Some ligands will have ionic areas on the molecule (which is what I suppose you mean by charges), such as an amino group (R-NH3+ at physiologic pH) or a carboxyl group (R-COO-). And amino acid side chains within the protein binding site can be like that as well. But the presence of a charged functional group is not necessary for ligand binding. You can have ion-dipole interactions (there would be a charged functional group with that), dipole-dipole (no charged group), hydrogen bonds (no charged group), and hydrophobic van der Waals interactions (no charged group) that all increase binding affinity. There probably are issues regarding the presence of water molecules as well (aqueous solubility), but that's a supposition on my part.

We really don't use the term bonding for the insertion of a ligand into a protein binding site. It's binding, a much more general term. You don't actually form a bond (covalent, ionic), but of course you can have a hydrogen bond, which are transient and reversible. The most important thing for a good fit, however, is a matching of the shape/conformation of the molecules. The hand in a glove analogy is a good one.