One way of interpreting the normal Aharonov-Bohm effect is that it implies that the electromagnetic potential is more "real" than the electromagnetic field. We tend to think of the electromagnetic field as what's "really" out there because if you stick a charge in an electric field you can measure the field strength by measuring the force the charge feels or you can measure the force on a moving charge and calculate the magnetic field. By contrast the electromagnetic potential is not uniquely defined which makes it feel like a mathematical construct. You can add any function with zero divergence to the electric potential and any function with zero curl to the magnetic potential and have a potential that behaves identically for any experiment you could do (these are different gauge choices in the parlance).
The classic Aharonov-Bohm effect is that you take a particle and pass it around a magnetic solenoid in a superposition of two states --- say one state goes to the left of the solenoid, the other goes to the right. In a sense the particle takes both paths simultaneously. The magnetic field is zero along both paths the particle takes. If you turn the solenoid on and off you would think it would make no difference because no matter what you do, the magnetic field the particle encounters is zero. The key, though, is that the magnetic potential is not zero along the particle's path, and the magnetic potential affects the particle's phase in measurable way when the two states recombine.
Sometimes I see interpretations that this means that there is a small but non-zero chance that the particle is found inside the solenoid where the magnetic field is non-zero, but I don't think that's right. You can construct the experiment in a way that the chance of this happening is negligible. The real implication seems to be that the electromagnetic potential is the more physically fundamental quantity even though it is not generally directly observable.
All the math for the Aharonov-Bohm effect applies for gravitational fields as well. But here it's not quite as "weird" a result because gravitational fields can't be shielded the same way magnetic fields can be. The particle always encounters the gravitational field along both paths, so it's not so spooky that the field has a measurable effect.