The XVP/ NAC003 protein associates with the plasma membrane through KR rich regions and translocates to the nucleus by changing phosphorylation status

ABSTRACT

Membrane localized transcription factors play essential roles in various plant developmental processes. The XVP/NAC003 protein is a NAC domain transcription factor associated with the plasma membrane and involved in the TDIF-PXY signaling during vascular development. We report here the mechanisms of XVP membrane localization and its nuclear translocation. Using a transient transformation approach, we found that XVP is associated with the plasma membrane through positively charged KR-rich regions. Mutagenesis studies found that the threonine amino acid at position 354 (T354) is critical for XVP translocation to the nucleus. In particular, the threonine to alanine mutation (T354A) resulted in a partial nucleus localization, while threonine to aspartic acid (T354D) mutation showed no effect on protein localization, indicating that dephosphorylation at T354 may serve as a nucleus translocation signal. This research sheds new light on the nucleus partitioning of plasma membrane-associated transcription factors.

Introduction

The proliferation of stem cells in vascular meristem is responsible for secondary growth, which produces secondary phloem and secondary xylem and contributes to the thickening of plant stem and root.¹ The maintenance of vascular cambium is regulated by a signaling pathway including a CLE peptide Tracheary Element Differentiation Inhibitory Factor (TDIF), its receptor TDIF Receptor/Phloem intercalated with Xylem (TDR/PXY), and a downstream transcription factor WOX4.^2–5 TDIF is encoded by the CLAVATA3/EMBRYO SURROUNDING REGION-RELATED 41 (CLE41) and CLE44 genes in the phloem and travels toward the cambium, where it binds the plasma membrane-localized receptor PXY.^3,4,6,7 The binding of TDIF to PXY promotes the expression of WOX4, as a result enhancing cell divisions in the cambium, repressing xylem cell specification, and regulating vascular patterning. The SOMATIC EMBRYO GENESIS RECEPTOR KINASEs (SERKs), including BRI1 associated Kinase 1 (BAK1)/SERK3, are PXY co-receptors of the TDIF peptide.⁸

We previously reported that XVP/NAC003 functions as a negative regulator of TDIF signaling and fine-tunes the function of the TDIF-PXY module in vascular cambium.⁹ Overexpression of XVP represses the functions of the TDIF peptide in cambium proliferation, vascular patterning, and xylem differentiation.^5,9 In the xvp nac048 double mutant, expression of WOX4 is enhanced with or without TDIF treatment, supporting XVP as a negative regulator of TDIF signaling.⁹ Further expression analyses showed that XVP promotes the expression of CLE44, the encoding gene of TDIF, while overexpression of either CLE44 or CLE41 represses XVP expression.⁹ Therefore, XVP is also involved in feedback regulation in TDIF signaling.^5,9

The XVP protein is localized on the plasma membrane and directly interacts with the TDIF co-receptor complex PXY-BAK1.⁹ As a transcription factor, XVP has a functional nuclear localization sequence and shows transcription activity.⁹ Membrane-bound NAC domain transcription factors are tethered to the plasma transmembrane through their transmembrane domains and function in different biological pathways.^10–12 Intriguingly, the XVP protein does not contain a transmembrane domain. In a previous study, we have found that the C-terminal domain (CTD) is responsible for XVP plasma membrane localization, although the mechanism is unknown.⁹ In this report, we found that the KR-rich regions in the CTD are critical for XVP plasma membrane association, and a threonine amino acid at position 354 is important for its nucleus translocation. Thus, this study identifies a new mechanism of membrane localization and nuclear translocation for NAC domain transcription factors.

Materials and methods

Plant materials and growth condition

The Nicotiana benthamiana plants were grown to 6–8 leaves stage for transient transformation experiments. Plants were grown in growth chambers with the following settings: photoperiod 16 h light and 8 h dark cycle; temperature, 22°C day: 20°C night; relative humidity 70–80%; and light intensity 150 μmol m⁻² s⁻¹.

Constructs

To overexpress XVP, the full-length coding sequence of the XVP was cloned using a high-fidelity polymerase and inserted into the pENTR TOPO-D vector. Primer sequences for cloning XVP are XVP Fw - CACCATGGAAACTCCTGTGGGTTTAAGA and XVP Re (TCAAGTTCTTGAGATGGAAGAACATAGGTG). The XVP gene insertion in the pENTR vector was confirmed by Sanger sequencing, and subcloned to the destination vector pGWB506 by LR reaction. Primers used to make truncations can be found in Supplementary Table 1. To generate mutations in the XVP sequence, an overlapping PCR strategy was used with the mutagenesis primers specified in Supplementary Table 1. All mutations were confirmed through Sanger sequencing before being subcloned to the expression vectors.

Transient transformation

The method for infiltration of N. benthamiana leaves has been described previously.⁹ In brief, Agrobacterium cultures were incubated in Luria–Bertani (LB) medium with appropriate antibiotics overnight and centrifuged at 1800 g for 10 min to collect bacterial cells. Them, pellets were resuspended in infiltration buffer, mixed with p19 suppressor culture, and infiltrated into N. benthamiana leaves with a 1-ml syringe.

Microscopic observation

Leaf samples were collected three days after infiltration by cutting next to the infiltration sites, placed on glass slides, mounted in ddH₂O, and covered with a cover glass. The samples were observed using a Nikon A1R Spectral Confocal (Nikon) with consistent settings. At least five leaf samples per construct were observed to determine a consensus localization pattern in the cell. Representative images were collected and used in this paper.

Results and discussion

Minimal C-terminal domain (CTD) sequence required for XVP membrane localization

In a previous study,⁹ we found that XVP/NAC003 is associated with the plasma membrane through its CTD. To investigate the minimal sequence required for XVP plasma membrane localization, we performed protein truncation and transient transformation analyses. The primers used for the truncation constructs and mutagenesis can be found in the material and method section. The fusion of an 84 amino acids CTD of XVP from position 311 to 394 (CTD311-394) with the GFP fusion resulted in an exclusive membrane-localized fluorescence signal (Figure 1a-c). Shortening of the CTD to 62 amino acids (CTD 333–394) showed the exact membrane localization as the construct with 84 amino acids (Figure 1d). However, further shortening either from the N-terminal (CTD351-394) or from the C-terminal (CTD333-387) resulted in changing localization (Figure 1e,f). The fusion proteins of these two constructs resulted in the formation of a punctuated fluorescent structure around the cell plasma membrane (Figure 1e,f). The nature of the punctuated structures is likely membrane-associated, and the membrane localization of XVP is significantly affected. These experiments indicated that the minimal sequence necessary for XVP plasma membrane localization is a 62 amino acids fragment at the C terminus.

Figure 1.

The minimal CTD sequence for XVP/NAC003 membrane localization. (a) The structure of XVP/NAC003. The numbers denote the beginning positions of each construct. (b) Diagrams showing constructs used for localization analysis. (c-f) Localization of different GFP-CTD constructs of 311–394 (c), 333–394 (d), 351–394 (e), and 333–387 (f). For each construct, representative GFP channel (left), bright light (middle), and over-lap (right) images were shown in this figure. Bar = 50 μm

Localization of XVP on plasma membrane relies on KR-rich domains

In the identified minimal CTD sequence of XVP, no transmembrane domain can be detected using multiple prediction tools, such as TMHMM (http://www.cbs.dtu.dk/services/TMHMM/) and TMpred (embnet.vital-it.ch/software/TMPRED_form.html). It has been reported that lysine and arginine (KR) rich regions may serve as a membrane hook, possibly due to the positive charge of these amino acids.^13,14 By analyzing the CTD domain of XVP, we can detect three regions with multiple K or KR amino acids (Figure 2a). To investigate if these three K/KR-rich regions affect plasma membrane localization, we performed mutagenesis and transient transformation experiments (Figure 2). Mutation of either lysine (K) to alanine (A) at positions 383 and 384 (Figure 2b) or modifications of the five Ks at three additional positions (343–345) showed no obvious effect on XVP plasma membrane localization (Figure 2c). These results indicate that these two K-rich regions play a limited role in XVP localization. In contrast, mutations of the KKKR sequence to AAAA at positions 367–370 formed punctuated structures in the cell, indicating that this region is essential for plasma membrane localization (Figure 2d). Combing the mutations at all three KR-rich regions of all nine amino acids (NAC3 9KR/AA) caused severe mislocalization of XVP protein and formed large aggregation in the cell (Figure 2e), indicating that the two K-rich regions contribute to proper XVP plasma membrane localization.

Figure 2.

The KR-rich domains are important for XVP/NAC003 membrane localization. (a) Protein sequences of XVP and its homologs. Critical positions for truncation or mutation were labeled. (b) Mutation of lysine (k) to alanine (A) at positions 383 and 384 showed no obvious effect on XVP localization (c) Mutations of five Ks at three additional positions (343–345) still showed no apparent effect on XVP localization. (d) Mutations of KKKR at positions 367–370 to AAAA affected XVP localization by forming punctuated structures in the cell. (e) Mutations of K/R at all nine positions caused severe mislocalization of XVP protein. Bar = 50 μm

A T354A mutation resulted in translocation of XVP to the nucleus

The XVP protein is a transcription factor and localized on the plasma membrane. Further, XVP interacts with the receptor kinase complex PXY-BAK1. Therefore, we reasoned that XVP could potentially be modified by the PXY receptor kinase, therefore translocating to the nucleus. Using a functional analysis tool for post-translational modification,¹⁵ we identified a threonine amino acid at position 91 (T91) as a potential target for phosphorylation. To investigate if T91 is essential for nuclear translocation, we performed a protein mutagenesis assay. The results indicated that mutation of T91 to either glutamic acid (T91E) or alanine (T91A) showed no effect on XVP localization (Figure 3a,b).

Figure 3.

The threonine to alanine mutation at position 354 (T354A) translocates XVP partially to the nucleus. (a and b) Mutation of threonine at position 91 to either glutamic acid (a) or alanine (b) showed no effect on XVP localization. (c and d) Mutation of threonine at position 354 to aspartic acid (T354D) showed no effect on XVP localization (c), but alanine (T354A) resulted in nucleus localization (d). For each construct, representative GFP channel (left), bright light (middle), and over-lap (right) images were shown in this figure. Bar = 50 μm

We further reasoned that KR-rich regions at CTD are critical for XVP plasma membrane localization. Either phosphorylation or dephosphorylation of T/S/Y amino acids around the KR- rich regions may affect its localization as previously reported.^13,16 Analysis of XVP and its homologs indicated that one threonine at position 354 (T354) is conserved among the closely related NAC domain TFs and could be a potential target for protein modification (Figure 2a). To investigate if T354 is necessary for XVP localization/translocation, we mutated this threonine to either aspartic acid (T354D) or alanine (T354A). The T354D mutation showed no effect on XVP localization (Figure 3c). Surprisingly, the T354A mutation resulted in nucleus translocation (Figure 3d). These experiments indicated that T354 is a potential dephosphorylation site and is critical for XVP nucleus translocation. The T354 is outside of the KKKR motif, which may associate with the membrane due to positive charge. It is possible that dephosphorylation at T354 may affect the configuration of the XVP protein, thus concealing the KKKR motif. It is also possible that T354 is part of the interacting interface with the plasma membrane.

This report identified a 61 amino acid fragment at the CTD as the minimal sequence responsible for XVP membrane localization. It is interesting to note that truncation of the last few amino acids from 388 to 394 also affects XVP plasma membrane localization. However, this sequence is not conserved among the closely related NAC domain TFs (Figures 1f and 2a), and the mechanism is currently unknown. Three K/KR-rich domains are identified in the minimal CTD region required for XVP plasma membrane localization. Among the three motifs, the KKKR domain is the single most crucial domain, while the other two K-rich domains contribute to the proper plasma membrane localization. Mutagenesis analysis indicated that dephosphorylation might play an role in XVP protein partitioning between the plasma membrane and the nucleus (Figure 3d). The mechanism of XVP plasma membrane localization is different from the NAC domain TFs with transmembrane motifs^10,11 or lipid anchored through S-palmitoylation.¹⁷ Proteins associated with the plasma membrane through KR-rich domains are likely regulated by phosphorylation or dephosphorylation.^13,16 It would be intriguing to identify the developmental or environmental cues for XVP protein modification and other components involved in the signaling pathway.

Funding Statement

This research was supported by National Science Foundation (IOS-1453048), and in part, by the USDA National Institute of Food and Agriculture, Hatch project no. CONS00990; and by grants to HW from the UConn Research Excellence Award.

Author contributions

KL and HW designed the experiments; KL, GC, SY, QD, LQ and HW performed the experiments; KL, and HW analyzed the data; and HW wrote the article.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed on the publisher’s website