I found a commercial reagent called in vivo-jetPEI® from Polyplus-Transfection that is able to delivery mRNA in vivo. I'm planning to do an experiment with it, but I'm worry about how to synthesise mRNA and purify them clean enough to inject to living body. I also found a recommended Kit called HiScribe T7 High Yield RNA Synthesis Kit that can help me to synthesise mRNA but I'm not sure if I can use the final product for an in vivo injection. Anyone have any experience with it?
You should purify your RNA after transcription from any buffer/enzyme contaminants. There are several kits available in the market for that. One is GeneJET DNA & RNA Cleanup and Concentration that uses standard silica-beds packed columns to absorb nucleic acids. You can then elute them in pure water and you should be ready to go. Once the RNA is purified you will use the vivo-jetPEI for the delivery.
My preferred method is to use the Ambion T7 Megascript kit, but I haven't made a full comparison to other kits. They should all be more or less the same. But the transcription itself may not be enough.
Be sure that the mRNA contains a 5' cap. This is needed to bind the translation machinery and produce protein. You can include this in one of 2 ways, a cap analog, or capping enzymes. The cap analog is easier, because you just put the analog in the transcription mix. But this reduces yield, because you typically need to replace 80% of the GTP with your analog. You also won't get 100% capping, the best case scenario would give you 80% capping, but typical results are often much lower than that. If you use an older cap analog, half of the caps will be inserted the wrong way, but ARCA will prevent that. The Capping Enzymes are usually more efficient, but not cheap, and adds an extra reaction and purification step to your protocol.
A good mRNA also needs a 3' PolyA Tail. This can be added enzymatically, or it can be built into the DNA template. The enzymatic method isn't hard, it just uses E. Coli PolyA Polymerase and ATP, and can often be added to a mRNA reaction without additional purification in between. But the tail length would be a little random. Building the tail into the DNA offers more control over length, and eliminates an additional reaction step. But ordering a DNA with 100 to 200 consecutive A's is a little difficult, because quality control is hard with consecutive bases.
Finally, the RNA needs to purified. The method alec_djinn describes isn't bad, but I prefer to use Phenol:Chloroform extraction and isopropanol precipitation. It's rigorous, but it takes more time and uses potentially dangerous chloroform. You can find details on the internet, or just ask me for instructions.
I've made lots of RNA this way, and injected that RNA into a lot of mice. But I'm skeptical of JET-PEI, or any PEI product, for in vivo use. Protecting mRNA in vivo is very hard, and the simple ionic interactions between PEI and mRNA fall apart in the bloodstream. PEI is probably better than nothing, but take a look at the data on that page. The RNA all goes to the lungs. This behavior is typical of a cationic particle that aggregates in the blood, and these aggregates get trapped in the narrow capillaries of the lungs. If this was normal PEI, that mouse would die. Also, they're using 50 µg doses of mRNA. I was doing my work with 1 µg doses, but I was using a different type of transfection reagent and hydrodynamic injection. I also don't see any numbers reported on their bioluminescent imaging data. It's pretty easy to make that data look better than it is by adjusting the scale.