Nuclear Import of Arabidopsis Poly(ADP-Ribose) Polymerase 2 Is Mediated by Importin-α and a Nuclear Localization Sequence Located Between the Predicted SAP Domains

Abstract

Proteins of the Poly(ADP-Ribose) Polymerase (PARP) family modify target proteins by covalent attachment of ADP-ribose moieties onto amino acid side chains. In Arabidopsis, PARP proteins contribute to repair of DNA lesions and modulate plant responses to various abiotic and biotic stressors. Arabidopsis PARP1 and PARP2 are nuclear proteins and given that their molecular weights exceed the diffusion limit of nuclear pore complexes, an active import mechanism into the nucleus is likely. Here we use confocal microscopy of fluorescent protein-tagged Arabidopsis PARP2 and PARP2 deletion constructs in combination with site-directed mutagenesis to identify a nuclear localization sequence in PARP2 that is required for nuclear import. We report that in co-immunoprecipitation assays PARP2 interacts with several isoforms of the importin-α group of nuclear transport adapters and that PARP2 binding to IMPORTIN-α2 is mediated by the identified nuclear localization sequence. Our results demonstrate that PARP2 is a cargo protein of the canonical importin-α/β nuclear import pathway.

Keywords: Arabidopsis thaliana; PARP2; Poly(ADP-Ribose) Polymerase; importin-α; nuclear localization sequence; nucleo-cytoplasmic transport.

FIGURE 1

PARP2 localizes to the nucleus and interacts with several importin-α isoforms. (A) Representative (n = 10) confocal microscopy images of Nicotiana benthamiana cells expressing PARP2:RFP or free RFP. Left image RFP channel. Right image overlay with bright field image. Scale bar 30 μm. Images were taken 72 h after infiltration. (B) GFP-tagged versions of IMPORTIN-α1, - α2, - α3, - α4, - α6, - α9 or free YFP were co-expressed with PARP2:RFP or RFP:HaRxL106 (positive control) in N. benthamiana. At 72 h post infiltration, GFP-tagged proteins or YFP were immunoprecipitated and co-precipitating PARP2:RFP was detected by an α-RFP western blot. The polyclonal α-GFP antibody detects also YFP that we used as negative control.

FIGURE 2

PARP2 nuclear targeting sequences are located in the N-terminal SAP domains. (A) Predicted domain structure of PARP2 based on homology modeling and sequence conservation. SAP, SAF-A/B, Acinus and PIAS domains; WGR, domain with conserved Trp, Gly and Arg residues; PRD, protein regulatory domain; CAT, catalytic domain. (B) Representative (n = 10) confocal microscopy images of N. benthamiana cells expressing the indicated GFP fusion proteins or free YFP. Left image GFP or YFP channel. Right image overlay with bright field image. Scale bar 30 μm. Images were taken 72 h after infiltration.

FIGURE 3

PARP2 amino acids 48–51 are essential for nuclear targeting and mediate interaction with IMPORTIN-α2. (A) Representative (n = 10) confocal microscopy images of N. benthamiana cells expressing the indicated GFP/RFP fusion proteins. Left image GFP or RFP channel. Right image overlay with bright field image. Scale bar 30 μm. Images were taken 72 h after infiltration. ‘AAAA’ indicates mutation of PARP2 amino acids 48–51 to quadruple Ala. ‘QMQL’ indicates mutation of PARP2 amino acids 48–51 to Gln-Met-Gln-Leu. ‘K92/93A’ indicates mutation of PARP2 Lys 92 and 93 to Ala. (B) Distribution of the RFP signal from PARP2-AAAA:RFP around a nucleus (n = 4). (C) IMPORTIN-α2:GFP was co-expressed with free RFP or the indicated RFP-tagged variants of the PARP2 SAP domains in N. benthamiana. At 72 h post infiltration, IMPORTIN-α2:GFP was immunoprecipitated and co-precipitating RFP-tagged proteins were detected by an α-RFP western blot. (D) Co-immunoprecipitation experiment as in (C) but with the wildtype and mutated full-length PARP2 sequences instead of the SAP domains.

FIGURE 4

The PARP2 SAP domains and IMPORTIN-α2 do not form a stable protein complex in vitro. (A) Size exclusion chromatography elution profiles of His6:ΔIBB-IMPORTIN-α2, His6:SUMO:SAP and a mixture of both proteins (1:2 molar ratio) as determined by the absorption at 280 nm. (B) Coomassie-stained SDS-PAGE of the elution fractions from (A). The molecular weight of His6:ΔIBB-IMPORTIN-α2 is 53 kDa. The molecular weight of the His6:SUMO:SAP fusion is 25 kDa.

FIGURE 5

Multiple sequence alignment showing the partial conservation of the KRKR motif (PARP2 amino acids 48–51) in PARP2 homologs from different plant species. Only the SAP domains are shown. The sequence alignment was generated with Clustal Omega (Sievers and Higgins, 2014).