Co-infection with Baculovirus Vectors for Multiple Protein Expression

A question that comes up from time to time is the use of co-infection with baculovirus vectors for multiple protein expression in insect cells.  The need arises if you have produced a number of baculovirus vectors that each synthesise a component of a multi-subunit complex, such as an enzyme or a virus like particle.  Intuitively, you know that it is probably necessary to add many virus particles per cell to achieve efficient infection with each recombinant.  But how many viruses can you add to each cell and still get each one into every cell in equal numbers?

There is no easy answer to this.  We don’t know of any experimental work published that has looked into the issue.  However, one paper that explored this statistically (Belyaev et al., 1995) concluded that if virus particles are distributed randomly across cells at a given multiplicity of infection, the Poisson distribution with mean determines the proportion of cells receiving a certain number of virions.  In a typical co-infection, it can be assumed that each virus type is distributed between cells independently.  For 2, 3, 4 or 5 virus co-infections at a moi of 5, the percentage of cells lacking one or more virus types is low (1.3, 2, 2.5 and 3.4% respectively). Unfortunately, if the optimal ratio of expressed proteins is provided by an equal ratio of infecting viruses, a minority of cells will receive such ratios as the numbers of virus types involved are increased.  In layman’s terms, or in words our CEO can understand (he hates statistics!), the more different viruses you add to cells the fewer will receive equal numbers of each.

Following on from this, with 2, 3, 4 or 5 virus types each at a moi of 5, theoretically the percentage of cells receiving an equal ratio of the virions is 12.78, 1.88, 0.29 and 0.05% respectively.  The consequence of this is that there will be an unbalanced synthesis of the individual proteins in the cells.  However, unless you can examine synthesis of your multiple protein targets in individual cells, which is not impossible with modern confocal microscopy techniques, you will not be aware of the problem other than a lack of your multiple component target complex.  Simple protein gels just inform about total expression in all cells.

If you do want to produce a multi-component protein complex in insect cells you should consider inserting all of your coding regions into a single virus, or at most only two.  Belyaev et al. (1995) also addressed this issue by constructing a recombinant baculovirus containing 5 sequences for expression.  Other methods are also available. We also considered the possibility for differential expression of foreign genes in a single recombinant baculovirus in a recent blog post.

Co-infection with baculovirus vectors for multiple protein expression is feasible, but as we have described above it does have problems of ensuring all cells receive equal numbers of virus particles.  You might ask if viruses can be added sequentially to overcome the problem.  We will talk about this problem in a future blog and why it is not such a good idea.


Belyaev et al. (1995) High level expression of five foreign genes by a single recombinant baculovirus.  Gene 156, 229-233.