Overexpression of the LcCUC2-like gene in Arabidopsis thaliana alters the cotyledon morphology and increases rosette leaf number

Plant materials, growth conditions, and treatments

Samples of leaves at different developmental stages (Tu et al., 2019) were harvested from an adult L. chinense tree in Xiashu, Jurong County, Jiangsu Province, China (119°13′20^′E, 32°7′8^′N) (Yang et al., 2014). Plant materials were maintained at −80 °C prior to analysis.

Wild-type (WT) A. thaliana (Columbia-0), transgenic A. thaliana, Nicotiana benthamiana (Ben) plants, and L. chinense seedlings were cultivated at 23 °C under a long-day photoperiod (16 h light, 8 h dark) in a growth chamber (70% relative humidity).

For the IAA treatment, L. chinense seedlings grown in the incubator for 5 months were sprayed with 200 µM IAA on leaves and then sampled at 48 h. Water was used as the control. The samples were placed in an ultra-low-temperature freezer for further analysis.

DNA, RNA extraction, and RT-qPCR

A DNAsecure Plant Kit (Tiangen) was used for DNA isolation from L. chinense. RNA was extracted from L. chinense and A. thaliana leaves using the RNAprep Pure Plant Kit (Tiangen). Subsequently, first-strand cDNA was synthesized using PrimeScript™ RT Master Mix (TaKaRa) in accordance with the manufacturer’s protocol. The LcCUC2L transcript levels in leaves of different developmental stages (Tu et al., 2019) were investigated by RT-qPCR. Moreover, we also detected LcCUC2L expression levels under IAA treatment. L. chinense Actin 97 was used as the internal control. In addition, the transcription levels of LcCUC2L and some other genes in WT and transgenic lines were examined using specific primers (Table S1). A. thaliana Actin 2 was used as the internal quantitative control. RT-qPCR was carried out using the SYBR Premix EX Taq kit (TaKaRa). We applied the 2^−ΔΔC_T method to analyze the data from relative quantification (Livak & Schmittgen, 2001).

Cloning and bioinformatics analysis

Based on the known transcriptome database (Ma et al., 2018), the CUC2 gene was identified in L. chinense. The full-length LcCUC2L gene was obtained using RT-PCR and RACE. The LcCUC2L gene was assembled with the middle region, 5^′ sequence and 3^′ sequence. ORF Finder was used for prediction of the open reading frame (ORF) and protein sequence. The structure of the LcCUC2L gene was analyzed by the online software GSDS (http://gsds.gao-lab.org/). Analysis of the physicochemical characteristics of the proteins was conducted online at the following website: (https://web.expasy.org/protparam/). PredictProtein was used to forecast protein secondary structure. Subcellular localization was predicted by applying WoLF PSORT. Protein conserved domain analysis of LcCUC2L was performed with NCBI (https://www.ncbi.nlm.nih.gov/cdd/?term). The amino acid sequences of the CUC2 gene in other plant species were acquired from GenBank (http://www.ncbi.nlm.nih.gov/genbank). Multiple sequence alignment of LcCUC2L and its homologous genes was performed with ClustalW (https://www.genome.jp/tools-bin/clustalw). The results were uploaded to ESPript 3.0 (http://espript.ibcp.fr/ESPript/cgi-bin/ESPript.cgi). A phylogenetic tree was constructed using MEGA 7.

After consulting the published L. chinense genome data (Chen et al., 2019), approximately 2 kb of genomic DNA upstream of the start codon of LcCUC2L was cloned. The PlantCARE (http://bioinformatics.psb.ugent.be/webtools/plantcare/html) was used to predict the cis-acting elements of ProLcCUC2L.

Subcellular localization of LcCUC2L

The whole coding sequence (CDS) of LcCUC2L without a stop codon was ligated to a linearized pBI121-eGFP vector. The recombinant vector was transformed into Agrobacterium tumefaciens GV3101 (ShiFeng). And the 35S::eGFP was used as the control. Positive clones and the P19 vector were cultured in Luria-Bertani (LB) medium at 28 °C. Next, the cells were collected and suspended in buffer containing 10 mM MgCl2 and 150 μM acetosyringone (AS). Two bacterial suspensions were mixed and injected into tobacco leaves (Sparkes et al., 2006). After 48 h of injection, the tobacco leaves were stained with the nuclear dye 4^′, 6-diamidino-2-phenylindole (DAPI) to visualize the nucleus. Subsequently, the fluorescent signals in the epidermis cells of tobacco leaves were examined with a confocal laser scanning microscope (LSM 710, Zeiss, Germany).

Plasmid construction and genetic transformation

The full-length ORF of LcCUC2L was cloned into the linearized overexpression vector pBI121. Similarly, the sequence of ProLcCUC2L was inserted at the HindIII/NcoI site of the pCAMBIA1301 vector by substituting the CaMV35S promoter, yielding the construct ProLcCUC2L::GUS. The Agrobacterium-mediated floral dip method was used for the genetic transformation of A. thaliana (Clough & Bent, 1998). Transgenic A. thaliana seeds were screened on $\frac{1}{2}$ MS medium containing antibiotics. Using leaf genomic DNA as templates, T1-positive transgenic plants were identified by PCR using specific primers. We acquired T3 transgenic plants through continuous screening for three generations.

Venation pattern analysis and scanning election microscopy

To compare cotyledon vein patterns of transgenic and WT A. thaliana plants, the cotyledons were bleached with 75% (v/v) ethanol and observed using stereomicroscopy. For cotyledon epidermal cell analysis, the cotyledons of transgenic and WT A. thaliana plants were fixed in FAA fixative solution (38% formalin, 50% ethanol, and glacial acetic acid). After critical point drying (CPD), the samples were coated with an Edwards E-1010 ion sputter golden coater and examined with a scanning electron microscope (FEI Quanta 200 FEG MKII).

Auxin content measurement

Basal rosette leaves samples of 0.6 g were harvested from 3-week-old WT and LcCUC2L transgenic A. thaliana plants and immediately frozen in liquid nitrogen. Determination of indole-3-acetic acid (IAA) in A. thaliana leaves was performed by ELISA (Huding, Shanghai, China). Three replicates were used for each sample.

Histochemical staining to detect GUS activity

The β-Galactosidase Reporter Gene Staining Kit (Leagene) was used for GUS histochemical staining in ProLcCUC2L transgenic and WT A. thaliana plants. In this study, seven-day-old seedlings and fourteen-day-old seedlings were processed. Moreover, leaves at different stages of development (10, 20 and 30 days of age) from plants planted in soil were subjected to GUS staining. All materials were treated with GUS staining solution and incubated overnight at 37 °C, followed by bleaching with 75% (v/v) ethanol.

Statistical analysis

Both the control and the experimental lines had three repetitions. The data are expressed as the mean ± standard deviation (SD). Statistical significance was determined with Student’s t-test. P < 0.05 was regarded as significant (*), and P ≤ 0.01 was considered highly significant (**).

Cloning and sequence analysis of LcCUC2L and ProLcCUC2L

The full length of the LcCUC2L gDNA sequence is 4,374 bp and consists of three exons and two introns (Fig. 1A). LcCUC2L (accession number: lcCUC2MW629054) contains a 1,017 bp ORF (Data S1), which encodes 338 amino acids and has a predicted molecular weight of 37.32 kDa and a theoretical PI of 8.45. The instability index was 44.46, indicating that the protein is unstable. The secondary structure of LcCUC2L was predicted to consist of helix (3.25%), loop (84.02%) and strand (12.72%) structures. Subcellular localization prediction showed that the LcCUC2L protein was mainly distributed in the cell nucleus. The amino acid sequences of the LcCUC2L protein in L. chinense and other plants were analyzed via multiple sequence alignment. Furthermore, the amino acid sequence of LcCUC2L was compared with the sequences of other CUC2 proteins in Arabidopsis lyrata subsp. lyrata, A. thaliana, Ananas comosus, Cardamine hirsute, Glycine max, and Zea mays. Similar to the N-terminus of other CUC2 proteins, the N-terminus of the LcCUC2L protein contains a conserved NAC domain, with five conserved subdomains (A, B, C, D, E), indicating that these proteins belong to the CUC subfamily of the NAC family. In addition, CUC genes harbor a variable CTD domain. The K motif (PSKKTKVPSTIS), V motif (EHVSCFS), L motif (SLPPL) and W-motif (WNY) can be observed in CUC2. In contrast, CUC1 does not have a K motif and there is only an L motif in CUC3 (Larsson et al., 2012). LcCUC2L harbors a variable CTD domain, and the K motif, L motif, V motif and W motif were observed in the LcCUC2L protein (Fig. 1B). The V motif of CUC1 and CUC2 has been identified as a miR164 recognition site (Rhoades et al., 2002). As shown in Fig. 1C, which indicated the LcCUC2L might be recognized by miR164. Moreover, phylogenetic tree analysis confirmed that CUC proteins could be classified into three clads: CUC1, CUC2, and CUC3. The LcCUC2L protein is most closely related to the Ananas comosus CUC2 protein (Fig. 1D). In conclusion, these results show that the sequence is highly homologous to CUC2, so we named it LcCUC2L.

To explore the function of LcCUC2L, a 2,028 bp (Data S2) sequence upstream of the translational initiation site (ATG) of the LcCUC2L gene, which contains the promoter sequence of LcCUC2L, was amplified using the DNA template of L. chinense. PlantCARE software was used to analyze the cis-acting elements of ProLcCUC2L. As shown in Table 1, some cis-acting elements are involved in light regulation, such as AE-box, ATCT-motif, Box4, G-Box, G-box, I-box, TCCC-motif, and chs-CMA1a. LTR and WUN motifs associated with low-temperature responsiveness and wound responsiveness, respectively, exist in the sequence. In addition, some hormone-related elements were identified, including the abscisic acid responsive regulatory motif (ABRE), auxin-responsive element (TGA-element) and MeJA-responsiveness elements (CGTCA-motif and TGACG-motif). The predicted results revealed that the core promoter elements of ProLcCUC2L contained TATA-box and CAAT-box.

Expression pattern analysis of LcCUC2L

The transcript levels of LcCUC2L were determined by RT-qPCR in leaves of different developmental stages (Fig. 2A). The highest expression levels of LcCUC2L were observed in the leaf bud, and the lowest abundance of LcCUC2L transcripts were in mature leaves (Fig. 2B, Table S2). The results indicated that LcCUC2L may play an important role in leaf bud development.

Based on cis-element sequence analysis of the LcCUC2L promoter, we predicted that LcCUC2L might be associated with auxin. To confirm the effect of auxin on LcCUC2L expression, we sprayed 200μM IAA on the leaves of L. chinense. The application of IAA significantly inhibited the expression of LcCUC2L (Fig. 2C, Table S3), which suggested that IAA represses LcCUC2L expression.

To further assess the expression pattern of LcCUC2L, ProLcCUC2L was fused to the GUS reporter and subsequently transferred into A. thaliana (hereafter ProLcCUC2L). After screening in media containing hygromycin and DNA detection, we obtained 8 transgenic lines (Fig. S1). Histochemical GUS staining was applied to detect the expression levels of the GUS gene in transgenic A. thaliana plants, which displayed inducible activity of ProLcCUC2L. In the 7-day-old and 14-day-old transgenic seedlings, GUS staining was mainly detected in the cotyledon lamina and roots (Fig. 2D). In contrast, low GUS activity was observed in the cotyledon petiole and hypocotyl of the transgenic seedlings. GUS activity was not detected in the true leaves of 14-day-old transgenic seedlings. These results indicated that LcCUC2L may have participated in cotyledon development in the transgenic A. thaliana seedlings.

Confocal images with DAPI nuclear staining (blue) were taken 48 h after transfection, showing GFP (green) expression that indicates the subcellular localization of LcCUC2L. Figure 2E shows that the green fluorescence of the control 35S::GFP was distributed throughout the cells. In contrast, the green fluorescence of the 35S::LcCUC2L-eGFP fusion protein was observed in the nucleus, suggesting that LcCUC2L localized to the nucleus. This result was fully consistent with the subcellular localization result of the AtCUC2 protein in A. thaliana (Taoka et al., 2004).

Overexpression of LcCUC2L regulates leaf development in A. thaliana

To investigate the functions of LcCUC2L, we obtained transgenic A. thaliana plants overexpressing LcCUC2L under the control of the CaMV 35S promoter. After screening on selective medium, seven positive transgenic plants were acquired. Subsequently, three high expression lines with the same phenotypes (OE 1, OE 3, and OE 5) were selected for phenotype analysis (hereafter 35S::LcCUC2L-OE1, 35S::LcCUC2L-OE3, and 35S::LcCUC2L-OE5) (Fig. 3F, Table S4). We found that true leaves began to emerge on or near the 7th day, as shown in Fig. 3A. The results revealed significant variation in the cotyledon phenotypes between WT and 35S::LcCUC2L plants. In contrast to WT cotyledons, the cotyledons of the 35S::LcCUC2L plants appeared as long strips and without petioles, and they had narrower blades compared to the WT cotyledons. Moreover, we compared vascular development between the WT and 35S::LcCUC2L cotyledons. Two kinds of veins exist in WT cotyledons: primary veins and subprime veins. Several subprime veins branch from the midvein and then unite to form areoles separated by veins (Figs. 3B, 3C) (Sieburth, 1999). In contrast, 35S::LcCUC2L plant cotyledons displayed an incomplete vascular pattern and had only a single central strand (Figs. 3B, 3C). In conclusion, the 35S::LcCUC2L cotyledons exhibited significant deficiencies in vascular development. To examine the cotyledon cell changes in the transgenic plants, we imaged the epidermal cells of 35S::LcCUC2L and WT with scanning election microscopy (SEM). Scanning election microscopy observations revealed that in WT cotyledons, the epidermal cells were organized in a specific pavement-like pattern (Fig. 3D) (Koyama et al., 2007). Remarkably, the shape and arrangement of the epidermal cells in the 35S::LcCUC2L cotyledons were similar to those of the WT petiole, with both being arranged regularly and exhibiting a rectangular shape. Taken together, the findings revealed large variations in cotyledon morphology between the 35S::LcCUC2L and WT A. thaliana plants, demonstrating that LcCUC2L affects cotyledon development in A. thaliana.

Ten days later, we transplanted seedlings of the homozygous transgenic lines and WT A. thaliana plants into nutrient soil. We observed that the rosette leaves of the 30-day-old 35S::LcCUC2L plants grew in clusters and were small and numerous (Figs. 3E, 3G). The 35S::LcCUC2L-OE1 and 35S::LcCUC2L-OE3 had 20 rosette leaves on average and the 35S::LcCUC2L-OE5 had 18 rosette leaves on average (Table S5), which was considerably more than the number in WT (Fig. 3H). These results indicated that LcCUC2L influences leaf development in A. thaliana.

LcCUC2L regulates leaf development by upregulating the expression of some genes related to auxin and leaf shape development

Many recent investigations have indicated that auxin plays a vital role in leaf development (Xiong & Jiao, 2019). The 35S::LcCUC2L plants amassed more IAA than the WT A. thaliana plants (Fig. 4A, Table S6). To explore whether LcCUC2L regulates leaf development by affecting auxin signaling, the expression levels of auxin biosynthesis genes and auxin transport genes were examined (Table S7). The transcript levels of the auxin biosynthetic genes YUCCA (AtYUC2 and AtYUC4) (Fig. 4B), an auxin influx carrier (AtAUX1), and efflux carriers (AtPIN1, AtPIN3, and AtPIN4) (Fig. 4C) were increased in LcCUC2L-expressing plants compared to WT A. thaliana plants. However, LcCUC2L overexpression had little effect on the expression level of AtYUC6. The results imply that the effects of LcCUC2L in leaf development might be enhanced by changes in auxin biosynthesis and polar transport. Additionally, we analyzed some genes related to leaf shape development, i.e., KNAT1, KNAT2, KNAT6 (for KNOTTED-like from A. thaliana), SHOOOTMERISTEMLESS (STM), and DPA4. We found that 35S::LcCUC2L plants presented much higher transcript levels of KNAT6, KNAT2, DPA4, and AtCUC2 than WT A. thaliana plants (Fig. 4D). The results indicate that the overexpression of LcCUC2L led to the upregulation of some genes related to leaf shape development and thus affected leaf development in A. thaliana.