- Email: [email protected]
Displacement chromatography of proteins using a retained pH front in a hydrophobic charge induction chromatography column
G Model
ARTICLE IN PRESS
CHROMA-356248; No. of Pages 7
Journal of Chromatography A, xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Displacement chromatography of proteins using a retained pH front in a hydrophobic charge induction chromatography column N.D.S. Pinto, Douglas D. Frey ∗ Department of Chemical, Biochemical and Environmental Engineering, University of Maryland Baltimore County, Baltimore, MD 21250, USA
a r t i c l e
i n f o
Article history: Received 7 December 2014 Received in revised form 29 January 2015 Accepted 29 January 2015 Available online xxx Keywords: Displacement chromatography Hydrophobic charge induction chromatography MEP HyperCel Chromatofocusing Proteins
a b s t r a c t The chromatographic separation of two proteins into a displacement train of two adjoined rectangular bands was accomplished using a novel method for hydrophobic charge induction chromatography (HCIC) which employs a self-sharpening pH front as the displacer. This method exploits the fact that protein elution in HCIC is promoted by a pH change, but is relatively independent of salt effects, so that a retained pH front can be used in place of a traditional displacer in displacement chromatography. The retained pH front was produced using the two adsorbed buffering species tricine and acetic acid. The separation of lysozyme and ␣-chymotrypsinogen A into adjoined, rectangular bands was accomplished with overall recoveries based on the total mass injected greater than 90 and 70%, respectively. The addition of urea to the buffer system increased the sharpness of the pH front by 36% while the yields of lysozyme and ␣-chymotrypsinogen A based on the total mass eluted increased from 76% to 99% and from 37% to 85%, respectively, when the purities of both proteins in their product fractions were fixed at 85%. The results demonstrate that the method developed in this study is a useful variant of HCIC and is also a useful alternative to other displacement chromatography methods. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Mixed-mode chromatography (MMC), which is sometimes also termed multimodal chromatography, employs a chromatographic column packing that exhibits more than one mode of interaction, with electrostatic interactions, hydrophobic interactions and hydrogen bonding being the most common of the interactions modes employed. The recent rapid growth in the number of publications on MMC shows the increasing interest in these chromatographic packings by the scientific community [1]. It is commonly thought that the synergism between the various interaction modes in MMC can yield higher selectivity and resolution when compared to single-mode chromatography. Burton and Harding [2–5] were the originators of hydrophobic charge induction chromatography (HCIC), which is one of the most common MMC methods used for protein purification. The mercaptoethyl-pyridine (MEP) ligands employed on the HCIC column packing exhibit hydrophobic interactions of nearly constant strength, but pH dependent electrostatic interactions. Protein adsorption is promoted mostly by the hydrophobic interactions that are present, while electrostatic repulsive interactions induce
∗ Corresponding author. Tel.: +1 410 455 3418; fax: +1 410 455 6500. E-mail address: [email protected] (D.D. Frey).
protein elution when the pH is lowered due to the presence of the ionizable groups on the column packing. An unretained pH step change is the most common method employed in HCIC to promote protein elution since the protein desorption process is mostly salt independent [5]. Other ways to promote partial or total protein elution with the MEP column packing have been accomplished by employing ethylene glycol, 2propanol, urea or guanidine [6,7]. Employing this type of partial elution can be useful to elute impurities in a washing step after protein binding provided that it does not significantly lower the yield of the proteins of interest. However, these latter techniques do not take full advantage of the dual adsorption modes of HCIC column packings since they do not utilize electrostatic interactions. Zhao and Sun [8] introduced displacement chromatography using the MEP column packing in an attempt to exploit its mixedmode characteristic. As is typical of displacement chromatography in general, these workers employed a displacer (in this case a quaternary ammonium salt) that is more strongly adsorbed than any of the protein mixture components to promote protein displacement into pure rectangular bands without the aid of a pH change. The second chromatographic mode possible in MMC – electrostatic repulsion via lowering the pH – was only used to remove the displacer, and in this way Zhao and Sun eliminated one of the key disadvantages of displacement chromatography. However, other disadvantages, such as the potential
http://dx.doi.org/10.1016/j.chroma.2015.01.087 0021-9673/© 2015 Elsevier B.V. All rights reserved.
Please cite this article in press as: N.D.S. Pinto, D.D. Frey, Displacement chromatography of proteins using a retained pH front in a hydrophobic charge induction chromatography column, J. Chromatogr. A (2015), http://dx.doi.org/10.1016/j.chroma.2015.01.087
G Model CHROMA-356248; No. of Pages 7 2
ARTICLE IN PRESS N.D.S. Pinto, D.D. Frey / J. Chromatogr. A xxx (2015) xxx–xxx
contamination of the proteins being separated by the displacer, were not addressed. Chromatofocusing is a chromatographic technique that exploits electrostatic interactions using retained pH gradients that are chemically produced inside an ion-exchange chromatographic column to accomplish protein separations. Sluyterman and coworkers [9,10] first described this technique in 1978. Since then, many of the shortcomings of the method as originally described have been alleviated, such as the elimination of polyampholyte buffers by using well-defined multicomponent buffer systems so that the method can be more effectively optimized [11–15]. Frey and co-workers [16] introduced a chromotofocusing technique for the displacement chromatography of proteins where a retained pH front was employed as a displacer in anion-exchange chromatography. This technique does not require an additional chemical species to act as a displacer since a change in the concentration of the hydrogen ion in effect plays the role of the displacer. In this way, many of the major disadvantages of displacement chromatography associated with traditional displacers are eliminated. In the method developed by Frey and co-workers the protein adsorption process was largely controlled by electrostatic interactions since the column packing employed was of the traditional single-mode type. The optimization of the adsorption and desorption of proteins through different retention mechanisms was therefore not possible. Furthermore, the method was also limited to the separation of proteins with isoelectric points within range of the pH gradient employed. In this study, a novel displacement chromatography technique for HCIC will be investigated where protein adsorption is due to hydrophobic interactions so that it can be optimized using additives such as urea, while protein desorption is due to electrostatic repulsion so that it is promoted by a tailored, retained pH front formed with adsorbing buffering species that acts as a displacer. It is shown in this work that the addition of urea to the buffer affects the buffering species and protein adsorption equilibrium beneficially so that the sharpness of the pH front is increased and the overlap between adjacent protein bands is minimized. By using the methods described here, the optimization of the adsorption and desorption mechanisms of proteins can be added as advantages of displacement chromatography, along with its previously well known advantages of high capacity and high resolution [17–19], while also avoiding the disadvantages commonly associated with a traditional displacer [16].
1.1. Protein and buffering species behavior in HCIC with a retained pH front Frey and co-workers [13–16,20,21] discussed the formation of retained pH gradients when strong-electrolyte, or both strong and weak-electrolyte, functional groups are present on an anionexchange column packing. In the method considered by these investigators, the buffering species and protein adsorption processes were largely controlled by electrostatic interactions since the functional groups on the column packing primarily exhibited only electrostatic interactions and were charged throughout the entire pH range employed. In contrast to the work of Frey and co-workers, the mercapto-ethylpyridine (MEP) ligands on the chromatographic column packings employed in the present study are uncharged and exhibit significant hydrophobic interactions due to the aromatic ring on the pyridine moiety at neutral or higher pH. However, at a pH below 7 the basic group present on the MEP ligand becomes charged and therefore exhibits electrostatic interactions due to its low acid–base dissociation constant of pKa = 4.8, as compared to traditional weak anion-exchange column packings where the pKa value is typically above 9.
Fig. 1. Schematic representation of the pH and concentration profiles at the column outlet when a retained pH front is produced using two buffering species, one of which is in the presaturation buffer while the other is in the elution buffer. Both presaturation and elution buffers are titrated to the desired pH with a strong base containing an inert cationic species. Figure is based on numerical simulations as described by Pinto [22].
Fig. 1 shows the general behavior of the buffering species and effluent pH profiles for a MEP HyperCel chromatographic column when a retained pH gradient is employed. This figure is based on numerical simulations using the method described by Frey et al. [13] and is discussed in more detail by Pinto [22]. The buffering species employed in the figure are either uncharged or negatively charged so that, in the latter case, they are the proper ionic form to adsorb electrostatically onto an anion-exchange column packing. In addition, the buffering species in the elution buffer has the lowest acid–base dissociation constant so that it exhibits the strongest binding to the column packing, while the buffering species in both the presaturation and elution buffers are titrated to the desired pH with a strong base which introduces an inert cationic species into the buffers. As shown in Fig. 1, an unretained pH front with a small pH change is formed under the conditions just described after one column void volume has passed through the column due to the elution of the unretained buffering species present in the presaturation buffer. The retained pH front shown, i.e., the shelf-sharpening pH front that is chemically produced inside the column, is associated with the adsorption of the buffering species in the elution buffer as the charge on the column packing increases due to the large change in pH at that front. Fig. 1 also shows that there is some uptake of the inert cationic species in the intermediate plateau, which indicates that the charge on the column packing is negligible above neutral pH. This latter observation is also confirmed by the fact that the pH values in the presaturation and intermediate plateaus are similar. Fig. 2 shows qualitatively the protein behavior caused by a retained pH front. In the top part of the figure, curves representing the velocities of protein concentration levels as a function of pH are superimposed on the axial pH profile in the column. In the bottom part of the figure, an illustration of the corresponding protein adsorption isotherm is shown. The velocity of a concentration front for either proteins or the buffering species is given by [23]:
vC =
vfluid
1 + ((1 − ˛)/˛)ε + ((1 − ˛)/˛)(1 − ε) q∗i /Ci∗
(1)
where vC is the velocity of a concentration front, vfluid is the interstitial fluid velocity, ˛ is the interstitial void volume in the bed, ε is the internal porosity of the particle, is the difference across the front, q∗i is the equilibrium value of the adsorbed concentration of
Please cite this article in press as: N.D.S. Pinto, D.D. Frey, Displacement chromatography of proteins using a retained pH front in a hydrophobic charge induction chromatography column, J. Chromatogr. A (2015), http://dx.doi.org/10.1016/j.chroma.2015.01.087
G Model CHROMA-356248; No. of Pages 7
ARTICLE IN PRESS N.D.S. Pinto, D.D. Frey / J. Chromatogr. A xxx (2015) xxx–xxx
3
filtered using polystyrene filters with 0.2-m pores obtained from Corning (Corning, NY, USA). A 12 × 1 cm I.D. Omnifit glass column (Kinesis, Malta, NY, USA) packed with MEP HyperCel chromatographic packing material (Pall Life Sciences, Port Washington, NY, USA) and having a volume of 9.4 mL was used for the displacement experiments. Fraction analysis was accomplished with a 300 × 7.8 mm I.D. G3000 SWXL size-exclusion column (Tosoh Biosciences, Montgomeryville, PA, USA). 2.2. Equipment
Fig. 2. Top: Schematic representation of the superposition of the velocities of protein concentration levels on the axial pH profile in the column with the mass-transfer resistances ignored so that the pH fronts are vertical steps. Bottom: illustration of corresponding protein adsorption isotherm at the intermediate plateau pH. Note that the horizontal axis in the top panel corresponds to the position along the column in the direction of flow so that the pH gradient shown is reversed in orientation as compared to the time-based representation at the column outlet shown in Fig. 1.
Experiments were performed using a LC Packings (now Thermo Scientific Dionex, Sunnyvale, CA, USA) Ultimate HPLC system which included a UV–vis absorbance detector with a flow cell having a 3L volume and 10-mm path length, a Famos autosampler, and a Probot fraction collector. Samples were filtered using nylon filters having 0.2-m pores obtained from Fisher Scientific (Pittsburgh, PA, USA). Sample injections with a 5-mL volume were performed manually using a Rheodyne model 7010 6-port valve and a 5-mL sample loop (Idex Health and Science, Oak Harbor, WA, USA). The column effluent pH was measured using a Model 450CD pH electrode (Sensorex, Garden Grove, CA, USA) with a flow cell having a 50-L internal volume and an Orion (now Thermo Scientific Orion, Beverly, MA, USA) Model 520A pH meter. Chromeleon software version 6.6 (Thermo Scientific Dionex) was used for data collection and instrument control. 2.3. Procedures for pH displacement experiments
Figure is adapted from Fig. 2 of Ref. [16] and is used with permission.
species i corresponding to Ci , and Ci is the concentration of species i in the liquid phase. As discussed by Frey and co-workers [16], Fig. 2 demonstrates that if the protein velocity curves of two different dilute proteins (represented, e.g., by the curve C1 ) intercept the retained pH front velocity, these two proteins will become focused on the pH front and will co-elute from the column as a single band. This focusing effect is due to the competition between the proteins and the other ionic species present for adsorption sites on the column packing. More specifically, individual proteins that lag behind in terms of their spatial position with respect to the center of the focus band containing the majority of a particular protein will be exposed to a slightly different pH than the main protein band. This will cause these proteins to change their net charge so that they rejoin the main focused band. An analogous situation applies to individual proteins that lag ahead of the center of the focused band. Finally, as also discussed by Frey and co-workers [16], as the amount of a particular protein in the feed sample is increased, eventually a rectangular band containing the protein and positioned in the intermediate pH plateau is formed inside the column at a concentration (i.e., the concentration C2 in Fig. 2) where the protein band velocity at the pH of the intermediate plateau matches the velocity of the pH front.
The retained pH gradients employed in the displacement experiments were obtained by performing a step change between the presaturant and elution buffers. Equilibration of the MEP HyperCel column was accomplished by using a presaturation buffer which flowed through the column until the column effluent stabilized at the presaturation pH. Proteins were dissolved in the presaturation buffer to form the sample. The column effluent was monitored at 280 nm and 1 mL fractions were collected. Column regeneration was accomplished with the presaturation buffer. 2.4. Procedures for fraction analysis Size-exclusion chromatography column was used for the analysis of the effluent fractions from the pH displacement experiments. The buffer employed for size exclusion chromatography consisted of 50 mM potassium phosphate and 150 mM sodium sulfate titrated with sodium hydroxide to pH 7.0. The UV absorbance of the column effluent was monitored at 280 nm, and 50 L samples of each fraction where employed at a flow rate of 1 mL/min. The protein concentration for each fraction was determined by numerical integration of the digitally recorded data to calculate the peak areas, which were then converted to mass quantities using a linear calibration curve relating area to mass. 3. Results and discussion
2. Experimental 2.1. Materials and columns The chemical species used in the buffer systems were tricine, acetic acid, potassium phosphate, hydrochloric acid, sodium sulfate, sodium hydroxide, sodium chloride and urea, all obtained from Sigma–Aldrich (St. Louis, MO, USA). Lysozyme and ␣chymotrypsinogen A were also obtained from Sigma–Aldrich. The buffer solutions were prepared using deionized water and were
The results presented here describe various aspects of the chromatography of proteins where retained pH gradients are employed together with a MEP HyperCel column packing. In Section 3.1, the chromatography of a single protein present in various amounts is studied. This section also establishes that it is possible to elute proteins bands from a MEP HyperCel column packing that are focused on a retained pH front. The flexibility of employing a MMC column packing is shown in Section 3.2 where the protein adsorption equilibrium is beneficially affected by employing urea to take advantage
Please cite this article in press as: N.D.S. Pinto, D.D. Frey, Displacement chromatography of proteins using a retained pH front in a hydrophobic charge induction chromatography column, J. Chromatogr. A (2015), http://dx.doi.org/10.1016/j.chroma.2015.01.087
G Model CHROMA-356248; No. of Pages 7 4
ARTICLE IN PRESS N.D.S. Pinto, D.D. Frey / J. Chromatogr. A xxx (2015) xxx–xxx
Fig. 3. Effluent pH and UV absorbance at 280 nm for the chromatography of various amounts of lysozyme using a MEP HyperCel column. The presaturant buffer contained 10 mM sodium hydroxide and 20 mM tricine titrated to pH 8.0 with hydrochloric acid. Elution was performed with 20 mM acetic acid titrated to pH 5.0 with sodium hydroxide. The flow rate was 0.5 mL/min.
of the fact that protein binding and elution are promoted by different interaction modes on the column packing. Finally, Section 3.3 explores protein displacement into rectangular, adjoined bands using a retained pH front. 3.1. Behavior of a single protein when pH displacement is used with MEP HyperCel The chromatographic band shape for a single protein can be used to infer its behavior in a displacement process involving several proteins; namely, whether a set of proteins will form adjoined rectangular bands which exit the column ahead of a retained, selfsharpening pH front. Fig. 3 shows the effluent pH profile for a MEP HyperCel column presaturated with 10 mM sodium hydroxide and 20 mM tricine that was titrated to pH 8.0 with hydrochloric acid and then eluted with 20 mM acetic acid titrated to pH 5.0 with sodium hydroxide. As shown, a self-sharpening pH front was produced that travels through the column at a velocity lower than the interstitial fluid velocity, which is consistent with the method described by Frey and co-workers [13–16]. More specifically, in order to form a retained self-sharpening pH front, it is necessary to employ adsorbed buffering species where the buffering species present in the elution buffer is more strongly retained than the buffering species in the presaturation buffer, since otherwise a proportionate pattern (i.e., non-self-sharpening) front would be formed. In the case of MEP HyperCel where the column packing functional groups are cationic in nature, the buffering species more strongly adsorbed is that with the lowest pKa value, which is acetic acid for the present case.
Fig. 4. Effluent pH and UV absorbance at 280 nm for the chromatography of various amounts of lysozyme using a MEP HyperCel column. The presaturant buffer contained 10 mM sodium hydroxide, 20 mM tricine and 4 M urea titrated to pH 8.0 with hydrochloric acid. Elution was performed with 20 mM acetic acid and 4 M urea titrated to pH 5.0 with sodium hydroxide. The flow rate was 0.5 mL/min.
Fig. 3 also illustrates the UV absorbance at 280 nm in the column effluent for the chromatography of feed samples with 1.0, 10 and 25 mg of lysozyme, which corresponds respectively to 0.11, 1.1, and 2.7 mg/mL based on the total column volume. Note that the difference in the retention time of the pH front for the chromatography of feed samples with 1.0 mg versus 10 and 25 mg of lysozyme is due to differences in the sample loop size used, i.e., 50 L and 5.0 mL respectively. In addition, the shapes of the protein bands shown in the figure demonstrate the desired trend of obtaining rectangular protein bands as the amount of protein chromatographed is increased. Note also that the presence of large amounts of protein affects the shape of the pH front to some extent, although a rectangular shape for the protein band is still produced. 3.2. Influence of urea on the shape of a single protein band with pH displacement In order to evaluate the influence of urea and its possible usefulness for the pH displacement of proteins when using the MEP HyperCel column packing, a single protein band study similar to the one performed in Section 3.1 was conducted, but where 4 M urea was added to the buffer system. Note that the protein secondary structure under these conditions is likely preserved since it has been reported that urea concentrations up to 4 M generally do not significantly alter this structure for proteins in general, and
Please cite this article in press as: N.D.S. Pinto, D.D. Frey, Displacement chromatography of proteins using a retained pH front in a hydrophobic charge induction chromatography column, J. Chromatogr. A (2015), http://dx.doi.org/10.1016/j.chroma.2015.01.087
G Model CHROMA-356248; No. of Pages 7
ARTICLE IN PRESS N.D.S. Pinto, D.D. Frey / J. Chromatogr. A xxx (2015) xxx–xxx
for lysozyme in particular [24–26]. Additional past studies have shown that partial unfolding especially of the tertiary structure of some proteins, including both lysozyme and ␣-chymotrypsinogen A, may occur in the presence of large amounts of urea, although this unfolding is generally reversible when urea is removed [27–30]. By comparing the chromatograms for 1.0 mg of lysozyme with and without urea present (i.e., the top portions of Figs. 3 and 4), it can be observed that employing urea improves the sharpness of the pH front, and consequently also the symmetry of the protein peak. The improved shape of the pH front is likely due to the reported ability of urea to alter the water structure and also lower the dielectric constant so that the electrostatic interactions between charged groups is increased, including those responsible for the formation of a retained, self-sharpening pH front [30]. Figs. 3 and 4 show that, for the case of feed samples with 10 and 25 mg of lysozyme, urea causes the protein bands to be more “rectangular,” i.e., the shapes of the protein bands in the presence of urea are wider and shallower than in its absence. This also indicates that the concentration of protein in the adsorbed phase is reduced when urea is added. This is because the velocities of the protein bands are similar for the two cases so that, according to Eq. (1), a reduction in the liquid-phase protein concentration in the band implies that the adsorbed-phase protein concentration is also reduced. This effect is possibly due to the ability of the hydrophobic regions of the protein to interact preferentially with urea in comparison to water so that the equilibrium adsorption isotherm is shifted in position when urea is added [30]. The alteration in the shape of the protein bands obtained with the addition of urea shown in Fig. 4 is likely to be beneficial in practice since, when several adjoined protein bands are present, the boundary regions between bands that consist of a protein mixture will contain a smaller fraction of the original protein sample. Moreover, a comparison of Figs. 3 and 4 indicates that the amount of urea added when the sample contains a protein mixture is limited by the constraint that the time corresponding to the entire displacement train is less than the retention time of the pH gradient. 3.3. pH displacement chromatography of lysozyme and ˛-chymotrypsinogen A A feed sample containing a total protein mass of 25 mg (which corresponds to 2.7 mg/mL based on the total column volume) with approximately equal amounts of lysozyme and ␣chymotrypsinogen A was used to verify that the method described here produced adjoined rectangular protein bands adjacent to a retained pH front produced inside a MEP HyperCel column. Lysozyme and ␣-chymotrypsinogen A were chosen for use in this study since their hydrophobic properties and pI values are similar to monoclonal antibodies, which are a class of proteins that are important in the biotechnology industry and that have been widely employed with the MEP HyperCel column packing [31]. Effluent pH and protein concentration profiles for the displacement chromatography of lysozyme and ␣-chymotrypsinogen A using the MEP HyperCel column packing and a self-sharpening pH front as the displacer, in the presence and absence of urea, are shown in Figs. 5 and 6, respectively. The flow rate employed was 0.1 mL/min and the buffer systems used were the same as previously described in Sections 3.1 and 3.2. The protein concentration profiles shown in the pictures were obtained using size-exclusion chromatography as described in Section 2.4. Figs. 5 and 6 illustrate that a self-sharpening, retained pH front can produce a protein displacement train using the MEP HyperCel column packing. In Fig. 5, the protein bands formed by ␣-chymotrypsinogen A and lysozyme in the presence of urea show minimal overlap with total recoveries of 74.8% and 94.0%, respectively, based on the total mass injected. In contrast, Fig. 6 shows
5
Fig. 5. Effluent pH and protein concentration profiles for the displacement chromatography of 25 mg of lysozyme and ␣-chymotrypsinogen A present in equal amounts. A 5 mL sample loop and a MEP HyperCel column were used. The column was presaturated with 10 mM sodium hydroxide, 20 mM tricine and 4 M urea titrated to pH 8.0 with hydrochloric acid. Elution was performed with 20 mM acetic acid and 4 M urea titrated to pH 5.0 with sodium hydroxide. Flow rate was 0.1 mL/min.
a higher overlap of the lysozyme and ␣-chymotrypsinogen A protein bands when urea was absent with total recoveries of 97.7% and 89.9%, respectively, based on the total mass injected. As illustrated in the figures, these differences appear to be largely due to the effect of urea on the protein concentrations in the effluent profile and the associated protein bandwidths, which are larger for the case of urea addition in comparison to the case without urea. In order to more completely assess the effect of urea on the chromatographic results shown in Figs. 5 and 6, quantitative measures were developed for the pH front sharpness and for the protein yield achieved at a given purity. In particular, the pH front sharpness was defined as the width of the pH front with urea present subtracted from the width of the pH front without urea present divided by the latter quantity. For this purpose the width of a pH front was determined using pH values that differed by 2.5% (based on the
Fig. 6. Effluent pH and protein concentration profiles for the displacement chromatography of 25 mg of lysozyme and ␣-chymotrypsinogen A present in equal amounts. A 5 mL sample loop and a MEP HyperCel column were used. The column was presaturated with 10 mM sodium hydroxide and 20 mM tricine titrated to pH 8.0 with hydrochloric acid. Elution was performed with 20 mM acetic acid titrated to pH 5.0 with sodium hydroxide. The flow rate was 0.1 mL/min.
Please cite this article in press as: N.D.S. Pinto, D.D. Frey, Displacement chromatography of proteins using a retained pH front in a hydrophobic charge induction chromatography column, J. Chromatogr. A (2015), http://dx.doi.org/10.1016/j.chroma.2015.01.087
G Model CHROMA-356248; No. of Pages 7 6
ARTICLE IN PRESS N.D.S. Pinto, D.D. Frey / J. Chromatogr. A xxx (2015) xxx–xxx
total pH change for the front) from the upstream and downstream plateau pH values. To assess the protein yield for a given purity, four product fraction cut points were located, one of which was upstream from the displacement train, one of which was downstream from the displacement train, and two of which were within the displacement train and near the boundary between the two individual protein bands. The positions of the latter two cut points were specifically determined so that one product fraction consisted of lysozyme with a purity of 85%, one product fraction consisted of ␣-chymotrypsinogen A also with a purity of 85%, and the third product fraction consisted roughly of equal amounts of the two proteins. The yields of the two proteins in the first two product fractions just mentioned were then calculated based on the total mass eluted. Using these definitions it was found that the addition of urea increased the pH front sharpness by 36% while the yields of lysozyme and ␣-chymotrypsinogen A increased from 76% to 99% and from 37% to 85%, respectively, when the purities of both proteins were fixed at 85%. Figs. 5 and 6 also indicate that lysozyme and ␣chymotrypsinogen A exhibit a selectivity reversal (i.e., the retention order is changed) in the presence and absence of urea for the conditions employed. The protein retention order for the case where urea is absent (i.e., Fig. 6) is the same as the retention order observed in the absence of urea by Zhao and Sun [8] at pH 5 when using the MEP HyperCel column packing, and with the retention order observed at pH 7 in hydrophobic interaction chromatography [32]. This retention order appears to be largely determined in these cases by the fact that lysozyme is less hydrophobic than ␣-chymotrypsinogen A and, because of its higher isoelectric point, it is also more strongly repelled electrostatically, so that lysozyme is less strongly adsorbed than ␣-chymotrypsinogen A. The trends just described in Figs. 5 and 6 are most likely due to the well-documented effects of urea of reducing the hydrophobic attractive forces, increasing the electrostatic repulsive forces, and partly unfolding reversibly one or the other of the proteins present [29]. The combined result of these three effects evidently is to reverse the adsorption affinities and therefore also the retention times of the two proteins used in Figs. 5 and 6. Note that the amount of protein used in these two figures is the same, and in the case of the Fig. 6 is not sufficient in quantity to produce fully rectangular bands. No attempt was made in this study to produce fully rectangular protein bands for the case of Fig. 6 since the very large protein amounts that would have been required were inconvenient to use, and since the conclusions obtained from this investigation were not dependent on forming fully rectangular protein bands for the specific case of Fig. 6. Fig. 7 shows the adsorption isotherms for lysozyme and ␣chymotrypsinogen A in both molar concentrations and mass concentration units obtained from fitting the data of Zhao and Sun [8] for these proteins at pH 5 to the Langmuir model. Even though the pH and ionic strength used by Zhao and Sun differ somewhat from the conditions used in this study, both studies employed acetate as the counter-ion and are similar enough so that comparisons are nevertheless useful. The operating lines that are also shown in Fig. 7 with and without urea present are determined by the ratio of qi /Ci observed for both lysozyme and ␣-chymotrypsinogen A in the displacement train for these two cases. This ratio was calculated by applying Eq. (1) to obtain the adsorbed-phase protein concentrations from the retained pH front velocities and the liquid-phase protein concentrations shown in the Figs. 5 and 6. As illustrated in the figure, urea addition produces a decrease in the slope of the operating line, suggesting that it has lowered the protein adsorption isotherms. This is consistent with the fact that the rectangular protein bands in Fig. 5 are smaller in height as compared to Fig. 6. Note also that the operating lines obtained in this study compare well with the
Fig. 7. Operating lines with and without urea present in the buffer system for lysozyme () and ␣-chymotrypsinogen A () correspond to the chromatograms in Figs. 5 and 6, respectively. The adsorption isotherms shown were obtained by fitting the data of Zhao and Sun [8] to the Langmuir model. The top panel is in mass concentration units while the bottom panel is in molar concentration units.
isotherms of Zhao and Sun [8] which indicates the consistency of the two studies. Fig. 7 also indicates that the relative position of the proteins on the operating lines with and without urea is reversed when comparing the plot having mass concentration units with the plot having molar concentration units. This is consistent with the retention order reversal shown in Figs. 5 and 6 since generally the former type of plot indicates the retention order as opposed to the latter type plot (see, e.g., the results of Evans et al. [33]). Fig. 7 is also consistent with the results of Zhao and Sun [8] who reported selectivity reversal between lysozyme and ␣-chymotrypsinogen A when using the MEP HyperCel column packing for the case where the pH was changed from 5.0 to 7.0. The results described in this section indicate that protein displacement when using a pH front as a displacer together with the MEP HyperCel column packing is a complex process influenced by a combination of electrostatic and hydrophobic interactions. This is shown by the fact that the ordering of protein bands in displacement chromatography can be modified either by changing the pH or the urea concentration. This also suggests that additional opportunities to customize the displacement of proteins through a synergetic optimization of the various interaction modes present may be possible, in which case the range of applications of the method can likely be broadened to apply to a wide variety of proteins. 4. Conclusions The chromatographic displacement of proteins to yield rectangular, adjoined protein bands was accomplished by using a self-sharpening pH front as the displacer and a MEP HyperCel column packing. The buffer system employed in the method utilized adsorbed buffering species to produce a retained pH gradient that traveled through the chromatographic column. By effectively employing a pH front as the displacer, many of the disadvantages
Please cite this article in press as: N.D.S. Pinto, D.D. Frey, Displacement chromatography of proteins using a retained pH front in a hydrophobic charge induction chromatography column, J. Chromatogr. A (2015), http://dx.doi.org/10.1016/j.chroma.2015.01.087
G Model CHROMA-356248; No. of Pages 7
ARTICLE IN PRESS N.D.S. Pinto, D.D. Frey / J. Chromatogr. A xxx (2015) xxx–xxx
associated with the traditional chemical displacer employed in displacement chromatography are eliminated, such as the contamination of the proteins in the displacement train by the displacer, column regeneration issues, as well as the displacer selection and toxicity. Moreover, since the method uses a single stepwise change in buffer composition as the column inlet and is primarily optimized by varying the buffer composition, it is simple to implement even on a large scale. In addition, the pH range employed is similar to the typical conditions employed for protein separations in the traditional HCIC method. The method therefore has few disadvantages compared to HCIC as performed without a displacement procedure while also preserving the advantages of displacement chromatography, such as its ability to concentrate proteins as they are being separated, and its high capacity and resolution. Exploratory experiments performed at loading amounts of 50 mg and 100 mg of lysozyme (data not shown), which corresponds to 5.4 and 10.8 mg/mL based on the total column volume, suggest that the potential exists for operating the method developed here at these higher loadings. However, under these conditions further work is required to find the optimal urea concentration to use since urea was found to have complex effects on the shapes of both the protein band and the retained pH front at high protein loadings. This type of complex optimization was considered outside of the scope of this work. reversal between lysozyme and ␣Selectivity chymotrypsinogen A caused by the presence and absence of urea was observed when using the MEP HyperCel column packing. This shows that protein elution in this case is a complex process that is influenced by a combination of electrostatic and hydrophobic interactions. It is evident from the work performed here that the mixed-mode characteristics of MEP HyperCel provides significant flexibility and optimization options for the design of the buffer system since it is possible in this case to produce retained pH fronts that promote the displacement of proteins with a pH range that can be tailored for a given purpose. The method can also employ urea and likely other chaotropic agents to beneficially affect the protein adsorption equilibrium. For example, it was determined that the addition of urea increased the yields of lysozyme and ␣-chymotrypsinogen A obtained by the method from 76% to 99% and from 37% to 85%, respectively, when the purities of both proteins were fixed at 85%. Lastly, the results described here may be useful to aid in the development of chromatographic processes that have not been directly addressed in this study, such as the method of selective displacement chromatography where the displacement train contains a subset of the proteins in the feed sample [34]. Acknowledgments Support from grants CBET 0854151 and 1264392 from the National Science Foundation is greatly appreciated. References [1] Y. Yang, X. Geng, Mixed-mode chromatography and its applications to biopolymers, J. Chromatogr. A 1218 (2011) 8813–8825. [2] S.C. Burton, D.R.K. Harding, High-density ligand attachment to brominated allyl matrices and application to mixed mode chromatography of chymosin, J. Chromatogr. A 775 (1997) 39–50. [3] S.C. Burton, D.R.K. Harding, Preparation of chromatographic matrices by free radical addition ligand to allyl groups, J. Chromatogr. A 796 (1998) 273–282.
7
[4] S.C. Burton, D.R.K. Harding, Hydrophobic charge induction chromatography: salt independent protein adsorption and facile elution with aqueous buffers, J. Chromatogr. A 814 (1998) 71–81. [5] S.C. Burton, D.R.K. Harding, Salt-independent adsorption chromatography: new broad-spectrum affinity methods for protein capture, J. Biochem. Biophys. Methods 49 (2001) 275–287. [6] T. Arakawa, Y. Kita, H. Sato, D. Ejima, MEP chromatography of antibody and Fcfusion protein using aqueous arginine solution, Protein Expr. Purif. 63 (2009) 158–163. [7] T. Arakawa, M. Futatsumori-Sugai, K. Tsumoto, Y. Kita, H. Sato, D. Ejima, MEP HyperCel chromatography II: binding, washing and elution, Protein Expr. Purif. 71 (2010) 168–173. [8] G. Zhao, Y. Sun, Displacement chromatography of proteins on hydrophobic charge induction adsorbent column, J. Chromatogr. A 1165 (2007) 109–115. [9] L.A.Æ. Sluyterman, O. Elgersma, Chromatofocusing isoelectric focusing on ionexchange columns: I. General principles, J. Chromatogr. 150 (1978) 17–30. [10] L.A.Æ. Sluyterman, J. Wijdeness, Chromatofocusing: isoelectric focusing on ionexchange columns: II. Experimental verification, J. Chromatogr. 150 (1978) 31–44. [11] M.T.W. Hearn, D. Lyttle, Buffer-focusing chromatography using multicomponent electrolyte elution systems, J. Chromatogr. 218 (1981) 483–495. [12] T.W. Hutchens, C.M. Li, P. Besch, Performance evaluation of a focusing buffer developed for chromatofocusing on high-performance anion-exchange columns, J. Chromatogr. 359 (1986) 169–179. [13] D.D. Frey, A. Barnes, J. Strong, Numerical studies of multicomponent chromatography using pH gradients, AIChE J. 41 (1995) 1171–1183. [14] D.D. Frey, Local-equilibrium behavior of retained pH and ionic strength gradients in preparative chromatography, Biotechnol. Prog. 12 (1996) 65–72. [15] J.C. Strong, D.D. Frey, Experimental and numerical studies of the chromatofocusing of dilute proteins using retained pH gradients formed on a strong-base anion-exchange column, J. Chromatogr. A 769 (1997) 129–143. [16] C.R. Narahari, J.C. Strong, D.D. Frey, Displacement chromatography of proteins using a self-sharpening pH front formed by absorbed buffering species as the displacer, J. Chromatogr. A 825 (1998) 115–126. [17] Cs. Hováth, Displacement chromatography: yesterday, today and tomorrow, in: F. Brunner (Ed.), The Science of Chromatography, Elsevier, Amsterdam, 1985, pp. 179–203. [18] S. Cramer, Displacement chromatography of proteins, in: A. Kenney, S. Fowell (Eds.), Practical Protein Chromatography (Methods in Molecular Biology), Human Press, New Jersey, 1992, pp. 259–272. [19] G. Guiochon, S.G. Shirazi, A.M. Katti, Fundamentals of Preparative and Nonlinear Chromatography, second ed., Academic Press, Boston, 1994. [20] R.C. Bates, X. Kang, D.D. Frey, High-performance chromatofocusing using linear and concave pH gradients formed with simple buffer mixtures: I. Effect of buffer composition on the gradient shape, J. Chromatogr. A 890 (2000) 25–36. [21] X. Kang, R.C. Bates, D.D. Frey, High-performance chromatofocusing using linear and concave pH gradients formed with simple buffer mixtures: II. Separation of proteins, J. Chromatogr. A 890 (2000) 37–43. [22] N.D.S. Pinto, Novel Hybrid Chromatofocusing Methods for Protein Purification, Ph.D. Dissertation, University of Maryland Baltimore County, Baltimore, MD, 2013. [23] P.C. Wankat, Rate-Controlled Separations, Elsevier, New York, 1990. [24] P. Gagnon, Fine-Tuning Selectivity on Ion Exchangers, Validated Biosystems, 1999, pp. 1, http://www.validated.com/revalbio/pdffiles/ionslect.pdf. [25] T.W. Hutchison, Chromatofocusing, in: J.-C. Janson, L. Ryden (Eds.), Protein Purification: Principles, High Resolution Methods, and Applications, first ed., VCH Publishers, New York, 1989, pp. 149–174. [26] S. Timasheff, G. Xie, Preferential interactions of urea with lysozyme and their linkage to protein denaturation, Biophys. Chem. 105 (2003) 421–448. [27] E.S. Parente, D.B. Wetlaufer, Effects of urea-thermal denaturation on the highperformance cation-exchange chromatography of ␣-chymotrypsinogen-A, J. Chromatogr. 314 (1984) 337–347. [28] Y. Hou, T.B. Hansen, A. Staby, S.M. Cramer, Effects of urea induced protein conformational changes on ion exchange chromatographic behavior, J. Chromatogr. A 1247 (2010) 7393–7400. [29] M.A. Holstein, S. Parimal, S.A. McCallum, S.M. Cramer, Effects of urea on selectivity and protein-ligand interactions in multimodal cation exchange chromatography, Langmuir 29 (2013) 158–167. [30] F. Khademi, A. Mostafaie, Effect of urea on protein separation by ion-exchange chromatography, J. Biochem. 147 (2010) 735–741. [31] S. Ghose, B. Hubbard, S.M. Cramer, Protein interactions in hydrophobic charge induction chromatography (HCIC), Biotechnol. Prog. 21 (2005) 498–508. [32] A. Katti, Y.-F. Maa, Cs. Horváth, Protein surface area and retention in hydrophobic interactions chromatorgraphy, Chromatographia 24 (1987) 646–650. [33] S.T. Evans, A. Freed, S.M. Cramer, Displacer concentration effects in displacement chromatography: implications for trace solute detection, J. Chromatogr. A 1216 (2009) 79–85. [34] A. Kundu, K.A. Barnthouse, S.M. Cramer, Selective displacement chromatography of proteins, Biotechnol. Bioeng. 56 (1997) 119–219.
Please cite this article in press as: N.D.S. Pinto, D.D. Frey, Displacement chromatography of proteins using a retained pH front in a hydrophobic charge induction chromatography column, J. Chromatogr. A (2015), http://dx.doi.org/10.1016/j.chroma.2015.01.087
ARTICLE IN PRESS
CHROMA-356248; No. of Pages 7
Journal of Chromatography A, xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Displacement chromatography of proteins using a retained pH front in a hydrophobic charge induction chromatography column N.D.S. Pinto, Douglas D. Frey ∗ Department of Chemical, Biochemical and Environmental Engineering, University of Maryland Baltimore County, Baltimore, MD 21250, USA
a r t i c l e
i n f o
Article history: Received 7 December 2014 Received in revised form 29 January 2015 Accepted 29 January 2015 Available online xxx Keywords: Displacement chromatography Hydrophobic charge induction chromatography MEP HyperCel Chromatofocusing Proteins
a b s t r a c t The chromatographic separation of two proteins into a displacement train of two adjoined rectangular bands was accomplished using a novel method for hydrophobic charge induction chromatography (HCIC) which employs a self-sharpening pH front as the displacer. This method exploits the fact that protein elution in HCIC is promoted by a pH change, but is relatively independent of salt effects, so that a retained pH front can be used in place of a traditional displacer in displacement chromatography. The retained pH front was produced using the two adsorbed buffering species tricine and acetic acid. The separation of lysozyme and ␣-chymotrypsinogen A into adjoined, rectangular bands was accomplished with overall recoveries based on the total mass injected greater than 90 and 70%, respectively. The addition of urea to the buffer system increased the sharpness of the pH front by 36% while the yields of lysozyme and ␣-chymotrypsinogen A based on the total mass eluted increased from 76% to 99% and from 37% to 85%, respectively, when the purities of both proteins in their product fractions were fixed at 85%. The results demonstrate that the method developed in this study is a useful variant of HCIC and is also a useful alternative to other displacement chromatography methods. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Mixed-mode chromatography (MMC), which is sometimes also termed multimodal chromatography, employs a chromatographic column packing that exhibits more than one mode of interaction, with electrostatic interactions, hydrophobic interactions and hydrogen bonding being the most common of the interactions modes employed. The recent rapid growth in the number of publications on MMC shows the increasing interest in these chromatographic packings by the scientific community [1]. It is commonly thought that the synergism between the various interaction modes in MMC can yield higher selectivity and resolution when compared to single-mode chromatography. Burton and Harding [2–5] were the originators of hydrophobic charge induction chromatography (HCIC), which is one of the most common MMC methods used for protein purification. The mercaptoethyl-pyridine (MEP) ligands employed on the HCIC column packing exhibit hydrophobic interactions of nearly constant strength, but pH dependent electrostatic interactions. Protein adsorption is promoted mostly by the hydrophobic interactions that are present, while electrostatic repulsive interactions induce
∗ Corresponding author. Tel.: +1 410 455 3418; fax: +1 410 455 6500. E-mail address: [email protected] (D.D. Frey).
protein elution when the pH is lowered due to the presence of the ionizable groups on the column packing. An unretained pH step change is the most common method employed in HCIC to promote protein elution since the protein desorption process is mostly salt independent [5]. Other ways to promote partial or total protein elution with the MEP column packing have been accomplished by employing ethylene glycol, 2propanol, urea or guanidine [6,7]. Employing this type of partial elution can be useful to elute impurities in a washing step after protein binding provided that it does not significantly lower the yield of the proteins of interest. However, these latter techniques do not take full advantage of the dual adsorption modes of HCIC column packings since they do not utilize electrostatic interactions. Zhao and Sun [8] introduced displacement chromatography using the MEP column packing in an attempt to exploit its mixedmode characteristic. As is typical of displacement chromatography in general, these workers employed a displacer (in this case a quaternary ammonium salt) that is more strongly adsorbed than any of the protein mixture components to promote protein displacement into pure rectangular bands without the aid of a pH change. The second chromatographic mode possible in MMC – electrostatic repulsion via lowering the pH – was only used to remove the displacer, and in this way Zhao and Sun eliminated one of the key disadvantages of displacement chromatography. However, other disadvantages, such as the potential
http://dx.doi.org/10.1016/j.chroma.2015.01.087 0021-9673/© 2015 Elsevier B.V. All rights reserved.
Please cite this article in press as: N.D.S. Pinto, D.D. Frey, Displacement chromatography of proteins using a retained pH front in a hydrophobic charge induction chromatography column, J. Chromatogr. A (2015), http://dx.doi.org/10.1016/j.chroma.2015.01.087
G Model CHROMA-356248; No. of Pages 7 2
ARTICLE IN PRESS N.D.S. Pinto, D.D. Frey / J. Chromatogr. A xxx (2015) xxx–xxx
contamination of the proteins being separated by the displacer, were not addressed. Chromatofocusing is a chromatographic technique that exploits electrostatic interactions using retained pH gradients that are chemically produced inside an ion-exchange chromatographic column to accomplish protein separations. Sluyterman and coworkers [9,10] first described this technique in 1978. Since then, many of the shortcomings of the method as originally described have been alleviated, such as the elimination of polyampholyte buffers by using well-defined multicomponent buffer systems so that the method can be more effectively optimized [11–15]. Frey and co-workers [16] introduced a chromotofocusing technique for the displacement chromatography of proteins where a retained pH front was employed as a displacer in anion-exchange chromatography. This technique does not require an additional chemical species to act as a displacer since a change in the concentration of the hydrogen ion in effect plays the role of the displacer. In this way, many of the major disadvantages of displacement chromatography associated with traditional displacers are eliminated. In the method developed by Frey and co-workers the protein adsorption process was largely controlled by electrostatic interactions since the column packing employed was of the traditional single-mode type. The optimization of the adsorption and desorption of proteins through different retention mechanisms was therefore not possible. Furthermore, the method was also limited to the separation of proteins with isoelectric points within range of the pH gradient employed. In this study, a novel displacement chromatography technique for HCIC will be investigated where protein adsorption is due to hydrophobic interactions so that it can be optimized using additives such as urea, while protein desorption is due to electrostatic repulsion so that it is promoted by a tailored, retained pH front formed with adsorbing buffering species that acts as a displacer. It is shown in this work that the addition of urea to the buffer affects the buffering species and protein adsorption equilibrium beneficially so that the sharpness of the pH front is increased and the overlap between adjacent protein bands is minimized. By using the methods described here, the optimization of the adsorption and desorption mechanisms of proteins can be added as advantages of displacement chromatography, along with its previously well known advantages of high capacity and high resolution [17–19], while also avoiding the disadvantages commonly associated with a traditional displacer [16].
1.1. Protein and buffering species behavior in HCIC with a retained pH front Frey and co-workers [13–16,20,21] discussed the formation of retained pH gradients when strong-electrolyte, or both strong and weak-electrolyte, functional groups are present on an anionexchange column packing. In the method considered by these investigators, the buffering species and protein adsorption processes were largely controlled by electrostatic interactions since the functional groups on the column packing primarily exhibited only electrostatic interactions and were charged throughout the entire pH range employed. In contrast to the work of Frey and co-workers, the mercapto-ethylpyridine (MEP) ligands on the chromatographic column packings employed in the present study are uncharged and exhibit significant hydrophobic interactions due to the aromatic ring on the pyridine moiety at neutral or higher pH. However, at a pH below 7 the basic group present on the MEP ligand becomes charged and therefore exhibits electrostatic interactions due to its low acid–base dissociation constant of pKa = 4.8, as compared to traditional weak anion-exchange column packings where the pKa value is typically above 9.
Fig. 1. Schematic representation of the pH and concentration profiles at the column outlet when a retained pH front is produced using two buffering species, one of which is in the presaturation buffer while the other is in the elution buffer. Both presaturation and elution buffers are titrated to the desired pH with a strong base containing an inert cationic species. Figure is based on numerical simulations as described by Pinto [22].
Fig. 1 shows the general behavior of the buffering species and effluent pH profiles for a MEP HyperCel chromatographic column when a retained pH gradient is employed. This figure is based on numerical simulations using the method described by Frey et al. [13] and is discussed in more detail by Pinto [22]. The buffering species employed in the figure are either uncharged or negatively charged so that, in the latter case, they are the proper ionic form to adsorb electrostatically onto an anion-exchange column packing. In addition, the buffering species in the elution buffer has the lowest acid–base dissociation constant so that it exhibits the strongest binding to the column packing, while the buffering species in both the presaturation and elution buffers are titrated to the desired pH with a strong base which introduces an inert cationic species into the buffers. As shown in Fig. 1, an unretained pH front with a small pH change is formed under the conditions just described after one column void volume has passed through the column due to the elution of the unretained buffering species present in the presaturation buffer. The retained pH front shown, i.e., the shelf-sharpening pH front that is chemically produced inside the column, is associated with the adsorption of the buffering species in the elution buffer as the charge on the column packing increases due to the large change in pH at that front. Fig. 1 also shows that there is some uptake of the inert cationic species in the intermediate plateau, which indicates that the charge on the column packing is negligible above neutral pH. This latter observation is also confirmed by the fact that the pH values in the presaturation and intermediate plateaus are similar. Fig. 2 shows qualitatively the protein behavior caused by a retained pH front. In the top part of the figure, curves representing the velocities of protein concentration levels as a function of pH are superimposed on the axial pH profile in the column. In the bottom part of the figure, an illustration of the corresponding protein adsorption isotherm is shown. The velocity of a concentration front for either proteins or the buffering species is given by [23]:
vC =
vfluid
1 + ((1 − ˛)/˛)ε + ((1 − ˛)/˛)(1 − ε) q∗i /Ci∗
(1)
where vC is the velocity of a concentration front, vfluid is the interstitial fluid velocity, ˛ is the interstitial void volume in the bed, ε is the internal porosity of the particle, is the difference across the front, q∗i is the equilibrium value of the adsorbed concentration of
Please cite this article in press as: N.D.S. Pinto, D.D. Frey, Displacement chromatography of proteins using a retained pH front in a hydrophobic charge induction chromatography column, J. Chromatogr. A (2015), http://dx.doi.org/10.1016/j.chroma.2015.01.087
G Model CHROMA-356248; No. of Pages 7
ARTICLE IN PRESS N.D.S. Pinto, D.D. Frey / J. Chromatogr. A xxx (2015) xxx–xxx
3
filtered using polystyrene filters with 0.2-m pores obtained from Corning (Corning, NY, USA). A 12 × 1 cm I.D. Omnifit glass column (Kinesis, Malta, NY, USA) packed with MEP HyperCel chromatographic packing material (Pall Life Sciences, Port Washington, NY, USA) and having a volume of 9.4 mL was used for the displacement experiments. Fraction analysis was accomplished with a 300 × 7.8 mm I.D. G3000 SWXL size-exclusion column (Tosoh Biosciences, Montgomeryville, PA, USA). 2.2. Equipment
Fig. 2. Top: Schematic representation of the superposition of the velocities of protein concentration levels on the axial pH profile in the column with the mass-transfer resistances ignored so that the pH fronts are vertical steps. Bottom: illustration of corresponding protein adsorption isotherm at the intermediate plateau pH. Note that the horizontal axis in the top panel corresponds to the position along the column in the direction of flow so that the pH gradient shown is reversed in orientation as compared to the time-based representation at the column outlet shown in Fig. 1.
Experiments were performed using a LC Packings (now Thermo Scientific Dionex, Sunnyvale, CA, USA) Ultimate HPLC system which included a UV–vis absorbance detector with a flow cell having a 3L volume and 10-mm path length, a Famos autosampler, and a Probot fraction collector. Samples were filtered using nylon filters having 0.2-m pores obtained from Fisher Scientific (Pittsburgh, PA, USA). Sample injections with a 5-mL volume were performed manually using a Rheodyne model 7010 6-port valve and a 5-mL sample loop (Idex Health and Science, Oak Harbor, WA, USA). The column effluent pH was measured using a Model 450CD pH electrode (Sensorex, Garden Grove, CA, USA) with a flow cell having a 50-L internal volume and an Orion (now Thermo Scientific Orion, Beverly, MA, USA) Model 520A pH meter. Chromeleon software version 6.6 (Thermo Scientific Dionex) was used for data collection and instrument control. 2.3. Procedures for pH displacement experiments
Figure is adapted from Fig. 2 of Ref. [16] and is used with permission.
species i corresponding to Ci , and Ci is the concentration of species i in the liquid phase. As discussed by Frey and co-workers [16], Fig. 2 demonstrates that if the protein velocity curves of two different dilute proteins (represented, e.g., by the curve C1 ) intercept the retained pH front velocity, these two proteins will become focused on the pH front and will co-elute from the column as a single band. This focusing effect is due to the competition between the proteins and the other ionic species present for adsorption sites on the column packing. More specifically, individual proteins that lag behind in terms of their spatial position with respect to the center of the focus band containing the majority of a particular protein will be exposed to a slightly different pH than the main protein band. This will cause these proteins to change their net charge so that they rejoin the main focused band. An analogous situation applies to individual proteins that lag ahead of the center of the focused band. Finally, as also discussed by Frey and co-workers [16], as the amount of a particular protein in the feed sample is increased, eventually a rectangular band containing the protein and positioned in the intermediate pH plateau is formed inside the column at a concentration (i.e., the concentration C2 in Fig. 2) where the protein band velocity at the pH of the intermediate plateau matches the velocity of the pH front.
The retained pH gradients employed in the displacement experiments were obtained by performing a step change between the presaturant and elution buffers. Equilibration of the MEP HyperCel column was accomplished by using a presaturation buffer which flowed through the column until the column effluent stabilized at the presaturation pH. Proteins were dissolved in the presaturation buffer to form the sample. The column effluent was monitored at 280 nm and 1 mL fractions were collected. Column regeneration was accomplished with the presaturation buffer. 2.4. Procedures for fraction analysis Size-exclusion chromatography column was used for the analysis of the effluent fractions from the pH displacement experiments. The buffer employed for size exclusion chromatography consisted of 50 mM potassium phosphate and 150 mM sodium sulfate titrated with sodium hydroxide to pH 7.0. The UV absorbance of the column effluent was monitored at 280 nm, and 50 L samples of each fraction where employed at a flow rate of 1 mL/min. The protein concentration for each fraction was determined by numerical integration of the digitally recorded data to calculate the peak areas, which were then converted to mass quantities using a linear calibration curve relating area to mass. 3. Results and discussion
2. Experimental 2.1. Materials and columns The chemical species used in the buffer systems were tricine, acetic acid, potassium phosphate, hydrochloric acid, sodium sulfate, sodium hydroxide, sodium chloride and urea, all obtained from Sigma–Aldrich (St. Louis, MO, USA). Lysozyme and ␣chymotrypsinogen A were also obtained from Sigma–Aldrich. The buffer solutions were prepared using deionized water and were
The results presented here describe various aspects of the chromatography of proteins where retained pH gradients are employed together with a MEP HyperCel column packing. In Section 3.1, the chromatography of a single protein present in various amounts is studied. This section also establishes that it is possible to elute proteins bands from a MEP HyperCel column packing that are focused on a retained pH front. The flexibility of employing a MMC column packing is shown in Section 3.2 where the protein adsorption equilibrium is beneficially affected by employing urea to take advantage
Please cite this article in press as: N.D.S. Pinto, D.D. Frey, Displacement chromatography of proteins using a retained pH front in a hydrophobic charge induction chromatography column, J. Chromatogr. A (2015), http://dx.doi.org/10.1016/j.chroma.2015.01.087
G Model CHROMA-356248; No. of Pages 7 4
ARTICLE IN PRESS N.D.S. Pinto, D.D. Frey / J. Chromatogr. A xxx (2015) xxx–xxx
Fig. 3. Effluent pH and UV absorbance at 280 nm for the chromatography of various amounts of lysozyme using a MEP HyperCel column. The presaturant buffer contained 10 mM sodium hydroxide and 20 mM tricine titrated to pH 8.0 with hydrochloric acid. Elution was performed with 20 mM acetic acid titrated to pH 5.0 with sodium hydroxide. The flow rate was 0.5 mL/min.
of the fact that protein binding and elution are promoted by different interaction modes on the column packing. Finally, Section 3.3 explores protein displacement into rectangular, adjoined bands using a retained pH front. 3.1. Behavior of a single protein when pH displacement is used with MEP HyperCel The chromatographic band shape for a single protein can be used to infer its behavior in a displacement process involving several proteins; namely, whether a set of proteins will form adjoined rectangular bands which exit the column ahead of a retained, selfsharpening pH front. Fig. 3 shows the effluent pH profile for a MEP HyperCel column presaturated with 10 mM sodium hydroxide and 20 mM tricine that was titrated to pH 8.0 with hydrochloric acid and then eluted with 20 mM acetic acid titrated to pH 5.0 with sodium hydroxide. As shown, a self-sharpening pH front was produced that travels through the column at a velocity lower than the interstitial fluid velocity, which is consistent with the method described by Frey and co-workers [13–16]. More specifically, in order to form a retained self-sharpening pH front, it is necessary to employ adsorbed buffering species where the buffering species present in the elution buffer is more strongly retained than the buffering species in the presaturation buffer, since otherwise a proportionate pattern (i.e., non-self-sharpening) front would be formed. In the case of MEP HyperCel where the column packing functional groups are cationic in nature, the buffering species more strongly adsorbed is that with the lowest pKa value, which is acetic acid for the present case.
Fig. 4. Effluent pH and UV absorbance at 280 nm for the chromatography of various amounts of lysozyme using a MEP HyperCel column. The presaturant buffer contained 10 mM sodium hydroxide, 20 mM tricine and 4 M urea titrated to pH 8.0 with hydrochloric acid. Elution was performed with 20 mM acetic acid and 4 M urea titrated to pH 5.0 with sodium hydroxide. The flow rate was 0.5 mL/min.
Fig. 3 also illustrates the UV absorbance at 280 nm in the column effluent for the chromatography of feed samples with 1.0, 10 and 25 mg of lysozyme, which corresponds respectively to 0.11, 1.1, and 2.7 mg/mL based on the total column volume. Note that the difference in the retention time of the pH front for the chromatography of feed samples with 1.0 mg versus 10 and 25 mg of lysozyme is due to differences in the sample loop size used, i.e., 50 L and 5.0 mL respectively. In addition, the shapes of the protein bands shown in the figure demonstrate the desired trend of obtaining rectangular protein bands as the amount of protein chromatographed is increased. Note also that the presence of large amounts of protein affects the shape of the pH front to some extent, although a rectangular shape for the protein band is still produced. 3.2. Influence of urea on the shape of a single protein band with pH displacement In order to evaluate the influence of urea and its possible usefulness for the pH displacement of proteins when using the MEP HyperCel column packing, a single protein band study similar to the one performed in Section 3.1 was conducted, but where 4 M urea was added to the buffer system. Note that the protein secondary structure under these conditions is likely preserved since it has been reported that urea concentrations up to 4 M generally do not significantly alter this structure for proteins in general, and
Please cite this article in press as: N.D.S. Pinto, D.D. Frey, Displacement chromatography of proteins using a retained pH front in a hydrophobic charge induction chromatography column, J. Chromatogr. A (2015), http://dx.doi.org/10.1016/j.chroma.2015.01.087
G Model CHROMA-356248; No. of Pages 7
ARTICLE IN PRESS N.D.S. Pinto, D.D. Frey / J. Chromatogr. A xxx (2015) xxx–xxx
for lysozyme in particular [24–26]. Additional past studies have shown that partial unfolding especially of the tertiary structure of some proteins, including both lysozyme and ␣-chymotrypsinogen A, may occur in the presence of large amounts of urea, although this unfolding is generally reversible when urea is removed [27–30]. By comparing the chromatograms for 1.0 mg of lysozyme with and without urea present (i.e., the top portions of Figs. 3 and 4), it can be observed that employing urea improves the sharpness of the pH front, and consequently also the symmetry of the protein peak. The improved shape of the pH front is likely due to the reported ability of urea to alter the water structure and also lower the dielectric constant so that the electrostatic interactions between charged groups is increased, including those responsible for the formation of a retained, self-sharpening pH front [30]. Figs. 3 and 4 show that, for the case of feed samples with 10 and 25 mg of lysozyme, urea causes the protein bands to be more “rectangular,” i.e., the shapes of the protein bands in the presence of urea are wider and shallower than in its absence. This also indicates that the concentration of protein in the adsorbed phase is reduced when urea is added. This is because the velocities of the protein bands are similar for the two cases so that, according to Eq. (1), a reduction in the liquid-phase protein concentration in the band implies that the adsorbed-phase protein concentration is also reduced. This effect is possibly due to the ability of the hydrophobic regions of the protein to interact preferentially with urea in comparison to water so that the equilibrium adsorption isotherm is shifted in position when urea is added [30]. The alteration in the shape of the protein bands obtained with the addition of urea shown in Fig. 4 is likely to be beneficial in practice since, when several adjoined protein bands are present, the boundary regions between bands that consist of a protein mixture will contain a smaller fraction of the original protein sample. Moreover, a comparison of Figs. 3 and 4 indicates that the amount of urea added when the sample contains a protein mixture is limited by the constraint that the time corresponding to the entire displacement train is less than the retention time of the pH gradient. 3.3. pH displacement chromatography of lysozyme and ˛-chymotrypsinogen A A feed sample containing a total protein mass of 25 mg (which corresponds to 2.7 mg/mL based on the total column volume) with approximately equal amounts of lysozyme and ␣chymotrypsinogen A was used to verify that the method described here produced adjoined rectangular protein bands adjacent to a retained pH front produced inside a MEP HyperCel column. Lysozyme and ␣-chymotrypsinogen A were chosen for use in this study since their hydrophobic properties and pI values are similar to monoclonal antibodies, which are a class of proteins that are important in the biotechnology industry and that have been widely employed with the MEP HyperCel column packing [31]. Effluent pH and protein concentration profiles for the displacement chromatography of lysozyme and ␣-chymotrypsinogen A using the MEP HyperCel column packing and a self-sharpening pH front as the displacer, in the presence and absence of urea, are shown in Figs. 5 and 6, respectively. The flow rate employed was 0.1 mL/min and the buffer systems used were the same as previously described in Sections 3.1 and 3.2. The protein concentration profiles shown in the pictures were obtained using size-exclusion chromatography as described in Section 2.4. Figs. 5 and 6 illustrate that a self-sharpening, retained pH front can produce a protein displacement train using the MEP HyperCel column packing. In Fig. 5, the protein bands formed by ␣-chymotrypsinogen A and lysozyme in the presence of urea show minimal overlap with total recoveries of 74.8% and 94.0%, respectively, based on the total mass injected. In contrast, Fig. 6 shows
5
Fig. 5. Effluent pH and protein concentration profiles for the displacement chromatography of 25 mg of lysozyme and ␣-chymotrypsinogen A present in equal amounts. A 5 mL sample loop and a MEP HyperCel column were used. The column was presaturated with 10 mM sodium hydroxide, 20 mM tricine and 4 M urea titrated to pH 8.0 with hydrochloric acid. Elution was performed with 20 mM acetic acid and 4 M urea titrated to pH 5.0 with sodium hydroxide. Flow rate was 0.1 mL/min.
a higher overlap of the lysozyme and ␣-chymotrypsinogen A protein bands when urea was absent with total recoveries of 97.7% and 89.9%, respectively, based on the total mass injected. As illustrated in the figures, these differences appear to be largely due to the effect of urea on the protein concentrations in the effluent profile and the associated protein bandwidths, which are larger for the case of urea addition in comparison to the case without urea. In order to more completely assess the effect of urea on the chromatographic results shown in Figs. 5 and 6, quantitative measures were developed for the pH front sharpness and for the protein yield achieved at a given purity. In particular, the pH front sharpness was defined as the width of the pH front with urea present subtracted from the width of the pH front without urea present divided by the latter quantity. For this purpose the width of a pH front was determined using pH values that differed by 2.5% (based on the
Fig. 6. Effluent pH and protein concentration profiles for the displacement chromatography of 25 mg of lysozyme and ␣-chymotrypsinogen A present in equal amounts. A 5 mL sample loop and a MEP HyperCel column were used. The column was presaturated with 10 mM sodium hydroxide and 20 mM tricine titrated to pH 8.0 with hydrochloric acid. Elution was performed with 20 mM acetic acid titrated to pH 5.0 with sodium hydroxide. The flow rate was 0.1 mL/min.
Please cite this article in press as: N.D.S. Pinto, D.D. Frey, Displacement chromatography of proteins using a retained pH front in a hydrophobic charge induction chromatography column, J. Chromatogr. A (2015), http://dx.doi.org/10.1016/j.chroma.2015.01.087
G Model CHROMA-356248; No. of Pages 7 6
ARTICLE IN PRESS N.D.S. Pinto, D.D. Frey / J. Chromatogr. A xxx (2015) xxx–xxx
total pH change for the front) from the upstream and downstream plateau pH values. To assess the protein yield for a given purity, four product fraction cut points were located, one of which was upstream from the displacement train, one of which was downstream from the displacement train, and two of which were within the displacement train and near the boundary between the two individual protein bands. The positions of the latter two cut points were specifically determined so that one product fraction consisted of lysozyme with a purity of 85%, one product fraction consisted of ␣-chymotrypsinogen A also with a purity of 85%, and the third product fraction consisted roughly of equal amounts of the two proteins. The yields of the two proteins in the first two product fractions just mentioned were then calculated based on the total mass eluted. Using these definitions it was found that the addition of urea increased the pH front sharpness by 36% while the yields of lysozyme and ␣-chymotrypsinogen A increased from 76% to 99% and from 37% to 85%, respectively, when the purities of both proteins were fixed at 85%. Figs. 5 and 6 also indicate that lysozyme and ␣chymotrypsinogen A exhibit a selectivity reversal (i.e., the retention order is changed) in the presence and absence of urea for the conditions employed. The protein retention order for the case where urea is absent (i.e., Fig. 6) is the same as the retention order observed in the absence of urea by Zhao and Sun [8] at pH 5 when using the MEP HyperCel column packing, and with the retention order observed at pH 7 in hydrophobic interaction chromatography [32]. This retention order appears to be largely determined in these cases by the fact that lysozyme is less hydrophobic than ␣-chymotrypsinogen A and, because of its higher isoelectric point, it is also more strongly repelled electrostatically, so that lysozyme is less strongly adsorbed than ␣-chymotrypsinogen A. The trends just described in Figs. 5 and 6 are most likely due to the well-documented effects of urea of reducing the hydrophobic attractive forces, increasing the electrostatic repulsive forces, and partly unfolding reversibly one or the other of the proteins present [29]. The combined result of these three effects evidently is to reverse the adsorption affinities and therefore also the retention times of the two proteins used in Figs. 5 and 6. Note that the amount of protein used in these two figures is the same, and in the case of the Fig. 6 is not sufficient in quantity to produce fully rectangular bands. No attempt was made in this study to produce fully rectangular protein bands for the case of Fig. 6 since the very large protein amounts that would have been required were inconvenient to use, and since the conclusions obtained from this investigation were not dependent on forming fully rectangular protein bands for the specific case of Fig. 6. Fig. 7 shows the adsorption isotherms for lysozyme and ␣chymotrypsinogen A in both molar concentrations and mass concentration units obtained from fitting the data of Zhao and Sun [8] for these proteins at pH 5 to the Langmuir model. Even though the pH and ionic strength used by Zhao and Sun differ somewhat from the conditions used in this study, both studies employed acetate as the counter-ion and are similar enough so that comparisons are nevertheless useful. The operating lines that are also shown in Fig. 7 with and without urea present are determined by the ratio of qi /Ci observed for both lysozyme and ␣-chymotrypsinogen A in the displacement train for these two cases. This ratio was calculated by applying Eq. (1) to obtain the adsorbed-phase protein concentrations from the retained pH front velocities and the liquid-phase protein concentrations shown in the Figs. 5 and 6. As illustrated in the figure, urea addition produces a decrease in the slope of the operating line, suggesting that it has lowered the protein adsorption isotherms. This is consistent with the fact that the rectangular protein bands in Fig. 5 are smaller in height as compared to Fig. 6. Note also that the operating lines obtained in this study compare well with the
Fig. 7. Operating lines with and without urea present in the buffer system for lysozyme () and ␣-chymotrypsinogen A () correspond to the chromatograms in Figs. 5 and 6, respectively. The adsorption isotherms shown were obtained by fitting the data of Zhao and Sun [8] to the Langmuir model. The top panel is in mass concentration units while the bottom panel is in molar concentration units.
isotherms of Zhao and Sun [8] which indicates the consistency of the two studies. Fig. 7 also indicates that the relative position of the proteins on the operating lines with and without urea is reversed when comparing the plot having mass concentration units with the plot having molar concentration units. This is consistent with the retention order reversal shown in Figs. 5 and 6 since generally the former type of plot indicates the retention order as opposed to the latter type plot (see, e.g., the results of Evans et al. [33]). Fig. 7 is also consistent with the results of Zhao and Sun [8] who reported selectivity reversal between lysozyme and ␣-chymotrypsinogen A when using the MEP HyperCel column packing for the case where the pH was changed from 5.0 to 7.0. The results described in this section indicate that protein displacement when using a pH front as a displacer together with the MEP HyperCel column packing is a complex process influenced by a combination of electrostatic and hydrophobic interactions. This is shown by the fact that the ordering of protein bands in displacement chromatography can be modified either by changing the pH or the urea concentration. This also suggests that additional opportunities to customize the displacement of proteins through a synergetic optimization of the various interaction modes present may be possible, in which case the range of applications of the method can likely be broadened to apply to a wide variety of proteins. 4. Conclusions The chromatographic displacement of proteins to yield rectangular, adjoined protein bands was accomplished by using a self-sharpening pH front as the displacer and a MEP HyperCel column packing. The buffer system employed in the method utilized adsorbed buffering species to produce a retained pH gradient that traveled through the chromatographic column. By effectively employing a pH front as the displacer, many of the disadvantages
Please cite this article in press as: N.D.S. Pinto, D.D. Frey, Displacement chromatography of proteins using a retained pH front in a hydrophobic charge induction chromatography column, J. Chromatogr. A (2015), http://dx.doi.org/10.1016/j.chroma.2015.01.087
G Model CHROMA-356248; No. of Pages 7
ARTICLE IN PRESS N.D.S. Pinto, D.D. Frey / J. Chromatogr. A xxx (2015) xxx–xxx
associated with the traditional chemical displacer employed in displacement chromatography are eliminated, such as the contamination of the proteins in the displacement train by the displacer, column regeneration issues, as well as the displacer selection and toxicity. Moreover, since the method uses a single stepwise change in buffer composition as the column inlet and is primarily optimized by varying the buffer composition, it is simple to implement even on a large scale. In addition, the pH range employed is similar to the typical conditions employed for protein separations in the traditional HCIC method. The method therefore has few disadvantages compared to HCIC as performed without a displacement procedure while also preserving the advantages of displacement chromatography, such as its ability to concentrate proteins as they are being separated, and its high capacity and resolution. Exploratory experiments performed at loading amounts of 50 mg and 100 mg of lysozyme (data not shown), which corresponds to 5.4 and 10.8 mg/mL based on the total column volume, suggest that the potential exists for operating the method developed here at these higher loadings. However, under these conditions further work is required to find the optimal urea concentration to use since urea was found to have complex effects on the shapes of both the protein band and the retained pH front at high protein loadings. This type of complex optimization was considered outside of the scope of this work. reversal between lysozyme and ␣Selectivity chymotrypsinogen A caused by the presence and absence of urea was observed when using the MEP HyperCel column packing. This shows that protein elution in this case is a complex process that is influenced by a combination of electrostatic and hydrophobic interactions. It is evident from the work performed here that the mixed-mode characteristics of MEP HyperCel provides significant flexibility and optimization options for the design of the buffer system since it is possible in this case to produce retained pH fronts that promote the displacement of proteins with a pH range that can be tailored for a given purpose. The method can also employ urea and likely other chaotropic agents to beneficially affect the protein adsorption equilibrium. For example, it was determined that the addition of urea increased the yields of lysozyme and ␣-chymotrypsinogen A obtained by the method from 76% to 99% and from 37% to 85%, respectively, when the purities of both proteins were fixed at 85%. Lastly, the results described here may be useful to aid in the development of chromatographic processes that have not been directly addressed in this study, such as the method of selective displacement chromatography where the displacement train contains a subset of the proteins in the feed sample [34]. Acknowledgments Support from grants CBET 0854151 and 1264392 from the National Science Foundation is greatly appreciated. References [1] Y. Yang, X. Geng, Mixed-mode chromatography and its applications to biopolymers, J. Chromatogr. A 1218 (2011) 8813–8825. [2] S.C. Burton, D.R.K. Harding, High-density ligand attachment to brominated allyl matrices and application to mixed mode chromatography of chymosin, J. Chromatogr. A 775 (1997) 39–50. [3] S.C. Burton, D.R.K. Harding, Preparation of chromatographic matrices by free radical addition ligand to allyl groups, J. Chromatogr. A 796 (1998) 273–282.
7
[4] S.C. Burton, D.R.K. Harding, Hydrophobic charge induction chromatography: salt independent protein adsorption and facile elution with aqueous buffers, J. Chromatogr. A 814 (1998) 71–81. [5] S.C. Burton, D.R.K. Harding, Salt-independent adsorption chromatography: new broad-spectrum affinity methods for protein capture, J. Biochem. Biophys. Methods 49 (2001) 275–287. [6] T. Arakawa, Y. Kita, H. Sato, D. Ejima, MEP chromatography of antibody and Fcfusion protein using aqueous arginine solution, Protein Expr. Purif. 63 (2009) 158–163. [7] T. Arakawa, M. Futatsumori-Sugai, K. Tsumoto, Y. Kita, H. Sato, D. Ejima, MEP HyperCel chromatography II: binding, washing and elution, Protein Expr. Purif. 71 (2010) 168–173. [8] G. Zhao, Y. Sun, Displacement chromatography of proteins on hydrophobic charge induction adsorbent column, J. Chromatogr. A 1165 (2007) 109–115. [9] L.A.Æ. Sluyterman, O. Elgersma, Chromatofocusing isoelectric focusing on ionexchange columns: I. General principles, J. Chromatogr. 150 (1978) 17–30. [10] L.A.Æ. Sluyterman, J. Wijdeness, Chromatofocusing: isoelectric focusing on ionexchange columns: II. Experimental verification, J. Chromatogr. 150 (1978) 31–44. [11] M.T.W. Hearn, D. Lyttle, Buffer-focusing chromatography using multicomponent electrolyte elution systems, J. Chromatogr. 218 (1981) 483–495. [12] T.W. Hutchens, C.M. Li, P. Besch, Performance evaluation of a focusing buffer developed for chromatofocusing on high-performance anion-exchange columns, J. Chromatogr. 359 (1986) 169–179. [13] D.D. Frey, A. Barnes, J. Strong, Numerical studies of multicomponent chromatography using pH gradients, AIChE J. 41 (1995) 1171–1183. [14] D.D. Frey, Local-equilibrium behavior of retained pH and ionic strength gradients in preparative chromatography, Biotechnol. Prog. 12 (1996) 65–72. [15] J.C. Strong, D.D. Frey, Experimental and numerical studies of the chromatofocusing of dilute proteins using retained pH gradients formed on a strong-base anion-exchange column, J. Chromatogr. A 769 (1997) 129–143. [16] C.R. Narahari, J.C. Strong, D.D. Frey, Displacement chromatography of proteins using a self-sharpening pH front formed by absorbed buffering species as the displacer, J. Chromatogr. A 825 (1998) 115–126. [17] Cs. Hováth, Displacement chromatography: yesterday, today and tomorrow, in: F. Brunner (Ed.), The Science of Chromatography, Elsevier, Amsterdam, 1985, pp. 179–203. [18] S. Cramer, Displacement chromatography of proteins, in: A. Kenney, S. Fowell (Eds.), Practical Protein Chromatography (Methods in Molecular Biology), Human Press, New Jersey, 1992, pp. 259–272. [19] G. Guiochon, S.G. Shirazi, A.M. Katti, Fundamentals of Preparative and Nonlinear Chromatography, second ed., Academic Press, Boston, 1994. [20] R.C. Bates, X. Kang, D.D. Frey, High-performance chromatofocusing using linear and concave pH gradients formed with simple buffer mixtures: I. Effect of buffer composition on the gradient shape, J. Chromatogr. A 890 (2000) 25–36. [21] X. Kang, R.C. Bates, D.D. Frey, High-performance chromatofocusing using linear and concave pH gradients formed with simple buffer mixtures: II. Separation of proteins, J. Chromatogr. A 890 (2000) 37–43. [22] N.D.S. Pinto, Novel Hybrid Chromatofocusing Methods for Protein Purification, Ph.D. Dissertation, University of Maryland Baltimore County, Baltimore, MD, 2013. [23] P.C. Wankat, Rate-Controlled Separations, Elsevier, New York, 1990. [24] P. Gagnon, Fine-Tuning Selectivity on Ion Exchangers, Validated Biosystems, 1999, pp. 1, http://www.validated.com/revalbio/pdffiles/ionslect.pdf. [25] T.W. Hutchison, Chromatofocusing, in: J.-C. Janson, L. Ryden (Eds.), Protein Purification: Principles, High Resolution Methods, and Applications, first ed., VCH Publishers, New York, 1989, pp. 149–174. [26] S. Timasheff, G. Xie, Preferential interactions of urea with lysozyme and their linkage to protein denaturation, Biophys. Chem. 105 (2003) 421–448. [27] E.S. Parente, D.B. Wetlaufer, Effects of urea-thermal denaturation on the highperformance cation-exchange chromatography of ␣-chymotrypsinogen-A, J. Chromatogr. 314 (1984) 337–347. [28] Y. Hou, T.B. Hansen, A. Staby, S.M. Cramer, Effects of urea induced protein conformational changes on ion exchange chromatographic behavior, J. Chromatogr. A 1247 (2010) 7393–7400. [29] M.A. Holstein, S. Parimal, S.A. McCallum, S.M. Cramer, Effects of urea on selectivity and protein-ligand interactions in multimodal cation exchange chromatography, Langmuir 29 (2013) 158–167. [30] F. Khademi, A. Mostafaie, Effect of urea on protein separation by ion-exchange chromatography, J. Biochem. 147 (2010) 735–741. [31] S. Ghose, B. Hubbard, S.M. Cramer, Protein interactions in hydrophobic charge induction chromatography (HCIC), Biotechnol. Prog. 21 (2005) 498–508. [32] A. Katti, Y.-F. Maa, Cs. Horváth, Protein surface area and retention in hydrophobic interactions chromatorgraphy, Chromatographia 24 (1987) 646–650. [33] S.T. Evans, A. Freed, S.M. Cramer, Displacer concentration effects in displacement chromatography: implications for trace solute detection, J. Chromatogr. A 1216 (2009) 79–85. [34] A. Kundu, K.A. Barnthouse, S.M. Cramer, Selective displacement chromatography of proteins, Biotechnol. Bioeng. 56 (1997) 119–219.
Please cite this article in press as: N.D.S. Pinto, D.D. Frey, Displacement chromatography of proteins using a retained pH front in a hydrophobic charge induction chromatography column, J. Chromatogr. A (2015), http://dx.doi.org/10.1016/j.chroma.2015.01.087