Why don't you "just" use an MMR deficient mouse? Are those the wrong substitutions? Otherwise use a NER mouse from Rotterdam or an ATR mouse from Hernandez-Capetillo

Yes, it's a possibility. But my goal is to use the existing mouse model ( we deliver a cre virus in the lung of multiallele floxed mice) so i can limit the DNArepair deficiency to the very same cells where i induce tumors with cre. therefore, engineering the cre virus is simpler and I dont have to cross a lot of mice to bring a new allele in the strain I need.

Which strains, precisely, would you suggest ? thanks

You should talk to Hein te Riele...I think he has mouse strains with cre alleles for MMR genes.

As far as I know most people in the DNA repair field work with mouse models, not viruses. Although cell culture work can possibly give insight into what construct you can use. But I am less familiar with that literature. I think these two reviews can help you decide what DNA repair pathway you will want to attack (Schumacher, 2008 Trends Genet and Hoeijmakers, 2001 Nature).

Thanks again for the insightful comments.

Indeed I realize it won't be possible to mimic tobacco carcinogens by simply knocking down/out a DNA repair component. I just wanted to use a genetic trick to switch the tumor genetic from being copy-number driven to mutation driven and ideally (very ideally) introduce a similar type of dna damage as tobacco carcinogens.

In my model, constant, fairly unspecific mutational burden shaping the tumor genome is desirable, because that is indeed what smokers experience. I guess the selective pressure will do the rest. The impaired DNA repair will occur only in infected cells.

In this sense, the genetic trick (e.g. an shOGG1- cre virus) is way simpler than exposing mice to carcinogens for prolonged time (dose/schedules are difficult to optimize and effects quite variable). It also solves the problem of systemic effects that I would encounter using a full KO mouse; I cannot afford the mouse to develop many unrelated tumors, and if one day I want to administer chemotherapy to already long-treated mice it might get confusing.

Finally, I think an shOGG1 will do. Ogg1 KO seems to be cause lung cancer in the long run and it's good to know that such genetic deficiency cannot be easily bypassed by other components of the same pathway.