Baculovirus Gene Mutations and Protein Expression

Have you ever wondered why the baculovirus we use for most protein expression purposes is called AcMNPV?  This is short for Autographa californica nucleopolyhedrovirus, which derives from the Latin name of the alfalfa looper, a pest of alfalfa crops.  The convention for naming baculoviruses is to use the host insect from which they were isolated.  This does mean that some viruses can effectively be named twice, if they are isolated from different insect species.  A good example is Trichoplusia ni nucleopolyhedrovirus, which is virtually identical to AcMNPV.  It is no surprise, therefore, that AcMNPV-based expression vectors replicate very well in cell lines derived from T. ni (also known as the cabbage looper, which is a pest of cabbage crops).  These cell lines, together with those derived from Spodoptera frugiperda (fall armyworm) e.g. Sf21 and Sf9, form the mainstay of hosts for expression of recombinant proteins in insect cells.  However, while we use Sf cells for both virus amplification and protein expression, we reserve T. ni cells for the latter purpose and never use them for making recombinant virus stocks.

Of course, if you inoculate T. ni cells with AcMNPV they will produce some infectious budded virus.  In comparison with virus stocks amplified in Sf cells, the yield of infectious virus may be 100-1000-fold lower.  This makes it useless for subsequent protein expression work as the infectious titre (plaque forming units per ml) is too low to achieve the high multiplicity of infection needed.

A further complication with using T. ni cells for propagating budded virus stocks is that the virus genome is prone to picking up mutations that can affect recombinant protein production.  The two most common mutation events seem to involve the insertion of transposable elements into the virus genome or a simple nucleotide addition in FP-25.  The latter mutation has the most serious consequences for protein production.  The 25 kDa protein product from FP-25 is required for very late expression of baculovirus polyhedrin, which is the promoter most commonly used in the expression vector system.  You might wonder why the odd mutation in a population of viruses can be such a problem.  After all, compared with RNA synthesis, DNA polymerases have proof reading mechanism that limit errors in nucleic acid copying.  Unfortunately, a side effect of FP-25 mutations is that the yield of budded virus in T.ni cells increases at least 10-fold, sometimes up to 100-fold.  You don’t have to be a maths genius to see that any mutation within FP-25 can rapidly become fixed within the virus population.

Although mutations to FP-25 can also occur when using Sf cells to amplify virus stocks, the increase in the titre of infectious virus is only about 3-4-fold.  This means that in contrast to T. ni cells it takes a lot longer for the mutation to become dominant in the virus population.  For most projects, you don’t need to go beyond a P4 virus stock, so there shouldn’t be any problem with a drop off in protein yield in Sf cells owing to FP-25 mutations.

We do sometimes hear of people having problems with loss of recombinant protein production over a period of time.  While we hesitate to attribute these to accumulation of mutations within FP-25, it is worth bearing in mind our comments above and trying to minimize the number of times you sequentially amplify a recombinant virus.

There is a way to avoid the effects of FP-25 mutations and that is to use the other very late baculovirus gene promoter (p10), which strangely is refractory to loss of a functional 25 kDa protein in insect cells.  Unfortunately, not many transfer vectors utilise the p10 promoter for expression, other than those (e.g. pOET5) that make use of them for dual synthesis of recombinant proteins in baculovirus-infected cells.  Therefore, given the focus on the use of polyhedrin-based expression vector systems it is worth bearing mind how successive passage of recombinant vectors in insect cells may result in undesirable changes in the virus genome.