Examination of Thermal Unfolding and Aggregation Profiles of a

Feb 16, 2015 - Screening for pharmaceutically viable stability from measurements of thermally induced protein unfolding and short-term accelerated str...
0 downloads 11 Views 5MB Size
Download PDF
Article pubs.acs.org/molecularpharmaceutics

Examination of Thermal Unfolding and Aggregation Profiles of a Series of Developable Therapeutic Monoclonal Antibodies Mark L. Brader,* Tia Estey, Shujun Bai,† Roy W. Alston, Karin K. Lucas,‡ Steven Lantz, Pavel Landsman, and Kevin M. Maloney Protein Pharmaceutical Development, Biogen Idec, 14 Cambridge Center, Cambridge, Massachusetts 02142, United States S Supporting Information *

ABSTRACT: Screening for pharmaceutically viable stability from measurements of thermally induced protein unfolding and short-term accelerated stress underpins much molecule design, selection, and formulation in the pharmaceutical biotechnology industry. However, the interrelationships among intrinsic protein conformational stability, thermal denaturation, and pharmaceutical stability are complex. There are few publications in which predictions from thermal unfolding-based screening methods are examined together with pharmaceutically relevant long-term storage stability performance. We have studied eight developable therapeutic IgG molecules under solution conditions optimized for largescale commercial production and delivery. Thermal unfolding profiles were characterized by differential scanning calorimetry (DSC) and intrinsic fluorescence recorded simultaneously with static light scattering (SLS). These molecules exhibit a variety of thermal unfolding profiles under common reference buffer conditions and under individually optimized formulation conditions. Aggregation profiles by SE-HPLC and bioactivity upon long-term storage at 5, 25, and 40 °C establish that IgG molecules possessing a relatively wide range of conformational stabilities and thermal unfolding profiles can be formulated to achieve pharmaceutically stable drug products. Our data suggest that a formulation design strategy that increases the thermal unfolding temperature of the Fab transition may be a better general approach to improving pharmaceutical storage stability than one focused on increasing Tonset or Tm of the first unfolding transition. KEYWORDS: developability, developable, stability, antibody therapeutics, thermal denaturation, protein aggregation, protein formulation, high-throughput screening

■

INTRODUCTION Monoclonal antibodies (mAbs) are a prominent class of therapeutics constituting the overwhelming majority of protein drugs currently in development. There are more than 30 mAb therapeutics approved for human use,1,2 and the emergence of biosimilar3 mAbs and more exotic mAb-based constructs such as bispecifics4 and single-domain molecules appears to be imminent.5 mAbs are extensively engineered to achieve favorable binding affinities, pharmacokinetics, effector function, immunogenicity, and pharmaceutical properties, including stability and solubility.6−8 mAbs are susceptible to a variety of degradation mechanisms, of which aggregation is one of the most pervasive.9,10 To achieve therapeutic viability, proteins must be robust to the multitude of stresses encountered during production, fillfinish, long-term storage, shipping, in-use handling, and administration.11 Poor protein stability can limit the therapeutic viability of an otherwise medically significant product. For many mAbs in the biotech pipeline, a minimal target product profile will require solution stability of at least 2 years at 5 °C plus a subsequent period of in-use stability ranging from several hours to several weeks at room temperature. Stabilization of the molecule against interfacial damage, freeze−thaw, shear, © 2015 American Chemical Society

agitation, temperature excursion, and photodegradation is also required. Prior to clinical candidate selection, candidate molecules should be screened for their likelihood of fulfilling manufacturing, dosing, and product stability requirements (developability). Significant therapeutic and commercial benefits can be realized if more favorable stability can be achieved: these include an improved safety profile, longer shelf life, greater supply chain flexibility, and less restrictive drug handling and administration requirements for the patient. The putative connection between protein conformational stability and pharmaceutical stability continues to play a major role in guiding protein engineering and formulation design efforts for therapeutic proteins. This approach is based on the premise that conditions that favor the native folded state of the protein will most effectively inhibit degradation pathways that proceed via unfolded or partially unfolded intermediates.12 Despite awareness that temperature-induced protein aggregation is often non-Arrhenius over even relatively small Received: Revised: Accepted: Published: 1005

November 6, 2013 January 23, 2015 February 16, 2015 February 16, 2015 DOI: 10.1021/mp400666b Mol. Pharmaceutics 2015, 12, 1005−1017

Article

Molecular Pharmaceutics temperature ranges relevant to drug product development13 and that rank order obtained at denaturation temperatures might not correspond to that at lower temperatures,14−18 automated high-throughput techniques that monitor protein conformation while applying a thermal ramp have become fashionable tools in the biotech industry’s quest for more efficient molecule design, developability assessment, and expanded formulation design space.19−23 Such methods measure onset (Tonset) and midpoint (Tm) temperatures of protein thermal unfolding transitions, thereby providing a convenient means to rank samples.20,22,24−27 Studies based on elevated temperatures and thermal denaturation screens have been applied extensively to support the development of mAb therapeutics. mAbs are multidomain proteins that have been shown to exhibit a wide range of conformational stabilities even within a given subclass.28 Engineering mAbs to increase domain stabilities indicated by Tm and thereby improve overall pharmaceutical stability is an established tenet of therapeutic antibody design and optimization.6 Isolated immunoglobulin heavy chain variable domains and isolated CH2 domains have been found to be highly aggregation prone; however, sequence modification resulting in greater thermal stability measured by Tm has been shown to significantly improve aggregation resistance.29 Case studies on mAbs are reported where screening for effects on the thermally induced unfolding profile has been central to identifying favorable formulation conditions.30−36 Several of these studies have used Tm values to screen and identify favorable excipients.32,35−37 For example, a strong inverse correlation between aggregation rate at 40 °C versus Tm for various excipient and formulation conditions was reported for an IgG1 mAb.32 Bajaj et al.38 examined in detail the significance of conformational versus colloidal stability indicated by the second virial coefficient (B22). These authors concluded that protein structural conformation and not B22 related to the longterm irreversible aggregation of an IgG2 mAb. Another study examined the stability of eight mouse mAbs and concluded that DSC was predictive of accelerated and long-term storage for six out eight of the mAbs studied.39 Cases in which conformational stability did not clearly predict physical stability of mAbs have also been reported. Matheus et al.40 found that Tm did not necessarily correspond to storage stability at 40 °C, measured by SE-HPLC and SDS-PAGE, for an IgG1 mAb. Fesinmeyer et al.41 identified significant differences in the agitation-induced and 45 °C accelerated stability of three IgG2 mAbs, yet these proteins were not readily differentiated by their Tm(1) values. Moreover, numerous studies have examined mAb aggregation kinetics and conformational stability under harsh conditions of acidic pH values (pH 3−4) and/or forced degradation temperatures in the range 45−73 °C.15,18,33,34,37,41−44 The use of very high temperature stress (viz. 65 °C) for identifying favorable mAb excipient conditions has occasionally been demonstrated.37,45 Although it is stated frequently that enhanced conformational stability should lead to decreased aggregation rates upon long-term storage, there are few published reports that have tested this hypothesis directly using pharmaceutically authentic stability protocols and formulation compositions.26 Consequently, there is a limited frame of reference to relate mAb accelerated screening data to long-term storage stability and pharmaceutical robustness. Despite the widespread application and apparent success of accelerated temperature- and Tm-based screening methods as a practical development approach, the connection between

conformational stability parameters obtained from thermal unfolding experiments and the pharmaceutical storage stability of mAbs seems to be ambiguous. In view of (i) the lack of published literature for mAbs reporting both the thermal unfolding parameters and corresponding pharmaceutical storage stability data, (ii) the questionable connection between stability toward thermal denaturation and stability at pharmaceutical storage temperatures, and (iii) apparent ambiguity regarding the practical utility of thermal unfolding-based methods applied to mAbs, we were prompted to examine the thermal unfolding and longterm aggregation profiles of a series of mAbs established as pharmaceutically developable molecules. We designed stable, pharmaceutically robust formulations, conducted long-term stability studies, and then characterized retrospectively the thermal unfolding profiles using DSC and intrinsic fluorescence collected simultaneously with static light scattering (SLS). Here, we report these thermal unfolding profiles recorded under common screening buffer (or reference formulation) conditions and under formulation conditions optimized individually and independently for each molecule.

■

MATERIALS AND METHODS Monoclonal Antibodies. The monoclonal antibodies used in this study were produced at Biogen Idec (Cambridge, MA). mAbs1−6 are IgG1 and mAbs7−8 are IgG4 subtypes. mAb6 and mAb7 are aglycosylated. For DSC and Optim1000 thermal scanning studies, mAb samples were prepared in a single common screening buffer and, respectively, in eight individually optimized formulation buffers. These solutions were prepared by dialysis using a 10 kDa NMWCO membrane at 2−8 °C for at least 24 h and were then diluted to 1.0 mg/mL using the corresponding dialysis buffers. The composition of the screening buffer was 20 mM sodium citrate, 150 mM NaCl at pH 6.0. Formulation excipients, buffers, and salts were USP grade (or higher) and purchased from Sigma-Aldrich. The screening buffer composition and the DSC data collection procedure used herein were selected to correspond to the conditions under which DSC data for 17 monoclonal antibodies were reported previously.28 The compositions of the individually optimized formulations are not disclosed, but they comprised typical generally recognized as safe (GRAS) pharmaceutical excipients. The term optimized formulation is used herein to denote that each formulation was designed to confer acceptable overall stability with respect to all stresses and degradation products applicable to a commercial therapeutic mAb. The optimized formulations were designed independently of one another and did not follow a common formulation design paradigm. Differential Scanning Calorimetry. Differential scanning calorimetry measurements were performed using a MicroCal capillary VP-DSC system (Northampton, MA) with Origin VPViewer2000, version 2.0.64, controlling software. Sample measurements in the DSC were made at a concentration of 1.0 mg/mL total protein. The thermograms were generated by scanning the temperature from 20 to 95 °C at a rate of 1 °C/ min using the medium feedback mode. Multipeak fitting/ deconvolution of the DSC thermograms and calculation of unfolding transition points (Tm values) as deconvoluted peak maxima were performed using Origin 7 software. The number of peaks in the deconvolution model was increased from 2 peaks until a satisfactory fit was reached as estimated by the chisquare test. The numeric data for the DSC thermograms and 1006

DOI: 10.1021/mp400666b Mol. Pharmaceutics 2015, 12, 1005−1017

Article

Molecular Pharmaceutics

Figure 1. Thermal unfolding profiles for four mAbs, mAb1, mAb2, mAb3, and mAb4, measured in a screening buffer (I) and in individually optimized formulations (II). The upper panels for each molecule show the measured DSC thermal unfolding profile (black) and the mathematically deconvoluted component transitions. The lower panels for each molecule show the thermal unfolding profiles detected by intrinsic fluorescence (purple) and static light scattering (brown). The vertical dashed lines indicate the thermal transition midpoints, and the vertical solid lines indicate the transition onsets.

order to achieve satisfactory nonlinear fits (estimated by their R2 values). In cases where multiple transitions were apparent from the Optim-1000 thermograms, the thermoscan data sets were divided into the corresponding subsets to apply satisfactory sigmoid fits to the individual transitions in a multitransition thermal unfolding/aggregation model. The midpoint temperatures of the thermal transitions were then obtained as solutions of the respective sigmoid fit equations at 1/2 × (Ymax − Ymin). The onset temperatures of the first thermal fluorescence or light scattering transitions were estimated as solutions of the respective sigmoid fit equations at 10% × (Ymax − Ymin), where Y is the corresponding output value (either scatter intensity or F350/330); Ymin is the saturation output value at the respective sigmoid start plateau and Ymax, the value at the sigmoid end plateau. Four run repeats of the same sample array at the same thermal scan rate demonstrated high precision and accuracy of measurements, with run-to-run variability of calculated onsets and midpoints not exceeding ±2 °C for the test mAb samples (not shown). Isothermal Storage Stability Studies. Formulation stability studies were conducted independently for each of the eight mAbs using type I glass vials with elastomeric stoppers in a variety of container closure and fill volume configurations. Vials were autoclaved and filled under sterile conditions representative of commercial drug product manufactured under cGMP (current good manufacturing practice) conditions. Samples at respective formulation concentrations were incubated in stability chambers maintained at 5 ± 3 and 25 ± 2 °C with 60% relative humidity and 40 ± 2 °C with 75% relative humidity and protected from light. SE-HPLC conditions (mobile phase, column, buffer, and run parameters) were

their deconvoluted components were then exported from Origin into Excel and SigmaPlot 12.3 software in order to perform Gaussian fitting of the first deconvolution component of each thermogram. This enabled estimation of the first thermal transition onset (Tonset), defined as 10% of (CPmax − CPmin), where CPmax is CP peak value and CPmin is the background value at the thermogram start. Intrinsic Fluorescence and SLS Thermal Scanning Measurements. The Optim-1000 instrument (Pall Corporation, Port Washington, NY) was used to perform thermal scans in fluorescence and static light scattering output modes. In a typical run, two multiple cuvette arrays (MCAs) of 16 cuvettes each and of 9 μL cuvette capacity were loaded with 1.0 mg/mL sample solutions. The total array of 16 duplicate samples was formed with the eight mAbs under two assay solution conditions (screening buffer and optimized formulation). The thermal ramping was then performed simultaneously for all array samples in a single 2 h run from 35 to 95 °C, at a scan rate of 30 °C per hour. Thermal scan data from protein intrinsic fluorescence and light scattering outputs were thus acquired during the single run at an excitation laser wavelength 266 nm. Upon the acquisition, the raw spectral data from the instrument were preprocessed into the thermoscan light scattering and fluorescence emission intensity ratio 350/330 nm (F350/330) output modes using the Primary Analysis suite of Optim-1000 analytical software (Avacta, York, UK). For further processing and analysis, the preprocessed data were then exported in numeric form into Microsoft Excel. For nonlinear (sigmoid) fits and calculations of thermal inflection temperatures (i.e., midpoint (Tm) and onset (Tonset) temperatures), the data were further exported into SigmaPlot 12.3 software in 1007

DOI: 10.1021/mp400666b Mol. Pharmaceutics 2015, 12, 1005−1017

Article

Molecular Pharmaceutics

Figure 2. Thermal unfolding profiles for four mAbs, mAb5, mAb6, mAb7, and mAb8, measured in a screening buffer (I) and in an individually optimized formulation (II). The upper panels for each molecule show the measured DSC thermal unfolding profile (black) and the mathematically deconvoluted component transitions. The lower panels for each molecule show the thermal unfolding profiles detected by intrinsic fluorescence (purple) and static light scattering (brown). The vertical dashed lines indicate the thermal transition midpoints, and the vertical solid lines indicate the transition onsets.

conditions, whereas four components were required to fit mAb4 and mAb7. A satisfactory three-component fit was obtained for mAb6 under the screening buffer conditions, whereas a four-component fit was required under the optimized formulation conditions. Thermal Unfolding Monitored by Intrinsic Fluorescence. Thermograms generated by monitoring F350/330 as a function of temperature are shown in the lower panels of Figures 1 and 2 (purple curves). Transition midpoints and onsets are given in Table 1. The basis of this detection method is the change in polarity of the microenvironment around the aromatic amino acids (predominantly tryptophan) as the protein unfolds, which leads to a change in the average emission wavelength of the fluorescence. Typically, a red shift occurs from 330 nm in a protein where a tryptophan is fully buried to 350 nm where it is exposed completely to water,46 resulting in an increase in F350/330. This red-shift behavior was indeed observed for seven of the eight mAbs studied here; however, mAb4 (Figure 1) exhibited a decrease in F350/330 upon heating at 70 °C followed by an increase upon further heating. For several of the mAbs, more than one F350/330 transition was resolved and fit to sigmoid curves. Simultaneous Monitoring by SLS and Intrinsic Fluorescence. The Optim1000 instrument provides the capability to record simultaneously the intrinsic fluorescence and static light scattering (SLS) outputs in a single run, which are presented together with the corresponding DSC thermograms in Figures 1 and 2. The lower panels of Figures 1 and 2 (orange curves) show the SLS signals for each protein over the course of the thermal ramp. For some of these samples (mAb1, mAb2, mAb5, mAb6), the SLS data could be fit with a single

used as per assay protocols developed individually for the respective mAb products (not disclosed herein). The SE-HPLC profiles were analyzed for relative content (%) of high molecular weight (HMW) soluble aggregates in the mAb product. The latter was determined as area-under-the-curve of the respective high molecular weight (HMW) peak relative to that of the main peak of the chromatogram assigned to the mAb product. Biological (pharmaceutical) activity of each product was assessed in the individual assay targeting the pharmaceutically relevant property of the product, typically a binding assay against the relevant model target. The individual assay protocols used in these assessments are not described herein.

■

RESULTS Differential Scanning Calorimetry. DSC thermograms measured for the eight mAbs are shown in the upper panels of Figures 1 and 2 together with the mathematical deconvolution of each profile into Gaussian components. Each mAb was evaluated under two solution conditions: screening buffer conditions (20 mM citrate, 150 mM NaCl, pH 6.0) and optimized formulation conditions, which refers to a formulation optimized specifically for each molecule to confer acceptable overall pharmaceutical stability. Thermograms recorded in the screening buffer are shown in the left panels (labeled I), and those recorded in optimized formulations are shown on the right panels (labeled II). Tm values obtained from the Gaussian components are listed in Table 1 together with transition onset values corresponding to the first thermal transition (Tonset). For mAb1, mAb2, mAb3, mAb5, and mAb8, satisfactory fits using three components were obtained under both formulation 1008

DOI: 10.1021/mp400666b Mol. Pharmaceutics 2015, 12, 1005−1017

Article

Molecular Pharmaceutics

Table 1. Thermal Unfolding Transition Onsets (Tonset) and Midpoints (Tm) Measured in Screening Buffer (20 mM Citrate, 150 mM NaCl, pH 6.0) and under Individually Optimized Formulation Conditions Using DSC, Intrinsic Fluorescence Emission Ratio at 350/330 nm (F350/330), and Scattering at 266 nm (Sc266)a

a

Standard deviations from duplicate runs are shown. Values without standard deviations were taken from a single run. Green-shaded cells indicate DSC transitions assigned tentatively to the Fab region. Aggregation rates (Agg rate) from linear fit to SE-HPLC measurements of aggregation upon storage at 40 and 25 °C are given in percent aggregate per month.

observation is supported by the corresponding DSC data that shows no enthalpy detected for the first fluorescence transition. These results are in accord with previous studies of IgG molecules that have found that the least conformationally stable regions of a multidomain protein are not necessarily the most aggregation-prone.47 This demonstrates the value of the SLS signal to help pinpoint which domain needs attention from a protein engineering or formulation optimization viewpoint. The Optim1000 instrument provides a convenient way to generate this data using a small quantity of protein. Pharmaceutical Storage Stability. It is beyond the scope of this article to present stability data for the entire set of pharmaceutical quality attributes. Nevertheless, aggregation and biological activity are critical quality attributes associated closely with the integrity of protein higher-order structure and therefore are central to overall pharmaceutical stability. Aggregation time-course profiles measured by SE-HPLC of the eight mAbs in their individually optimized formulations stored quiescent in glass vials at 5, 25, and 40 °C are shown in Figure 3. Overall, these data show that each of these molecules exhibits adequate pharmaceutical stability characteristics with respect to aggregation during long-term storage. After 2 years at 5 °C and after 6 months at 25 °C, aggregate levels for all of these molecules increase by less than 2%. An inspection of linear fits to these data indicate that the 40 °C data conform quite well to zero-order kinetics, as do the 25 °C data (although less so than the 40 °C data). Aggregation rates calculated as monthly degradation rates from the slopes of linear fits are

transition, and for others (mAb3, mAb4, mA7, mAb8), two or three transitions were apparent. Because SLS is sensitive to apparent molecular weight, the rapid increase in SLS at the transition onset indicates that aggregation begins to occur at that temperature under the specific conditions of this experimental measurement. This signal adds an orthogonal indication of thermal stability such that aggregation and spectral microenvironments of the intrinsic fluorophores are probed simultaneously as the protein unfolds. Because these signals correspond to distinct structural and kinetic phenomena, the transitions detected by each method are not expected to coincide. From close inspection of Figures 1 and 2, it is apparent that many of these molecules show significant spectroscopic perturbations at temperatures where no increase in SLS has occurred. Similarly, comparison of the SLS thermal profiles with the DSC profiles shows that for certain molecules a domain unfolds before an increase in SLS signal occurs (e.g., mAb2 and mAb7). These observations highlight the value of recording the SLS signal, as it is possible to distinguish spectroscopic- and DSC-detected transitions that may not be associated with an aggregation event (e.g., the unfolding of an aggregation-relevant mAb domain). One example of this is mAb4, which shows an initial intrinsic fluorescence transition over the range of ∼45−60 °C with no concomitant increase in SLS. However, the second fluorescence transition (Tm(2) = 66.4 °C, screening buffer) corresponds closely to first the SLS transition (Tm(1) = 67.0 ± 1, screening buffer). This 1009

DOI: 10.1021/mp400666b Mol. Pharmaceutics 2015, 12, 1005−1017

Article

Molecular Pharmaceutics

Figure 3. Storage stability data measured by SE-HPLC plotting the increase in % aggregates at (a) 40 °C with linear regression lines, (b) 25 °C with linear regressions lines, and (c) 5 °C with lines connecting data points.

Figure 4. Bioactivity of mAbs in optimized formulations after storage at 25 °C for 6 months (black bars) and 40 °C for 3 months (gray bars). Data is not shown for mAb8 or mAb5 at 40 °C. 1010

DOI: 10.1021/mp400666b Mol. Pharmaceutics 2015, 12, 1005−1017

Article

Molecular Pharmaceutics included in Table 1. In contrast, several of the 5 °C data sets appear to depart from zero-order kinetics, and the time-course profiles are generally more irregular. The aggregate quantities formed at 5 °C are relatively small, and the impact of analytical and experimental variability is greater for the 2 year duration of this time-course; thus, the overall relative error in the 5 °C SEHPLC data is expected to be much higher. Storage stability data for mAbs1−7, as measured by percent biological activity in the respective assay specified for each of the mAbs (see Materials and Methods), are shown in Figure 4. The levels of biological activity after storage, respectively, at 25 °C for 6 months and after 3 months at 40 °C are shown (except for mAb5, for which the 40 °C data was not collected). These data show that after 6 months at 25 °C mAb1, mAb2, mAb3, mAb4, mAb6, and mAb7 retained ≥90% of initial biological activity and mAb5 retained 81%. After 3 months at 40 °C, the biological activity values ranged from 56% for mAb2 to 97% for mAb7. A typical bioassay specification for biologic products requires the product to retain at least 80% biological activity at end of shelf life. Applying this criterion, it can be concluded that good pharmaceutical stability could be claimed for these molecules for several months at room temperature and for at least 2 years at 5 °C (data not shown). It is noteworthy that the range of stabilities observed across this series of molecules is quite large at 40 °C for both the SE-HPLC data and the biological activity after 3 months. These data show that a mAb candidate molecule exhibiting an aggregation rate as high as 3.6% per month at 40 °C can achieve good storage stability at 5 and 25 °C.

examples of rational formulation strategies utilizing excipients to increase ΔGu with a resulting favorable impact on aggregation rates.50 Thus, there is a valid logic for utilizing a read-out of protein conformational stability as a practical approach to screening and ranking formulations and candidate molecules.24,25,51 However, using thermal denaturation as the basis for assessing conformational stability introduces some complications. It is evident from inspection of the modified Gibbs−Helmholtz equation (eq 1) why two proteins with the same Tm could have different thermodynamic stabilities at a temperature lower than the Tm because this will depend on ΔCp, the heat capacity change between the native and denatured forms, which could be different. ΔGu(T ) = ΔHm(1 − T /Tm) + ΔCp[(T − Tm) − T ln(T /Tm)]

(1)

where ΔHm is the value of ΔHu at Tm. It is also worth noting that although changes in free energy are widely used to evaluate the effect of temperature on the stability of proteins, the hydrophobic interaction, its temperature dependence, and its role in stabilizing proteins is still an evolving and incompletely understood aspect of fundamental biochemistry.52 Temperatures relevant to protein pharmaceuticals span a wide range: −70 and −20 °C for frozen bulk storage, 5 °C for long-term storage, and room temperature and above for patient use. There is much scope for the temperature dependence of the hydrophobic interaction to influence the outcome of the thermal denaturation process in an unpredictable manner, especially if one extends this range further for accelerated stability testing (e.g., ∼40−70 °C). There are additional reasons why prediction of low-temperature stability from accelerated temperatures is not straightforward. Proteins can aggregate via a variety of mechanisms that may change as temperature changes. Not all aggregation mechanisms proceed via an unfolded or partially unfolded state.53 The multidomain nature of mAbs further complicates the unfolding process, giving rise to the possibility of interactions between the multiple transition states or domains. The relationship between conformational stability and aggregation propensity is also more complex, as the least conformationally stable domain may not necessarily be the aggregation-prone domain. In the context of a pharmaceutical drug product, stability refers to meeting predefined acceptance criteria (specifications) pertaining to quality attributes. The desire for rapid screening (