My 4D world's equivalent of Tiktaalik, the ancestor of all terrestrial endoskeletal animals, is a tetralaterally symmetric octopod with two limb girdles holding four limbs in the front and four limbs in the back. (Note that because this is in 4 dimensions, the planes of radial symmetry in which the limbs are attached are fully perpendicular to the plane formed by the spinal axis and the vertical axis.)
Among 3D animals, it seems pretty much universal for jointed limbs to have a joint with maximal freedom for the shoulder / hip (a ball joint for tetrapods, and a compound hinge joint for arthropods), then a simple hinge join that can only rotate in a single plane (i.e., an elbow or knee), and then variously complicated stuff for the terminal segment of the limb--i.e., extended tarsals/metatarsals than turn the wrist/ankle into an additional or reversed elbow/knee equivalent, all sorts of different foot/fin/wing structures, etc. But all the specialized stuff aside, you pretty much always see a maximally generic joint with the body, then one or two simple hinges.
It is easy to generalize that to 4 spatial dimensions. The only big change it that you need at least 4, rather than 3, muscle groups to control the full range of motion of the shoulder / hip joints. But is that actually the ideal structure for a 4D limb? Do we need full range of motion of the shoulder / hip in 4D, or can it be limited to a 3D range (being a hybrid between a hinge and ball joint that only exists in 4 dimensions) with only 3 controlling muscle groups? And if we do need the full geometrically possible range in the shoulder and hip, how about the elbow and knee? Can those remain simple hinges anyway, or do they need greater freedom as well?