Research
© The Authors (2009) New Phytologist  (2009) 182: 863–877   863  Journal compilation © New Phytologist  (2009) www.newphytologist.org   863
BlackwellPublishingLtdOxford, UKNPHNew Phytologist0028-646X1469-8137© The Authors (2009).Journalcompilation© New Phytologist (2009)282210.1111/j.1469-8137.2009.02822.xMarch200900863???877???OriginalArticle XX XX 
Early gene expression programs accompanying trans-differentiation of epidermal cells of Vicia faba  cotyledons into transfer cells
Stephen J. Dibley, Yuchan Zhou, Felicity A. Andriunas, Mark J. Talbot, Christina E. Offler, John W. Patrick and
David W. McCurdy 
School of Environmental and Life Sciences, The University of Newcastle, Callaghan, New South Wales 2308, Australia 
Summary
extensive wall ingrowths that enhance plasma membrane transport of nutrients.
Here, we investigated transcriptional changes accompanying induction of TC
development in adaxial epidermal cells of cultured Vicia faba cotyledons.
• Global changes in gene expression revealed by cDNA-AFLP were compared
between adaxial epidermal cells during induction (3 h) and subsequent building
(24 h) of wall ingrowths, and in cells of adjoining storage parenchyma tissue, which
do not form wall ingrowths.
• A total of 5795 transcript-derived fragments (TDFs) were detected; of these, 264
TDFs showed epidermal-specific changes in gene expression and a further 207 TDFs
were differentially expressed in both epidermal and storage parenchyma cells. Genes
involved in signalling (auxin/ethylene), metabolism (mitochondrial; storage product
hydrolysis), cell division, vesicle trafficking and cell wall biosynthesis were specifically
induced in epidermal TCs. Blockers of auxin action and vesicle trafficking inhibited
ingrowth formation and marked increases in cell division accompanied TC development.
• Auxin and possibly ethylene signalling cascades induce epidermal cells of V. faba
cotyledons to trans-differentiate into TCs. Trans-differentiation is initiated by rapid
de-differentiation to a mitotic state accompanied by mitochondrial biogenesis driving
storage product hydrolysis to fuel wall ingrowth formation orchestrated by a modified
vesicle trafficking mechanism.
Email: David.McCurdy@newcastle.edu.au
Key words: cDNA-AFLP, trans- differentiation, transfer cells, Vicia faba, wall ingrowths.
Introduction
Transfer cells (TCs) are characterized by wall ingrowths that protrude into the cytoplasm forming a complex labyrinth (Talbot  et al ., 2001) that acts as a scaffold for an amplified plasma membrane enriched in nutrient transporters (Offler et al  ., 2003). Transfer cells trans -differentiate from diverse cell types in response to developmental cues, stress or other factors (Offler  et al  ., 2003). Despite their importance in nutrient exchange in plants and, consequently, plant development (Offler et al  ., 2003), little is known of the identity of genes that orchestrate their induction and building of their wall labyrinths.
Several genes nominated as TC-specific have been identified in basal endosperm TCs of developing maize kernels. These
include four defensin-like genes, BETL1-4  (Thompson et al  ., 2001), a novel cell wall-related protein, MEG1 (Thompson et al ., 2001; Gutiérrez-Marcos etal ., 2004), and a transcriptional activator, ZmMRP-1 (Gómez et al ., 2002). The transcriptional activator has been shown to activate BETL and MEG1 promoters (Gutiérrez-Marcos et al  ., 2004), and the ZmMRP-1 promoter itself is active in regions of active transport between source and sink tissues (Barrero et al ., 2009). Furthermore, a putative type-A response regulator gene, ZmTCRR-1, was shown to be specifically expressed in the basal endosperm layer of maize (Muñiz et al  ., 2006). While these studies provide important insights, our wider understanding of the molecular processes underlying TC development remains poor.
 
www.newphytologist.org  Journal compilation © New Phytologist  (2009)
Research864
2000) and thus are not readily accessible for experimental analysis. However, this issue is obviated by the readily accessible epidermal cells of Vicia faba  (Faba bean) cotyledons. During  in planta cotyledon development, abaxial epidermal cells trans - differentiate to form TCs but adaxial epidermal cells do not (Offler et al  ., 1997). However, when cotyledons are cultured  with their adaxial surface in contact with nutrient agar, adaxial epidermal cells form small papillate wall ingrowths within 3 h (Wardini et al ., 2007b), and functional TCs with a complex, transporter-rich wall labyrinth by 48 h (Offler et al  ., 1997; Farley et al  ., 2000; Wardini et al  ., 2007a). Thus, the V. faba  cotyledon culture system provides a large population of  developing TCs that share the same induction event, and importantly, are morphologically and functionally equivalent to TCs that form in planta .
To analyse transcriptional changes accompanying induction and development of TCs in adaxial epidermal cells of V. faba  cotyledons, we developed an efficient and simplified cDNA- amplified fragment length polymorphism (AFLP) procedure incorporating nonsaturating PCR cDNA amplification to profile transcripts derived from isolated epidermal cells. We show that large-scale changes in gene expression occur within 3 h of TC induction, with many genes being induced, upregulated or rapidly switched off specifically in the adaxial epidermal cell layer undergoing trans -differentiation. Of particular interest among genes exhibiting upregulated and selective expression in trans -differentiating adaxial epidermal cells were suites involved in auxin signalling, cell division, vesicle trafficking associated with cell wall biosynthesis, mitochondrial biogenesis and storage product hydrolysis. Blocking auxin action or vesicle trafficking  inhibited wall ingrowth formation, thus confirming these
processes as key participants in TC development. Enhanced numbers of mitotic figures present in 3-h cultured adaxial epidermal cells demonstrated that these cells underwent rapid de-differentiation on exposure to inductive conditions. Collectively, these observations provide new insights into the early gene expression events leading to the induction and formation of TCs.
Materials and Methods
Plant material, cotyledon culture and tissue processing
Vicia faba  L. (cv. Fiord) plants were grown in controlled glasshouse and growth cabinet conditions (Talbot etal ., 2001). At harvest, cotyledons of 80–120 mg FW were removed surgically  from their seed coats and either fixed immediately in ice-cold ethanol and acetic acid (3 : 1, v : v) for 1 h at 4°C or cultured adaxial face down on modified Murashige and Skoog (MS) media containing 50 mm glucose and 50 mm fructose (Farley  et al  ., 2000) for 3 h or 24 h and then fixed (see earlier). Fixed tissue was processed rapidly by rinsing briefly in distilled  water before isolating sheets of adaxial epidermal cells as epidermal peels. The adhering ‘tag’ of parenchyma tissue (Fig. 1a) was surgically removed and each epidermal peel snap frozen in liquid nitrogen. Light and scanning electron microscopy (SEM) observations revealed that most epidermal cells in the peels were sheared along their anticlinal walls and their cellular contents remained mostly intact (Fig. 1b,c). Following peeling, 1-mm thick discs of storage parenchyma  tissue, free of epidermal cells, were collected from 3-h cultured cotyledons using a 5-mm diameter cork borer and immediately  snap frozen.
 
 Journal compilation © New Phytologist  (2009) www.newphytologist.org
Research 865
RNA extraction and cDNA amplification
Epidermal peels, from a minimum of three cotyledons per treatment, were pooled and total RNA extracted using an RNeasy RNA isolation kit (Qiagen). Total RNA was extracted from corresponding storage parenchyma discs using an RNeasy RNA isolation kit following treatment with Trizol reagent (Invitrogen). Extracted RNA was reverse transcribed with Powerscript reverse transcriptase using 3′-rapid amplification of cDNA ends (RACE) CDS primer A and SMART II A  oligonucleotide (Clontech; see the Supporting Information, Table S1) to generate fully transcribed first-strand cDNA  tagged with short sequences complementary to the SMART II A oligonucleotide at both the 5′ and 3′ ends. This first strand cDNA was purified with phenol–chloroform and ethanol precipitated using linear polyacrylamide as a carrier (Ambion, Austin TX, USA) and then used as template for full-length cDNA amplification following the Super SMART PCR cDNA synthesis kit (Clontech) protocol. Amplification parameters were optimized empirically by electrophoretic analysis of small aliquots of PCR products from every two cycles. The resulting double-stranded amplified cDNA were purified and adjusted to equal total amounts by comparing  amplification of V. faba GAPDH1 and elongation factor 1-α  
(VfGAPDH1-FP/RP and VfEF1α-FP/RP primer pairs, respectively; see Table S1) by semiquantitative PCR.
RNA fingerprinting with cDNA-AFLP
The cDNA-AFLP fingerprinting reactions were carried out using a protocol modified from Bachem et al  . (1998). Briefly, equal amounts of amplified cDNA from each experimental sample were digested with  Mse I and  ApoI (NEB, Ipswich, MA, USA). The resulting digestion fragments were ligated to enzyme-specific adaptors (Milioni et al  ., 2002, and Table S1) using T4 DNA ligase (MBI Fermentas, Burlington, Canada). Fragments ligated to the ApoI adaptor, biotinylated at the 5′ terminus, were collected following binding to streptavidin- coated paramagnetic Dynabeads (Dynal, Oslo, Norway). A  1/10 dilution of this ligation reaction was preamplified using  primers targeted to the adaptor sequences (Table S1). A 1/100 dilution of preamplification product was used for each selective PCR determined by two specified bases at the 3′ end of each primer extending into the fragment sequence. In contrast to Bachem et al  . (1998), PCR primer concentration was 250 nm to allow fragment visualization by silver staining. The PCR  reactions were incubated at 94°C for 10 min followed by 13 cycles of 94°C for 30 s, 65°C for 30 s and 72°C for 1 min,  with the annealing temperature dropping by 0.6°C each cycle. The reactions were completed by 23 cycles of 94°C for 30 s, 56°C for 30 s and 72°C for 1 min. All 256 possible PCR  primer combinations were tested. Products from these reactions were run on 16 cm-long 5% polyacrylamide gels for 4.5 h at 40 mA and stained using a rapid silver staining 
procedure (Qu et al  ., 2005). Fragments were visualized on a  Molecular Imager gel documentation XR system (Bio-Rad). Relative band intensities were determined using quantity 
one  software (version 4.6.3; Bio-Rad), and a band was classified as differentially expressed if its intensity showed ≥ 5-fold temporal change.
Transcript-derived fragment (TDF) extraction, verification and sequencing
Each TDF of interest on silver-stained polyacrylamide gels  was stabbed with a sterile 200 µl pipette tip and incubated in 15 µl of 10 mm  Tris-HCl (pH 8.0) for 30 min at room temperature. Each fragment was reamplified using 5 µl of the fragment extract as template and subjected to the selective PCR cycle program with an additional seven cycles at an annealing temperature of 56°C. Reamplified products were separated on agarose gels, DNA bands were extracted using  the Wizard gel and PCR purification kit (Promega) and cloned directly into the TA cloning vector pGEM-T Easy  (Promega). Clones were sequenced using T7 primer and BDT sequencing chemistry (Invitrogen). Gene homology analysis  was performed using the blast  program (Altschul et al  ., 1997) at the NCBI website (http://www.ncbi.nlm.nih.gov/ BLAST/) and the TIGR database (http://www.jcvi.org/) with default parameters. The TDF sequences were searched against blastx and blastn of NCBI or blastx and blastn of TIGR. Promoter sequences (a maximum of 2 kb up-stream of the  ATG start codon) of each Arabidopsis orthologue were screened for regulatory cis -elements by Athena (O’Conner et al ., 2005).
Expression of selected TDFs was validated by quantitative real-time PCR, with Platinum Taq polymerase and SYTO9 dye (Invitrogen), on unamplified cDNA produced from independently isolated RNA. Reactions were performed using  a Corbett RotorGene 6000 with fluorescence acquisition through the green channel. Expression quantification utilized the ‘two standard curve’ method as described in the Corbett Rotor-Gene 6000 software package (version 1.7), using V. faba  elongation factor 1-α (VfEF1-α  ) standard curve to normalise expression.
Scanning electron microscopy of treated cotyledons
Cotyledon cultures were established as described earlier except that sister cotyledons were divided between culture media with or without the specified treatment (Wardini et al  ., 2007b) or prepared under green light. After 15 h, adaxial epidermal peels were prepared, washed in 2% (w : v) NaOCl for 3 h and subsequently dehydrated at 4°C through a 10% step-graded ethanol–distilled H2O series, changed at 30-min intervals. Peels were critical point-dried with liquid CO2 in a  critical-point drier (Balzers Union, Liechtenstein) and secured outer face down onto sticky tabs to reveal the cytoplasmic face of their outer periclinal cell walls. Samples were sputter-coated
www.newphytologist.org  Journal compilation © New Phytologist  (2009)
Research866
 with gold to a thickness of 20 nm in a sputter-coating unit (SPI Suppliers, West Chester, PA, USA), and viewed at 15 kV   with a Philips XL30 SEM.
Analysis of cell division
Cotyledons were cultured and fixed as described earlier. After  washing briefly in phosphate-buffered saline (PBS), adaxial epidermal peels were collected and stained with 1 µg ml−1
4,6-diamidino-2-phenylindole (DAPI) for 5 min. Epidermal peels were rinsed 2 × 5 min in PBS and mounted in Mowiol (Calbiochem, San Diego, CA, USA) with 0.1% (w : v)  p- phenylenediamine. Tissue was viewed with a Zeiss Axiophot epifluorescence microscope equipped with a 50 W short-arc mercury lamp and a UV (365–420 nm) filter (Osram). Mitotic indices were estimated as percentages of cells containing  mitotic profiles from at least 100 cells scored per replicate.
Results
Isolation of RNA and amplification of cDNA from epidermal peels
Recovery of total RNA obtained from either single or pooled (maximum of five) epidermal peels was not sufficient to yield reliable banding patterns using standard cDNA-AFLP protocols (data not shown). These protocols typically use up to 100 µg  of total RNA for starting material (Bachem  et al  ., 1998)
compared with nanogram amounts retrieved from epidermal peels. We therefore incorporated a nonsaturating PCR-based cDNA amplification step based on procedures developed for cDNA microarray analysis of small tissue samples (Hertzberg  et al ., 2001; see the Materials and Methods section). Figure 1d shows that transcript-derived fragments (TDFs), generated by  selective PCR of amplified cDNA, were consistent between technical repeats and detected temporal changes in selective gene expression (Fig. 1e). Using this modified procedure, we  were able to profile gene expression in adaxial epidermal cells of freshly isolated cotyledons (no culture) or those cultured for 3 h and 24 h, and to compare these profiles with those of  storage parenchyma cells from 3-h cultured cotyledons to identify changes in gene expression occurring specifically in adaxial epidermal cells and therefore likely to be related to TC induction and development (Table 1). Expression profiles deduced from our cDNA-AFLP approach (Table 1) were verified using real-time PCR on unamplified cDNA (see Fig. S1 and associated text). This analysis demonstrated that conclusions of cell-specific expression profiles could be drawn with confidence but distinction between induced and upregulated gene expression was less clear.
cDNA-AFLP analysis of transcriptional regulation accompanying induction and development of TCs
 Analysis of the 256 primer combinations containing two basepair overhangs yielded a total of 5795 TDFs, ranging in
Table 1 Categories of verified gene expression profiles identified in adaxial epidermal cells of Vicia faba cotyledons induced to form transfer cells (TCs)
Expression profilea, b
  Number (%) of TDFs
  s  p   e   c
   i   f   i  c
Induced 69 (15) Late-Induced 22 (5) Early Transient-Induced 21 (4) Up-Regulated 30 (6) Rapidly Switched-Off 102 (22) Gradually Switched-Off 20 (4)
   E   p
  a   n
  p   a   r  e   n   c
   h   y   m   a
Induced 116 (25) Late-Induced 13 (3) Early Transient-Induced 19 (4) Up-Regulated 19 (4) Rapidly Switched-Off 26 (6) Gradually Switched-Off 14 (3)
Total number of TDFs 471 (100)
 
 Journal compilation © New Phytologist  (2009) www.newphytologist.org
Research 867
size from 50 bp to 500 bp, from adaxial epidermal cells and storage parenchyma tissue. From this pool of TDFs, 756 demonstrated differential expression, defined here as a ≥ 5-fold change (up or down) in band intensity detected on silver-stained gels (see the Materials and Methods section). Of these, a total of 471 fragments were verified as true cDNA-AFLP fragments by extracting each band from the polyacrylamide gel and reamplifying using the original selective primer pair. An analysis of 234 V. faba cDNAs present in GenBank revealed that 72% were cut at least once by both ApoI and Mse I (data  not shown). Applying this percentage to the 471 differentially  expressed TDFs identified (Table 1), we estimate that TC formation may involve differential expression of c . 650 different genes. This estimate compares well with the numbers of  developmentally regulated genes detected during tracheary  element formation in cultured  Zinnia mesophyll cells by  cDNA-AFLP (562; Milioni  et al ., 2002) or by microarray  (523; Demura et al ., 2002) analyses. Furthermore, of the 471 differentially expressed TDFs, our approach identified 142 TDFs (Table 1 and hence an estimated 195 genes totally) that  were induced or upregulated specifically in epidermal cells during TC formation. This number of genes displaying  epidermal-specific changes in expression is within the range of  preferentially expressed genes reported for epidermal cells of  maize coleoptiles (130; Nakazono et al ., 2003) and Arabidopsis  stems at defined stages of development (180; Suh et al ., 2005). These comparisons support the conclusion that our cDNA-  AFLP study has successfully identified the majority of genes being differentially regulated specifically in epidermal cells during TC formation. Moreover, this conclusion is supported by the finding that genes known to be expressed exclusively in epidermal layers, such as fiddlehead  (Pruitt et al  ., 2000) and B1-type cyclin (Boudolf et al  ., 2004), were identified in the epidermal-specific cohort of TDFs (Table 2).
Temporal patterns of expression were classified as ‘Induced’, ‘Late-Induced’, ‘Early Transient-Induced’, ‘Up-Regulated’, ‘Rapidly Switched-Off’ and ‘Gradually Switched-Off’ (Table 1). Of those genes displaying epidermal-specific changes in expression, approximately equal numbers were either induced/ upregulated or switched-off rapidly or gradually (Table 1). Responses of differential gene expression were typically rapid,  with 85% of differential expression occurring within 3 h of  culture (Table 1).
Ontology-deduced functions of induced epidermal- specific genes relate to cell wall biosynthesis, metabolism and protein synthesis/metabolism and are potentially regulated by auxin and/or ethylene
Cotyledon culture induces the formation of TCs in adaxial epidermal cells but not in cells of the underlying storage parenchyma tissue (Farley   et al ., 2000; Talbot et al  ., 2007). Consequently, attention was focused on identifying, by  homology searching (see the Materials and Methods section),
the 112 TDFs showing epidermal-specific, induced expression (Induced, Late-Induced, Early Transient-Induced; Table 1). Genes showing this expression pattern are more likely to be directly related to TC development, compared with those associated with stress responses which are expected to be expressed comparably in the adjacent storage parenchyma tissue. Functional classifications were determined by searching blast similarity matches (blast expectation values [E ] of ≤ 10–3) through the Gene Ontology (http://www.geneontology.org) and KEGG BRITE (http://www.genome.jp/kegg/brite.html) databases, with confirmation by reference to the literature. This process enabled TDFs to be placed into one of nine predicted functional groups (Table 2; groupings based on the categories used by Milioni et al  ., 2002). Of the 112 TDFs, 44 (39%) returned no significant match to any database entry  (data not shown) and a further 15 (23%) matched database entries for hypothetical or unknown proteins (Table 2; Fig. 2a). The remaining 68 TDFs showed significant alignments and  were placed in functional groups. The major groups were metabolism, energy and storage (12 TDFs; 18% of 68), protein synthesis and metabolism (11 TDFs; 16%), cell  wall and vesicle trafficking (9 TDFs; 13%), and transcription (6 TDFs; 9%) (Table 2; Fig. 2a).
The development of TCs in tomato roots is regulated by auxin and ethylene (Schikora & Schmidt, 2001, 2002).  Accordingly, promoter regions of Arabidopsis orthologues of identified V. faba  TDFs (Table 2) were screened using   Athena (O’Conner et al ., 2005) for the presence of auxin- and ethylene-regulatory cis -element sequences. Of the 48 Arabi- dopsis orthologues identified, 24 (50%) contained at least one repeat of the auxin-responsive element,  AuxRe   (TGTCTC; Guilfoyle & Hagan, 2007) in its corresponding promoter region, while nine (19%) contained at least one ET-responsive element (GCC -box; Ohme-Takagi & Shinshi, 1995, and see Table 2). These percentages are substantially higher than the 41% and 9% for AuxRe and the GCC -box elements, respec- tively, found by searching all promoter regions in the Arabi- dopsis genome using the Data Mining application of Athena.
Genes encoding hypothetical and unknown proteins are abundant in those rapidly switched off within 3 h of cotyledon culture
Of the 102 TDFs whose epidermal-specific expression was rapidly switched-off (Table 1), 34 were selected for sequencing  based on their size (c . 150–400 bp) and band intensity on the silver-stained gels. Of this cohort only three returned no significant hits, and of those exhibiting low E values (Table 3; Fig. 2b), a substantial proportion (14 TDFs; 45%) matched hypothetical and unknown proteins suggesting the possibility  of novel functions linked with trans -differentiation of TCs. The proportion of genes distributed among the various functional groupings was generally similar to that observed for induced genes (compare Fig. 2b with 2a).
 Journal compilation © New Phytologist  (2009) www.newphytologist.org
Research 871
Similar to the cohort of induced genes, promoter analysis of the pool of Arabidopsis orthologues closely matching  V. faba  TDFs, which were rapidly switched off, revealed an increase in the frequency of  AuxRe   and GCC -box motifs  within the promoter regions of these identified genes (65% and 24%, respectively; Table 3).
Testing key functional pathways predicted by cDNA- AFLP gene discovery – light, auxin, vesicle trafficking and cell division
Predicted functions of V. faba genes deduced from ontology  searches of databases, which were rapidly and specifically  induced in adaxial epidermal cells, indicated possible light (e.g. Gravitropic in the light (GIL1) and Constans -like 3 (COL-3)) and auxin-mediated (e.g. GH1 and AuxRe  promoter
motifs) signalling pathways leading to wall ingrowth induction (Table 2). The operation of these predicted signalling pathways  were tested experimentally by culturing cotyledons under green light or in the presence of the competitive auxin inhibitor,
 p-chlorophenoxyisobutyric acid (PCIB; Oono et al  ., 2003). Cotyledon culture in the absence of an early light signal had no significant effect on wall ingrowth induction (Fig. 3e). By  contrast, PCIB reduced numbers of adaxial epidermal cells forming wall ingrowths by 60% (Fig. 3e).
More than 10% of genes showing induced, epidermal-specific expression encoded proteins predicted to be involved in vesicle trafficking and cell wall synthesis (Fig. 2a), for example,  ADP-ribosylation factor 1 (ARF1), YKT61 and a pectin methylesterase inhibitor (Table 2). To examine a requirement for vesicle trafficking in wall ingrowth formation, cotyledons  were cultured in the presence of Brefeldin A, a potent inhibitor
 
www.newphytologist.org  Journal compilation © New Phytologist  (2009)
Research872
of vesicle formation (Ritzenthaler et al  ., 2002). Under these conditions, wall ingrowth formation was abolished (93% inhibition; Fig. 3e), demonstrating an absolute require- ment for vesicle trafficking in wall ingrowth deposition.
Induction of a mitotic cyclin, an endonuclease and chromatin assembly factor C specifically in epidermal cells (Table 2) suggested activation of the cell cycle upon TC induction. Comparisons of mitotic index in adaxial epidermal cells showed a dramatic rise in mitotic rates following cotyledon culture, rising from 0.5 to 7.4 in the first 3 h (Fig. 3f).
Discussion
 We used experimental induction of adaxial epidermal TCs in V. faba  cotyledons to reveal transcriptional changes accompanying  trans -differentiation of epidermal cells into functional TCs (Tables 2 and 3). Rapid (< 3 h) epidermal-specific induction of genes (Table 1) is consistent with the finding of Wardini et al . (2007b) that all biosynthetic machinery required to form  wall ingrowths is transcribed within 1 h following exposure of 
cotyledons to inductive signal(s). Concurrently there is an equal number of genes rapidly switched off (122 TDFs; Table 1) upon exposure to culture, reflecting a major change in the epidermal transcriptome associated with trans - differentiation of epidermal TCs. Generic responses, including  those to abiotic stress, may be distinguished from those peculiar to trans -differentiation of epidermal TCs by analysing genes specifically induced in these cells (Epidermal-specific; Table 1). This assumption is supported by the absence of gene functions associated with generic stress responses from this cohort of  genes specifically induced in adaxial epidermal cells upon cotyledon culture (Table 2). The relative distribution of these genes among functional categories (Fig. 2a) matches those reported for tracheary element formation (Milioni et al ., 2002) except for expression of transporter genes and those linked  with cell division. Expression of transporter genes (e.g. ammonium transporter, P-type and vacuolar H+-ATPases; Table 2) is consistent with TC function (Offler et al  ., 2003) and further supports our conclusion that the experimental approach used here has enabled identification of gene
 
 Journal compilation © New Phytologist  (2009) www.newphytologist.org
Research 873
Signalling TC induction – role for light, auxin and ethylene?
The extent of wall ingrowth formation in phloem parenchyma  and companion cell TCs of  Arabidopsis   and pea leaves, respectively, has been shown to be dependent on incident light flux densities (Amiard  et al  ., 2005). For the V. faba  cotyledon system, exposure of their adaxial epidermal cells to  white light upon cotyledon removal from seed coats may  initiate a photomorphogenic signal cascade. In this context, induction of homologues of CONSTANS-like  (COL-3; Datta  et al  ., 2006) and Gravitropic in the Light  (GIL1; Allen et al  ., 2006) and downregulation of B-EXPANSIN   (Tepperman et al  ., 2004) is consistent with a phytochrome-driven response (Tepperman et al  ., 2004; Tables 2, 3). COL-3, in contrast to most COLs that function in flowering responses, has been shown to control vegetative growth patterns (Datta et al  ., 2006) that might include cell wall formation. However, rates of wall ingrowth initiation in adaxial epidermal TCs were found to be independent of a light signal (Fig. 3). Whether a  light signal affects the extent of wall ingrowth formation in committed adaxial epidermal TCs (Amiard et al ., 2005) remains to be determined. Indeed, upregulation of GIL1 (Table 2), that renders auxin transport nonpolar in the dark (Allen et al  ., 2006), provides a link between light and auxin signals possibly  mediating induction of wall ingrowth formation.
Elevated auxin levels are known to enhance formation of  TCs in rhizodermal cells of a number of species, including  tomato (Schikora & Schmidt, 2001). An indication that auxin levels are elevated in adaxial epidermal cells of cultured cotyledons is provided by the induced expression of a MtN21 homologue (Table 2), a signature gene for elevated levels of  auxin in developing tissues (Busov   et al ., 2004). Observed profiles of selective gene expression in adaxial epidermal cells (Table 2) indicate that elevated auxin levels could arise from altered transport and/or enhanced biosynthesis. Inhibitory  effects of flavonoids on auxin transport (Peer & Murphy, 2007) could be relieved by their enhanced metabolism through induced expression of flavonoid 3-O-galactosyltransferase  (Miller et al ., 2002) and glutathione-S-transferase  (Smith et al  ., 2003). Induction of GIL1  and an aminopeptidase   (Table 2) could impact on auxin transport by randomly relocalizing PIN1 proteins around plasma membranes of cotyledon cells (Murphy   et al ., 2005; Allen  et al  ., 2006). These effects on auxin transport, combined with enhanced auxin biosynthesis by induced expression of a nitrilase   (Table 2), catalysing  hydrolysis of indole-3-acetonitrile into active indole-3-acetic acid (IAA; Vorwerk  et al  ., 2001), could alter patterns of auxin distribution to drive wall ingrowth formation. High auxin concentrations could account for the transient induction of an early-response auxin gene, GH1 homologue (Table 2), belonging 
to the Aux/IAA gene family of transcriptional regulators (Guilfoyle et al  ., 1993). The Aux/IAA proteins interact with auxin response factors (ARFs) to confer various auxin responses alone or in combination by binding to AuxRe  motifs (Guilfoyle & Hagan, 2007). These motifs are enriched (54 vs 41%) among Arabidopsis orthologues of the genes identified in our cDNA-AFLP screen (Tables 2, 3), indicating a potentially  important role for auxin in orchestrating wall ingrowth formation. This conclusion is supported by finding that PCIB, an auxin analogue that inhibits auxin action by competitively  binding with auxin receptors (Oono  et al  ., 2003), reduced numbers of adaxial epidermal cells forming wall ingrowths in cultured cotyledons (Fig. 3).
The proposition that ethylene may contribute to TC induction in V. faba  cotyledons arises from finding a 2.7-fold enrichment of ethylene responsive cis -elements in promoter regions of Arabidopsis  orthologues of differentially expressed V. faba genes (Tables 2 and 3). This proposition is supported by the finding that 1-aminocyclopropane-1-carboxylic acid (ACC, an ethylene precursor) enhanced TC formation in root epidermal cells of tomato (Schikora & Schmidt, 2002) and adaxial epidermal cells of V. faba cotyledons (F. A. Andriunas et al ., unpublished).
Guided by the presence of  AuxRe and GCC -box motifs (Tables 2, 3), significant downstream targets of auxin and ethylene signalling pathways inducing TC development could include cellular metabolism ( AuxRe ), cell division ( AuxRe ) and vesicle trafficking/cell wall biosynthesis (GCC -box). These phenomena are discussed in the following sections.
Transfer cell induction coincides with increases in cell division
 
www.newphytologist.org  Journal compilation © New Phytologist  (2009)
Research874
to the B1 subgroup. The B1-CDKs drive the G2/M transition in mitosis (Francis, 2007), and are expressed preferentially in epidermal cells (Boudolf et al  ., 2004). A B1-CDK dependent arrest at the G2/M phase accounts for the ability of these cells to rapidly (within 3 h) enter mitosis upon exposure to the inductive signal (Fig. 3f).
In addition to the epidermal-specific induction of mitosis, induction of twoβ-1,3-glucanases (Table 2) suggests reinitiation of cytokinesis during trans -differentiation of epidermal TCs. Both induced genes are Family 17 glycoside hydrolases (Minic & Jouanin, 2006), with the Arabidopsis  orthologue of V245B (Table 2) ascribed with an ancestral function in cell division/ cell wall remodelling (Doxey et al  ., 2007). In this instance, the β-1,3-glucanase may be a candidate for performing a specialized role during cell plate formation or, alternatively, participating  in wall remodelling events required to achieve the unique morphology of reticulate wall ingrowths.
Modification of energy metabolism during transfer cell development
Induced genes selectively expressed in adaxial epidermal cells contributing to energy metabolism (Table 2) included components of the mitochondrial electron transport chain (nad 7 , cob, cox1) and Kreb cycle (aconitase , malate dehydrogenase ). Expression of mitochondrial-encoded nad 7 , cob  and cox1 (Table 2) are insensitive to altered oxygen tensions resulting  from cotyledon excision (Rolletschek et al  ., 2003) but reflect expression profiles linked with mitochondrial biogenesis (Howell et al  ., 2007). This process is possibly orchestrated by  chromatin assembly factor C (CAF-C; Table 2), which is known to influence mitochondrial numbers in yeast through the Ras/cAMP pathway (Ruggieri et al  ., 1989; Rigoulet et al  ., 2004). Consistent with this conclusion, mitochondrial matrix  densities and cristae formation increase along with mitochondrial numbers in adaxial epidermal cells undergoing wall ingrowth development (Farley et al  ., 2000). Induced expression of  NADH-dependent malic enzyme and aconitase (Table 2) is suggestive that mitochondrial activity has switched to an anaplerotic mode to meet demand for intermediates consumed in various synthetic processes underpinning wall ingrowth construction.
Given that sugar demand exceeds supply during the trans -differentiation of epidermal TCs in planta  (Harrington et al  ., 1997) and in vitro (Wardini et al  ., 2007a), carbon skeletons are likely to be sourced from reserves. In this context, a profile of genes potentially involved in remobilization of storage compounds were induced (Table 2), including  those remobilizing lipids (triacylglycerol lipase, hydroxysteroid dehydrogenase, aconitase and malate dehydrogenase) and starch (isoamylase and glyceraldehyde-3-phosphate dehydrogenase). Oil body breakdown through triacylglyceride lipase activity   would provide free fatty acids to enter glyoxysomes as described for germinating seeds (Eastmond, 2006). Within glyoxysomes,
fatty acid molecules are oxidized to acetyl-CoA and enter the glyoxylate cycle to produce C4 precursors which can be used for energy generation or fed through gluconeogenesis into an array of biosynthetic pathways, including cell wall component biosynthesis (see previous section). An alternative energy  source may be derived through starch hydrolysis catalysed by  isoamylases (Hussain et al  ., 2003) in combination with a  plastid glyceraldehyde-3-phosphate dehydrogenase (Table 2) forming part of a carbon pathway before plastid/cytosol exchange of carbon skeletons.
Vesicle trafficking is essential in wall ingrowth deposition
It is not surprising that genes involved in vesicle trafficking  and cell wall biogenesis are induced in epidermal cells undergoing trans -differentiation into TCs. Genes induced or switched off specifically in epidermal TCs (Tables 2, 3) can be presumed to act as regulators of papillate ingrowth deposition at defined loci (Talbot et al  ., 2001; Fig. 3). The absence of key   wall-building genes from Table 2 (listing genes expressed in epidermal cells but not in storage parenchyma) such as celluloses and sucrose synthases is explained by their generic expression in all cells undergoing expansion at this stage of  cotyledon development (Borisjuk et al  ., 1995).
Modification of vesicle trafficking upon TC induction is evident through the induction of vesicle-targeting genes ADP- ribosylation factor 1 ( ARF1), a SNARE (YKT61) and a clathrin coat adaptor subunit (Table 2). The ARFs are considered central to orchestrating asymmetrical vesicle trafficking to effect polarity in plant cells (Xu & Scheres, 2005), a characteristic consistent with wall ingrowth deposition in adaxial epidermal cells of cotyledons. However, Class 1 ARFs (Table 2) act as intracellular regulators of trafficking, being primarily localized to the Golgi and subpopulations of post-Golgi vesicles (Matheson et al  ., 2008). Induction of YKT61, a SNARE located in the cis -Golgi cisternae (Chen et al  ., 2005), together with ARF1, is indicative of enhanced protein trafficking between Golgi and endoplasmic reticulum (ER). Upregulation of ARF1, but not SAR1, the GTPase responsible for assembly of COP11 protein coats directing vesicle budding from the ER (Memon, 2004), points to ARF1 as a key regulator of vesicle trafficking activity  during wall ingrowth formation. This conclusion is supported by inhibiting this process when cotyledons were cultured in the presence of Brefeldin A (Fig. 3d,e). Interestingly, this result suggests that the contribution of cellulose synthase/sucrose synthase complexes to building papillate wall ingrowths (Talbot et al ., 2007) also depends upon vesicle trafficking.
 
 Journal compilation © New Phytologist  (2009) www.newphytologist.org
Research 875
 Arabidopsis  (Chen et al ., 2005). Since large numbers of secretory  vesicles are associated with developing wall ingrowths in TCs (Wardini et al  ., 2007b), it is possible that, together with the early transiently upregulated putative ras-GTPase-activating  protein (Table 2), ARF1/YKT61/AP-3 could constitute a  specialized gene complex facilitating increased trans -Golgi vesicle delivery to the plasma membrane in a polarized manner.
Genes encoding wall components and modifying enzymes are not well represented in the list of genes specifically induced or switched-off in epidermal cells (Tables 2, 3), consistent with  wall ingrowths being compositionally equivalent to primary  cell walls (Vaughn et al  ., 2007). Some exceptions are β-N - acetylhexosaminidase (Table 3), UDP-glucosyltransferase and pectin methylesterase inhibitor (Table 2). Induction of a V-  ATPase (Table 2) in the trans-Golgi network could increase synthesis of cell wall components and trafficking to the membrane (Brüx et al  ., 2008). Late induction of a pectin methylesterase inhibitor (PMEI) is consistent with maintaining  extensibility of developing wall ingrowths. Wall ingrowths are rich in pectins, and for V. fabacotyledon epidermal TCs these pectins are esterified (Vaughn et al  ., 2007 and references cited therein). Pectin methylesterases (PMEs) have a major role in pectin remodelling (Pelloux et al ., 2007) through de-esterification decreasing wall extensibility (Röckel et al ., 2008). Therefore, late induction of PMEI (Table 2) suggests a role in maintaining exten- sibility of developing wall ingrowths as they commence branching  and fusing to form a fenestrated layer (Talbot et al  ., 2001).
Conclusions
Extensive, rapid and cell-specific transcriptional regulation underpins trans -differentiation of adaxial epidermal cells of  V. faba  cotyledons into TCs. Auxin, possibly in combination  with ethylene, functions as an inductive signal to initiate wall ingrowth formation. The induction of TCs initiates re-entry  into a division cycle coincidental with modification of vesicle trafficking and wall assembly machinery specifically in these cells. The rapid stepped increase in metabolic demand by the trans -differentiating epidermal cells for intermediates to support these biosynthetic activities is met by remobilization of lipid and starch stores processed through an enhanced anaplerotic pathway in newly formed mitochondria. Inhibition of pectin de-esterification in wall ingrowths could confer sufficient mechanical flexibility to form the characteristic fenestrated wall layers. The insights generated from these findings open new opportunities for further studies to expand our understanding of signalling pathways inducing, and metabolic machinery responsible for constructing, the intricate  wall ingrowths of TCs.
Acknowledgements
 We thank Kevin Stokes for raising healthy experimental material and acknowledge funding of this project from Australian
Research Council Discovery Project grants DP0556217 and DP0664626.
References
 Allen T, Ingles PJ, Praekelt U, Smith H, Whitelam GC. 2006.
Phytochrome-mediated agravitropism in Arabidopsis hypocotyls requires
GIL1 and confers a fitness advantage. Plant Journal  46: 641–648.
 Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W,
Lipman DJ. 1997. Gapped blast and psi-blast: a new generation of
protein database search programs. Nucleic Acids Research 25: 3389–3402.
 Amiard V, Much KE, Demmig-Adams B, Ebbert V, Turgeon R, Adams
 WWI. 2005. Anatomical and photosynthetic acclimation to the light
environment in species with differing mechanisms of phloem loading.
Proceedings of the National Academy of Sciences, USA  102: 12968–12973.
Bachem CWB, Oomen RJFJ, Visser RGF. 1998. Transcript imaging with
cDNA-AFLP: a step-by-step protocol. Plant Molecular Biology Reporter  16:
157–173.
Barrero C, Royo J, Grijota-Martinez C, Faye C, Paul W, Sanz S, Steinbiss
H-H, Hueros G. 2009. The promoter of ZmMRP-1, a maize transfer
cell-specific transcriptional activator, is induced at solute exchange surfaces
and responds to transport demands. Planta  229: 235–247.
Borisjuk L, Weber H, Panitz R, Manteuffel R, Wobus U. 1995.
Embryogenesis of Vicia faba L.: histodifferentiation in relation to starch
and storage protein synthesis. Journal of Plant Physiology  147: 203–218.
Boudolf V, Barrôco R, Engler JDA, Verkest A, Beeckman T, Nandts M,
Inzé D, De Veylder L. 2004. B1-type cyclin-dependent kinases are
essential for the formation of stomatal complexes in  Arabidopsis thaliana . Plant Cell  16: 945–955.
Brüx A, Liu T-Y, Krebs M, Stierhof Y-D, Lohmann JU, Miersch O,
 Wasternack C, Schumacher K. 2008. Reduced V-ATPase activity in the
trans -Golgi network causes oxylipin-dependent hypocotyl growth
inhibition in Arabidopsis . Plant Cell  20: 1088–1100.
Busov VB, Johannes E, Whetten RW, Sederoff RR, Spiker SL, Lanz-Garcia
C, Goldfarb B. 2004. An auxin-inducible gene from loblolly pine (Pinus taeda L.) is differentially expressed in mature and juvenile-phase shoots and
encodes a putative transmembrane protein. Planta  218: 916–927.
Chen Y, Shin YK, Bassham DC. 2005.  YKT6 is a core constituent of
membrane fusion machineries at the Arabidopsis trans -Golgi network.
 Journal of Molecular Biology  350: 92–101.
Datta S, Hettiarachchi GHCM, Deng X-W, Holm M. 2006.  Arabidopsis CONSTANS-LIKE3 is a positive regulator of red light signaling and root
growth. Plant Cell  18: 70–84.
Demura T, Tashiro G, Horiguchi G, Kishimoto N, Kubo M, Matsuoka N,
Minami A, Nagata-Hiwatashi M, Nakamura K, Okamurua Y  et al .
2002. Visualization by comprehensive microarray analysis of gene
expression programs during transdifferentiation of mesophyll cells into
xylem cells. Proceedings of the National Academy of Sciences, USA  99:
15794–15799.
Doxey AC, Yaish MW, Moffatt BA, Griffith M, McConkey BJ. 2007.
Functional divergence in the Arabidopsis β-1,3-glucanase gene family
inferred by phylogenetic reconstruction of expression states. Molecular Biology & Evolution 24: 1045–1055.
Eastmond PJ. 2006. SUGAR-DEPENDENT1encodes a patatin domain
triacylglycerol lipase that initiates storage oil breakdown in germinating
 Arabidopsis seeds. Plant Cell  18: 665–675.
Farley SJ, Patrick JW, Offler CE. 2000. Functional transfer cells differentiate
in cultured cotyledons of Vicia faba L. seeds. Protoplasma  214: 102–117.
Francis D. 2007. The plant cell cycle – 15 years on.New Phytologist  174:
261–278.
Gómez E, Royo J, Guo Y, Thompson R, Hueros G. 2002. Establishment of
cereal endosperm expression domains: identification and properties of a
maize transfer cell-specific transcription factor,  ZmMRP-1. Plant Cell  14:
599–610.
www.newphytologist.org  Journal compilation © New Phytologist  (2009)
Research876
Guilfoyle TJ, Hagan G. 2007. Auxin response factors. Current Opinion in Plant Biology  10: 453–460.
Guilfoyle TJ, Hagen G, Li Y, Ulmasov T, Liu ZB, Strabala T, Gee M. 1993.
 Auxin-regulated transcription. Australian Journal of Plant Physiology  20:
489–502.
Gutiérrez-Marcos JF, Costa LM, Biderre-Petit C, Khbaya B, O’Sullivan
DM, Wormwald M, Perez P, Dickinson HG. 2004. maternally expressed  gene1 is a novel maize endosperm transfer cell-specific gene with a maternal
parent-of-origin pattern of expression. Plant Cell  16: 1288–1301.
Haritatos E, Medville R, Turgeon R. 2000. Minor vein structure and sugar
transport in Arabidopsis thaliana . Planta  211: 105–111.
Harrington GN, Nussbaumer Y, Wang X-D, Tegeder M, Franceschi VR,
Frommer WB, Patrick JW, Offler CE. 1997. Spatial and temporal
expression of sucrose transport-related genes in developing cotyledons of
Vicia faba L. Protoplasma  200: 35–50.
Hertzberg M, Sievertzon M, Aspeborg H, Nilsson P, Sandberg G,
Lundeburg J. 2001. cDNA microarray analysis of small plant tissue
samples using a cDNA tag target amplification protocol. Plant Journal  25:
585–591.
Howell KA, Cheng K, Murcha MW, Jenkin LE, Millar H, Whelan J. 2007.
Oxygen initiation of respiration and mitochondrial biogenesis in rice.
 Journal of Biological Chemistry  282: 15619–15631.
Hussain H, Mant A, Seale R, Zeeman S, Hinchliffe E, Edwards A,
Hylton C, Bornemann S, Smith AM, Martin C et al . 2003. Three
isoforms of isoamylase contribute different catalytic properties for the
debranching of potato glucans. Plant Cell  15: 133–149.
Matheson LA, Suri SS, Hanton SL, Chatre L, Brandizzi F. 2008. Correct
targeting of plant ARF GTPases relies on distinct protein domains. Traffic   9: 103–120.
Memon AR. 2004. The role of ADP-ribosylation factor and SAR1 in
vesicular trafficking in plants. Biochimica et Biophysica Acta 1664:
9–30.
Milioni D, Sado P-E, Stacey NJ, Roberts K, McCann MC. 2002. Early gene
expression associated with the commitment and differentiation of a plant
tracheary element is revealed by cDNA-amplified fragment length
polymorphism analysis. Plant Cell  14: 2813–2824.
Miller KD, Stommer J, Taylor LP. 2002. Conservation in divergent
solanaceous species of the unique gene structure and enzyme activity of a
gametophytically-expressed flavonol 3-O -galactosyltransferase. Plant  Molecular Biology  48: 233–242.
Minic Z, Jouanin L. 2006. Plant glycoside hydrolases involved in cell
 wall polysaccharide degradation. Plant Physiology & Biochemistry   44: 435–449.
Muñiz LM, Royo J, Gómez E, Barrero C, Bergareche D, Hueros G. 2006.
The maize transfer cell-specific type-A response regulator ZmTCRR-1 appears to be involved in intercellular signalling. Plant Journal  48:
17–27.
Murphy AS, Bandyopadhyay A, Holstein SE, Peer WA. 2005.
Endocytotic cycling of PM proteins. Annual Review of Plant Biology  56:
221–251.
Nakazono M, Qiu F, Borsuk LA, Schnable PS. 2003. Laser-capture
microdissection, a tool for the global analysis of gene expression in specific
plant cell types: identification of genes expressed differentially in epidermal
cells or vascular tissues of maize. Plant Cell  15: 583–596.
O’Conner TR, Dyreson C, Wyrick JJ. 2005. Athena: a resource for rapid
visualization and systematic analysis of  Arabidopsis promoter sequences.
Bioinformatics  21: 4411–4413.
Offler CE, Liet E, Sutton EG. 1997. Transfer cell induction in cotyledons
of Vicia faba L. Protoplasma  200: 51–64.
Offler CE, McCurdy DW, Patrick JW, Talbot MJ. 2003. Transfer cells:
cells specialized for a special purpose. Annual Review of Plant Biology  54:
431–454.
proteins that interact with an ethylene-responsive element. Plant Cell  7:
173–182.
1 (ARF1) regulates recruitment of the AP-3 adaptor complex to
membranes. Journal of Cell Biology  142: 391–402.
Oono Y, Ooura C, Rahman A, Aspuria ET, Hayashi K, Tanaka A,
Uchimiya H. 2003. p-Chlorophenoxyisobutyric acid impairs auxin
response in arabidopsis root. Plant Physiology  133: 1135–1147.
Peer WA, Murphy AS. 2007. Flavonoids and auxin transport: modulators or
regulators? Trends in Plant Science  12: 556–563.
Pelloux J, Rustérucci C, Mellerowicz EJ. 2007. New insights into pectin
methylesterase structure and function. Trends in Plant Science  12:
267–277.
Pruitt RE, Vielle-Calzada J-P, Ploense SE, Grossniklaus U, Lolle SJ. 2000.
FIDDLEHEAD , a gene required to suppress epidermal cell interactions in
 Arabidopsis , encodes a putative lipid biosynthetic enzyme. Proceedings of the National Academy of Sciences, USA 97: 1311–1316.
Qu LJ, Li XY, Wu GQ, Yang N. 2005. Efficient and sensitive method of
DNA silver staining in polyacrylamide gels. Electrophoresis  26: 99–101.
Rigoulet M, Aguilaniu H, Avéret N, Bunoust O, Camougrand N,
Grandier-Vazeille X, Larsson C, Pahlman I-L, Manon S, Gustafsson L.
2004. Organization and regulation of the cytosolic NADH metabolism in
the yeast Saccharomyces cerevisiae . Molecular and Cellular Biochemistry   256–257: 73–81.
Ritzenthaler C, Nebenführ A, Movafeghi A, Stussi-Garaud C, Behnia L,
Pimpl P, Staehelin LA, Robinson DG. 2002. Reevaluation of the effects
of brefeldin A on plant cells using tobacco bright yellow 2 cells expressing
Golgi-targeted green fluorescent protein and COPI antisera. Plant Cell  14:
237–261.
Röckel N, Wolf S, Kost B, Rausch T, Greiner S. 2008. Elaborate spatial
patterning of cell-wall PME and PMEI at the pollen tube tip involves
PMEI endocytosis, and reflects the distribution of esterified and
de-esterified pectins. Plant Journal  53: 133–143.
Rolletschek H, Weber H, Borisjuk L. 2003. Energy status and its control on
embryogenesis of legumes. Embryo photosynthesis contributes to oxygen
supply and is coupled to biosynthetic fluxes. Plant Physiology  132:
1196–1206.
Ruggieri R, Tanaka K, Nakafuku M, Kaziro Y, Toh-EA, Matsumoto K.
1989. MSI1, a negative regulator of the RAS-cAMP pathway in
Saccharomyces cerevisiae . Proceedings of the National Academy of Sciences, USA 86: 8778–8782.
Schikora A, Schmidt W. 2001. Acclimative changes in root epidermal cell
fate in response to Fe and P deficiency: a specific role for auxin?
Protoplasma  218: 67–75.
Schikora A, Schmidt W. 2002. Formation of transfer cells and H+-ATPase
expression in tomato roots under P and Fe deficiency. Planta  215:
304–311.
Smith AP, Nourizadeh SD, Peer WA, Xu J, Bandyopadhyay A, Murphy AS,
Goldsbrough PB. 2003. Arabidopsis AtGSTF2 is regulated by ethylene and
auxin, and encodes a glutathione S-transferase that interacts with
flavonoids. Plant Journal  36: 433–442.
Suh MC, Samuels AL, Jetter R, Kunst L, Pollard M, Ohlrogge J, Beisson F.
2005. Cuticular lipid composition, surface structure, and gene expression
in Arabidopsis stem epidermis. Plant Physiology  139: 1649–1665.
Talbot MJ, Franceschi VR, McCurdy DW, Offler CE. 2001. Wall ingrowth
architecture in epidermal transfer cells of Vicia faba cotyledons.
Protoplasma  215: 191–203.
Talbot MJ, Wasteneys GO, Offler CE, McCurdy DW. 2007. Cellulose
synthesis is required for deposition of reticulate wall ingrowths in transfer
cells. Plant & Cell Physiology  48: 147–158.
Tepperman JM, Hudson ME, Khanna R, Zhu T, Chang SH, Wang X,
Quail PH. 2004. Expression profiling of phyB mutant demonstrates
substantial contribution of other phytochromes to red-light-regulated
gene expression during seedling de-etiolation. Plant Journal  38:
725–739.
Thompson RD, Hueros G, Becker HA, Maitz M. 2001. Development and
 
 Journal compilation © New Phytologist  (2009) www.newphytologist.org
Research 877
 Vaughn KC, Talbot MJ, Offler CE, McCurdy DW. 2007. Wall ingrowths
in epidermal transfer cells of Vicia faba cotyledons are modified primary
 walls marked by localized accumulations of arabinogalactan proteins.
Plant & Cell Physiology  48: 159–168.
 Vorwerk S, Biernacki S, Hillebrand H, Janzik I, Müller A, Weiler EW,
Piotrowski M. 2001. Enzymatic characterization of the recombinant
 Arabidopsis thaliana nitrilase subfamily encoded by the NIT 2/NIT 1/NIT 3-gene cluster. Planta  212: 508–516.
 Wardini T, Talbot MJ, Offler CE, Patrick JW. 2007a. Role of sugars in
regulating transfer cell development in cotyledons of developing Vicia faba seeds. Protoplasma  230: 75–88.
 Wardini T, Wang X-D, Offler CE, Patrick JW. 2007b. Induction of wall
ingrowths of transfer cells occurs rapidly and depends upon gene
expression in cotyledons of developing Vicia faba seeds. Protoplasma  231:
15–23.
FACTOR 1 function on epidermal cell polarity. Plant Cell  17:
525–536.
Supporting Information
 Additional supporting information may be found in the online version of this article.
Fig. S1 Transcript-derived fragment (TDF) expression patterns in unamplified cDNA using real-time PCR.
Table S1 Oligonucleotide primer sequences used for cDNA  synthesis and amplification, cDNA-amplified fragment length polymorphism (AFLP) and real-time PCR verification of  TDF expression
Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting information supplied by  the authors. Any queries (other than missing material) should be directed to the New Phytologist  Central Office.
About New Phytologist 
• New Phytologist  is owned by a non-profit-making charitable trust  dedicated to the promotion of plant science, facilitating projects from symposia to open access for our Tansley reviews. Complete information is available at www.newphytologist.org .
• Regular papers, Letters, Research reviews, Rapid reports and both Modelling/Theory and Methods papers are encouraged. We are committed to rapid processing, from online submission through to publication ‘as-ready’ via Early View – our average submission to decision time is just 29 days. Online-only colour is free, and essential print colour costs will be met if necessary. We also provide 25 offpr ints as well as a PDF for each article.
• For online summaries and ToC alerts, go to the website and click on ‘Journal online’. You can take out a personal subscription to the journal for a fraction of the institutional price. Rates start at £139 in Europe/$259 in the USA & Canada for the online edition (click on ‘Subscribe’ at the website).