Okay, just shooting in the dark here. (I'm given to speculation and painting myself into difficult corners, but... whatever. Ain't writing a thesis here.)
Here's, perhaps, a (grossly simpole) model. There are, let's say, five genes that regulate the size of the upper canines. Let's say there is some variety in each of these alleles, so that variation in each, with all the others held constant, causes an increase in tooth size, but these alleles are intially rare enough that the occasion of an individual being homozygous for them at two different loci is very low.
Nonetheless, there occasionally is an individual who is homozygous for big teeth at two of these loci, and it's got 6-inch teeth instead of the mean 3-inch teeth. Now, let's say this is really advantageous, and this individual is in a small, isolated population, so that after some number of generations most of the individuals are homozygous for both of these alleles.
Now, being homozygous for a big-tooth allele at a third locus bumps the tooth-length from 6 to 8 inches...
And homozygotes for big teeth at a fourth locus have 10 inch teeth.
And homozygotes at the fifth locus have 12 inch teeth.
Just a very crude model. Not a lot of morphological traits are influenced by one or two genes.
As to the real question -- could there be mutations which increase the susceptibility of a genome to mutation -- it's conceivable, I suppose, but I don't know of any examples. Perhaps something that regulates meiosis could have an effect. I just read in a text about regions of DNA that appear to have a propensity for moving around the genome, but like any mutagen they're likely to cause a great deal more harm than good.
There is apparently a mutation (perhaps a common one?) in DNA polymerase that increases the overall mutation rate. However...
What they found is that the higher mutation rate (the mutator strains) does not necessarily accelerate the pace of fitness evolution. Only small population coupled with high mutation rate will speed up the fitness evolution. Figure 1 & 2. This is because in a small population, the clonal interference is reduced within the population. Thus, the population spends most of its time waiting for beneficial mutations. In addition, mutators are very common in asexual populations, since they are most likely to hitchhike to high frequency by introducing mutations at a higher rate and thereby create beneficial mutations more often. It is very likely that to have both mutators and beneficial mutations coupled in one E. coli. They will not accelerate fitness evolution due to clonal interference. Within the population, there will be more beneficial mutations as a result of the increased mutation rate by the mutator mutation. However within the population, all beneficial mutation are competing against each other to be the fittest. Thus the process still remains the same. Clonal interference also imposes a speed limit on adaptive evolution in asexual populations, because two or more beneficial mutations that arise in different lineages cannot be combined into the same lineage. (That what sex is good for) Thus the rate of evolution will be capped, or independent of the mutation supply rate.
Not sure how this would apply to animals (since it appears here to be a bacterial thing). A researcher I worked for claimed to have found mutator-like strains of Plasmodium falciparum (malaria), which is a eukaryote but still unicellular and, so far as I know, primarily an asexual reproducer.
This is something I've mulled over, but I don't know enough to posit any mechanism. I mean, you could have mutations in the enzymes responsible for correcting mutations, but that's not likely to be a cell that lives very long, let alone happening in the germline and getting passed on.
(Mayhaps I'll be able to say more when I'm done with genetics, animal development, and bacteriology, but I very much doubt it.)