It's mostly an issue of semantics,in retroviruses the LTR function as the only transcription regulatory element so it function as a both promoter and enhancer.

Thanks a lot Yoav. Then, the expression of the vector could work without CMV promoter?

I have a two vectors;

one of them with CMV promoter- Moloney murine sarcoma virus 5´LTR and packaging signal-puromycin-CMV-cloning site-Moloney murine sarcoma virus 3´LTR.

the other one has CMV promoter- Moloney murine sarcoma virus 5´LTR and packaging signal- a protein of interest and Moloney murine sarcoma virus 3´LTR. So, my doubt is that if this protein of interest after integration in the host genome will be able to express without have the CMV just forward.

I can't be 100% sure without seeing the map of your specific vectors but from how many vectors are built I think I can give you a relatively reliable answer.

The first CMV promoter is there in order to generate the initial viral genomic RNA that will be packaged, the LTR need to be part of the transcribed sequence in this case.

Then, post infection, in your first vector the Puromycin gene would be expressed from the LTR while your protein of interest would be expressed from the second CMV promoter.

In your second vector the protein of interest would be expressed from the LTR.

thanks again for your time and your help. I have to choose between this two vectors. It is suppose that both are the same with the differences i have told you before, but I didn´t understand well why one of them has one CMV and the othe one has two. When I worked with lentivirus I always have a promoter such as CMV or SFFV just forward my protein of interest, but this vectors are new for me, so I have started to have doubts. 

Thanks Yoav!