A note on Mutazyme and Manganese

Can Mutazyme be powered up?
Adding manganese works poorly, but stronger variants seem to be known.
(see also discussion about Mutazyme in Part I and II)
Mutazyme is great as it is not as biased as manganese mutagenesis which requires imbalanced nucleotides to correct, but annoyingly cannot reach a really high mutation rate. The maximum number of doublings is lower than the number of PCR cycles, so that imposes a limit to about 20 doublings. Assuming an error-rate of 0.9/kb/doubling, the hard limit is about 20 mutations per kb, which means that a 100 AA protein cannot be mutated to more than an average of 7 nucleotide mutations, which means 4 amino acid changes* (70% variants have between 2–5 mutations).

*) 58% of nucleotide mutations are amino acid mutations if all mutations are equal. This is an empirically determined figure.

Manganese

So what happens when manganese is used to increase the error-rate?

Curiously, the Mutazyme PCR reaction fails if 1+ mM manganese chloride is added. It gives a faint band, not multi-banding. Mutazyme is simply a homologue of Pyrococcus furiosus polymerase that lacks the proof-reading activity. The dNTP concentration is the same as normal PCR and the primers only twice higher. But Mutazyme does not seem to behave with StEP either or with more than 40 cycles, so it is not a very happy protein, so this is not too surprising.

This is not really the end of the world as the mutation rate hoped for is too high for epPCR (redundancy issues and frameshifts), while degenerate primer based strategies are more recommended, but it's a shame it does not stack.

Genetics

So what exactly is mutazyme?
It is actually exo– Phusion, which is Pfu DNA polymerase B with sso7d fused to the C-terminus.
The patents are horrible to read but that is what they seem to suggest.
The proof-reading domain (exo) is known to be killed with a D215A mutation.
Potentially it could be made more mutagenic with D473G (Biles and Colloney, 2004), but the numbers don't match so the 471–3 loop must be unaltered.
Also, I have been told USER cloning does not work with Mutazyme, so it is not V93Q like Phusion U (Nørholm, 2010).