How come vaccine viruses are so safe while normal viruses are so dangerous?

One method that has been used extensively to generate worthwhile vaccines is that of forcing an initially disease-causing virus to replicate inside a non-natural cell, for example imagine forcing the human specific measles virus to replicate within cells from a chicken. Over time these viruses - all with extremely high mutation rates - will evolve and adapt to the conditions within a chicken cell while at the same time losing it's ability to survive within human cells. The use of these live-attenuated viruses as vaccines has led to the dramatic reduction in a number of important human - and livestock - viruses - the likes of measles, mumps, polio and rinderpest. 

These vaccine viruses usually retain their ability to infect and replicate within their host (one of the reasons why they are so good at protecting us) yet fail to cause significant disease past the odd fever. However, in some cases the use of these vaccines has led to a number of cases where they caused serious illness. Hence, with the powerful immunity these vaccines generate comes the important potential chance of causing disease and therefore the ability to understand and predict how a particular vaccine will behave following administration is key to continuing the safe use of live attenuated vaccines. Sadly, our knowledge of the mechanisms behind virulence and attenuation are largely unknown.

Vaccination of the man behind the mumps vaccine - Maurice Hilleman's daughter, Kirsten, while her half-sister, Jeryl Lynn, looks on. The MuV vaccine strain Jeryl-Lynn was originally developed from the virus which infected Hilleman's daughter of the same name through non-natural replication. http://www.historyofvaccines.org/

One example of this can be found with the case of the mumps virus (MuV) vaccine. Interested in mumps? see here. Before the widespread use of the vaccine, MuV was responsible for the majority of cases of aseptic meningitis (inflammation of the lining of the brain not caused by bacteria) in the western world - this virus is highly adapted in entering the human central nervous system and replicating within the epithelial cells that line its inner layer, making it particularly dangerous. Because of this, any vaccine batch produced must go through rigorous pre-clinical testing, the majority of which is carried out in primates. Yet sometimes, vaccine viruses will lead to the development of aseptic meningitis following immunization. We still do not adequately understand the molecular basis of the ability of MuV to cause disease in the nervous system - something that may facilitate the development of new and improved MuV vaccines.

Sauder et al, publishing in Journal of Virology, explore the genetic basis of MuV neuropathogenesis through the generation of around 30 chimeric viruses - see below - comprising genes of a neurovirulent MuV (strain: 88-1961) and a highly attenuated MuV (vaccine strain Jeryl-Lynn 5). Through the analysis of the potential for these viruses to cause disease in a rat model of mumps meningitis they were able to assess the contribution of specific genes - or combinations of genes - in either increasing or decreasing its ability to cause disease.

Chimeric mumps viruses (combinations of attenuated - JL and virulent - 88). What effect will each have on pathogenesis?

MuV has a single-stranded RNA genome of 15,384 nucleotides and encoded within this one molecule are 7 genes through which at least 9 different proteins are expressed.  These proteins - and hence their corresponding genes - govern the basic biology of this virus: building of the virus particle, receptor binding, cell entry, transcription and replication and finally, cell exit. N, P and L forming the replication apparatus of the virus while M, F, SH and HN are involved in particle assembly, entry and exit. It is these same proteins that are responsible for the ability of MuV to cause disease in humans, specifically aseptic meningitis. Any understanding of the ability of a virus to cause disease must identify its key molecular - and genetic - components hence, what particular genes along the MuV genome are responsible for causing aseptic meningitis in humans? Is it those that allow the virus to enter the cell? Those involved in replication? or those involved in building the virus particle? Sauder et al sought to try and convert an attenuated virus into a virulent one and vice versa - and in doing so uncover the biology behind MuV pathogenesis.

Can we transform the virulent 88 strain to an attenuated virus by inserting combinations of attenuated JL genes?

Above shows the results from the initial experiment comparing disease caused by the different viruses: This is where the added different genes from the attenuated virus to the virulent to see whether or not this weakened the viruses ability to cause disease. As you can see, no individual genes or combinations added to the 88-1961 virus caused it to be as attenuated as the vaccine strain, suggesting that in this case with MuV attenuation is a complex, polygenic trait involving many genes. Neither the transfer of all replication proteins (N, P and L) nor the assembly proteins (M, F, SH and HN) resulted in complete attenuation. The most dramatic effect was seen with N and M transfers - although the reason why was not addressed. 

Can we transform the attenuated JL strain to a virulent virus by inserting combinations of pathogenic 88 genes.

Following on from this, the group tried to see whether they could turn the attenuated into the virulent virus through carrying out the reverse of the above experiment although this time the results were not the same as no genes or combinations resulted in anywhere near the levels of pathogenesis seen for the 88-1961 virus. Even with the addition of the previously effective N and M combination.
 

These results tell us a number of things about MuV attenuation and virulence, firstly: it's a lot more complex than we might have previously thought! - this is not all down to one gene but a few of them working together in combination. Secondly, it doesn't work both ways - the mechanisms behind how a virus causes and doesn't cause a disease are different as the same genes couldnt do both. And lastly these results point us to some interesting avenues of future research into virus attenuation - what are the specific molecular roles each of these genes play in turning the virulent virus into an attenuated one? For example, why do the attenuated N and M have such a drastic effect? More work will be undoubtedly be done to determine these questions and this may allow us to rationally attenuate viruses in the future using these combinations of genes.

Sauder CJ, Zhang CX, Ngo L, Werner K, Lemon K, Duprex WP, Malik T, Carbone K, & Rubin SA (2011). Gene-specific contributions to mumps virus neurovirulence and neuroattenuation. Journal of virology, 85 (14), 7059-69 PMID: 21543475