ESTRUCTURA DEL MANUAL DE FUNCIONES 48

Initial shake flask fermentations of E.coli BL21(DE3) pET21a expressing HCV proteinase were carried out in LB media at 37°C. These fermentations were undertaken to produce material for investigation of the HCV proteinase system, and applications of proteomic technology within fermentation and downstream processing. SDS-PAGE analysis indicated that HCV proteinase had a molecular weight of approximately 22kDa, and was expressed inclusion bodies. As is commonly found when expressing recombinant proteins within an E.coli host, a large percentage of the recombinant target is expressed in an insoluble form, and can be found in the pelleted fraction of the lysed cell. When activity is retained within the insoluble form of the protein, protocols can be developed that can solubilise the protein and retain the activity. For example, one strategy is to modify the protein at the genetic level by addition of an affinity tag to the recombinant target. This can increase the percentage of soluble form of the protein, and increase the ease of downstream processing steps (often only one column required) (Mohr & Pommerening, 1985). These tags can be removed with relative ease by cleaving them from the protein post purification (e.g. factor Xa), providing that the cleavage site for the tag is available to the removing agent. Tagging of proteins in this way represents an ideal way for capture of recombinant proteins, but can also produce problems when using the product for further study. For example, the presence of a tag can compromise crystallographic data. With these considerations in mind, a protein was chosen that was untagged for this study.

One of the most common methods of increasing the level of a recombinant protein in a soluble form is to decrease the rate at which it is expressed. Slowing the rate at which a protein is synthesised can reduce the formation of inclusions bodies within the cell. Ways in which protein expression rate can be reduced include the reduction of temperature at which the protein is induced, and reduction of the concentration of induction agent (e.g. IPTG). In the case of HCV proteinase, the temperature of induction was lowered. In the first instance the temperature was lowered from 37°C to 30°C. This reduction did not significantly increase the level of soluble HCV proteinase, as indicated by SDS-PAGE investigation. A further reduction to 26°C also failed to increase the level, but decreasing the induction temperature to 22°C produced increased amounts of HCV proteinase in the soluble extract. These variations in induction temperature and increases in product

production are also accompanied by an increase in fermentation length (i.e. the productivity), which have to be considered.

The investigation of the HCV proteinase production at this small scale produced interesting results when the protein ‘ladders’ within the SDS-PAGE gels were examined (Figure 3.3). This basic investigation of the E.coli proteome indicated some significant shifting of proteomic output by the cell by altering the temperature. Proteomic shifts will therefore result in changes in the composition of host protein contamination of the product. These variations in protein content could then be exploited downstream, as individual proteins that may, for example, contaminate active fractions, can be reduced or eliminated at an early stage of the process. In order to investigate the effect of the fermentation temperature on the proteome of

E.coli, cells expressing HCV proteinase needed to be investigated by 2-DE.

With this in mind, a set of experiments was devised that exerted more control on the growth and expression of the recombinant protein, so that the proteome from these cells could be investigated. The cell cultures were then increased in scale, with 4 fermentations conducted in identical bioreactors. This allowed control of dissolved oxygen content, temperature, pH and impeller speeds. Additional comparability was also gained as all fermentations were inoculated from the same starter culture. The cells were grown to a predetermined optical density and then induced with 1mM IPTG. As would be expected the growth of the 4 fermentations did not vary significantly before induction as they were grown under the same conditions and inoculated from the same culture. Following induction growth varied as the temperature was modified according to the experimental design. At different points of induction, HCV proteinase production was monitored using a specific assay. Soluble and active HCV proteinase production was greatest at 22°C (49pg/mL) and 26°C (17pg/mL). There was no active protein produced in the fermentations at 30°C and 37°C. This may be due to the formation of inactive inclusion bodies at the higher temperatures as was seen in the previous shake flask cultures (Section 3.2). This implied that promotion of active recombinant protein expression was dependent upon the operating temperature, with lower temperatures producing increased amounts compared to higher temperatures.

The biomass from these fermentations was recovered using a membrane filter and secondly by centrifugation. Using this biomass, samples were normalised and then investigation by 2-DE was begun. As described in this chapter (section

3.5.1) different methods of sample preparation, and running conditions were investigated in order to view the maximum number of proteins within the proteome, and also produce gels in which spots could be differentiated and analysed. Following this lengthy procedure, the proteomes of the biomass from the 4 fermentations were investigated using 2-DE. After gel images were captured, spots were detected, their intensities normalised, and then matched throughout the experiments. This analysis revealed that the protein composition of the cells at the point of harvest were greatly influenced by fermentation temperature, and that protein levels fluctuated under these varying conditions. Analysis by eye reveals many changes, but in order to quantitate the levels of proteins within each proteome, spot intensities needed to be calculated (using Phoretix 2D), and then a method devised in order to make sense of this data. The method used was to group the protein intensities according to their isoelectric point and molecular weight, and then to compare them across the proteomes. Grouping this data within a ‘master grid’ allowed comparison of areas within the E.coli proteome.

Following the investigation of operating temperature on the E.coli proteome, HCV proteinase was isolated in order to view the protein on a 2-DE gel. This also gave an opportunity to examine proteins that were co-eluting within the active fraction, and also to examine if HCV proteinase could be viewed on a 2-DE gel following negative results with whole lysate material. This revealed that under the conditions at which the 2-DE was run, HCV proteinase was not taken up within the lEF gel. Smearing within the gel at the cathodic end of the gel and at the molecular weight of HCV proteinase also compounded this view. In order to examine if the protein exhibited a higher isoelectric point that was above the maximum pi of the lEF gel, a pH 6-11 lEF gel was run. Again this gel showed smearing at the anion electrode (results not shown). Further investigation of higher pH gels were not run due to limitations of the pre-cast lEF gels. The experimental pi of HCV proteinase was therefore not determined.

Overall the HCV proteinase system has provided a platform on which to develop the proteomic capability for this project. This in turn has also led to development of a novel way of analysing the spot data. The windows of host protein contamination that were produced provided an ideal way of comparing E.coli host proteins under different conditions (operating temperature). This however still had limitations in the way that data could be viewed. In order to provide an output from this analysis that could provide rapid answers, a comparison of these operating

conditions needed to be developed in a more user friendly way. As development of the HCV proteinase system has shown there is a solubility issue surrounding its expression, and also in viewing the protein as a spot within a 2-DE gel. With this in mind, it was decided to investigate a new system using the techniques and expertise that were developed with HCV proteinase.

In the following chapters of this thesis a different protein was selected for analysis (glutathione S-transferase) due to the considerations outlined. Therefore future work concerning data analysis methods is considered at a later point.

CHAPTER 4. A PROTEOMIC APPROACH FOR SELECTION OF FERMENTATION