Opalescence of an IgG1 Monoclonal Antibody Formulation is Mediated by Ionic Strength and Excipients

April 1, 2009

Can increase in ionic strength result in higher viscosity?

ABSTRACT

Opalescence is a phenomenon that has been observed in several commercially available monoclonal antibodies, both in liquid and reconstituted lyophilized formulations. In this article, we demonstrated that an increase in the ionic strength of a monoclonal antibody formulation (MAb1) led to opalescence and higher viscosity. When the ionic strength was reduced, no opalescence in the MAb1 formulation was observed. The removal of polysorbate-80 (PS-80) from the formulation resulted in an increase in opalescence in NaCl-containing formulations, whereas it had no effect on formulations lacking NaCl. Differential scanning calorimetry with MAb1 formulations containing increasing amounts of NaCl indicated that formulations with higher ionic strength present a lower apparent melting temperature. Opalescent MAb1 formulations placed on stability remained unchanged after four weeks at 4 °C, whereas at 45 °C, an increase in dimers was observed. Using multi-angle light scattering, the MAb1 formulation was found to have a negative second virial coefficient.

There are more than 23 therapeutic monoclonal antibodies that are commercially available for treating a wide variety of diseases.1 Of these products, several have an opalescent appearance either as liquids or following reconstitution of a lyophilized formulation.2 Opalescence is defined as exhibiting a play of colors like that of the opal, or having a milky iridescence.3

(Merck & Co.)

Factors influencing the opalescence of an IgG1 monoclonal antibody formulation include the concentration of the antibody as well as formulation components such as buffers, ionic strength, excipients, and pH. It was demonstrated that an increase in the concentration of the antibody and a decrease in temperature (5 °C) of an IgG1 formulation resulted in opalescence.4

It is well known that ionic strength can affect the behavior of proteins in solution. In most cases, at low and high salt concentrations either salting-in or salting-out occurs, respectively.5 Salting-in is observed when electrostatic interactions between the salt ions and charged residues of the protein are favorable.5 Salting-out occurs when the salt ions are excluded from the protein, which is mainly caused by unfavorable interactions between the salts and hydrophobic regions of the protein.5 There are examples, however, in which proteins are soluble at high salt concentrations.5–7

The ionic strength can also mediate protein–protein interactions. One example is with the protein b-lactoglobulin, which is predominantly monomeric at pH 3 in the absence of salt, but is dimeric in the presence of salt.8 Additionally, the type of salt may also affect the stability of the protein. Bovine serum albumin is stabilized against thermal unfolding with NaSCN and NaClO4, both kosmotropic salts, yet destabilized by chaotropic salts at high ionic strength.9,10 In other examples, aggregation was decreased in the presence of NaCl for both recombinant factor VIII SQ and recombinant keratinocyte growth factor, whereas aggregation increased in the presence of NaCl with recombinant human granulocyte colony stimulating factor.7,11

In the following article, we sought to determine whether ionic strength and excipients mediated the opalescence of an IgG1 formulation (MAb1). It was demonstrated that MAb1 formulations become opalescent as the ionic strength of the formulation is increased. Conversely, MAb1 formulations without salt lack opalescence. The second virial coefficient of MAb1 was negative. Stability studies indicated that opalescent MAb1 formulations have increased amount of irreversible dimers at elevated temperatures.

MATERIALS AND METHODS

Materials and Reagents

MAb1 is a fully human IgG1 monoclonal antibody. MAb1 was purified from Chinese hamster ovary (CHO) cells by Bioprocess Research and Development, Merck Research Laboratories (Whitehouse Station, NJ). The MAb1 formulation contains 24 mg/mL IgG1 in a formulation containing a buffer, NaCl, and polysorbate-80 (PS-80), pH 6. Polysorbate-80 was from Croda Incorporated (Mill Hall, PA). NaCl, KCl, MgCl2, KSCN, Na3PO4, CsCl, Na2SO4, hexamethylenetetramine, and hydrazine sulfate were from Sigma (St. Louis, MO). The filter used was a 0.22 μm Millex GV from Millipore (Billerica, MA).

METHODS

Opalescence Determination

Opalescence of the samples was assessed according to the European Pharmacopoeia (EP) 5.0 (2.2.1). Opalescence reference suspensions were made using hexamethylenetetramine and hydrazine sulfate. Samples were evaluated at a volume of 11 mL in 20 mL glass vials by comparing them to the reference suspensions in diffused daylight against a black background.

Turbidity Measurements

The opalescence of MAb1 was measured by subtracting the optical density (OD) at 350 nm from the OD at 550 nm using an HP spectrophotometer as previously described.12

Size Exclusion High Performance Liquid Chromatography

Size exclusion high performance liquid chromatography (SEC–HPLC) was performed using a TSKgel 3000SWXL column from Tosoh Corporation (Tokyo, Japan) with a mobile phase consisting of 25 mM phosphate, 0.3 M NaCl, pH 7.0. The flow rate was 0.5 mL/min. The temperature of the column was maintained at 25 °C. The samples were detected at 230 nm.

Dynamic Light Scattering (DLS)

The particle sizes of samples tested were measured by dynamic light scattering (DLS) using a Zetasizer Nano System from Malvern Instruments (Malvern, UK). A total of 10 measurements for each sample were taken and the results were reported as the Zave, the average hydrodynamic size.

Viscosity Measurements

Viscosities were measured using a LVDV-III Ultra Programmable Rheometer by Brookfield Engineering (Middleboro, MA). Sample volumes of 0.5 mL were tested for each measurement.

Determination of the Osmotic Second Virial Coefficient Values

The osmotic second virial coefficient was obtained using a Dawn Heleos multi-angle light scattering (MALS) instrument from Wyatt Technology Corporation (Santa Barbara, CA). The formulations containing buffer and 150 mM NaCl (no PS-80) were diluted and injected into the instrument as follows: 1:2.5 (8.20 mg/mL), 1:5 (4.10 mg/mL), 1:10 (2.05 mg/mL), and 1:20 (1.03 mg/mL). The Rayleigh equation was used by the computer software to generate data on the Zimm plot. The software also calculated the molecular weight and particle size of the sample.

The Rayleigh equation is described as follows (Malvern Instruments):13

(Equation 1)

in which C is the particle concentration (g cm-3 ), Rθ is the the Rayleigh ratio—the ratio of scattered light to incident light of the sample (cm-1), M is the sample molecular weight (g mol-1 ), A2 is the 2nd virial coefficient (cm3 mol g-2), P(θ) is the angular dependence of the sample scattering intensity, and K is the optical constant, which is calculated using Equation 1.1:

(Equation 1.1)

in which NA is Avogadro's constant (mol-1), λo is the Laser wavelength (cm), no is the solvent refractive index, and dn/dc is the the differential refractive index increment (cm3 g-1).

For particles that are smaller than the incident light (laser) used to measure the particle size, Equation 1 can be reduced to the linear form shown in the following:13

(Equation 2)

RESULTS

Visual Observations

MAb1 in a liquid formulation consisting of 24 mg/mL of IgG1 in a buffer, 150 mM NaCl, and PS-80, pH 6.0 in glass vials was inspected visually at 25 °C and was determined to be opalescent (Figure 1). The MAb1 formulation was equivalent to a Reference III standard [18–30 nephelometric turbidity units (NTU)] according to the EP (5.0, 2.2.1, Clarity and degree of opalescence of liquids). Following filtration of MAb1 through a 0.22-μm sterilizing filter, the opalescence remained unchanged, indicating it was not caused by insoluble aggregates (data not shown).

Figure 1

Particle Size

The particle size of MAb1 was evaluated beginning at a concentration of 20 mg/mL of IgG1 in a buffer, 150 mM NaCl, and PS-80, pH 6.0 in glass vials and was diluted in the same formulation to approximately 1 mg/mL (Figure 2). It was determined that MAb1 had a larger apparent hydrodynamic radius at higher concentrations, and that it decreased as the formulation was diluted.

Figure 2

SEC–HPLC was then performed on MAb1 at a concentration of 1 mg/mL. It was determined that the sample was mostly monomeric (>99%), with a small amount of dimer present (~1%) (Figure 3). In independent experiments, it was demonstrated that IgG monomers have a retention time of approximately 16 to 19 min on a TSKgel 3000SWXL column (data not shown).

Figure 3

Increase in Ionic Strength Correlates to an Increase in Opalescence

To determine if the ionic strength contributed to opalescence, NaCl (one of the excipients in the formulation) was tested in a range of 0 to 200 mM in a formulation consisting of 24 mg/mL IgG1 and PS-80, pH 6.0. A visual observation of the vials indicated that as the molarity of NaCl increased, the opalescence also increased (Figure 4). In the absence of NaCl, no opalescence was observed and correlated to at or below a Reference I standard (3–6 NTU) according to EP 5.0, 2.2.1. These results indicated that an increase in the ionic strength played a major role in contributing to the opalescent appearance of MAb1 (Figure 4).

Figure 4

Effect of Excipients on the Opalescence of MAb1

To determine if other components of the formulation influenced the opalescence of the MAb1 formulation, PS-80 was tested. Four formulations consisting of 24 mg/mL IgG1 with and without PS-80 and NaCl were generated in glass vials and are depicted in Figure 5. Opalescence was strongest in the formulation without PS-80 in the presence of 150 mM NaCl. This sample was more opalescent than the Reference III standard according to EP 5.0, 2.2.1. When PS-80 was added to this formulation, the opalescence was reduced (Figure 5). PS-80 had no effect on formulations that did not contain NaCl (Figure 5). The results indicated that PS-80 plays an important role in mediating the opalescence of MAb1 formulations containing NaCl.

Figure 5

Opalescence and Impact of Different Salts

It was next evaluated whether salts other than NaCl could affect the opalescence of MAb1. Different salts were selected based on their position in the Hofmeister series.14,15 Salts that were evaluated included KCl, MgCl2, KSCN, Na3PO4, CsCl, and Na2SO4. They were selected because they yield ions that are strong chaotropes, such as SCN¯, or strong kosmotropes, such as PO43-. Chaotropes have destabilizing effects and promote salting-in of proteins in solution, whereas kosmotropes have stabilizing and salting-out effects on proteins.14,15 Each of these salts were added to MAb1 at a concentration of 150 mM in the absence of PS-80. The IgG1 concentration was maintained at 24 mg/mL. The OD for each sample was measured and the results are depicted in Figure 6. It was determined that addition of these salts resulted in a similar level of opalescence as was observed with 150 mM NaCl (Figure 6). Ionic strength rather than the specific ion is therefore the major contributor to the opalescence of MAb1 (Figure 6).

Figure 6

Impact of Ionic Strength on Viscosity

To further explore the effect of ionic strength on the MAb1 formulation, the viscosity of formulations with increasing amounts of NaCl and concentration were evaluated. It was determined that the viscosity increased in a concentration-dependent manner (Figure 7). Additionally, as the ionic strength was increased with NaCl, the viscosity also increased (Figure 7). MAb1 at 100 mg/mL in 150 mM NaCl had a viscosity of ~20 cP (Figure 7). The most significant change in viscosity (~90 cP) occurred in the 100 mg/mL sample containing 500 mM NaCl (Figure 7).

Figure 7

Differential Scanning Calorimetry of MAb1 with Increasing Amounts of NaCl

Differential scanning calorimetry (DSC) was performed on MAb1 at a concentration of 1 mg/mL IgG1, PS-80, pH 6.0 with increasing amounts of NaCl. It was determined that the first transition (Tm1 = 69.7 °C) shifted to a lower temperature (67.1 °C) as the amount of NaCl was increased to 250 mM. The second transition (Tm2 = 88.1 °C) also decreased (85.5 °C) as the NaCl concentration was increased to 250 mM (Figure 8).

Figure 8

Stability Studies

It was also investigated whether the ionic strength had an effect on the stability of MAb1 formulations. MAb1 formulations at 24 mg/mL IgG1 containing increasing amounts of NaCl, a buffer, and PS-80 were placed on stability for four weeks at 4 °C and 45 °C. Samples were analyzed using SEC–HPLC at different time points. In formulations stored for four weeks at 4 °C, there were no changes detected (data not shown). After four weeks at 45 °C in the absence of NaCl, approximately 2.5% dimers were present (Figure 9). However, in samples containing 250 mM NaCl that were incubated for four weeks at 45 °C, the amount of dimers increased to approximately 5% (Figure 9). A systematic trend of increasing dimers was observed with an increase in NaCl concentration (Figure 9).

Figure 9

Osmotic Second Virial Coefficient of the MAb1 Formulation

The osmotic second virial coefficient is commonly indicated by the term A2 or B22. When the second virial coefficient is measured using light scattering, it reflects protein–protein interactions as well as contributions from protein–cosolute interactions and protein nonideality.16,17 A positive B22 value correlates to net repulsive forces between solute molecules, and negative B22 values indicate net attractive forces between solute molecules.16,18,19 Several methods can be used to measure the B22 value of a formulation, one of which is the use of multi-angle light scattering with dilutions of the formulation. This method was used to obtain second virial coefficient values for MAb1.

A Zimm plot of MAb1 in the formulation containing a buffer and 150 mM NaCl is depicted in Figure 10. PS-80 was removed from the formulation to prevent potential interference from micelles. It was determined that this formulation had a negative second virial coefficient (–4.3 x 10-5 mol mL/g2). The molecular weight calculated for MAb1 was 1.46 x 105 g/mol (146 kDa), which is close to the molecular weight of MAb1, which is 148 kDa.

Figure 10

DISCUSSION

The purpose of this study was to investigate whether the ionic strength and excipients in the MAb1 formulation were associated with the opalescence observed. We identified that the ionic strength of the MAb1 formulation played a major role in mediating opalescence.

Effects of Ionic Strength

The opalescence of MAb1 was demonstrated to increase with ionic strength. In the MAb1 formulation containing NaCl, opalescence was pronounced (Reference III opalescence standard) whereas in its absence, the solution was clear (Reference I opalescence standard). It was also observed that the apparent hydrodynamic radius of the MAb1 formulation was larger at higher concentrations (20 mg/mL) and lower at more dilute concentrations (1 mg/mL). The reason for the larger particle size at higher concentrations of MAb1 is not known, and will require further investigation.

We explored whether salts other than NaCl had an impact on opalescence. Salts were selected based on their position in the Hofmeister series, and included KCl, MgCl2, KSCN, Na3PO4, CsCl, and Na2SO4.14,15 Opalescence was observed in the presence of all salts tested, and no specific ion effects were observed.

As the concentration of NaCl was increased, the electrostatic shielding becomes saturated and it is possible that interactions between the hydrophobic regions of MAb1 become dominant. An example of electrostatic shielding and enhancement of hydrophobic effects in the presence of NaCl was demonstrated in studies using lysozyme.5,20

There are other examples of the effect of ionic strength on proteins. It was demonstrated that when recombinant factor VIII SQ was placed in a solution containing 100 mM NaCl at pH 7, the solution turned opalescent within the first hour.7 However, after the first hour, the protein precipitated out of solution.7 At higher concentrations of NaCl, the precipitate could be dissolved, and activity was observed to recover.7

The interactions between MAb1 molecules described in these studies were not strong enough to precipitate the MAb. It is not known, however, if the increase in opalescence correlates to molecular self-association of MAb1 molecules. Sukumar and colleagues have attributed opalescence of an IgG1 formulation to Rayleigh scatter and indicated that opalescence is not caused by noncovalent association.4

Effects of Excipients

PS-80 has been widely used in liquid parenteral formulations to reduce aggregation caused by surface interactions, to prevent denaturation at the air–liquid interface, and to reduce agitation and temperature-induced aggregation.12,21 In this study, it was demonstrated that PS-80 was able to reduce the opalescence of the MAb1 formulation containing 150 mM NaCl. It is possible that PS-80 partially disrupted the hydrophobic–hydrophobic interactions that are dominant following the initial shielding of electrostatic charges.

Effects of Viscosity

We also explored the effect concentration and NaCl had on the viscosity of MAb1 solutions. It was determined that the viscosity increased as the protein concentration increased. Similar results were obtained by Liu and colleagues for an IgG at a concentration range of 30 to 125 mg/mL in a formulation containing 16 mM histidine, 266 mM sucrose, and 0.03% PS-20, pH 6.22 The viscosity of MAb1 solutions also increased with NaCl concentration. The increase in viscosity of MAb1 solutions with increasing NaCl, however, was in contrast to the results obtained with an IgG by Liu and colleagues.22 In their studies, increasing the molarity of NaCl resulted in a decrease in viscosity.22 The differences in the viscosity increase or decrease in the presence of NaCl when comparing MAb1 to the IgG studied by Liu and colleagues may be caused by sequence-specific variability in the complementary determining regions (CDRs) in the heavy and light chains of the molecules.

Differential Scanning Calorimetry

Salts can affect proteins by several mechanisms and can either increase or decrease thermal stability.23 The dominant mechanism of the interaction of salts with proteins and the amplitude of the stabilizing or destabilizing effect depend on the type and concentration of the salt, but also on other solvent conditions.24

The thermal unfolding of MAb1 presents two main transitions. Based on the amplitude of these transitions, the first transition corresponds to the unfolding of the CH2 domain in the Fc fragment and the second transition corresponds to the unfolding of the Fab fragment and CH3 domain in the Fc fragment of the IgG molecule.25 The apparent melting temperature for both transitions is slightly decreased as the concentration of the salt is increased. In the absence of salt, the second transition presents a shoulder that may reflect a reduced overlap between CH3 and Fab unfolding. This observation suggests that the Fab fragment is more sensitive to the presence of salt than the CH3 domain.

Second Virial Coefficient

In the presence of NaCl, MAb1 had a negative B22. It is not known if the negative second virial coefficient of MAb1 was associated with the opalescence observed or was caused by the properties of the MAb. It could be hypothesized that net attractive interactions are associated with opalescence, however additional studies will be required to test the idea. It would also be interesting to evaluate MAb1 formulations lacking NaCl to determine if the second virial coefficient changes.

Studies with other proteins have shown that the second virial coefficient decreases as the concentration of NaCl increases.26 For example, it was demonstrated that the second virial coefficient of the peptide enfuviritide decreases with an increase in salt.26 In other studies, including one with an IgG, it was demonstrated that the second virial coefficient decreases as the concentration of NaCl increases.22,27

Stability of Opalescent IgG Formulations

It was determined that the ionic strength of the MAb1 formulation does not affect the stability at 4 °C after four weeks. However, at elevated temperatures for four weeks in the presence of NaCl, irreversible dimer formation occurs, as demonstrated by SEC–HPLC. Interestingly, there was no significant change in high-order aggregate formation as the amount of NaCl was increased (data not shown).

The stability of an opalescent IgG1 formulation has been previously described by Sukumar and colleagues. In those studies, it was determined that the hydrodynamic size of formulations ranging from 0.5 to 50 mg/mL at 25 °C remained unchanged up to 300 minutes.4

An example in which NaCl induced dimerization of an IgG occurred was with the monoclonal antibody rhuMAb (VEGF).28 In the presence of 1M NaCl, a decrease in the kD (the dissociation constant) for rhuMAb (VEGF) was observed, which resulted in the formation of dimers.28 In the same study, it was demonstrated that another salt, CaCl2, had the same effect as NaCl, in that the increased ionic strength resulted in dimerization of rhuMAb (VEGF).28 These results parallel the study described here, in which higher levels of NaCl result in dimer formation. In separate studies with an IgG, it was demonstrated that the percentages of IgG aggregates increased with increasing ionic strength.29

Overall, the studies reported here demonstrate that the ionic strength of a formulation as well as the excipients, play an important role in the opalescent appearance of an IgG1 monoclonal antibody. These factors may be relevant to the appearance and stability of other IgG formulations under stressed conditions.

ACKNOWLEGDEMENTS

Craig McKelvey and Yang Wang are acknowledged for helpful discussions.

Ning Wang is a research biochemist, Binghua Hu is a research biochemist, Roxana Ionescu is a research fellow, Henryk Mach is a senior investigator, when this article was written, Joyce Sweeney was a senior investigator, Christopher Hamm is a research biochemist, Marc J. Kirchmeier is an associate director, and Brian K. Meyer is a research fellow, all in Bioprocess Analytical and Formulation Sciences, Bioprocess Research and Development, at Merck Research Laboratories, Merck & Co., West Point, PA, 215.652.3992, brian_meyer@merck.com

REFERENCES

1. Wang W, Singh S, Zeng DL, King K, Nema S. Antibody structure, instability, and formulation. J Pharm Sci. 2007 Jan;96(1):1–26.

2. Physicians' Desk Reference, 60th Edition. 2007. Montvale NJ, Medical Economics.

3. Webster's Encyclopedic Unabridged Dictionary. 1989. New York, NY, Dilithium Press, Ltd.

4. Sukumar M, Doyle BL, Combs JL, Pekar AH. Opalescent appearance of an IgG1 antibody at high concentrations and its relationship to noncovalent association. Pharm Res. 2004;21:1087–93.

5. Curtis RA, Prausnitz JM, Blanch HW. Protein–protein and protein–salt interactions in aqueous protein solutions containing concentrated electrolytes. Biotechnol Bioeng. 1998;57:11–21.

6. Arakawa T, Bhat R, Timasheff SN. Preferential interactions determine protein solubility in three-component solutions: the MgCl2 system. Biochem. 1990;29:1914–23.

7. Fatouros A, Osterberg T, Mikaelsson M. Recombinant factor VIII SQ—inactivation kinetics in aqueous solution and the influence of disaccharides and sugar alcohols. Pharm Res. 1997;14:1679–84.

8. Sakurai K, Oobatake M, Goto Y. Salt-dependent monomer-dimer equilibrium of bovine beta-lactoglobulin at pH 3. Protein Sci. 2001;10:2325–35.

9. Chi EY, Krishnan S, Randolph TW, Carpenter JF. Physical stability of proteins in aqueous solution: mechanism and driving forces in nonnative protein aggregation. Pharm Res. 2003;20:1325–36.

10. Yamasaki M, Yano H, Aoki K. Differential scanning calorimetric studies on bovine serum albumin: II. Effects of neutral salts and urea. Int J Biol Macromol. 1991;13:322–8.

11. Chi EY, Krishnan S, Kendrick BS, Chang BS, Carpenter JF, Randolph TW. Roles of conformational stability and colloidal stability in the aggregation of recombinant human granulocyte colony-stimulating factor. Protein Sci. 2003;12:903–13.

12. Mahler HC, Muller R, Friess W, Delille A, Matheus S. Induction and analysis of aggregates in a liquid IgG1-antibody formulation. Eur J Pharm Biopharm. 2005;59:407–17.

13. Molecular Weight Theory. Zetasizer Nano Series Manual 0317. Chapter 15; 2008. Malvern UK, Malvern Instruments.

14. Broering JM, Bommarius AS. Evaluation of Hofmeister effects on the kinetic stability of proteins. J Phys Chem B. 2005;109:20612–9.

15. Zhang Y, Cremer PS. Interactions between macromolecules and ions: The Hofmeister series. Curr Opin Chem Biol. 2006;10:658–63.

16. Alford JR, Kendrick BS, Carpenter JF, Randolph TW. Measurement of the second osmotic virial coefficient for protein solutions exhibiting monomer-dimer equilibrium. Anal Biochem. 2008;377:128–33.

17. Winzor DJ, Deszczynski M, Harding SE, Wills PR. Nonequivalence of second virial coefficients from sedimentation equilibrium and static light scattering studies of protein solutions. Biophys Chem. 2007;128:46–55.

18. Bajaj H, Sharma VK, Badkar A, Zeng D, Nema S, Kalonia DS. Protein structural conformation and not second virial coefficient relates to long-term irreversible aggregation of a monoclonal antibody and ovalbumin in solution. Pharm Res. 2006;23:1382–94.

19. George A, Wilson WW. Predicting protein crystallization from a dilute solution property. Acta Crystallogr D Biol Crystallogr. 1994;50:3–5.

20. Valente JJ, Verma KS, Manning MC, Wilson WW, Henry CS. Second virial coefficient studies of cosolvent-induced protein self-interaction. Biophys J. 2005;89:4211–8.

21. Shi L, Sanyal G, Ni A et al. Stabilization of human papillomavirus virus-like particles by non-ionic surfactants. J Pharm Sci. 2005;94:1538–51.

22. Liu J, Nguyen MD, Andya JD, Shire SJ. Reversible self-association increases the viscosity of a concentrated monoclonal antibody in aqueous solution. J Pharm Sci. 2005;94:1928–40.

23. Baldwin RL. How Hofmeister ion interactions affect protein stability. Biophys J. 1996;71:2056–63.

24. Ionescu RM, Shi L. Thermal unfolding of proteins. The encylopedia of industrial biotechnology: bioprocess, bioseparation, and cell technology. 2008. John Wiley & Sons, Inc. In Press.

25. Ionescu RM, Vlasak J, Price C, Kirchmeier M. Contribution of variable domains to the stability of humanized IgG1 monoclonal antibodies. J Pharm Sci. 2008;97:1414–26.

26. Payne RW, Nayar R, Tarantino R et al. Second virial coefficient determination of a therapeutic peptide by self-interaction chromatography. Biopolymers. 2006;84:527–33.

27. Rosenbaum DF, Zukoski CF. Protein interactions and crystallization. J Crystal Growth. 1996;169:752–8.

28. Moore JM, Patapoff TW, Cromwell ME. Kinetics and thermodynamics of dimer formation and dissociation for a recombinant humanized monoclonal antibody to vascular endothelial growth factor. Biochem. 1999;38:13960–7.

29. Kameoka D, Masuzaki E, Ueda T, Imoto T. Effect of buffer species on the unfolding and the aggregation of humanized IgG. J Biochem. 2007;142:383–91.