Two Paralogous Genes Encoding Auxin Efflux Carrier Differentially Expressed in Bitter Gourd (Momordica charantia)

The phytohormone auxin regulates various developmental programs in plants, including cell growth, cell division and cell differentiation. The auxin efflux carriers are essential for the auxin transport. To show an involvement of auxin transporters in the coordination of fruit development in bitter gourd, a juicy fruit, we isolated novel cDNAs (referred as McPIN) encoding putative auxin efflux carriers, including McPIN1, McPIN2 (allele of McPIN1) and McPIN3, from developing fruits of bitter gourd. Both McPIN1 and McPIN3 genes possess six exons and five introns. Hydropathy analysis revealed that both polypeptides have two hydrophobic regions with five transmembrane segments and a predominantly hydrophilic core. Phylogenetic analyses revealed that McPIN1 shared the highest homology to the group of Arabidopsis, cucumber and tomato PIN1, while McPIN3 belonged to another group, including Arabidopsis and tomato PIN3 as well as PIN4. This suggests different roles for McPIN1 and McPIN3 in auxin transport involved in the fruit development of bitter gourd. Maximum mRNA levels for both genes were detected in staminate and pistillate flowers. McPIN1 is expressed in a particular period of early fruit development but McPIN3 continues to be expressed until the last stage of fruit ripening. Moreover, these two genes are auxin-inducible and qualified as early auxin-response genes. Their expression patterns suggest that these two auxin transporter genes play a pivotal role in fruit setting and development.


Introduction
Auxins are plant hormones that mediate many aspects of plant growth and development. Auxins stimulate cell enlargement, stem growth, cell division in the cambium, differentiation of phloem and xylem, root initiation and growth of flower parts. Furthermore, these hormones delay leaf senescence and are involved in apical dominance and leaf abscission. Auxin can induce fruit setting and growth in some fruits, assimilate partitioning and promote flowering in Bromeliads and female traits in dioecious flowers [1]. As early as 1926, Cholodny and Went proposed that the gravitropic curvature of plant organs was dependent upon lateral transport of the plant hormone indole-3-acetic acid (IAA)-auxin-from the upper to the lower side of a responding organ, resulting in different auxin concentration on the two sides of the organ. Different auxin levels result in different elongation rates and, thus, change the growth curvature [2].
As auxin is synthesized in young leaves of the shoot system and transported downward to the root tip, auxin moves in a basal-to-apical (acropetal) direction in the root [3]. The strongly polar process of auxin movement through the components of cells is thought to involve both passive diffusion and active transport. Therefore, the existence of a set of asymmetrically distributed molecules is required for the influx and efflux of auxin [1]. In this directional transport, auxin is mobilized by adenosine triphosphate-driven transporter (ATP Binding Cassette subfamily B; ABCBs), gradient-driven carriers (PIN-FORMED; PINs) and amino acid/auxin permease family of proton-driven symporters (AUXIN RESISTANT1/LIKE AUX1, AUX1/LAXes). There are at least 15 different carriers belonging to these three classes of auxin transporters in Arabidopsis [4,5].
In Arabidopsis, the PIN family consists of eight members and is divided into two subclades: full-length PINs (AtPIN1, 2, 3, 4 and 7) with polar localization on plasma membrane (PM) and short PINs (AtPIN5, 6 and 8) localized to endomembranes with a shorter central hydrophilic loop [6]. AtPIN1 acts as the auxin efflux carrier in Arabidopsis vascular tissue [7]. Re-localization of the AtPIN1 plays a role in phototropic responses [8]. Direct involvement of AtPIN1 in organogenesis and embryogenesis has also been demonstrated [9]. AtPIN2/EIR1/AGR1 mediates auxin reflux flows from the root apex [10]. AtPIN3 functions in the lateral redistribution of auxin and is present in the stem endodermal layer and in columella cells of the root cap [11]. Moreover, AtPIN3 functions during gravity response in roots and in hypocotyl by polarization [12]. Mobilization of auxin out of the root meristem is dependent on AtPIN3 and AtPIN4 [13]. AtPIN7 plays a role in forming and maintaining apical−basal auxin gradients that are essential for establishing embryonic polarity and functions in root acropetal auxin transport. This first presents in columella cells, and relocates during gravitropic stimulation [13,14]. AtPIN5 and AtPIN8 are localized to the endoplasmic reticulum (ER) and mediate auxin flow from the cytosol to the lumen of the ER to regulate subcellular homeostasis of auxin [6,15]. AtPIN6 at the PM and ER mediates both transportation of auxin across the PM and intracellular homeostasis of auxin during primary, lateral and adventitious root organogenesis and other developmental processes, such as shoot apical dominance [16,17].
Very little information is known about the molecular basis of auxin transport during fruit set and development until recently. Ten tomato PIN (SlPIN1 to 10) genes were identified and their expression patterns are heterogeneous [18]. The expression of tomato auxin efflux carrier genes SlPIN1 (referred as SlPIN9 in [19]) and SlPIN2 (referred as SlPIN4 in [19]) was highest in the very young fruit 2 and 4 days after anthesis, respectively. This suggests that these two proteins play important roles in the fruit setting. Furthermore, the expression of the SlPIN2 promoter in transgenic tomatoes was restricted to ovules and young seeds 0 and 4 days after anthesis, which suggests that SlPIN2 play roles in exporting auxin synthesized in young seeds. SlPIN1, which had relatively high expression in the whole early tomato fruit and declined to lower levels by 15 days after anthesis, might be responsible for the subsequent transport of auxin into peduncles [20]. Eight SlPIN genes were expressed in fruit and could be divided into three classes: SlPIN5 in seeds/locular tissue; SlPIN1, SlPIN4, SlPIN7 and SlPIN8 in placenta; and SlPIN6 in pericarp [18]. Specific silencing of SlPIN4, which is much more highly expressed in ovary and fruit than any other SlPIN genes, leads to parthenocarpic fruits. This result indicates that SlPIN4 plays a major role in the auxin regulation of tomato fruit set [19].
To study the relationship between auxin polar transport and fruit development in Cucurbitaceae, we isolated cDNAs encoding the auxin efflux carrier homologue using AtPIN1 and AtPIN2 gene fragments as probes for screening. Here, we report a primary feature of putative auxin efflux carrier genes from bitter gourd and their gene expression patterns during bitter gourd fruit development.

Isolation and Characterization of Auxin Efflux Carrier cDNAs from Bitter Gourd
To isolate the fruit-related auxin efflux carrier genes, the cDNA library constructed by poly(A) + RNA extracted from fruit of bitter gourd was screened using Arabidopsis AtPIN1 and AtPIN2 fragments as probes. Among the cDNA clones obtained, pMAEC28, pMAEC43 and pMAEC93 were selected for complete nucleotide sequence analysis. Properties of these three cDNAs and deduced polypeptides are shown in Table 1. The cDNA in pMAEC28 is probably derived from an alternative splicing product, which possesses a 432 bp intron ( Figure S1). The length of the 5 -untranslated region (5 -UTR) in pMAEC43 cDNA is 222 bp shorter than that in pMAEC28. After neglecting the intron and the longer 5 -UTR, pMAEC28 cDNA shares 99.7% identity of the nucleotide sequence with pMAEC43 cDNA. Therefore, pMAEC43 cDNA was used as the probe for Southern and Northern blot analysis. The encoded polypeptide by pMAEC93 cDNA possesses 27 more residues than pMAEC43 cDNA does. Hydropathy analysis revealed that both these polypeptides have two hydrophobic regions with five N-terminal (residues 1-151) and five C-terminal (residues 461-605) transmembrane segments as well as a predominantly hydrophilic core extending from residues 152-460 ( Figure 1). Furthermore, amino acid sequence analysis using the program SignalP (V3.0) indicates the possibility of a signal peptide spanning from residues 1-30. The cleavage site of a potential amino-terminal signal peptide is between Trp 30 and Trp 31 Figure 2).

Genome Organization of Auxin Efflux Carrier Genes in Bitter Gourd
Both cDNAs in pMAEC43 and pMAEC93 were used as gene-specific probes to perform Southern blot analyses. According to the results of Southern blot analyses, there was no BamHI and

Genome Organization of Auxin Efflux Carrier Genes in Bitter Gourd
Both cDNAs in pMAEC43 and pMAEC93 were used as gene-specific probes to perform Southern blot analyses. According to the results of Southern blot analyses, there was no BamHI and SalI cutting sites in the gene body of McPIN1 and McPIN3. These two genes are located in different fragments after being digested with SalI, SalI and BamHI, and BamHI ( Figure

Genome Organization of Auxin Efflux Carrier Genes in Bitter Gourd
Both cDNAs in pMAEC43 and pMAEC93 were used as gene-specific probes to perform Southern blot analyses. According to the results of Southern blot analyses, there was no BamHI and

Gene Expression of Auxin Efflux Carrier during Fruit Ripening in Bitter Gourd
According to the results of Northern blot analysis, 2.3 kb of mRNA of McPIN1 was abundant in flowers. There were smaller amounts in roots and stem, while it was not detectable in young or mature leaves. On the contrary, the mRNA of McPIN3 that was 2.5 kb in length was detected in all tissues of bitter gourd, including roots, stems, leaves and flowers ( Figure 4A). Maximum McPIN1 mRNA accumulated in both staminate and pistillate flower, while maximum McPIN3 mRNA was found in the staminate flower ( Figure 4B). To understand the expression pattern of McPIN1 and McPIN3 genes in fruit development and ripening, the mRNA of bitter gourd harvested at different days after pollination was extracted to perform Northern blot analysis. It was classified into four stages: the first stage being 6 days after pollination; the second stage being 12 days after pollination; the third stage being 18 days after pollination; and the last stage being 24 days after pollination. The growth of fruit is halted and the ripening starts at stage 3. Accumulation of mRNA for both McPIN1 and McPIN3 genes decreased throughout the fruit development. McPIN1 mRNA was detectable only in the early stages (stage 1 and stage 2) of fruit development ( Figure 5A), whereas the hybridization signal corresponding to the expression of McPIN3 was still detectable at stage 4 ( Figure 5B).

Gene Expression of Auxin Efflux Carrier during Fruit Ripening in Bitter Gourd
According to the results of Northern blot analysis, 2.3 kb of mRNA of McPIN1 was abundant in flowers. There were smaller amounts in roots and stem, while it was not detectable in young or mature leaves. On the contrary, the mRNA of McPIN3 that was 2.5 kb in length was detected in all tissues of bitter gourd, including roots, stems, leaves and flowers ( Figure 4A). Maximum McPIN1 mRNA accumulated in both staminate and pistillate flower, while maximum McPIN3 mRNA was found in the staminate flower ( Figure 4B). To understand the expression pattern of McPIN1 and McPIN3 genes in fruit development and ripening, the mRNA of bitter gourd harvested at different days after pollination was extracted to perform Northern blot analysis. It was classified into four stages: the first stage being 6 days after pollination; the second stage being 12 days after pollination; the third stage being 18 days after pollination; and the last stage being 24 days after pollination. The growth of fruit is halted and the ripening starts at stage 3. Accumulation of mRNA for both  were quantified by using LabWorks bioimaging system. The expression level in each treatment was normalized to the 26S and 18S loading control; the maximum level was set to 100%, and all other organ points were normalized to it. Relative expression was determined in triplicate measurements in three independent biological replicates. Error bars represent the standard deviation.  were quantified by using LabWorks bioimaging system. The expression level in each treatment was normalized to the 26S and 18S loading control; the maximum level was set to 100%, and all other organ points were normalized to it. Relative expression was determined in triplicate measurements in three independent biological replicates. Error bars represent the standard deviation.   (B) Hybridization signals were quantified by using LabWorks bioimaging system. The expression level in each treatment was normalized to the 26S and 18S loading control; the level at 0 h was set to 100%, and all other time points were normalized to it; (C) Gene expression in response to different concentrations of IAA incubated for 0.5 h in bitter gourd. The expression level in each treatment was normalized to the 26S and 18S loading control; the level at control (CK) was set to 100%, and all other concentration points were normalized to it. Relative expression was determined in triplicate measurements in three independent biological replicates. Error bars represent the standard deviation.

Effects of Exogenous Auxin on Promoter Activity of Auxin Efflux Carrier Genes from Bitter Gourd in Transgenic Tobacco
β-Glucuronidase (GUS) activity in both transgenic tobacco seedlings was detected only in leaf veins and were stronger in McPIN1 pro ::GUS compared to that in McPIN3 pro ::GUS plants. The promoter activity was induced by 1-naphthaleneacetic acid (NAA) and indole-3-butyric acid (IBA) in the mesophyll of newborn leaves. This was enhanced by 20 µM of NAA and 100 µM of 2,4-dichlorophenoxyacetic acid (2,4-D) in the roots of McPIN1 pro ::GUS transgenic tobacco. All auxins treated in this study had a slightly suppressive effect on McPIN3 promoter activity (Figure 7).

Discussion
This is the first report on the expression patterns of auxin efflux carrier PIN-like genes in monoecious Cucurbitaceae. The gene expression of CsPIN1 from cucumber was only investigated in roots and it appears to be involved in peg formation [24]. Previous studies did not examine the gene expression of AtPIN1 or other AtPIN genes from Arabidopsis in the capsule. Expression patterns of ten tomato SlPIN genes were investigated in flower buds and developing fruits [18,19] since the tomato

Discussion
This is the first report on the expression patterns of auxin efflux carrier PIN-like genes in monoecious Cucurbitaceae. The gene expression of CsPIN1 from cucumber was only investigated in roots and it appears to be involved in peg formation [24]. Previous studies did not examine the gene expression of AtPIN1 or other AtPIN genes from Arabidopsis in the capsule. Expression patterns of ten tomato SlPIN genes were investigated in flower buds and developing fruits [18,19] since the tomato is a hermaphrodite plant similar to Arabidopsis. Here, we characterized that McPIN genes from bitter gourd belong to a gene family. A similar situation occurred for the genome of other plants, such as eight AtPINs for Arabidopsis [9], at least two genes for pea [25], five for potato, six for soybean, eight for wheat, nine for rice [9], ten for Medicago truncatula [26], and ten for tomato [18]. All the deduced amino acid sequences of McPIN1, McPIN2 and McPIN3 exhibited conserved regions in 5 -and 3 -end corresponding to the hydrophobic regions containing transmembrane domains similar to the other auxin efflux carrier proteins, such as CsPIN1−6 ( Figure 1). Short PIN genes might not express or express traces in developing fruits of bitter gourd so we did not obtain any related cDNA clones. The variability in the length of AtPIN proteins was due to differences in the amino acids in the central region, such as AtPIN5, AtPIN6 and AtPIN8, which lack a central loop domain [9]. Additionally, while compared to the putative amino-terminal signal peptide in AtPIN2, which is presumably cleaved at the end of the second putative transmembrane domain between Ser 60 and Ser 61 [10], the potential cleavage site in McPINs is more typical because it is located between the first and the second predicted transmembrane domains (Figure 1). According to the phylogenetic analysis, McPIN1 and McPIN3 share a relatively high sequence similarity with SlPIN1 and SlPIN3/4 [18][19][20], respectively ( Figure 2 and Table 2). It indicates that these two auxin efflux carrier proteins in bitter gourd might play similar roles in the fruit development as SlPINs. Furthermore, there are different roles for McPIN1 and McPIN3 in auxin transport involved in the fruit development of bitter gourd. There is no doubt that the closest orthologous relationships occur between the McPINs from bitter gourd and CsPINs from cucumber (Figure 2), which both belong to the monoecious Cucurbitaceae. The numbers of exons and introns in McPIN1 and McPIN3 (Figure 3) are the same as in AtPIN3 and AtPIN4 [11]. There are five exons in AtPIN1 [7] and nine exons in AtPIN2 [27].
The expression level of McPIN1 is much higher in staminate and pistillate flowers than in roots and stems, while McPIN1 mRNA can barely be detected in young or old leaves. However, the expression pattern of McPIN3 (Figure 4) is similar to that of AtPIN1, in which mRNA was detected in all Arabidopsis organs tested, including cotyledons, flowers, roots, rosette leaves, seedlings, inflorescence axes and siliques [7]. Tomato SlPIN4 was found to be expressed abundantly in sepal, petal, ovary and young leaves, while it was expressed slightly in mature leaves, roots, and stems. Furthermore, the expression of SlPIN4 increased gradually during the flower development to the flowering stage [19]. The accumulation of SlPIN1 and SlPIN9 mRNA was greatest in the ovary, while that of SlPIN3 and SlPIN7 was greatest in the petals. The expression levels of the other SlPINs were one tenth that of SlPIN4 [19]. The gene expression pattern of McPIN3 in different organs of bitter gourd is similar to that of SlPIN4 in tomatoes.
Both McPIN1 and McPIN3 expression were enhanced by exogenous auxin in the fruit of bitter gourd 30 min after treatment ( Figure 6). Auxin-mediated cell elongation, one of the fastest hormonal responses known (with a lag period of 15-25 min), is associated with rapid changes in the expression of a select set of early genes [28,29], such as PS-IAA4/5 and PS-IAA6 from peas [30]. Therefore, McPIN1 and McPIN3 genes qualify as the early auxin-response genes [31]. It suggests that the regulatory functions of McPIN1 and McPIN3 in auxin-induced events are responsible for auxin transportation and cell growth in fruit. On the other hand, only the expression of McPIN1 promoter was enhanced in the roots of transgenic tobacco (Figure 7). In Arabidopsis, the expression pattern of AtPIN2 in seedlings revealed a root specificity, which correlated well with the root-specific alterations detected in agr1/eir1/pin2 mutants [10,32]. No AtPIN2 transcripts were detected in other parts of seedlings or mature plants, although they could be found in hypocotyl and cotyledon tissues of etiolated seedlings at a level 10−fold lower than in roots [32]. Gene expression of MtPIN2 was limited to nodulated roots, while transcripts of all other expressed genes were detected in both shoots and roots of Medicago truncatula [26]. Tomato SlPINs were also found to be expressed in roots, although this was at lower levels [18][19][20]. Different group McPIN1 classified by the amino acid sequence-based dendrogram ( Figure 2) and the expression patterns (Figures 4 and 7) of AtPIN2 indicated that another McPIN protein is responsible for auxin transport in roots.
The differences of gene expression between McPIN1 and McPIN3 are significant during fruit development and ripening. McPIN1 was expressed in a particular period of early fruit development but McPIN3 continued to be expressed until the last stage of fruit ripening ( Figure 5). The gene expression of SlPIN4 also decreased during the fruit development in tomatoes [19]. In the fruit of tomatoes, the expression level of SlPIN4 mRNA was higher in the central part of the fruit (columella, locular tissue) than the outer part of the fruit (exocarp) [18,19]. Moreover, silencing of SlPIN4 in tomato enhanced precocious fruit development and resulted in parthenocarpic fruits [19]. As we know, auxins stimulate ovule growth and fruit setting is initiated by successful pollination. Specific inhibitors for polar auxin transport, such as 2,3,4-triiodobenzoic acid (TIBA), resulted in accumulation of auxin in the cucumber ovary and production of seedless parthenocarpy fruit without pollination [33]. According to the similar amino acid sequences and expression pattern to SlPIN4, McPIN3 could play the same role as SlPIN4 in fruit setting. Since no parthenocarpic fruits have been found in bitter gourd, it would be useful to further investigate the application of exogenous auxin to the unpollinated ovary of bitter gourd or gene silencing that is specific to McPIN1 or McPIN3.

Construction and Screening of cDNA Library and Genomic Library
Poly(A) + RNA was isolated from bitter gourd fruit 16 days after pollination. Double-stranded cDNAs were prepared according to the method described by Kimmel and Berger [36] and inserted into the EcoRI and XhoI sites of λZAPII using the cDNA Synthesis Kit supplied by Stratagene (La Jolla, CA, USA), which was used to construct the cDNA library. Approximately 1.5 × 10 6 amplified plaques were screened with AtPIN1 (517-967 of AF089085) and AtPIN2 (651-1531 of AF086907) gene fragments synthesized from using Arabidopsis as a probe. Hybridization was performed as described in Sambrook et al. [37]. Rescue for the putative clones was accomplished by in vivo excision according to the manufacturer's protocol (Stratagene) to produce derivatives as pBluescript SK(−) plasmids.
For construction of the genomic library, genomic DNA was extracted from leaves of bitter gourd as described by Jofuku and Goldberg [38]. After partial digestion by Sau3AI, DNA fragments of 15-25 kb were collected and inserted into BamHI sites of the λEMBL3 replacement lambda phage vector (Stratagene, La Jolla, CA, USA). Hybridization was performed as described in Sambrook et al. [37] using the pMEC43 and pMAEC93 cDNA as probes.

Nucleotide and Amino Acid Sequence Analysis
Sequencing was performed using an ABI PRISM 377 DNA Sequencer according to the manufacturer's instructions. Amino acid sequences were aligned using the AlignX program of Vector NTI Suite 10.0 Software (Thermo Fisher Scientific Inc., Waltham, MA, USA). The cleavage site of a potential signal peptide was analyzed by SignalP 3.0 [21]. Polypeptide topology prediction has been performed using the HMMTOP-Server [39].

Northern Blot Analysis
Total RNA (20 µg) was denatured by glyoxylation [42] and electrophoresed through a 1% (w/v) agarose gel. Transfer of RNA onto a nylon membrane (Micron Separation Inc.) and subsequent hybridization were carried out by conventional methods, as instructed by the manufacturer. The amount of ribosomal RNA was used as a loading control. Relative expression was determined in triplicate measurements in three independent biological replicates.

Promoter Activity Analysis
Wild-type tobacco (Nicotiana tabacum L. cv. Wisconsin 38) was transformed to generate McPIN1 pro ::GUS and McPIN3 pro ::GUS transgenic tobacco. Homozygous transgenic seeds of tobacco (R2) were germinated on the selective media (half-strength Murashige and Skoog (1/2 MS) media [43] containing 100 mL L −1 of kanamycin). Transgenic tobacco seedlings were grown in a culture room with cycles of 16-h light and 8-h darkness at 25 ± 2 • C. Four-week-old seedlings were treated with IAA, NAA, IBA, and 2,4-D at 20 or 100 µM for 24 h and analyzed by GUS staining. GUS histochemical staining was performed as described by Jefferson et al. [44].

Conclusions
Auxin exerts a major influence on transcriptional regulation in plants and other organisms. Auxin efflux carriers are thought to provide an adaptive advantage by allowing auxin transport to tissues where auxin acts to regulate physiological processes. Our approach to understanding the functions of auxin action is to study the regulation of carrier protein genes and then work backward toward the central auxin function. In this study, we investigated the gene expression patterns of the auxin efflux carrier genes during fruit development of bitter gourd. We found that the auxin efflux carrier genes might be involved in the regulation of fruit development. This finding provides novel insights into roles of auxin efflux carriers and a better understanding of fruit growth and the development of juicy melon at the molecular level. Since the function of SlPIN4 gene has been elucidated previously [19], gene expression during monoecious flower development, protein localization in juicy fruit and gene function of McPINs will be further investigated.