Figure 1 A and B (bottom): Separation of two proteins, Cytochrome C (-) and Ribonuclease A (-), on CEX resin (A) and Nuvia cPrime Resin (B).

Chromatographic analysis incorporates two different phases—the stationary phase and the mobile phase, which contains the protein of interest. The stationary phase is usually composed of resins that are functionalized inert supports that can differentially bind to the mobile phase components and provide separation. Based on the physiochemical characteristics of resins, they can provide single modes of separations, such as size exclusion (SEC), affinity, ion exchange (IEX), hydrophobic interaction (HIC) or a combination of any two or more of these modes leading to a multi-modal or mixed-mode (MM) separation.

Not all resins are created equal

A wide array of resins with diverse properties is available to help with complex purification requirements and challenges. However, no single resin currently on the market can encapsulate all the properties needed for complete one-step purification. Hence, typically a series of purification steps are performed with different resins contributing to the workflow to achieve the final purification goal. 

The choice of the appropriate resin for each step of the workflow depends on multiple factors. The source of the initial feed from which the biomolecule is to be purified is an important consideration. Sources such as bacterial expression systems, cell culture supernatants, native serum and plasma introduce different types and/or quantities of impurities, calling for different purification strategies. The initial salt concentration of the buffer containing the biomolecule of interest also contributes to the selection process. For example, IEX requires the initial load’s salt concentration to be low, whereas HIC often requires high salt-containing loads. On the other hand, proteins can be eluted from IEX columns at high salt concentrations and from HIC columns at low salt concentrations.

The physiochemical properties of the biomolecule of interest are also very important selection criteria. Some proteins exhibit a phenomenon called “salting-out” at the high ionic concentrations required for HIC purifications. Some proteins have a potential to lose their activity in a relatively short time period, warranting the use of resins like IEX and HIC, which allow high flow rates. The same resin type (e.g., IEX resin) can have variable bead chemistries leading to different properties, which could lead to different outcomes for the final purification goal. For example, if you are performing high throughput purification, it is critical to select a resin that can perform the purification while maintaining its flow properties. 

The initial load volume is also a consideration point. SEC columns can help with small volume separations but are not compatible with large volume loads. Another important consideration for SEC columns is that they cannot be used with highly concentrated protein samples. This excludes them as a viable choice for the initial step in process purification of such samples. 

Finally, the level of purity to be attained also determines the workflow and the resins to be used. Proteins purified for therapeutic needs often require the most stringent purification workflows to attain the maximum purity and retain the required activity. The bottom line is to get the highest biomolecule purity in the fewest steps possible.

Mixed-mode resins/media

Mixed-mode resins combine the functionality of two or more unimodal resin types. This allows a single resin to bind biomolecules through two or more interaction modes, which can potentially decrease the number of required purification steps in the workflow, require less sample material, offer better separation in the same time, and help in separating impurities that are very similar to the target biomolecule. Mixed-mode resin-based purification can be performed in both flow through and bind/elute mode, just like purification using unimodal resins. Since this method is a combination of two or more purification chemistries, it usually saves processing time and effort in the long run, since the eluate from one column does not have to be processed (diluted/concentrated/desalted) before loading on the second column.

Mixed-mode resins also offer unique selectivities that cannot be matched by traditional unimodal resins. For example, an experiment was performed to purify two proteins—Cytochrome C (pI = 10.7) and Ribonuclease A (pI = 8.7)—using the same conditions with a traditional CEX resin and a mixed-mode resin. The binding was done with Buffer A (20 mM Na acetate + 150 mM NaCl (pH 5.5)). After washing with Buffer A, the elution was done with a progressively increasing gradient of Buffer B (20 mM Na phosphate + 1 M NaCl (pH 7.0), up to 25 percent B in Buffer A over a total of 30 column volumes. 

Creation of this buffer gradient on the packed column showed the impact of buffer pH and conductivity simultaneously on the two proteins to be separated. The proteins did not show good separation on the CEX resin (Figure 1A). They were eluted at almost the same time, with Cytochrome C showing only a slightly higher retention time than Ribonuclease A. The higher retention time is explained by Cytochrome C’s slightly more basic nature, which would lead to a stronger electrostatic interaction with the cation exchange resin. In contrast, good separation was achieved on the Nuvia cPrime Resin (Figure 1B). The retention time of Ribonuclease A was considerably different than Cytochrome C, leading to the proteins eluting at different time points. 

Based on these results, Ribonuclease A is more hydrophobic than Cytochrome C and hence shows better retention on Nuvia cPrime Resin (Bio-Rad), which is a multi-modal resin that allows both cation exchange and hydrophobic interactions. This result emphasizes the importance of selecting the most optimal resin for efficient purification.

Selecting the optimal purification conditions

Although the pI and molecular mass of a protein can generally be predicted by its sequence—other properties such as physiochemical and conformational properties that can affect its interaction with different resins and impurities cannot be predicted easily. This leaves a gap in determination of ideal purification conditions. An alternative approach to creating buffer gradients in order to determine the best conditions for efficient purification is to use a Design of Experiment (DOE) approach. The working conditions established by a small-scale DOE can then be used for process purification of the target.

To illustrate this point, a DOE approach was designed for a diverse set of proteins with varying pI. The binding capacity and recovery of these proteins on Nuvia cPrime Resin were assessed under 11 different experimental conditions of buffer pH and NaCl concentrations. The optimal conditions predicted by DOE (Table 1) suggest that these proteins interact by different modes with the resin under the explored conditions and these modes cannot be completely predicted by their pI. 

Test Protein   pI  Optimal Conditions Predicted by DOE
Bovine serum albumin  4.7 Binding: 10 mM NaCl, pH 4.0
Elution: 1,000 mM NaCl, pH 8.0
Bovine carbonic anhydrase  5.9 Binding: 10 mM NaCl, pH 4.6
Elution: 1,000 mM NaCl, pH 8.0
Conalbumin  6.9 Binding: 10 mM NaCl, pH 4.0
Elution: 505 mM NaCl, pH 6.0
Lactoferrin  9.2 Binding: 205 mM NaCl, pH 4.0
Elution: 1,000 mM NaCl, pH 8.0
mAbX   9.5  Binding: 300 mM NaCl, pH 4.6
Elution: 800 mM NaCl, pH 8.0

During in-depth analysis, maximal binding of Bovine serum albumin and Bovine carbonic anhydrase was seen at the lowest pH and NaCl levels tested, whereas complete recovery was attained at high pH and NaCl levels—indicating that the main interaction between these proteins and the resin involves the cation exchange (electrostatic interaction) of the Nuvia cPrime Resin. Lactoferrin and mAbX are basic proteins and therefore expected to interact strongly by electrostatic interactions. However, the optimal binding capacity for these proteins was seen at substantial NaCl concentrations, suggesting that the binding of these proteins is actually enhanced by hydrophobic interactions. 

High pH and salt levels were needed for complete protein elution, indicating that charge-charge interaction was involved in association of these proteins with Nuvia cPrime Resin. Thus, a simple DOE screening can predict the chromatographic behavior of diverse proteins on a given resin to a large extent.

Selecting the optimal workflow

Since each resin works optimally within a set range of technical parameters, complete purification of complex proteins, especially antibodies, warrants the need for multiple consecutive purification steps with different resins. A typical workflow would involve two or three steps for the capture, intermediate and/or polish purification of the biomolecule. Each of the selected resins and the workflow conditions have to be orchestrated for a smooth and efficient purification process. Usually a protein A-based affinity resin can be used for the initial capture. 

Alternatively, a high-capacity cation exchanger can also be used for non-affinity capture. The potential of acidic proteins, DNA or other negatively charged impurities in the feed binding to an anion exchange resin and thereby decreasing its binding capacity should be taken into consideration before selecting AEX for the capture step. Such resins work well for the intermediate purification steps. Mixed-mode resins can be used in both intermediate and polishing steps of purification.

Another consideration, while selecting the resins for each purification step, is to ensure that the process scale up from small to production scale is aligned to maintain the process economy and product quality. Ultimately, the final choice of resin for each purification step crucially depends on the biomolecule to be purified and the anticipated impurities. Some complex proteins cannot be purified completely with off-the-shelf resins.

Since a considerable number of resins with different chemistries, features and requirements are available in the market, the process of selecting ideal resins can be confusing. The rule to live by is to select resins that can provide maximal target purification in the shortest amount of time, with the least effort and at the lowest cost possible.