Analytical Biochemistry 380 (2008) 331332

Contents lists available at ScienceDirect

Analytical Biochemistry
journal homepage: www.elsevier.com/locate/yabio

Notes & Tips

Estimation of protein aggregation propensity with a melting point apparatus

Andrei A. Raibekas *
Department of Pharmaceutics, Amgen, Thousand Oaks, CA 91320, USA

a r t i c l e

i n f o

a b s t r a c t
A method for studying protein aggregation with an automated melting point apparatus is described. The method employs thermal ramping and can generate a series of protein aggregation curves. The midpoint aggregation curve-associated temperature (Ta) is used to evaluate the difference between the curves where the lower Ta value corresponds to a higher aggregation propensity. The applicability of the method was demonstrated with human interleukin-1 receptor antagonist (IL-1ra) as a protein aggregation model. The method could be employed for rapid evaluation of various factors such as mutations, buffers, and excipients inuencing protein aggregation propensity under the thermal stress. 2008 Elsevier Inc. All rights reserved.

Article history: Received 26 March 2008 Available online 22 May 2008

Nonnative protein aggregation is related to protein instability and can be induced and inuenced by a variety of factors such as protein structural rearrangements, high pressure, temperature, and the presence of metals and ions [15]. Numerous methods for detection of protein aggregates or deposits in vivo and in vitro using absorbance, uorescence, luminescence, and solidstate nuclear magnetic resonance (NMR)1 have already been described in the literature [6]. Among them, a protein aggregation process measured by an increase in light scattering (turbidity) signal has been employed extensively in various kinetic studies [4,5,7,8]. Recently, Vedadi and coworkers [9] reported a new high-throughput approach to assess protein thermally induced aggregation propensity using a thermal gradient (ramping) mode in combination with light scattering detection provided by a specially designed multiwell format instrument, StarGazer (Harbinger Biotech, Toronto, Canada). Here I describe an alternative method for a quick evaluation of the protein aggregation propensity. The method uses both gradient thermal ramping and light scattering (turbidity) detection capabilities of a conventional melting point apparatus MPA 100 (Stanford Research Systems, Sunnyvale, CA, USA). Similarly to Vedadi and coworkers approach, it identies a narrow temperature region where the protein aggregation can occur under continuously increasing thermal stress. Human interleukin-1 receptor antagonist (IL-1ra), serving in this study as a protein aggregation model, is a 17-kDa monomeric protein with a regulatory role in inammation and functioning of the adaptive immune system [10]. It is known that an IL-1ra tendency to aggregate at elevated temperatures can be inuenced by its concentration and the buffer solution conditions [5,11].

* Fax: +1 805 376 8505. E-mail address: andreir@amgen.com 1 Abbreviations used: NMR, nuclear magnetic resonance; IL-1ra, interleukin-1 receptor antagonist; EDTA, ethylenediaminetetraacetic acid. 0003-2697/$ - see front matter 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.ab.2008.05.023

Puried recombinant human IL-1ra was obtained from Amgen manufacturing facility (Thousand Oaks, CA, USA). The 1 to 10 mg/ ml protein sample solutions were prepared from 220 mg/ml protein stock solution in PSE buffer (10 mM sodium phosphate, 140 mM NaCl and 0.5 mM ethylenediaminetetraacetic acid [EDTA]), pH 6.5 by dilution with the corresponding volumes of the same buffer. All chemicals were of the highest grade available. Typically, a 30 ll sample solution was placed inside of the top 20 mm portion of the glass capillary (1.5 90 mm, Stanford Research Systems) using a 1 to 200 ll gel loading tip and an automatic pipettor. To avoid introduction of the air bubbles and to achieve an optimal sample load, the loading tip was inserted to approximately 20 mm deep into the capillary and sample injection was gently performed while the loading tip was simultaneously and slowly moved backward to the top (open) end of the capillary. After the sample was placed inside the top end of the capillary, by holding its top end with two ngers, the capillary was vigorously shaken three to ve times sliding the sample from the top to the bottom end. Next, the loaded capillaries (total of three per run) were analyzed using an OptiMelt automated melting point system (model MPA100) equipped with the built-in digital camera and Digital Image Processing (DIP) technology (Stanford Research Systems). The instruments capillary holder was automatically preheated to 35 C, and then the capillaries were placed inside the holder. Immediately after that, an automated 30 min thermal ramping process was initiated and conducted at 1 C/min in a temperature range from 35 to 65 C. The image data processing and real-time display were performed using the MeltView software program (Stanford Research Systems). All signal data were automatically normalized to the maximum intensity (1 intensity unit) at the end of the each run. The results of experiment using four different concentrations (in triplicate) of the IL-1ra are shown in Fig. 1. Increasing the protein concentration led to a uniform left

332

Notes & Tips / Anal. Biochem. 380 (2008) 331332

1.0 0.8

Intensity

0.6 0.4 0.2 0.0 45

10

mg/ml

Ta

50

55

60

65

Temperature, (C)
Fig. 1. IL-1ra aggregation data generated by automated melting point system. Sample protein concentrations are 1, 2, 5, and 10 mg/ml in PSE buffer (pH 6.5). Overlaid traces of triplicate samples are shown to illustrate the methods reproducibility. All signal data are automatically normalized to the maximal intensity (1 intensity unit). The arrow indicates a curve midpoint temperature value (Ta).

10

ear decrease in Ta. The lower Ta value in this case should indicate a higher aggregation propensity [9]. To conrm that the IL-1ra concentration-dependent curve shifts produced by this method were indeed linked to the higher aggregation propensity, the same set of samples was submitted to a traditional plate reading assay where the IL-1ra aggregation kinetics was measured by light scattering (turbidity) over time at constant elevated temperatures [5]. The instrument was a temperature-controlled plate reading spectrophotometer, SpectraMax M5 (Molecular Devices, Sunnyvale, CA, USA). The experimental results using 1, 2, 5, and 10 mg/ml IL-1ra and 53 C revealed a series of saturation curves that clearly displayed protein concentration dependence (data not shown). The observed kinetic behavior was typical of IL-1ra [5] and consisted of three separate phases: lag, linear growth, and plateau. In accord with the data generated by the thermal ramping method, the increase in protein concentration resulted in a gradually shorter lag, stiffer growth phases, and higher plateaus, suggesting an overall increase in aggregation propensity. Another study currently in progress will address the link between these two different thermal methods and the IL-1ra aggregation behavior. In summary, a new method for estimating protein aggregation propensity using a melting point apparatus has been described. In general, this thermal ramping-based technique offers a number of conveniences such as small sample load (30 ll), thermal precision, reproducibility, and applicability within a broad temperature range ($35400 C). By using a capillary format and a specially designed holder, it minimizes sample heat transfer and evaporation problems and allows a relatively short (30 min per run) estimation of the narrow temperature region to cause visible protein aggregation. The method has the potential to become an effective screening tool for mutations, ligands, and buffer excipients inuencing protein stability and aggregation processes. References

IL-1ra (mg/ml)
1

51

52

53

54

55

56

57
[1] M. Fandrich, M.A. Fletcher, C.M. Dobson, Amyloid brils from muscle myoglobin, Nature 410 (2001) 165166. [2] M.B. Seefeldt, Y.S. Kim, K.P. Tolley, J. Seely, J.F. Carpenter, T.W. Randolph, Highpressure studies of aggregation of recombinant human interleukin-1 receptor antagonist: Thermodynamics, kinetics, and application to accelerated formulation studies, Protein Sci. 14 (2005) 22582266. [3] V.N. Uversky, J. Li, A.L. Fink, Metal-triggered structural transformations, aggregation, and brillation of human alfa-synuclein: A possible molecular NK between Parkinsons disease and heavy metal exposure, J. Biol. Chem. 276 (2001) 4428444296. [4] D.S. Maclean, Q. Qian, C.R. Middaugh, Stabilization of proteins by low molecular weight multi-ions, J. Pharm. Sci. 91 (2002) 22202229. [5] A.A. Raibekas, E.J. Bures, C.C. Siska, T. Kohno, R.F. Latypov, B.A. Kerwin, Anion binding and controlled aggregation of human interleukin-1 receptor antagonist, Biochemistry 44 (2005) 98719879. [6] R. Wetzel (Ed.), Amyloid, Prions, and Other Protein Aggregates, Methods in Enzymology, vol. 309, Academic Press, San Diego, 1999. [7] C.E. Smith, M.A. Lauffer, Polymerization-depolymerization of tobacco mosaic virus protein: VIII. Light-scattering studies, Biochemistry 6 (1967) 24572465. [8] K. Wang, B.I. Kurganov, Kinetics of heat- and acidication-induced aggregation of rey luciferase, Biophys. Chem. 106 (2003) 97109. [9] M. Vedadi, F.H. Niesen, A. Allali-Hassani, O.Y. Fedorov, P.J. Finerty Jr., G.A. Wasney, et al., Chemical screening methods to identify ligands that promote protein stability, protein crystallization, and structure determination, Proc. Natl. Acad. Sci. USA 103 (2006) 1583515840. [10] W.P. Arend, The balance between IL-1 and IL-1ra in disease, Cytokine Growth Factor Rev. 13 (2002) 323340. [11] J.R. Alford, B.S. Kendrick, J.F. Carpenter, T.W. Randolph, High concentration formulations of recombinant human interleukin-1 receptor antagonist: II. Aggregation kinetics, J. Pharm. Sci. (2007), doi:10.1002/jps.21205.

Ta
Fig. 2. Graphic relationship between protein concentration and its aggregation parameter, Ta, shown as a semilog plot. Samples are 1 to 10 mg/ml IL-1ra in PSE buffer (pH 6.5).

shift of the aggregation curve by as much as 5 C toward the lower temperature region. The protein aggregation process, a gradual changing in the sample clear-to-turbid phase transition, was monitored by a built-in digital camera and resulted in a plot resembling an inverse protein melting curve [5,9]. The curve shifting pattern (but not the shift direction) was similar to that observed by Vedadi and coworkers in their experimental system where the protein sulfotransferase (Sult1C1) was titrated with its ligand, PAP [9]. The observed correlation between the curve shift and the IL-1ra concentration was further evaluated by introducing a Ta parameter (Fig. 1), that is, the temperature value corresponding to the midpoint intensity value (0.5) of the aggregation curve [9]. Fig. 2 shows a plot of the native protein concentration (1 to 10 mg/ml samples with 1 mg/ml increments) versus its respectively derived Ta value. The semilog plot follows a linear inverse relationship; that is, the protein concentration factor here is exponentially linked to the lin-