Skip to main content
NCBI home page
Plant Physiology logoLink to Plant Physiology
. 2022 Feb 16;189(1):301–314. doi: 10.1093/plphys/kiac050

Transcriptomic and functional analysis provides molecular insights into multicellular trichome development

Mingming Dong 1, Shudan Xue 2, Ezra S Bartholomew 3, Xuling Zhai 4, Lei Sun 5, Shuo Xu 6, Yaqi Zhang 7, Shuai Yin 8, Wenyue Ma 9, Shuying Chen 10, Zhongxuan Feng 11, Chao Geng 12, Xiangdong Li 13, Xingwang Liu 14,15,16,†, Huazhong Ren 17,18,19,✉,†
PMCID: PMC9070826  PMID: 35171294

Abstract

Trichomes, the hair-like structures located on aerial parts of most vascular plants, are associated with a wide array of biological processes and affect the economic value of certain species. The processes involved in unicellular trichome formation have been well-studied in Arabidopsis (Arabidopsis thaliana). However, our understanding of the morphological changes and the underlying molecular processes involved in multicellular trichome development is limited. Here, we studied the dynamic developmental processes involved in glandular and nonglandular multicellular trichome formation in cucumber (Cucumis sativus L.) and divided these processes into five sequential stages. To gain insights into the underlying mechanisms of multicellular trichome formation, we performed a time-course transcriptome analysis using RNA-sequencing analysis. A total of 711 multicellular trichome-related genes were screened and a model for multicellular trichome formation was developed. The transcriptome and co-expression datasets were validated by reverse transcription-quantitative PCR and in situ hybridization. In addition, virus-induced gene silencing analysis revealed that CsHOMEOBOX3 (CsHOX3) and CsbHLH1 are involved in nonglandular trichome elongation and glandular trichome formation, respectively, which corresponds with the transcriptome data. This study presents a transcriptome atlas that provides insights into the molecular processes involved in multicellular trichome formation in cucumber and can be an important resource for future functional studies.

Transcriptomic and functional analysis of cucumber (Cucumis sativus L.) cotyledon trichomes provides a framework for understanding the regulatory network of multicellular trichome formation.

Introduction

Trichomes are the hair-like epidermal protrusions that cover most aerial plant tissues. Morphologically, trichomes can be divided into unicellular and multicellular structures based on their cell number. Trichomes can also be classified as secretory glandular and nonglandular based on their ability to biosynthesize, store, and secrete specialized metabolites. In many species, trichomes play important roles in plant adaptation and resilience to biotic and abiotic stresses, such as insect and pathogen deterrence and UV-B radiation protection (Karabourniotis et al., 1992; Mauricio and Rausher, 1997; Kennedy, 2003). Trichomes also influence the economic value of some plant species. Sweet wormwood (Artemisia annua) contains glandular secreting trichomes that biosynthesize artemisinin, an economically important and well-established drug used for the treatment of malaria (Matias-Hernandez et al., 2017). Additionally, the cucumber (Cucumis sativus L.) fruit is covered by large nonglandular trichomes (called spines) that influence its marketability and commercial value, making it a key trait in cucumber production (Liu et al., 2016).

The nonglandular and unicellular trichomes found in the model plant Arabidopsis (Arabidopsis thaliana) have been extensively studied, with >40 genes known to regulate various trichome developmental processes (Schellmann and Hulskamp, 2005; Pattanaik et al., 2014). Arabidopsis trichome development is divided into six sequential stages: initiation (Stage 1), tubular formation (Stage 2), branching points formation (Stage 3), extension and orientation of the branches (Stages 4 and 5), and maturation (Stage 6) (Szymanski et al., 1998; Mathur et al., 1999). Although majority of terrestrial plants contain multicellular trichomes, our understanding of the regulatory networks of multicellular trichome development remains largely unknown and the intact developmental stages have not been clearly defined. In tomato (Solanum lycopersicum), Slwoolly (Slwo), encoding a homeodomain-leucine zipper (HD-Zip) IV transcriptional factor (TF), regulates the initiation of multicellular trichomes. Several Slwo putative orthologs, namely Nbwoolly (Nbwo), GhHD1, CmGLABROUS (CmGL), and CsGLABROUS3 (CsGL3), also control trichome initiation in Nicotiana benthamiana, cotton (Gossypium hirsutum), melon (Cucumis melo), and cucumber, respectively (Yang et al., 2011; Walford et al., 2012; Pan et al., 2015; Cui et al., 2016; Zhu et al., 2018). However, PROTODERMAL FACTOR2, a Slwo putative ortholog in Arabidopsis, did not function in unicellular trichomes initiation but was found to regulate shoot epidermal cell differentiation (Abe et al., 2003). Several transcription factors, AaMIXTA1, AaHD1, and AaGLANDULAR TRICHOME-SPECIFIC WRKY2, are known to initiate the formation of glandular trichome in sweet wormwood (Yan et al., 2017, 2018; Xie et al., 2021). Additionally, SlMYC1, GoPIGMENT GLAND FORMATION (GoPGF), and AaTRICHOME AND ARTEMISININ REGULATOR1 (AaTAR1) have been shown to regulate glands formation in tomato, cotton (Gossypium spp), and sweet wormwood, respectively (Tan et al., 2015; Ma et al., 2016; Xu et al., 2018; Hua et al., 2020; Li et al., 2020). These studies suggest that the mechanisms regulating multicellular trichome development are similar in several plant species but differ from unicellular trichomes in Arabidopsis.

With the advancement of next-generation sequencing technology, transcriptome analysis has become a powerful strategy to explore various life processes. Recently, several transcriptome profiling studies have discovered genes expressed in mature multicellular trichomes in tobacco (Nicotiana tabacum) (Yang et al., 2015; Nautiyal et al., 2020), medicinal cannabis (Cannabis sativa L.) (Braich et al., 2019), mint (Mentha spp.) (Akhtar et al., 2017), sweet wormwood (Soetaert et al., 2013), and cucumber (Chen et al., 2014; Zhao et al., 2015). These studies utilized samples from young leaves/fruits with well-developed trichomes, which does not capture the entire developmental process of multicellular trichomes. Therefore, stage-specific transcriptome data, especially at the early developing stages of multicellular trichomes, are needed.

In this study, we reported a comprehensive time-coursed transcriptomic analysis that captures the complete and continuous developmental process of both glandular and nonglandular multicellular trichomes in cucumber. This process was divided into five sequential stages and gave insights into multicellular trichome formation. The results establish a potential model of multicellular trichome formation and provide valuable molecular information for further studies on multicellular trichome development.

Results

Stages of multicellular trichomes development in cucumber

To investigate genes involved in multicellular trichome development, it is important to first identify the intact and consecutive trichome developmental process. Initial observations of the aerial parts of cucumber plants indicate that trichomes were already fully developed, even on tissues barely visible to the naked eyes. Furthermore, trichomes developed at various time intervals on different tissues, which led to difficulties in identifying continuous developmental stages (Supplemental Figure S1). However, using cotyledons as a model may be advantageous, since trichomes on cotyledons develop synchronously. Therefore, we sampled the cotyledons every 6 h after seed soaking (HAS) (Figure 1A), with the entire process occurring within 114 HAS (Figure 1, B1–B18).

Figure 1.

Figure 1

Open in a new tab

Developmental series of cucumber cotyledonal trichomes from 12 to 114 HAS. A, Cucumber seedling from 12 to 114 HAS. Numbers below the plants indicated the HAS. The images were digitally extracted for comparison. Bar = 1 cm. B1–B18, SEM images of cucumber cotyledonal adaxial surface from 12 to 114 HAS. Numbers in the lower left corner indicated the HAS. Bar = 35 um.

On cucumber cotyledons, we observed two types of multicellular trichomes, in common with trichomes on other tissues: nonglandular trichomes with sharply pointed tips found on the adaxial surface, and glandular trichomes with glandular heads distributed mainly on the margins and veins (Supplemental Figure S2). Based on morphological observations, multicellular trichome development can be divided into five stages: (I) initiation, (II) first division, (III) tip head formation/glandular head transition, (IV) elongation/glandular head formation, and (V) base formation/active metabolism. In Stage I (18–36 HAS), multicellular trichomes were initiated by the protuberating and enlarging of the epidermal surface (Figure 2, A–C). During Stage II (36–60 HAS), trichomes underwent a periclinal bipartition to two cells (Figure 2, C–E), which is further validated by propidium iodide staining (Supplemental Figure S3). In these early stages, we observed no visible difference between glandular and nonglandular trichomes. However, in Stage III (60–72 HAS), the distal cells of nonglandular trichomes changed to form a sharp pointed tip (Figure 2, E and F), while those of glandular trichomes continued to expand (Figure 2, E and J). In Stage IV (72–96 HAS), the nonglandular trichomes were noticeably elongated with fewer cell divisions (Figure 2, F–H), while a four-celled glandular head formed in glandular trichomes (Figure 2, J–L). In Stage V (96–114 HAS), the base of nonglandular trichomes underwent multiple rounds of cell division to form a bulged base (Figure 2, H and I), while no distinct morphological change was observed in glandular trichomes. However, the head of glandular trichomes appeared darker and wrinkled, suggesting possible secretion of specialized metabolites (Figure 2, L and M).

Figure 2.

Figure 2

Open in a new tab

Stage division of multicellular trichome development. SEM images of developing nonglandular (A–I) and glandular (A–E and J–M) multicellular trichomes. Numbers in the lower left corner indicated the HAS. Bar = 10 um.

The generation of multicellular trichomes time-course transcriptome data

To gain insights into multicellular trichome development, we generated time-coursed transcriptome data by sampling the proximal ends of cotyledons at 18, 36, 60, 72, and 96 HAS (Supplemental Figure S4, A–E). To further narrow down the range of MTRGs, we also sampled the shoot apical meristem of RIL-46 W (CsGL3) and RIL-46 M (csgl3) (Supplemental Figure S4, F and G), because: (1) csgl3 shows a fully glabrous phenotype without other developmental deformities (Pan et al., 2015); (2) trichomes at different developmental stages were observed on the shoot apical meristem; and (3) the lack of trichomes on csgl3 can affect the expression of MTRGs.

A total of 92 million high-quality reads was generated and mapped to cucumber 9930 reference genome using HISAT (Kim et al., 2015). An average of 91% reads were uniquely mapped (Supplemental Table S1) and used to calculate normalized gene expression level as fragments per kilobase of transcript per million mapped reads (FPKM). A comparison of the biological replicates showed that the expression values were highly associated. Therefore, the average FPKM value was used for expression analysis. Based on the adopted cutoff parameter (fold change ≥2 and false discovery rate [FDR] < 0.05), we identified 12,746 differentially expressed genes (DEGs) in time-coursed transcriptome data using DESeq version 2 (Love et al., 2014). Furthermore, 711 genes were differentially expressed between RIL-46 W and RIL-46 M and were annotated as MTRGs (Supplemental Table S2).

Co-expression modules of MTRGs

To better uncover the underlying mechanism of multicellular trichome formation, we clustered the 711 MTRGs into 16 co-expression modules according to their expression levels using the k-means clustering algorithm (Figure 3 and Supplemental Table S3). From the results, genes in modules 1–6 were specifically expressed in one of the five developmental stages and may play an important function for their corresponding stages.

Figure 3.

Figure 3

Open in a new tab

Expression patterns of MTRGs. For each gene, the FPKM normalized value (by dividing gene expression at different time points with the maximum observed FPKM) is shown. Low expression level is in blue and high expression level is in red (see the color bar at the top of the figure). Red (1) and blue (0) represent the maximal and zero FPKM detected between five time points, respectively. The name of each module is shown on the left.

Expression analysis proves the robustness of the transcriptome and co-expression data set

To validate the RNA-sequencing (RNA-seq) dataset, we used reverse transcription-quantitative PCR (RT-qPCR) to verify the expression levels of three genes (CsGL3, CsTINY BRANCHED HAIR/CsTBH, and CsMYB60) known to regulate cucumber trichome development. In accordance with the glabrous phenotype of csgl3 (Figure 4A), CsGL3 was highly expressed at 18 HAS (Figure 4B) and belongs to module 1, which is consistent with our assumption that genes from module 1 are associated with multicellular trichome initiation. The spontaneous mutant cstbh exhibited tiny aborted multicellular trichome with round top cells rather than sharp-pointed or inflated glandular head cells (Figure 4C). CsTBH showed peak expression at 60 HAS and belonged to module 3 (Figure 4D), consistent with our assumption that genes from this module are related to tip formation or glandular head transition. CsMYB60 was reported to control the synthesis of flavonols and anthocyanins in multicellular trichomes (Liu et al., 2018). Overexpression of CsMYB60 resulted in black trichomes without morphological defects (Liu et al., 2018), suggesting that CsMYB60 may function in Stage V. Our transcriptome data and RT-qPCR results show that CsMYB60 was mainly expressed at Stage V and belonged to module 5 (Supplemental Figure S5C).

Figure 4.

Figure 4

Open in a new tab

Expression patterns of known and predicted stage-specific TFs. A and C, SEM images of trichomes in csgl3 (A) and cstbh (C). Bar = 10 μm. B and D, Comparison of CsGL3 (B) and CsTBH (D) expression levels from RNA-seq (FPKM) and RT-qPCR. CsTUA served as the internal control. Three biological replicates were performed. Error bars represent SD from three biological repeats. The maximum relative expression level was set as “1.” E1–J5, in situ hybridization performed at five developmental stages. CsGL3 was expressed in the trichome mother cell (blue arrow in E1) at Stage I but was undetectable at other stages (black arrow in E2–E5). Transcript accumulation of CsTBH was localized in the multicellular trichome especially the head cell at Stage III (blue arrows in G3), but was undetectable at other stages (black arrow in G2, G4, and G5). CsWIP6, CsHOX3, and CsMYB60 were highly expressed in developing trichomes at Stages II (blue arrow in F2), IV (blue arrows in H4), and V (blue arrow in I5), respectively. However, they were undetectable at other stages (black arrow in F3–F5, H2, H3, H5, and I2–I4). Antisense probes of each gene were used for detecting gene transcripts and sense probes of CsGL3 were used as control (J1–J5). Bar = 50 um.

There is no previous report on genes involved in Stages II and IV of cucumber multicellular trichome development. However, from the transcriptome data, it was evident that CsWIP6 (Csa5G365160) was strongly expressed in module 2. Similarly, the RT-qPCR expression level of CsWIP6 was also reported to be higher at 36 HAS (module 2) (Supplemental Figure S5A). In cotton, GhHOX3 is known to promote trichome and fiber elongation (Shan et al., 2014), which should associate with module 4. Therefore, we used a cucumber GhHOX3 putative ortholog, CsHOMEOBOX3 (CsHOX3) (Csa1G031750), to validate the expression pattern of genes associated with module 4. Our results clearly indicated that CsHOX3 is mainly expressed at 72 HAS (module 4) (Supplemental Figure S5B). The expression of CsGL3, CsWIP6, CsTBH, CsHOX3, and CsMYB60 was further analyzed by in situ hybridization experiments. The results show that each gene was mainly expressed in trichomes at their corresponding stages (Figure 4, E1–J5). In summary, the expression dynamics of known genes and genes identified here indicate the high quality and reliability of our dataset and the feasibility of the cluster set.

Candidate genes associated with different developmental stages of multicellular trichomes

To better understand the morphological changes and functional transitions involved in multicellular trichome development, we analyzed the biological processes in each co-expression module, especially modules 1–6.

In module 1, 80 genes were linked to multicellular trichome initiation, including CsGL3. This stage is characterized by the radial expansion out of the plane of cotyledons, suggesting that cell wall plasticity is needed (Cosgrove, 2005). Of 80 genes in module 1, nine were associated with cell wall loosening/modification, including three Xyloglucan endotransglucosylase (XTH) genes: CsXTH2 (Csa1G422480), CsXTH3 (Csa1G422470), and CsXTH5 (Csa5G602120). These XTH genes are known to reversibly or irreversibly loosen existing cell wall material, which enables cell expansion (Rose et al., 2002; Kushwah et al., 2020). Therefore, the expression of XTH genes in this stage suggests a possible role in multicellular trichome initiation. Two B-type cyclin (CYCB) genes (Csa6G454370 and Csa3G778340) were also found in module 1. CYCB proteins were recently reported to negatively regulate trichome initiation through interacting with Slwo or its homologs in tomato, N. benthamiana, and soybean (Glycine max) (Gao et al., 2017; Liu et al., 2020; Wu et al., 2020). Our results also support the important role of CYCB in trichome initiation.

The primary difference between unicellular and multicellular trichome is cell number, with unicellular trichome undergoing three or four rounds endoreduplication rather than mitosis. CYCB proteins have been reported to regulate the gap 2/mitosis transition stage, which is a key in determining if cells undergo endoreduplication or mitosis (Inze and De Veylder, 2006). Schnittger et al. (2002) reported that ectopic CYCLIN B1;2 expression in Arabidopsis trichome led to multicellular structure, while CYCB genes were not expressed in wild-type unicellular trichomes. Here, multicellular trichome expansion and division stage were linked to 56 genes within module 2, including a CYCB gene, CsCycB-like (Csa6G425780). Unlike Csa6G454370 and Csa3G778340, which were mainly expressed in Stage I, CsCycB-like was highly expressed in Stage II and later stages. Therefore, we speculate that CsCycB-like may play a major role in the transition of unicellular trichomes to multicellular trichomes in cucumber.

Cell wall organization and stiffening can be critical in determining trichome morphogenesis (Ivakov and Persson, 2013). In module 3, 42 genes were linked to multicellular trichome tip head formation (for nonglandular multicellular trichomes) or glandular head transition (for glandular multicellular trichomes), including CsTBH. CsWALLS ARE THIN1-LIKE (CsWAT1-like, Csa1G288020), and a pectin methylesterase (PME) gene, CsPME3 (Csa7G447990), were highly expressed in this module. AtWALLS ARE THIN1 (AtWAT1), the homolog of CsWAT1-like, is known to function in secondary cell wall formation and rigidity (Ranocha et al., 2010). Likewise, PMEs have been shown to affect cell wall rigidity through linear demethylesterification or random demethylesterification of pectin (Micheli, 2001). Additionally, a gene encoding proline-rich protein (PRP), CsPRP-like (Csa2G176690), with a similar expression pattern as CsTBH, was also identified. PRPs are known to play important roles in cell wall structures (Bradley et al., 1992). Therefore, we speculate that CsPRP-like may also participate in multicellular trichome tip formation or glandular head transition.

Multicellular trichome elongation (for nonglandular multicellular trichomes) or glandular head formation (for glandular multicellular trichomes) were linked to 41 genes within module 4. In this module, Csformin8 (Csa6G404250), encoding a filamentous actin (F-actin) binding gene, was highly expressed. Organized F-actins have been reported to be essential for both unicellular and multicellular trichome cell growth in Arabidopsis and tomato (Szymanski et al., 1999; Kang et al., 2016). Formin proteins are well-known regulators of actin dynamics and have been identified in various eukaryotic organisms, including fungi, animals, and plants (Deeks et al., 2002). Thus, combining its expression dynamics, Csformin8 may play a role in multicellular trichome elongation through nucleating F-actin organization. In glandular multicellular trichomes, bHLH TFs were proven to control glandular cell differentiation. In cotton, no glandular cells and gossypol were formed in knockout GhPGF plants (Ma et al., 2016; Chalvin et al., 2020). Likewise, knockout of SlMYC1 in tomato resulted in abnormal glandular head of type VI trichomes (Xu et al., 2018; Hua et al., 2020). In module 4, CsbHLH1 (Csa7G071650) was specifically expressed in the glandular head forming stage, suggesting a bHLH gene may also play an important role in glandular trichome formation in cucumber. In addition, seven lipid-related genes, including CsFATTY ACID DESATURASE5-LIKE1 (CsFAD5-like1; Csa4G006140), CsFAD5-like2 (Csa4G006150), CsFATTY ACID HYDRATASE-LIKE (Csa6G079750), Cs4-COUMARATE COA LIGASE3 (Csa2G433350), CsLIPOXYGENAS2 (Csa4G286940), Cs3-KETOACYL COA SYNTHASE19 (Csa5G177700), and CsGDSL-like (Csa2G346080), were highly expressed in this module, suggesting the complex lipid metabolism may be coupled with glandular head formation.

In modules 5 and 6, 139 genes were linked to multicellular trichome base formation (for nonglandular trichomes) or active metabolism (for glandular trichomes), including a C2H2 TF, CsTUBERCULE (CsTU, Csa5G577350). CsTU was reported to be essential for the cucumber fruit warty trait, which is the protuberance under the fruit spine (Yang et al., 2014). In line with the morphology of the leaf trichome base, it is assumed that CsTU may also affect epidermal growth at the base of nonglandular multicellular trichomes, leading to the bulge phenotype. MYB TFs play key roles in various life processes, including development, biotic and abiotic stress, and metabolism. In modules 5 and 6, four MYB TFs were identified, including two homologs expressed in tobacco glandular trichomes at late developmental stages (Nautiyal et al., 2020), reflecting the important role of MYB TFs in multicellular trichome formation. Interestingly, 56 of the 139 genes in modules 5 and 6 were barely expressed in csgl3 (FPKM < 1), suggesting that these genes may play a role in the synthesis of specialized metabolism in glandular trichomes or the formation of bulge base of nonglandular trichomes.

A total of 353 genes were highly expressed in more than one of the five stages, as shown in modules 7–16, which indicates common biological processes within stages. For example, CsROP9 (Csa1G696470), a member of a plant-specific family of Rho-related GTPases (ROPs), was highly expressed at Stages III and IV (module 13). ROPs are known to control pollen tube tip growth and unicellular trichome sharp branch morphogenesis in Arabidopsis (Fu et al., 2002; Cheung et al., 2003). CsROP9 expression was consistent with the morphological changes in multicellular trichomes observed within Stages III and IV, such as sharp-pointed head formation and rapid elongation. Furthermore, CsERF1A (Csa2G006270), a homolog of AaTAR1, was mainly expressed in Stages IV and V (module 7). Knockdown of AaTAR1 caused abnormal glands formation and dramatically reduced artemisinin content (Tan et al., 2015). These results suggest that genes that regulate glands formation may also influence the synthesis of specialized metabolites.

Co-expression network analysis of glandular trichomes

Glandular trichomes have the unique capacity to biosynthesize, store, and secrete specialized metabolites. They are described as bio-factories and are good targets for metabolic engineering studies. Therefore, understanding the gene regulatory networks of glandular trichomes can improve breeding efforts. Since glandular heads are formed at Stage IV, we suspect that genes in this stage are likely to be involved in the biosynthesis of specialized metabolites. Therefore, we constructed a co-expression network containing genes highly expressed in Stages IV and V (Figure 5; Supplemental Table S4). From the network, CsAOC3 (Csa5G366670), encoding an enzyme involved in jasmonic acid biosynthesis, was shown to be a major hub gene. Several other jasmonate acid (JA)-related genes, CsJAZ10 (Csa6G091930), CsMYC2 (Csa3G011620), and CsME3 (Csa7G074970), were also found in this network. Additionally, two gibberellin (GA) 20-oxidase genes, Csa5G172270 and Csa6G476070, were also included in this network. These results suggest that JA together with GA is essential for the synthesis of specialized metabolites in glandular trichomes.

Figure 5.

Figure 5

Open in a new tab

The correlation network of genes highly expressed at Stages IV and V. Genes with more edge numbers are shown in red, whereas blue means fewer edge numbers. Six plant hormone-related genes are shown in large circles.

Functional analysis of MTRGs by VIGS

Recently, a virus-induced gene silencing (VIGS) system for cucurbits was developed by Fang et al. (2021), which has the potential for functional analysis of screened MTRGs. To test the suitability of VIGS system, we conducted preexperiments using CsPDS and CsGL3 as positive controls. Four of 10 Tobacco ringspot virus 2 (TRSV2)::CsPDS plants showed photo-bleaching on the fourth true leaf, while three of the ten TRSV::CsGL3 plants showed decreased trichome density on the fourth true leaf. These results indicate that the system is suitable for functional verification of MTRGs. Some drawbacks of the VIGS system were the infection methods, which require the use of N. benthamiana as an intermediate host to cultivate the ringspot virus, and relatively late onset time. To overcome these limitations, germinating seeds were used instead of seedlings and agro-inoculation was used instead of rub-inoculation. Interestingly, the first true leaf had the corresponding phenotype and the silencing efficiency was 60% (Supplemental Figure S6).

We then verified the functions of several stage-specific TFs (CsGL3, CsTBH, CsHOX3, and CsbHLH1) involved in multicellular trichome formation. Results were obtained from the adaxial surface and petiolate of the first true leaf of three independent plants. The corresponding gene expression levels for each test plant were also recorded (Supplemental Figure S7). TRSV::CsGL3 infected plants showed an almost glabrous phenotype which was similar to its null mutant (Figure 6, B1–B4, and F). Trichomes on TRSV::CsTBH infected plants exhibit blunt head cells, which were not rounded as its null mutant (Figure 6, C1–C4). The trichomes density was also increased by 30% in TRSV::CsTBH infected plants. We speculate that there may be a feedback regulation that increases trichome initiation when no mature trichomes develop. TRSV::CsHOX3 reduced the length of nonglandular trichomes but did not influence glandular trichome formation (Figure 6 D1–D4 and G), which is similar with its putative orthologs GhHOX3 and SlHD8 (Shan et al., 2014; Hua et al., 2021). Conversely, TRSV::CsbHLH1 decreased the density of glandular trichomes but had no significant influence on nonglandular trichome formation (Figure 6, E1–E4). Although the bHLH TFs, GoPGF and SlMYC1, are known to control glands formation in cotton and tomato, CsbHLH1 does not belong to the same subfamily (Supplemental Figure S8). Therefore, a different mechanism may be involved in glandular trichome formation in cucumber. Altogether, these results demonstrate that the MTRGs identified in this study are reliable and can be utilized in future multicellular trichome research.

Figure 6.

Figure 6

Open in a new tab

VIGS analysis of CsGL3, CsTBH, CsHOX3, and CsbHLH1 in cucumber. A1–E4, Phenotypes of gene silenced plants: TRSV::00 (A1–A4), TRSV::CsGL3 (B1–B4), TRSV::CsTBH (C1–C4), TRSV::CsHOX3 (D1–D4), and TRSV::CsbHLH1 (E1–E4). First and third columns are pictures of adaxial surface of first true leaf; second and fourth columns are pictures of petiole of first true leaf. The blue arrows indicate glandular trichomes. Bars represent 1 mm. F, Density (per mm2) of different trichome types in different silenced plants. G, The length of nonglandular trichomes in TRSV:00 and TRSV::CsHOX3 infected plants. For (F) and (G): error bars represent SD from three biological repeats. Different letters and ** indicate statistically significant differences between groups (one-way ANOVA with post-hoc Tukey’s honest significant difference test; P < 0.01).

A model for multicellular trichome development

Based on our findings and previous reports, we proposed a model for multicellular trichome development in cucumber (Figure 7). First, during the initiation stage, a CsGL3 determined trichome mother cell expands relative to its surrounding cells. XTH genes associated with cell wall loosening/modification are also involved in cell expansion. Next, CsCycB-like mediates cell mitosis, which differentiates unicellular and multicellular trichomes. In Stage III, CsTBH and other cell wall organization-related genes regulates the changes in the head cell of multicellular trichomes. In Stage IV, the organization of F-actin is essential for the elongation of nonglandular trichomes, which is similar to unicellular trichome elongation. For glandular trichomes, bHLH TFs play important roles in gland formation. During the last stage, nonglandular trichome development involves a CsTU mediated bulge base formation, whereas glandular trichomes synthesize specialized metabolites under the control of JA and GA signaling.

Figure 7.

Figure 7

Open in a new tab

A proposed developmental model for glandular and nonglandular multicellular trichomes. CsGL3 and XTHs are associated with multicellular trichome initiation. CsCycB-like possibly mediates cell mitosis for the first division of multicellular trichomes. CsTBH and other cell wall organization-related genes regulate the changes in the head cell of multicellular trichomes. CsHOX3 mediates nonglandular multicellular trichome elongation and CsbHLH1 is involved in glandular head formation. CsTU plays a role in the base formation of nonglandular multicellular trichomes. JA and GA possibly participate in the synthesis of specialized metabolites in glandular multicellular trichomes.

Discussion

Cucumber cotyledon trichomes: a good model for studying multicellular trichome development

As a hair-like structure widely located on the aerial tissues of most terrestrial plants, the development of trichomes is accompanied by the continuous division and differentiation of various plant organs. Therefore, trichomes at different developmental stages, with most in later developmental stages, can be observed in a single plant organ, such as a young leaf. This makes it difficult to observe the trichome consecutive developmental process and to accurately sample stage-specific developing trichomes. In this study, we determined that trichome development follows a strict and uniform time sequence on cucumber cotyledons. Additionally, the large size of cucumber seeds makes it relatively simple to observe and sample the intact developmental process. These factors enabled us to generate time-course transcriptome data of multicellular trichomes using cucumber cotyledons. The cucumber glabrous mutant, csgl3, which is morphologically similar to WT plants but lacks trichomes, was used to narrow down the range of MTRGs. As expected, previously reported genes that control multicellular trichomes development in cucumber were detected in our data. Several homologous genes, known to control trichome development in other species, such as GhHOX3 and AaTAR1, were also detected with corresponding expression patterns. These findings indicate that trichomes on cucumber cotyledons are a good model for studying multicellular trichome development.

The similarities and differences between unicellular and multicellular trichome formation

The genetic mechanisms of unicellular trichomes development have been well studied. However, our understanding of the regulatory network involved in multicellular trichomes development is still in its infancy. Also, several studies suggest that the genes involved in unicellular trichomes development may be different from that of multicellular trichomes. Ectopic expression of a maize (Zea mays) R gene, encoding a bHLH TF, triggers excess trichomes differentiation in Arabidopsis, whereas it did not affect tobacco and tomato trichomes (Lloyd et al., 1992). Likewise, the heterologous overexpression of GLABROUSGL1, an R2R3 MYB gene crucial for unicellular trichome development in Arabidopsis, failed to alter the trichome phenotype of tobacco. On the other hand, overexpressing the MIXTA gene from Antirrhinum majus or G. hirsutum did not alter trichome production in Arabidopsis but produced supernumerary trichomes in tobacco (Payne et al., 1999). In our transcriptome data, Csa5G148680, the putative ortholog of AtGL1, was not detected at any trichome development stages or shoot apical meristem, indicating that it may not be the regulator of multicellular trichome formation in cucumber. These results suggest that the regulators of multicellular and unicellular trichome formation are probably different.

Unicellular trichomes and nonglandular multicellular trichomes share some common morphological features, such as a sharp pointed tip and papillae on mature trichomes. Therefore, it was speculated that they may share similar mechanisms regarding their morphogenesis. To test this hypothesis, we analyzed the transcriptome data from Arabidopsis unicellular trichomes (Marks et al., 2009) and found that 113 homologs of MTRGs were also preferentially expressed in Arabidopsis (expression level is more than two folds in trichomes compared with processed shoots) (Supplemental Table S5). Gene ontology (GO) enrichment term of the 113 genes indicated that “cell morphogenesis involved in differentiation” was one of the significantly enriched groups, which suggests they do have some similar morphogenesis (Supplemental Figure S9). Furthermore, the cytoskeleton of top cell in tomato multicellular trichomes was similar to the sharp tips in Arabidopsis unicellular trichomes (Szymanski et al., 1999; Chang et al., 2019). Altogether, the results indicate that unicellular and multicellular trichomes may have different regulatory networks but share partially similar morphogenesis.

Genes involved in multicellular trichome development

Here, we conducted a time-course transcriptome analysis of multicellular trichomes and reported several genes involved in multicellular trichome development. To determine the possibility of these genes being used as targets for future studies, we first verified the reliability of the transcriptome with known genes. The results indicate that their expression patterns were in line with their mutant phenotype. Genes reported to be involved in multicellular trichome formation in other species were also found in our transcriptome data. The representative gene, CsHOX3, was verified to be highly expressed in elongating trichomes, which was consistent with the function of GhHOX3 in cotton and SlHD8 in tomato (Shan et al., 2014; Hua et al., 2021). Of the MTRGs identified here, the function of CsbHLH1 was tested by VIGS. From the results, CsbHLH1 plays a function in glandular trichome formation, which is in line with its expressional patterns. Altogether, these findings indicated that the genes reported here are excellent targets for further functional analysis.

Conclusion

In this study, we observed and divided multicellular trichome formation into five consecutive stages: (I) initiation, (II) first division, (III) tip head formation/glandular head transition, (IV) elongation/glandular head formation, and (V) base formation/active metabolism. According to time-course transcriptome analysis, we screened 711 MTRGs and proposed a model for multicellular trichome development. The VIGS-based functional analysis, together with the fact that functional known genes in cucumber (CsGL3, CsTBH, etc.) and cucumber homologs from other species (CsHOX3, CsERF1A, etc.) were shown in our data set with the expression patterns in line with their functions, indicate that our transcriptome data are solid a resource for future functional studying in multicellular trichome development.

Materials and methods

Plant material collection and RNA-seq

Seeds from cucumber (C. sativus L.) inbred line 3,542 were first soaked in water (initially at 60°C and then naturally cooling to room temperature) for 1 h. After soaking, seeds were cultivated at a constant temperature of 28°C under 12-h light/12-h dark conditions. The blade bases of cotyledons were sampled at 18, 36, 60, 72, and 96 HAS (Supplemental Figure S4A). This experiment included five seedlings per replicate (at all five stages) with three biological replicates. Cucumber inbred lines RIL-46 W (CsGL3) and RIL-46 M (a spontaneous glabrous mutant found in the F6 RIL-46 M, csgl3) were planted in a growth incubator at 28°C under 16-h light and 23°C under 8-h dark conditions. The shoot apical meristem of the two lines was sampled at 15 d after planting (two biological replicates with five plants per replicate).

Total RNA was extracted using a Huayueyang RNA extraction kit (Huayueyang, Beijing, China). RNA-seq libraries were constructed according to the manufacturer’s protocol of the Vazyme mRNA-seq library preparation kit (Vazyme, Nanjing, China) and were sequenced on an Illumina HiSeq X sequencer.

Read mapping and DEG analysis

The raw paired-end reads were subjected to preprocessing where low-quality reads were trimmed using the SolexaQA version 2.5 software (Cox et al., 2010). High-quality reads were mapped to the cucumber 9930 reference genome version 2 (http://cucurbitgenomics.org/organism/2) (Huang et al., 2009) using Hisat version 2-2.0.4 (Kim et al., 2015) with default settings. The bam files of uniquely mapped reads were used as inputs for the Cufflinks (version 2.20) software (Ghosh and Chan, 2016) and gene expression was estimated in terms of FPKM. Genes with at least two-fold change in expression (FPKM ≥ 2) and an FDR ˂ 0.05 were considered DEGs using DEseq version 2 (Love et al., 2014).

Scanning electron microscopy

Scanning electron microscopy (SEM) of wild-type cucumber trichomes (line 3542) was performed on cotyledons at several time intervals (12, 18, 24, 30, 36, 42, 48, 54, 60, 66, 72, 78, 84, 90, 96, 102, 108, and 114 HAS). The cotyledons of csgl3 and cstbh were also sampled for SEM at 114 HAS. Samples were processed as previously described (Chen et al., 2014) and observed using Hitachi S-4700 SEM with a 2-kV accelerating voltage.

Gene co-expression and description

Normalization of genes expression values was achieved by dividing their expression values at different time points with the maximum expression value between the five time points (e.g. we divide the expression value of each stage by the expression value of the first stage to get the normalized expression value of CsGL3). Based on the normalized expression values, the MeV (version 4.9) software (http://www.tm4.org/mev.html) with the k-means method was used for co-expression analysis. Cucumber gene description was obtained from the cucurbit genome database (http://cucurbitgenomics.org/organism/2).

RT-qPCR analysis

Total RNA was isolated using a Huayueyang RNA extraction kit (Huayueyang) and then reverse transcribed using PrimeScript reagent Kit with gDNA Eraser (TaKaRa, Shiga, Japan) following the manufacturers’ protocol. RT-qPCR was conducted in 96-well plates with an ABI 7500 Real-Time PCR System (Applied Biosystems, Waltham, MA, USA) using SYBR Premix Ex Taq (TaKaRa). Three biological replicates and three technical replicates were performed for each combination of cDNA samples and primer pairs. The cucumber α-TUBULIN served as the internal control gene (Wan et al., 2010). The gene-specific RT-qPCR primers are listed in Supplemental Table S6.

In situ hybridization

In situ hybridization was performed as described by Zhang et al. (2013). The primers corresponding to tested genes were added by T7 and SP6 promoter sequence for antisense and sense probe, respectively. In vitro transcription was conducted with either T7 or SP6 RNA polymerase to generate the antisense or sense probe using the purified (polymerase chain reaction) PCR product as the template. The primers used for probe synthesis were listed in Supplementary Table S6.

Gene regulatory network inference

The gene regulatory networks were constructed using the Pearson correlation coefficient. Only correlations ≥0.9 and P.adj <0.05 were visualized in Cytoscape (Shannon et al., 2003).

VIGS assay and phenotypic observation

TRSV-based VIGS was used to analyze the potential roles of MTRGs in cucumber. The VIGS experiments were performed according to previous reports (Fang et al., 2021) with an improved injection method. In short, the unique 300–500 bp CDS sequences of each target gene (primers are shown in Supplemental Table S6) were insert into the Sna BI restriction site of pTRSV2 and then transformed into Agrobacterium tumefaciens GV3101::pMP90. Germinating cucumber seeds, when primary roots reached ∼1 cm, were infected with mixed pTRSV1 and pTRSV2 (containing different target fragments) by vacuum-infiltration under minus 900 kPa condition for 5 min. The seeds were then put on half MS solid medium with 100 μM acetosyringone for 5 d. After that, the seedlings were planted in half Hoagland solution for another 15 d. The adaxial surface of the first true leaf was used to count trichome density and trichomes on the petiole of the first true leaf were used to count trichome length.

Propidium iodide staining

Cotyledons at several time intervals (24, 36, and 48 HAS) were sliced and dipped into propidium iodide buffer (100 μg/mL) for 3 min. The samples were then imaged by a Leica SP8 confocal microscope.

Phylogenetic analysis

Gene information for protein alignments was downloaded from TAIR (https://www.arabidopsis.org/) and National Center for Biotechnology Information (https://www.ncbi.nlm.nih.gov/). The phylogenic analysis was conducted using the neighbor-joining method with 1,000 bootstrap replicates in MEGA version 5.

GO analysis

GO terms of genes were taken from Ensembl BioMarts. The GO enrichment analysis was performed with topGO32. The GO terms with a P-value ˂ 0.05 were considered significantly enriched.

Statistical analysis

For all expression level analyses and phenotypic analyses, statistical significance was determined by one-way ANOVA with post-hoc Tukey’s honest significant difference test; P < 0.01.

Accession numbers

The generated raw reads have been uploaded to National Center for Biotechnology Information under accession number PRJNA783663. Sequence data from this article can be found in the Cucurbit Genomics Database (http://cucurbitgenomics.org/) under accession numbers: CsGL3 (Csa6G514870), CsTBH (Csa3G748220), CsHOX3 (Csa1G031750), CsbHLH1 (Csa7G071650), and CsPDS (Csa4G011080).

Supplemental data

The following materials are available in the online version of this article.

Supplemental Figure S1. Developing trichomes on young leaves.

Supplemental Figure S2. Distribution of trichomes on cucumber cotyledons.

Supplemental Figure S3. First division of multicellular trichome at Stage II.

Supplemental Figure S4. Examples of sampling parts.

Supplemental Figure S5. RT-qPCR analysis of three stage-specific TFs.

Supplemental Figure S6. Analysis of VIGS efficiency in cucumber via CsPDS.

Supplemental Figure S7. Gene expression levels in the silenced plants.

Supplemental Figure S8. Phylogenetic analysis of CsbHLH1 along with bHLH TFs from Arabidopsis and other TFs involved in trichome development.

Supplemental Figure S9. GO term enrichment of MTRGs also related to unicellular trichome formation in Arabidopsis.

Supplemental Table S1. Summary of RNA-seq read mapping results.

Supplemental Table S2. Expression level of DEGs.

Supplemental Table S3. List of MTRGs and co-expression module set.

Supplemental Table S4. All 621 connections of gene correlation network in Stages IV and V.

Supplemental Table S5. List of MTRGs also related to unicellular trichome formation in Arabidopsis.

Supplemental Table S6. Primers used in this study.

Supplementary Material

kiac050_Supplementary_Data
Click here for additional data file. (4.5MB, zip)

Acknowledgments

We are grateful to Prof. Yiqun Weng (University of Wisconsin–Madison, Madison, USA) for supplying the cucumber seeds of RIL-46 W (CsGL3) and RIL-46 M (csgl3), and revision of the manuscript.

Funding

This study was supported by the National Natural Science Foundation of China (31830080), the 111 project (B17043), the Project of Yazhouwan Scientific and Technological Administration of Sanya (SYND-2021-18), and the Project of Beijing Agricultural Innovation Consortium (BAIC01).

Conflict of interest statement. The authors declare that they have no conflicts of interest.

Contributor Information

Mingming Dong, Department of Vegetable Science, College of Horticulture, China Agricultural University, Beijing 100193, China.

Shudan Xue, Department of Vegetable Science, College of Horticulture, China Agricultural University, Beijing 100193, China.

Ezra S Bartholomew, Department of Vegetable Science, College of Horticulture, China Agricultural University, Beijing 100193, China.

Xuling Zhai, Department of Vegetable Science, College of Horticulture, China Agricultural University, Beijing 100193, China.

Lei Sun, Department of Vegetable Science, College of Horticulture, China Agricultural University, Beijing 100193, China.

Shuo Xu, Department of Vegetable Science, College of Horticulture, China Agricultural University, Beijing 100193, China.

Yaqi Zhang, Department of Vegetable Science, College of Horticulture, China Agricultural University, Beijing 100193, China.

Shuai Yin, Department of Vegetable Science, College of Horticulture, China Agricultural University, Beijing 100193, China.

Wenyue Ma, Department of Vegetable Science, College of Horticulture, China Agricultural University, Beijing 100193, China.

Shuying Chen, Department of Vegetable Science, College of Horticulture, China Agricultural University, Beijing 100193, China.

Zhongxuan Feng, Department of Vegetable Science, College of Horticulture, China Agricultural University, Beijing 100193, China.

Chao Geng, Department of Plant Pathology, Shandong Provincial Key Laboratory of Agricultural Microbiology, College of Plant Protection, Shandong Agricultural University, Tai’an 271018, China.

Xiangdong Li, Department of Plant Pathology, Shandong Provincial Key Laboratory of Agricultural Microbiology, College of Plant Protection, Shandong Agricultural University, Tai’an 271018, China.

Xingwang Liu, Department of Vegetable Science, College of Horticulture, China Agricultural University, Beijing 100193, China; Engineering Research Center of Breeding and Propagation of Horticultural Crops, Chinese Ministry of Education, Beijing 100193, China; Sanya Institute of China Agricultural University, Sanya 572019, China.

Huazhong Ren, Department of Vegetable Science, College of Horticulture, China Agricultural University, Beijing 100193, China; Engineering Research Center of Breeding and Propagation of Horticultural Crops, Chinese Ministry of Education, Beijing 100193, China; Sanya Institute of China Agricultural University, Sanya 572019, China.

 

These authors contributed equally (M.D. and S.X.)

M.D., S.Xue., X.Liu, and H.R. designed the research. M.D. and S.Xue. designed the experimental approach and performed the experiments. X.Z., L.S., S.Xu., Y.Z., S.Y., S.C., and Z.F. helped with RNA-seq data analysis. W.M., C.G., and X.Li helped with VIGS experiments. M.D. drafted the manuscript, with input from X.Liu, E.S.B., and H.R. All authors contributed to ideas, discussed the results, and edited the manuscript.

The authors responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (https://academic.oup.com/plphys/pages/general-instructions) are: Huazhong Ren (renhuazhong@cau.edu.cn) and Xingwang Liu (liuxw01@cau.edu.cn).

References

  1. Abe M, Katsumata H, Komeda Y, Takahashi T (2003) Regulation of shoot epidermal cell differentiation by a pair of homeodomain proteins in Arabidopsis. Development  130:  635–643 [DOI] [PubMed] [Google Scholar]
  2. Akhtar MQ, Qamar N, Yadav P, Kulkarni P, Kumar A, Shasany AK (2017) Comparative glandular trichome transcriptome-based gene characterization reveals reasons for differential (-)-menthol biosynthesis in Mentha species. Physiol Plant  160:  128–141 [DOI] [PubMed] [Google Scholar]
  3. Bradley DJ, Kjellbom P, Lamb CJ (1992) Elicitor-induced and wound-induced oxidative cross-linking of a proline-rich plant cell wall protein—a novel, rapid defense response. Cell  70:  21–30 [DOI] [PubMed] [Google Scholar]
  4. Braich S, Baillie RC, Jewell LS, Spangenberg GC, Cogan NOI (2019) Generation of a comprehensive transcriptome atlas and transcriptome dynamics in medicinal cannabis. Sci Rep  9:  16583. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Chalvin C, Drevensek S, Dron M, Bendahmane A, Boualem A (2020) Genetic control of glandular trichome development. Trends Plant Sci  25: 477–487 [DOI] [PubMed] [Google Scholar]
  6. Chang J, Xu ZJ, Li M, Yang MN, Qin HY, Yang J, Wu S (2019) Spatiotemporal cytoskeleton organizations determine morphogenesis of multicellular trichomes in tomato. PLoS Genet  15:  24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Chen C, Liu M, Jiang L, Liu X, Zhao J, Yan S, Yang S, Ren H, Liu R, Zhang X (2014) Transcriptome profiling reveals roles of meristem regulators and polarity genes during fruit trichome development in cucumber (Cucumis sativus L.). J Exp Bot  65:  4943–4958 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cheung AY, Chen CYH, Tao LZ, Andreyeva T, Twell D, Wu HM (2003) Regulation of pollen tube growth by Rac-like GTPases. J Exp Bot  54:  73–81 [DOI] [PubMed] [Google Scholar]
  9. Cosgrove DJ (2005) Growth of the plant cell wall. Nat Rev Mol Cell Biol  6:  850–861 [DOI] [PubMed] [Google Scholar]
  10. Cox MP, Peterson DA, Biggs PJ (2010) SolexaQA: at-a-glance quality assessment of Illumina second-generation sequencing data. BMC Bioinformatics  11:  6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Cui JY, Miao H, Ding LH, Wehner TC, Liu PN, Wang Y, Zhang SP, Gu XF (2016) A new glabrous gene (csgl3) identified in trichome development in cucumber (Cucumis sativus L.). PLoS ONE  11:  e0148422. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Deeks MJ, Hussey PJ, Davies B (2002) Formins: Intermediates in signal-transduction cascades that affect cytoskeletal reorganization. Trend Plant Sci  7:  492–498 [DOI] [PubMed] [Google Scholar]
  13. Fang L, Wei XY, Liu LZ, Zhou LX, Tian YP, Geng C, Li XD (2021) A tobacco ringspot virus-based vector system for gene and microRNA function studies in cucurbits. Plant Physiol  186:  853–864 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Fu Y, Li H, Yang Z (2002) The ROP2 GTPase controls the formation of cortical fine F-actin and the early phase of directional cell expansion during Arabidopsis organogenesis. Plant Cell  14:  777–794 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Gao S, Gao Y, Xiong C, Yu G, Chang J, Yang Q, Yang C, Ye Z (2017) The tomato B-type cyclin gene, SlCycB2, plays key roles in reproductive organ development, trichome initiation, terpenoids biosynthesis and Prodenia litura defense. Plant Sci  262:  103–114 [DOI] [PubMed] [Google Scholar]
  16. Ghosh S, Chan CKK (2016) Analysis of RNA-Seq data using tophat and cufflinks. In  Edwards D, ed, Plant Bioinformatics: Methods and Protocols, Ed 2, Vol 1374. Humana Press Inc., Totowa, NJ, pp 339–361 [DOI] [PubMed] [Google Scholar]
  17. Hua B, Chang J, Xu Z, Han X, Xu M, Yang M, Yang C, Ye Z, Wu S (2021) HOMEODOMAIN PROTEIN8 mediates jasmonate-triggered trichome elongation in tomato. New Phytol  230:  1063–1077 [DOI] [PubMed] [Google Scholar]
  18. Hua B, Chang J, Wu M, Xu Z, Zhang F, Yang M, Xu H, Wang LJ, Chen XY, Wu S (2020) Mediation of JA signaling in glandular trichomes by the woolly/SlMYC1 regulatory module improves pest resistance in tomato. Plant Biotechnol J  19: 375–393 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Huang SW, Li RQ, Zhang ZH, Li L, Gu XF, Fan W, Lucas WJ, Wang XW, Xie BY, Ni PX, et al. (2009) The genome of the cucumber, Cucumis sativus L. Nat Genet  41:  1275–1281 [DOI] [PubMed] [Google Scholar]
  20. Inze D, De Veylder L (2006) Cell cycle regulation in plant development. Annu Rev Genet  40:  77–105 [DOI] [PubMed] [Google Scholar]
  21. Ivakov A, Persson S (2013) Plant cell shape: modulators and measurements. Front Plant Sci  4:  439. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kang JH, Campos ML, Zemelis-Durfee S, Al-Haddad JM, Jones AD, Telewski FW, Brandizzi F, Howe GA (2016) Molecular cloning of the tomato Hairless gene implicates actin dynamics in trichome-mediated defense and mechanical properties of stem tissue. J Exp Bot  67:  5313–5324 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Karabourniotis G, Papadopoulos K, Papamarkou M, Manetas Y (1992) UltravioIet-B radiation absorbing capacity of leaf hairs. Physiol Plant  86:  414–418 [Google Scholar]
  24. Kennedy GG (2003) Tomato, pests, parasitoids, and predators: Tritrophic interactions involving the genus Lycopersicon. Annu Rev Entomol  48:  51–72 [DOI] [PubMed] [Google Scholar]
  25. Kim D, Landmead B, Salzberg SL (2015) HISAT: A fast spliced aligner with low memory requirements. Nat Methods  12:  357–360 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kushwah S, Banasiak A, Nishikubo N, Derba-Maceluch M, Majda M, Endo S, Kumar V, Gomez L, Gorzsas A, McQueen-Mason S, et al. (2020) Arabidopsis XTH4 and XTH9 contribute to wood cell expansion and secondary wall formation. Plant Physiol  182:  1946–1965 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Li B, Liang S, Alariqi M, Wang F, Wang G, Wang Q, Xu Z, Yu L, Zafar MN, Sun L, et al. (2020) The application of temperature sensitivity CRISPR/LbCpf1 (LbCas12a) mediated genome editing in allotetraploid cotton (G. hirsutum) and creation of nontransgenic, gossypol-free cotton. Plant Biotechnol J  19:  221–223 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Liu M, Zhang C, Duan L, Luan Q, Li J, Yang A, Qi X, Ren Z (2018) The colour determination of cucumber fruit spines. J Exp Bot  70:  69–84 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Liu S, Fan L, Liu Z, Yang X, Zhang Z, Duan Z, Liang Q, Imran M, Zhang M, Tian Z (2020) A Pd1-Ps-P1 feedback loop controls pubescence density in soybean. Mol Plant  13: 1768–1783 [DOI] [PubMed] [Google Scholar]
  30. Liu X, Bartholomew E, Cai Y, Ren H (2016) Trichome-related mutants provide a new perspective on multicellular trichome initiation and development in cucumber (Cucumis sativus L). Front Plant Sci  7:  1187. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Lloyd AM, Walbot V, Davis RW (1992) Arabidopsis and nicotiana anthocyanin production activated by maize regulators R and C1. Science  258:  1773–1775 [DOI] [PubMed] [Google Scholar]
  32. Love MI, Huber W, Anders S (2014) Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol  15:  38. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Ma D, Hu Y, Yang C, Liu B, Fang L, Wan Q, Liang W, Mei G, Wang L, Wang H, et al. (2016) Genetic basis for glandular trichome formation in cotton. Nat Commun  7:  10456. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Marks MD, Wenger JP, Gilding E, Jilk R, Dixon RA (2009) Transcriptome analysis of Arabidopsis wild-type and gl3-sst sim trichomes identifies four additional genes required for trichome development. Mol Plant  2:  803–822 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Mathur J, Spielhofer P, Kost B, Chua NH (1999) The actin cytoskeleton is required to elaborate and maintain spatial patterning during trichome cell morphogenesis in Arabidopsis thaliana. Development  126:  5559–5568 [DOI] [PubMed] [Google Scholar]
  36. Matias-Hernandez L, Jiang W, Yang K, Tang K, Brodelius PE, Pelaz S (2017) AaMYB1 and its orthologue AtMYB61 affect terpene metabolism and trichome development in Artemisia annua and Arabidopsis thaliana. Plant J  90:  520–534 [DOI] [PubMed] [Google Scholar]
  37. Mauricio R, Rausher MD (1997) Experimental manipulation of putative selective agents provides evidence for the role of natural enemies in the evolution of plant defense. Evolution  51:  1435–1444 [DOI] [PubMed] [Google Scholar]
  38. Micheli F (2001) Pectin methylesterases: cell wall enzymes with important roles in plant physiology. Trend Plant Sci  6:  414–419 [DOI] [PubMed] [Google Scholar]
  39. Nautiyal AK, Gani U, Sharma P, Kundan M, Fayaz M, Lattoo SK, Misra P (2020) Comprehensive transcriptome analysis provides insights into metabolic and gene regulatory networks in trichomes of Nicotiana tabacum. Plant Mol Biol  102:  625–644 [DOI] [PubMed] [Google Scholar]
  40. Pan Y, Bo K, Cheng Z, Weng Y (2015) The loss-of-function GLABROUS 3 mutation in cucumber is due to LTR-retrotransposon insertion in a class IV HD-ZIP transcription factor gene CsGL3 that is epistatic over CsGL1. BMC Plant Biol  15:  302. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Pattanaik S, Patra B, Singh SK, Yuan L (2014) An overview of the gene regulatory network controlling trichome development in the model plant, Arabidopsis. Front Plant Sci  5:  259. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Payne T, Clement J, Arnold D, Lloyd A (1999) Heterologous myb genes distinct from GL1 enhance trichome production when overexpressed in Nicotiana tabacum. Development  126:  671–682 [DOI] [PubMed] [Google Scholar]
  43. Ranocha P, Denance N, Vanholme R, Freydier A, Martinez Y, Hoffmann L, Kohler L, Pouzet C, Renou JP, Sundberg B, et al. (2010) Walls are thin 1 (WAT1), an Arabidopsis homolog of Medicago truncatula NODULIN21, is a tonoplast-localized protein required for secondary wall formation in fibers. Plant J  63:  469–483 [DOI] [PubMed] [Google Scholar]
  44. Rose JKC, Braam J, Fry SC, Nishitani K (2002) The XTH family of enzymes involved in xyloglucan endotransglucosylation and endohydrolysis: current perspectives and a new unifying nomenclature. Plant And Cell Physiol  43:  1421–1435 [DOI] [PubMed] [Google Scholar]
  45. Schellmann S, Hulskamp M (2005) Epidermal differentiation: Trichomes in Arabidopsis as a model system. Int J Dev Biol  49:  579–584 [DOI] [PubMed] [Google Scholar]
  46. Schnittger A, Schobinger U, Stierhof YD, Hulskamp M (2002) Ectopic B-type cyclin expression induces mitotic cycles in endoreduplicating Arabidopsis trichomes. Curr Biol  12:  415–420 [DOI] [PubMed] [Google Scholar]
  47. Shan CM, Shangguan XX, Zhao B, Zhang XF, Chao LM, Yang CQ, Wang LJ, Zhu HY, Zeng YD, Guo WZ, et al. (2014) Control of cotton fibre elongation by a homeodomain transcription factor GhHOX3. Nat Commun  5:  5519. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N, Schwikowski B, Ideker T (2003) Cytoscape: A software environment for integrated models of biomolecular interaction networks. Genome Res  13:  2498–2504 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Soetaert SSA, Van Neste CMF, Vandewoestyne ML, Head SR, Goossens A, Van Nieuwerburgh FCW, Deforce DLD (2013) Differential transcriptome analysis of glandular and filamentous trichomes in Artemisia annua. BMC Plant Biol  13:  14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Szymanski DB, Jilk RA, Pollock SM, Marks MD (1998) Control of GL2 expression in Arabidopsis leaves and trichomes. Development  125:  1161–1171 [DOI] [PubMed] [Google Scholar]
  51. Szymanski DB, Marks MD, Wick SM (1999) Organized F-actin is essential for normal trichome morphogenesis in Arabidopsis. Plant Cell  11:  2331–2347 [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Tan H, Xiao L, Gao S, Li Q, Chen J, Xiao Y, Ji Q, Chen R, Chen W, Zhang L (2015) TRICHOME AND ARTEMISININ REGULATOR 1 is required for trichome development and artemisinin biosynthesis in Artemisia annua. Mol Plant  8:  1396–1411 [DOI] [PubMed] [Google Scholar]
  53. Walford SA, Wu Y, Llewellyn DJ, Dennis ES (2012) Epidermal cell differentiation in cotton mediated by the homeodomain leucine zipper gene, GhHD-1. Plant J  71:  464–478 [DOI] [PubMed] [Google Scholar]
  54. Wan HJ, Zhao ZG, Qian CT, Sui YH, Malik AA, Chen JF (2010) Selection of appropriate reference genes for gene expression studies by quantitative real-time polymerase chain reaction in cucumber. Anal Biochem  399:  257–261 [DOI] [PubMed] [Google Scholar]
  55. Wu ML, Cui YC, Ge L, Cui LP, Xu ZC, Zhang HY, Wang ZJ, Zhou D, Wu S, Chen L, et al. (2020) NbCycB2 represses Nbwo activity via a negative feedback loop in tobacco trichome development. J Exp Bot  71:  1815–1827 [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Xie L, Yan T, Li L, Chen M, Ma Y, Hao X, Fu X, Shen Q, Huang Y, Qin W, et al. (2021) The WRKY transcription factor AaGSW2 promotes glandular trichome initiation in Artemisia annua. J Exp Bot  72:  1691–1701 doi:10.1093/jxb/eraa523 [DOI] [PubMed] [Google Scholar]
  57. Xu J, van Herwijnen ZO, Dräger DB, Sui C, Haring MA, Schuurink RC (2018) SlMYC1 regulates type VI glandular trichome formation and terpene biosynthesis in tomato glandular cells. Plant Cell  30:  2988–3005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Yan T, Chen M, Shen Q, Li L, Fu X, Pan Q, Tang Y, Shi P, Lv Z, Jiang W, et al. (2017) HOMEODOMAIN PROTEIN 1 is required for jasmonate-mediated glandular trichome initiation in Artemisia annua. New Phytol  213:  1145–1155 doi:10.1111/nph.14205 [DOI] [PubMed] [Google Scholar]
  59. Yan T, Li L, Xie L, Chen M, Shen Q, Pan Q, Fu X, Shi P, Tang Y, Huang H, et al. (2018) A novel HD-ZIP IV/MIXTA complex promotes glandular trichome initiation and cuticle development in Artemisia annua. New Phytol  218:  567–578 doi:10.1111/nph.15005 [DOI] [PubMed] [Google Scholar]
  60. Yang C, Gao Y, Gao S, Yu G, Xiong C, Chang J, Li H, Ye Z (2015) Transcriptome profile analysis of cell proliferation molecular processes during multicellular trichome formation induced by tomato WoV gene in tobacco. BMC Genomics  16:  868. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Yang CX, Li HX, Zhang JH, Luo ZD, Gong PJ, Zhang CJ, Li JH, Wang TT, Zhang YY, Lu YE, et al. (2011) A regulatory gene induces trichome formation and embryo lethality in tomato. Proc Natl Acad Sci USA  108:  11836–11841 [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Yang X, Zhang W, He H, Nie J, Bie B, Zhao J, Ren G, Li Y, Zhang D, Pan J, et al. (2014) Tuberculate fruit gene Tu encodes a C2H2 zinc finger protein that is required for the warty fruit phenotype in cucumber (Cucumis sativus L.). Plant J  78:  1034–1046 [DOI] [PubMed] [Google Scholar]
  63. Zhang XL, Zhou Y, Ding L, Wu ZG, Liu RY, Meyerowitz EM (2013) Transcription repressor HANABA TARANU controls flower development by integrating the actions of multiple hormones, floral organ specification genes, and GATA3 family genes in Arabidopsis. Plant Cell  25:  83–101 [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Zhao JL, Wang YL, Yao DQ, Zhu WY, Chen L, He HL, Pan JS, Cai R (2015) Transcriptome profiling of trichome-less reveals genes associated with multicellular trichome development in Cucumis sativus. Mol Genet Genomics  290:  2007–2018 [DOI] [PubMed] [Google Scholar]
  65. Zhu HY, Sun XF, Zhang Q, Song PY, Hu QM, Zhang XJ, Li X, Hu JB, Pan JS, Sun SR, et al. (2018) GLABROUS (CmGL) encodes a HD-ZIP IV transcription factor playing roles in multicellular trichome initiation in melon. Theor Appl Genet  131:  569–579 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

kiac050_Supplementary_Data
Click here for additional data file. (4.5MB, zip)

Articles from Plant Physiology are provided here courtesy of Oxford University Press

RESOURCES