Expression Analysis of an R3-Type MYB Transcription Factor CPC-LIKE MYB4 (TRICHOMELESS2) and CPL4-Related Transcripts in Arabidopsis

The CAPRICE (CPC)-like MYB gene family encodes R3-type MYB transcription factors in Arabidopsis. There are six additional CPC-like MYB sequences in the Arabidopsis genome, including TRYPTICHON (TRY), ENHANCER OF TRY AND CPC1 and 2 (ETC1 and ETC2), ENHANCER OF TRY AND CPC3/CPC-LIKE MYB3 (ETC3/CPL3), and TRICHOMELESS1 and 2 (TCL1 and TCL2). We independently identified CPC-LIKE MYB4 (CPL4), which was found to be identical to TCL2. RT-PCR analysis showed that CPL4 is strongly expressed in shoots, including true leaves, but not in roots. Promoter-GUS analyses indicated that CPL4 is specifically expressed in leaf blades. Although CPC expression was repressed in 35S::ETC1, 35S::ETC2 and 35S::CPL3 backgrounds, CPL4 expression was not affected by ETC1, ETC2 or CPL3 over-expression. Notably, several chimeric transcripts may result from inter-genic alternative splicing of CPL4 and ETC2, two tandemly repeated genes on chromosome II. At least two chimeric transcripts named CPL4-α and CPL4-β are expected to encode complete CPC-like MYB proteins.


Introduction
Epidermal cell differentiation in Arabidopsis, including root-hair and trichome cell formation, has been used as a model system to analyze plant cell fate determination. Several regulatory factors are known to be involved in epidermal cell differentiation events. The glabra 2 (gl2) and werewolf (wer) mutants induce an increased number of root-hair cells [1,2]. The GL2 gene encodes a homeodomain leucine-zipper protein, and the WER gene encodes an R2R3-type MYB transcription factor that activates GL2 expression [1][2][3][4]. The GL1 and MYB23 genes encode R2R3-type MYB genes that are closely related to WER and are also involved in epidermal cell fate determination in Arabidopsis. The GL1 gene promotes trichome formation, and the gl1 mutant phenotype shows a reduced number of leaf trichomes [5]. GL1 is expressed in developing trichomes [6]. Constitutive expression of the MYB23 gene under the control of the CaMV 35S promoter induces ectopic trichome formation [7]. MYB23 and WER are preferentially expressed in non-hair cells in Arabidopsis roots [8,9]. GLABRA3 (GL3) and ENHANCER OF GLABRA3 (EGL3) encode basic helix-loop-helix (bHLH) transcription factors that affect non-hair cell differentiation in a redundant manner [10]. There are two other bHLH genes, AtMYC1 [11] and TRANSPARENT TESTA8 (TT8) [12], that are in the same gene family as GL3 and EGL3 [13]. The TRANSPARENT TESTA GLABRA1 (TTG1) gene encodes a WD40-repeat protein that regulates non-hair cell formation and trichome differentiation [14,15]. Using a yeast two-hybrid system, GL3 and EGL3 were shown to interact with WER [10] and with a WD40 protein (TTG1) [16][17][18].
The CAPRICE (CPC) gene encodes an R3-type small MYB protein, and the cpc mutation strongly reduces the formation of root hairs [19]. Previously, we reported that the CPC protein moves from non-hair cells to root-hair cells and represses GL2 expression [20,21]. In addition, we proposed a model in which the CPC gene could have arisen by evolution from the WER gene [22]. A protein complex including WER, GL3/EGL3 and TTG1 induces GL2 expression [2,3,14,19,23]. The CPC protein disrupts this protein complex by competitively binding with WER, leading to repression of GL2 expression [24,25]. Arabidopsis has several additional CPC-like MYB sequences in its genome, including TRYPTICHON (TRY), ENHANCER OF TRY AND CPC1 and 2 (ETC1 and ETC2), ENHANCER OF TRY AND CPC3/CPC-LIKE MYB3 (ETC3/CPL3), and TRICHOMELESS1 and 2 (TCL1 and TCL2) [26][27][28][29][30][31][32][33]. Trichome cluster formation on leaves of the try mutant indicates that the TRY protein has a regulatory role in trichome differentiation [26,34]. ETC1 and ETC2 have been identified and their function with CPC and TRY genetically examined [27][28][29]. TCL1 and TCL2 negatively regulate trichome formation on the inflorescence stems and pedicels [32,33]. We have recently identified a seventh CPC-like MYB gene between At2g30430 and ETC2 (At2g30420) independently of Gan et al. [33], and have named it CPC-LIKE MYB4 (CPL4). In this paper, we examine the expression of the CPL4 gene in Arabidopsis. Notably, between CPL4 and ETC2, there were several chimeric transcripts generated through alternative splicing. Around 14% of the protein coding genes of Arabidopsis are annotated as producing multiple transcript variants through alternative splicing [35]. Our study proposes that inter-genic alterative splicing also characterizes the CPC-like MYB gene family.
We did not detect CPL4 expression in Arabidopsis seedling roots (Figure 1c). Through this assay, we confirmed the expression of a seventh CPC-like MYB gene that had not been previously recognized.

Promoter-GUS Analysis
To analyze CPL4 expression at the tissue level, we made CPL4 promoter-GUS fusions. CPL4p::GUS was expressed in young true leaves and cotyledons of 7-day-old seedlings ( Figure 2a). CPL4p::GUS expression was also observed in two-week-old rosette leaves (Figure 2b). Previously, we showed that the expression patterns of CPC-like MYB genes could be roughly classified into two groups [31]. CPCp::GUS, TRYp::GUS and ETC1p::GUS are expressed mainly in roots and trichomes, and ETC2p::GUS and CPL3p::GUS are expressed in young leaves and mainly in guard cells [31]. Thus, GUS expression by the CPC-like MYB family is found in tissues throughout the entire plant body. We did not detect    In addition to the spatial expression patterns, CPC expression was regulated by the CPC MYB protein itself as shown in 35S::CPC transgenic lines [24]. The mechanism for this repression may be by negative feedback, thereby contributing to the root-hair and non-hair cell differences. Because CPC protein shares high amino acid sequence homology with the other CPC-like MYBs, we checked the effect of ETC1, ETC2 or CPL3 overexpression on CPC expression ( Figure 3). Semi-quantitative RT-PCR analyses showed that CPC expression was clearly repressed in the 35S::ETC1, 35S::ETC2 and 35S::CPL3 backgrounds ( Figure 3). This result suggests the existence of a similar feedback loop resulting in the proper trichome distribution on leaves. To compare the regulation of gene expression, we also performed RT-PCR analyses using CPL4-specific primers. In plants harboring 35S::ETC1, 35S::ETC2 and 35S::CPL3, CPL4 was expressed at almost the same level as wild-type (Col-0) ( Figure 3). These results suggest that CPL4 is not involved in a feedback regulatory cascade controlled by CPC-like MYBs. The CPC-like MYB family members are thought to have evolved by gene duplication [31] and have diverged to different regulatory functions in the course of evolution. The CPC-like genes retain some functional redundancy that may represent intermediate stages of regulatory specification [37,38]. Thus, CPL4 may have acquired specific functions different from that of CPC.   (Figures 1a and 4a). Both genes are composed of three exons (Figures 1a and  4a). As shown in Figure 1b, CPL4 shares high amino acid sequence homology with ETC2. Therefore, the forward primer beginning with the predicted start codon of ETC2 can also act as the forward primer beginning with the start codon of CPL4. Therefore, it was necessary to carefully design primer pairs specific for ETC2 and CPL4. Gene-specific RT-PCR was performed for ETC2 and CPL4 using rosette leaves from 12-day-old Arabidopsis seedlings. Both primer pairs designed for ETC2 (RT124/RT125) and for CPL4 (RT341/RT342) amplified specific, single gene products of the expected sizes (320 bp and 240 bp) (Figure 4b). Using the CPL4-specific forward primer (RT341) and ETC2-specific reverse primer (RT125) as a primer pair, two DNA fragments named CPL4-related Chimera 1 and Chimera 2 were amplified (Figure 4c). An abundantly expressed band (Chimera 1) was estimated to be approximately 450 bp, and a less expressed band (Chimera 2) was approximately 280 bp in size (Figure 4c). To confirm the existence of CPL4-related chimeric gene expression, we repeated the RT-PCR experiment using new RNA samples and another primer pair (RT341/RT318). As a result, the existence of Chimera 1 and 2 cDNA products was confirmed.    To determine the precise structure of CPL4-related chimeric transcripts (Chimera 1 and 2), we cloned these RT-PCR products and subsequently sequenced the constructs. Chimera 1 included two different amplicons, named Chimera 1-1 and Chimera 1-2 (Figures 4c and 5b). Chimera 2 also included two different amplicons, named Chimera 2-1 and Chimera 2-2 (Figures 4c and 5b). As shown in Figure 5a, CPL4 and ETC2 lie in a tandem orientation (Figure 5a). All chimeric transcripts may result from alternative splicing of CPL4 and ETC2. Chimera 1-1 and Chimera 1-2 contained the first two exons of CPL4 and all three exons of ETC2 (1, 2, 4, 5 and 6 exons) (Figure 5b). Both chimeric transcripts (Chimera 1-1 and Chimera 1-2) contain a TAGTT additional sequence between the second exon of CPL4 and the first exon of ETC2. This linker sequence "TAGTT" was just upstream of the 5′-UTR region of the first exon of ETC2. Although Chimera 1-1 included the first intron of ETC2, Chimera 1-2 did not (Figure 5b). Chimera 2-1 contained the first and second exons of CPL4 and the third exon of ETC2 (1, 2 and 6 exons) (Figure 5b). The "TAGTT" linker insertion of Chimera 1-1 and 1-2 was produced by GT-AG splicing between the second intron of CPL4 and an upstream region of the first exon of ETC2. This linker sequence included a "TAG" sequence that could function as a stop codon. Therefore, Chimera 1-1 and 1-2 are expected to produce a truncated MYB protein that includes only the first two exons of CPL4. On the other hand, transcripts of Chimera 2-1 and 2-2 are expected to encode complete CPC-like MYB proteins. Chimera 2-1 contained the first two exons of CPL4 and the third exon of ETC2 (exons 1, 2 and 6) (Figure 5b). Chimera 2-2 contained the first CPL4 exon, the first 16 bp of the second exon of CPL4, the second exon of ETC2 lacking the first 16 bp, and the third exon of ETC2 (exons 1, 2, 5 and 6) ( Figure 5b). Thus, we renamed Chimera 2-1 and 2-2 to CPL4-α and CPL4-β, respectively (Figure 6a). CPL4-α shares 90% identity with ETC2 and 83% identity with CPL4 at the nucleotide level. CPL4-β shares 89% identity with ETC2 and 84% identity with CPL4 at the nucleotide level (Figure 6b). CPL4-α shares 86% identity with ETC2 and 76% identity with CPL4 at the amino acid level. CPL4-β shares 82% identity with ETC2 and 80% identity with CPL4 at the amino acid level (Figure 6c). Both CPL4-α and CPL4-β contain the conserved amino acid signature [D/E]Lx2[R/K]x3Lx6Lx3R that is required for the interaction with R/B-like bHLH transcription factors [39]. Their results strongly suggest that CPL4-α and CPL4-β serve as CPC-like MYB proteins harboring similar functions of ETC2 and/or CPL4.

MYB
Our study raises the possibility that there are unexpected alternative splicing sites spanning two homologous genes. In Arabidopsis, there are many examples of genes where alternative splicing provides a regulatory mechanism controlling aspects of development and other processes including flowering [40]. Although many genes produce alternatively spliced transcripts, the critical role of alternative splicing is poorly understood. This work provides additional evidence for the function of alternative splicing.

Plant Materials and Growth Conditions
Arabidopsis thaliana ecotype Columbia (Col-0) was used as the wild type in this study. Seeds were surface-sterilized and sown on 1.5% agar plates as described previously [41]. Construction of the 35S::CPL3 transgenic line was described previously [31]. Seeded plants were kept at 4 °C for 2 days and then incubated at 22 °C under constant white light (50-100 μmol· m −2 · s −1 ).

Gene Constructs
Sequences of all primers used in this paper are listed in Table 1. All PCR-generated constructs were completely sequenced following isolation of the clones to check for amplification-induced errors. Table 1. Primer sequences used in this study.

Transgenic Plants
Plant transformation was performed by a floral dip method [43], and transformants were selected on 0.5 × MS agar plates containing 50 mg/L kanamycin. Homozygous transgenic lines were selected for kanamycin resistance. At least twelve T1 lines were isolated for each construct and at least six T2 and three T3 lines were selected on the basis of their segregation ratios for kanamycin resistance.

RNA Isolation and Semi-Quantitative RT-PCR
Total RNA was extracted using RNeasy Plant Mini Kits (Qiagen, Valencia, CA, USA). On-column DNase I digestion was performed during RNA purification following the protocol described in the RNeasy Mini Kit handbook. First-strand cDNA was synthesized from 1 µg total RNA in a 20 µL reaction mixture using the Prime Script RT Regent Kit (Takara).
Semi-quantitative RT-PCR reactions were conducted as described by Kurata et al. [44]. The CPL4 fragment was amplified with RT341/RT342 primer pairs. The CPC fragment was amplified with RT128/RT129 primer pairs. The ETC2 fragment was amplified with RT124/RT125 primer pairs. EF was amplified with the EF1α-F/EF1α-R primer pair as described by Kurata et al. [21]. To determine the sequence of the PCR products, we cloned them into the pT7 blue T-vector (Novagen) and subsequently sequenced the constructs.

Microscopy
Root GUS activity was observed using an Olympus Provis AX70 microscope. At least five individual primary roots of 7-day-and two-week-old seedlings were analyzed for root GUS activity. To observe leaf GUS activity, an Olympus SZH binocular microscope was used. At least five cotyledons of 7-day-old seedlings and two-week-old rosette leaves were analyzed for leaf GUS activity.

Conclusions
In this paper, we investigated the expression of CPL4 and CPL4-related chimeric transcripts in Arabidopsis using promoter-GUS and RT-PCR analyses. The results showed that CPL4 is specifically expressed in leaf blades, and CPL4 expression was not affected by ETC1, ETC2 or CPL3 overexpression. We hypothesize that the proteins created by inter-genic alternative splicing between CPL4 and ETC2 act in a redundant manner with other CPC-like MYB family proteins. Much work remains to determine the function of CPC-like MYB gene family members. By overexpressing 35S::CPL4-α and 35S::CPL4-β and investigating the localization of CPL4::CPL4-α:GFP and CPL4::CPL4-β:GFP in transgenic plants, we should be in a better position to elucidate the roles of these CPC-like MYB genes in Arabidopsis epidermal cell differentiation.