1- If you really want to teach, go to the interviews, even if you have no chance of getting the job. The fact is, the post-docs are probably better qualified than you, and so is junior faculty, but at least you get your name out there. You never know when someone suddenly can't teach, and there is no-one to replace them. They'll remember you.
2- This is not a seminar or a scientific talk. You need to forget about everything you've learned about presentation so far. You need lots of text on screen and to go slowly. If you talk without text, the students freak out: they write down everything. I swear some of them even wrote down my jokes. Giving handouts helps, but you don't want to give away too much or they won't come to class. I found 25 slides was a good size for a 1:20 lecture. I know it sounds crazy, since as a talk that would last me no more than 30 min, but it forces you to take the time to explain. I found that repeating myself each time with a slightly different figure, and highlighting different aspects worked well to get the point across. And it's better to finish early than to cram too much into one lecture.
3- Include anecdotes and factoids into your lecture but only if it is directly relevant. I found the students really related to them but it's also a slippery slope. For example I contrasted a certain aspect of physiology which differs between humans and most other mammals, only to have many students give it as an example of human physiology in the exam. It's clear that it's best not to confuse the student with superfluous facts or give exceptions to rules.
4- Don't ask questions directly to the class. I tried it once or twice just to see, and no one bit. But I know from having been a student, that the question is either too easy, in which case you don't want to sound stupid by answering it, or it's too hard and then you look like you're trying to show off. Rhetorical questions are fine. In any case, if the material is interesting the students will come up with questions by themselves. Giving them examples of questions that they could find on the exam is also a good preparation.
5- Exam questions are surprisingly hard to write. For multiple choices I abhor the questions which include combinatorial answers such as a) ... b)... c)... d) a+b, e) a+c. I always thought it was laziness on the part of the professor, so I made them as clear as possible. No point in tricking the students. The long answer questions are much more sensitive. I found that the language used must lend itself only to one interpretation. For example there is a slight difference in asking which hormones are important for implantation (of the embryo), versus which hormones are important during implantation.