I'm going to use topGO package to check GO enrichment in my data. I have a my so called "gene universe" and an list of genes of interest. But if a some genes of interest doesn't have any GO terms what will happen?  Will the software just skip those genes  or skip but introduce corresponding correction on initial number of genes of interest? To my understanding if sufficient part of my genes of interest will lack the GO terms it could lead me to false conclusion about enrichment. Right?

Genes that have no GO terms are invisible to the process. They neither contribute to the universe of available GO terms, nor to the set of interesting genes. It's as if they don't exist, which in this context is true.

If you have a large set of genes that could hypothetically have a GO term, but do not for whatever reason, and a lot of those genes are in your set of interesting genes, then you may conclude that the GO term of interest is not enriched when in fact it might be. But that isn't a problem with topGO, but a problem with the level of annotation for your organism of interest.