Edited by Ivan SadowskiKeywords: Gpn1 Gpn3 Gpn1-Gpn3 interaction Gpn1-Gpn3 nucleocytoplasmic shuttling Interdependent protein levels shRNA a b s t r a c t Gpn1 and Gpn3 are GTPases individually required for nuclear targeting of RNA polymerase II. Here we show that whereas Gpn3-EYFP distributed between the cytoplasm and cell nucleus, it was mainly cytoplasmic when coexpressed with Gpn1-Flag. Gpn3-Flag retained Gpn1-EYFP in the cytoplasm. However, Gpn3-EYFP/Gpn1-Flag nucleocytoplasmic shuttling was revealed after inhibiting nuclear export with leptomycin B. All Gpn3-EYFP coimmunoprecipitated with Gpn1-Flag, and all Gpn1-EYFP with Gpn3-Flag. Importantly, most endogenous Gpn1 and Gpn3 also associate. Gpn1-Gpn3 interaction was essential to maintain steady-state protein levels of both GTPases. We propose that most Gpn1 and Gpn3 associate, are mobilized, and function as a protein complex. Structured summary of protein interactions:GPN3 physically interacts with GPN1 by anti tag coimmunoprecipitation (1, 2) GPN3 and GPN1 colocalize by fluorescence microscopy (View interaction)

XAB1/Gpn1 is a GTPase that associates with RNA polymerase II (RNAPII) in a GTP-dependent manner. Although XAB1/Gpn1 is essential for nuclear accumulation of RNAPII, the underlying mechanism is not known. A XAB1/Gpn1-EYFP fluorescent protein, like endogenous XAB1/Gpn1, localized to the cytoplasm but it rapidly accumulated in the cell nucleus in the presence of leptomycin B, a chemical inhibitor of the nuclear transport receptor Crm1. Crm1 recognizes short peptides in substrate proteins called nuclear export sequences (NES). Here, we employed site-directed mutagenesis and fluorescence microscopy to assess the functionality of all six putative NESs in XAB1/Gpn1. Mutating five of the six putative NESs did not alter the cytoplasmic localization of XAB1/Gpn1-EYFP. However, a V302A/L304A double mutant XAB1/Gpn1-EYFP protein was clearly accumulated in the cell nucleus, indicating the disruption of a functional NES. This functional XAB1/Gpn1 NES displays all features present in most common and potent NESs, including, in addition to Φ1-Φ4, a critical fifth hydrophobic amino acid Φ0. Therefore, in human Gpn1 this NES spans amino acids 292-LERLRKDMGSVAL-304. XAB1/Gpn1 NES is remarkably conserved during evolution. XAB1/Gpn1 NES was sufficient for nuclear export activity, as it caused a complete exclusion of EYFP from the cell nucleus. Molecular modeling of XAB1/Gpn1 provided a mechanistic reason for NES selection, as functionality correlated with accessibility, and it also suggested a mechanism for NES inhibition by intramolecular masking. In conclusion, we have identified a highly active, evolutionarily conserved NES in XAB1/Gpn1 that is critical for nucleo-cytoplasmic shuttling and steady-state cytoplasmic localization of XAB1/Gpn1.

Gpn3 is required for RNA polymerase II (RNAPII) nuclear targeting. Here, we investigated the effect of a cancer-associated Q279* nonsense mutation in Gpn3 cellular function. Employing RNAi, we replaced endogenous Gpn3 by wt or Q279* RNAi-resistant Gpn3R in epithelial model cells. RNAPII nuclear accumulation and transcriptional activity were markedly decreased in cells expressing only Gpn3R Q279*. Wild-type Gpn3R localized to the cytoplasm but a fraction of Gpn3R Q279* entered the cell nucleus and inhibited Gpn1-EYFP nuclear export. This property and the transcriptional deficit in Gpn3R Q279*-expressing cells required a PDZ-binding motif generated by the Q279* mutation. We conclude that an acquired PDZ-binding motif in Gpn3 Q279* caused Gpn3 nuclear entry, and inhibited Gpn1 nuclear export and Gpn3-mediated RNAPII nuclear targeting.