, see Box 1). The comparison revealed that the two methods were in close agreement (Fig. 2, Table 2). Both indicated that the control NISTmAb samples were almost entirely monomeric, while the heat stress resulted in significant aggregation, with both methods reporting that about 20% of the species in the sample were either dimers or trimers. Figure 2. MP and AUC agreed on the proportions of monomers, dimers, and trimers present in the heat-stressed vs. control NISTmAb sample. Comparison of aggregation in control and heat-stressed NISTmAb samples, as measured by MP and AUC. For both methods, only the aggregate levels up to trimer were compared because the AUC data were poorly resolved for tetramers. Error bars on MP data denote standard deviation for triplicate measurements. The * denotes that the AUC measurements were converted to molar ratio (see Box 1). For quantification, see Table 2. Table 2. Percentages of monomer, dimers, and trimers present in heatstressed vs. control NISTmAb samples. Data shown is quantification of populations (molar ratio) for measurements presented in Fig. 2. For MP data, the mean and standard deviation (STD) are given for triplicate measurements. The * denotes that the AUC measurements were converted to molar ratio (see Box 1). No Heat 20 min at 80°C MP (% mean ± STD) AUC* (%) MP (% mean ± STD) AUC* (%) Monomer 98.5 ± 0.2 98.6 78.7 ± 1.3 76.9 Dimer 1.5 ± 0.2 1.1 13.4 ± 1.1 13.6 Trimer 0.1 ± 0.0 0.3 7.9 ± 0.3 9.5 38 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Before comparing results from AUC vs. MP, it is important to understand differences in how each technique works. In particular, AUC is a bulk technique: It monitors the collective sedimentation of groups of molecules with similar size and shape. In comparison, MP performs a single-molecule measurement of interferometric contrast, producing a mass distribution. To illustrate the implications of this difference, we consider a hypothetical sample containing 10 molecules of monomer and 10 molecules of dimer (Fig. 3). A particle-counting technique, such as MP, would report that this sample contains 50% monomers and 50% dimers – the molar ratio. A technique that measures UV absorbance, such as AUC, would report that the sample contains 33% monomers and 67% dimers – the mass ratio. As biomolecules tend to function based on their molar concentration, not their mass concentration, the molar concentration tends to be a more intuitive measure. Box 1. Understanding the differences between MP and AUC In AUC, molar extinction coefficients can be used to convert absorbance profiles into molar ratio readouts. For this to be accurate, the extinction coefficient must be known, and all the molecules in a given peak must have the same extinction coefficient (which may not always be the case). Fig 3. MP vs. AUC: MP reports the molar ratio while AUC reports the mass ratio. Particle-counting methods (such as MP) report the molar ratio of different species present in a sample, while UV absorbance methods (such as AUC) report the mass ratio. As this example shows, dimers produce twice the signal of the monomers when UV absorbance is measured, due to their having double the mass. Before comparing data between these two types of methods, this difference must first be taken into account. 39 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Conclusion Overall, for detecting and quantifying aggregates in the dimer and trimer range in these forced degradation studies, the MP results were comparable to those from AUC. MP was also able to resolve and quantify higher-order aggregates where AUC could not. Meanwhile, although DLS detected the presence of aggregates, the different aggregate species could not be resolved from the DLS data. These results, coupled with its practical advantages, make MP a powerful method for quantifying antibody aggregation for several reasons: • Column free: Like AUC and DLS, MP does not require a chromatography column. • Superior resolution: Compared to DLS and even AUC, MP has more power to resolve higher-mass species. • Low sample consumption: Because it uses little sample, MP can be used often, enabling e.g. forced degradation studies under different conditions (heat stress, buffer composition, pH etc.). MP consumes up to ~10,000x less sample than AUC. • Rapid measurement: An MP measurement takes one minute, making it possible to test several replicates of different conditions in a practical timeframe. MP is more than 300x faster than AUC. • Straightforward analysis: With the StreamlineMP Antibody Stability Module, it is easy to organize and perform the analysis of large MP datasets, like those produced by forced degradation studies. This software was used to analyze the MP data from this study, reducing the analysis time significantly. These strengths make MP an ideal technique to derisk antibody development and production with early, frequent, in-house analysis of antibody samples under different conditions. Materials and Methods • NISTmAb was obtained from Agilent. • AUC: Sedimentation velocity experiments were performed using a ProteomeLab XL-1 ultracentrifuge (Beckman Instruments) using an AnTi-60 rotor at 40000 rpm and 20°C (performed by Oxford University Innovations, UK). Heated and unheated samples were allowed to equilibrate at 20°C for 1 h prior to the start of the experiment. Absorbance data were collected at 280 nm and fit as a function of sedimentation coefficient distributions [c(s) analysis] in SedFIT. Because the peaks >trimer could not be resolved easily and their predicted mass deviated from the expected mass of multimers, only the contributions of monomers, dimers and trimers were quantified for comparison to MP data. The integrated absorbance of monomers, dimers and trimers (i.e. area under the peaks) was divided by 1, 2 and 3, respectively, to convert into molar distribution, and % abundance was calculated by this equation: %X-mer = (molar distribution X-mer) / (molar distribution monomer + molar distribution dimer + molar distribution trimer) * 100 • DLS: A ZetaSizer Ultra (Malvern) was used to collect DLS data for heated and unheated samples. Five replicates were collected for every sample. Data were fit using the ZS XPLORER software (v 4.0.0.683). • MP: Heated and unheated samples were diluted on sample well cassettes and MassGlass UC slides (Refeyn Ltd.) for final measurement at 5 nM with PBS, equivalent to 15ng. For mass calibration, MassFerence™ P1 (Refeyn Ltd) in PBS was used. Data was collected with AcquireMP v. 2024.2.0 software using default settings. Normalized density plots were generated with DiscoverMP v. 2025. 1.0 software. Data analysis was performed with StreamlineMP v. 2025.1.0a0. Percentages of monomers, dimers and trimers were calculated based on the number of counts of each population divided by the total counts of monomers, dimers and trimers. Measured were performed in triplicate. References 1 Refeyn, MassFluidix HC Product Data Sheet https://www.refeyn.com/massfluidix-hc-system 2 NIST (2025). NIST Monoclonal Antibody Reference Material 8671. https://www.nist.gov/programs-projects/nist-monoclonalantibodyreference- material-8671 40 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Characterization of forced antibody degradation with automated mass photometry Aggregation is a major issue for antibodies and can have a detrimental impact on product efficacy and safety1. In general, a low propensity to form aggregates (from dimers to visible precipitates) during the manufacturing process and in storage is desirable. Forced degradation studies are a commonly used method to assess the physicochemical stability of biopharmaceuticals, including monoclonal antibodies, bispecific antibodies and antibody-drug conjugates2. Analytical techniques play a vital role in such studies, ensuring that the right candidates are selected in the early stages of the drug development life cycle. MP is a convenient tool for providing rapid insights into changes in molecular behavior, from the aggregation and degradation of proteins to the characterization of AAV capsids3. MP can tolerate a wide range of buffer conditions and only requires very small quantities of sample for analysis. Refeyn’s TwoMP Auto is an automated MP system that can be used for a range of different applications – providing operators with greater ease of use and walkaway automation. In this application note, we describe the use of the TwoMP Auto as part of a forced degradation study involving a panel of therapeutic antibodies provided by an industrial partner. MP characterizes antibody stability in response to stress Data presented in this application note was part of a broader study that aimed to assess the effect of three stressors – light, pH and peroxide, on a panel of 14 antibodies. That was to establish whether fragmentation and/or aggregation occurred following exposure to those stress conditions. Each condition was measured in triplicate, resulting in a total of 168 measurements (including control measurements). The TwoMP Auto automated and simplified the user workflow by permitting the measurement of up to 24 samples per run, which typically takes ~90 minutes to complete and requires only picogram quantities of each antibody. The system can analyze the data whilst it is being recorded and the associated software is intuitive, so the user can start gleaning results easily, with no complicated data analysis routines necessary. Representative MP data for two of the antibodies from the panel tested are shown in Figs. 1-2 and Tables 1-2. They behave differently in response to the same stress conditions, when compared to the relevant control sample. More pronounced fragmentation is observed for ‘Antibody 1’ (Fig. 1, Table 1), particularly in response to stress condition 2, whereas a greater degree of aggregation is observed overall for ‘Antibody 2’ (Fig. 2, Table 2), indicative of these antibodies having very different physiochemical properties. The data also show the relative abundance of each distinct population of molecules in each antibody sample. MP is a single molecule technique that can detect and count each molecule in a sample. Therefore, the data acquired includes information on rare monomers, dimers and other species. Fig. 1 Exposure of ‘Antibody 1’ samples to three different stress conditions. The ‘Stress 1’, ‘2’, and ‘3’ are known to be light, pH change, or peroxide but details of which condition corresponds to which name were not disclosed by the industrial partner. Samples were measured at 10 nM concentration using 5 μl on the TwoMP Auto. Results show that ‘stress 2’ promotes the greatest degree of fragmentation of the antibody, as shown by an increase in the % fraction of ‘low molecular mass’ species (LMM), relative to the control sample. Mass photometry (MP) is a label-free bioanalytical tool that can assess antibody fragmentation and/or aggregation in response to induced stressors, such as light, pH and peroxide. The TwoMP Auto allows rapid, automated measurements of multiple samples with high reproducibility, across a wide mass range. 41 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Table 1. Fragmentation and aggregation products in Antibody 1 sample upon exposure to different stressors. Low molecular mass fraction (30 - 90 kD) corresponds to [fragmented antibody], monomer mass fraction (90 - 170 kDa) to [intact antibody] and high molecular mass fraction (> 170 kDa) to [aggregated antibody]. Table 2. Fragmentation and aggregation products in Antibody 2 sample upon exposure to different stressors. Low molecular mass fraction (30 - 90 kDa) corresponds to [fragmented antibody], monomer mass fraction (105 -200 kDa) to [intact antibody], dimer (250 - 340 kDa) and trimer (400 - 480 kDa) mass fractions to [aggregated antibody]. Condition % Fragmented antibody % Intact antibody % Aggregated antibody Control 9% (± 0.7) 90% (± 0.6) 1% (± 0.3) Stress 1 10% (± 1.1) 88% (± 1.1) 2% (± 0.1) Stress 2 27% (± 1.8) 70% (± 1.0) 4% (± 0.9) Stress 3 14% (± 2.2) 81% (± 2.1) 5% (± 0.3) Condition % Fragmented antibody % Intact antibody % Aggregated antibody (dimer) % Aggregated antibody (trimer) Control 4% (± 0.8) 95% (± 0.7) 1% (± 0.1) 0% Stress 1 5% (± 0.4) 85% (± 0.3) 8% (± 0.4) 2% (± 0.2) Stress 2 7% (± 1.5) 91% (± 1.4) 2% (± 0.1) 0% Stress 3 5% (± 0.8) 86% (± 0.5) 8% (± 0.3) 1% (± 0.1) Summary When performing forced degradation analyses, there may be many degradation products or variants generated; as such, several different analytical tools spanning a wide mass range may be needed to ensure all relevant data are captured. MP is ideally suited for use in this setting and would be a complementary approach to other techniques such as SECMALS, for example. MP is a column-free approach and works with a wide range of buffers, meaning very little sample optimization is required. Information about a sample can be captured using only picogram quantities of material, as in this study. Combined with user friendly operation and intuitive data analysis, this technique can provide new insights quickly, thereby facilitating rapid data-based decision making to support process development or developability assessments, for example. Consequently, MP, particularly using the TwoMP Auto, which permits the user to load samples and walk away, can form part of a suite of analytical methods to provide an in-depth characterization of biopharmaceuticals such as monoclonal and bispecific antibodies. Fig. 2 Exposure of ‘Antibody 2’ samples to three different stress conditions (light, pH change, or peroxide). Samples were measured at 10 nM concentration using 5 μl on the TwoMP Auto. Results show that stress conditions 1 and 3 promote aggregation (specifically, the formation of dimers and trimers), relative to the control sample. 20 Testimonials “Mass photometry provides a fast screening tool to investigate mRNA integrity and size.” De Vos et al. (2024), J Chromatogr A “The data confirm the great potential of [mass photometry] technology... as a fast and simple orthogonal method that provides insights into the homogeneity and stability of mRNA samples.” Camperi et al. (2024), Anal Chem Unit 9, Trade City, Sandy Lane West, Oxford OX4 6FF, United Kingdom ©2024 Refeyn Ltd For information on products, demos and ordering, write to [email protected] Samux and Refeyn are registered trademarks of Refeyn Ltd. refeyn.com @refeynit Refeyn Refeyn About Refeyn Refeyn pioneers analytical instruments that put molecular mass measurement capabilities within easy reach for scientists. Refeyn’s unique products measure the mass of individual proteins, nucleic acids, complexes and viruses directly in solution – providing vital insights for scientific discovery, R&D and therapeutics production. Our instruments feature mass photometry technology, which uses light to quantify the mass of single particles in solution without labels, and macro mass photometry technology, which uses light to characterize large viral vectors. Providing intuitive data in minutes, mass photometry technologies help scientists solve their research questions, optimize R&D processes and focus on innovation. 20 Testimonials “Mass photometry provides a fast screening tool to investigate mRNA integrity and size.” De Vos et al. (2024), J Chromatogr A “The data confirm the great potential of [mass photometry] technology... as a fast and simple orthogonal method that provides insights into the homogeneity and stability of mRNA samples.” Camperi et al. (2024), Anal Chem Unit 9, Trade City, Sandy Lane West, Oxford OX4 6FF, United Kingdom ©2024 Refeyn Ltd For information on products, demos and ordering, write to [email protected] Samux and Refeyn are registered trademarks of Refeyn Ltd. refeyn.com @refeynit Refeyn Refeyn About Refeyn Refeyn pioneers analytical instruments that put molecular mass measurement capabilities within easy reach for scientists. Refeyn’s unique products measure the mass of individual proteins, nucleic acids, complexes and viruses directly in solution – providing vital insights for scientific discovery, R&D and therapeutics production. Our instruments feature mass photometry technology, which uses light to quantify the mass of single particles in solution without labels, and macro mass photometry technology, which uses light to characterize large viral vectors. Providing intuitive data in minutes, mass photometry technologies help scientists solve their research questions, optimize R&D processes and focus on innovation. References 1 Vázquez-Rey & Lang, Biotechnol Bioeng, 2011. https://doi.org/10.1002/bit.23155 2 Blessy, Patel et al., JPA, 2014. https://doi.org/10.1016/j.jpha.2013.09.003 3 Refeyn, How does MP work? [Blog], 2021. https://refeyn.com/post/how-does-mass-photometry-work 42 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Quantifying antibody aggregation at nano- and micromolar sample concentrations with the mass photometry Antibody Stability Module Antibody aggregation can be measured quickly and accurately with mass photometry (MP). This tech note demonstrates this analysis, automated with Refeyn’s Antibody Stability Module for the StreamlineMP software suite. It provides accurate information about aggregation for samples at both nanomolar and micromolar concentrations. Aggregation is a key quality attribute that is essential to monitor during the development and production of therapeutic antibodies. It can be easily measured using MP, a technique that measures the mass of individual proteins in solution in just a few minutes and while using as little as 30 ng of sample. On top of these advantages, MP is currently the only analytical technique that can measure antibody samples in the nanomolar concentration range, while other frequently used techniques like size exclusion chromatography (SEC) are limited to micromolar concentrations. As IV-delivered antibodies1 are frequently in concentrations as low as 0.002 mg/mL (nM range), the measurement conditions of MP are especially relevant for in-use studies of these therapeutics. MP2 can also measure samples in the μM range by using Refeyn’s MassFluidixTM HC rapid dilution addon, 3,4 which facilitates comparisons with SEC and other orthogonal techniques. To streamline antibody characterization, Refeyn introduced a dedicated Antibody Stability Module for its StreamlineMP software suite. The Antibody Stability Module automatically analyzes large datasets and reports the relative proportions of monomers, dimers, trimers and higher-order oligomers – reducing analysis times by 80% or more. The module makes it easy to compare antibody types, stress conditions and more, and ensures the consistency of analysis parameters across users. In this tech note, we demonstrate the use of MP and the Antibody Stability Module for automated analysis of the aggregation of three sets of antibodies: The NIST monoclonal antibody standard (NIST mAb) and commercially available samples of IgG2 and IgG4. All samples were measured in triplicate. Fig . 1 MP quantifies aggregation in three antibody standards. Samples of IgG2 (A), IgG4 (B) and NIST mAB (C) were analyzed using MP and the Antibody Stability Module. Each sample was measured in triplicate; one representative mass histogram is shown for each sample. The x-axis shows mass in kDa, and the y-axis shows total counts for each bin. The dashed vertical lines indicate the mass limits used by the software to distinguish monomers, dimers, trimers and higher-order oligomers. In each case, monomers (orange peaks) were the dominant species. To showcase the capabilities of MP for antibody characterization, we measure each sample at nanomolar concentration (the typical concentration for MP) and at micromolar concentration (the typical concentration for SEC). The micromolar concentration measurement is performed using Refeyn’s MassFluidixTM HC rapid dilution add-on. 43 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Table 1. Summary of antibody aggregation results. An overview of the results from measuring the antibody standards (IgG2, IgG4 and NIST mAb) with MP and the Antibody Stability Module is shown. The percentages of monomers, dimers, trimers and higher-order aggregates are given for each sample (mean ± SD for triplicate measurements). The ‘nM’ column contains the results from standard MP measurements at low concentration (~5 nM), while the ‘μM’ column contains measurements after rapid dilution from ~7 μM to ~5 nM using the MassFluidix HC add-on. Sample Monomer (%) Dimer (%) Trimer (%) Higher-order (%) nM μM nM μM nM μM nM μM IgG2 97.30 ± 0.17 97.35 ± 0.07 2.47 ± 0.12 2.45 ± 0.07 0.13 ± 0.06 0.15 ± 0.07 0.10 ± 0.00 0.05 ± 0.07 IgG4 97.03 ± 0.15 96.05 ± 0.64 2.83 ± 0.21 3.70 ± 0.57 0.13 ± 0.12 0.20 ± 0.00 0.00 ± 0.00 0.05 ± 0.07 NIST mAb 98.50 ± 0.20 95.95 ± 3.20 1.40 ± 0.17 3.60 ± 2.75 0.07 ± 0.06 0.40 ± 0.35 0.03 ± 0.06 0.05 ± 0.06 Aggregation analysis with antibody standards Aggregation was assessed using MP in samples of three antibody standards: NIST mAb (an IgG1), and IgG2 and IgG4 (Absolute Antibody). The results show that the standards had low levels of aggregation (around 3%), mainly appearing as dimer formation, and that MP was able to detect and quantify this aggregation (Fig. 1, Table 1). Measuring aggregation in μM concentration samples The optimal concentration for MP experiments is 100 pM – 100 nM, which is significantly lower than the concentration used in other techniques that assess antibody aggregation, such as SEC. To demonstrate the application of MP to more concentrated samples, we repeated the measurements with the same samples at micromolar concentration with the MassFluidix HC add-on for the TwoMP, which uses microfluidics to dilute samples up to 10,000-fold in under 40 ms. The speed of the dilution means that the measurement is completed before significant changes to the composition of the sample can occur. The results show that the trends in aggregation were conserved between samples at high (~7 μM) or low (~5 nM) concentration (Table 1). This finding indicates that antibody aggregation can be meaningfully assessed with MP for samples at lower (nanomolar) as well as higher (micromolar) concentrations. 44 www.refeyn.com DISCOVERY CANDIDATE CHARACTERIZATION PROCESS DEVELOPMENT Conclusion MP is a rapid, reliable technique for accurately quantifying antibody aggregation. MP analysis of antibody aggregation is streamlined with the Antibody Stability Module, a specialized analysis module for Refeyn’s StreamlineMP software suite. The module automates the analysis process, dramatically reducing the time and effort required to assess aggregation. Furthermore, we have shown that MP can be used to measure samples at micromolar and nanomolar concentrations, and that trends in antibody aggregation are conserved between both concentration ranges. Characterizing samples at low concentrations is a unique capability of MP, allowing measurement conditions close to those found in IV-delivered antibody therapeutics. Meanwhile, measurements of samples at concentrations in the M range typically used in SEC are enabled by the MassFluidix HC add-on. MP results are comparable to those obtained by SEC5 with lower sample consumption. Therefore, the TwoMP mass photometer plus MassFluidix HC and the Antibody Stability Module represent a powerful and versatile analytical tool. They can streamline the development and production of antibody based therapeutics. References Experimental details • Measurements were performed on a TwoMP mass photometer and MassGlass UC sample carrier slides (Refeyn, Ltd.) • Data were analyzed using the Antibody Stability Module, Development Version 5 • For all samples, the expected mass value used in the Antibody Stability Module analysis was 150 kDa, with corresponding default mass limits set by the software • Calibrations were performed using the calibrant MassFerenceTM P1 (Refeyn, Ltd.) • For the measurements at nanomolar concentration, samples were diluted to ~5 nM prior to measurement • For the measurements at micromolar concentration, samples at 1 mg/mL (~6.67 M), underwent rapid dilution (using the MassFluidix HC add-on) for a MP measurement at ~5 nM • Samples were diluted in 1x Dulbecco’s Phosphate-Buered Saline (DPBS) • NIST mAb was obtained from Agilent Technologies, Inc. (Stockport, Cheshire, UK. Product code: 5191-5744) • IgG2 and IgG4 samples were obtained from Absolute Antibody, Ltd. (Redcar, Cleveland, UK) 1 Hong et al. PDA JPST 2024. https://doi.org/10.5731/pdajpst.2023.012860 2 Morar-Mitrica et al. MAbs 2015. https://doi.org/10.1080%2F19420862.2015.1046664 3 Refeyn, MassFluidix HC Product Data Sheet https://www.refeyn.com/massfluidix-hc-system 4 Refeyn, Application note: MP analysis of samples at micromolar concentration https://www.refeyn.com/analysis-of-low-affinity-protein-interactions 5 Refeyn, Technical Note: Assessing protein samples by MP and size exclusion chromatography https://www.refeyn.com/mass-photometry-vs-sec-in-protein-analysis 45 www.refeyn.com 18 www.refeyn.com TwoMP SamuxMP KaritroMP Technology Mass photometry Macro mass photometry Parameters measured Mass Diameter (size) Contrast (mass proxy) Concentration range 100 pM – 100 nM** 1011 particles/mL 108 – 109 particles/mL Particle type measured (Mass or size range) Proteins (30* kDa – 5 MDa) Nucleic acids (200 – 10k b; 100 – 5k bp) AAVs (500 kDa – 6 MDa) Large viruses (e.g. AdV) (40 – 150 nm) Consumables Sample carrier slides MassGlass™ UC for protein analysis, MassGlass NA for nucleic acid analysis, MassGlass UC MassGlass KV Calibrant MassFerence™ P1 (For protein measurements within 90 – 1000 kDa) MassFerence P2 SizeFerence™ Add-ons High-concentration samples Automation TwoMP Auto*** SamuxMP Auto*** Software Data acquisition AcquireMP Analysis Standard DiscoverMP DiscoverMPK StreamlineMP Antibody stability module GMP environments GMP software package * Using slides hand cleaned according to Refeyn protocol. The lower limit for Refeyn’s pre-cleaned MassGlass UC slides is 50 kDa without further cleaning. ** The MassFluidix HC add-on expands the TwoMP sample concentration range up to the tens of micromolar. *** Autonomous measurement of 24 samples in as little as 90 min. Refeyn mass photometers are Class 1 laser products. Table 3. Refeyn’s end-to-end mass photometry solutions enable rapid, user-friendly analysis of mRNA as well as other sample types. i Please note that in practice, exact specifications are sample and buffer dependent. * Using slides hand cleaned according to Refeyn protocol. The lower limit for Refeyn’s pre-cleaned MassGlass UC slides is 50 kDa without further cleaning. ** The MassFluidix HC add-on expands the TwoMP sample concentration range up to the tens of micromolar. *** Autonomous measurement of 24 samples in as little as 90 min. Refeyn mass photometers are Class 1 laser products. Mass photometry end-to-end solutions Our vision is to accelerate discovery through innovation, empowering the latest scienti c breakthroughs in basic research and transforming biotherapeutic development and manufacturing. Get in touch to speak with one of our mass photometry experts. Refeyn Headquarters (UK) Unit 9, Trade City Sandy Lane West Oxford OX4 6FF United Kingdom Phone: +44†1865 800175 Refeyn Headquarters (USA) 21 Hickory Drive Suite 200A Waltham MA 02451 USA Phone: + 1 (971) 200 1370 EMEA: Paul Davies Email:†[email protected] USA: Candi Mach Email: [email protected] APAC and Global: Guillaume Kohen Email: [email protected] About Refeyn pioneers measurement capabilities unique products acids, complexes Our instruments light to quantify labels, and macro characterize large mass photometry questions, optimize For information on products, demos and ordering, please email: [email protected] MassFerence, MassFluidix, MassGlass, Karitro, Samux, SizeFerence and Refeyn are registered trademarks of Refeyn Ltd. © 2026 Refeyn Ltd V1 JAN-2026