Human Enzyme PADI4 Binds to the Nuclear Carrier Importin α3

Abstract

PADI4 is a peptidyl-arginine deiminase (PADI) involved in the conversion of arginine to citrulline. PADI4 is present in macrophages, monocytes, granulocytes, and several cancer cells. It is the only PADI family member observed within both the nucleus and the cytoplasm. PADI4 has a predicted nuclear localization sequence (NLS) comprising residues Pro56 to Ser83, to allow for nuclear translocation. Recent predictors also suggest that the region Arg495 to Ile526 is a possible NLS. To understand how PADI4 is involved in cancer, we studied the ability of intact PADI4 to bind importin α3 (Impα3), a nuclear transport factor that plays tumor-promoting roles in several cancers, and its truncated species (ΔImpα3) without the importin-binding domain (IBB), by using fluorescence, circular dichroism (CD), and isothermal titration calorimetry (ITC). Furthermore, the binding of two peptides, encompassing the first and the second NLS regions, was also studied using the same methods and molecular docking simulations. PADI4 interacted with both importin species, with affinity constants of ~1–5 µM. The isolated peptides also interacted with both importins. The molecular simulations predict that the anchoring of both peptides takes place in the major binding site of Impα3 for the NLS of cargo proteins. These findings suggest that both NLS regions were essentially responsible for the binding of PADI4 to the two importin species. Our data are discussed within the framework of a cell mechanism of nuclear transport that is crucial in cancer.

Deciphering the Binding of the Nuclear Localization Sequence of Myc Protein to the Nuclear Carrier Importin α3

“…This may seem counterintuitive at first, as it could have been expected that they would rather flank the position of fragments F6–F8 on each respective side, roughly superimposing to the backbone of their crystallographic NLS counterparts. We have observed the same behavior in other similar cases [ 18 , 33 ], and raised the hypothesis that fragments adjacent to the core NLS binding motif having a good affinity for the same protein anchoring site could be convenient in the first steps of the recognition process or contribute to hampering the dissociation of the NLS when bound, thus favoring/disfavoring the kinetics of association/dissociation. Nevertheless, the region encompassing fragments F6–F8 is the sole expected to accommodate in that position at the end of the binding reaction, due to its slightly more favorable affinity for that protein pocket compared to the other fragments.…”

“…The main difference in the binding between the two importin species is due to the presence of the 60-residue-long IBB ( Figure 1 B), which is competing with the NLS for the NLS-anchoring region. Such a variation between the affinities of either Impα3 or ΔImpα3 for the same NLS was not as large as the one we detected for other NLSs explored [ 18 , 19 , 20 , 33 ]; furthermore, the values of the affinities for NLS-Myc of either Impα3 or ΔImpα3 were much more favorable (i.e., had a smaller dissociation constant) than those measured for other NLSs [ 18 , 19 , 20 , 33 ]. Thus, it seems that the sequence of NLS-Myc is much more optimized for binding to Impα3/ΔImpα3 than those of the other proteins we tested.…”

“…This slower dissociation rate ( k off ) is the reason of the higher affinity (smaller K d ) of the NLS-Myc peptide towards ΔImpα3. We have not observed such small dissociation rates in other NLSs from other proteins for which the kinetic rates have been measured [ 33 ], which suggests some difference in the release mechanism (e.g., the presence of a metastable state, or a kinetic barrier). Our docking simulations showed that (i) the residues flanking the core NLS of Myc have a slightly lower affinity for the same anchoring location of the core region itself; and (ii) a secondary binding site for the NLS of Myc may also exist for Impα3/ΔImpα3 in close proximity to the main binding site, although its lower affinity with respect to the latter suggests it should not be significantly populated under ordinary conditions.…”

“…Human c-Myc is an IDP with a largely disordered structure ( Figure 1 A) [ 35 ]. Previous studies on other NLSs of IDPs [ 18 , 19 , 20 ] or, alternatively, of folded proteins [ 15 , 16 , 33 , 40 ] have suggested an inhibitory action of the IBB of Impα3: this domain hampers the anchoring of the NLS of the corresponding cargo protein into the major NLS-binding region of Impα3 [ 12 , 41 ]. In this respect, from BLI measurements, we have observed an affinity one-order of magnitude higher in the binding of NLS-Myc towards ΔImpα3 compared to Impα3, in agreement with the findings obtained for other NLSs assayed before [ 18 , 19 , 20 , 33 ].…”

“…Previous studies on other NLSs of IDPs [ 18 , 19 , 20 ] or, alternatively, of folded proteins [ 15 , 16 , 33 , 40 ] have suggested an inhibitory action of the IBB of Impα3: this domain hampers the anchoring of the NLS of the corresponding cargo protein into the major NLS-binding region of Impα3 [ 12 , 41 ]. In this respect, from BLI measurements, we have observed an affinity one-order of magnitude higher in the binding of NLS-Myc towards ΔImpα3 compared to Impα3, in agreement with the findings obtained for other NLSs assayed before [ 18 , 19 , 20 , 33 ]. The main difference in the binding between the two importin species is due to the presence of the 60-residue-long IBB ( Figure 1 B), which is competing with the NLS for the NLS-anchoring region.…”