The rapid accumulation of defensive transgene products in plants only on pathogen invasion has clear advantages over their constitutive synthesis. In this study, two antimicrobial peptides from the skin secretions of frogs, MsrA2 (N-methionine-dermaseptin B1) and temporin A, were evaluated for engineering pathogen-induced disease resistance in plants. Both peptides inhibited plant-specific pathogens in vitro at micromolar concentrations that were not toxic to plant protoplasts. The plant-optimized nucleotide sequences encoding MsrA2 and temporin A were transcriptionally fused to the inducible win3.12T poplar promoter, which had strong systemic activity in response to fungal infection, and introduced into tobacco (Nicotiana tabacum L. cv. Xanthi). Transgene expression was very low in the leaves of unstressed plants; however, it was strongly increased after pathogen challenge or wounding. The pathogen responsiveness of the win3.12T promoter was found to be universal rather than species specific, with high activity in response to all pathogens tested. On induction, the amount of MsrA2 was up to 6–7 µg per gram of fresh leaf tissue. Most importantly, the induced accumulation of MsrA2 and temporin A in transgenic tobacco was sufficient to confer resistance to a variety of phytopathogenic fungi, such as Fusarium solani, F. oxysporum, Alternaria alternata, Botrytis cinerea, Sclerotinia sclerotiorum, the oomycete Pythium aphanidermatum and the bacterium Pectobacterium carotovorum. The accumulation of these peptides in transgenic plants did not alter the normal phenotype of tobacco. Thus, the expression of MsrA2 and temporin A in a pathogen-inducible manner enables the development of tobacco, and possibly other plant species, with wide-spectrum disease resistance, which can reduce the use of pesticides and the associated environmental risks.
Pathogenic fungi, oomycetes and bacteria are the leading causes of plant diseases, contributing to more than one-third of total crop losses throughout the world. The application of chemicals to control plant diseases continues to play a key role in meeting world food needs. However, the constant emergence of new, more tolerant, pathogen strains requires more pesticide applications, which, in turn, have a highly negative long-term impact on human health and the environment. The logical way to fight plant diseases effectively is to create plants that are resistant to these pathogens. Modern gene transfer technologies and the ability of plant tissue to regenerate from a somatic cell have revolutionized plant breeding programmes, providing a unique opportunity to introduce practically any gene of interest, from any source, into plant genomes without the loss of existing valuable traits of established cultivars.
The objective of our study was to compare the antimicrobial activities of two cationic peptides, MsrA2 and temporin A, and to test whether the accumulation of these AMPs in plants in a pathogen-inducible manner has the potential to enhance host resistance against phytopathogens. Both MsrA2 and temporin A belong to a large group of linear amphipathic α-helical AMPs. The MsrA2 peptide contains the full-length amino acid sequence of dermaseptin B1 (31 residues), preceded by methionine. The parental dermaseptin B1 is a larger precursor of dermaseptin b (27 amino acids) that was found in the skin secretion of the arboreal frogs Phyllomedusa sauvagei and Phyllomedusa bicolor (Mor et al., 1994; Fleury et al., 1998). Dermaseptin B1 has the best antimicrobial potency of all the dermaseptins, inhibiting numerous species of bacteria and fungi in vitro (Strahilevitz et al., 1994). Despite the activity on microbial membranes, dermaseptins do not lyse erythrocytes or other mammalian cells (Strahilevitz et al., 1994). Our group has shown previously that the constitutive CaMV 35S-driven expression of the MsrA2-encoding gene in potato increases plant resistance to a variety of potato-specific pathogens (Osusky et al., 2005). However, the use of the viral promoter and the rather limited number of transgenic lines in that report prompted us to extend the study of MsrA2 expression in plants. Another AMP, temporin A, is one of the smallest AMPs found in nature (only 13 amino acids), and was originally isolated from the skin secretion of the European red frog Rana temporaria. It has antibacterial activity, mostly against Gram-positive bacteria (Simmaco et al., 1996; Harjunpääet al., 1999). By contrast with structurally similar haemolytic peptides from wasp venom, temporin A is not toxic to human red blood cells (Argiolas and Pisano, 1984). Despite its good antimicrobial properties, the expression of temporin A in plants has not been reported previously.
In this study, the plant-optimized nucleotide sequences encoding MsrA2 and temporin A peptides were introduced into tobacco. Furthermore, their expression was regulated by the wound-inducible win3.12T poplar promoter that has powerful systemic activity in response to fungal infection (Yevtushenko et al., 2004, 2005). We report that both MsrA2 and temporin A peptides have strong antimicrobial activity against plant pathogens in vitro, inhibiting the germination of conidia and bacterial growth at physiologically low concentrations that are non-toxic to plant cells and comparable with the win3.12T-regulated accumulation of MsrA2 in vivo. We present data on transgene expression in response to wounding and pathogen infections, and the relationships between the number of transgene insertions, the level of AMP transcripts and the amount of accumulated AMPs in plants. Finally, we show that, in plants, the combination of a strong pathogen-inducible promoter with the genes encoding MsrA2 and temporin A peptides generates a level of transgene expression and AMP accumulation that is sufficient to confer resistance to fungal, oomycete and bacterial pathogens.
Antimicrobial activity of MsrA2 and temporin A in vitro
AMP activity in vitro was evaluated to obtain preliminary information on whether these peptides might be useful to provide disease resistance in plants. Chemically synthesized MsrA2 and temporin A peptides were tested against the following important plant pathogens: Fusarium solani, F. oxysporum, Alternaria alternata, Botrytis cinerea and Pectobacterium carotovorum. Dose–response curves were plotted for each peptide–pathogen combination using the results obtained after 24 h of incubation of fungal conidia or bacterial cells with different peptide concentrations (Figure 1). The curve-derived inhibitory AMP concentrations are summarized in Table 1, where IC50 indicates the concentration of peptide sufficient to inhibit the germination of 50% conidia (or reduce the number of bacterial colonies by 50%) and MIC (minimal inhibitory concentration) is defined as the lowest peptide concentration that completely inhibits the germination of conidia (or bacterial growth). All tested pathogens were sensitive to MsrA2 and temporin A at low micromolar peptide concentrations. MsrA2 was found to be more effective than temporin A against both fungal and bacterial pathogens, requiring less than half of the temporin A concentration to inhibit the germination of conidia; the difference between these AMPs with regard to the inhibition of bacteria was even higher (Table 1). Although these pathogens represent evolutionarily distant groups of organisms (fungi vs. bacteria), belong to different genera and differ in the size and morphology of their conidia, the values of the inhibitory concentrations of each AMP were relatively similar, with the exception of Gram-negative P. carotovorum, which needed a larger amount of temporin A to inhibit its growth (Table 1). The observation of fungal cultures grown at different peptide concentrations revealed that, although sublethal levels of AMPs (below MIC) did not stop the germination of all conidia, they significantly delayed the mycelial growth of the conidia that germinated, and often caused many abnormalities of fungal morphology, such as hyperbranching, shortening and swelling of the hyphae. The effect of MsrA2 and temporin A on the conidia germination of some fungal species is shown in Figure 2.
Table 1. Antimicrobial activities of MsrA2 and temporin A peptides in vitro against phytopathogenic fungi and bacteria. Average IC50 and MIC values were derived from dose-response curves obtained by analysis of three independent experiments
Temporin A (µm)
When the above samples were incubated for longer periods, none of the pathogens could resume growth at concentrations above the corresponding MICs, even after more than 1 week of incubation, and very little additional germination or growth was observed at concentrations below the corresponding MICs but above IC50. The latter suggests that both AMPs were relatively resistant to degradation by secreted pathogen proteinases, or that the pathogens could not recover from the initial damage. Boiling of MsrA2 and temporin A for 10 min did not inactivate them, although some loss of antimicrobial activity was detected for both peptides: about 20% for MsrA2 and less than 10% for temporin A. The higher tolerance of temporin A to heat inactivation may be explained by its smaller size compared with MsrA2 (14 vs. 32 amino acids, respectively).
Phytotoxicity of MsrA2 and temporin A
The application of MsrA2 and temporin A to control plant diseases in vivo is feasible only if these peptides have no deleterious effect on the host organism. To address this issue, we evaluated the viability of non-transgenic tobacco leaf protoplasts cultivated in liquid medium in the presence of AMPs. Because of the absence of the protective cell wall, plant protoplasts are much more sensitive than any other plant tissue to the possible phytotoxicity of AMPs (Yevtushenko et al., 2005), and, by contrast with pollen, they are available from in vitro plants throughout the year. Protoplast viability was determined by staining cell culture with neutral red or Evan's blue dye: the former accumulated in the vacuoles of living cells, whereas the latter stained dead cells (Figure 3). After incubation of freshly isolated mesophyll protoplasts with different amounts of AMPs for 24 h, no visible toxic effects on protoplast viability were observed up to concentrations of 5 µm MsrA2 and 30 µm temporin A. Above these levels, the protoplast viability was gradually reduced, and none of the protoplasts survived at peptide concentrations equal to or higher than 15 µm MsrA2 and 80 µm temporin A. Three weeks after the addition of AMPs to the protoplast culture, the overall rate of cell division and the number of calli formed correlated with the viability of protoplasts observed in the same cultures after 24 h, suggesting that prolonged incubation of plant cells with AMPs did not reduce cell viability, although we cannot exclude the possibility of reduced AMP concentration caused by released plant proteases. Interestingly, when AMPs were added to 5-day-old protoplast culture, the viability of the plant cells, which had already formed new cell walls and started first cell divisions, was found to be similar to that of freshly isolated protoplasts. Although MsrA2 was about twice as potent as temporin A against plant pathogens, it was also more toxic to plant protoplasts: the concentrations of MsrA2 that affected protoplast viability were up to six times lower than those of temporin A. However, the minimal phytotoxic concentrations of each AMP were higher than the peptide levels required to completely inhibit the germination of conidia and bacterial growth: twofold for MsrA2 and sixfold for temporin A.
Transformation with AMP-encoding genes and characterization of transgenic plants
The nucleotide sequences encoding MsrA2 and temporin A were inserted into plant expression vectors under the control of the 823-bp pathogen- and wound-inducible win3.12T promoter from poplar (Figures 4a,b and 5a,b), and introduced into tobacco (Nicotiana tabacum L. cv. Xanthi) via Agrobacterium-mediated transformation. Over 50 plantlets regenerated on kanamycin-containing medium were excised from the green calli in each experiment, rooted in the presence of selective agent and tested for the absence of Agrobacterium. All regenerated plants were morphologically normal, except for two lines that had symptoms of vitrification. Polymerase chain reaction (PCR) analyses with primers designed to amplify the promoter–transgene region showed the presence of transgenes in all regenerated plants (Figures 4c and 5c). Ten lines with each DNA construct, win3.12T-MsrA2 or win3.12T-TA, were randomly selected from the primary transformants, propagated in vitro and used for further studies at the stage of fully developed 4–5-week-old plants. Southern blot analyses confirmed the integration of the AMP expression cassettes into the plant genomes, and revealed a large difference in the number of transgene insertions among the lines, ranging from one to 11 copies per genome (Figures 6a and 7a).
Expression pattern of MsrA2 and temporin A genes in plants and accumulation of MsrA2 peptide
The next step was to examine whether the introduced AMP genes were expressed in plants, and to determine the levels of MsrA2 and temporin A transcripts generated from the win3.12T promoter. Six transgenic lines representing the full spectrum of transgene copy numbers (from single to the highest) were selected for each AMP construct and used in Northern blot analyses. Because the win3.12T promoter is induced by wounding (Yevtushenko et al., 2004), the accumulation of AMP messenger RNAs (mRNAs) was analysed by hybridization of transgene-specific antisense probes with total RNA extracted from both unwounded and wounded leaf tissues. As predicted, the activity of the win3.12T promoter in the absence of stimuli was very low. By contrast, the majority of lines showed a significant level of transgene expression on wounding (Figures 6b and 7b). No AMP mRNA was detected in non-transgenic controls. As the total RNA was isolated from the same leaf, both before and after wounding, and Northern analyses were independently repeated with similar results, we believe that this provided an accurate picture of the quantitative accumulation of win3.12T-regulated AMP transcripts in different transgenic lines. Comparison of the signal intensity of the mRNA bands with the Southern data revealed an interesting correlation between the wound-induced accumulation of AMP transcripts (Figures 6b and 7b) and the number of transgene insertions (Figures 6a and 7a). In lines with one to five transgene copies, the level of induced AMP mRNA was directly proportional to the number of transgene insertions, with the highest AMP expression found in lines Twm10 and Twt8 that contained a maximum of five transgene copies. These data were similar to those found in our previous studies of the win3.12T promoter (Yevtushenko et al., 2004, 2005), in which the highest number of transgene insertions was four or five. However, a further increase in the number of transgene copies to seven and eight per genome (lines Twm12, Twt2 and Twm3, Twt5) resulted in a gradual decrease in win3.12T-driven AMP transcripts (Figures 6b and 7b). In line Twt11, with 11 transgene insertions (the highest copy number in this study), the wound-induced accumulation of AMP mRNA was hardly detectable, being practically the same as in the absence of stimuli (Figure 7b).
The win3.12T promoter has strong systemic activity in the aerial parts of plants in response to fungal infection (Yevtushenko et al., 2004, 2005). Because transgenic win3.12T plants in those studies were tested only against one species of pathogen, namely F. solani, we wanted to be sure that this promoter was also responsive to infections caused by other pathogens, including fungi, oomycetes and bacteria. Therefore, plants of a representative transgenic line Twt8, which had the highest wound-induced accumulation of transgenic mRNA among the win3.12T-TA lines, were inoculated with different pathogens and examined for any change in the expression level of temporin A transcripts. Northern blot analysis of leaf RNA confirmed the high induction of transgene expression in response to pathogen infection in all transgenic plants, regardless of pathogen species, indicating that the pathogen responsiveness of the win3.12T promoter was universal rather than species specific (Figure 8).
To verify that the AMP transcripts were correctly processed into peptides and not degraded by host proteases, protein extracts were prepared from leaves of control and transgenic MsrA2 plants (both before and after pathogen infection), separated by acid urea polyacrylamide gel electrophoresis (AU-PAGE) and subjected to Western blot analysis with polyclonal antibodies raised against MsrA2 peptide. The MsrA2 peptide was detected in all transgenic lines that were infected with a pathogen, whereas no signal was found in protein samples from non-transformed controls (Figure 9). In addition, no MsrA2 was detected in a representative transgenic line Twm10 in the absence of stress stimuli (Figure 9), although a low level of transgene expression was observed in leaf RNA of unstressed transgenics (Figure 6b). If there was MsrA2 peptide in the unstressed transgenic plant, it was beyond the sensitivity of our detection system. Variations in the amount of MsrA2 peptide among the transgenic lines roughly reflected the different levels of MsrA2 expression found through Northern blot analysis (Figure 6b). Comparison with synthetic MsrA2 of known concentration revealed that the pathogen-induced accumulation of MsrA2 peptide in the leaves of transgenic tobacco was about 3–7 µg per gram of fresh tissue.
Evaluation of disease resistance
Detached leaves of fully developed transgenic (10 lines of each construct) and control plants were challenged with different pathogens, observed daily for the development of disease symptoms and scored for resistance as described in the ‘Experimental procedures’ section. Seven species of major tobacco pathogens were used in plant resistance bioassays: F. solani, F. oxysporum, A. alternata, B. cinerea, Sclerotinia sclerotiorum, Pythium aphanidermatum and P. carotovorum. With most pathogens, susceptible leaves developed visible disease lesions within 2–3 days after inoculation. Although lesions grew over time in both transgenic and untransformed leaves, their size and growth in most transgenic lines were significantly reduced relative to the controls. Lesion development varied among the leaves and depended on two key factors: (i) the level of AMP expression in a given transgenic line, and (ii) the virulence of a particular pathogen. Leaves from transgenic lines Twm10 (MsrA2 plants) and Twt8 (temporin A plants), which had the highest transgene expression and AMP accumulation among the lines, showed the lowest disease severity of all leaves tested (Figure 10). Even when lesions appeared on the leaves from Twm10 and Twt8 plants, they were confined strictly to the site of inoculation. By contrast, non-transgenic leaves exhibited large expanded lesions with rotting tissues and water-soaking symptoms (Figure 10). Both lines were scored as resistant against highly virulent strains of F. solani, F. oxysporum, B. cinerea and P. carotovorum, and highly resistant against less virulent A. alternata, S. sclerotiorum and P. aphanidermatum. Among other transgenic lines, leaves of Twm6, Twm8, Twm12, Twt2 and Twt4 plants were moderately resistant, with a total lesion area not exceeding 50% of that in non-transgenics, whereas leaves of Twm1, Twm3, Twt3, Twt5 and Twt11 plants were found to be susceptible to most pathogens, with no significant difference from the controls. Of the MsrA2- and temporin A-expressing plants (by comparing the best lines Twm10 and Twt8, each containing five transgene copies), the former showed higher antimicrobial resistance in some experiments, although the difference was not statistically significant (P > 0.1, by Student's t-test).
In another series of experiments, disease resistance of whole plants was evaluated by growing them in vitro in the presence of each of seven pathogens. These stringent biotests were based on the ability of resistant plants to survive infections much longer than non-transgenic plants. Because of differences in pathogen virulence, the time for development of visible disease symptoms in non-transgenic plants after contact with a pathogen varied from 1–2 days (Fusarium species, B. cinerea, high inocula of P. carotovorum) to more than 1 week (A. alternata, P. aphanidermatum). Nevertheless, all infected control plants exhibited severe wilt, discoloration of leaves and rotting of the entire plant within a much shorter period than most of the transgenic lines (Figure 11). With fungal infections, localized browning of the plant stem at the base, where initial contact between host and actively growing mycelia occurs, served as the first indicator of plant susceptibility. Similar to the bioassays of detached leaves, plants of transgenic lines Twm10 and Twt8 showed the strongest resistance of all the transgenics: they survived infection with F. solani, F. oxysporum and B. cinerea for more than 12 days (five to six times longer than non-transgenic plants). Moreover, no disease symptoms were observed in these lines after inoculation with A. alternata or P. aphanidermatum: the plants remained green, healthy and continued to grow even after 1.5 months of cultivation with the pathogens. In the other lines, Twm6, Twm8, Twm12, Twt2 and Twt4 plants were moderately resistant, and Twm1, Twm3, Twt3, Twt5 and Twt11 were scored as susceptible or moderately resistant, depending on the virulence of the pathogen. Overall, the observed spectrum of disease-free survival among the plants was similar to that of the detached leaves, and directly correlated with the data from Northern (MsrA2 and temporin A plants) and Western (MsrA2 plants) blot analyses: the best disease resistance was found in those transgenic lines that had the highest level of induced transgene expression and AMP accumulation.
In this study, the antimicrobial activities of two amphibian skin AMPs, MsrA2 (dermaseptin B1 derivative) and temporin A, were evaluated for the genetic engineering of pathogen-resistant plants. MsrA2 and temporin A have a variety of characteristics which suggest that they can be successfully used to confer resistance to a broad spectrum of plant pathogens. First, and most importantly, both MsrA2 and temporin A belong to an evolutionarily ancient group of highly efficient defensive peptides that target not a specific receptor, but the major structural features of the microbial membrane that distinguish microorganisms from plants and animals (Zasloff, 2002). Because the resistance of microorganisms to AMPs would require significant redesign of the microbial membrane, including changes to its lipid composition (Zasloff, 2002), the emergence of MsrA2- or temporin A-resistant pathogens is highly improbable. This defence mechanism has helped the host species (frogs) thrive in a pathogen-infested environment for millions of years. Second, it was shown here that both MsrA2 and temporin A have strong antifungal and antibacterial activities in vitro against plant-specific pathogens at concentrations (2.5–5.5 µm, MICs) that are within the typical micromolar range observed for the accumulation of recombinant proteins in transgenic plants. Third, the accumulation of MsrA2 and temporin A peptides in transgenic tobacco had no deleterious effects on plant growth and development, even in those lines with the highest level of transgene expression. The low phytotoxicity of MsrA2 and temporin A was also confirmed in highly sensitive in vitro bioassays with tobacco leaf protoplasts: the peptide concentrations that affected the viability of plant protoplasts were at least several times higher than those that completely inhibited the germination of conidia and bacterial growth. Intact plant tissues and organs were even more tolerant to AMPs than protoplasts. Fourth, MsrA2 and temporin A appear to be resistant to degradation by pathogen proteases, as prolonged incubation with the pathogens did not result in a loss of antimicrobial activity of AMPs (at least under the conditions used in this study). In addition, these peptides were heat stable and retained their antimicrobial activities after boiling. Both MsrA2 and temporin A showed greater resistance to digestion by proteases than did cecropin A–melittin (CEMA) hybrid AMP, described in our previous report (Yevtushenko et al., 2005). The stability of AMPs – their resistance to relatively rapid degradation by pathogen- or plant-secreted proteases – is vital for the generation of antimicrobial action during peptide–pathogen interaction. One of the reasons for the low disease resistance, reported in some studies on AMP-expressing transgenic plants, was the degradation of the peptides (mostly cecropin-based) by endogenous host proteases (Hightower et al., 1994; Allefs et al., 1995; Florack et al., 1995).
Apart from AMP properties, the choice of a suitable promoter with an appropriate level of spatio-temporal activity is an important consideration for the successful engineering of disease-resistant plants. In most previous studies, the expression of AMP genes in plants was controlled by the strong constitutive CaMV 35S promoter and its derivatives. However, continuous synthesis and accumulation of AMPs in the absence of pathogen invasion, i.e. when no AMP is needed, will result in permanent long-term exposure of the host plant to AMPs, which may negatively affect plant functions at otherwise non-phytotoxic AMP concentrations. In addition, constitutive expression of defensive compounds may cause an undesirable selection pressure on pathogen populations (Jouanin et al., 1998). Most importantly, constitutive promoters, especially those of non-plant origin, seem to be frequently associated with silencing effects (Matzke and Matzke, 1995; Puddephat et al., 1996), thereby reducing the amount of AMPs in plants to a level insufficient to kill the pathogen. Indeed, recent data on the expression of a synthetic cecropin B analogue (Shiva-1) in plants (Yi et al., 2004) have suggested that the low level of AMPs in transgenic plants and, subsequently, the low disease resistance observed in some studies on the expression of natural and synthetic cecropins in plants, is caused not only by the proteolytic activity of host proteases, but also by the use of the constitutive viral promoter. Although a number of studies on the expression of AMP genes under the control of a constitutive promoter have been successful (Gao et al., 2000; Ponti et al., 2003; Osusky et al., 2004; Rajasekaran et al., 2005), the use of an inducible plant promoter that activates only in response to pathogen invasion or pest attack in a predictable spatio-temporal manner has clear advantages, and is preferable for the regulation of AMP transgenes in plants. In the present study, we used the inducible 823-bp win3.12T promoter from hybrid poplar (Populus trichocarpa × Populus deltoides) to direct the expression of MsrA2- and temporin A-encoding genes in tobacco. This promoter is a truncated version of the original 1352-bp promoter of the wound-inducible win3.12 gene (Hollick and Gordon, 1993, 1995), which encodes a Kunitz-type proteinase inhibitor and has a developmental pattern of expression similar to that of storage protein genes. Previously, we have shown that the win3.12T poplar promoter has strong systemic activity in the aerial parts of heterologous transgenic plants (potato and tobacco) in response to mechanical wounding (mimicking pathogen invasion or insect chewing) and infection with the pathogenic fungus F. solani (Yevtushenko et al., 2004, 2005). Here, we confirmed the high accumulation of win3.12T-driven MsrA2 and temporin A transcripts in transgenic leaves after wounding or pathogen infection. Moreover, the pathogen-induced activity of the win3.12T promoter was not found to be species specific, but rather universal, with high responsiveness to all species of phytopathogens used in this work. Earlier (Yevtushenko et al., 2005), comparative analysis of the promoter sequence for the presence of putative pathogen-responsive cis-acting elements revealed five clustered copies of a W-box motif with a TGAC core sequence that confers elicitor/pathogen responsiveness in other plant promoters (Yang et al., 1999; Yu et al., 2001). The W-box is the binding site for members of the WRKY family of transcription factors (Rushton et al., 1996). Activation of the win3.12T promoter by different pathogens suggests that the pathogen-responsive regulatory elements of this promoter interact with a host transcription factor (most likely, a WRKY DNA-binding protein) common to all pathogen-induced responses, irrespective of pathogen nature.
In this study, the ability of transgenic tobacco to withstand pathogen infections was directly proportional to the level of pathogen-induced expression and accumulation of MsrA2 or temporin A: transgenic lines with the highest levels of transgene expression survived both fungal (F. solani, F. oxysporum, B. cinerea) and bacterial (P. carotovorum) infections much longer than lines with low transgene expression or non-transgenic controls. Furthermore, almost no disease symptoms were observed in bioassays of the best lines with pathogens A. alternata, S. sclerotiorum and P. aphanidermatum. The amount of pathogen-induced accumulation of MsrA2 peptide in the transgenic line with the highest score of resistance (Twm10) was 6–7 µg per gram of fresh leaf tissue (~1.8–2.0 µm). The induced level of MsrA2 was higher than IC50 for all pathogens tested in vitro (1.0–1.6 µm), and very close to the MICs for most pathogens (2.5 µm). It should be noted, however, that situations in vitro and in vivo are substantially different, and the antimicrobial potency of AMPs in plants may be partially inhibited by certain intracellular compounds (salts, divalent ions, etc.). By contrast, a plant's endogenous defence compounds produced during plant–pathogen interactions can compensate for the reduced AMP activity in vivo. Nevertheless, 1.8–2.0 µm of MsrA2 was sufficient to confer resistance of transgenic tobacco against all phytopathogens tested. We could not determine the induced level of temporin A in plants because of the absence of reliable primary antibodies for this peptide. However, because temporin A was more stable than MsrA2 in vitro, and the levels of induced MsrA2 and temporin A transcripts in the lines with highest transgene expression were comparable (lines Twm10 and Twt8, each with five transgene copies), the amount of pathogen-induced accumulation of temporin A peptide in the leaves of the most resistant line Twt8 was probably 2 µm or higher.
When comparing MsrA2 and temporin A, MsrA2 had twofold greater antimicrobial potency than temporin A in vitro, but temporin A was up to six times less toxic to plant protoplasts. Although, in some plant resistance bioassays, transgenic plants with MsrA2 peptide performed slightly better than plants with temporin A, no statistically significant advantage of the use of one peptide over another was observed. A sufficient level of MsrA2 or temporin A, rather than the type of AMP, was the decisive factor for the ability of transgenic plants to resist pathogens. It remains to be tested whether these two peptides could act synergistically in plants.
Comparison of the win3.12T-regulated expression pattern with previous studies (Yevtushenko et al., 2004, 2005) confirmed that the level of induced transcription from this promoter did not depend on the nature of the transgene or host organism. However, it correlated with the number of T-DNA insertions: the accumulation of induced AMP transcripts increased proportionally to the transgene copy number in lines that contained one to five transgene insertions, reaching the maximum level in plants with five transgene copies per genome, and then gradually decreased to the unchanged level as the transgene copy number continued to increase. The decrease in the transcripts in plants that contained more than five transgene copies per genome was probably a result of post-transcriptional gene silencing, and related to mRNA instability; according to the model, when the accumulation of similar transcripts crosses a certain threshold, a sequence-specific degradation system is activated to degrade this mRNA (Matzke and Matzke, 1995; Minocha, 2000). Other reports have shown positive (Gendloff et al., 1990), negative (reviewed by Finnegan and McElroy, 1994) or no (Peach and Velten, 1991) correlation between transgene copy number and transgene expression. The predictable expression pattern makes the win3.12T promoter a valuable tool in transgenic research.
In conclusion, our data suggest that both MsrA2 and temporin A combine powerful antimicrobial activities with low phytotoxicity at physiologically relevant concentrations, and the expression of MsrA2 and temporin A peptide genes in a pathogen-inducible manner provides an efficient mechanism to enhance plant resistance against a wide range of pathogenic fungi, oomycetes and bacteria. To our knowledge, this is the first report on the expression of native temporin A in plants. Further efforts are in progress to test this combination of an inducible promoter with AMPs in other plant species, particularly in poplar, from which the win3.12T promoter originated.
Plant material and growth conditions
Tobacco plants (N. tabacum L. cv. Xanthi) were grown aseptically in Magenta GA7 vessels (Magenta, Chicago, IL, USA) on hormone-free Murashige and Skoog medium (Murashige and Skoog, 1962) containing 2% sucrose and 0.7% agar (Difco, Sparks, MD, USA) at pH 5.8. All plant cultures were maintained at 24 °C using a 16-h light photoperiod at a light intensity of 25 µmol/m2/s.
Fungi F. solani, F. oxysporum, A. alternata, B. cinerea and S. sclerotiorum, the oomycete P. aphanidermatum and the bacterium P. carotovorum (also called Erwinia carotovora) were used in this study. All cause major tobacco diseases. The fungal and oomycete cultures were grown on either V8 medium [10% V8® juice, 5 mm 2-(N-morpholino)ethanesulphonic acid (MES), 15 g/L Difco™ agar, pH 6.4] or potato dextrose agar [PDA; 24 g/L Difco™ potato dextrose broth (PDB) powder, 5 mm MES, 15 g/L Difco™ agar, pH 6.4] in low light at room temperature, and transferred to the fresh medium monthly or stored at 4 °C for several months. The bacterial culture was grown for 1–2 days at 26 °C on a shaker to mid-logarithmic phase in liquid antibiotic-free Luria–Bertani (LB) medium (Sambrook et al., 1989). The bacterial cells were collected by centrifugation at 2000 g for 10 min, re-suspended in sterile 10 mm MgCl2, and equal volumes of bacterial suspension were dispensed into Eppendorf tubes and stored at –80 °C as glycerol stocks of known concentration. The number of colony-forming units (cfu) in the frozen bacterial stock culture was determined by plating serial dilutions from a representative tube on LB agar plates (Sambrook et al., 1989).
Antimicrobial activity of purified MsrA2 and temporin A
MsrA2 and temporin A peptides were synthesized by the Fmoc [N-(9-fluorenyl)methoxycarbonyl] solid-phase method with a peptide synthesizer (AnaSpec, San Jose, CA, Canada), with more than 95% purity as determined by analytical high-performance liquid chromatography. Lyophilized peptides were reconstituted in sterile dH2O to stock concentrations of 300 and 60 µm, and stored as 100-µL aliquots at –20 °C until use.
The antifungal activities of MsrA2 and temporin A in vitro were evaluated by quantifying the number of conidia germinated in the presence of the peptide. The conidial suspension was prepared as described previously (Yevtushenko et al., 2005), adjusted to 105 conidia/mL in a 50-fold diluted PDB (Difco), dispensed into 100-µL aliquots in 35-mm tissue culture plates (Sarstedt, St-Leonard, Quebec, Canada) and incubated with different concentrations of the peptide at room temperature in low light. The germination of conidia was assessed after 24 h of incubation, using a haemocytometer and a Nikon TE300 inverted microscope (Nikon, Tokyo, Japan). As in Jacobi et al. (2000), conidia were considered to have germinated only if the newly grown hyphae were at least twice the length of the conidia. Each experiment was replicated three to four times.
To evaluate the antibacterial activities of MsrA2 and temporin A in vitro, 100-µL aliquots of P. carotovorum cell suspension (104 cfu in 1 mL of 50-fold-diluted LB medium) were incubated with different concentrations of the peptide in Eppendorf tubes at 26 °C for 24 h, and then plated on LB agar plates (Sambrook et al., 1989). After overnight incubation at 26 °C, bacterial colonies were counted and compared with the control P. carotovorum plates.
Phytotoxicity of the peptides
Mesophyll protoplasts were isolated from fully expanded leaves of 3–4-week-old non-transgenic tobacco plants. The leaves were cut into strips, 1 mm in width, and incubated in a maceration enzyme solution (6 mL in a 9-cm Petri dish) containing 0.2% (w/v) Cellulase (Sigma, St. Louis, MO, USA), 0.1% (w/v) Driselase (Sigma), 0.4 m sucrose, 0.1 m glycine, 10 mm CaCl2 and 10 mm MES (pH 5.6). After 14–16 h at 25 °C in the dark, the incubation mixture was diluted with three volumes of 0.5 m sucrose and 10 mm CaCl2, shaken gently to release the protoplasts and filtered through an 80-µm nylon sieve into two 10-mL centrifuge tubes. One millilitre of W5 solution (Medgyesy et al., 1980) was carefully added to each tube as the top layer. After centrifugation at 350 g for 8 min, the viable protoplasts were collected from the sucrose–W5 interphase, washed with 7p medium (Kao and Michayluk, 1975) in a new tube, pelleted by centrifugation and resuspended gently in 2–3 mL of fresh 7p medium. The protoplasts were quantified with a haemocytometer and cultivated at a density of 105/mL in 100-µL aliquots of liquid 7p medium in 35-mm tissue culture plates (Sarstedt) at 24 °C in low light (5 µmol/m2/s). The cell culture was challenged with different concentrations (5-µm increments) of MsrA2 or temporin A. After incubation of the cells with the peptides for 24 h, 2 µL of Evan's blue (0.5% in 0.7 m sorbitol) or neutral red dye (0.1% in 0.7 m sorbitol) was added to each aliquot of the protoplast culture, and allowed to diffuse in the medium for 1 h. Evan's blue stained dead cells, whereas neutral red accumulated in the vacuoles of living cells. The percentage of viable protoplasts was estimated using a Nikon TE300 inverted microscope.
Two plant transformation vectors designed to express AMP genes under the control of an 823-bp pathogen- and wound-inducible promoter win3.12T (Yevtushenko et al., 2004) were constructed: pwin3.12T-MsrA2 contained the dermaseptin B1 derivative MsrA2 gene (Osusky et al., 2005), whereas pwin3.12T-TA contained the gene encoding methionine-preceded temporin A. To make these vectors, the XbaI/EcoRI DNA fragments containing either the MsrA2 (388 bp) or the temporin A (331 bp) gene were ligated into the corresponding sites of the plasmid pwin3.12T-GUS in place of the deleted β-glucuronidase (GUS)-containing region. The pwin3.12T-GUS vector is a derivative of pBI121 (Jefferson et al., 1987), in which the CaMV 35S promoter was replaced by the win3.12T promoter; details of its construction are given in Yevtushenko et al. (2004). The resulting vectors had a transcriptional fusion between the win3.12T promoter and the coding region of either the MsrA2 or temporin A gene, followed by the nopaline synthase (nos) 3′-untranslated region (polyadenylation signal). The correct insertion and full nucleotide sequence of the win3.12T promoter and the AMP genes were verified by restriction mapping and by DNA sequence analysis (Applied Biosystems, Foster City, CA, USA).
The T-DNA region of both vectors also contained the neomycin phosphotransferase (NPTII) gene, driven by the nos promoter, for use as a kanamycin resistance marker for the selection of transgenic plants. The constructs were maintained in Escherichia coli DH5α, and introduced into Agrobacterium tumefaciens MP90 by the freeze–thaw method (Holsters et al., 1978). The presence of the vectors in the antibiotic-selected bacterial clones was confirmed by PCR using promoter- and gene-specific primers.
Plant transformation and selection of transgenics
Transformation, selection and regeneration of transgenic tobacco plants were carried out as described in Yevtushenko et al. (2005).
Mechanical wounding, inoculation of plants with pathogens and tissue sampling
All treatments were conducted on well-developed young tobacco plants (3–4 weeks old), grown aseptically on antibiotic-free medium (see ‘Plant material and growth conditions’ section).
Mechanical wounding of tobacco leaves and tissue sampling for RNA extractions were performed as described previously (Yevtushenko et al., 2004).
For fungal infection, two 1-cm2 agar blocks of fungal mycelia were placed 2 cm from the stem of a tobacco plant, and cultivated for several days. One to two days after the fungal mycelia reached the plant stem (disease symptoms not yet visible), leaves from the middle part of the plant were collected for RNA and protein extractions. For oomycete infection, the agar blocks of P. aphanidermatum culture were placed in direct contact with the plant stem. For bacterial infection, P. carotovorum cells were adjusted to the concentration of 105 cfu in 1 mL of sterile dH2O, and 100 µL of the bacterial suspension was spread on to the base of each plant to cover the stem surface. Similar to the fungal and oomycete infection studies, leaves from the middle part of the plant were collected 1–2 days after inoculation before any disease symptoms appeared.
PCR analysis of transgenic plants
Tobacco DNA was isolated from the leaves of transformed and control plants using a GenElute™ Plant Genomic DNA Kit (Sigma). Each PCR mixture contained 200 ng of plant DNA, Taq PCR Master Mix (Qiagen, Valencia, CA, USA) and specific primers (0.4 µm each) in a final volume of 50 µL. All amplifications were performed with manual hot start and the following profile: 94 °C for 3 min, then 30 cycles of 94 °C for 30 s, 57 °C for 30 s, 72 °C for 60 s, followed by a final 10-min extension at 72 °C prior to halting the reaction at 4 °C. Primers for plants with the pwin3.12T-MsrA2 construct amplified the 934-bp full-length promoter–transgene region: the forward primer 5′WIN(D) (5′-AAGCTTCCAACATCAATGA-3′, 19-mer) had the 5′ end nucleotide sequence of the win3.12T promoter, and the reverse primer 3′MsrA2 (5′-TTACTGCGAGATGGTGTCGGCTA-3′, 23-mer) was complementary to the 3′ end of the MsrA2 gene. Primers for plants with the pwin3.12T-TA construct amplified the 877-bp full-length promoter–transgene region: the forward primer was the same as above, i.e. 5′WIN(D), whereas the reverse primer 3′TA (5′-TTACAGGATTCCCGAGAGA-3′, 19-mer) was complementary to the 3′ end of the temporin A gene. PCRs performed on the plasmids pwin3.12T-MsrA2 and pwin3.12T-TA served as positive controls.
Southern and Northern blot analyses
For Southern analysis, 4 µg of tobacco DNA from each line was digested with HindIII, electrophoresed and hybridized with a 32P-labelled win3.12T promoter probe, as described in Yevtushenko et al. (2005). Because the HindIII site is unique in the T-DNA region of both vectors, and the T-DNA integrates into the plant genome randomly, each transgene copy produced a band of unique mobility, indicating the number of transgene insertions in the genomes of respective transformants.
For Northern analysis, total RNA was isolated from tobacco leaves using an RNeasy® Plant Mini Kit (Qiagen). To ensure quantitative and reproducible results, in another set of experiments, the total RNA was isolated using TRIzol® Reagent (Invitrogen, Carlsbad, CA, USA). The RNA samples (30 µg per lane) were denatured, electrophoresed, checked for equal loading and RNA integrity, transferred to a Biodyne B nylon membrane (Pall, Pensacola, FL, USA) and hybridized with the 32P-labelled antisense DNA strand of either the MsrA2 or temporin A gene. The antisense DNA probes were prepared by linear PCR amplification using a Strip-EZ™ PCR Kit (Ambion, Austin, TX, USA) and probe-specific reverse primers. A purified XbaI/EcoRI DNA fragment from pwin3.12T-MsrA2 or pwin3.12T-TA plasmid contained the full-length transgene at the 5′ end, and was used as a template for run-off PCR probe synthesis at a concentration of 50 ng of the transgene in 100 µL of PCR mixture. The reverse primer for MsrA2 probe synthesis, 3′MsrA2, was complementary to the 3′ end of the MsrA2 gene, and its nucleotide sequence has been described in the ‘PCR analysis of transgenic plants’ section. The reverse primer for temporin A probe synthesis, 3′Nost1 (5′-GAACGATCGGGGAAATTCGAGC-3′, 22-mer), was complementary to the untranslated region downstream of the temporin A gene; the italic part of its sequence corresponds to the 5′ end of the nos polyadenylation region. The parameters of PCR were the same as described in the ‘PCR analysis of transgenic plants’ section, except that the number of cycles was increased to 40. The probes were purified from unincorporated nucleotides with NICK™ columns (GE Healthcare, Uppsala, Sweden), and used at a concentration of 2 × 106 cpm in 1 mL of hybridization solution. After hybridization in PerfectHyb™ Plus buffer (Sigma) at 65 °C for 1 day, the membranes were washed twice with constant agitation in 2 × standard saline citrate (SSC), 0.1% sodium dodecylsulphate (SDS) at room temperature for 5 min, twice in 1 × SSC, 0.1% SDS at 65 °C for 10 min, and then exposed to BioMax MR film (Kodak, Rochester, NY, USA) at –80 °C using an intensifying screen. The membranes were also analysed with a PhosphorImager (Molecular Dynamics, Sunnyvale, CA, USA).
Protein extractions and immunodetection of MsrA2 peptide
Non-transgenic and MsrA2-containing plants were inoculated with the pathogenic fungus F. solani as described above. For protein extracts, 200 mg of fresh leaf tissue from non-stressed and pathogen-infected plants was ground using liquid nitrogen, transferred to a 1.5-mL microtube with 0.4 mL of 0.5 m HCl, vortexed and then centrifuged twice at 14 000 g for 15 min at 4 °C to remove insoluble cell debris. To precipitate the proteins, the supernatant was transferred to a new 2-mL microtube, mixed with 1.5 mL of cold acetone and incubated at –20 °C overnight. After centrifugation at 14 000 g for 30 min at 4 °C, the supernatant was discarded and the protein pellet was dried in a SpeedVac Concentrator (Savant, Farmingdale, NY, USA) for 20 min at room temperature, and then dissolved in 50 µL of buffer containing 2.5 m urea, 30 mm formic acid and protease inhibitor cocktail (Sigma), vortexed and incubated for 30 min at 65 °C. The protein concentration was determined by the Bradford assay (Bradford, 1976). Equal protein samples were mixed with equal volumes of 2 × sample buffer (8 m urea, 10% acetic acid, Pyrodin Y dye), and separated by AU-PAGE for 2 h at 100 V, using 5% acetic acid as running buffer. The AU-PAGE gel was prepared by mixing 7 mg of thiourea, 4 mL of 30% acrylamide/0.2% bis-acrylamide, 1 mL of 43% acetic acid, 2 mL of 10 m urea, 1 mL of sterile dH2O and 45 µL of 30% H2O2. Because MsrA2 is a positively charged peptide that migrates towards the negative cathode, the polarity of the electrodes during electrophoresis was reversed. The separated proteins were electrotransferred to a Hybond-P polyvinylidene difluoride (PVDF) membrane (GE Healthcare, Little Chalfont, Buckinghamshire, UK) in a Mini Trans-Blot apparatus (Bio-Rad, Hercules, CA, USA) for 1 h at 100 V, using 0.7% acetic acid as transfer buffer. After blocking nonspecific binding sites with 5% (w/v) skim milk (Difco) in PBS-T buffer (80 mm Na2HPO4, 20 mm NaH2PO4, 100 mm NaCl, 0.1% Tween 20), the membrane was incubated with MsrA2-specific rabbit polyclonal antibodies (1 : 2000 dilution in PBS-T buffer) for 2 h at room temperature, washed according to the membrane manufacturer's instructions and then incubated with goat anti-rabbit IgG-horseradish peroxidase (1 : 100 000 dilution in PBS-T buffer) for 1 h. Detection of immobilized MsrA2–antibody complexes was performed with ECL Plus Western Blotting Reagents (GE Healthcare), and the images were recorded on BioMax MR film (Kodak).
Plant resistance bioassays
For the bioassay of detached leaves, fully expanded tobacco leaves, about 5–7 cm in length, were excised from aseptically grown 4–5-week-old plants and placed on sugar-free MS agar medium (Murashige and Skoog, 1962) with the adaxial side up. To test resistance to fungal or oomycete infections, a 1-cm2 agar block of freshly grown culture was placed in the centre of the detached leaf with the living cells side down, in direct contact with the plant tissue. To test resistance to bacterial infection, P. carotovorum cells were adjusted to a concentration of 105 cfu in 1 mL of sterile dH2O, and 100 µL of the bacterial suspension was spread on the adaxial side of the leaf. The leaves were cultivated under standard plant growth conditions, and observed daily for disease development. The impact of disease on detached leaves was evaluated by measuring the area of the decayed tissue, 1, 2 and 3 weeks after inoculation, and scored as a percentage of the average necrotic area formed in transgenic leaves relative to the average in untransformed controls: susceptible, decayed tissue in transgenic leaves was more than 71% of that in the control, with severe symptoms of leaf dehydration and chlorosis; moderately resistant, decayed tissue comprised 31%–70% of that in untransformed leaves; resistant, decayed tissue comprised 1%–30% of that in untransformed leaves; highly resistant, decayed tissue was less than 1% of that in control leaves, with no visible disease development. The experiments were performed with at least eight leaves of each transgenic line, and the results were evaluated by analysis of variance to detect significant differences in disease severity.
For the bioassay of whole plants, 4–5-week-old transgenic (10 lines of each construct) and control tobacco plants were challenged with the pathogens as described in the ‘Mechanical wounding, inoculation of plants with pathogens and tissue sampling’ section, and cultivated for several weeks. The plants were observed daily for disease symptoms, and scored for resistance as described previously (Yevtushenko et al., 2005), using the following criteria: susceptible, plants developed severe wilt symptoms, such as brown stems and dehydrated leaves, within the same time period as non-transgenic plants; moderately resistant, two- to threefold longer period to develop wilt symptoms; resistant, four- to fivefold longer period to develop wilt symptoms; highly resistant, more than sixfold longer period before any disease symptoms appeared. The bioassays were replicated three to five times with each transgenic line.
We thank Dr Zamir Punja (Simon Fraser University, BC, Canada) for providing phytopathogenic fungi. This research was supported by grants from the National Center of Excellence and the Advanced Foods and Materials Network to S.M.