I tried so many times but didn't get anything. It is a tag with the gene located around 66bp on this tag. There is a problem in the case of the C-terminal tag not in being in the N-terminal. Please advise.
3kb products are not impossible to get. If you have tried optimizing the PCR conditions and it still have no product you might want to think about designing a new primer set (if possible).
if the gene is well described, and sequences likewise, new primer sets are an option or gradient checks. if the gene is not well-described and sequenced, consider reviewing analogous genes from different organisms and input those into CodeHop which then will design primers sets in Blockmaker. (Blockmaker > Designed degenerate PCR primers based on multiple protein sequences) Both are great programs and have helped me amplify multiple 3kb+ analogous genes using sequences from different organisms.
3 kb should not be a problem for most enzymes, but if difficulties continue despite optimization, numerous "long PCR" enzymes are available. As an example, I have used "Elongase" from Invitrogen in the past.
3kb should be straightforward. As alluded to above, 'touching down' onto the annealing temperature is one way to go. Alternatively, if the gene is known, try making it in pieces and either splicing it together (boring) or using overlap PCR (exciting!)
If you face the problem with only the C terminal tagged version, it suggests the primer is the problem. If you cant redesign the primer due to cloning constraints, then consider amplifying it in two amplifications and using Gibson assembly to Join it together. You could even amplify most of the gene and perform a gibson assembly using synthetic oligonumcleotides to introduce the tag. Check out the NEB site for gibson assembly kits, they are easy to follow and I have been using them with some succsess. If like you say you have optomised the PCR and managed to amplify it with an N terminal tag, it suggests the primer or primer tag combination is the problem
We do tagged PCR with two forward primers. One long primer containing tag and complementary sequences from gene of interest for initial tethering of primer to template.... run first 5-10 cycles using long forward primer and reverse primer. After 5-10 cycles add shorter forward primer (20 bp) which is complementary to 5' end of tag sequence. 3 kb product size is not a problem.