I guess there is some complementarity between the primers, due to the sequence surrounding your target site. This is only a guess as I never faced this problem.

goodluck

Have another look at your primers. How long are they? Have you tried to extend their length and GC content? Another idea is to end your primer with stabilizing bases (G/C) and increase the annealing temperature. Hope it helps.

Cheers,

Hi Andrew,

Some one had the same question and here is my answer:

Your primer likely forms multimers by finding some complementarity within itself. 

Solution: Just give 5 more colonies for sequencing, it is unlikely that this could have happened to every clone. If all clones show this, you have a problem to solve. Even if one is correct, just move on with your problem. These anomalies are sort of common and hence always give 3 or more clones for sequencing. 

If all clones have this issue, then slight increase in annealing temp or the many variations of hot starts may help. Reducing primer concentration and number of cycles may help too. I feel one of your 15 clones will get you your mutant. At this time just sequence more clones. 

To Andrew,

I always give more than 3 clones for sequencing and sequence the mutated position first. If it is OK, then I ask for the entire gene to be sequenced. Saves money when you have longer inserts. Monika's suggestion to increase annealing temp is a good one. Having too much primer is an additional suspect.

Here is what I used do to avoid this: I add the template and primer and buffer and then heat to about 75C for 2minutes. Then cool on ice for 2 minutes and then add the remaining components. The idea is to break down any primer dimers/multimers and also denature the template just enough. This step allows a majority of the primers to find the right target in the template so that your first cycle of amplification is  very efficient. Hence the subsequent cycles will be more accurate too (that is the hope). 

Hope this helps