[en] A literature compilation of nuclear star cluster (NSC) masses is used to study the correlation between global and NSC properties. A comparison of observational data to the predictions of semi-analytical galaxy formation models places constraints on the co-evolution of NSCs, massive black holes (MBHs), and host galaxies. Both data and theoretical predictions show an increased scatter in the NSC scaling correlations at high galaxy masses, and we show that this is due to the progressively more efficient ejection of stars from NSCs caused by MBH binaries in more massive stellar spheroids. Our results provide a natural explanation of why in nucleated galaxies hosting an MBH, the ratio $(M_{\mathrm{N}\mathrm{S}\mathrm{C}}+M_{\mathrm{M}\mathrm{B}\mathrm{H}})/M_{\mathrm{b}\mathrm{u}\mathrm{l}\mathrm{g}\mathrm{e}}$ (with $M_{\mathrm{b}\mathrm{u}\mathrm{l}\mathrm{g}\mathrm{e}}$ as the host spheroid’s mass) shows significantly less scatter than $M_{\mathrm{N}\mathrm{S}\mathrm{C}}/M_{\mathrm{b}\mathrm{u}\mathrm{l}\mathrm{g}\mathrm{e}}$, and suggest that the formation of MBHs and NSCs are not mutually exclusive, as is also supported by observations of co-existing systems. Both MBHs and NSCs represent generic products of galaxy formation, with NSCs being destroyed or modified by the merger evolution of their companion MBHs.