Newton's Third Law, as it is most commonly worded, reads:
Every action has an equal and opposite reaction.
Note that there is nothing in this statement about the action and reaction forces being "on a line between the two objects." This is because this addition is not in general true, both because there are explicit counterexamples (take any non-central force, like friction, for instance) and because there are situations where a "line between the two objects" makes no sense (for example, in the case where a charge interacts with a field).
In any case, if we want to understand why the properly-written form of Newton's Third Law always applies, we have to make sure that the terms here are properly defined.
When we say action and reaction in Newton's Third Law, we are referring to a pair of forces that a pair of interacting objects (sometimes called an "action-reaction pair") exert on each other. A force is defined as an interaction in which momentum is transferred over time from object A to object B (where the convention is to refer to "the force exerted by A on B"). The magnitude of each component of the force is precisely the rate of momentum transfer in that direction, which is why we write $\vec{F}=\frac{d\vec{p}}{dt}$. (Incidentally, this is equivalent to the general case of Newton's Second Law.)
So, if we were being precise about our terminology, Newton's Third Law would read:
If objects A and B form an action-reaction pair, then the force exerted by A on B has the same magnitude and the opposite direction as the force exerted by B on A.
Using the definition of force, this statement reads:
If objects A and B form an action-reaction pair, then the rate at which the momentum of A is changing due to interaction with B has the same magnitude and the opposite direction as the rate at which the momentum of B is changing due to interaction with A.
In plainer language:
When objects A and B interact, any momentum lost by A due to that interaction is gained by B, and vice versa.
So, once we have properly defined action and reaction, Newton's Third Law is essentially a statement that momentum doesn't appear or disappear in an interaction. Applied over any possible interaction, this is a restatement of conservation of momentum.
So the next question one might ask is, "Why is momentum conserved in the first place?" The answer to this relies on a very deep result by celebrated mathematical physicist Emmy Noether. Noether's theorem essentially states that every conserved quantity of a system arises from a fundamental symmetry of that system. In the case of momentum, that symmetry is translational symmetry. So momentum is conserved because the fundamental laws of physics do not change as you change position in the universe.
Of course, you could also ask, "Why don't the laws of physics change as you change position in the universe?" I don't know that there is a widely-accepted answer for this question as of yet, other than the fact that we have observed it to be the case, and that dropping this assumption generally leads to theories in which it is very difficult or impossible to calculate much of anything.
Addendum:
Many times, you will hear that "Newton's Third Law is violated" in certain situations (a common one that tends to pop up is the case of two moving electric charges). This is a misconception borne from the fact that oftentimes people attach additional assumptions to Newton's Third Law, which are not in general true. Most of the time, this additional assumption is a restriction on the types of objects that can form an action-reaction pair. For example, in the case of two moving charges, the people who say that Newton's Third Law is violated are implicitly assuming that the two charges are the action-reaction pair. This is not true; in reality, the charge and the electromagnetic field are the action-reaction pair. The only thing that Newton's Third Law guarantees is that for every object that is feeling a force, there exists some generalized abstract object that forms an action-reaction pair with it.