you might try amplifying 2 overlapping shorter sequencesthen ligating them

I tried amplifying the overlapping sequence and ligate them by Givson assembly, but always failed. The squenching results showed either none of the TOPORS gene insertion or partial insertion. Additionally, there are not many E.coli transformants since the PCR fragments were ligated together with plasmid vector.

Hello! It seems like you are proceeding in decent and standard manner. The DMSO is kind of a bummer. I know it is a standard PCR optimization reagent, but I would recheck the sequence and redesign primers with a higher Tm. Also, there are polymerase for long reads and some GC rich. I think you may prefer the GC tolerant polymerase. 3kb is not considered too big. The NEB Phusion work at higher temperature which could be good or bad, as you want to make sure you still have three steps in your PCR. denature, anneal, and extend. Phusion is pretty good so I would try new primers first, than GC tolerant (NEB) polymerase. So you can use DNAworks and optimize codons at the same time. This worthwhile if you need to modify the transcript. 3kb is kind of expensive, but people do it. Good luck!!

I forgot to mention you can use PrimerBLAST and optimize those primers to PCR amplify that gene. I match the Tm of the primers within. 2 degrees C. I hope you get it.