ZmCTLP1 is required for the maintenance of lipid homeostasis and the basal endosperm transfer layer in maize kernels

Abstract

Maize kernel weight is influenced by the unloading of nutrients from the maternal placenta and their passage through the transfer tissue of the basal endosperm transfer layer (BETL) and the basal intermediate zone (BIZ) to the upper part of the endosperm. Here, we show that Small kernel 10 (Smk10) encodes a choline transporter-like protein 1 (ZmCTLP1) that facilitates choline uptake and is located in the trans-Golgi network (TGN). Its loss of function results in reduced choline content, leading to smaller kernels with a lower starch content. Mutation of ZmCTLP1 disrupts membrane lipid homeostasis and the normal development of wall in-growths. Expression levels of Mn1 and ZmSWEET4c, two kernel filling-related genes, are downregulated in the smk10, which is likely to be one of the major causes of incompletely differentiated transfer cells. Mutation of ZmCTLP1 also reduces the number of plasmodesmata (PD) in transfer cells, indicating that the smk10 mutant is impaired in PD formation. Intriguingly, we also observed premature cell death in the BETL and BIZ of the smk10 mutant. Together, our results suggest that ZmCTLP1-mediated choline transport affects kernel development, highlighting its important role in lipid homeostasis, wall in-growth formation and PD development in transfer cells.

Keywords: ZmCTLP1; choline transporter; kernel development; maize; plasmodesmata.

Fig. 1

Phenotypic features of the maize smk10 mutant. (a) Mature smk10 × B73 F2 ear (wild‐type (WT) and smk10 kernels segregate in a 3 : 1 ratio). Black arrows indicate smk10 kernels. Bar, 2 cm. The magnified sections are indicated by dotted boxes. Bar, 2 mm. (b, c) Mature wild‐type and smk10 kernels. Mature WT and smk10 kernels (b) and their longitudinal sections (c). Bars: (b) 1 cm (c) 0.2 cm. (d) Leaf phenotypes of 10‐d‐old WT and smk10 seedlings. Bar, 3 cm. (e) Root phenotypes of 2‐d‐old WT and smk10 seedlings. Bar, 1 cm. (f, g) Longitudinal sections of developing WT (f) and smk10 (g) kernels at 4–12 d after pollination (DAP). Bars: (f, g) 1 mm.

Fig. 2

Longitudinal sections of wild‐type (WT) and smk10 kernels in maize. (a, b) Comparison of the developing CZ in WT (a) and smk10 (b) kernels at 12 d after pollination (DAP). CZ, conducting zone. Bars, 100 μm. (c, d) Comparison of the developing aleurone layer (AL) in WT (c) and smk10 (d) kernels at 12 DAP. The AL is indicated by a red arrow. EN, endosperm; PE, pericarp. Bars, 20 μm. (e, f) Comparison of the developing basal endosperm transfer layer (BETL) and basal intermediate zone (BIZ) in WT and smk10 kernels at 12 DAP. Black arrows indicate wall in‐growths (WIG) of reticulate and flange BETL cells and BIZ cells in the WT (e). Wall in‐growths (WIG) were stunted in smk10 (f). PC, placento‐chalazal region. Bars, 50 μm. (g) The lengths of developing BIZ cells in wild‐type and smk10 longitudinal sections. Values are means ± SE, n = 200 cells (**, P < 0.01, Student’s t‐test).

Fig. 3

Positional cloning and identification of maize smk10 mutant. (a) The smk10 locus was mapped to a 300‐kb region on chromosome 2 using a high‐resolution map consisting of 3264 F3 individuals derived from a cross between smk10 and Mo17. The red triangle denotes the site of a mutation in the predicted intron‐exon splicing site. (b) Zm00001d001803 was targeted with a specific gRNA located in the first exon. Two independent events with different fragment deletions were generated with the CRISPR/Cas9 system. (c) Neighbour‐joining phylogenetic tree using Arabidopsis thaliana paralogues of AtCTL1 and maize paralogues of ZmCTLP1 (with four human and mouse protein sequences as an outgroup). Red and green triangles indicate ZmCTLP1 and AtCTL1, respectively. (d) pH‐dependent choline uptake mediated by ZmCTLP1 in Xenopus oocytes. The means ± SE from five independent experiments are presented (**, P < 0.01, Student’s t‐test). (e) Subcellular localisation of ZmCTLP1 in maize protoplasts. Bar, 10 μm. (f) Expression pattern of ZmCTLP1 in various tissues measured by qRT‐PCR. Actin was used as the endogenous control gene. Three biological replicates of each tissue type were analysed, and the values are reported as means ± SE.

Fig. 4

Changes in choline and lipid homeostasis accompany abnormal maize kernel development in the smk10 mutant at 12 d after pollination (DAP). (a) Choline contents in wild‐type (WT) and smk10 kernels. The means ± SE of four replicates per genotype are presented (**, P < 0.01, Student’s t‐test). (b) Total lipid contents in WT and smk10 kernels. (c) Twenty‐four classes of membrane lipids were measured in WT and smk10 kernels. (d) The absolute contents of 423 lipid species from 24 classes were measured. The y‐axis shows the P‐values associated with t‐tests between the WT and smk10. (e) The relative contents of 24 classes of membrane lipids in WT and smk10 kernels. (f) The relative contents of 423 lipid species from 24 classes were measured. The y‐axis shows the P‐values associated with t‐tests between the WT and smk10. Cer, ceramides; CL, cardiolipins; DAG, diacylglycerols; DGDG, digalactosyl diacylglycerols; FFA, free fatty acids; GluCer, glucosylceramides; LPA, lyso‐PA; LPC, lyso‐PC; LPE, lyso‐PE; LPS, lyso‐PS; MGDG, monogalactosyl diacylglycerols; PA, phosphatidic acids; PC, phosphatidylcholines; PE, phosphatidylethanolamines; PG, phosphatidylglycerols; PhytoCer, phytoceramides; PhytoCer‐OHFA, phytoceramides with hydroxylated fatty acyls; Phyto‐GluCer, phyto‐glucosylceramides; PhytoSph, phytosphingosines; PI, phosphatidylinositols; PS, phosphatidylserines; S1P, sphingosine‐1‐phosphate; Sph, sphingosines; TAG, triacylglycerols. In (b, c, e, f) the means ± SE of five replicates per genotype are presented (**, P < 0.01, *, P < 0.05; Student’s t‐test).

Fig. 5

ZmCTLP1 is essential for plasmodesmata (PD) development in maize. (a–d) Aniline blue staining of callose shows the numbers and locations of PD in 12 d after pollination (DAP) wild‐type (WT) (a) and smk10 (c) kernels. (b, d) Magnified sections of (a) and (c) indicated by dotted boxes. PD is indicated by a red arrow. Bars, 50 μm. (e, f) Longitudinal sections of the basal endosperm transfer layer (BETL) and basal intermediate zone (BIZ) from 12 DAP WT (e) and smk10 (f) kernels. Bars, 20 μm. (g, h) The magnified sections of (e) and (f) indicated by dotted boxes. Bars, 2 μm. (i, j) The magnified sections of (g) and (h) indicated by dotted boxes. Bars, 500 nm. (k, l) The magnified sections of (i) and (j) indicated by dotted boxes. PD structure and numbers are shown. Red arrows indicate PD. Bars, 100 nm. (m) Number of PD in 10 independent longitudinal sections of BIZ cells from 12 DAP WT and smk10 kernels. The number of PDs differed significantly between the WT and smk10. The means ± SE are presented (**, P < 0.01; Student’s t‐test).

Fig. 6

Gene Ontology analysis of the DEGs between 6 and 12 d after pollination (DAP) wild‐type (WT) and smk10 maize kernels based on RNA‐seq data. (a, b) The most significantly enriched GO terms in the DEGs of 6 DAP kernels (a) and 12 DAP kernels (b) and their associated P‐values are shown. Lower x‐axis, −log₁₀ (P‐value); upper x‐axis, the number of genes with a given GO term. (c) Expression of basal endosperm transfer layer (BETL) and basal intermediate zone (BIZ)‐specific genes quantified by RNA‐seq analysis. BIZ‐specific genes were obtained from Li et al. (2014). The log₂ (fold‐change) values between smk10 and the WT were calculated from RNA‐seq data and are shown as a heat map. (d) Tissue‐specific expression patterns of transport‐related genes associated with kernel development quantified by RNA‐seq analysis. PC (purple), placento‐chalazal region; PED (green), pedicel. The log₂ (fold‐change) values between smk10 and the WT were calculated from RNA‐seq data and are shown as a heat map (right). (e) Immunoblot analysis showing Mn1 protein accumulation in 4, 8, and 12 DAP WT and smk10 kernels. α‐Tubulin served as the loading control. (f–h) In situ hybridization of ZmSWEET4c using antisense probes in sections of smk10 and WT kernels at 12 DAP. (f, g) ZmSWEET4c‐antisense probes show the expression of ZmSWEET4c in the BETL of smk10 and wild‐type kernels. (h) The ZmSWEET4c‐sense probe served as the negative control. The BETL is indicated by a black arrow. Bars, 200 μm.

Fig. 7

Cell death in wild‐type (WT) and smk10 maize kernels analysed by the TUNEL assay. Green fluorescent dots indicate the TUNEL signals of fragmented DNA, and red fluorescent dots indicate nuclei stained with PI. (a, b) Cell death detected by the TUNEL assay in 8 d after pollination (DAP) WT (a) and smk10 (b) kernels. (c, d) Cell death detected by the TUNEL assay in 12 DAP wild‐type (c) and smk10 (d) kernels. BETL, basal endosperm transfer layer; BIZ, basal intermediate zone; PC, placento‐chalazal region. Bars, 50 μm.

Fig. 8

Working model for the control of maize kernel development by ZmCTLP1. (a) In wild‐type kernels, ZmCTLP1 maintains lipid homeostasis, normal plasmodesmata (PD) numbers, Mn1 and ZmSWEET4c expression and the function of wall in‐growths (WIG) in transfer cells through choline transport. (b) In smk10 kernels, altered choline uptake, transport and/or metabolism affect lipid homeostasis and reduce the number of PDs. The basal endosperm transfer cells undergo premature cell death. DET, differentially expressed transporter; SV, secretory vesicle.