Capsicum annuum CCR4-associated factor CaCAF1 is necessary for plant development and defence response

Authors


  • School of Biological Sciences and Technology, Chonnam National University, Kwangju 500-757, Korea.

(fax: 82 2 873 2056; e-mail: doil@snu.ac.kr or fax: 82 42 860 4309; e-mail sol6793@kribb.re.kr).

Summary

The CCR4-associated factor 1 (CAF1) protein belongs to the CCR4-NOT complex, which is an evolutionary conserved protein complex and plays an important role in the control of transcription and mRNA decay in yeast and mammals. To investigate the function of CAF1 in plants, we performed gain- and loss-of-function studies by overexpression of the pepper CAF1 (CaCAF1) in tomato and virus-induced gene silencing (VIGS) of the gene in pepper plants. Overexpression of CaCAF1 in tomato resulted in significant growth enhancement, with increasing leaf thickness, and enlarged cell size by more than twofold when compared with the control plants. A transmission electron microscopic analysis revealed that the CaCAF1-transgenic tomato plants had thicker cell walls and cuticle layers than the control plants. In addition to developmental changes, overexpression of CaCAF1 in tomato plants resulted in enhanced resistance against the oomycete pathogen Phytophthora infestans. Additionally, microarray, northern and real-time polymerase chain reaction analyses of CaCAF1-transgenic tomato plants revealed that multiple genes were constitutively upregulated, including genes involved in polyamine biosynthesis, defence reactions and cell-wall organogenesis. In contrast, VIGS of CaCAF1 in pepper plants caused significant growth retardation and enhanced susceptibility to the pepper bacterial spot pathogen Xanthomonas axonopodis pv. vesicatoria. Our results suggest roles for plant CAF1 in normal growth and development, as well as in defence against pathogens.

Introduction

Plants have basal defence responses that function in both compatible and incompatible interactions with pathogens. These defence responses are associated with a number of early and late events in plant cells. Plants have evolved many different mechanisms to defend against pathogen invasion, such as the generation of reactive oxygen species (Alvarez et al., 1998), ion flux change (Kuchitsu et al., 1997), synthesis of phytoalexins (Hain et al., 1993), expression of a series of pathogenesis-related (PR) proteins (Park et al., 2001) and callose deposition (Donofrio and Delaney, 2001). In many cases of host–pathogen interaction, the expression of a large number of genes is changed dramatically (Hammond-Kosack and Jones, 1997). Understanding the roles of these transcriptionally regulated genes following pathogen infection will be the prerequisite for understanding the whole picture of pathogen defence in plants.

Previously, we found that the expression of Capsicum annuumCCR4-associated factor 1 (CaCAF1) in the leaves of chilli pepper increased following infection by the soybean pustule pathogen Xanthomonas axonopodis pv. glycines 8ra (Xag 8ra) (Lee et al., 2004). CCR4-assocaited factor 1 (CAF1), also called CCR4-NOT transcription complex subunit 7, belongs to the CCR4-NOT complex, which is an evolutionarily conserved protein complex involved in the control of transcription and mRNA decay (Liu et al., 2001). The CCR4-NOT-CAF1 complex is likely to have a fundamental function in cellular homeostasis (Berthet et al., 2004). The CCR4-NOT complex has two major forms: one of approximately 1.9 MDa in length and the other of approximately 1.0 MDa in length (Liu et al., 2001). The larger complex contains five different NOT proteins and several other proteins (Liu et al., 1998). The smaller complex is the core complex and harbours several identified and unidentified components (Liu et al., 1998; Chen et al., 2001). In the core complex, CAF1 is a part of the CCR4p complex that binds to the central region of NOT1 (Chen et al., 2001).

Several biological roles have been proposed for CAF1 (Liu et al., 1997; Hata et al., 1998; Bai et al., 1999; Berthet et al., 2004). Although the roles of CAF1 within CCR4-NOT are not clearly established, CAF1 is a key subunit of the CCR4-NOT complex, enabling its formation, and is essential for the function of the complex (Berthet et al., 2004). In Saccharomyces cerevisiae, CAF1 plays an important role within this complex, acting as a bridge between the CCR4 and NOT proteins (Bai et al., 1999). CAF1 is required for glucose de-repression of gene expression, and also works as a major factor for mRNA decay in association with DHH1p in S. cerevisiae (Hata et al., 1998). CAF1 plays a critical role in cell-cycle control in association with the cell-cycle kinase DBF2 (Liu et al., 1997). In mammalian cells, ectopic overexpression of CAF1 exerted an anti-proliferative effect (Bogdan et al., 1998). Mice lacking murine CAF1 are viable, but males are sterile (Berthet et al., 2004).

Although several studies have reported the functional roles of CAF1 in yeast and mammals, and have emphasized the importance of CAF1 for cellular homeostasis, there has been no report on the role of CAF1 in plants. Therefore, we investigated the roles of the plant CAF1 gene using gain- and loss-of function studies in plants of the Solanaceae family. Overexpression of CaCAF1 in tomato plants and the reduction of gene expression of CaCAF1 by virus-induced gene silencing (VIGS) in pepper plants enabled us to estimate the roles of CaCAF1 in both plant defence and development. The regulons under the direct and indirect control of CaCAF1 were analysed using RNA gel and microarray data. Here, we report the roles of the CAF1 gene in plant development and biotic stress resistance.

Results

Sequence analysis, genomic organization and expression of CaCAF1

CaCAF1 was isolated from a chilli pepper cDNA library prepared from pepper leaves inoculated with the non-host pathogen Xag 8ra. The full-length cDNA of CaCAF1 (NCBI accession number DQ 672569) is 891 bp in length, and contains one open reading frame encoding a protein of 266 amino acid residues with a calculated molecular mass of 30 kDa. A search against the protein databases revealed that the CaCAF1 protein contains an RNaseD domain (DEDD) that may function in polyA-specific exonuclease activity. Phylogenetic analysis revealed that CaCAF1 is most closely related to the Arabidopsis thaliana CAF1 (NCBI accession number CAB88994), with 73% identity at the amino acid level, followed by Oryza sativa CAF1 (NCBI accession number XP468264) with 52% identity (Figure 1a). CaCAF1 shares 49 and 48% identity at the amino acid level with CAFs from Homo sapiens (NCBI accession number AAP36213) and Drosophila pseudoobscura (NCBI accession number EAL30121), respectively (Figure 1a).The genomic organization of CaCAF1 was investigated by Southern blot analysis using the 5′-untranslated region (5′-UTR) of CaCAF1. Only a single band was detected in each restriction digestion of the genomic DNA (data not shown); however, there is an expressed sequence tag (EST) with 85% amino acid homology with CaCAF1 in the database at the Solanacea project (http://sol.pdrc.re.kr/). These data suggest that the genome of C. annuum contains at least two copies of CAF1 genes. The expression of CaCAF1 was detected in various organs of pepper plants, including leaves, stems, roots and flowers. Expression was especially high in stems and roots, whereas it was low in flowers (Figure 1b).

Figure 1.

 Isolation and characterization of Capsicum annuum CCR4-associated factor 1 (CaCAF1).
(a) Comparison of the amino acid sequences of CaCAF1 and CAFs from other organisms. The amino acid sequence of CaCAF1 (NCBI accession no. DQ672569) was aligned with those of Arabidopsis thaliana (NCBI accession no. CAB88994), Oryza sativa (NCBI accession no. XP468264), Drosophila pseudoobscura (NCBI accession no. EAL30121) and Homo sapiens (NCBI accession no. AAP36213). The conserved residues are boxed in black or light grey based on the degree of conservation. The RNaseD DEDD domain residues are indicated by the arrows. Alignment was performed using the clustalw multiple alignment program (http://www.ebi.ac.uk/clustalw/) (Chenna et al., 2003).
(b) Tissue-specific expression of CaCAF1 in pepper plants. Total RNA extracted from various tissues was hybridized with 32[P]dCTP-labelled full-length CaCAF1 cDNA. Equal loading was identified by staining rRNA with ethidium bromide.

Overexpression and silencing of CaCAF1

To evaluate the roles of CaCAF1, tomato plants (Solanum esculentum cv. ‘MicroTom’) were transformed to express the full-length cDNA of CaCAF1 under the control of the CaMV 35S promoter. A total of 15 independently transformed plants were acquired, and four T2 lines (lines 2, 3, 4 and 8) were selected based on the constitutively high overexpression of CaCaF1 under normal growth conditions. The expression levels of CaCAF1 in the leaves of T3 lines and equal loading of total RNA were documented (Figure 2a, upper panels). All T3 lines showed much higher expression of CaCAF1 than control plants transformed with an empty vector. Enhanced growth was observed in all CaCAF1 overexpressing T3 tomato lines compared with the control plants (Figure 2a, middle panel). CaCAF1 overexpressing lines 2, 3, 4 and 8 had plant weights 1.40, 1.33, 1.25 and 1.52 times higher, respectively, than the control plants at the flowering stage (Figure 2a, lower panel). There seems to be a correlation between the CaCAF1 expression levels and plant weights. However, we did not digitalize the expression levels to calculate the ratios of the expression levels to plant weights because of the lack of knowledge about the effect of the the CaCAF1 site in the tomato genome, and about the mechanisms for enhanced growth.

Figure 2.

 Phenotypes of CaCAF1-transgenic ‘MicroTom’ and CaCAF1-silenced ‘ECW-30R’ pepper plants.
(a) Expression of CaCAF1 and phenotypes in the T3-transgenic ‘MicroTom’ lines 2, 3, 4 and 8. (Upper panels) RNA gel blot analysis detected the expression level of CaCAF1. Equal loading was identified by staining rRNA with ethidium bromide. (Middle panel) Control and CaCAF1-transgenic lines at the flowering stage. (Lower panel) Plant weight was measured at the flowering stage. Data are expressed as the mean ± SE of 10 plants per line.
(b) Effect of silencing CaCAF1 in ‘ECW-30R’ pepper plants by virus-induced gene silencing. (Upper panels) real-time polymerase chain reaction analysis of CaCAF1 expression levels in the control and CaCAF1-N and -C plants. The actin gene was used to identify the use of equal quantities of total RNA to make cDNA. (Middle panel) Photograph of the control and CaCAF1-silenced pepper plants taken after 35 days of VIGS treatment. (Lower panel) Plant height measured after 35 days of VIGS treatment. Data are expressed as the mean ± SE of 10 plants per line.

The loss-of-function phenotype of CaCAF1 in ‘ECW-30R’ pepper plants was investigated using the VIGS method. CaCAF1-N and CaCAF1-C vectors were constructed to silence specifically the N-terminal and C-terminal regions of CaCAF1, respectively. Silencing of CaCAF1 in ‘ECW-30R’ plants with the CaCAF1-N or -C constructs significantly decreased the expression of CaCAF1 compared with control plants silenced with the GFP gene (Figure 2b, upper panel). The equal usage of total RNA to make cDNA was verified through PCR for the pepper actin gene (Figure 2b, middle panel). When the phenotypes of pepper plants were observed after 35 days of VIGS treatment, CaCAF1-N- and CaCAF1-C-silenced pepper plants showed significantly retarded growth (Figure 2b, middle and lower panels). CaCAF1-N- and -C-silenced pepper plants showed 26.7 and 33.3% reduction in plant height, respectively (Figure 2b, lower panel).

Cytological analysis of transgenic tomato plants over expressing CaCAF1

The mechanisms underlying enhanced plant growth in the CaCAF1-transgenic tomato plants were further investigated by measuring cell size and cell number. Cross sections of CaCAF1-transgenic and control plant leaves were prepared and observed under a light microscope (Figure 3). Microscopic observations revealed that all CaCAF1-transgenic tomato lines had significantly thicker leaves than control plants (Figure 3a–d). When the length of mesophyll cells was measured, transgenic lines had cells at least twice as long as control plants (Figure 3e). Similar results were obtained for the size of protoplasts isolated from leaves (data not shown). However, enlarged cell size in the leaves of transgenic tomato lines resulted in a decrease in cell density (total number of cells per unit area), rather than increasing or maintaining the cell density (Figure 3f).

Figure 3.

 Leaf morphology of control and CaCAF1-transgenic ‘MicroTom’ lines. Images were acquired using light microscopy for (a) control, (b) CaCAF1-2, (c) CaCAF1-3 and (d) CaCAF1-4 lines. Scale bar is 100 μm.
(e) Length of mesophyll cells in the control and CaCAF1-transgenic lines 2, 3 and 4.
(f) Numbers of cell per 550 μm, including mesophyll and upper and lower epidermal cells. Cell counts were performed at least three times independently; data are expressed as the mean ± SE of the three replicates.

A transmission electron microscope was used to observe changes in the cell wall and cuticle region of CaCAF1-transgenic tomato plants. All CaCAF1 overexpressing lines had significantly thicker cell walls and cuticles than control plants (Figure 4a–d). The cell walls of the CaCAF1-transgenic lines 2, 3 and 4 were 1.83, 1.65 and 2.04 times thicker than those of control plants, respectively (Figure 4e).

Figure 4.

 Morphology of the cell wall and cuticle of the leaf epidermal cells from control and CaCAF1-transgenic ‘MicroTom’ lines. The shape of the cell wall and cuticle was analysed by transmission electron microscopy. Leaves (30-days old) were cross-sectioned for imaging.
(a) Control plants;
(b–d) CaCAF1-transgenic lines. Abbreviations: cu, cuticle; cw, cell wall; ct, cytoplasm. The size of the scale bar is 0.5 μm.
(e) Thickness of cell walls in the control and CaCAF1-transgenic lines 2, 3 and 4. Data are expressed as the mean ± SE of the three replicates.

Enhanced resistance of CaCAF1-transgenic plants to Phytophthora infestans

Phytophthora infestans, the cause of tomato late blight disease, was spray-inoculated onto control and CaCAF1-transgenic plants to test whether ectopic expression of CaCAF1 changed the level of disease resistance. Disease symptoms were visible in both the control and transgenic tomato plants within 2 days of P. infestans spraying. However, the development of disease symptoms was significantly delayed in all the CaCAF1-transgenic lines (2, 3, 4 and 8) compared with the control plants (Figure 5a, upper panel). In the control plants, disease symptoms were observed over approximately 40% of the total leaf area, whereas in the transgenic tomato lines, symptoms were observed over only 10–25% of the total leaf area (Figure 5a, lower panel). Because CaCAF1-transgenic lines showed increased resistance to P. infestans, we analysed the expression of two important PR genes, PR1 and PR6, in control and CaCAF1-transgenic lines. Interestingly, under normal growth conditions, PR1 and PR6 were constitutively expressed at higher levels in the CaCAF1-transgenic tomato plants than in the control plants (Figure 5b).

Figure 5.

 Response of CaCAF1-transgenic ‘MicroTom’ plants to Phytophthora infestans infection.
(a) Reduced infection frequency in CaCAF1-transgenic ‘MicroTom’ 3 days after P. infestans infection. The control and CaCAF1-transgenic lines 2, 3, 4 and 8 were used. (Upper panel) Symptoms of infection in the control and CaCAF1-transgenic ‘MicroTom’ plants. (Lower panel) Disease index measured on 10 plants per line infected with P. infestans. Data are expressed as the mean ± SE.
(b) Expression level of CaCAF1 and pathogenesis-related (PR) genes LePR1 and LePR6 in the control and CaCAF1-transgenic lines under normal growth conditions. Expression of CaCAF1, LePR1 and LePR6 was analysed by RNA gel blot analyses.

Responses of CaCAF1-silenced plants to bacterial pathogens

To determine how a loss of function of CaCAF1 affects the defence response, we infiltrated X. axonopodis pv. vesicatoria (Xav) race 1 and race 3 intoCaCAF1-silenced ‘ECW-30R’ pepper plants. ‘ECW-30R’ pepper plants are compatible with Xav race 3 and incompatible with Xav race 1; thus, they respond to Xav race 3 by inducing disease and to Xav race 1 by inducing a hypersensitive response (HR). We measured colony-forming units (CFU) per cm2 in ‘ECW-30R’ challenged with Xav race 3 or documented HR induced by the infiltration of Xav race 1. Interestingly, Xav race 3 grew much faster in both VIGS lines, CaCAF1-silenced ‘ECW-30R’ by the CaCAF1-N or -C construct, than in the control plants for 6 days post-inoculation (Figure 6a). These data indicate that the CaCAF1-silenced lines had lower levels of resistance to Xav race 3 than the control plants. Blocked symptoms of HR were also observed in both the CaCAF1-N and -C silenced lines inoculated with different concentrations of Xav race 1 for 24 h, suggesting a lower level of resistance by the silencing of CaCAF1 in ‘ECW-30R’ pepper plants (Figure 6b).

Figure 6.

 Responses of CaCAF1-silenced ‘ECW-30R’ pepper plants to Xanthomonas axonopodis pv. Vesicatoria (Xcv) race 1 and 3.
(a) Bacterial growth in control, CaCAF1-N or -C silenced ‘ECW-30R’ pepper plants. CFUs of Xav race 3 were counted 0, 2, 4 and 6 days after infiltration. Data are expressed as the mean ± SE of three replicates comprised of three subreplicates.
(b) HR responses of CaCAF1-N or -C silenced ‘ECW-30R’ pepper plants inoculated with Xav race 1. Xav race 1 was infiltrated into 10 leaves per line, comprised of two leaves from five independent plants. The HR was observed 24 h after infiltration.

Altered gene expression in ‘MicroTom’ plants overexpressing CaCAF1

To elucidate the molecular basis of enhanced plant growth and disease resistance in CaCAF1-transgenic lines, global gene expression was analysed using microarrays. Microarray analyses of the control and CaCAF1-transgenic line 2 identified 68 genes out of 8700 total embedded genes that were upregulated by more than 2.8-fold. A 2.8-fold increase represents log2(2.8) = 1.5; thus, genes were selected based on the inequality log2(C2/C1) > 1.5 (C2 = expression level of a gene in transgenic plants; C1 = expression level of a gene in control plants). Natural variation could lead to values of log2(C2/C1) of <1.5; therefore, we selected only genes where log2(C2/C1) > 1.5 because these should represent truly higher expression rather than experimental error. These upregulated genes can be categorized into several distinctive functional groups (Table 1): cell-wall modification/biosynthesis (two genes), signalling components (10 genes), metabolism (22 genes), defence/stress response (two genes), photosynthesis (three genes), transcription factors (six genes), transport (three genes) and unknown functions (20 genes).

Table 1.   Upregulated genes in unstressed CaCAF1-transgenic tomato plants identified from a cDNA microarray embedding 8700 genes in total. Upregulated genes are those that were increased in expression by more than 2.8-fold in the CaCAF1-transgenic line 2 compared with the control plants. The induction change indicates the gene expression level in transgenic plants divided by expression in control plants
Functional categoryGene identificatione-ValueSGN no.Induction changePutative function
Cell-wall modification/ biogenesisCellulose synthase homologue3.00E−119U1451753.48Cell-wall biogenesis
Pectinesterase (pectin methylesterase) family2.00E−166U1440183.31Cell-wall biogenesis
Signalling componentPutative cell division control protein9.00E−108U1511213.13Cell division
Auxin-regulated protein7.60E−93U1473183.26Hormone responsive
Cig30U1452383.45Cytokinin inducible
Allene oxide cyclase5.00E−106U1474227.94Hormone responsive (JA)
Receptor kinase related1.30E−66U1510742.90Kinase
Putative serine/threonine protein kinase 1.30E−08U1510612.87Kinase
Putative Ca2+-dependent lipid-binding protein 2.00E−126U1473792.85Signalling
Armadillo repeat-containing protein1.40E−62U1564613.68Signalling
WD-40 repeat protein family8.70E−41U1526003.88Signalling
Tetratricopeptide repeat-containing protein2.40E−80U1477036.40Signalling
MetabolismS-adenosylmethionine synthetase 20U1441473.00Hormone synthesis
Arginine decarboxylase0U1436864.98Polyamine biosynthesis
Cinnamic acid 4-hydroxylase0U1438094.02Metabolism
β-Amylase7.00E−178U1442143.94Metabolism
Obtusifoliol-14-demethylase5.90E−63U1438743.43Metabolism
Putative 3-ketoacyl-CoA reductase 12.00E−122U1446855.06Metabolism
Anthocyanidin 3-O-glucosyltransferase0U1453713.52Metabolism
Phytoene synthase (EC 2.5.1.-)3.80E−42U1433963.45Metabolism
Phytone synthase 1, chloroplast precursor0U1434043.32Metabolism
PAPS-reductase-like protein0U1555678.29Metabolism
Putative epimerase/dehydratase 0U1448284.86Carbohydrate metabolism
Omega-3 fatty acid desaturase0U1445333.94Lipid/fatty acid metabolism
Putative acyl-CoA synthetase0U1431292.97Lipid/fatty acid metabolism
Fatty acid desaturase family protein2.00E−180U1477563.28Lipid/fatty acid metabolism
Bifunctional lysine-ketoglutarate reductase/saccharopine dehydrogenase0U1439303.14Amino acid metabolism
Serine acetyltransferase 43.00E−164U1453343.95Amino acid metabolism
Isoleucine-tRNA ligase homologue F24 G24.120 2.00E−112U1536055.39Amino acid metabolism
Putative peroxidasee-151U1438413.49Oxygen and radical metabolism
Probable phospholipid hydroperoxide glutathione peroxidase1.40E−93U1451323.20Oxygen and radical metabolism
Phenylalanine-tRNA synthetase-related protein0U1460684.53Protein biosynthesis
Cysteine proteinase5.80E−37U1527694.08Protein degradation
Proteasome subunit alpha type 46.00E−122U1457823.29Protein degradation
Defence/stress responseSimilar to hsr203 J0U1455793.81Defence
Chaperonin CPN60-2, mitochondrial precursor (HSP60-2)0U1453654.93Stress response
PhotosynthesisPutative early light-induced protein1.10E−45U1444104.88Light response
Plastidic aldolase-U1555553.32Photosynthesis
Thioredoxin peroxidase 11.60E−87U1449013.14Photosynthesis
Transcription factorsDof zinc finger protein3.10E−36U1498052.98Transcription
Myb-related transcription factor4.00E−68U1478925.74Transcription
Zinc finger protein1.10E−66U1504853.06Transcription
WRKY family transcription factor2.10E−83U1458104.65Transcription
CHP-rich zinc finger protein3.60E−34U1502242.91Transcription
Putative zinc transporter1.00E−109U1484903.12Transcription
TransportBoron transporter1.70E−94U1464193.00Transport
Putative CMP-sialic acid transporter8.00E−148U1497373.24Transport
Golgi SNARE protein, putative1.70E−66U1497833.33Transport
UnknownRemorin 12.40E−57U1454066.32Unknown
ENSANGP000000006402.10E−26U1596402.85Unknown
Expressed protein5.40E−55U1461894.30Unknown
Expressed protein2.30E−65U1469042.86Unknown
Expressed protein2.10E−08U1512823.79Unknown
Expressed protein6.60E−84U1515472.88Unknown
Expressed protein3.40E−91U1532223.75Unknown
Expressed protein9.10E−11U1561913.51Unknown
Expressed protein1.50E−05U1583782.85Unknown
Conserved hypothetical protein4.40E−25U1495633.51Unknown
Hypothetical protein7.00E−19U1469503.04Unknown
Hypothetical protein1.10E−96U1558203.49Unknown
No hits found U1508792.92Unknown
No hits found U1516473.14Unknown
No hits found U1517633.78Unknown
No hits found U1556092.97Unknown
No hits found U1560784.10Unknown
Unknown protein4.60E−45U1471243.39Unknown
Unknown protein3.00E−179U1435173.52Unknown
Unknown protein5.00E−06U1455903.37Unknown

To confirm the microarray analysis, RNA gel blot and real-time polymerase chain reaction (RT-PCR) analyses were used on selected genes involved in cell-wall/cell elongation and defence. RNA gel blot analysis showed that expression levels of tomato cinnamic acid-4-hydroxylase (C4H), arginine decarboxylase (ADC) and S-adenosyl methionine synthetase 2 (SAMS2) were highly upregulated in all the transgenic lines (Figure 7a). Furthermore, RT-PCR analysis of transgenic line 2 revealed that the cellulose synthase isolog, pectinesterase, and auxin-regulated protein genes were upregulated (Figure 7b).

Figure 7.

 Expression analysis of the genes upregulated by the overexpression of Capsicum annuum CCR4-associated factor 1 (CaCAF1) in ‘MicroTom.’
(a) Expression levels of cinnamic acid 4-hydroxylase (LeC4H), arginine decarboxylase (LeADC) and S-adenosyl methionine synthetase 2 (LeSAMS2) were measured using RNA gel blot analyses.
(b) Expression levels of genes associated with defence (allene oxide cyclase), growth (auxin-regulated protein) and cell-wall biosynthesis (pectinesterase and cellulose synthase isologue) were detected using real-time polymerase chain reaction.

Discussion

The role of CAF1 in plants was previously unknown, although there have been in-depth studies of the role of CAF1 in yeast and mammals because of its importance to cellular homeostasis. Previous studies in our laboratory indicated that chilli pepper responded to a non-host pathogen Xav 8ra by increasing CaCAF1 expression (Lee et al., 2004). CaCAF1, which was isolated from a chilli pepper cDNA library, contains one open reading frame encoding a protein with 266 amino acids that shares high homology with CAF1 in many eukaryotic organisms (Figure 1). Because the functional roles of CAF1 in plants have not been studied previously, we elucidated the roles of CaCAF1 using CaCAF1 overexpressing transgenic plants and CaCAF1-silenced plants via the VIGS technique.

CaCAF1 seems to be an important gene regulating the growth of plants. Overexpression of CaCAF1 in tomato plants resulted in increased plant heights accompanying with increased cell sizes and thickened cell walls (Figures 2a, 3, and 4). Silencing of CaCAF1 by VIGS in pepper plants, however, resulted in dramatically decreased plant height (Figure 2b). Both overexpression and silencing of CaCAF1 exerted effects on plant growth. Therefore, CaCAF1 seems to regulate the growth of plants. Interestingly, we found that overexpression of CaCAF1 in tomato plants also resulted in a more than doubled leaf thickness (Figure 3). Increased leaf thickness was induced by enlarged leaf cells and thicker cell walls. These results suggest that overexpression of CaCAF1 enhances plant growth through cell elongation in transgenic plants, and that turgor pressure, which should be lost when cells elongate, may be maintained by the thickened cell wall.

We also found that CaCAF1-transgenic tomato plants exhibited enhanced resistance to the oomycete pathogen P. infestans. Although the precise defensive role of the CaCAF1 gene product was not adequately elucidated, the thickened cell wall and cuticle could inhibit zoospore penetration, thereby increasing resistance to P. infestans infection. Silencing of CaCAF1 in pepper plants resulted in enhanced susceptibility to a virulent bacterial pathogen, Xav race 3, and also retarded HR to an avirulent pathogen, Xav race 1 (Figure 6). These results indicate that the CaCAF1 gene product plays an important role in increasing resistance to various types of pathogens. To better understand the underlying molecular basis for the growth and disease regulation abilities of CaCAF1, we analysed changes in gene expression profiles caused by ectopic expression of CaCAF1 in tomato plants. We found that the expression of 68 genes increased by more than 2.8-fold (Table 1). There are several interesting genes that might cause increased growth and thickened cell walls. Genes that might increase growth include ADC and SAMS (Table 1). ADC (EC 4.1.1.19) is an important enzyme responsible for putrescine biosynthesis (Walters et al., 2002). SAMS (EC 2.5.1.6) is an enzyme that catalyzes the formation of S-adenosylmethionine, which is the key intermediate for the biosynthesis of polyamines (Evans and Malmberg 1989) and also plays a central role in transmethylation reactions (Tabor and Tabor 1984). Polyamines stimulate many reactions involved in the synthesis of DNA, RNA and proteins (Walters, 2000). Therefore, the enhanced growth of plants and the elongation of mesophyll cells could be induced by the accumulation of polyamines caused by the upregulation of ADC and SAMS. The increase in cell-wall thickness and resistance to P. infestans might be caused by the upregulation of several genes, e.g. cellulose synthase homologue, pectinesterase, cinnamic acid 4-hydroxylase and peroxidase genes. The upregulation of these genes in CaCAF1-transgenic plants might contribute to thickening cell walls by forming cellulose by the cellulose synthase homologue, by enlarging cells and thickening cell walls by pectinesterase, by synthesizing lignins for the cell wall by cinnamic acid 4-hydroxylase, and by fortificating the cell wall by peroxidases.

Gene expression of CaCAF1 was upregulated by the exogenous application of salicylic acid (SA) and methyl jasmonate (MJ) (data not shown). Overexpression of CaCAF1 in transgenic tomato plants increased the expression of PR1 and PR6 regulated by SA and the expression of ADC and SAMS regulated by jasmonic acid (JA), suggesting that CaCAF1 might regulate several genes through SA- and/or JA-mediated signalling pathways. The expression of PR1 and PR6 depends on SA accumulation (Shin et al., 2001), and PR1 and PR6 are upregulated in CaCAF1-transgenic plants under normal growth conditions (Figure 5b). Exogenous application of MJ increased the expression of ADC (Biondi et al., 2001) and enzymatic activity (Walters et al., 2002), which eventually led to a significant reduction in powdery mildew infection. Exogenous application of MJ also increased the expression of SAMS2 (Imanishi et al., 1998). RNA gel blot analyses demonstrated that the mRNA accumulation of ADC and SAMS2 occurred in CaCAF1-transgenic plants (Figure 7a). Interestingly, we observed the strong upregulation in CaCAF1-transgenic plants of allene oxide cyclase (Table 1 and Figure 7b), which is involved in the biosynthesis of JA (Hause et al., 2003). CaCAF1, under the regulation of JA, may determine the level of JA by inducing the expression of allene oxide cyclase. These data strongly suggest that the CaCAF1-mediated defence reaction is possibly mediated through an SA- and JA-mediated defence signalling pathway.

In conclusion, overexpression of CaCAF1 in plants could activate several distinctive biological processes, such as increasing cell size, thickening cell walls and inducing polyamine defence mechanisms. Disease resistance against a broad range of pathogen infections and rapid growth might be induced by these biological processes. Alteration of plant height, and the enlarged cell size and thickened cell walls caused by the modulation of CaCAF1 expression, suggests that the CAF1 gene product is necessary for normal growth and development and for defence in plants.

Experimental procedures

Plant materials

Tomato (S. esculentum L. cv. ‘MicroTom’) seeds were obtained from the National Horticultural Research Institute (RDA, http://www.hridir.org/countries/korea/PROVCOUN/rural_development_administration_rda/index.htm). The seeds were surface-sterilized in 1% (v/v) NaOCl at 25°C for 10 min, followed by three washes with sterilized distilled water for 5 min. Seeds were then germinated on MS agar medium (Murashige and Skoog, 1962) and kept in a plant growth chamber under a 16-h photoperiod at 25°C for 2 weeks before being used for transformation.

For pathogen treatment, ‘MicroTom’ and pepper (C. annuum cv. ‘ECW-30R’) plants were grown in pots containing a sterile mixture of soil/vermiculite (2:1, v/v) and maintained under a 16-h photoperiod at 25°C for 4 weeks.

Plant expression vector construction and Agrobacterium-mediated transformation

CaCAF1 full-length cDNA (NCBI accession number DQ672569) was cloned into the plant expression pMBP1 vector. The recombinant plasmids contained the Cauliflower mosaic virus 35S promoter, a nopaline synthase terminator and a neomycin phosphotransferase gene. The recombinant vector was then transferred to Agrobacterium tumefaciens strain LBA 4404 by the freeze–thaw method (An et al., 1987). Cotyledons from 2-week-old ‘MicroTom’ plants germinated on MS media were used for co-cultivation with A. tumefaciens to generate transgenic plants, as described by Van Roekel et al. (1993).

Gel blot analysis

Total RNA was prepared using TRIzol® reagent (Gibco-BRL, http://www.invitrogen.com/) according to the manufacturer’s instructions. Total RNA (10 μg) from the empty vector-transformed ‘MicroTom’ plants and four independent T3 transgenic lines were separated on 1% formaldehyde agarose gel. The gel was blotted onto a nylon membrane and the membrane was hybridized to the 32P-labelled cDNA probes for CaCAF1, LePR1, LePR6, tomato cinnamic acid 4-hydroxylase, arginine decarboxylase and S-adenosyl methionine synthetase 2. All sequence information was obtained from the SOL Genomics Network (http://www.sgn.cornell.edu) at Cornell University (Ithaca, NY, USA). Equivalent RNA loading was demonstrated by ethidium bromide staining of rRNA on the gel.

Gene expression analysis by RT-PCR

Tomato cDNA was prepared using 1 μg of total RNA, 0.5 μg of oligo (dT) and 200 U of SuperScript II (Invitrogen, http://www.invitrogen.com). The tomato actin gene was amplified as the internal equal-loading control using a suitable primer set (forward, 5′-GATTTGCTGGTGATGATGCTCCTC-3′; reverse, 5′-CTAGCATACAGAGAAAGCACAGCC-3′). All other gene-specific primers used in this experiment were designed based on sequences in the Sol Genomic Network. The upregulated genes and the primer sets for RT-PCR were cellulose synthase isolog (SGN-U145175, 5′–GGGACAAAGGATACTGGAGTC-3′ and 5′-GCTTGTCTAGCTGAAGATGGA-3′), pectinesterase (pectin methylesterase) family (SGN-U144018, 5′–TCCCACGTCTTGCAGAACTTC-3′ and 5′-TTGGCCTAGCTGCATTTGATG-3′), auxin regulated protein (SGN-U147318: 5′–GATTCTGATGTGAGGCCTCTC-3′ and 5′–GCCAATGCCTTCGTTAGCTGC-3′), and allene oxide cyclase (SGN-U147422: 5′-AAGGAGATCGCTATGAAGCTG-3′ and 5′–CCCTTCCTCACAAGCTTTAGC-3′).

Semi-quantitative RT-PCR in a total volume of 30 μl was carried out using 1 μl of cDNA and 10 pmol of each oligonucleotide primer. The PCR profile was 3 min at 94°C; then 30 cycles of 94°C for 20 sec, 55°C for 30 sec, and 72°C for 1 min; followed by a final extension for 7 min at 72°C. The amplified products were separated on 1% agarose gel and visualized by ethidium bromide staining.

Hybridization and scanning of the cDNA microarray

The microarray experiments were replicated four times. Target solutions were prepared for the experiments constructed using total RNA prepared from CaCAF1-transgenic tomato plants (line 2) and control plants that were 4 weeks old. Total RNA (100 μg) was labelled by direct incorporation of Cy-3- or Cy-5-conjugated dUTP (Amersham Pharmacia Biotech, http://www.amersham.com), following the protocol of the Pat Brown Lab (http://cmgm.stanford.edu/pbrown/protocols/4_yeast_RNA.html). The combined labelled probes were purified using the QIAquick PCR Purification Kit (Qiagen, http://www.qiagen.com) and further concentrated to a final volume of 5 μl. For hybridization, the concentrated 5 μl of probes were mixed with 10 μl of formamide (Sigma-Aldrich, St. Louis, USA) and 5 μl of 2× hybridization solution (Amersham Pharmacia Biotech). The combined solution was denatured for 3 min at 95°C, then applied to tomato microarray slides representing 8700 genes (provided by Dr J. Giovanoni at Cornell University) and incubated in a water bath at 42°C for 16 h. The microarray slide was then washed at 55°C with each of 1× SSC/0.2% SDS, 0.1× SSC/0.2% SDS and 0.1× SSC for 10 min. The signal was scanned with an Axon GenePix 4000A scanner (Axon, http://www.axonglobal.com) to generate a TIFF image, and the Photomultiplier tube (PMT) voltage was adjusted to yield a Cy-3/Cy-5 signal intensity as close to 1.0 as possible. The spot intensities were measured using the axon genepix Pro 4.0 image analysis software, and global normalization was applied using the calculated ratio of the median factor (Lee et al., 2004). Several quality control methods were used. Firstly, all the spots flagged as ‘bad’ or ‘not found’ by the image analysis software were removed. Secondly, all the spots with regression square values <0.6 were removed. Third, all the spots with a signal-to-noise ratio <2.5 were removed. Finally, all the spots with validated data from replicated experiments were collected and averaged for further use. Upregulated genes that increased their expression levels by more than 2.8-fold were selected from the transgenic tomato plants. The genes were classified semi-automatically using the MIPS MatDB (http://mips.gsf.de/proj/thal/index.html) and KEGG (http://www.genome.jp/kegg/kegg2.html) databases.

Silencing CaCAF1 by the VIGS method

A stock solution of A. tumefaciens for VIGS was prepared as described by Liu et al. (2002). A 221-bp section of CaCAF1-N (silencing the CaCAF1 N-terminal side) and a 146-bp section of CaCAF1-C (silencing the CaCAF1 C-terminal side) were prepared by PCR with specific primers. The primers were as follows: CaCAF1-N (forward, 5′-AATGGATCCATCCTTTTCAGGATTCTGTTG-3′; reverse, 5′-CGCAAGCTTTTAGCCTTCAACAACTTGTACTG-3′) and CaCAF1-C (forward, 5′–AAAGGATCCGTGGCATGCATTTCAGAAAATG-3′ and reverse, 5′-CGCAAGCTTGAATAAATTAAATAACTTTTATT-3′). The amplified fragments were digested with BamH1 and HindIII (the underlined portions of the primers) and inserted into a pTRV2 vector. The pTRV2 vector with CaCAF1-N or CaCAF1-C was then introduced into A. tumefaciens strain GV3101. The recombinant A. tumefaciens strain GV3101 containing TRV:CaCAF1-N or -C construct was used for VIGS experiments according to the method described by Chung et al. (2004) and Ryu et al. (2004). Briefly, bacteria were grown at 28°C on Luria Bertani (LB) media with appropriate antibiotics, harvested by centrifugation, re-suspended into Agrobacterium inoculation buffer (10 mm MgCl2, 10 mm 2-(N-morpholine)-ethanesulphonic acid (MES), pH 5.6, 150 μm acetosyringone) to a final OD600 of 0.15, and incubated at 22°C with shaking for 4 h. The Agrobacterium with TRV1 and with TRV2:GFP, TRV2:PDS, TRV2:CaCAF1-N or TRV2:CaCAF1-C were mixed at a 1:1 ratio and then infiltrated into the cotyledons of 7-day-old ‘ECW-30R’ seedlings using a needleless 1-mL syringe. Each construct was used for silencing in 20 seedlings, and the seedlings were used after 4–6 weeks of VIGS treatment. All experiments were repeated at least twice.

Microscopic analysis

Leaves (30-days old) from the control and CaCAF1-transgenic lines 2, 3, 4 and 8 were observed under a light microscope and a transmission electron microscope. Leaves were cross-sectioned for imaging. The middle part of the leaf lamina was sectioned for transmission electron microscopy. The cell-wall structure was observed by transmission electron microscopy according to the methods of Rhee et al. (1998).

Pathogen inoculation

‘MicroTom’ plants (4-weeks old) were used for inoculation with P. infestans. P. infestans was grown on PDA media at 20°C for 5 days in darkness. Sporangia of P. infestans were harvested in sterile water and stimulated to release zoospores by incubation at 4°C for 2–3 h. After filtration through muslin, the resultant suspension was observed under a light microscope for quantification of zoospores and was then used for inoculation. After spraying the inoculum (a suspension containing 2 × 104 sporangia mL−1 of P. infestans using a fine glass optimizer), plants were placed in a moist chamber at 18°C. Ten independent plants from each line were infected, and disease symptoms appeared within 2–3 days after infection.

The bacterial pathogen Xav race 3 causing bacterial spot disease on ‘ECW-30R’ pepper plants was grown in YPD medium at 30°C. Suspensions (108 CFU.mL−1) of Xav race 3 were infiltrated into leaf mesophyll tissues of 4-week-old control (GFP-silenced) and CaCAF1-N- and -C-silenced ‘ECW-30R’ pepper plants using a needleless 1-mL syringe. To monitor bacterial growth in leaf tissues, the leaf tissues were ground with sterile water in a microfuge tube and spread on YPD media. The growth of bacterial colonies was measured based on the number of CFU on the selective medium 2, 4 and 6 days after infiltration. Xav race 1 (incompatible) was infiltrated at concentrations of OD600 = 0.02 and 0.04 into leaf mesophyll tissues of 4-week-old controls and CaCAF1-N- and -C-silenced ‘ECW-30R’ pepper plants. HR was visually documented 24 h after infiltration.

Acknowledgements

This work was supported by grants from CFGC (CG1431) and PDRC of the 21st Century Frontier Research Programs funded by MOST of the Korean Government.

Ancillary