Pluto is inclined in a way that makes it slightly off the orbital plane of all the other planets and sun. Could you have a planet that was inclined at 90 degrees, or at least close to that? What would make something like that happen (i.e. it collided with something and somehow the force was enough to take it all the way up to that type of orbit, or could it be a captured planet?) A planet couldn't form that way from the protoplanetary disk, could it? If something crashed into it hard enough to make it do that, how big would the planet orbiting have to be? The thing that crashed into it?
While still not totally clear if the question ask for an inclined orbit or an inclined rotation, I will offer my two bits for both questions possible.
First, Pluto for example - while no one yet was able to reconstruct what happened after all as far as I know - can have several major options to get his orbit after all.
I can think of the possibility that he didn't start his carrier as a planet(like object) after all, but was a giant moon of one of our gas giants for some time, until... good question. But if he had a huge altitude every time his parent planet hat its closest approach to another gas giant, a tiny momentum of pull might have happened. I have to admin that this scenario is pretty unlikely, but still...
At one point he might just got the notch to make him escape his parent planet after all and went off for adventures of his own. Maybe he would dwell around between Jupiter and Saturn for some time, until he get a random gravity assist (break or accelerate) that would change his orbit once more. And because its unlikely that he went in exactly at the same plane as the pushing random gas giant, he got a free of charge lift away from the planets plane after all.
A wild theory, I know.
Other options were named here already: a "small" celestial body like Pluto could get thrown off course by an impact, especially if this happens at his descending or ascending node (where its effect would be most effective). He could have met a rouge planet aeons ago, that stole some orbital speed away so orbit became eccentric and/or inclined. And last but not least, it could be a huge alien structure hauled to this place a long time ago; as soon as it detects ftl travel from earth, it will call in a huge artillery strike at this location because the aliens don't want us fooling around in that peaceful galaxy.
To be honest, I can't offer much to this topic. Still, a huge impact may have the ability to add a minor vector to the spin, but that wont work with huge planets.
Further what I will call "bowling ball effect" because I cant recall the scientific name may. A minor or major spot of heavy whatever located way outside the rotational center will cause some... well, it wont rotate as good as before. Can be supplied by random impact encounters. Now add some moons and a non circular spin and... well, maybe not.
For Uranus I can offer a theory someone who claimed to be studying astronomy told me years ago. After all, even we on earth are afraid that our magnet field will shift some day. Some say this is caused by the core who think it might be a good time to change its rotation. However, because we are sitting on a pile of heavy molten stone and metal, the magnetic field alone wont be enough to have a huge impact on rotation direction. BUT if you are a piece of gas (in different forms), sudden change of the cores rotation direction may be able to pull the light stuff at the outside hard enough to make it copy your new way of spinning, just by applying magnetic forces. Chances are good that below the surface metallic hydrogen still is spinning the way it did some aeons ago. But this should cause atmospheric trouble beyond imagination...
Up to this day I never heard this last theory again, so use with caution.
So... happy planet spinning. Oh, you might use the universal sandbox to find out if the "from moon to planetoid with shitty orbit" explanation from me can be true; I do no have this toy sadly.
Rogue planets are a likely cause: a planet formed around another star, flung out of orbit by a supernova or the passing of another large star or some-such, and then captured by the sun. They could end up at any angle.
Perhaps two suns collide, each with planetary disks at right angles to each-other, their planets captured by the resultant bigger star.
I don't think this is possible via a collision. The amount of energy it would take to alter the orbital plane this much would be well in excess of the planet's gravitational binding energy: it would come apart and the debris would disperse.
It's conceivably possible via a capture of a rogue planet, ejected from another star system, or by perturbation from a star passing close by. The later will likely change the orbits of everything orbiting our star.
That no one mentioned Uranus? Shame on you! (not including you, Blackknght)
The fact is, we have just such an example in our solar neighbourhood. Unfortunately, a lot remains unknown about the icy giant. But there was a theory that early on the planet was hit by a protoplanet (a precursor to the planets we call neighbours today) near its magnetic north (I think, don't quote me on that) pole.
Without large close moons, the axis can be torqued gradually due to other bodies naturally. I've seen animations that show that without the moon Earth would wobble beyond 90° over billions of years. The same seems to be true of Mars.
As for forming, you have angular momentum in each piece’s rotation and orbital motion. When they collide the hit might be at any point on the body near the pole, perhaps. This changes the vector at which torque is referenced and the same closing velocity could impart a spin in any direction.
The accretion tends to go with the general flow on average. But the last few big things coming together can show the result of luck, as with Venus. Just look at how much angular momentum you can impart with a collision, and compare that with the spinning posesed by both bodies already.
For changing the asis of an already formed planet, you need to look at precession over a long time, not a quick job.
I really recommend the SETI Weekly Seminar series for learning about stuff like this. There have been many presentations on planet formation and planetary system dynamics.