Whenever I write to a professor or researcher who I don't know in person, I make sure to write a formal letter. For the first letter, I introduce myself briefly (I'm an undergraduate at university of michigan, my research interests are mainly higher category theory, categorical homotopy theory, and commutative algebra, etc.) (just don't make it too long) and explain why I'm writing him or her. I make sure that I've done my homework (see the MO FAQ/howtoask page) and explain why I'm writing (oftentimes, it's that nobody in the department is working on the things I'm trying to learn, so it's very hard for me to just ask for some help in person).
I make sure that the recipient knows that I understand that he or she is probably very busy, and that I really appreciate him or her taking the time to read my e-mail. Then I ask the questions I'd like to ask. I think it's best to keep all of your communication very formal until you've met in person. The professor/researcher you're emailing is doing you a big favor by explaining a point that is probably trivial in the grand scheme of things.
I don't know if these are necessarily always the right things to do, but I've always gotten a good response by doing it this way. Most professors and researchers seem to be very happy that you're reading their work, and as long as you're not annoying them, they seem pretty happy to discuss it.