We get asked to do many service projects for customers where they want to produce a modified form of a protein for structural/functional studies. Frequently, these represent examples that have already been expressed in other studies, often to high levels, so the expectation is that yields will be good in these new projects. Everyone sets off on the road to discovery with high hopes. Usually, these are fulfilled and both parties are happy!
Imagine the chagrin, therefore, when occasionally the first results come up on the protein gels that show much lower and occasionally zero levels of protein expression. We all scratch our heads, repeat the expression tests, try different conditions (multiplicity of infection), time to harvest, cell line, incubation temperature of cells etc. Usually, this solves the problem but occasionally fail to enhance protein production significantly. So what is going on in these rare instances?
Unfortunately, as so often in science, there is no easy answer. We would dearly love to be able to wave a magic wand and consign poor protein expression levels to history. However, what has emerged as a common thread linking the projects where protein expression is low is the fact that one or more mutations have been made to the gene target prior to expression. Further, in most projects where genes are synthesized from scratch, the codon usage is optimized for invertebrate cells because, well that is what people do! So immediately we have two variables in a synthetic construct that may affect levels of protein expression. Whether or not codon optimization is necessary remains conjecture. It can affect mRNA secondary structure, which can be a factor in the efficiency of translation. Perhaps it is not surprising that if you alter the gene target, which nature has evolved over millions of years, you get unintended consequences – such as poor expression levels. Changes in your gene may affect protein structure, rendering post translational processing difficult or promoting degradation within the cell.
So what’s to be done? Well, if you want to make a particular mutation in a gene target for further analysis of protein function, you may be wise to include a control gene construct that expressed the native protein. This would tell you if your desired gene mutation is affecting the levels of recombinant protein. At this point we can hear you say that you can’t afford to make two copies of the same gene. Actually, it can be cheaper than you think, as many synthetic gene providers offer minor variants of a target at less than the cost of the complete gene. If you have several regions to modify, you may consider designing what we like to call “trapdoor” restriction enzyme sites to permit subsequent modification of these sequences.
Although making multiple gene constructs may appear to be time consuming and expensive, if you use a system such as OET’s flashBAC™ vectors for baculovirus expression, the process is actually very easy. With flashBAC™ you just have to mix it your transfer vector prior to cotransfection of insect cells, wait five days and you have a P0 stock of recombinant virus. A further five days to amplify a working virus stock and you can start testing expression of your gene target. Therefore, comparison of a number of modified gene targets becomes very easy.
To sum up, we suggest that if you are considering the expression of a gene with any modifications which deviate from the native sequence that you have the original protein made in parallel to determine the effect of these changes. The jury remains out on the effect of codon optimization on protein expression in insect cells but we are hoping to do some work in the next 12 months to gain some insight to this issue.
Finally, before beginning any protein expression project, even before contacting us, do your homework in the library (sorry, online!) and search out past publications. Delve back in time and see what others have done with the same or similar proteins. If there are many publications describing successful expression and subsequent purification, often with crystallographic studies, the chances are that your project has a good prospect of a positive outcome. If your target has few such reports, it is possible that others have tried with little success –negative results are rarely published.
Comments