Also I undestand that some viruses (retro-viruses?) have special parts of their genetic information with high mutation rates and that this helps them avoid being eliminated. It is also my understanding that other parts are much more stable -resistant to mutation to help them keep theior essential features.
There are several things going one here. First I would like to address a simple principle:
1. what works stays, what doesn't work is eliminated.
There is no active mechanism as such in place that sees in the future and tries to keep things that might be good for the future. The high mutation rates are therefore not a means of safeguarding against an uncertain future, but are the result of a highly stressful past! A large conceptual difference that has to be addressed although of course a lot of people will say that it is obvious.
2. mechanical constraints.
I know that in the Eukaryotic DNA there are regions that have a higher mutation rate than other regions. And I am talking about actual mutation rates, not the selection that will take place after the mutation has taken place. Some regions just see more mutations than others.
I do not know if this is also the case in viruses. Their genome is rather small. One would expect that it would be more difficult to have regions that have different qualities that raise or lower mutations rates.
Certain regions in the viral genome do evolve faster than others, regions such as the ones responsible for virion assembly and genome packaging usually evolve slower and are used for phylogenetic studies. But in this case it is not a reflection of mutation rate, but structural importance.
Fast mutation rates of RNA viruses
There are different types of viruses (like for instance the broad classes of DNA vs RNA viruses - however, in reality the complexity of the grouping is of course much more prominent), and some classes have higher mutation rates just because of how they work. RNA viruses mutate faster than DNA ones. Some RNA viruses mutate faster than other RNA virus families.
Genetic variation can be induced in the RNA virus through various mechanism, such as mutation, homologous and nonhomologous recombination, and genome segment reassortment. Different virus families use these mechanisms in different ways.
Especially in RNA viruses errors in genetic duplication are common, but variable among the different classes! The error rates in the duplication of RNA viruses can vary between 10-[sup]3[/sup] to 10[sup]-5[/sup] (sorry, can't find superscript). The error rate is the amount of substitutions per round of copying.
The error rate of duplication in DNA viruses is much lower.
The main reason for this high mutation rate during duplication is the low efficiency of the proofreading capability of the RNA replicases and transcriptases.
3. the principle of arms race/environmental stress
For the virus the high mutation rates aren't necessarily bad. A high mutation rate guarantees the presence of a high degree of variation. In times of environmental stress (ass in the host is not cooperating nicely) this can lead to rapid evolution.
However, with mutation rates that are too high you can pass the so-called
error threshold. It's value depends largely on the information that needs to be maintained and value of the mutated genome compared to the original. Passing the error threshold will mean that essential genetic information will be lost. Not good.
RNA viruses replicate and mutate very close to this error threshold. Simply increasing the mutation rate by experimental means can push such a virus beyond the threshold into doom and extinction.
general reference:
Domingo and Holland. Annual Review of Microbiology Vol. 51: 151-178
