Purine is a remarkable substance, given Nature has chosen it as the scaffold for two nucleobases from DNA/RNA: adenine (A) and guanine (G). Its structure also appears in several other substances of biological interest, by substitution on the fused ring system, like these examples shown in Wikipedia:
The base ring system has 4 nitrogen atoms and 5 carbon atoms. Then I wonder, it's not hard to imagine isomers or at least similar compounds just by changing the positions of the nitrogen atoms around the rings, with a few constraints. The five-membered ring needs at least one N to maintain aromaticity in the pyrrole ring. It's better to have this N at the position where a glycosidic bond is expected, if we want to incorporate it in some kind of nucleic acid. Along the edge connecting the two fused cycles, it's better to have C, to avoid N atoms with four bonds. So we are left with the following template:
We can distribute the remaining 3 N and 3 C atoms among the 6 free positions in $6!/((6-3)! \cdot 3!) = 20$ different ways.
So my question is, what's special in purine, that caused its scaffold to be picked among all the other possibilities?