Precipitation as an Alternative to Chromatography in the Insulin Manufacturing Process

Peer-Reviewed Article submitted: May 29, 2015 Article accepted: Aug. 19, 2015

Martin McCarthy/Getty Images

Abstract

Recombinant human insulin production using Pichia pastoris typically involves the expression of an insulin precursor as a single chain that is subsequently secreted into fermentation media. The first step in the downstream process after cell separation typically involves capture by ion-exchange chromatography. In the capture step, the protein is concentrated by ion-exchange chromatography, which also results in the removal of a significant proportion of the process-related impurities, such as pigments and host-cell proteins (HCPs) (1). The capture step requires an armamentarium of materials, including chromatographic resin media, various buffers, and highly skilled personnel to handle the process. Although chromatography is a widely used step in this process, there exists a strong impetus to evaluate alternative process technologies from the perspective of process economics.

Modern downstream processing of biologics has elucidated several specific protein precipitation techniques that could considerably reduce the number of purification steps within a process. Some of the specific precipitants (2) used for these techniques include polyelectrolytes, affinity ligands, metal ions, and protein-binding dyes.

The authors evaluated two precipitation strategies-one using polyelectrolyte polyvinyl sulfonic acid (PVS) and the other using zinc chloride-as alternatives to the conventionally used chromatographic process. The authors also compared the benefits of these strategies with ion-exchange chromatography as they relate to product purity, reduction of process-related impurities, and cost of operation.

Materials and methods
Cell-free supernatant used for precipitation trials was generated in the laboratory by a fermentation process using a P. pastoris strain engineered to secrete a human insulin precursor. Polyvinyl sulfonic acid (PVS), zinc chloride, phenol, sodium hydroxide, orthophosphoric acid used for the precipitation trials were procured from MilliporeSigma. The standard Pierce BCA Protein Assay Kit (Thermo Fisher Scientific) was used to analyze the total protein content in the samples. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS–PAGE) was performed using Thermo Fisher Scientific’s Novex NuPAGE 12% Bis-Tris precast gel (1mm*10 well).

A Synergy HT microplate reader (BioTek) was used to measure the optical density of samples. Agilent’s 1260 Infinity HPLC-Chip/MS was used to check the purity profile and specific protein content of the samples.

HPLC analytical method
A high-performance liquid chromatography (HPLC) method-operable at alkaline pH-was developed to analyze post-precipitation samples, because the polyvinyl sulfonate/insulin complex precipitated at acidic pH. A Zorbax C18 column from Agilent (4.6*150mm) was used for purification. Buffer A consisted of 100mM tris and 15mM magnesium chloride and buffer B consisted of 100% acetonitrile (HPLC grade). An in-house qualified insulin standard with concentration of 4 mg/mL was used to calculate the product content in post-precipitation samples.

All precipitation samples were analyzed using a 30-minute reverse-phase HPLC method, and the HPLC column was maintained at 25°C. Elution was done with a gradient formed by mixing buffers A and B as follows: 85–60% B (0–15 minutes); 60–20% B (15–20 minutes); 20% B (18–23 minutes); 20–85 % B (23–25 minutes); 85% B (25–30 minutes). The flow rate was maintained at 1 mL/min, and the column effluent was monitored at 215 nm.

BCA Assay for protein estimation
The authors followed the standard protocol described for the Pierce BCA Protein Assay Kit (3) to establish a protein standard curve. The protein samples were diluted to approximately 1.5 g/L before performing the assay. The BCA reagent (4), mixed with the protein sample (total volume being 200 µl), was pipetted into a 96-well cell-culture plate and absorbance was recorded at 562 nm using a Synergy HT Multi-Detection Microplate Reader from BioTek.

SDS–PAGE
The authors followed the standard protocol for the NuPAGE kit (5). MES running buffer (2-(N-morpholino)ethanesulfonic acid), de-staining solution, sample buffer (5X), staining dye were prepared as per the procedures mentioned in the protocol. The samples were diluted to 1 g/L of product concentration. Five volumes of sample were mixed with one volume of sample buffer, and the mixture was incubated at 85 °C for 15 minutes. Next, 25 µl of the sample was loaded into respective wells of the precast gel. The gel was run at 175V for 60 minutes. Post-completion, the gel was rinsed in destaining buffer for 10 minutes followed by staining for three hours. The gel was destained for nine hours and examined for protein bands.

Experimental procedurePVS-based precipitation
PVS concentrations of 0.10%, 0.25%, 0.50%, and 1% v/v were screened at pH 2.5, 3.5, and 4.5 to determine the optimal conditions for precipitation. PVS was added to clarified cell-free supernatant under mixing and was allowed to mix for approximately 8–10 minutes. The pH was adjusted using concentrated orthophosphoric acid. The mixture was further mixed for 15 minutes and centrifuged at 7000 g for 25 minutes. The centrifuged supernatant was analyzed for product loss. The entire process was carried out at approximately 20–25 °C. The pellet obtained was dissolved in 1M tris buffer and checked for various process attributes such as product purity, specific protein content, and pigment content.

Zinc-based precipitation
Phenol concentration was screened at constant zinc chloride and pH to determine the optimal concentration required for the precipitation process. The clarified cell-free supernatant was treated with concentrations of 0.125%, 0.5%, and 0.75% v/v of 100% phenol and allowed to mix for 10 minutes. This was followed by addition of 5% v/v of 4% w/v zinc chloride solution and mixed for 5–7 minutes. The pH of the precipitation mixture was adjusted to 4.5 by slow addition of 1N hydrochloric acid. Post-pH adjustment, the precipitation mixture was allowed to settle at 2–8 °C for 10–12 hours. The supernatant was decanted, and the remaining slurry was centrifuged at 6000 g for 25 minutes at 23+2 °C. The obtained pellet was dissolved in 1M tris to measure various process attributes such as product purity, specific protein content, and pigment content.

Results and discussion
The development of a precipitation process for a protein to maximize yield and product purity requires the screening and optimization of various factors, such as the nature and concentration of the precipitant, the pH, the temperature, the ionic strength, and the dielectric constant of the medium. The precipitation of an insulin precursor in the cell-free supernatant (CFS) stage strongly depends on the buffer composition and various other factors that are present in the supernatant.

Polyelectrolyte-mediated precipitation
Polyelectrolytes are essentially charged polymers that can either be anionic or cationic in nature. Polyelectrolytes dissociate in aqueous medium, which leads to the formation of protein polyelectrolyte complexes (6). In addition to electrostatic forces, complex formation is also influenced by hydrogen bonds and hydrophobic forces. This phenomenon-polyelectrolyte-mediated precipitation-has been used in the protein purification process.

The earlier work by McDonald et al. on antibody precipitation (7) demonstrated the successful replacement of the chromatography step with polyelectrolyte precipitation. Such precipitation is able to remove host-cell proteins, DNA, and other impurities, and results in a product with a purity that matches that of a product obtained with the chromatographic process. The precipitation of antibodies has no influence on the activities of the antibodies, which enables the iterative use of the antibodies in additional protein purification processes. These antibody results led the authors to evaluate a similar precipitation strategy for the purification process of a human insulin precursor.

Figure 1: Effect of concentration of polyvinyl sulfonic acid (PVS) and pH on product recovery.

Various factors affecting polyelectrolyte precipitation of an insulin precursor include: conductivity of the medium, pH of the medium, and polyelectrolyte concentration. In the case of an insulin precursor, preliminary experiments revealed that precipitation did not occur above the conductivity of 25 mS/cm. However, there exists a flexibility of using higher concentration of polyelectrolytes to counter the effects of higher conductivity of the medium. The effect of pH (Figure 1) can be explained by the charge neutralization of the proteins as the pH increases toward the isoelectric point (pI). Based on the results (see Figure 1), PVS concentration of 0.5% v/v and pH range of 2.5–3.5 were found to be optimal for protein recovery by precipitation. At higher concentrations of polyelectrolytes, the electrostatic repulsion effect between the excess of free polyelectrolyte and polyelectrolyte bound to the protein increases the solubility of the insulin precursor.

Zinc phenol-mediated precipitation
One of the fundamental methods of precipitating insulin precursor molecules involves the use of metal ions. Zn⁺² is the most preferred metal ion because it is specific to insulin-like molecules (8). Zinc induces the hexamerization of an insulin precursor (9) and thus, stabilizes the molecule. Precipitation using zinc ions was also evaluated as a substitute for the chromatography step. Hexamerization of insulin precursor requires 0.3 moles of Zn⁺² for every mole of insulin precursor. In the current study, 5% v/v of 4% zinc chloride solution is added for precipitation, which is in excess of molar requirement for hexamerization of an insulin precursor. In the authors’ unpublished study, it was found that a molar excess of insulin precursor is required to ensure complete precipitation of an insulin precursor.  

Table I: Effect of phenol concentration on recovery of an insulin precursor with a constant zinc concentration.

In the case of zinc phenol-mediated precipitation (precipitant concentration being mainly phenol), the pH of the medium and the protein concentration before addition of precipitants play a key role in recovery of insulin precursor. Based on the results (see Table I), 0.5% v/v of phenol was found to be optimal for precipitation of an insulin precursor. A higher-than-optimal amount of phenol has shown to incur higher product losses due to an increase in solubility of the insulin precursor.

Product quality assessment after precipitation
The product quality obtained through the precipitation method was comparable to the quality of a product purified through traditional chromatography routes.

Figure 2: Sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) for precipitation feed and precipitate samples. Lane 1: Cell-free supernatant (CFS) Lane 2: Clarified CFS Lane 3: Polyelectrolyte precipitate re-dissolved Lane 4: Zinc-based precipitate re-dissolved Lane 5: Ion-exchange elution pool Lane 6: Human insulin drug substance re-dissolved Lane 7: Protein ladder

Certain parameters were monitored to assess the quality of medications produced using precipitation methods compared with products obtained through chromatography. Host-cell proteins (HCPs), host-cell pigments, and product purity (as measured by HPLC) were studied. The SDS–PAGE (see Figure 2) results show a comparable band pattern for the chromatography route and the polyelectrolyte precipitation technique. The optical densities at 600 nm, 450 nm, and 652 nm and the specific protein contents were recorded for the three strategies (the two precipitation strategies in addition to traditional chromatography, as compared in Table II). Reductions in the color of the solution at a constant product concentration indicate the removal of different pigments. The optical densities and specific protein contents of the CFS at various stages are tabulated in Table II. Reduction in optical density is indicative of the removal of the host-cell pigments by the respective strategies. Specific protein is a measure of the amount of insulin precursor relative to the total protein content in the sample.

The product purities of the three strategies measured by HPLC were comparable (see Table II). Considering the purity of the CFS under study, all the strategies gave a significant increase in overall purity. This could be connected to the increase in specific protein content as well after employing various strategies. Enrichment of the specific protein content was accomplished by precipitation strategies, although chromatography is a superior technique.

Removal of pigments prior to subsequent purification steps helps to improve purification efforts. Whether these pigments are associated with the insulin precursor is still in question, but there are instances where these pigments have affected the chromatographic purification of proteins (10). In a separate unrelated study, these pigments-which are understood to be alcohol-oxidase crystalloids-were shown to interact with hydrophobic proteins (such as

CLICK IMAGE TO ENLARGE Table II: Optical densities and specific protein contents at various stages of the precipitation process. HPLC=high-performance liquid chromatography. HPLC=high-performance liquid chromatography.

HPLC=high-performance liquid chromatography. human growth hormone), resulting in altered charge and polarity characteristics of the molecule (11). These pigments considerably affect overall process economics, leading to a lower binding capacity of target proteins, reduced lifespan of the chromatographic media, reduced yields, and lower product purity.

Cost considerations of precipitation
Use of a precipitation strategy is economical compared with the use of an ion-exchange chromatography method. The costs of the major raw materials used to process 10 g of an insulin precursor using PVS-mediated precipitation, zinc phenol-mediated precipitation, and ion-exchange chromatography are compared in Table III. Even though ion-exchange resins can be reused for 80–100 cycles, the cost of processing by ion-exchange chromatography is 3.5 times higher than it is with the PVS-based precipitation process and 10 times higher than it is in zinc phenol-mediated precipitation. Hence, precipitation provides a significant cost advantage in terms of raw materials.

CLICK IMAGE TO ENLARGE Table III: Major raw material cost comparison of polyvinyl sulfonic acid (PVS)-based precipitation and zinc phenol-mediated with ion-exchange chromatography to process 10 grams of insulin precursor.

Conclusion 
Precipitation offers several advantages over chromatography. Precipitation is cheaper, facilitates high-throughput processing, and leads to higher concentration of proteins. Chromatographic processes are limited by the binding capacity of the chromatographic media, a higher cost of goods, and a larger volume of elution pools. In the present context, considering the example of enzyme reaction after recovery of an insulin precursor through various strategies, the authors conclude that use of a precipitation step offers the flexibility of working at higher concentrations and at the buffer compositions suited for an enzyme reaction that is associated with higher yield and purity. Chromatography limits the choice of eluents, which may be composed of chemicals that could affect subsequent processing steps.

The current study did not evaluate a hybrid approach, which would include the use of chromatography and precipitation for purification. A hybrid approach that employed a precipitation step, however, may be a desirable approach in the future for successive chromatography steps.

Purification of biologics typically involves several chromatography steps. Thus, precipitation strategies discussed here reduce the number of chromatography steps required in a purification process by serving to replace capture chromatography. A hybrid strategy could also offer significant cost advantages by reducing the number of chromatographic steps involved in the purification process.

In terms of infrastructure requirement, precipitation involves the use of stirred tanks for mixing and the solid-liquid separation by either using centrifugation or filtration. In contrast, the chromatographic processes require a complex infrastructure of chromatographic systems integrated to buffer tanks, which significantly limits the throughput of the process.

Recommendations
Results obtained in the current study are promising in terms of cost reduction of capture step of purification process without compromising on the quality and functionality of the product. Purification by precipitation offers the opportunity for cost reduction in capture steps without compromising the quality and functionality of the product. Therefore, precipitation is a promising method that could serve as an alternative to expensive chromatography techniques.

References
1. R.L. Fahrner et al., Biotechnol. Gen. Eng. Rev. 18 (1), pp. 301–327 (2001).
2. M.Q. Niederauer and C.E. Glatz, “Selective Precipitation,” in Bioseparation Advances in Biochemical Engineering/Biotechnology, G.T. Tsao, Ed. (Springer-Verlag, Berlin, 1st ed., 1992), 47, pp. 159–188.
3. T. Adilakshami and R.O. Laine, J. Biol. Chem. 277, pp. 4147–4151 (2002).
4. P.K. Smith et al., Anal. Biochem. 150, pp. 76–85 (1985).
5. I.M. Szalo et al., Clin. Diagn. Lab. Immunol. 11 (3), pp. 532–537 (2004).
6. M. Braia et al., J. Chrom. 873 (2), pp. 139–143 (2008).
7. P. McDonald et al., Biotechnol. Bioeng. 102, pp. 1141–1151 (2009).
8. M. Sahyun, J. Biol. Chem. 138, pp. 487–490 (1941).
9. G.D. Smith et al., Proc. Natl. Acad. Sci. USA 81, pp. 7093–7097 (1984).
10. S.A. Minyasab et al., “A method of purifying human growth hormone and purified growth hormone thereof,” US Patent Application WO2010134084 A1, Nov. 2010.
11. L.M. Damasceno et al., Protein Expr. Purif. 37 (1), pp. 18–26 (2004). 

ALL FIGURES ARE COURTESY OF THE AUTHORS.

Article DetailsBioPharm International
Vol. 29, No. 1
Pages: 30–35
Citation: When referring to this article, please cite it as M. Buddha, S. Rauniyar, S. Qais, D. Goudar, S.S. Kandukuri, S. Mahajan, S. Siddik, and P. Hazra, "Precipitation as an Alternative to Chromatography in the Insulin Manufacturing Process," BioPharm International29 (1) 2016.