Modified pea apyrase has altered nuclear functions and enhances the growth of yeast and Arabidopsis

Abstract

Apyrases (NTPDases) regulate growth and development in multiple eukaryotic organisms and function in multiple sub-cellular locales. An earlier report showed that the ectopic expression of psNTP9 (PS), a chromatin-associated pea (Pisum sativum) apyrase, enhanced the uptake of inorganic phosphate (Pi) and increased the growth of yeast and Arabidopsis. In this follow-up study, we generated a modified form of PS, abbreviated DM ("double mutant"), in which two-point mutations, S208L and P216R, were introduced into its DNA-binding domain. Ectopic expression of DM increased the growth of yeast and Arabidopsis, the seed yield of Arabidopsis, and the Pi content of yeast and Arabidopsis grown in Murashige-Skoog media beyond that effected by PS. Both the PS and DM proteins co-purified with nuclei and chromatin-associated proteins from yeast and Arabidopsis, and expression of their transgenes in these model organisms produced gene expression profiles that would be expected to promote increased growth and Pi uptake. Chromatin immunoprecipitation (ChIP)-seq analyses showed that PS and DM have largely different binding sites on yeast chromatin, including sites in promoters of numerous genes that are differentially-expressed in PS and DM transgenic lines. These results are consistent with the hypothesis that the effects of ectopically expressing the pea apyrase in yeast and in Arabidopsis are mediated, at least in part, by its activities in the nucleus that impact transcription.

Keywords: DNA-binding; apyrase; calmodulin; phosphate; point mutation; seed yield.

Figure 1

PS and DM transcript and protein levels, and growth of yeast pho84 mutant NS219 constitutively expressing PS and DM. (A) Left panel, relative transcript abundance in different colonies (C) of empty-vector (EV)-expressing pYES2, PS-expressing pYES2, and DMexpressing PYES2. The qRT-PCR samples were from 72 h induced pho84 yeast cells, under the same induction condition as in panels B and C, with four biological repeats. Relative transcript levels were normalized using the expression levels of UBC6 for qRT-PCR. P value ≤ 0.01. Right panel, top row, immunoblot indicating relative staining of PS and DM by monoclonal antibody 8B6 in extracts of the same colonies as in left panel; bottom row, relative protein staining by Coomassie Brilliant Blue (CBB) of proteins extracted from the same colonies as in left panel. (B) Growth curves for the mutant transformed with the pYES2 vector alone, with pYES2:PS, pYES2:DM and pYES2:PHO84 (positive control). Growth was in low-Pi medium (100 µM). (C) Growth curves for the mutant transformed with the pYES2 vector alone, pYES2:PS, or with pYES2:PS-NLS. Values represent the means ± SE of three independent transformants, each with three technical repeats. An asterisk indicates that the mean value was significantly different (*p < 0.05, **p < 0.01, ***p < 0.001) from that of the pYES2 vector control and # indicates mean value of pYES2:DM was significantly different (p ≤ 0.05) from that of the pYES2:PS when analyzed by one-way ANOVA (Tukey’s Multiple Comparison Test).

Figure 2

Structure models of apyrases psNTP9 and psNTP9-DM. (A) Hinged structure of the enzyme, connecting the N-terminal Domain I (green) and C-terminal Domain II (blue) is shown for psNTP9. (B) Conserved regions ACR1-5, as reported for the Trifolium apyrase, line the catalytic cleft of this enzyme (ACR1-yellow, ACR2-green, ACR3-light blue, ACR4-pink, ACR5-orange). Catalytic residues in ACR1-5 (red), identified by Summers et al. (2017) are labeled using psNTP9 numbering (see Supplementary Figure S1 ). CaM binding domain 2 (PCBD2 (D); putative CaM binding domain 1 (PCBD1 (E); and PCBD1-DM (F), which is mutated (S208R, P216R) in psNTP9-DM (C) are shown, rotated 90° to the right from orientations within the enzyme structures. Communication between PCBD1, located at the hinge position in the enzyme, and N-terminal ACR4 and catalytic residue D195, which descends into the catalytic cleft, are illustrated by extended structures for PCBD1 in these enzymes. Helical wheel representation and sequence for residues in CaM-binding PCBD2 (G) illustrates the amphipathic character of this a-helix (dotted line). C-terminal sequences comprising the a-helix of PCBD1 in psNTP9 (E) and psNTP9-DM (F) are colored bright green; their helical wheels and corresponding sequences are shown below (H, I).

Figure 3

PS and DM overexpression significantly enhances the phosphate (Pi) content of yeast. Total Pi contents of yeast cells transformed with empty vector (EV), PS or DM. Yeast cells were grown for 72 h on 0.1 mM Pi, as in Figure 1 , and transferred to 1 mM Pi medium for 15 min, then their Pi contents were measured. Data are means ± S.E., n = 4 biological replicates. Letters indicate significant differences (p < 0.05), as determined by one-way analysis of variance (ANOVA) with post-hoc Tukey honest significant difference (HSD) testing.

Figure 4

PS and DM overexpression significantly enhances the phosphate (Pi) content and seed yield of Arabidopsis seedlings. (A) Total Pi contents of 7-d-old Arabidopsis PS2 and DM4 transgenic lines, compared with WT Col-0 seedlings. Data are for means ± SE for two independent experiments, each with 3-5 biological replicates (n=10-20 seedlings). Significant differences (*p = 0.09, **p<0.05) were calculated using the Student’s t-test. (B) Comparison of seed weight (g) per plant of wild type (WT), empty vector (EV), PS- and DM-overexpressing Arabidopsis plants grown in greenhouses. Different letters above the bars indicate that the mean value was significantly different (p ¾ 0.05) from that of other samples when analyzed by twoway ANOVA (Bonferroni post-tests; n ≥ 16).

Figure 5

Immunoblot analysis of total extracts and of nuclei purified from etiolated Col-0 (WT) seedlings, and from etiolated Col-0 seedlings expressing PS (PS2), DM (DM4), or modified versions of PS missing either their N-terminal signal peptide (PS-N), or their nuclear localization sequence (PS-NLS). The positive control was highly purified pea apyrase. The immunostaining used apyrase-specific antibody (Anti-8B6) or Anti-histone3 antibody. Coomassie-Brilliant Blue (CBB) staining was used for the loading control. The 8b6 signal relative intensities, with Pea apyrase set as 1.0, are given beneath the stained bands.

Figure 6

Immunoblots showing the presence of PS and DM in yeast nuclei and in yeast extracts. (A) nuclei purified from the NS219 mutant transformed with either an empty vector (EV), with PS, or with DM. (B) yeast extracts at different stages of the purification of chromatin-associated proteins from yeast transformed with either an empty vector, with PS, or with DM. These stages included whole cell extract, chromatin supernatant 1, chromatin supernatant 2, the final chromatin pellet, and the supernatant of this pellet (Supernatant 3). Histone 3 antibody was used as a positive control for nuclear expression, and exoribonuclease Xrn1 antibody was used as a negative control for cytoplasmic expression.

Figure 7

ChIP-seq identification of PS and DM binding sites in a region of yeast chromosome I. Mapped single-end reads (green,+ strand; red, - strand) coverage along the reference genome is shown. (A) Peak calling in the uninduced and induced PS samples, based on the distribution of unique reads mapped to the reference genome (normalized genome coverage) in test and input control samples. Peaks in the uninduced test samples represent non-specifically immunoprecipitated DNA fragments. Overlapping peaks in uninduced and induced test samples are removed and putative PS binding sites are retained. Gene annotation shows the locations of PS binding sites, relative to genes. PS peak “A” is located in a region of divergent promoters for upstream gene LSD1 (antisense strand) and downstream gene PSK1 (sense strand). PS peak “B” is located with the PSK1 gene and is > 1,000 bp upstream from the TSS for LSD1. It would not be classified as a potential regulatory site for LSD1. (B) Identification of DM peaks in the same region of chromosome I, using the approach described in part A. Mapped reads tracks are not shown. DM peak “C” is located within gene LSD1, > 1,000 bo from the TSS of downstream gene PSK1. In contrast, DM peak “D”, like PS peak “A”, lies within divergent promoters for LSD1 and PSK1 and partially overlaps with the putative PS “A” binding site. High-confidence (p ≤ 0.05) final peak assignments for PS and DM are shown in tracks. Peaks of varying lengths are represented by brown arrows. Mapped single-end reads are green (+ strand) or red (- strand).

Figure 8

Venn analysis of potentially-regulated target genes with PS-specific, DM-specific, or PS+DM shared binding sites in their promoters. The large number of genes in the PS+DM overlap group with DM is due to the selection of DM as the reference (see Figure 7C ). Numbers of DEG included in PS-specific or DM-specific target cohorts. Total numbers of DEG are indicated in parentheses.

Figure 9

Discovery of a Yap3p-like motif in PS-specific binding sites in promoters of phosphate metabolism genes induced in PS yeast. (A) Locations of PS and DM binding sites and potentially-regulated target genes. PS-specific binding sites in promoters are enclosed in red boxes. (B) Yap3p-like motifs within PS binding sites and Logo representation of their sequences, compared with the discovered STREME-1 motif and Yap3p motif consensus sequences.