Necrotic upper tips1 mimics heat and drought stress and encodes a protoxylem-specific transcription factor in maize

Abstract

Maintaining sufficient water transport during flowering is essential for proper organ growth, fertilization, and yield. Water deficits that coincide with flowering result in leaf wilting, necrosis, tassel browning, and sterility, a stress condition known as "tassel blasting." We identified a mutant, necrotic upper tips1 (nut1), that mimics tassel blasting and drought stress and reveals the genetic mechanisms underlying these processes. The nut1 phenotype is evident only after the floral transition, and the mutants have difficulty moving water as shown by dye uptake and movement assays. These defects are correlated with reduced protoxylem vessel thickness that indirectly affects metaxylem cell wall integrity and function in the mutant. nut1 is caused by an Ac transposon insertion into the coding region of a unique NAC transcription factor within the VND clade of Arabidopsis NUT1 localizes to the developing protoxylem of root, stem, and leaf sheath, but not metaxylem, and its expression is induced by flowering. NUT1 downstream target genes function in cell wall biosynthesis, apoptosis, and maintenance of xylem cell wall thickness and strength. These results show that maintaining protoxylem vessel integrity during periods of high water movement requires the expression of specialized, dynamically regulated transcription factors within the vasculature.

Keywords: flowering; maize; protoxylem; vasculature; water transport.

Fig. 1.

The phenotype of the nut1 mutant. (A) The nut1-m1 mutant (Right) shows wilted necrotic leaf tips and a blasted tassel compared to wild type (WT, Left). (B) Comparison of uppermost leaves enclosing the tassel of WT versus a second allele nut1-1234 mutant. Leaves on the Right were separated from each apex shown on the Left. Dye uptake and movement assay using sheaths (C) and floral stems (D). Red arrows indicate blockages in nut1. (D) Transverse floral stem sections taken at the same distance (1 cm) above the dye level. Dye uptake in nut1 vascular bundles is reduced in stems of both alleles, indicating defective water transport. (E) nut1 ear length and silk elongation are greatly reduced. (Scale bar, 5 cm in A, B, and E and 2 mm in C and D.)

Fig. 2.

Characterization of the nut1 gene and expression. (A) nut1 gene model and alleles. nut1-m1 is caused by an Ac insertion into the third exon of a NAC transcription factor. The nut1-s1234 allele has an 8-bp insertion causing a premature stop codon. A Nut1′ wild-type revertant restoring the reading frame was isolated from nut1-m1, proving that the mutant phenotype is caused by the insertion. Red arrows indicate the location of primers for qRT-PCR. Blue dash lines correspond to the region of the NAC domain. Blank rectangles represent the UTRs, blue rectangles represent the exons, and lines represent introns. (B) Tissue specificity of nut1 expression as indicated by qRT-PCR of 6-wk-old plants. (C) Western blot using a NUT1 antibody showing protein accumulation in wild-type stems, but not in either allele. Bottom is ACTIN loading control. (D) Quantification of nut1 expression by qRT-PCR of 1- to 5-wk-old roots (mixed samples containing primary, seminal, and crown roots; purple bars), dissected shoot apices (pink bars), and aerial nodes (n) and internodes (i) from entire 11-wk flowering plants (turquoise bars) starting from the base (node 10) to the uppermost node below the tassel (node 1). Nodes 6n and 7n normally produce ears in this genetic background, and floral induction normally occurs in node 8. R, root. (E and F) RWC measurement in leaves (E) and sheaths (F) of 11-wk-old and 4-wk-old wild type (turquoise lines) and mutant (orange lines). Numbers on x axis refer to leaf number counting from tassel #1 (E) or apex (F) going toward the ground. The asterisks indicate statistical significance by t test at P < 0.05. The error bars in B, D, and F are SD from three replicates in B and D and five replicates in E and F.

Fig. 3.

NUT1 accumulates in protoxylem precursor cells. (A) NUT1 immunolocalization of transverse sections of wild-type stem (A) and sheath (B) showing protein accumulation in early initiating protoxylem cells (red arrow). Control probing of nut1-m1 mutant is Inset in A and shows no expression. All tissues were from 6-wk-old florally induced plants. (C) NUT1 immunolocalization on longitudinal sections of shoot apex. (D) Magnification of the region framed by box in C, showing nuclear localization in protoxylem provascular strand (red arrow). (E) NUT1 immunolocalization on transverse sections of primary root tip. nut1 mutant root is Inset. (F) Magnification of the region framed by box in E, showing nuclear localization in root protoxylem (red arrows) differentiating in a series outwards from metaxylem. (G and H) Schematic of progression of protoxylem initiation during sheath development. After the initial protoxylem undergoes cell death the adjacent distal cell differentiates into the next one. Pa, parenchyma; Bs, bundle sheath; Tc, tracheid cells; Ph, phloem; Pp, primary protoxylem; Sp, secondary protoxylem. (I) Sheath vascular bundle showing NUT1 accumulation in future protoxylem (red arrows). (J and K) After the primary protoxylem cell undergoes cell death and has formed a secondary cell wall (gray arrows), the adjacent NUT1-expressing cell becomes the next protoxylem. (K and L). At later stages, NUT1 was not detected in initiating metaxylem cells (black arrowheads). (Scale bar, 50 µm in A–F and 25 µm in I–L.) P, protoxylem; M, metaxylem.

Fig. 4.

Genome-wide overview of the regulatory network downstream of nut1. (A) Genome-wide distribution of NUT1 binding peaks revealed by DAP-seq. (B) Distribution of NUT1 binding peaks relative to gene models, showing strong enrichment within 1 kb upstream of TSSs. (C) Enriched motifs within the NUT1 binding peaks. The TTGCTT motif shows up as the most enriched core sequence, and the second most enriched is a 17-nt motif consisting of two partially reverse complements with high similarity to the core sequence. E-value was calculated by MEME. (D) Localization of the motifs relative to the NUT1 peak summits. (E) Overlap of all NUT1 target genes and DEGs in the nut1 mutant. P < 0.0001 by hypergeometric test. F is the same as E with up- and down-regulated DEGs separated. (G and H) Functional categories of down- (G) and up-regulated (H) DEGs using agriGO v2 (39). Greater than 23% of all cell wall metabolism genes are DEGs. Bubble size indicates the number of differentially expressed gene counts in the corresponding GO category, and x axis indicates the relative ratio to all genes within that GO category.

Fig. 5.

NUT1 targets genes involved in cellulose biosynthesis and apoptosis. (A) Heat maps showing that genes involved in cellulose biosynthesis and programmed cell death were differentially expressed in nut1. Gene names highlighted in red were also bound by NUT1 via DAP-seq. The gradient color scale indicates the log₂ value of expression fold change (blue is down- and red is up-regulated). (B) NUT1 binds preferentially to the 3′ end of the top two cellulose synthase genes (cesa12 and cesa6) that also show reduced expression in nut1. (C) NUT1 binds to the 5′ end and third intron, respectively, of the top two programmed cell death genes (xcp1 and dcd) that show reduced expression in the mutant. (D and E) ChIP qPCR validating NUT1 binding to the genic regions of cesa12 (D) and xcp1 (E). ChIP was conducted using NUT1 antibody, and IgG was used as negative control. The positions of the p1 and p2 primers used for ChIP-qPCR are indicated by red arrows in B and C. The error bars represent SD from three biological replicates.

Fig. 6.

Reduced nut1 protoxylem vessel thickness caused by defective secondary cell walls. (A) Transverse section of a wild-type and nut1 sheath vascular bundle by cryofracture SEM. (Scale bar, 20 µm.) (B) Quantitative comparison of secondary cell wall thickness of protoxylem. Red lines indicate positions where measurements were taken for each replicate. A total of 78 to 93 vascular bundles from at least seven individual plants were measured. Different letters indicate statistically significant differences at P < 0.05 by Tukey’s test. (C) Confocal imaging of plastic section of vascular bundles from second apical sheath of wild-type and nut1 flowering plants. Protoxylem cellulose was stained using Direct Red 23. (Scale bar, 50 µm.) (D) Quantitative comparison of the cellulose content of protoxylem as determined by normalized fluorescence intensity in C. For each replicate, 80 and 97 measurements were taken from four wild-type and nut1 mutant plants, respectively. Different letters indicate statistically significant differences at P < 0.05 by Tukey’s test. (E) Confocal imaging of xylan immunolocalization using the LM10 monoclonal antibody. Comparison of plastic sections of second apical sheath tissue from WT and nut1-m1 flowering plants. Dark panels show fluorescence intensity, while light panels are the overlay with brightfield image. (Scale bar, 20 µm.) (F) Quantitative comparison of xylan content of wild-type and nut1 protoxylem as determined by the normalized fluorescence intensity in E. For each replicate, measurements were taken on 98 individual protoxylem cells from three wild-type and nut1 mutant plants, respectively. Statistically significant differences were detected at P < 0.05 by Tukey’s test. (G) Cleared sheath tissue stained with basic fuchsin for wild type (Left) and nut1 (Right) in areas where dye movement is blocked (as indicated by arrows in Fig. 1C). Lignified protoxylem (p) annular thickenings are visible in between both metaxylem vessels. Vessel junction (white arrow) was associated with blocked vessel transport. (Scale bar, 50 µm.) (H) Quantitative comparison of annular thickening density between wild-type and nut1 protoxylem. All of the fluorescence images used for pairwise comparison were taken under identical microscope settings. (I) Transverse serial sections of W22 (Left) and nut1 node#1 sheath tissue subjected to dye movement assay similar to the experiment shown in Fig. 1C. A nut1 vascular bundle with limited dye movement was fixed, sectioned, and stained with Toluidine Blue below, near and above the dye stoppage point, respectively. The presence of transverse cell walls (red arrows) in the metaxylem near the dye stoppage points indicate the presence of perforation plates at vessel junctions, while the metaxylem above them appear collapsed with altered cell wall morphology (red asterisk). Wild-type W22 node#1 sheath with normal dye movement was used as a control. (Scale bar, 25 µm.)