Molecular Background of Pi Deficiency-Induced Root Hair Growth in Brassica carinata - A Fasciclin-Like Arabinogalactan Protein Is Involved

Abstract

Formation of longer root hairs under limiting phosphate (P) conditions can increase the inorganic P (Pi) uptake. Here, regulatory candidate genes for Pi deficiency-induced root hair growth were identified by comparison of massive analysis of cDNA ends (MACE) provided expression profiles of two Brassica carinata cultivars (cv.) differing in their root hair response to Pi deficiency: cv. Bale develops longer root hairs under Pi deficiency, but not cv. Bacho. A split-root experiment was conducted for the differentiation between locally and systemically regulated genes. Furthermore, plants were exposed to nitrogen and potassium deficiency to identify P-specific reacting genes. The latter were knocked out by CRISPR/Cas9 and the effect on the root hair length was determined. About 500 genes were differentially expressed under Pi deficiency in cv. Bale, while these genes did not respond to the low P supply in cv. Bacho. Thirty-three candidate genes with a potential regulatory role were selected and the transcriptional regulation of 30 genes was confirmed by quantitative PCR. Only five candidate genes seemed to be either exclusively regulated locally (two) or systemically (three), whereas 25 genes seemed to be involved in both local and systemic signaling pathways. Potassium deficiency affected neither the root hair length nor the expression of the 30 candidate genes. By contrast, both P and nitrogen deficiency increased the root hair length, and both affected the transcript levels in 26 cases. However, four genes reacted specifically to Pi starvation. These genes and, additionally, INORGANIC PHOSPHATE TRANSPORTER 1 (BcPHT1) were targeted by CRISPR/Cas9. However, even if the transcript levels of five of these genes were clearly decreased, FASCICLIN-LIKE ARABINOGALACTAN PROTEIN 1 (BcFLA1) was the only gene whose downregulation reduced the root hair length in transgenic hairy roots under Pi-deficient conditions. To the best of our knowledge, this is the first study describing a fasciclin-like arabinogalactan protein with a predicted role in the Pi deficiency-induced root hair elongation.

Keywords: BcFLA1; CRISPR/Cas9; K deficiency; MACE; N deficiency; hairy root; local and systemic response; split-root.

FIGURE 1

Shoot dry matter (dm) (A), shoot P concentration (B), AtIPS2 expression (C), root hair length (D) and representative root hairs (E, scale = 0.5 mm) of B. carinata cultivar (cv.) Bale and cv. Bacho as affected by P supply in the 1st experiment (MACE experiment). Small letters denote significant differences at P < 0.05 between cv.s at the same P-level; capital letters between P treatments of the respective cv. (Tukey test). Columns represent means and bars indicate SE; n = 4 (A,B) and 8 (D) biological replications, whereas 3 technical replications of pooled samples were realized for (C).

FIGURE 2

Relative number of normalized tags assigned to the selected candidate genes for increased root hair length under Pi deficiency (A–F) in B. carinata cv. Bale and cv. Bacho affected by P supply determined by MACE.

FIGURE 3

Relative expression of selected candidate genes for increased root hair length under Pi deficiency (A–E) in B. carinata cv. Bale and cv. Bacho affected by P supply determined by qPCR in samples obtained from the MACE (1st) experiment (exp.) and the independently conducted 2nd exp. AtUBC9 was used as an endogenous control. Significance was calculated according to Steibel et al. (2009). Small letters denote significant differences at P < 0.05 between cv.s at the same P-level; capital letters between P treatments of the respective cv. Columns represent means; n = 3.

FIGURE 4

Shoot (A) and root (B) dm of B. carinata cv. Bale affected by P supply in a split-root system. Different letters denote significant differences at P < 0.05 (Tukey test). Columns represent means and bars indicate SE; n = 3.

FIGURE 5

P_total concentration in shoot (A) and root (B) as well as P_inorganic concentration in shoot (C) and root (D) of B. carinata cv. Bale affected by P supply in a split-root system. Different letters denote significant differences at P < 0.05 (Tukey test). Columns represent means and bars indicate SE; n = 3.

FIGURE 6

Root hair length of B. carinata cv. Bale affected by P supply in a split-root system. Different letters denote significant differences at P < 0.05 (Tukey test). Columns represent means and bars indicate SE; n = 30.

FIGURE 7

Relative expression of candidate genes for increased root hair length during Pi deficiency (A–E) in B. carinata cv. Bale affected by P supply in a split-root system. AtUBC9 was used as an endogenous control. Significance was calculated according to Steibel et al. (2009). Different letters denote significant differences at P < 0.05. Columns represent means; n = 3.

FIGURE 8

Shoot dm (A) and root hair length (B) affected by nutrient deficiency. Significant differences at P < 0.05 are denoted by small letters between cv.s for the same nutrient treatment and capital letters between nutrient treatments for the same cv. (Tukey test); columns represent means and bars indicate SE; n = 6 (A) and n = 12 (B), respectively.

FIGURE 9

Relative expression of candidate genes for increased root hair length during Pi deficiency (A–E) in B. carinata cv. Bale and cv. Bacho affected by nutrient deficiency determined by qPCR. AtUBC9 and BcEF-1-a1 were used as endogenous controls. Significance was calculated according to Steibel et al. (2009). Significant differences at P < 0.05 are denoted by small letters between cv.s for the same nutrient treatment and capital letters between nutrient treatments for the same cv. Columns represent means; n = 3.

FIGURE 10

Relative candidate gene expression (A) and relative root hair length of the transgenic hairy roots (B) after targeting the respective candidate genes for increased root hair length under Pi deficiency by CRISPR/Cas9. Horizontal lines represent the average wildtype expression and root hair length, respectively. Stars denote significant differences, determined according to Steibel et al. (2009) (A) and Tukey test (B) (significance codes: 0 ‘^∗∗∗’ 0.001 ‘^∗∗’ 0.01 ‘^∗’ 0.05). Columns represent means for n = 4–8 (A) and 5–14 (B), respectively. Bars in (B) indicate SE.

FIGURE 11

CRISPR/Cas9-induced deletions and insertions in BcFLA1. gRNA regions (gray), PAM sequence (blue), number of bp inserted (red) and number of bp deleted (on the right). Numbers on the left indicate the respective transgenic roots, whereby ‘L’ indicates a long fragment (small deletion), ‘M’ a medium fragment and ‘S’ a small fragment (large deletion). Nucleotides used for differentiation of BcFLA1b from BcFLA1a (violet, underlined). CR = CRISPR, /../ = break.

FIGURE 12

Validation of the CRISPR/Cas9-induced gene-editing of BcFLA1 by fragment analysis by PAGE. A, C, G and T = Sanger sequencing reactions with the corresponding dideoxynucleotides; c, control; b, blank (no template control). Lane numbers indicate the respective transgenic roots. Numbers close to the bands indicate the bp deleted (–) and inserted (+) compared to the wildtype size (0). Brackets indicate indels with approximated size, which did not occur during sequencing.

FIGURE 13

Relative expression of candidate genes for Pi deficiency-induced root hair growth in bcfla1 roots grown under Pi-deficient conditions. AtUBC9 was used as an endogenous control. Significance was calculated according to Steibel et al. (2009) (significance codes: 0 ‘^∗∗∗’ 0.001 ‘^∗∗’ 0.01 ‘^∗’ 0.05). Columns represent means; n = 3.

FIGURE 14

Model of the potential position and working mechanism of BcFLA1.