|Home | About | Journals | Submit | Contact Us | Français|
Because of its rapid growth rate, relative ease of transformation, sequenced genome and low gene number relative to Arabidopsis, the tropical fruit tree, Carica papaya, can serve as a complementary genetic model for complex traits. Here, new phenotypes and touch-regulated gene homologues have been identified that can be used to advance the understanding of thigmomorphogenesis, a multigenic response involving mechanoreception and morphological change.
Morphological alterations were quantified, and microscopy of tissue was conducted. Assays for hypocotyl anthocyanins, lignin and chlorophyll were performed, and predicted genes from C. papaya were compared with Arabidopsis touch-inducible (TCH) and Mechanosensitive channel of Small conductance-like genes (MscS-like or MSL). In addition, the expression of two papaya TCH1 homologues was characterized.
On the abaxial side of petioles, treated plants were found to have novel, hypertrophic outgrowths associated with periderm and suberin. Touched plants also had higher lignin, dramatically less hypocotyl anthocyanins and chlorophyll, increased hypocotyl diameter, and decreased leaf width, stem length and root fresh weight. Papaya was found to have fewer MSL genes than Arabidopsis, and four touch-regulated genes in Arabidopsis had no counterparts in papaya. Water-spray treatment was found to enhance the expression of two papaya TCH1 homologues whereas induction following touch was only slightly correlated.
The novel petiole outgrowths caused by non-wounding, mechanical perturbation may be the result of hardening mechanisms, including added lignin, providing resistance against petiole movement. Inhibition of anthocyanin accumulation following touch, a new phenotypic association, may be caused by diversion of p-coumaroyl CoA away from chalcone synthase for lignin synthesis. The absence of MSL and touch-gene homologues indicates that papaya may have a smaller set of touch-regulated genes. The genes and novel touch-regulated phenotypes identified here will contribute to a more comprehensive view of thigmomorphogenesis in plants.
Plants respond to diverse mechanical perturbations such as developmental weight alterations, wind and rain, or insect and herbivore activity. The way in which plants respond defines particular touch-regulated reactions (reviewed by Jaffe et al., 2002; Braam, 2005). The closure of the bi-lobed leaves of the Venus flytrap (Dionaea muscipula) and the folding of the leaflets of the sensitive plant (Mimosa pudica) are thigmonastic responses that occur rapidly and are generally independent of the direction of stimulus. Thigmotropism, on the other hand, is directional growth determined by stimulus position and describes such phenomena as the ability of roots to grow around objects in the soil. Finally, the alteration of plant morphology and composition resulting from prolonged mechanical stimulation is referred to as thigmomorphogenesis (Jaffe, 1973).
The most common morphological change associated with thigmomorphogenesis is the development of shorter, more compact plants caused by decreased elongation. However, significantly different thigmomorphogenetic responses between and within species is indicative of genetic diversity regulating this phenomenon. Cucurbita pepo, Pisum sativum and Triticum aestivum exhibit little or no changes in elongation following mechanical stimulation of internodes while other species are significantly affected (Jaffe, 1973). Mechanical stimulation of cauliflower, lettuce and celery results in grossly smaller plants but increases hypocotyl thickness in lettuce and decreases hypocotyl thickness in cauliflower (Biddington and Dearman, 1985). Within 11 Arabidopsis populations treated with wind, eight populations developed a highly branched, bushy phenotype, two populations were intermediate and one population failed to develop the phenotype (Pigliucci, 2002). In response to mechanical perturbation, two Populus hybrids differed in stem height to diameter ratio, flexure rigidity and elasticity (Pruyn et al., 2000).
The identification of touch-regulated genes has provided insight into the proteins and signalling network that contribute to the diverse morphological changes associated with thigmomorphogenesis. The first touch-induced genes identified encode for a calmodulin (TCH1), calmodulin-like proteins (TCH2 and TCH3) and a xyloglucan endotransglucosylase/hydrolase (TCH4) (Braam and Davis, 1990). The transcript levels of these genes increase within 10–30 min following touch treatment (Braam and Davis, 1990). Because three of the TCH genes code for calmodulin (CaM) or CaM-like (CML) proteins, it was thought that these proteins play a central role in touch signal transduction. More recently, microarray analysis was used to show that 2·5 % of the Arabidopsis genome is at least 2-fold up-regulated 30 min following touch treatment (Lee et al., 2005). Only TCH4, from the original TCH genes, was found among 60 genes that are up-regulated at least 10-fold. These highly up-regulated genes code for proteins from diverse functional categories including kinases, transcription factors and putative disease resistance proteins (Lee et al., 2005). The specific functions of CaM or CaM-like proteins, therefore, are likely to be integrated into a complex touch-signalling pathway.
In part because of the complex signalling pathway, the ability to assign genes to phenotypes associated with thigmomorphogenesis has proven difficult. Analysis of changes in cellular composition, however, sheds some light on the underlying biochemical pathways. de Jaegher et al. (1985) found that rubbing the internodes of Bryonia dioica results in increased lignin, decreased cellulose and increased activities of phenylalanine ammonia-lyase and cell wall peroxidases. Furthermore, the chemical inhibition of lignin synthesis significantly counteracts the reduction of internode elongation, suggesting a direct role of lignification in thigmomorphogenetic growth responses (de Jaegher and Boyer, 1987). Alteration of chlorophyll content has also been associated with mechanical perturbation. Brushing increases chlorophyll per leaf weight in celery and lettuce but decreases chlorophyll levels in cauliflower (Biddington and Dearman, 1985). Interestingly, two mechanosensitive ion channel-like proteins from Arabidopsis, MSL2 and MSL3, were recently found to localize to the plastid envelope, and insertional mutagenesis of both genes was found to affect plastid size and shape (Haswell and Meyerowitz, 2006). It is possible that these proteins may contribute to the regulation of chlorophyll content.
The primary sensory mechanism(s) responsible for mechanoreception, which ultimately regulate thigmomorphogenesis, are just beginning to be revealed (reviewed by Telewski, 2006). A role of calcium in mechanotransduction was demonstrated by expressing transgenic apoaequorin in Nicotiana plumbaginifolia (Knight et al., 1991). Rapid accumulation of [Ca2+]i following touch indicated that a Ca2+-permeable, stretch-activated (SA) channel(s) may play a primary role in touch sensing and signalling. To investigate this possibility, Nakagawa et al. (2007) identified an Arabidopsis cDNA (MCA1) that functionally complements a lethal Ca2+-permeable (SA) channel associated mutant (mid1) in Saccharomyces cerevisiae. MCA1 localized to the plasma membrane and was capable of mediating Ca2+ uptake in yeast (Nakagawa et al., 2007). Evidence also suggests that MCA1 may contribute to touch sensing. In Chinese hamster ovary cells, MCA1 increased [Ca2+]i in response to stretch treatment, and the primary roots of mca1-null plants were found to be compromised in the ability to transition from soft agar to penetrate hard agar (Nakagawa et al., 2007). Although MCA1 may contribute to an initial touch response, its contribution to thigmomorphogenesis remains to be determined.
To comprehensively identify the biochemical events underlying thigmomorphogenesis, complementary model systems must be studied to define conserved pathways and reveal areas of divergence. The tropical fruit tree Carica papaya has many attributes that make it an attractive system for genetic studies, including a relatively small and sequenced genome (Arumuganathan and Earle, 1991; Ming et al., 2008), an established transformation system (Fitch et al., 1992), a short generation period, and relatively few genes. In this study, both well-known and novel phenotypes associated with thigmomophogenesis are revealed which offer new models for genetic and physiological investigation. In addition, a lower number of touch gene homologues in papaya, likely resulting from a lack of recent genome duplication (Ming et al., 2008), highlight its potential utility in mechanotransduction research.
Carica papaya L. (‘SunUp’) seeds were surface-sterilized and pretreated with 1·0 m KNO3, as previously described (Porter et al., 2007), before being planted in 3·5-in. (9 cm) pots filled with pre-moistened Sunshine® Mix 1 potting soil (Sun Gro Horticulture, Vancouver, BC, Canada). Pots were covered with a humidity dome to facilitate germination. Plants were grown at approx. 22–25 °C with a 16-h light/8-h dark photoperiod. Light was provided by high-pressure sodium Ceramalux Agro 430 W (SON AGRO 430W) and metal halide 400 W (MH400/U) lamps (Philips, Somerset, NJ, USA) producing PAR at approx. 175·0 µmol of quanta s−1 m−2 as measured using a LI-COR® Biosciences (Lincoln, NE, USA) light meter model LI-250. Touch treatment (Braam and Davis, 1990) was either performed at a specified time before tissue collection or three times per day (morning, noon and evening) by gently bending the stems of each plant approx. 45° forward, backward, left and right. Touch treatment was initiated upon emergence of cotyledons. Water spray treatment was conducted using a standard spray bottle adjusted for medium mist. Ten spray pumps were applied to each plant at a distance of approx. 30·5 cm.
Plant height was measured from the base of the stem to the shoot tip. The fourth true leaf was measured at the point of maximum width. Hypocotyl diameter was measured at the base using a 4-in. digital caliper (Woodcraft, Parkersburg, WV, USA). Root systems were thoroughly washed, removed from the shoots just above the first lateral root, blot dried and weighed.
Fresh petiole samples were harvested and placed immediately in 20-mL glass scintillation vials filled with antifreeze fixative solution composed of 10 % dimethyl sulfoxide, 4 % paraformaldehyde, 1 % TWEEN® 80 and 0·05 m sodium cacodylate, pH 7·5. Vials were loosely capped and placed under vacuum O/N to displace air with fixative. Petioles were cut to approx. 0·8 cm and mounted vertically in Leica Cryocompound (Leica Microsystems Inc., Bannockburn, IL, USA). A Leica CM1850 cryostat (Leica Microsystems Inc., Bannockburn, IL, USA) was used to obtain 20-μm tissue sections that were placed directly on microscope slides and rinsed with distilled H2O to remove residual Leica Cryocompound. Sections were either stained with Sudan IV or observed directly using standard light microscopy procedures.
For every 0·1 g of hypocotyl tissue, 1·0 mL of EtOH containing 1 % HCl was used to extract anthocyanins. Tissue was ground using a mortal and pestle until all pigments were extracted. One millilitre of extract was added to a microcentrifuge tube and centrifuged for 5 min at approx. 16 000 g. Absorbance of the supernatant was quantified at 530 nm using a DU® Series 700 UV/Vis Scanning Spectrophotometer (Beckman Coulter, Inc., Fullerton, CA, USA).
Chlorophyll was extracted from hypocotyls in 1·0 mL of 80 % acetone per 0·1 g of sample using a mortar and pestle for tissue homogenization. One millilitre from each extraction was centrifuged for 5 min at approx 16 000 g, and the supernatant was transferred to a quartz cuvette for spectrophotometry. Absorbance was quantified at three wavelengths (A663, A645 and A720), and the following equation was used to calculate total chlorophyll concentration:
Hypocotyl lignin analysis was conducted by the Agricultural Diagnostic Services Center, University of Hawai'i at Mānoa. To determine acid detergent lignin, 72 % sulfuric acid (ANKOM Technology, Macedon, NY, USA) was used.
Seedling shoots were collected, immediately frozen in liquid nitrogen and thoroughly ground using a mortar and pestle. RNA extraction was conducted using an RNeasy® Plant Mini Kit (Qiagen, Valencia, CA, USA). cDNA of genes 33·179 and 34·217 was amplified by RT-PCR, as previously described (Porter et al., 2007), using the following primers: 33·179F 5′-TGGCGATGGTTGCATCACCACCAAG-3′, 33·179R 5′-GACGAATTCCTCATAATTAATCT GA-3′, 34·217F 5′-CGGGGATGGTTGCATAACAACGAAA-3′ and 34·217R 5′-GACAAACTC CTCATAGTTGATTTGG-3′. Amplified fragments were cloned using the pGEM®-T Vector System (Promega Corporation, Madison, WI, USA). Sequencing was conducted using a 3700xl DNA analyser (Applied Biosystems, Foster City, CA, USA). Probes were labelled using the Rediprime™ II DNA Labeling System (GE Healthcare, Piscataway, NJ, USA). Standard northern blotting methods were used (Sambrook and Russell, 2001). Membranes were exposed to a Phosphor Screen (PerkinElmer®, Wellesley, MA, USA) that was scanned using a Cyclone™ Storage Phosphor System (PerkinElmer®). Signal intensity was quantified using Opti-Quant (Version 4·0) software (PerkinElmer®).
Papaya genome sequence was provided by the University of Hawai'i at Mānoa's Advanced Studies of Genomics, Proteomics and Bioinformatics (ASGPB) (http://asgpb.mhpcc.hawaii.edu/papaya/). Nucleic acid and protein sequences were compared with the database using BLASTN and BLASTP, respectively (Altschul et al., 1997). EVidenceModeler (EVM) and Program to Assemble Spliced Alignments (PASA) were used by the ASGPB to predict protein-coding genes from papaya genome sequence (Haas et al., 2008). Nucleic acid and amino acid sequence comparison was conducted using BLAST2 Sequences (Tatusova and Madden, 1999). CLUSTAL W was used for sequence alignments (Thompson et al., 1994), and TreeTop was used for phylogenetic tree prediction (Yushmanov and Chumakov, 1988). Protein domains were identified using Pfam (Finn et al., 2006).
Following 25 d of mechanical perturbation, papaya seedlings were stunted (approx. 46 % decrease in height) and had significantly increased (approx. 36 %) hypocotyl diameter (Fig. 1A–C). Touch treatment also retarded the accumulation of papaya root biomass by approx. 59 % and leaf width by approx. 42 % (Fig. 1D, E). Such developmental changes resemble the complex thigmomorphic responses to mechanical stimulation collectively referred to as hardening or strengthening (Whitehead, 1962; Jaffe and Forbes, 1993; Goodman and Ennos, 2001). The observed changes in stem, hypocotyl, leaf and root morphology of papaya are most likely the result of this process.
Seedlings of papaya cultivar ‘SunUp’ normally have dark, purple hypocotyls, but surprisingly, mechanical stimulation effectively blocked accumulation of anthocyanins to near undetectable levels (Fig. 1F–H). This phenomenon could be visually observed (Fig. 1F), and without exception, all touch-treated plants across three independent experiments exhibited loss of anthocyanin accumulation. Touched plants were also found to have approx. 12 % higher concentrations of lignin (Fig. 1I), while chlorophyll concentrations were found to be less than half of untouched seedlings (Fig. 1J).
After 25 d of touch-treatment, tissue outgrowths were observed on the abaxial side of petioles, occurring predominately on the first true leaves. Early-stage cellular outgrowth had narrow eruptions (<1 mm; Fig. 2A) that progressed to become larger and more amorphous in later stages (Fig. 2B, C). Microscopic examination of early-stage outgrowth (Fig. 3A) revealed a surface layer of dead epidermis and subepidermal cells (DE) immediately beneath which were rows of cells with thin, periclinal walls (Fig. 3B). The outermost layer of dead cells plus the periclinal files of cells supports the hypothesis that the thin-walled cells are phellem and phellogen cells. The former isolates external cells from nourishment and leads to their death (Fahn, 1990). The periderm (P) is subtended by hypertrophied cells of the cortex (Fig. 3B). Webb (1984) observed a similar type of periderm, including hypertrophied cortical cells, in cultured roots of Dioon edule exposed to light.
To confirm the presence of periderm, the tissue was stained. In regions of the petiole with normal growth, Sudan IV stained the lipid-rich epidermal cuticle (C) in the outer tangential walls (Fig. 3C) but not the inner walls of epidermal cells or any walls of subepidermal cells (Fig. 3C). In regions with a surface layer of dead cells, however, subepidermal cells were stained by Sudan IV (Fig. 3D), indicating the presence of suberin (S) which is characteristically associated with periderm formation (Fahn, 1990). It is likely, therefore, that the observed petiole tissue outgrowth is non-wounding, touch-induced periderm similar to cork normally associated with wounding (Fahn, 1990). Importantly, the phellogen cells observed in the petiole outgrowths have thinner walls as was previously reported in the roots of Dioon edule treated with light (Webb, 1984). Consequently, these cells express most but not all of the traits of periderm formation. More mature, late-stage outgrowth encompassed the entire sides of petioles (Fig. 3E) and displayed extensive cork formation with thickened cell walls (T) below the area of active periderm formation (Fig. 3F). Control petioles lacked anatomical features associated with touch-induced cellular outgrowths (Fig. 3G) or a layer of dead cells (Fig. 3H).
Recent completion of 3× coverage of the ‘SunUp’ papaya genome (Ming et al., 2008) allowed for comparison of predicted papaya open reading frames (ORFs) to ORFs from Arabidopsis that were previously shown to be touch-regulated (Lee et al., 2005). Fifty-one Arabidopsis ORFs with the highest expression ratios following 30 min of touch treatment were compared with papaya (Table 1). Of the original TCH genes, only TCH4 was among the top fifty-one, so TCH1, TCH2 and TCH3 were added to the analysis to explore homology of these genes as well. Four of the touch up-regulated ORFs in Arabidopsis were found to have no counterpart in papaya, two of which had expression ratios among the top ten (At3g04640 and At4g27652). All four of these Arabidopsis ORFs lack Pfam domain matches (Table 1). In addition, the Arabidopsis ORF with the third highest expression ratio (At3g01830) was found to have only minimal homology, at the amino acid level, to a predicted papaya ORF. The ORFs from Arabidopsis with the first (At4g29780) and fourth (At1g17420) highest touch expression ratios, however, were among the ORFs with the highest amino acid sequence similarity to papaya.
Interestingly, the gene with the highest similarity to papaya at the nucleic acid level was TCH1 (At2g41110). TCH1, also known as CAM2, codes for a calmodulin and was found to be highly related to two predicted papaya ORFs, 33·179 and 34·217 (for alignment, see Fig. S1 in Supplementary data, available online). The coding sequences of 33·179 and 34·217, designated Cp-TCH1-1 and Cp-TCH1-2, respectively, are 85 % and 84 % identical to TCH1, respectively, and 85 % identical to each other. Expression of both Cp-TCH1-1 and Cp-TCH1-2 was confirmed by RT-PCR and sequence analysis (data not shown). Following touch treatment, the transcript abundance of both ORFs showed a slight positive correlation across four time points (0, 10, 20 and 30 min) (Fig. 4A, B). Water spray treatment resulted in significant up-regulation of both Cp-TCH1-1 and Cp-TCH1-2 (Fig. 4C, D) as was previously reported for TCH1 (Braam and Davis, 1990). Homologues of TCH1 in papaya appear to be regulated as their Arabidopsis counterpart, especially in response to water spray treatment.
Recently, two Mechanosensitive channel of Small conductance-like (MscS-like or MSL) proteins in Arabidopsis, MSL9 and MSL10, were found to provide mechanosensitive ion channel activities in the plasma membrane of root cells (Haswell et al., 2008). In total, ten MSL genes have been identified in Arabidopsis (Haswell and Meyerowitz, 2006), and the sequences of these genes were used to screen the papaya genome. Seven MSL homologues were identified in papaya, each having significant sequence similarity to one or more ArabidopsisMSL gene(s) (Table S1 in Supplementary data, available online). As is the case for disease resistance genes, papaya contains a smaller number of MSL genes. MSL proteins have been categorized into two classes (Pivetti et al., 2003). ArabidopsisMSL1–3 belong to Class I proteins which are thought to localize to the mitochondria and chloroplast. MSL4–10 belong to Class II proteins which lack an organelle-targeting sequence (reviewed by Haswell, 2007). Phylogenetic analysis indicates that, while papaya and Arabidopsis appear to have an equal number of Class I proteins, papaya only has four proteins with significant similarity to the Class II proteins (Fig. 5).
To avoid damage, trees have had to evolve mechanisms to overcome the physico-mechanical stresses associated with wind (Telewski, 1995). Wind-regulated growth is especially important for papaya, which has a semi-woody, hollow trunk capable of reaching heights of 4–5 m. Touch and wind are both perceived as mechanical perturbation and have been associated with regulation of similar genes (Braam and Davis, 1990) and increases in cytosolic Ca2+ (Knight et al., 1991, 1992). The pathways regulating the novel touch responses described in this study, including hypertrophic outgrowths and altered anthocyanin biosynthesis, are likely a product of wind-associated selection pressure. Increased hypocotyl diameter, decreased height and root biomass and increased lignin content occurring after touch treatment of papaya (Fig. 1A–D, I) resemble the morphological changes associated with responses to wind as previously observed with pine trees (Jacobs, 1954; Meng et al., 2006). The rate of response, relative to that of woody trees, may correlate with papaya's rapid rate of development. Quantification of the rate of touch-regulated lignification relative to treatment dose in papaya and woody trees such as elm and pine (Telewski and Pruyn 1998; Meng et al., 2006) may be particularly useful for exploring these developmental differences. In addition, because of its taxonomic isolation as the only member of the genus Carica (Aradhya et al., 1999; Kim et al., 2002), detailed analysis of papaya's response to wind, relative to woody trees, is likely to reveal unique genetic and morphological differentiation.
As far as is known, this is the first report of mechanical perturbation that dramatically influences anthocyanin levels. The biosynthesis, accumulation and transport of anthocyanins are dependent on localization and subcellular conditions as well as multiple enzymes and substrates of the phenylpropanoid pathway. It is possible that touch treatments affected some of these processes. Because touch stimulation regulates numerous genes (Lee et al., 2005) and hormones (Biro and Jaffe, 1984; Sanyal and Bangerth, 1998; Stelmach et al., 1998; Weiler et al., 1993), several explanations can account for the dramatic down-regulation of anthocyanins observed here (Fig. 1F–H). In the phenylpropanoid pathway, hydroxycinnamoyl-CoA shikimate/quinate hydroxycinnamoyl transferase and caffeoyl-CoA O-methyltransferase are in pathways leading to lignin biosynthesis, and both enzymes compete with chalcone synthase (CHS) for the same substrate, p-coumaroyl CoA. Because CHS activity is required for anthocyanin production, it is possible that touch-driven lignin production (Fig. 1I) monopolizes p-coumaroyl CoA, deprives CHS of its substrate, and sharply decreases anthocyanin biosynthesis. Reduction of CHS and competition for the common CHS substrates, malonyl-CoA and p-coumaroyl CoA, were found to lead to alterations in anthocyanin biosynthesis in flowers (Courtney-Gutterson et al., 1994; Fischer et al., 1997). Other studies have also reported touch-regulated lignin production, as was found in Bryonia dioica (red bryony) calli and internodes, but not in association with anthocyanin variation (de Jaegher et al., 1985; Bourgeade et al., 1989).
Decreased auxin transport has been observed in mechanically stimulated Phaseolus vulgaris (French bean; Erner and Jaffe, 1982) in which ethylene production increased at the touch sites (Biro and Jaffe, 1984). Similarly, increased ethylene production and decreased polar auxin transport occurred in apple shoots following bending (Sanyal and Bangerth, 1998). Hydroxycinnamoyl transferase-silenced Arabidopsis plants accumulate higher levels of flavonoids, have altered lignification, decreased size, and inhibition of auxin transport (Besseau et al., 2007). The flavonoids quercetin and kaempferol were also shown to be effective auxin transport inhibitors in Cucurbita pepo hypocotyls (Jacobs and Rubery, 1988). Future studies will measure the effects of mechanostimulation of papaya on flavonoid content, ethylene production, auxin levels, stem elongation and anthocyanin accumulation.
Other studies have reported both increases and decreases in chlorophyll of leaves in response to mechanical perturbation. Cauliflower leaves had less chlorophyll (Biddington and Dearman, 1985), but celery, lettuce and tomato leaves were found to have more chlorophyll following mechanical stimulation (Mitchell et al., 1975; Biddington and Dearman, 1985). When comparing such treatments, however, the smaller leaf and organ sizes, as well as tissue density, of touched plants must be considered (Mitchell et al., 1975; Biddington and Dearman, 1985). Hypocotyls from touched papaya plants had increased radial expansion (Fig. 1C), which potentially could have diluted chlorophyll concentrations as measured per fresh weight (Fig. 1J). Absolute chlorophyll measurements, however, demonstrate that touch reduces total chlorophyll (Fig. S2 in Supplementary data, available online).
The touch-regulated petiole-specific outgrowth (Fig. 2A–C) is a new thigmomorphic phenotype characterized by parenchyma cell hypertrophy and cork formation. Callose deposition also occurs in response to mechanical perturbation. The synthesis and deposition of callose has been associated with plant–insect interaction (Hao et al., 2008), plant–pathogen interaction (Hamiduzzaman et al., 2005), wounding (Scherp et al., 2001; Jacobs et al., 2003) and non-wounding mechanical perturbation (Jaffe et al., 1985). Callose was rapidly deposited (within minutes) in sieve tubes at the sieve plate and pit fields of cell walls of P. vulgaris internodes in response to mechanical perturbation (Jaffe et al., 1985). A previous study in papaya suggests that phellogen and callose production can occur together. It was shown that papaya fruit inoculated with the fungal pathogen Colletotrichum gloeosporioides accumulate callose and a phellogen-like cell layer beneath C. gloeosporioides appressoria (Stanghellini and Aragaki, 1966). Callose deposition may also be associated with petiole hypertrophy and periderm formation of touch-treated papaya (Fig. 3A, B, D–F). Identification of callose in this tissue would provide additional evidence that pathogen and touch responses share common pathways, both of which can lead to the production of periderm.
Ethylene, a hormone associated with pathogen defence (Mur et al., 2008) that has been shown to be touch-regulated (Biro and Jaffe, 1984) may contribute to the stimulation of cork formation. Ethrel-treated loblolly pine (Pinus taeda) were found to have increased radial growth, decreased shoot elongation and stimulated bark production (Telewski et al., 1983).
Regardless of the pathways involved, the question of why touch-regulated petiole outgrowth is isolated to papaya seedling petioles remains unanswered. It is possible that this outgrowth is the result of hardening mechanisms, including added lignin, reacting with the high-degree nyctitropic movement associated with this tissue. In addition, in Arabidopsis, it was shown that increased [Ca2+]c, caused by gravistimulation, occurs primarily in hypocotyls and petioles (Toyota et al., 2008). Similarly, papaya petioles may be a site of active touch response.
It was shown that some highly touch-inducible Arabidopsis genes, such as At3g04640 and At4g27652, have no counterparts in papaya while others, such as At4g29780 and At1g17420, have significant sequence similarity to ORFs in papaya (Table 1). Others such as CML40 (At3g01830) have limited homology (Table 1). Missing counterparts may suggest a genetic basis for unique, species-specific thigmomorphic responses or compensation by other genes. Limited homology, however, may indicate the presence of functional homologues. Although a highly related homologue to the Arabidopsis calmodulin-like protein CML40 (McCormack et al., 2005) is not present in papaya, it is possible that a weakly related CML protein could serve the same functional role. Ultimately, when a papaya microarray is created, the same microarray expression analysis conducted to identify Arabidopsis touch-induced genes (Lee et al., 2005) will need to be conducted in papaya. Papaya has fewer genes than Arabidopsis and, notably, fewer disease-resistance genes (Ming et al., 2008), the expression of which are enriched following touch treatment in Arabidopsis (Lee et al., 2005). A smaller set of touch-regulated genes may facilitate the identification and analysis of regulators of mechanotransduction.
Arabidopsis TCH1 (calmodulin, CAM2) was found to have the highest nucleic acid similarity to papaya Cp-TCH1-1 and Cp-TCH1-2 (Table 1, and Fig. S1 in Supplementary data, available online). Arabidopsis has seven highly related CAM genes (McCormack et al., 2005), but only TCH1 has been reported to be touch induced (Braam and Davis, 1990; Lee et al., 2005). The reason papaya has two touch-regulated CAM2-like genes and is lacking other touch homologues remains to be determined. The generally high expression level of CAMs in Arabidopsis indicates a requirement for abundant CAM protein (McCormack et al., 2005). Having two touch-induced CAMs in papaya may be an alternate strategy for ensuring an adequate supply of CAM protein.
Plant MSL proteins have been identified that are related to the MscS (Mechanosensitive channel of Small conductance) family of proteins from bacteria. First identified in the plasma membrane of giant Escherichia coli spheroplasts (Sukharev et al., 1993), these channels have a small conductance of approx. 1 nS and open in response to stretching of the cell membrane. The phylogenetic relationship of Arabidopsis MSL proteins and papaya homologues alludes to functional similarities. Papaya ORFs 126·38 and 28·80 are most similar to the Class II proteins MSL9 and MSL10 (Fig. 5, and Table S1 in Supplementary data, available online), which localize to the plasma membrane of Arabidopsis root cells and are required for mechanosensitive channel function (Haswell et al., 2008). Predicted papaya ORFs 21·18 and 55·26 are most related to MSL2 and MSL3 (Fig. 5 and Table S1) and may function in the regulation of plastid size and shape (Haswell and Meyerowitz, 2006). Localization of papaya MSL protein may provide additional evidence of functional similarity.
Carica papaya is the fifth angiosperm to be sequenced (Ming et al., 2008), and as more species are added to this list, comparative genomics will allow multiple systems to be used synergistically to dissect complex, developmental processes. In this study, novel thigmomorphogenetic phenotypes demonstrate an association between mechanical perturbation, anthocyanins and periderm development. In addition, fewer touch-regulated gene homologues were found in papaya, relative to Arabidopsis. The phenotypes identified in this study may now be used as markers to complement the functional characterization of these genes.
Supplementary information is available online at www.aob.oxfordjournals.org and includes the following. Supplementary dataset: the nucleic acid and amino acid sequences of the predicted Carica papaya ‘SunUp’ open reading frames (ORFs) described in this study. Table S1: Comparison (blastp E-values) of predicted Carica papaya open reading frames to MscS-like proteins from Arabidopsis (Pivetti et al., 2003). Fig. S1: Multiple sequence alignment of Arabidopsis gene TCH1 (AT2G41110) and two C. papaya TCH1 homologues, Cp-TCH1-1 (33·179) and Cp-TCH1-2 (34·217). Fig. S2: Quantification of total chlorophyll from hypocotyls of papaya seedlings following 25 d of touch treatment.
The authors thank the Advanced Studies of Genomics, Proteomics and Bioinformatics (ASGPB), University of Hawai'i at Mānoa, for providing the predicted C. papaya open reading frames used for this study. Funding for this project was provided by the United States Department of Agriculture (USDA), Cooperative State Research Education and Extension Service (CSREES), Tropical and Subtropical Agriculture Research (T-STAR No. 2006-34135-17684) to D.A.C. and Y.J.Z.