I guess micro injections may not be a simple answer. You have to inject the sequence within the genome and that to have to fuse with the main genome. I mean that because a vector has the expression machinery, which by simple promoter may not be achieved.
Question is not clear therefore you may not get the answer you wish to get. You can target gfp using gene targetting proteins to desired location in any genome and insert whereever you want by cleavage induced homologous recombination. This will allow you to generate the fusion of interest and express your protein of interest under the native promoter or any desired promoter you supply with. The problem with these techniques is that they are rather difficult to apply at the moment.
either the delivery of mRNA, in this case you can use a transfection reagent dedicated to mRNA delivery such as RmesFect,
or, for me the best option will be to deliver directly the protein right into the cytoplasm by a transfection reagent. 2 to 3 h after transfection, the protein is delivered into the cells. The important point is that you can deliver active proteins (such as enzyme) and also tagged protein for live cell imaging.
Protein delivery using Pro-DeliverIN reagent works very well as demonstrated by several papers; of whom the ones by Pérez LM. (Diabetes. 2013) :http://www.ncbi.nlm.nih.gov/pubmed/23423565) and Maucksch C., (J Stem Cell Reg Med. 2013): http://www.pubstemcell.com/monthly/008030300132.htm)
Do not hesitate to contact me if you need more information,
I will suggest the electroporation. you may read this article for further informations :Nanochannel electroporation delivers precise amounts of biomolecules into living cells