Advertisement

Screening for protein–protein interactions with asymmetrical flow field-flow fractionation

Published:February 24, 2021DOI:https://doi.org/10.1016/j.xphs.2021.02.026

      Abstract

      We describe a new method for screening protein-protein interaction of biopharmaceutical molecules at dilute concentrations to predict development issues at high concentration. The method is based on Asymmetrical Flow Field-Flow Fractionation (AF4) measurements using well known effects of protein-protein attraction on the fractionation profile due to elevated protein concentrations occurring close to the membrane. We explore the effect for 4 different monoclonal antibodies and show that the profiles obtained are quite different. Interestingly, we find that the recovery in AF4 correlates with the diffusion interaction parameter, which is a standard method for the analysis of protein-protein attraction. The results are insensitive to the protein concentration and buffer composition of the sample solution and only depend on the absolute amount of protein loaded and on the running buffer. This makes the method highly suitable for developability assessment in a compound discovery workflow.

      Introduction

      During the last decade, several methods to assess developability of biopharmaceuticals early in the discovery process have been described.
      • Jarasch A
      • Koll H
      • Regula J T
      • Bader M
      • Papadimitriou A
      • Kettenberger H.
      Developability assessment during the selection of novel therapeutic antibodies.
      • Jain T
      • Sun T
      • Durand S
      • Hall A
      • Houston N R
      • Nett J H
      • Sharkey B
      • Bobrowicz B
      • Caffry I
      • Yu Y
      • Cao Y
      • Lynaugh H
      • Brown M
      • Baruah H
      • Gray L T
      • Krauland E M
      • Xu Y
      • Vásquez M
      • Wittrup KD
      Biophysical properties of the clinical-stage antibody landscape.
      • Wolf Pérez A-M
      • Sormanni P
      • Andersen J S
      • Sakhinini L I
      • Rodriguez-Leon I
      • Bjelke J R
      • Gajhede A J
      • De Maria L
      • Otzen D E
      • Vendruscolo M
      • Lorenzen N
      In vitro and in silico assessment of the developability of a designed monoclonal antibody library.
      These techniques are applied in industrial research laboratories to ensure that new therapeutic molecules not only have the right biological function, but also the robustness required for production and logistics at commercial scale. Developability assessment covers many different areas such as assessment of chemical stability, physical stability, expression yields and immunogenicity. Due to the high concentration often needed for antibody formulations (100-150 mg/ml), physical stability is often considered as the primary development risk. Common physical stability liabilities include high viscosity, precipitation, aggregation and phase separation which can compromise successful purification, stability in formulation and drug administration. Consequently, particular focus has been given to detect protein-protein attraction, which has been established to be a good predictor of high viscosity, aggregation, precipitation and liquid-liquid phase separation.
      • Connolly B D
      • Petry C
      • Yadav S
      • Demeule B
      • Ciaccio N
      • Moore J M R
      • Shire S J
      • Gokarn YR.
      Weak interactions govern the viscosity of concentrated antibody solutions: high-throughput analysis using the diffusion interaction parameter.
      • Starr C G
      • Tessier P M
      Selecting and engineering monoclonal antibodies with drug-like specificity.
      • Kingsbury J S
      • Saini A
      • Auclair S M
      • Fu L
      • Lantz M M
      • Halloran K T
      • Calero-Rubio C
      • Schwenger W
      • Airiau C Y
      • Zhang J
      • Gokarn YR.
      A single molecular descriptor to predict solution behavior of therapeutic antibodies.
      The ‘gold standard’ is the diffusion interaction parameter kD,
      • Connolly B D
      • Petry C
      • Yadav S
      • Demeule B
      • Ciaccio N
      • Moore J M R
      • Shire S J
      • Gokarn YR.
      Weak interactions govern the viscosity of concentrated antibody solutions: high-throughput analysis using the diffusion interaction parameter.
      which has been shown to correlate well with high concentration viscosity.
      • Connolly B D
      • Petry C
      • Yadav S
      • Demeule B
      • Ciaccio N
      • Moore J M R
      • Shire S J
      • Gokarn YR.
      Weak interactions govern the viscosity of concentrated antibody solutions: high-throughput analysis using the diffusion interaction parameter.
      ,
      • Yadav S
      • Shire S J
      • Kalonia D S
      Viscosity behavior of high-concentration monoclonal antibody solutions: correlation with interaction parameter and electroviscous effects.
      Another method is AC-SINS, which is well suited for high throughput screening.
      • Wu J
      • Schultz J S
      • Weldon C L
      • Sule S V
      • Chai Q
      • Geng S B
      • Dickinson C D
      • Tessier P M
      Discovery of highly soluble antibodies prior to purification using affinity-capture self-interaction nanoparticle spectroscopy.
      ,
      • Liu Y
      • Caffry I
      • Wu J
      • Geng S B
      • Jain T
      • Sun T
      • Reid F
      • Cao Y
      • Estep P
      • Yu Y
      • Vásquez M
      • Tessier P M
      • Xu Y.
      High-throughput screening for developability during early-stage antibody discovery using self-interaction nanoparticle spectroscopy.
      These methods are highly useful but have some shortcomings: AC-SINS is currently limited to monoclonal antibodies (mAbs) and kD measurements require double-digit concentrations of high purity, which can be a significant limitation in a compound discovery workflow.
      Asymmetrical Flow Field-Flow Fractionation (AF4) was developed more than 30 years ago.
      • Wahlund K-G
      • Giddings J C
      Properties of an asymmetrical flow field-flow fractionation channel having one permeable wall.
      AF4 is a well-known analytical technique for separation by size and the method and its applications has been discussed in detail in several reviews.
      • Wahlund K-G.
      Chapter 18 Asymmetrical flow field-flow fractionation in.
      • Wahlund K-G.
      Flow field-flow fractionation: critical overview.
      • Yohannes G
      • Jussila M
      • Hartonen K
      • Riekkola ML
      Asymmetrical flow field-flow fractionation technique for separation and characterization of biopolymers and bioparticles.
      • Fraunhofer W
      • Winter G.
      The use of asymmetrical flow field-flow fractionation in pharmaceutics and biopharmaceutics.
      • Leeman M
      • Storm M U
      • Nilsson L.
      Practical applications of asymmetrical flow field-flow fractionation (AF4): a review.
      • Gigault J
      • Pettibone J M
      • Schmitt C
      • Hackley VA
      Rational strategy for characterization of nanoscale particles by asymmetric-flow field flow fractionation: a tutorial.
      • Cao S
      • Pollastrini J
      • Jiang Y.
      Separation and characterization of protein aggregates and particles by field flow fractionation.
      In brief, the separation takes place in a flow channel, where one wall is a membrane permeable to the elution buffer, such that the flow along the channel is superimposed by a flow across the channel, which concentrates the molecules to be analyzed near the membrane. The size separation is due to a complex interplay between cross-flow, molecular diffusion and the (parabolic) profile of flow along the channel. During the first step of AF4 the sample is injected and focused into a narrow band. The cross-flow forces the sample molecules towards the accumulation wall and causes an increased concentration in a layer near the membrane.
      • Wahlund K-G
      • Giddings J C
      Properties of an asymmetrical flow field-flow fractionation channel having one permeable wall.
      The concentration at the accumulation wall may be 10-200 times higher than the sample injection concentration.
      • Caldwell KD
      • Brimhall S L
      • Gao Y
      • Giddings JC
      Sample overloading effects in polymer characterization by field-flow fractionation.
      ,
      • Arfvidsson C
      • Wahlund K-G.
      Mass overloading in the flow field-flow fractionation channel studied by the behaviour of the ultra-large wheat protein glutenin.
      The high concentration will affect the mean layer thickness, sample interaction with the membrane, intermolecular interactions, and viscosity.
      • Wahlund K-G.
      Flow field-flow fractionation: critical overview.
      ,
      • Caldwell KD
      • Brimhall S L
      • Gao Y
      • Giddings JC
      Sample overloading effects in polymer characterization by field-flow fractionation.
      • Arfvidsson C
      • Wahlund K-G.
      Mass overloading in the flow field-flow fractionation channel studied by the behaviour of the ultra-large wheat protein glutenin.
      • Litzén A
      • Wahlund K-G.
      Effects of temperature, carrier composition and sample load in asymmetrical flow field-flow fractionation.
      Hence, in the practice of AF4 it is often observed that the elution profile is different from the theoretical. Excess sample load lead to changes in the migration velocity, which affects peak elution times as well as peak symmetry, so called mass overloading effects.
      • Wahlund K-G.
      Flow field-flow fractionation: critical overview.
      ,
      • Caldwell KD
      • Brimhall S L
      • Gao Y
      • Giddings JC
      Sample overloading effects in polymer characterization by field-flow fractionation.
      ,
      • Litzén A
      • Wahlund K-G.
      Effects of temperature, carrier composition and sample load in asymmetrical flow field-flow fractionation.
      • Benincasa MA
      • Giddings JC
      Separation and molecular weight distribution of anionic and cationic water-soluble polymers by flow field-flow fractionation.
      • Benincasa MA
      Chapter 27 Synthetic polymer-water soluble in.
      The peak maximum retention times can increase or decrease, and the peak asymmetry can be tailing or fronting depending on sample type (e.g. particles, polymers, proteins) and the solvent as well as ionic strength used in the analysis.
      • Cao S
      • Pollastrini J
      • Jiang Y.
      Separation and characterization of protein aggregates and particles by field flow fractionation.
      • Caldwell KD
      • Brimhall S L
      • Gao Y
      • Giddings JC
      Sample overloading effects in polymer characterization by field-flow fractionation.
      • Arfvidsson C
      • Wahlund K-G.
      Mass overloading in the flow field-flow fractionation channel studied by the behaviour of the ultra-large wheat protein glutenin.
      • Litzén A
      • Wahlund K-G.
      Effects of temperature, carrier composition and sample load in asymmetrical flow field-flow fractionation.
      • Benincasa MA
      • Giddings JC
      Separation and molecular weight distribution of anionic and cationic water-soluble polymers by flow field-flow fractionation.
      • Benincasa MA
      Chapter 27 Synthetic polymer-water soluble in.
      • Moon MH
      • Park I
      • Kin Y.
      Size characterization of liposomes by flow-field flow fractionation and photon correlation spectroscopy, effect of ionic strength and pH of carrier solution.
      In the case of proteins, the effect of concentration-dependent self-association becomes important.
      In a recent publication Marioli et al,
      • Marioli M
      • Kok W T
      Recovery, overloading, and protein interactions in asymmetrical flow field-flow fractionation.
      looked in detail on protein interactions and overloading effects. They found that the effect of sample load on the fractograms could be rationalized by the higher local viscosity close to the membrane and was not only dependent on the operational conditions during the AF4 separation, but also dependent on protein characteristics. We hypothesized that this response to the high concentrations inherent in AF4 might be useful for screening of biopharmaceuticals and investigated a set of different mAbs.

      Materials and methods

      MAbs 1, 2, 3 and 4 are different human monoclonal antibodies of the IgG4 isotype raised against different targets. They have been expressed in mammalian cell lines and been purified as a minimum with both a Protein A purification and a size-exclusion chromatography step yielding a purity above 95 % as measured with SE-HPLC (data not shown). mAb 2 has previously been characterized in detail, and is denoted as WT in references
      • Wolf Pérez A M
      • Sormanni P
      • Andersen J S
      • Sakhnini L I
      • Rodriguez-Leon I
      • Bjelke J R
      • Gajhede A J
      • De Maria L
      • Otzen D E
      • Vendruscolo M
      • Lorenzen N
      In vitro and in silico assessment of the developability of a designed monoclonal antibody library.
      ,
      • Kopp M R G
      • Wolf Pérez A M
      • Zucca M V
      • Palmiero U C
      • Friedrichsen B
      • Lorenzen N
      • Arosio P
      An accelerated surface-mediated stress assay of antibody instability for developability studies.
      and as mAb-C in reference.
      • Neergaard M S
      • Kalonia D S
      • Parshad H
      • Nielsen A D
      • Møller E H
      • van de Weert M.
      Viscosity of high concentration protein formulations of monoclonal antibodies of the IgG1 and IgG4 subclass - prediction of viscosity through protein-protein interaction measurements.
      Theoretical pI is 6.9, 8.1, 8.4 and 8.4 for mAb 1-4 respectively. The mAbs were buffer exchanged to 20 mM HEPES, pH 7.6 including 10, 50 or 150 mM NaCl.
      AF4 was carried out on an Eclipse™ DualTec™ system controlled by VOYAGER CDS®, part of the VISION Chromatography Software Hub version 2.0.1.2 (Wyatt Technology Europe GmbH, Germany). The AF4 system was coupled to an Agilent 1200 series HPLC system (Agilent Technologies, Santa Clara, CA, USA), which consisted of a 1260 isocratic pump, UV detector and autosampler. An 18-angle static light scattering detector (DAWN™ HELEOSII™) and a differential refractive index detector (Optilab™ T-rEX) both from Wyatt Technology Corporation, Santa Barbara, CA, USA were coupled to the system. A short channel (effective length 145 mm) with a spacer height of 490 μm (W490) and regenerated cellulose membranes with a cutoff of 10 kDa (Wyatt Technology Europe GmbH, Germany) as bottom semi permeable wall were used. Focusing was performed with a cross-flow of 1.5 mL/min for 1 min. This was followed by separation at detector flow 1 mL/min and constant cross-flow of 2 mL/min for 22 min. Finally, a 5 min period without cross-flow and detector flow at 1 mL/min was applied to elute remaining sample. ASTRA® software version 7.3.1.9 was used to integrate the main peak, i.e. sample eluting during constant cross-flow of 2 mL/min and elution after cross-flow set to zero was not included (full end-to-end fractograms for the four samples with lowest recovery are shown in Figure S1 in Supplementary Information). The recovery was calculated using the UV 280 nm absorption trace, multiplying by the extinction coefficient and normalizing by the injected amount. Sample concentrations of 1 mg/ml was used and prepared by dilution from stock solutions. 20 mM HEPES, pH 7.6 including either 10, 50 or 150 mM sodium chloride was used as running buffer. Sample injection of 30 µg, i.e. 30 µL of 1 mg/mL was used in general, exceptions to this are stated in the text and figure legends.
      DLS data were acquired at 25°C by adding 30 μL of each sample in triplicates to a 384 well plate Corning® 3540 (Corning, NY, USA). Each well was measured 40 times with 5 s acquisition time. A DynaPro™ Plate Reader and Dynamics™ software version 7.8.0.26 (Wyatt Technology Corp., Santa Barbara, CA, USA) was used to collect and analyze the data. The interaction parameter, kD was calculated as the slope of the diffusion coefficient D versus concentration, divided by the D0 (the extrapolation of D to zero concentration).
      • Yadav S
      • Shire S J
      • Kalonia D S
      Viscosity behavior of high-concentration monoclonal antibody solutions: correlation with interaction parameter and electroviscous effects.
      Sample series of 0.5-20 mg/ml was prepared by dilution of stock solutions.

      Results and discussion

      We first investigated the difference between the mAb compounds, as well as the effect of different running buffers. In Fig. 1 overlaid fractograms of the four different mAbs, analyzed under identical conditions are presented. The mAbs display quite different profiles, i.e. elution time and peak shape, and mAb4 show obvious signs of overloading effects. mAb3 was then analyzed in running buffers containing various concentrations of sodium chloride, 10-150 mM. Again, the profiles are quite distinct and show a clear effect of ionic strength of the running buffer Fig. 2. The amount mAb injected for obtaining the results presented in Fig. 1 and Fig. 2 was 30 µg (30 µL of 1 mg/ml mAb). This amount was chosen to obtain a concentration close to the membrane that mimics the concentration regime normally used for mAb formulations, i.e. ≥ 100 mg/mL (a simulated concentration profile is shown in Figure S2 in Supplementary Information). The experimental results clearly show that AF4 can differentiate between compounds with different biophysical properties and capture the effect of the running buffer.
      Fig. 1
      Fig. 1Fractogram of mAb 1 (black), 2 (blue), 3 (green) and 4 (red) under identical conditions. Injection volume was 30 µL of sample concentration 1 mg/mL and 20 mM HEPES, pH 7.6 including 150 mM NaCl was used as running buffer.
      Fig. 2
      Fig. 2Fractogram of mAb3 in 3 different running buffers. Injection volume of 30 µL with concentration of 1 mg/ml using 20 mM HEPES, pH 7.6 including 150 mM NaCl (black), 50 mM NaCl (blue) and 10 mM NaCl (green) as running buffer.
      Next, we investigated the effect of sample concentration and sample buffer. In a compound screening workflow, it is highly preferably that the response is independent of these parameters. Fig. 3 panel A shows the results of analyzing mAb1 at loads of approximately 5, 10, 20 and 40 µg, using either constant injection volume (of various mAb concentrations) or various injection volumes (of constant mAb concentration). The fractograms are essentially identical in both cases, thus the total sample load is the determining factor and the focusing step normalizes the starting conditions. Fig. 3 panel B shows the results of analyzing mAb1 and mAb4 in the same running buffer, with the same sample load, but with different compositions of the sample buffer. It is seen that the results are virtually independent of the sample buffer. Hence the focusing step of the AF4 is providing a buffer exchange and hence determining the outcome of the fractogram and therefore initial sample buffer is not important.
      Fig. 3
      Fig. 3Robustness of AF4. A: mAb1 analyzed at 4 different sample loads. Full lines: 2 µl sample with concentrations 2.5, 5, 10 and 20 mg/ml. Dashed lines: 5, 10, 20 and 40 µl samples with concentration 1.0 mg/ml. B: mAb1 and mAb4 analyzed with different buffers in the applied sample. Full lines: 20 mM HEPES, pH 7.6, 10 mM NaCl. Dashed lines: 20 mM HEPES, pH 7.6, 150 mM NaCl. Sample load 30 µg. Running buffer 20 mM HEPES, pH 7.6, 150 mM NaCl.
      Finally, we considered how the complex shape of the fractogram can be reduced to one or a few parameters. In Fig. 4 we have plotted the recovery and retention time versus kD for mAb1-4 analyzed under different buffer conditions. Both parameters correlate will with kD. We favor using the recovery since it can be calculated even for very poor fractograms and has e.g. been suggested for accessing stability of IgG by AF4.
      • Ma D
      • Martin N
      • Tribet C
      • Winnik F M
      Quantitative characterization by asymmetrical flow field flow fractionation of IgG thermal aggregation with and without polymer protective agents.
      The retention time is clearly also very informative. Parameters such as peak asymmetry and broadness cannot be reliable calculated in cases of low recovery (data on the peak parameters discussed are presented in Table 1 in Supplementary Information). Low recovery and high retention time correlate with highly negative kD, indicative of strong protein-protein attraction. In other words, the AF4 analysis provide similar results as kD analysis, but with a single sample injection and with buffer exchange and concentration step as part of the AF4 method. In addition, the sample purity is expected to be less important, since the UV measurement is not as sensitive to large particles as DLS. Obviously the fractogram contains more information, but the full analysis is outside the scope of this short report.
      Fig. 4
      Fig. 4Recovery of AF4 (left) and peak retention time (right) versus kD. mAb1-4 in 20 mM HEPES, pH 7.6; 10, 50 or 150 mM NaCl. AF4 sample load 30 µg. mAb 1(black circles), mAb 2 (blue squares), mAb 3 (green triangles) and mAb 4 (red revered triangles).
      In conclusion, we present a novel application of AF4 for screening of biopharmaceuticals for protein-protein interactions at high concentrations. The method is straightforward to implement with standard AF4 instruments, robust towards variation in sample concentration and buffer and well adapted to practical compound screening workflows. The AF4 method correlates to the widely used interaction parameter kD obtained from DLS. Further efforts to characterize the correlation between AF4 and other developability parameters are under way in our laboratory.

      Acknowledgements

      The authors are grateful for the outstanding technical laboratory work carried out by Janni Larsen.

      Appendix. Supplementary materials

      References

        • Jarasch A
        • Koll H
        • Regula J T
        • Bader M
        • Papadimitriou A
        • Kettenberger H.
        Developability assessment during the selection of novel therapeutic antibodies.
        J Pharm Sci. 2015; 104: 1885-1898
        • Jain T
        • Sun T
        • Durand S
        • Hall A
        • Houston N R
        • Nett J H
        • Sharkey B
        • Bobrowicz B
        • Caffry I
        • Yu Y
        • Cao Y
        • Lynaugh H
        • Brown M
        • Baruah H
        • Gray L T
        • Krauland E M
        • Xu Y
        • Vásquez M
        • Wittrup KD
        Biophysical properties of the clinical-stage antibody landscape.
        PNAS. 2017; 114: 944-949
        • Wolf Pérez A-M
        • Sormanni P
        • Andersen J S
        • Sakhinini L I
        • Rodriguez-Leon I
        • Bjelke J R
        • Gajhede A J
        • De Maria L
        • Otzen D E
        • Vendruscolo M
        • Lorenzen N
        In vitro and in silico assessment of the developability of a designed monoclonal antibody library.
        mAbs. 2019; 11: 388-400
        • Connolly B D
        • Petry C
        • Yadav S
        • Demeule B
        • Ciaccio N
        • Moore J M R
        • Shire S J
        • Gokarn YR.
        Weak interactions govern the viscosity of concentrated antibody solutions: high-throughput analysis using the diffusion interaction parameter.
        Biophys J. 2012; 103: 69-78
        • Starr C G
        • Tessier P M
        Selecting and engineering monoclonal antibodies with drug-like specificity.
        Curr Opin Biotechnol. 2019; 60: 119-127
        • Kingsbury J S
        • Saini A
        • Auclair S M
        • Fu L
        • Lantz M M
        • Halloran K T
        • Calero-Rubio C
        • Schwenger W
        • Airiau C Y
        • Zhang J
        • Gokarn YR.
        A single molecular descriptor to predict solution behavior of therapeutic antibodies.
        Sci Adv. 2020; 6: eabb0372
        • Yadav S
        • Shire S J
        • Kalonia D S
        Viscosity behavior of high-concentration monoclonal antibody solutions: correlation with interaction parameter and electroviscous effects.
        J Pharm Sci. 2012; 101: 998-1011
        • Wu J
        • Schultz J S
        • Weldon C L
        • Sule S V
        • Chai Q
        • Geng S B
        • Dickinson C D
        • Tessier P M
        Discovery of highly soluble antibodies prior to purification using affinity-capture self-interaction nanoparticle spectroscopy.
        Protein Eng Des Sel. 2015; 28: 403-414
        • Liu Y
        • Caffry I
        • Wu J
        • Geng S B
        • Jain T
        • Sun T
        • Reid F
        • Cao Y
        • Estep P
        • Yu Y
        • Vásquez M
        • Tessier P M
        • Xu Y.
        High-throughput screening for developability during early-stage antibody discovery using self-interaction nanoparticle spectroscopy.
        MAbs. 2014; 6: 483-492
        • Wahlund K-G
        • Giddings J C
        Properties of an asymmetrical flow field-flow fractionation channel having one permeable wall.
        Anal Chem. 1987; 59: 1332-1339
        • Wahlund K-G.
        Chapter 18 Asymmetrical flow field-flow fractionation in.
        in: Schimpf M E Caldwell K Giddings J C. Field-Flow Fractionation Handbook. Wiley-Interscience, New York2000: 279-294
        • Wahlund K-G.
        Flow field-flow fractionation: critical overview.
        J Chromatogr A. 2013; 1287: 97-112
        • Yohannes G
        • Jussila M
        • Hartonen K
        • Riekkola ML
        Asymmetrical flow field-flow fractionation technique for separation and characterization of biopolymers and bioparticles.
        J Chromatogr A. 2011; 1218: 4104-4116
        • Fraunhofer W
        • Winter G.
        The use of asymmetrical flow field-flow fractionation in pharmaceutics and biopharmaceutics.
        Eur J Pharm Biopharm. 2004; 58: 369-383
        • Leeman M
        • Storm M U
        • Nilsson L.
        Practical applications of asymmetrical flow field-flow fractionation (AF4): a review.
        LCGC. 2015; (December): 642-651
        • Gigault J
        • Pettibone J M
        • Schmitt C
        • Hackley VA
        Rational strategy for characterization of nanoscale particles by asymmetric-flow field flow fractionation: a tutorial.
        Anal Chim Acta. 2014; 80
        • Cao S
        • Pollastrini J
        • Jiang Y.
        Separation and characterization of protein aggregates and particles by field flow fractionation.
        Curr Pharm Biotechnol. 2009; 10: 382-390
        • Caldwell KD
        • Brimhall S L
        • Gao Y
        • Giddings JC
        Sample overloading effects in polymer characterization by field-flow fractionation.
        J Appl Polym Sci. 1988; 36: 703-719
        • Arfvidsson C
        • Wahlund K-G.
        Mass overloading in the flow field-flow fractionation channel studied by the behaviour of the ultra-large wheat protein glutenin.
        J Chromatogr A. 2003; 1011: 99-109
        • Litzén A
        • Wahlund K-G.
        Effects of temperature, carrier composition and sample load in asymmetrical flow field-flow fractionation.
        J Chromatogr. 1991; 548: 393-406
        • Benincasa MA
        • Giddings JC
        Separation and molecular weight distribution of anionic and cationic water-soluble polymers by flow field-flow fractionation.
        Anal Chem. 1992; 64: 790-798
        • Benincasa MA
        Chapter 27 Synthetic polymer-water soluble in.
        in: Schimpf ME Caldwell K Giddings JC Field-Flow Fractionation Handbook. Wiley-Interscience, New York2000: 407-432
        • Moon MH
        • Park I
        • Kin Y.
        Size characterization of liposomes by flow-field flow fractionation and photon correlation spectroscopy, effect of ionic strength and pH of carrier solution.
        J Chromatogr A. 1998; 813: 91-100
        • Marioli M
        • Kok W T
        Recovery, overloading, and protein interactions in asymmetrical flow field-flow fractionation.
        Anal Bioanal Chem. 2019; 411: 2327
        • Wolf Pérez A M
        • Sormanni P
        • Andersen J S
        • Sakhnini L I
        • Rodriguez-Leon I
        • Bjelke J R
        • Gajhede A J
        • De Maria L
        • Otzen D E
        • Vendruscolo M
        • Lorenzen N
        In vitro and in silico assessment of the developability of a designed monoclonal antibody library.
        mAbs. 2019; 11: 388-400
        • Kopp M R G
        • Wolf Pérez A M
        • Zucca M V
        • Palmiero U C
        • Friedrichsen B
        • Lorenzen N
        • Arosio P
        An accelerated surface-mediated stress assay of antibody instability for developability studies.
        mAbs. 2020; 12e815995
        • Neergaard M S
        • Kalonia D S
        • Parshad H
        • Nielsen A D
        • Møller E H
        • van de Weert M.
        Viscosity of high concentration protein formulations of monoclonal antibodies of the IgG1 and IgG4 subclass - prediction of viscosity through protein-protein interaction measurements.
        Eur J Pharm Sci. 2013; 49: 400-410
        • Ma D
        • Martin N
        • Tribet C
        • Winnik F M
        Quantitative characterization by asymmetrical flow field flow fractionation of IgG thermal aggregation with and without polymer protective agents.
        Anal Bioanal Chem. 2014; 406: 7539-7547