Expression in CHO-3E7 cells with an optimised method key to producing a difficult to express protein

Background

In this case study Emma Cains describes a series of experiments that ultimately resulted in a 10-fold increase in expression with a ‘difficult to express protein’. The initial work involved comparing expression in HEK293-6E and CHO-3E7 cells, followed by the use of a high cell density protocol with the CHO cell line. Different signal sequences were also evaluated throughout the study and gave a further boost to expression in the final conditions. This 10-fold optimisation was key to being able deliver this protein to our customer to support their ongoing structural biology project.

Sygnature Discovery currently offers CHO (CHO-3E7), HEK (HEK293-6E) and ExpiHEK (Expi293F) expression systems for transient mammalian protein production.

In this study, the protein of interest had been successfully produced several times in HEK293-6E cells at Sygnature Discovery, but at a low expression level (average 120µg/L) and therefore large culture volumes were required (20L) to deliver protein for a crystallography campaign, which was both expensive and unsustainable for the project. The expression had previously been optimised to some extent in HEK cells, and it had been demonstrated that adjusting the proportion of “coding DNA” and “stuffer DNA” increased the expression of protein.

The aim of our follow up experiments described below was to see if we could build on this data and find a way to increase the expression and purified yield of the target protein.

Initial Comparison of Cell Lines and DNA Levels

To start with, an experiment was carried out at small scale (50ml) to compare levels of protein expression in CHO-3E7 cells vs HEK293-6E cells, alongside some DNA optimisation conditions for the CHOs. CHO expression was carried out in the F17 standard conditions supplied by the NRC, at three different coding and stuffer DNA ratios (1, 2 & 3). Expression was followed by a single step purification and run on SDS-PAGE (Figure 1).

Figure 1 – SDS-PAGE of purified small scale protein expression. Arrow marks protein of interest. Numbers 1-3 represent three different ratios of coding and stuffer DNA

The results showed that adjusting proportions of ‘coding’ and ‘stuffer’ DNA modify the expression level of the protein in CHO cells. Coding:stuffer DNA ratio ‘1’ showed higher expression in the CHO-3E7 cells than in the HEK293-6E line.

Confirmation of Expression and Comparison of Signal Sequences

An increased scale expression (500mL) was followed by a multi-step purification process to confirm expression and purity levels for both cell lines. This used the CHO-3E7 cells with optimal DNA ratio ‘1’ above.

All expression to this point had used DNA with the native signal sequence for this protein (ss1). We have observed with other secreted proteins that changing the signal sequence can improve or reduce expression. A construct using an alternative signal sequence ss2 (IgK) was tested here alongside the native signal sequence.
Material was purified by a multi-step purification process, followed by quantification using absorbance at 280nm (Figure 2). The A280 data showed very little difference in expression using the different signal sequences for either cell line. The CHO expression levels were approximately 4-fold higher than the HEK293-6E levels, confirming the improvement seen at small scale.

Figure 2 – Yields of protein from HEK & CHO expression with native (ss1) and Igk (ss2) signal sequences

Equal amounts of protein were loaded onto SDS-PAGE to compare protein purity (Figure 3).

Figure 3 – SDS-PAGE of final multi-step purified protein from 500ml HEK&CHO expression with native (ss1) and IgK (ss2) signal sequences. Arrow marks protein of interest

The SDS-PAGE also showed CHO derived material demonstrated higher purity than seen from the HEKs.

Scale Up

Based on the above data, it was decided to scale up the expression using the optimum DNA conditions in CHO cells using the IgK signal sequence (ss2).
Previous expression grows of the protein had been in HEK293-6E cells at the 20L scale. Due to the increased yield with the CHO cells, expression grows were reduced in size to 10L culture.
The CHO expression grows gave an average of 3-fold higher expression per litre than the previous HEK batches (Figure 4), meaning that a higher total yield was obtained from half the culture volume.

Figure 4 – Yields of purified protein in µg/L following multi-step purification of protein from large scale expression batches in HEK293-6E and CHO-3E7 cells

CHO Expression Method Comparison

The above work all used the standard F17 expression method supplied with the cell line. We have also successfully increased protein expression in these CHO cells using versions of the high-density cell expression method in BalanCD Transfectory CHO (BCDT) medium as reported in Stuible et al., 2018.
To test if this method could increase yields of this specific protein, three small scale (50ml) transfections were carried out in CHO-3E7 cells. One was carried out with the standard F17 method used above, and two used the higher cell density method fed with differing feed methods A and B. All three used the optimised coding:stuffer DNA ratio ‘1’.
Expression was followed by a single step purification and run on SDS-PAGE (Figure 5).

Figure 5 – SDS-PAGE of single step purified protein using standard F17 method, and a high cell density method fed with 2 feed regimes, A and B. Arrow marks protein of interest.

Results suggested some increase in yield with the high cell density BCDT method, with feed B being slightly better than feed A.
Larger 500ml expression grows were carried out to confirm this, using optimised DNA conditions as above (ratio ‘1’), one with native signal sequence (ss1) and one with the IgK signal sequence (ss2), both using feed method B. Material was purified by a multi-step purification regime .
The size exclusion traces from the final purification step of these CHO high density BCDT grows overlayed with traces from the HEK and CHO F17 grows, are shown in Figure 6.

Figure 6 – Size Exclusion traces from final purification step of 500ml cultures.

Pooled peak fractions were quantified using absorbance at 280nm and compared to data from HEK cells and CHOs in the F17 method (Figure 7):

Figure 7 – Protein yields from CHO high density method compared to HEK293-6E and CHO F17 methods.

The yields from the CHO BCDT high density expression grows were both higher than for the CHO F17 method (2 to 3-fold), at 1132 and 874 µg/L compared to 432 and 444 µg/L. This is approximately 10-fold higher than the HEK expression at 108 and 92 µg/L, which is a large increase.
The use of the IgK signal sequence has not had much effect on the yield obtained.
SDS-PAGE of equal protein loads of the two preparations is shown in Figure 8. The purity may be reduced compared to the F17 material, due to this being a prolonged culture and increasing amounts of contaminating host cell proteins in the medium, even though viability remains high. Depending on project requirements, a balance can be achieved between yield and contaminating proteins by optimising the harvest day.

Figure 8 – SDS-PAGE of protein multi-step purified from CHO high density method using native (ss1) and IgK (ss2) signal sequences.

Conclusion

CHO-3E7 cells were successfully used to increase the yield of a difficult to express/low expressing protein compared to HEK293-6E expression.
Switching cell lines provided an approximate 3-fold increase in yield, allowing large scale grows to be halved in volume, whilst increasing the yield. An alternative high cell density expression method was tested with the CHO-3E7 cells, providing a further increase in yield. At the 500mL scale, these changes have provided an approximate 10-fold increase in yield compared to initial HEK293-6E expression. These changes have been key to being able to supply this particular protein for crystallography, with the added benefits of reduction in timelines and cost for the customer.

An optimised version of this high-density method is now central to our standard CHO-3E7 expression protocol for any new protein target. This method has been used successfully for expression from small volumes of only a few mL up to a 20L wavebag.

This case study highlights our commitment to be constantly striving to improve our protocols, especially when faced with a difficult protein to produced. If you have a project that could benefit from this approach then please don’t hesitate to get in touch info@peakproteins.com

Reference

Stuible, M et al; Optimization of a high-cell-density polyethylenimine transfection method for rapid protein production in CHO-EBNA1 cells. Journal of Biotechnology 281 (2018)