Seeds of Pueraria montana var. lobata (Kudzu) were collected at Nagoya city, Japan and stored at room temperature for 3 years. Seeds were washed with a neutral detergent and tap water, and sterilized with 2% NaClO solution for 5 min. The hard seed coat was cut with a knife for imbibition. After 10 min further sterilization, they were washed three times with autoclaved water in a clean bench, and were aseptically cultured on 0.8% agar medium in a 50 ml tube in the dark at 24˚C for 6 to 14 days. Lactuca sativa (lettuce) seedlings were prepared as described previously [16]. Briefly, lettuce seeds (Great Lakes) in a small bag of Miracloth were washed with a neutral detergent and tap water, and sterilized with 1.5% NaClO solution for 15 min and washed three times with autoclaved water. They were aseptically cultured on 0.8% agar medium in a 15 ml tube in the light condition (60 µE s⁻¹) at 24˚C for 5 - 7 days.

Optimal combination of cell wall degrading enzymes and osmotic condition for Kudzu protoplast isolation was determined from preliminary experiments using 14 day-old seedlings, using 24 combinations of six kinds of enzymes, i.e., Cellulase R10, Cellulase RS, Hemicellulase, Driselase 20, Macerozyme R10, Pectolyase Y-23 [17], in 0.4 M to 0.8 M mannitol solution. Cotyledons of 6 day-old Kudzu seedlings aseptically grown, were cut and treated 20 - 24 hrs in the enzyme combination, 1% each of Cellulase R10 and Driselase 20, in 0.6 M mannitol solution in a 24-well plastic culture plate at static condition at 27˚C in the dark. Protoplasts of lettuce cotyledons were isolated in 1% each of Cellulase RS and Macerozyme R10 [16] in 0.6 M mannitol solution in a 100 ml flask. After filtration with a mesh size of 80 µm (Kudzu) or 63 µm (lettuce), the protoplasts were washed three times with 0.6 M mannitol solution by centrifugation at 900 rpm (100 g) for 5 min.

Effects of different concentrations of 2,4-dichlorophenoxyacetic acid (2,4-D), 0, 0.1, 1, 10, 100 µM, and benzyladenine (BA), 0, 0.1, 1, 10 µM, were investigated using 50 µl each liquid Murashige and Skoog’s (MS) [32] basal medium containing 3% sucrose and 0.6 M mannitol in each well of a 96-well culture plate (Falcon No. 3072). Five µl of Kudzu protoplasts in 0.6 M mannitol solution (0.6 - 70 × 10⁴/ml) were put into the 50 µl medium. Initial protoplasts densities of culture were described to be 0.6 - 70 × 10³/ml. Numbers of non-spherically enlarged protoplasts and divided protoplasts of Kudzu were counted under an inverted microscope periodically (2 to 12 days of culture). Protoplast growth was described as the percentage of initial plated numbers of protoplasts in each well.

Fifty µl of liquid MS basal medium containing 1 µM 2,4-D and 0.1 µM BA, 3% sucrose and 0.6 M mannitol was put in a well of 96-well culture plate. Five µl each of protoplast suspensions of Kudzu and lettuce at 10 times of final protoplast densities in 0.6 M mannitol solution were added to each well. Final protoplast densities were 0.6 - 30 × 10³/ml for Kudzu and 6 - 100 × 10³/ml for lettuce, respectively. Numbers of non-spherically enlarged and divided protoplasts were counted under an inverted microscope. After 6 days of co-culture, lettuce protoplast growth was calculated by deduction of the numbers of dark-yellow non-spherically enlarged Kudzu protoplasts from the total numbers of enlarged protoplasts of lettuce and Kudzu. After 10 days of co-culture, the numbers of divided lettuce protoplasts were counted. Percentages of control values without Kudzu protoplasts were calculated and then, the percentages of control at different densities of lettuce protoplasts (6, 12, 25, 50, 100 × 10³/ml) were averaged with standard error (SE). Yellow pigment accumulation of lettuce protoplasts after 28 days of co-culture was analyzed using digital image analysis described in 2.6.

Isoflavones, puerarin and daidzein, were dissolved at 10 mg/ml and diluted in filter-sterilized DMSO (Milipore PTFE membrane). And one µL of each solution was put into 50 µl of the same medium described in 2.4. except for 0.4 M mannitol. Final concentration of DMSO was 2% as previously described [19]. Final protoplast densities of lettuce were the same as in 2.4 (6 - 100 × 10³/ml). Numbers of non-spherically enlarged and divided protoplasts of lettuce were counted after 4 and 8 days of culture. Percentages of control values without isoflavone were calculated and then, the percentages of control at different densities of lettuce protoplasts (6, 12, 25, 50, 100 × 10³/ml) were averaged with SE. Yellow pigment accumulation of lettuce protoplasts after 27 days of co-culture was analyzed as in 2.6.

Image analysis of yellow pigment accumulation of lettuce protoplasts was performed after 3 weeks to 1.5 months of culture as described previously [17] [22] [23] [24] [25] [27] [29] [31]. Digital image of a 96-well culture plate was scanned using a scanner (Epson GTX-970). Image analysis by software Image J [33] was performed. An image was selected from the blue channel. A horizontal straight line was drawn at the center of the wells. The plot profile of the line was analyzed. Using excel software, we determined the average of blue plot values for each well. The yellow value was converted by deduction of each averaged blue value from the highest blue value (control). The yellow value at each concentration of isoflavones was deduced as a control. The yellow values of Kudzu protoplasts at each density were not deduced or deduced as noted in the text. The % yellow value of control without protoplasts of Kudzu or isoflavones was calculated at each lettuce protoplast density. Finally, the percentages of control were averaged with SE at different cell densities of lettuce (6, 12, 25, 50, 100 × 10³/ml).

Isolated protoplasts of Kudzu cotyledons were non-green and around 40 µm diameter (similar to Figure 1(a)). No difference was observed during 1 day of culture. In the optimal hormonal conditions, non-spherical enlargement (Figure 1(f)) after 2 days of culture and cell division (Figure 1(g)) after 3 days of culture were observed.

Figure 2(a) shows the effects of 2,4-D and BA on non-spherical enlargement of Kudzu protoplasts after 2 days of culture at 17 × 10³/ml. All hormonal conditions tested gave cell enlargement, while 0.1 - 10 µM of 2,4-D were better than without 2,4-D. Enlargement occurred without BA. Combination of 10 µM 2,4-D and 0.1 µM BA was the best condition. Figure 2(b) shows the effects of 2,4-D and BA on the cytoplasm-rich cell division of Kudzu protoplasts after 5 days of culture at 17 × 10³/ml. Although enlarged protoplasts (including vacuolated cells) can be observed in all hormonal conditions, cytoplasm-rich division can be observed in both 2,4-D and BA containing media. High division efficiency was observed at 2,4-D 10 µM and BA 0.1 - 10 µM (8.9%, 5.9%, 7.7%). After 12 days of culture, large brown colored colony development was observed at the similar hormonal conditions as of Figure 2(b). At higher initial protoplast densities (35, 70 × 10³/ml) (Figure 2(c) and Figure 2(d)), division occurred after 5 days of culture at 1 - 10 µM 2,4-D and BA 0.1 - 10 µM. Although 10 µM each of 2,4-D and BA was the best condition, the percentages of division were lower (5.0, 2.9%) than those at 17 × 10³/ml (Figure 2(b)).

Figure 3 shows the effects of different protoplast densities of Kudzu at 1 µM 2,4-D and 0.1 µM BA condition. Good growth of Kudzu protoplasts was obtained at 4.2 × 10³/ml up to 17 × 10³/ml of protoplast densities. After 2 days of culture, efficiency of highest density (70 × 10³/ml) was inhibitory for non-spherical enlargement (E). After 5 days of culture, non-spherical enlargement (E) occurred at this highest density, however, much less division (D) was observed. In another independent experiments for lower protoplast densities (0.6, 1.2, 2.5, 5, 10 × 10³/ml), after 6 days of culture, growth (E + D) of 10%, 7%, 18%, 13%, 16% were obtained, respectively. Average % was 13 ± 2 (SE).

The lettuce cotyledon protoplasts were green and ca.30 µm diameter (Figure 1(d)) and a little smaller than of Kudzu protoplasts (Figure 1(a)) at the 1^st day of co-culture. Lettuce protoplasts did not enlarge before 3 days of culture, when the Kudzu protoplast started to grow without lettuce. Non-spherically enlarged or divided Kudzu protoplasts without lettuce had no yellow color (Figure 1(f), Figure 1(g).

Figure 4 shows the inhibition of lettuce protoplast growth by co-cultured Kudzu protoplasts at very low densities (0.6 - 10 × 10³/ml) in the hormonal condition of 1 µM 2,4-D and 0.1 µM BA.

After 6 days of co-culture, dark yellow Kudzu protoplasts can be distinguished under an inverted microscope (Figure 1(b), Figure 1(c)). From the total numbers of non-spherically enlarged protoplasts, such Kudzu-specific numbers were deducted and described as lettuce-specific non-spherically enlarged lettuce protoplasts. After 10 days of co-culture, only the numbers of divided cells including developed colonies of lettuce protoplasts were counted. Cell division of lettuce protoplasts and of Kudzu was distinguished easily under an inverted microscope. Almost total inhibition of lettuce division was observed at 10 × 10³/ml of Kudzu protoplasts, which was similar density to the lowest density of lettuce tested in co-culture (6 × 10³/ml). Kudzu protoplasts clearly inhibited growth of lettuce protoplasts at two stages, i.e., cell wall formation stage after 6 days and cell division stage after 10 days of co-culture.

At the lowest lettuce protoplast density tested (6 × 10³/ml), Kudzu specific growth was prominent at high protoplast densities of Kudzu (10, 30 × 10³/ml). Averaged counted numbers of lettuce protoplast growth at control without Kudzu were 85 ± 21 (SE) at 6 days of culture (91% was non-spherically enlargement) and 48 ± 8 (SE) at 10th day of culture, respectively.

As shown in the scanned image of 96-well plate co-cultured for 28 days (Figure 5(a)), at the highest Kudzu density, 30 × 10³/ml, brown colored Kudzu growth was prominent. Though, lettuce protoplasts inhibited the brown colored Kudzu growth at density dependent manner. At high Kudzu density (30 × 10³/ml), the calculated yellow values increased at lowest lettuce density (6 × 10³/ml). But decrease of yellow color at the high Kudzu density (10 × 10³/ml) and low lettuce density (6 × 10³/ml) was clearly shown in Figure 5(a). On line b1 of yellow pigment accumulation (Figure 5(b)), when the yellow values of Kudzu protoplasts at each density were not deducted, less inhibition of yellow pigment accumulation by Kudzu was observed compared with inhibition on cell wall formation and cell division stages (Figure 5). In line b2 of yellow pigment accumulation (Figure 5(b)), where each zero control value without lettuce protoplasts was deducted from each yellow value of co-cultured well, inhibition by Kudzu was similar to that at 6 days of co-culture (Figure 4).

Two putative allelochemicals, puerarin and daidzein of Kudzu were tested using protoplast co-culture method with digital image analysis. As shown in Figure 6, big differences of effects on three stages of lettuce protoplast growth were found between puerarin and daidzein. Figure 7 shows the scanned image of 96-well plate in Figure 6(c). In the daidzein co-culture with lettuce protoplasts, many crystals were found in spherical and non-spherical protoplasts at 100 µM (Figure 8), which almost totally inhibited cell division after 8 days of co-culture (Figure 6(b)). At higher concentrations of daidzein, many crystals were prominent in the culture medium. No such crystals of puerarin were observed. In contrast, stimulation of cell division was observed at high concentrations of puerarin (Figure 6(b)). At up to 10 µM, the 20% inhibition rate of daidzein was not different among the three stages of lettuce growth (Figures 6(a)-(c)). At 100 µM, inhibition was the strongest at the cell division stage (Figure 6(b)). Less inhibition was observed on early cell wall formation stage (Figure 7(a)), or yellow pigment accumulation stage (Figure 6(c)).

Averaged counted numbers of lettuce protoplast growth at zero control without isoflavone were 50 ± 12 (SE) at 4 days of culture (Figure 6(a)) and 24 ± 2.7 (SE) at 8 days of culture (Figure 6(b)), respectively.

In the protoplast cultures of Kudzu, optimal hormonal condition for cell division and colony development was a combination of 10 µM 2,4-D and 0.1 - 10 µM BA in MS basal medium containing 0.6 M mannitol. 2,4-D only medium was good for early cell wall formation, but both 2,4-D and BA were needed for further growth. Such a result was repeatedly obtained in experiments using different protoplast densities (17 - 70 × 10³/ml) (Figure 2).

As for the lettuce cotyledon protoplast growth, lower concentrations of 2,4-D (0.1 - 1 µM) and BA (0.1 - 1 µM) were optimal [16]. The combination of 1 µM 2,4-D and 0.1 µM BA in MS basal medium, was used for co-culture with recipient lettuce cotyledons protoplasts in bioassay of allelopathy of many test plants [16] [18] [19] [24] and of Kudzu in this paper. It was a suboptimal condition for Kudzu protoplast growth. MS basal medium was not optimal for protoplast growth of lettuce instead ammonium nitrate-free MS basal medium was optimal [16]. However, the hormonal condition and basal medium for co-culture bioassay of allelopathy were not necessary the optimal condition for each test plant protoplasts as studied with Mucunapruriens [16], Derris indica [19],Leucaena leucocephala and M. gigantea [18], and bamboo species [24]. At sub-optimal hormonal condition for cell division of test plants protoplasts, a similar strong allelopathic activity was observed.

Kudzu protoplasts could grow in a wide density range (0.6 - 70 × 10³/ml) at sub-optimal hormonal conditions, though quantitative evaluation was more difficult at high densities under an inverted microscope, which was much low density compared with the optimal high density of Mucunapruriens. Inhibition of lettuce protoplast growth was observed at lower densities than optimal high densities (300 × 10³/ml or more) for M. pruriens [16].

About 50% inhibition of lettuce protoplast division by 2.5 × 10³/ml Kudzu and almost total inhibition by 10 × 10³/ml (10⁴/ml) Kudzu indicates very strong inhibition (Figure 4). Compared with the effective protoplast density of other plant species reported, the grade of inhibition of Kudzu was as high as the protoplasts of an invader plant, L. leucocephala (1.5 × 10³/ml, 20 × 10³/ml) and M. gigantea (1.5 × 10³/ml, 20 × 10³/ml) [18]. Inhibition by epicotyl protoplasts of V. villosa (1 × 10³/ml, 3 × 10³/ml) [17] , which was three times stronger inhibition than of Kudzu, was the strongest among those tested using protoplast co-culture method.

Compared with Arabidopsis leaves (20 × 10³/ml, 100 × 10³/ml) [23], one tenth lower density was effective in Kudzu, while a similar pattern of inhibition among three growth stages (less inhibition on yellow pigment accumulation compared to that on cell wall formation or cell division) was observed. An anthocyanin accumulating Sonneratia ovata (10 × 10³/ml, 100 × 10³/ml<) showed much lower activity, and different patterns (no inhibition on the yellow pigment accumulation stage) were observed among three growth stages [25].

Mutual inhibition was observed by lettuce protoplasts on Kudzu protoplast growth. In early cell wall formation stage in co-culture, the number of the dark-yellow non-spherically enlarged protoplasts of Kudzu was deducted from the total number of non-spherically enlarged protoplasts, which included both lettuce and Kudzu reacted. Division of lettuce protoplasts was detected under an inverted microscope. However, yellow pigment accumulation of lettuce protoplasts at low densities (6, 12 × 10³/ml) is difficult to separate from the brown colored colony development of Kudzu at high densities (10, 30 × 10³/ml). Similar mutual inhibition by bamboo protoplasts and by lettuce protoplasts has been reported [24]. Strong inhibition of yellow pigment accumulation by Sasakurilensis was calculated as in line b2 in Figure 5(b), which shows strong inhibition at the cell wall formation stage similar to Kudzu. Such a calculation method (Subtraction of the control yellow values without lettuce protoplasts, at each density of test plant protoplasts or at each concentration of allelochemicals) had been employed in the previously reported papers using the DIA-PP method [17] [22] - [27] [29] [31].

Yellow pigment accumulation at late growth stage (after 3 weeks of culture) of lettuce protoplast might be related to the synthesis and degradation of secondary metabolites. A carotenoid was extracted to hexane fraction from lettuce protoplasts that had accumulated the yellow pigment [22]. Recently, a very different pattern of inhibition at three growth stages of lettuce by a carotenoid-accumulating callus of Avicenniaalba has been reported [27]. Very recently, unique allelopathic activities of three carotenoids were investigated at three growth stages of lettuce protoplast using the DIA-PP method [34].

Though, lettuce protoplast has a broad range of optimal osmotic condition, e.g., 0.4 M to 0.8 M mannitol, better and rapid growth is observed in low 0.4 M mannitol condition. We selected 0.6 M in co-culture with Kudzu as it was the optimal osmotic condition for Kudzu protoplasts. However, we used 0.4 M for testing the allelochemicals, as dissolving solvent DMSO inhibits lettuce protoplast growth at a concentration higher than 2% [19]. We found no differences in the inhibition patterns among different osmotic conditions. Though, growth stages of lettuce proceed earlier under lower osmotic conditions.

Complete inhibition of lettuce protoplast division by daidzein at 100 µM (Figure 6(b)) is in the group of chemicals with strong allelopathic activity studied using protoplast co-culture method. Similar strong inhibition has been observed by,e.g., canavanine in Viciavillosa [18] ; L-DOPA in Mucunapruriens [16]; isoflavonoid, rotenone in Derris indica [19]; cinnamic acid [26] and tulipalin A [31] in Spiraeathunbergii. In contrast, coumarin in Prunusyedoensis was 25% of control at 100 µM [20]. An anthocyanin, cyanin in Sonneratiaovata inhibited growth to 40% of the control [25], respectively. Although, purine alkaloid, caffeine was the strongest in the group of related metabolites [28] [29], no inhibition was observed at 100 µM. Stimulation effect of trigonelline was reported [35].

This is the first study showing the strong inhibitory allelopathic activity of isoflavone, daidzein, at the cellular level, using the protoplast co-culture method with digital image analysis. Among the three growth stages of lettuce, inhibition at the cell division stage was the strongest. Such inhibitory patterns are similar to those of the Kudzu protoplast in co-culture. By contrast, puerarin, which is a glucoside of daidzein, was not inhibitory, and showed a rather stimulatory effect at high concentrations at all three growth stages of lettuce protoplasts (Figure 6).

Differences in the stimulation and inhibition patterns among three growth stages were observed among plant species and putative allelochemicals studied [17] [23] [24] [25] [26] [27] [30] [31], which must reflect the cellular action site of each chemical.

In the present study, strong allelopathic activity of cotyledon protoplasts of Kudzu was found in co-culture with lettuce protoplasts (Figure 4). Such strong inhibitory activity is consistent with the data of the plant box method, i.e., 20% to 28% growth of control [13] [15], which measured the effect of the intact root of Kudzu on the growth of lettuce seedlings. These were stronger compared with leaves of kudzu, 48% to 65% growth of control, using the sandwich method [10] [13].

High content of isoflavone, puerarin, in the roots (e.g., 13 mg/g dry weight, 5 mg/g dry weight) and immature cotyledon-derived callus or suspension cultured cells (1 - 2 mg/g dry weight), but not in the leaves of Kudzu had been reported [2] [3] [4]. These phenomena suggested the possibility of puerarin as an allelochemical in Kudzu protoplasts.

However, in this report, puerarin itself was not likely the allelochemical of Kudzu protoplasts since no inhibitory activity of puerarin was observed up to 780 µM in the protoplast co-culture method (Figure 6). By contrast, crystal of daidzein was incorporated into the lettuce protoplasts (Figure 8), and showed strong inhibitory activity on lettuce protoplast growth at 100 µM (Figure 6). In Kudzu plant tissues, content of daidzein is not high (e.g., 0.07 mg/g dry weight) compared with that of puerarin, though the extraction method affects their contents [2] [3] [4] [5] [6]. In Kudzu root prepared after water extraction, the glycoside, puerarin content decreased and aglycon, daidzein content increased [6]. When the water content of Kudzu tissue is predicted to be 90%, concentration of daidzein is calculated to be ca. 30 µM, which showed inhibition of division of lettuce protoplasts (Figure 6(b)). Even 100 times lower content of daidzein than puerarin might still be effective as an allelochemical in Kudzu plant. These findings suggest the need of further study on the allelopathic activities of other isoflavones of different aglycon types, e.g., daidzein, genistein, glycitein and their glucosides, e.g., daidzin, genistin, glycitin, and glycosides, whose low content had been found in different organs and at different growth stages of Kudzu plants [2] [3] [4] [5] [6] and in the tissue culture of Kudzu [7] [8]. Several isoflavones are under investigation using the DIA-PP method. Crystal found in co-cultured protoplasts is also an interesting theme.

On the other hand, in another isoflavone-containing plant,Trifolium pratense (red clover) [36] [37], activities of isoflavones themselves were not so high when red clover was used as a recipient, and isoflavonoid-origin phenolic acids in the soil environment were discussed as germination retardant [36]. Soluble phenolics in leaves, roots and soil of Kudzu were reported as allelochemicals using lettuce and radish as recipient plants [38]. In soybean, different amounts of isoflavones and glycosides were reported in sprouts [39], leaves and roots [40], of low allelopathic activities [15]. Recently, a high content of daidzein, which was not found in leaves, was found in the rhizosphere soil, which affects soybean growth through rhizosphere bacterial community [40] [41]. Xanthoxin, which is known as an intermediate of a growth retardant hormone, abscisic acid (ABA) synthesis, was reported as an allelochemical of leaves of Kudzu using cress as a recipient plant [42]. Using the protoplast co-culture method, differences in activities among different recipient plants are known, e.g., inhibition activity of L-DOPA, which is the allelochemical in Mucuna pruriens, was weak on rice protoplasts [16]; Exogenously supplied ABA was strongly inhibitory on Prunus yedoensis protoplasts, in which coumarin was an allelochemical [20], but ABA was stimulatory on lettuce protoplasts [16]. As we found that Kudzu protoplasts can grow at low densities, effects of chemicals, such as coumarin and hinokitiol, were investigated using only 500 Kudzu protoplasts in a well, at optimal hormonal condition of Kudzu (2,4-D 10 µM and BA 1 µM) [43].

Therefore, a single allelochemical in an allelopathic plant might not be the only cause of the allelopathy of the whole plant. The volatile compounds [13] [31], and metabolites in rhizosphere [40] [41] might also be considered. The protoplast co-culture method with digital image analysis using lettuce as a recipient for bioassay of allelopathy will contribute to identification of the underlying mechanism(s) of allelopathy at a cellular level.