PCR can be very sensitive to seemingly minor and irrelevant process details.  When your reaction is "right on the edge" between working and not working, it can happen that it works on DNA prepared on one day but not on the same DNA prepared on another day.  It can stop working when DNA has been stored for a different length of time, or it can vary between persons or between thermocyclers. 

So, to cut a long story short, it sounds like it's probably premature to credit the breeding pair, as there are probably many (in my view) more plausible process details that would be causing the variability, and it would take significant study to investigate them.  I mean there could be a genetic distinction in these mice that does the trick, yes.  But it's just so unlikely compared with process details.  After all PCR is known to be finicky in this way.  The saying goes "pray to the PCR gods" for good reason.  

So for your purposes I think the question would be how to make a PCR generally more robust.  A comprehensive answer would be more than I can handle right now, and there are plenty of threads about that on RG.  As a starting point, maybe try lowering your annealing temperature by a few oC.  If that doesn't bring false bands up, it might just push your reaction from "on the edge" into "robust" territory.  

If that doesn't do the trick, one would  need to know more details of your sequence and your method to start troubleshooting for real. 

PCR can be very sensitive to seemingly minor and irrelevant process details.  When your reaction is "right on the edge" between working and not working, it can happen that it works on DNA prepared on one day but not on the same DNA prepared on another day.  It can stop working when DNA has been stored for a different length of time, or it can vary between persons or between thermocyclers. 

So, to cut a long story short, it sounds like it's probably premature to credit the breeding pair, as there are probably many (in my view) more plausible process details that would be causing the variability, and it would take significant study to investigate them.  I mean there could be a genetic distinction in these mice that does the trick, yes.  But it's just so unlikely compared with process details.  After all PCR is known to be finicky in this way.  The saying goes "pray to the PCR gods" for good reason.  

So for your purposes I think the question would be how to make a PCR generally more robust.  A comprehensive answer would be more than I can handle right now, and there are plenty of threads about that on RG.  As a starting point, maybe try lowering your annealing temperature by a few oC.  If that doesn't bring false bands up, it might just push your reaction from "on the edge" into "robust" territory.  

If that doesn't do the trick, one would  need to know more details of your sequence and your method to start troubleshooting for real. 

Hi Kimberly,

Is it possible that you are working with heterozygous knock out mice? Weak bands on the gel could be explained by a reduced amount of template in your PCR. Are you checking for both alleles (wild type and mutant)?

However, some small differences during the experimental steps are more likely.

Best,

Boas

Totally agreed with John, PCR on genomic DNA is particularly tricky and can strongly depend on the gDNA extraction method as well as the operator.

My first suggestion is to waste a bit more time to obtain a fairly clean gDNA (so having at least a salting out with saturated NaCl) and avoid the dirty preps with just boiling and EtOH precipitation after the PK digestion.

Secondly, if you feel that the PCR is "solid" enough and your gDNA is fairly clean, be aware that too much gDNA heavily inhibits the Taq polymerase. Sounds counter-intuitive, but that's how it is! Try diluting 1:10 some of your gDNAs that give weak bands and re-run the PCR... you might be surprised... If nothing changes or it gets worse, then your PCR conditions definitely need some optimization...

Good luck!