Nuclear translocation of proteins and the effect of phosphatidic acid

Abstract

Transport of proteins containing a nuclear localization signal (NLS) into the nucleus is mediated by nuclear transport receptors called importins, typically dimmers of a cargo-binding α-subunit and a β-subunit that mediates translocation through the nuclear pore complexes (NPCs). However, how proteins without canonical NLS move into the nucleus is not well understood. Recent results indicate that phospholipids, such as phosphatidic acid, play important roles in the intracellular translocation of proteins between the nucleus and cytoplasm.

Abbreviations

NLS

nuclear localization signal

NPCs

nuclear pore complexes

PA

phosphatidic acid

PtdIns(5)P

Phosphatidylinositol-5-bisphosphate

PLD

phospholipase D

The movement of proteins into the nucleus is a fundamental process critical to the basic function and regulation of eukaryotic cells. Many proteins that move into the nucleus contain a conserved short cluster of basic residues called nuclear localization sequence (NLS).^1-4 These proteins are translocated into the nucleus through nuclear pore complexes (NPCs) by receptor-mediated import pathways.⁵ However, some proteins that move into the nucleus do not possess the apparent canonical NLS and/or they are translocated into the nucleus only under specific cues, such as stress or hormonal stimulations. In addition, some cytosolic metabolic enzymes move into the nucleus where they have regulatory functions. Recent studies indicate that phospholipid mediators, such as phosphatidic acid (PA), play important roles in regulating the nuclear movement of proteins.

Receptor-mediated Pathways for Protein Import to the Nucleus

The classic receptor-mediated import pathways involve recognition by the importin (IMP) α/β heterodimer, where the adaptor IMPα subunit, which possesses NLS-binding site, binds the NLS, and the receptor IMPβ subunit targets the IMPα/NLS-bearing protein complex to the nuclear pore in eukaryotes.^1,6,7 This process also requires the cytoplasmic RanGDP to translocate the transport complex to NPC, and RanGTP to release IMPα/NLS-bearing protein into the nucleoplasm.^1,8,9 Alternatively, IMPβ itself or related homologs fulfills the function of both IMPα and β in binding NLSs, targeting them to the NPC.^10-13 In the model plant Arabidopsis thaliana, the IMPα recognizes the NLSs and also exhibits the nuclear envelope localization typical of IMPβ in other eukaryotic systems, mediates nuclear import pathway independent of IMPβ or IMPα/β.¹⁴

The prototype of a monopartite NLS is the sequence motif PKKKRKV (basic residues are underlined) of the simian virus 40 (SV40) large T-antigen, whereas the prototypical bipartite NLS is that of nucleoplasmin (KRPAATKKAGQAKKKKL).^15,16 NLSs of nuclear proteins are collected in the database of NLSdb, containing 114 experimentally determined NLSs obtained through an extensive literature search.¹⁷

Conditional Movement of Proteins Into the Nucleus

While some proteins are designated to go to the nucleus, some proteins are translocated into the nucleus only under specific conditions, such as environmental or hormonal cues. For example, phytochromes are red and far-red photoreceptors that regulate plant growth and development in response to light, and their nuclear movement upon light activation is regarded as a key regulatory step in phytochrome signaling.^18-20 Phytochromes exist in 2 photo-interconvertible conformational states: a biologically inactive Pr form and a biologically active Pfr form. Phytochrome-B (PhyB) is the prominent photo-stable phytochrome involved in red-light-sensing and shade-avoidance responses.²¹ PhyB contains NLS, but in the Pr form, the NLS in the C-terminal PAS-related domain (PRD) is masked through direct interactions with the phyB's N-terminal bilin lyase domain (BLD) and PHY subdomains of the N terminus. In the Pfr form, structural rearrangements in BLD and PHY lead to weaker interactions with the PRD, allowing the exposure of the NLS for translocation.²²

On the other hand, the photoreceptor phytochrome-A (phyA) does not contain a specific NLS. The nuclear import of light-dependent phyA depends on 2 specific plant proteins, namely FAR-RED ELONGATED HYPOCOTYL1 (FHY1) and FHY1-LIKE (FHL) in planta, and the import is mediated by the conformer- (Pfr>Pr) dependent interaction of these proteins.^23-25 PhyA regulates germination and seedling establishment by mediating very low fluence (VLFR) and far-red high irradiance (FR-HIR) responses in Arabidopsis. In darkness, phyA is a soluble cytosolic protein in the biologically inactive Pr form, but upon light, a portion of phyA forms into biologically active Pfr and then rapidly moves to nucleus.^20,26,27

The concomitant nuclear translocation is another pathway for proteins with or without canonical NLS for movement to the nucleus. Arabidopsis transcription factors, E2F and DP, are not predominantly localized to the nucleus, and the complete nuclear localization of some members is driven by the co-expression of their partner proteins. The nuclear localization of AtE2F1/3 and AtDPa depends on their interaction and occurs concomitantly in a cooperative manner.²⁸ The GFP fusion protein with single AtE2F1, AtE2F3, or AtDPa did not exhibit an exclusive nuclear localization except when AtE2F1 or AtE2F3 was co-expressed with AtDPa, indicating that neither E2F1/3 nor DPa has an autonomous activity for nuclear localization.²⁸ Even the addition of the SV40 NLS to AtE2F3 did not cause the nuclear localization, whereas it improved the efficiency of the nuclear localization when DPa was co-expressed.²⁸ The concomitant nuclear translocation conferred by the DPa interaction would be an effective dimerization partner of both AtE2F1 and AtE2F3.²⁸

Another case for concomitant nuclear translocation is the CCAAT-box binding nuclear factor Y (NF-Y) that consists of 3 different subunits, A, B, and C. All Arabidopsis NF-YBs lack NLS and their transport into the nucleus is proposed to be dependent on their interaction with NF-YC, while NF-YA and NF-YC are imported into the nucleus independently.²⁹ Interaction of the representatives of the NF-Y subunits, NF-YB10 and NF-YC9, in the cytoplasm is essential for the translocation of NY-FB into the nucleus and follows a piggyback mechanism.²⁹

Hexokinases (HXKs) function as glucose sensors and the catalytic and sensory functions of Arabidopsis HXK1 (AtHXK1) can be uncoupled.³⁰ The AtHXK1-GFP fusion protein was found to be in mitochondria, but a small amount of AtHXK1 is also associated with nuclei.^31,32 Nuclear AtHXK1 controls the expression of specific photosynthetic genes as a transcriptional co-repressor by creating a nuclear complex core with the vacuolar H⁺-ATPase B1 and the 19S regulatory particle of proteasome subunit.^31,33 Nuclear AtHXK1 also enhances the degradation of ETHYLENE-INSENSITIVE3 (EIN3), a key transcriptional regulator in ethylene signaling to mediate glucose signaling into or out of the nucleus and cross talk with ethylene signaling.³² In ethylene signaling, ETHYLENE-INSENSITIVE2 (EIN2) functions upstream EIN3. EIN2 is localized at the ER membrane.³⁴ The ethylene signal promotes the cleavage of the carboxyl end of EIN2 (CEND) from ER-located EIN2, and facilitates its nuclear localization to stabilize EIN3 protein.³⁵

Post-translational modifications of proteins under stress conditions also facilitate protein nuclear translocation. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) has been largely reported not only to be a ubiquitous enzyme involved in glycolysis, but also behaves as a moonlighting (namely nuclear localization and DNA binding) protein in both animal and plant cells. Redox modifications of GAPDH cysteine regulate its catalytic activity, and also confer its non-metabolic function. S-nitrosylated modifications of GAPDH have a profound effect on interaction with other proteins and eventually change its subcellular localization, which can transmit an apoptotic stimulus in animal cells.^36-39 The acetylation of 3 lysine residues was also proposed for the nuclear translocation of GAPDH after the apoptotic stimulus in animal cells.⁴⁰ Plant cytosolic GAPDHs, referred as GAPCs, can be translocated to the nucleus under particular, oxidizing conditions for multifunctional properties.^41-43 GAPC accumulates in nuclei upon stress in plants, and has a motif with some similarity to NLS at the N-terminus with 6 basic amino acids (MADKKIRIGINGFGRIGRLVAR), which overlaps with the NAD-binding domain (GXGXXG).⁴⁴ Their interaction between oxidized GAPC and PLD-δ did not show the role for oxidized GAPC's translocation. In Arabidopsis leaves, oxidized GAPC interacts with PLDδ and promote PA production and stomatal closure.⁴⁵ PA also binds to oxidized GAPC and promotes its proteolysis, and the GAPC2 and its cleavage product are present both in the cytosol and the nucleus.⁴¹ However, it is unclear whether the GAPC's NLS has the nuclear localization function.

Lipid Effects on Proteins Movement Into the Nucleus

In addition to protein-protein interaction and posttranslational modifications, lipid protein interactions are involved in mediating the intracellular translocation between the nucleus and cytoplasm. Phosphatidylinositol-5-bisphosphate [PtdIns(5)P] promotes protein translocation to the nucleus⁴⁶ and also from the nucleus to the cytoplasm.^47-49 For example, in Arabidopsis, PtdIns(5)P promotes translocation of an Arabidopsis trithorax homolog (ATX1), a histone 3 lysine 4 (H3K4me3) trimethyltransferase, from the nucleus to the cytoplasm.⁴⁷

PA is a lipid mediator involved in various cellular processes, and one of the key mechanisms of action by PA is its direct interaction with proteins. The PA binding can regulate the target protein activity and intracellular location. PA can promote or inhibit protein localization to the nucleus, depending on the specific protein. For example, PA in the endoplasmic reticulum (ER) binds to the transcriptional repressor Opi1p (OVERPRODUCER OF INOSITOL PROTEIN 1) and keeps it out of the nucleus in yeast⁵⁰ (Fig. 1). This PA tethering relieves the transcriptional suppression so membrane lipid synthesis will increase. The de novo lipid biosynthesis will consume PA, which releases Opi1p that enters the nucleus to suppress lipid production^50,51 In plants, PA tethers ABSCISIC ACID INSENSITIVE 1 (ABI1) to the plasma membrane, decreasing the translocation of ABI1 into the nucleus where it binds to ARABIDOPSIS THALIANA HOMEOBOX PROTEIN 6 (ATHB6), a transcription factor that negatively regulates ABA responses^52,53 (Fig. 1).

Figure 1.

Phospholipid effects on protein intracellular translocation between nucleus and cytoplasm reported in plant and yeast. In plant and yeast, phosphatidic acid (PA) can promote or inhibit protein localization to the nucleus, depending on the specific protein. In Arabidopsis, PA can directly interact with a MYB transcription factor WEREWOLF (WER), and that the interaction is necessary for WER's nuclear localization to regulate the root hair development. PA tethers its binding protein ABI1 to the plasma membrane, resulting in prevention of ABA INSENSTIVE 1 (ABI1) from translocation into the nucleus where it binds to ARABIDOPSIS THALIANA HOMEOBOX PROTEIN 6 (ATHB6), a transcription factor that negatively regulates ABA responses. In Saccharomyces cerevisiae, PA in the endoplasmic reticulum (ER) binds to the transcriptional repressor OVERPRODUCER OF INOSITOL PROTEIN 1 (Opi1p) and keeps it out of the nucleus to regulate phospholipid metabolism. Another phospholipid in Arabidopsis, phosphatidylinositol-5-bisphosphate [PtdIns(5)P] promotes translocation of an Arabidopsis trithorax homolog (ATX1), a histone 3 lysine 4 (H3K4me3) trimethyltransferase, from the nucleus to the cytoplasm to regulate gene expression.

PA can directly interact with the MYB transcription factor WEREWOLF (WER) that does not have a detectable nuclear localization signal based on its amino acid sequence.^54,55 The interaction is necessary for WER's nuclear localization⁵⁵ (Fig. 1). The PA interaction may tether WER to the nuclear membrane and this tethering is necessary for the localization of the proteins from the cytosol to the nucleus. These studies also propose that PA exists in nuclear membrane.

The different effects of PA are likely to be determined by the location and source of PA on the membranes, as well as other factors associated with the PA-binding proteins. In the PA-Opi1p interaction, PA on the ER is metabolized for de novo lipid biosynthesis^50,51, whereas in the PA-WER interaction, phospholipase D (PLD) contributes to the production of PA⁵⁵. The spatial and temporal production of PA by PLD may result in the different effects of PA on the nuclear localization of its target proteins, and PA produced by nucleus-associated PLD may regulate WER's nuclear localization.

The Arabidopsis genome has PLDs, PLDα(3), β(2), γ(3), δ, ε, ζ(2), with different reaction requirements.⁵⁶ The subcellular localization for some PLDs is still unknown. PLDγ was detected in the nuclear fraction from subcellular fractions of Arabidopsis leaves,⁵⁷ and PLDβ1 is predicted to localize in the nucleus (http://www.arabidopsis.org/). The nucleus-associated PLDs may contribute to the PA production on and in the nucleus. In addition, PA can be produced by other reactions, such as the activation of phospholipase C followed by diacylglycerol kinases. Further studies are needed to determine how PA mediates the nuclear localization of proteins and whether the PA-protein interactions constitute a mechanism for the nuclear translocation of proteins that lack NLS.

Funding Statement

The work was supported as part of the Center for Advanced Biofuels Systems (CABS), an Energy Frontier Research Center funded by the US. Department of Energy, Office of Science, Basic Energy Sciences under Award DE-SC0001295 and supported by the National Science Foundation (NSF) under Award IOS-0818740. HY acknowledges support by Chinese Scholarship Council.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

The authors thank Jennifer Potratz for critically reading the manuscript.