In principle it is definitely possible. Multicopy Pichia strains have been shown to exhibit genetic instability, for instance; see e.g. Article Can too many copies spoil broth?

What samples are you using for PCR? Cells from the finished culture, or cells from the working cell bank? Also, have you done PCR with primers specific for the heterologous gene instead of simply primers flanking the homology regions?

Hope it helps!

Thanks Alejandro... I am using aox1 primers and haven't tried using gene specific primers. but can that solve the problm?

Also I have another doubt... under methanol induction does the pichia genome aox1 gene also over express?

If the problem is that the strain is indeed losing the heterologous gene somehow then no, PCR won't solve the problem. But...

Suppose that for some reason the band containing the target gene is amplified less efficiently than the wild-type gene (different %GC, presence of secondary structures, longer amplicon, what have you). In that case the yield gap between each amplicon will increase with each amplification cycle. If you stop the reaction early enough, or the initial amount of template was low enough, you might have a situation like what you describe, and the situation might change with seemingly minute changes to reaction conditions. So you would end up with what you have (the target amplicon appears sometimes, and some others it doesn't).

The most rigurous approach would of course be relative quantification using qPCR. If the strain is "losing" the gene (or if your cell banks are contaminated with cells of the wild type host which overgrow the recombinants once methanol becomes the only carbon source -this would be especially plausible in mutS hosts) you should observe a decline with time in the copy number of the heterologous gene.

And yes, the AOX1 gene product is also overexpressed under methanol induction (that is one of the reasons the pAOX1 promoter was selected by Cregg & co).

Hope it helps!

Thanks a lot.. it was indeed helpful.. I'll try out your suggestion..

Is there a possibility that after losing the heterologous gene the clone still remains his + ? Since my clones are able to grow on minimal dextrose media but does not contain the gene.