The columns of $U$ corresponding to nonzero singular values (the left singular vectors) span the column space of $A$, and they are orthonormal, so they form an orthonormal basis for the column space. Similarly, the columns of $V$ corresponding to nonzero singular values (the right singular vectors) span the row space of $A$ (equivalently, the column space of $A^T$), giving an orthonormal basis for the row space.