You haven't given enough details to understand what you are doing. Are you preparing a single plasmid with both mutations, or are you trying to do sequential gene replacements in C. glutamicum? And is this gene replacement of a wild type allele on the chromosome or integrating a gene that is not present in the host?

Maybe, the combinations of the punctual mutations are resulting non-viable cells and that is the reason you get always back the wild type genotype. It is not usual to do two or more mutations at the same time. It is better to make sequencially. If you have one, you can try to get the second one and it should be easier to verify if the combination are killing the bacterium. In the secondary metabolism is rare to see this effect but in the metabolic engineering of the first metabolism could be more common.

Michael J. Benedik I am trying to do sequential gene replacements, one after another (the plasmids contain 1kb homologous arms with a single point mutation in the middle of the sequence). The creation of the first mutant strain containing the mutation Nr. 1 worked. Now I want to insert the mutation Nr. 2 and I just get the WT genotype all the time no matter which insert DNA I use.

@Carolina: these mutations should rather make the bacterium grow faster. I am doing re-engineering from an evolved strain, and I have tested their impact in single mutant strains.

Are the two mutations fairly close to each other? If you are trying to make the double mutant in two steps, then you are requiring one of the recombination events to occur between the two mutations. Depending upon their distance, this might be a low frequency event. Assuming this is the case, then you probably just need to screen a lot.

It might prove easier to create the multiple mutant alleles in E. coli first and then just transform the entire region into your strain.

Michael J. Benedik the first and second mutation are 700 kb apart. The third mutation is 1500 kb apart from the second.