Plant Cell Reports

, Volume 29, Issue 3, pp 231–238

Heterologous expression of taro cystatin protects transgenic tomato against Meloidogyne incognita infection by means of interfering sex determination and suppressing gall formation


  • Yuan-Li Chan
    • Institute of Plant BiologyNational Taiwan University
    • Agricultural Biotechnology Research CenterAcademia Sinica
  • Ai-Hwa Yang
    • Tainan District of Agricultural Improvement and Extension Station, Council of Agriculture
  • Jen-Tzu Chen
    • AVRDC-The World Vegetable Center
  • Kai-Wun Yeh
    • Institute of Plant BiologyNational Taiwan University
    • Institute of Plant BiologyNational Taiwan University
    • Agricultural Biotechnology Research CenterAcademia Sinica
    • Academia Sinica Biotechnology Center in Southern Taiwan
Original Paper

DOI: 10.1007/s00299-009-0815-y

Cite this article as:
Chan, Y., Yang, A., Chen, J. et al. Plant Cell Rep (2010) 29: 231. doi:10.1007/s00299-009-0815-y


Plant-parasitic nematodes are a major pest of many plant species and cause global economic loss. A phytocystatin gene, Colocasia esculenta cysteine proteinase inhibitor (CeCPI), isolated from a local taro Kaosiang No. 1, and driven by a CaMV35S promoter was delivered into CLN2468D, a heat-tolerant cultivar of tomato (Solanum lycopersicum). When infected with Meloidogyneincognita, one of root-knot nematode (RKN) species, transgenic T1 lines overexpressing CeCPI suppressed gall formation as evidenced by a pronounced reduction in gall numbers. In comparison with wild-type plants, a much lower proportion of female nematodes without growth retardation was observed in transgenic plants. A decrease of RKN egg mass in transgenic plants indicated seriously impaired fecundity. Overexpression of CeCPI in transgenic tomato has inhibitory functions not only in the early RKN infection stage but also in the production of offspring, which may result from intervention in sex determination.


PhytocystatinCeCPIRoot-knot nematodeTomatoGiant cells


Root-knot nematodes (RKN, Meloidogyne spp.) are one of the major plant-parasitic nematodes with the hosts of most flowering plant species, and cause global loss of agricultural crop yields annually (Bird and Kaloshian 2003). Meloidogyneincognita, the most widespread RKN species, is found in most of the tomato-growing area in the world and has impact effects on tomato yields (Castagnone-Sereno 2006). RKN is a sedentary parasite, whereby only the mobile second-stage (J2) juveniles possess the capacity to enter the plants by invading the elongation zone of roots. Following entrance, the juveniles first migrate to the root apex and later to the vascular cylinder to establish permanent feeding sites. The juveniles become sedentary in the feeding sites for nutrient uptake and undergo three molts to develop into adults with accompanying shape alteration (Atkinson et al. 1996). The males with a fusiform shape migrate out of the roots, whereas the immobile females become enlarged and have a saccate shape to produce the egg mass (Wyss et al. 1992; Williamson and Gleason 2003). The most obvious symptom in RKN-infected plants is the formation of galls in the roots, which subsequently affects the uptake of nutrients and water (Milligan et al. 1998) and thus decreases the yield of plants and increases the plant sensitivity to other pathogens (Castagnone-Sereno et al. 1992).

Several strategies have been applied to control RKN infection in agriculture, including the use of chemical nematicides. Although the application of chemical nematicides is effective, alternative strategies are required due to the concerns of high chemical toxicity to humans and the contamination of the environment (El-Alfy and Schlenk 2002). One of non-chemical alternatives is the use of soil solarization; however, Wang and McSorley (2008) suggested that the temperature at deeper soil depth was more important than the time period of heating soil in terms of suppressed nematode growth (Wang and McSorley 2008). Although solarization has been considered as an effective strategy for nematode management (Rosskopf et al. 2005), long time of treatments limit its application in agriculture. Another alternative is to produce RKN-resistant cultivars through genetic engineering. Previous studies have explored several genetic-engineering strategies, including avoidance of nematode invasion, disruption of feeding cell formation, and arrest of nematode growth or development (Fuller et al. 2008). Although the former two approaches seem more efficient for crop protection, the major limitation is lack of a promoter with specific activity in feeding cells. In contrast, the last approach, focusing on toxicity to nematodes or destruction of their digestibility, is easier to achieve. The major digestive enzymes in nematodes have been identified as serine and cysteine proteinases (Koritsas and Atkinson 1994); thus, the inhibitors of the two proteinases have been chosen as the targets for engineering RKN resistance in plants. Hepher and Atkinson (1992) reported that overexpression of cowpea trypsin inhibitor, an inhibitor of serine proteinase, retarded the early growth and affected the sexual fate of potato cyst nematode Globodera pallida in transgenic plants (Hepher and Atkinson 1992). Similar results were observed with the overexpression of sweet potato trypsin inhibitor in sugar beet hairy roots; the high trypsin inhibitory activity conferred high resistance to beet cyst nematode Heterodera schachtii (Cai et al. 2003). However, a recent study regarding serine protease inhibitors isolated from soybean suggested that they may not play an important role in the resistance to soybean cyst nematode infection (Rashed et al. 2008).

In addition to trypsin inhibitors, plant cysteine proteinase inhibitors (phytocystatin) are the other targets that limit the nutrient uptake of nematodes. A number of phytocystatin genes have been isolated from various plant species, including potato (Turrà et al. 2009) and taro (Yang and Yeh 2005), which have been shown to be involved in plant defense systems against biotic or abiotic stresses. However, only the rice phytocystatin gene Oc-1 has been demonstrated to be involved in the nematode defense mechanism (Urwin et al. 1995). Overexpression of Oc-I and a variant, Oc-IΔD86, in transgenic tomato hairy roots reduced the size of G. pallida females, with Oc-IΔD86 displaying even higher resistance than Oc-I (Urwin et al. 1995). In addition, transgenic Arabidopsis with overexpression of Oc-IΔD86 conferred the growth retardation of H. schachtii and M.incognita (Urwin et al. 1997). Although Oc-IΔD86 has conferred resistance to various nematode species, there appears to be a lack of literature regarding its application to transgenic tomato plants.

Tomato (Solanum lycopersicum Mill.) is one of the most economically important vegetable plant species because of its high nutritive value and diversified use. The worldwide production of tomato reached 126 million tons in 2007 ( Tomato is also one of the popular vegetables in Taiwan, and the production in 2007 was about 9.6 metric tons ( However, the growth and development of tomato are severely affected by environmental stresses. Tomato mainly grows in temperate areas, and a temperature below 15°C or above 30°C will cause a fertilization defect, thus reducing reproduction. In addition, infection with phytopathogens, including fungi, bacteria, insects, and nematodes, is another environmental cue restricting its growth and development. Due to the high temperature and humidity in Taiwan, the production of tomato is seriously reduced by heat and numerous pathogens (Chen et al. 2003). Therefore, the development of heat- and pathogen-tolerant tomato cultivars is desirable in Taiwan.

Over the past 30 years, AVRDC-The World Vegetable Center, in Taiwan, has developed a number of tomato cultivars, including CLN2468D, the plant material utilized in the present study, with heat and multiple pathogen resistance (Chen et al. 2003). Chen et al. (2003) reported that CLN2468D is highly resistant to tomato mosaic virus, tomato yellow leaf curl virus and Pseudomonas solanacearum. However, CLN2468D is susceptible to RKN infection.

In the present study, a taro phytocystatin gene, Colocasia esculenta cysteine proteinase inhibitor (CeCPI), was delivered into a tomato cultivar CLN2468D and the resistance of the T1 generation to RKN infection was evaluated. CeCPI comprises an ORF with 618 nt and encodes a protein of 29 kDa. It also contains two consensus motifs of phytocystatin: one is the reactive site Gln-Val-Val-Ser-Gly and the other is an ARFAV sequence (Yang and Yeh 2005). Amino acid sequence alignment showed that CeCPI shares 50.5% similarity with Oc-I, and the most similar region is the N terminus of CeCPI (Yang and Yeh 2005; Wang et al. 2008). Although CeCPI has shown anti-fungal activity in vitro (Yang and Yeh 2005; Wang et al. 2008), no in vivo study has validated its functions against nematode infection. In this research, we reported that transgenic tomato overexpressing CeCPI exhibited enhanced resistance to RKN infection by means of diminished gall numbers, decreased proportion of female nematodes, and reduced egg masses. In addition, the present study also demonstrated that CeCPI is the first phytocystatin to be involved in gall formation and sex determination of RKN post-infection.

Materials and methods

Plant materials

The nematode-susceptible tomato (Solanum lycopersicum Mill.) cultivar CLN2468D was used as the material for gene transformation. Seeds were kindly provided by AVRDC-The World Vegetable Center, Tainan, Taiwan. Seeds were surface sterilized with 1% NaOCl following the standard procedure and subsequently germinated on Murashige and Skoog (MS) basal medium with a 16-h photoperiod at 22°C.

Plasmid construction and transformation

DNA fragment amplification and sequence verification of CeCPI were done as described previously (Yang and Yeh 2005). The fragment was re-amplified with the addition of restriction enzyme sites XbaI and BamHI. A nos poly (A) sequence was subcloned into pUC18 as a SmaI and KpnI fragment to form pUC18/nos. The amplified CeCPI was digested with XbaI and BamHI and subcloned into pUC18/nos to form pUC18/CeCPI-nos. The cauliflower mosaic virus (CaMV35S) promoter was excised from pBI221 as a SphI and XbaI fragment and subcloned into pUC18/CeCPI-nos to generate pUC18/35S-CeCPI-nos. The cassette including the CaMV35S promoter, CeCPI and nos poly (A) was excised as a SphI and SmaI fragment and cloned into pCAMBIA1300 (Center for the Application of Molecular Biology of International Agriculture, Canberra, Australia) to obtain p35S-CeCPI (Fig. 1a). The p35S-CeCPI DNA was transformed into CLN2468D through Agrobacterium-mediated transformation (Chan et al. 2005).
Fig. 1

CeCPI integrated in the genome of transgenic tomato. a Schematic diagram of the structure of the binary vector p35S-CeCPI used in the transformation. HPTII: hygromycin phosphotransferase II cDNA sequence; CeCPI: taro cystatin cDNA. p35S: cauliflower mosaic virus 35S promoter; nos: nopaline synthase terminator sequence. b Southern blot analysis of tomato plants. Ten microgram of genomic DNA isolated from wild-type (WT) and transgenic plants (lines 3, 5, and 19) was digested with XbaI and underwent analysis with a DIG-labeled CeCPI cDNA probe

Molecular characterization of transgenic tomato plants

Preliminary screening of putative T0 transgenic plants was conducted by genomic PCR with a 5′ primer, 5′-TCTAGAATGGCCTTGATGGGGGGC-3′ located at the 5′ end of CeCPI, and a 3′ primer, 5′-TCGCAAGACCGGCAACAGGATTC-3′ located at the 5′ end of nos poly (A). For the T1 generation, genomic DNA isolated from leaves underwent Southern blot hybridization as described previously (Chan et al. 2005). In order to detect the expression level of CeCPI, total RNA was extracted from roots as described by Knapp and Chandlee (1996) and analyzed by northern blot hybridization. The full length of CeCPI cDNA fragment labeled with digoxigenin (DIG) was used as the probe for hybridization, and the signal was detected with the use of CDP-Star (New England BioLabs) and the FujiFilm luminescent image analyzer LAS-1000plus. Total protein was extracted from the roots of transgenic plants and analyzed on SDS-PAGE as described previously (Yang and Yeh 2005). The extracted total proteins then underwent western blot analysis with the use of antibody against purified CeCPI recombinant protein raised in New Zealand white rabbits as described previously (Yang and Yeh 2005).

Nematode assay

Second-stage juveniles (J2) of RKN were hatched from the egg mass collected from RKN-infected mungbean roots and used as inoculum for the infection. Seeds of transgenic tomato lines were selected on MS medium supplemented with 15 ppm hygromycin, whereas seeds of wild-type plants were sown on MS medium only. After germination at 22°C for 2 weeks, root weight of nine seedlings was recorded, while another six seedlings were transplanted to sandy soil. After acclimation in a walk-in chamber at 22°C for 2 weeks, each seedling was inoculated with approximately 1,000 J2 nematodes as described previously (Yaghoobi et al. 1995). Experiments involved three trials with six replicates in each trial. In order to analyze the penetration efficiency of RKN in test plants, the root tissues were collected 2 days post-infection at 28°C and stained with acid fuchsin as described previously (Atkinson et al. 1996) to calculate RKN numbers. While, for analysis of gall and egg mass production, the roots of test plants were collected 4 weeks post-infection at 28°C. The egg mass was counted on staining with 0.05% erioglaucin (Sigma) as a blue color (Omuega et al. 1988). The root tissues were then stained with acid fuchsin and nematodes were observed under a microscope as described previously (Atkinson et al. 1996) to analyze nematode growth and development. Based on the shapes, the observed RKNs were divided into 3 groups as described previously (Atkinson et al. 1996). In addition, the length and size of the observed RKNs were also noted. The experiment involved four replicates with 50 RKNs in each replicate.

Statistical analysis

Data were expressed as mean ± SD. Data analysis involved use of ANOVA, and means were compared by Scheffe’s test at the 5, 1, and 0.1% probability levels.


High expression of CeCPI protein in transgenic tomato

After hygromycin selection and regeneration, 13 putative transformants were obtained and identified as transgenic plants through PCR amplification of full length of CeCPI and the hygromycin-resistant gene hptII (data not shown). Wild-type and transgenic plants did not differ in morphological characteristics (data not shown). Transgenic lines 3, 5, and 19 were selected for further study due to the production of high seed set. Southern blot analysis of T1 generation revealed one to two insertions of CeCPI in the tomato genome (Fig. 1b). The highest expression of CeCPI, analyzed through northern blot hybridization, was observed in transgenic line 19 and the lowest expression in line 3 (Fig. 2a). Western blot analysis revealed the most abundant CeCPI protein in line 19 and the least in line 3 (Fig. 2b). Thus, all of the transgenic lines are independent transformants with abundant recombinant CeCPI proteins.
Fig. 2

High expression of CeCPI detected in transgenic tomato. a Northern blot analysis of CeCPI transcripts in tomato plants. Ten micrograms of total RNA extracted from the root tissues of wild-type (WT) and transgenic plants (lines 3, 5, and 19) was analyzed with use of the DIG-labeled CeCPI cDNA probe. Equal RNA loading was demonstrated by ribosomal RNA (rRNA) expression. b Western blot analysis of CeCPI in tomato plants. One hundred micrograms of total protein extracted from the roots of WT and transgenic plants was assayed with anti-CeCPI polyclonal antibody

Decreased gall numbers in transgenic tomato post-infection

No significant difference was observed in the root mass between wild-type and T1 transgenic plants grown on selection medium supplemented with hygromycin (Fig. 3a), indicating that hygromycin did not affect root growth of transgenic plants. In addition, there was no remarkable difference in the penetration efficiency of RKN between transgenic and wild-type plants 2 days post-infection (Fig. 3b). The result reveals that CeCPI did not inhibit RKN invasion. Following RKN invasion, galls were induced on the roots of wild-type and transgenic plants (Fig. 3c). However, less than 50% of gall numbers was observed in transgenic lines in comparison with wild-type plants. Although the least gall numbers were exhibited in line 3, no considerable difference was discovered among the three transgenic lines (Fig. 3c). The results suggest that nematode invasion may not be suppressed by the abundance of CeCPI; however, the numbers of root galls was dramatically reduced in transgenic plants.
Fig. 3

Transgenic tomato with overexpressed CeCPI repressed gall formation but not RKN penetration. a The fresh weight of roots of test plants, including 9 replicates in each line, were recorded 2 weeks post-germination on MS medium without hygromycin (wild type, WT) or with hygromycin (transgenic lines 3, 5 and 19). b The nematode numbers inside the roots was estimated in WT and transgenic T1 plants inoculated with M. incognita for 2 days. c Total number of galls was counted in WT and transgenic T1 plants post-infection. Means of the samples were analyzed by ANOVA. **Significant difference at P < 0.01

Reduced proportion of female RKN in transgenic plants

RKNs with fusiform, saccate, and enlarged saccate shapes were observed in both wild-type (Fig. 4a) and transgenic plants (Fig. 4b) 4 weeks post-infection. On analysis of RKN size, the length of fusiform RKNs did not significantly differ between the wild type and transformants (Fig. 4c). Similar results were detected for female RKNs with saccate and enlarged saccate shapes (Fig. 4c). Thus, the growth of RKNs with various shapes was not affected by CeCPI. However, the proportion of female nematodes, including those with saccate and enlarged saccate shapes, in transgenic plants was much lower than that in wild-type plants (Fig. 4d). Especially in line 5, only half of the RKNs were females. In contrast, the proportion of RKNs with a fusiform shape in transgenic plants was much higher than that in wild-type plants. Thus, the ratio of female to male RKNs was highly reduced in transgenic tomato.
Fig. 4

Low proportion of female M. incognita in transgenic tomato post-infection. RKN with various shapes was observed in wild-type (a) and transgenic plants (b). F fusiform, S saccate and ES enlarged saccate. c The length of RKNs with a fusiform shape or the size of the saccate shape was estimated in test plants. d The proportion of RKNs with a fusiform or saccate shape was detected in test plants. TS percentage of female nematodes with saccate and enlarged saccate shapes in the nematode group, TF percentage of nematodes with fusiform shape in the nematode group. *, **, and ***Significant difference at P < 0.05, 0.01, or 0.001, respectively

Suppressed reproductivity of RKN in transgenic tomato

After infection with RKN, much more egg masses were observed in wild-type plants (Fig. 5a) than they were in transgenic plants (Fig. 5b). Among the transgenic lines, line 19 showed the highest number of egg masses, which was only one-third of the numbers in wild-type plants, and line 5 displayed the least number of egg masses (Fig. 5c). The results demonstrated that CeCPI produced in transgenic tomato inhibited the reproduction of RKNs post-infection.
Fig. 5

Low abundance of egg masses in transgenic tomato post-infection. The number of egg masses in wild-type and transgenic plants were counted. Egg masses were observed in a wild-type or b transgenic root systems. c The number of egg masses in transgenic or wild-type plants was estimated. **, ***Significant difference at P < 0.01 or 0.001, respectively


Overexpression of a phytocystatin gene, CeCPI, isolated from an important staple food of Taiwan aborigines, the taro (C. esculenta), in transgenic tomato confers resistance to RKN infection. Transgenic tomato showed less than one-third of the egg masses produced in wild-type plants, which demonstrated a serious defect in the production of RKN offspring. Several generations of RKN can infect one crop plant within its life cycle (Ehwaeti et al. 2000) to cause severe impairment in the growth and development. Therefore, the reduction in egg mass is extremely important for engineering RKN resistance. On the basis of the model established by Ehwaeti et al. (2000), a decline in 60% of multiplication in three generations provided by a single transgene may be sufficient for effective resistance to RKN infection (Ehwaeti et al. 2000). In comparison with the model, transgenic tomato plants in the present study exhibit much better resistance since the egg mass production was decreased to 30% in only one generation. Although such results were not observed in transgenic Arabidopsis overexpressing Oc-IΔD86 against RKN infection (Urwin et al. 1997), a recent study of Oc-IΔD86 expression in potato identified a 50% decrease in RKN reproduction in transgenic plants (Lilley et al. 2004). It is difficult to evaluate whether better resistance can be provided by CeCPI or Oc-IΔD86 since different plant species were used in the studies.

The production of the egg mass is closely related to the number of female RKNs since the reproduction mode of M.incognita is parthenogenesis (as reviewed in Castagnone-Sereno 2006). However, juveniles develop as males rather than females under adverse environmental conditions (Triantaphyllou 1973). One of environmental cues to influence the sex determination is nutrient deficiency, which can be resulted from the inhibition of digestibility (as reviewed in Abad et al. 2003). In the digestion process of RKN, cysteine proteinases play an important role (Koritsas and Atkinson 1994), and its activity has been shown to be inhibited by the binding of phytocystatins to the active sites (Walker et al. 1998; Shingles et al. 2007). A cysteine proteinase, Mi-cpl-I, was isolated from RKN, and its expression was detected in the intestine of J2 and female nematodes thus suggesting its involvement in the digestion of RKNs (Neveu et al. 2003). Therefore, the sex determination of RKN in transgenic tomato may be influenced by nutrient shortage caused by the interference of CeCPI with the enzymatic activity of cysteine proteinases in the intestine of RKNs. In the present study, a dramatic decrease in the proportion of female RKNs was observed in transgenic tomato as compared with the wild type, demonstrated further that the sex determinism was greatly influenced by the overexpression of CeCPI. The results also revealed that the decreased egg mass production in transgenic tomato plants indeed resulted from the reduced frequency of female RKNs. Atkinson et al. (1996) reported a growth retardation and higher proportion of female RKNs in transgenic tomato hairy roots with overexpression of Oc-I in comparison with the control. In contrast, no growth arrest of female RKNs was noticed in the present study. Therefore, the mechanism of CeCPI involved in the engineering of RKN resistance may differ from that of Oc-I.

Overexpression of CeCPI also caused a pronounced reduction in gall numbers in transgenic tomato plants. Galls are composed of giant cells, the sites for permanent stay and a nutrition source of RKNs (Williamson and Hussey 1996). Therefore, the gall numbers are considered as an indicator of the capacity of RKN invasion in the early stage of infection. In the present study, galls were formed in transgenic tomato roots as well as in the wild type, which indicates that CeCPI did not function as an anti-invasion agent. The results were consistent with previous studies of Oc-I or Oc-IΔD86 used for genetically engineering nematode resistance (Urwin et al. 1995; Urwin et al. 1997). However, the present study showed that gall numbers were greatly reduced in transgenic tomato lines, an observation not found in previous reports of the engineering of Oc-IΔD86 for RKN resistance in transgenic potato or Arabidopsis (Urwin et al. 1997; Lilley et al. 2004). One of the possible reasons may be the reduced number of RKNs entered into roots. However, results in the present study indicated that the penetration efficiency of RKNs was not significantly different between transgenic and wild-type plants. The results demonstrated that CeCPI did not influence the penetration of RKNs. The possible mechanism behind the process is still unclear yet, and further study is required to elucidate the function of CeCPI in the gall formation.

Phytocystatin is present in various plant species, including potato tubers (Rodis and Hoff 1984), rice, and maize (Richardson 1991). A study regarding the toxicity of Oc-IΔD86 demonstrated that daily consumption of 10 mg recombinant protein per kg body weight for 28 consecutive days was not toxic to mammals (Atkinson et al. 2004), which confirmed its biosafety. Therefore, the manipulation of RKN with genetic engineering of phytocystatins in crop plants is a promising alternative in agriculture. In addition to Oc-IΔD86, the taro CeCPI studied in the present article is another effective anti-nematode phytocystatin for RKN management, functioning not only in the early stage of RKN infection but also in the reproduction stage to reduce the population of the next generation. Moreover, our study also demonstrated that CeCPI is the first phytocystatin involved in the formation of root galls and intervention of sex determination.


Kai-Wun Yeh and Ming-Tsair Chan contributed equally to this study. This study belongs to the TW-SOL project (The International Solanaceae Genome Project—Taiwan) and was supported by grants from the National Science and Technology Program for Agricultural Biotechnology [numbers 95AS-6.2.1-ST-a1 (46), 96AS-1.2.1-ST-a2 (13), and 97AS-1.2.1-ST-a3 (13)] to Dr. Kai-Wun Yeh.

Copyright information

© Springer-Verlag 2010