Signal Transduction in the Carnivorous Plant Sa rra cenia purpurea’ · Plant Physiol. (1997) 115: 1461-1471 Signal Transduction in the Carnivorous Plant Sa rra cenia purpurea’

Plant Physiol. (1997) 115: 1461-1471

Signal Transduction in the Carnivorous Plant Sa rra cenia purpurea’

Regulation of Secretory Hydrolase Expression dwring Development and in Response to Resources

Daniel R. Gallie* and Su-Chih Chang Department of Biochemistry, University of California, Riverside, California 92521 -01 2 9

~~ - ~

Carnivory in plants has developed as an evolutionary adaptation to nutrient-poor environments. A significant investment of the resources of a carnivorous plant i s committed to producing the traps, attractants, and digestive enzymes needed for the carnivory. The coskbenefit ratio of carnivory can be improved by either maximiz- ing the prey capture rate or by reducing the metabolic commitment toward carnivory. Using the pitcher plant Sarracenia purpurea, we have investigated whether the expression of the hydrolytic enzymes needed for digestion is regulated in response to the presence of prey. Expression of protease, RNase, nuclease, and phosphatase activities could be induced in the fluid of nonactive traps by the addition of nucleic acids, protein, or reduced nitrogen, suggesting that hydrolase expression i s induced upon perception of the appropriate chemical signal. Hydrolase expression was also developmentally controlled since expression commenced upon opening of a trap, increased for severa1 days, and in the absence of prey largely ceased within 2 weeks. Nevertheless, the traps remained competent to induce expression i n response to the appropriate signals. These data suggest that in young traps hydrolase expression i s developmentally regulated, which is later replaced by a signal transduction mechanism, and they demonstrate the ability of a carnivorous species to respond to the availability of resources.

Although plants are autotrophic with respect to chemical energy and reduced carbon, they must scavenge nitrogen and minerals from their environment, usually from the soil through uptake by their roots. Some plants have evolved alternative means to improve the acquisition of nitrogen, e.g. through symbiosis with members of the genera Rhizo- bium and Bradyrkizobium, resulting in the reduction of molecular nitrogen (for review, see Nap and Bisseling, 1990; Verma, 1992). Carnivory represents an alternative and efficient means to acquire nutrients and has evolved independently throughout the plant kingdom (Albert et al., 1992). Since carnivorous species are photosynthetically competent and are usually found in sunny areas, it is thought that the goal of plant carnivory is not carbon but

This work was supported by grant no. 95-37100-1618 from the U.S. Department of Agriculture National Research Initiative Com- petitive Grants Program.

* Corresponding author; e-mail [email protected]; fax 1- 909 -787-3590.

1461

rather reduced nitrogen and minerals that, in the marginal habitats of most carnivorous plants, either are present in low amounts (Plummer, 1963) or are subject to acquisition by bacteria or other plant species (Christensen, 1976; Chapin and Pastor, 1995). Because carnivorous plants can survive in the absence of prey, the purpose of carnivory may be to improve the competitive advantage of the plant through increased biomass or through flowering and seed set (Darwin, 1878; Dore Swamy and Mohan Ram, 1969; Gibson, 1983). In the evolution of carnivory these species have developed mechanisms to attract, capture, digest, and absorb their prey. However, the machinery required for carnivory requires a diversion of plant resources that must yield a greater return than its original investment to justify the expenditure. One means by which these species could improve the cost:benefit ratio of carnivory would be to regulate hydrolase expression so that they are induced only in response to the presence of prey. Such regulation would require that the plant be able to perceive signals associated with the prey and be able to transduce this information to induce hydrolase expression.

Sarracenia purpurea, a pitcher plant common to North America, has modified leaves that form open, fluid- containing pitchers. Secretions produced at the lip of the pitcher attract insects, some of which fall into the pitcher, drown, and are subsequently digested. During the early development of the plant, the pitcher produces attenuated traps with large, flat phyllodia. These immature leaves do not function well as traps and, consequently, they are used primarily as photosynthetic organs. This developmental stage is followed by the production of large, fully developed traps with attenuated phyllodia as the plant becomes competent for carnivory. Even though a trap can live in excess of 1 year, the bulk of the prey is caught within the first 50 d after a trap opens (Fish and Hall, 1978; Cresswell, 1991). Although the presence of open traps containing digestive fluid should have facilitated the biochemical characterization of the secreted hydrolases and the plant’s ability to respond to nutrient stimulus, little is known about hydrolase expression and its regulation in S. purpurea, either developmentally or in response to nutrients in the trap.

Abbreviation: CFU, colony-forming unit(s).

www.plantphysiol.orgon May 3, 2020 - Published by Downloaded from Copyright © 1997 American Society of Plant Biologists. All rights reserved.

http://www.plantphysiol.org

1462 Gallie and Chang Plant Physiol. Vol. 11 5, 1997

In this report we followed the expression of severa1 hydrolytic enzymes, including proteases, RNases, nucleases, and phosphatases, during trap development and in response to nutrient stimulus in the traps of S. purpuyeu spp. purpureu. Hydrolytic activities appeared in the trap fluid commensurate with the opening of a trap and, in the absence of nutrient stimulus, expression decreased substantially within 2 weeks. Nevertheless, those traps that had ceased to express hydrolytic enzyme activities could still induce expression of hydrolases after the addition of nucleic acids, protein, or reduced nitrogen to the traps. These findings support the idea that this carnivorous species has evolved a signal transduction mechanism to opti- mize the benefit of carnivory by reducing its associated costs.

MATERIALS AND METHODS

Plant Growth Conditions

Surruceniu purpurea spp. purpureu (Carolina Biological Supply, Burlington, NC) was grown in terraria under a daily regimen of 12 h of light at 24°C. For the developmental studies, 15 mL of sterile water containing 100 @mL ampicillin, 100 pg/mL carbenicillin, and 250 pg/mL cefotaxime was added after the opening of a trap, and 1-mL samples that were removed at time points were replaced with 1 mL of additional water. For the stimulus-response studies, only older traps that had ceased the developmentally regulated expression of hydrolases, as confirmed by the appropriate enzyme assays, were used. The traps were emptied and washed with sterile water, and 10 mL of sterile water containing 100 pg/mL ampicillin, 100 pg/mL carbenicillin, and 250 pg/mL cefotaxime was added. After 1 week, 600 pg of RNA purified from wheat germ, 1 mg of herring sperm DNA (Sigma), 1 mg of BSA (Boehringer Mannheim), 50 mM NH,CI, 50 mM sodium phosphate, 50 mM KC1, or 50 mM NaCl was added to traps, and 1-mL samples were removed at 2-d intervals.

RNase and DNase Activity Cels

RNase activity gels were performed using SDS-PAGE gels containing 100 pg/mL wheat (Triticum aestivum L.) rRNA as described (Blank and McKeon, 1991) with minor modifications. After electrophoresis, the SDS was removed by incubating the gels in 25% isopropanol containing either 10 mM sodium acetate, pH 5.5 (for acidic RNase activity), or 10 mM imidazole, pH 7.5 (for neutral RNase activity). The gels were then washed for 10 min in either 10 mM sodium acetate, pH 5.5, or 10 mM imidazole, pH 7.5, depending on the assay, to remove the isopropanol and allow renatur- ation of enzyme activity (Blank et al., 1982). In situ degradation of the RNA was carried out by incubating the gels at 50°C for 20 to 30 min either in 100 mM imidazole, pH 7.5, for the detection of neutral RNase activities or 100 mM sodium acetate, pH 5.5, for the detection of acidic RNase activities. The gels were then washed for 10 min in either the 10 mM imidazole buffer (for the neutral RNase activities) or the 10 mM sodium acetate buffer (for the acidic

RNase activities) to allow any residual digested RNA fragments to diffuse from the gel. The gel was stained for 10 min in toluidine blue, followed by a wash in the appropriate buffer to remove excess dye, and the bands representing RNases were observed as cleared regions in the gel.

DNase activity determination was performed using SDS- PAGE gels containing 10 pg/mL single-stranded herring sperm DNA as described (Compton and Cidlowski, 1987) with minor modifications. After electrophoresis, the SDS was removed by washing the gels twice in 25% isopropanol containing 10 mM Tris, pH 7.5, for 10 min. The gels were then washed twice for 10 min each in 10 mM Tris, pH 7.5, 2 mM MgC1, to remove the isopropanol. The gels were incubated at 50°C for 60 to 90 min in 40 mM Tris, pH 7.5,2 mM MgC1, to permit in situ degradation of the DNA. The gels were then washed for 30 min in the same buffer to remove the digested DNA fragments from the gel. The gels were stained with ethidium bromide for 10 min followed by a wash in water to remove the excess stain. The bands representing nucleases were visualized by UV illumina- tion. The gels were incubated overnight in 0.1% SDS at room temperature and stained once again with ethidium bromide to confirm that the observed bands did not result from masking by an abundant protein band.

Protease and Chitinase Assays

Samples were assayed for protease activity in a reaction containing 30 mM sodium acetate, pH 6.0, 10 mM MgCl,, 0.5 mM DTT, and 10 mg of hide powder azure (Sigma) at 37°C. The extent of the digestion of the hide powder was determined by measuring the amount of released azure at A,,,. Each assay was performed in duplicate and the average reported.

Chitinase activity was assayed in a reaction containing 30 mM sodium acetate, pH 6.0, 10 mM MgCl,, 0.5 mM DTT, and 10 mg of chitin azure or 4-methylumbelliferyl p-D- N,N’,N”-triacetylchitotriose (Sigma) at 37°C. The extent of the digestion of the chitin azure was determined spectrofluorometrically by measuring the amount of released azure at or by measuring the amount of released 4-methylumbelliferone fluorometrically at an excitation wavelength of 365 nm and an emission wavelength of 455 nm in a TKO 100 DNA fluorometer (Hoefer Scientific, Inc., San Francisco, CA) when 4-methylumbelliferyl p-D- N,N’,N”-triacetylchitotriose was used as the substrate.

Phosphatase Assay

Phosphatase activity was assayed in 20 mM Tris, pH 8.0 (or 30 mM sodium acetate, pH 4.5, for the acidic activity), 10 mM MgCl,, and 16 mM p-nitrophenyl phosphate (Sigma) at 37°C. The reaction was stopped by the addition of 0.2 M

NaCO,, and the extent of the phosphatase activity was determined spectrofluorometrically by measuring the amount of released p-nitrophenyl at A,,,. Each assay was performed in duplicate and the average reported.



Regulation of Secretory Hydrolase Expression in Sarracenia purpurea 1463

RESULTS

Hydrolytic Enzyme Expression IsDevelopmentally Regulated

To determine the expression patterns of nucleases andother secreted hydrolytic enzymes produced during thedevelopment of an opened trap, 15 mL of sterile water wasadded immediately after the opening of a trap, and 1-mLsamples were taken at regular intervals during the first 6weeks. The samples were assayed for RNases using activitygels, allowing the visualization of the individual activitiesand the determination of their enzyme requirements. Fouractivities of molecular masses 26, 28, 29, and 30 kD wereobserved in the trap fluid upon opening (Fig. 1A, d 0). Thelevel of all of the activities increased for the first 4 d, butbeginning at d 10 (Fig. 1 A) after the opening of the trap, thelevels of the p26 and p28 activities began to decrease. By d28 (Fig. 1A), the p30 activity had disappeared, and by d 35,the p29 activity had declined. These activities have anacidic pH optimum, since they were not observed when thegels were assayed at a neutral pH (Fig. ID). However, aneutral activity of 175 kD was observed (Fig. ID, d 7-42)initiating at 7 d after trap opening. Note that a solid line atapproximately 22 kD is sometimes observed in the gels(e.g. Fig. 1, D or F). This represents a wheat RNase that ispresent at a low level in the RNA purified from wheat

germ, which is the source of the RNA for the activity gels.This wheat RNase can be distinguished from RNases in thesamples by its size and the fact that it is not confined toindividual lanes but rather runs as a uniform line across thegel-

To determine whether these activities are true RNases(i.e. an enzyme that uses single- or double-stranded RNAas the substrate) or are nucleases (i.e. an enzyme that useseither RNA or DNA as the substrate), the samples wereanalyzed using DNase activity gels and the results com-pared with those observed with the RNase activity gels.DNase activity gels differ from RNase activity gels only inthat DNA is used in place of RNA in the resolving gel.Substantial pi75 activity was observed even at 4 d aftertrap opening (Fig. 2A, d 4-42), confirming it as a nuclease,and p29 and p30 activities were also observed (Fig. 2A).These data demonstrate that the p29, p30, and p!75 activ-ities are nucleases, although it is possible that there areRNases of 29 and 30 kD in size, in addition to the observednucleases of 29 and 30 kD, which are not resolved from thenucleases by the activity gels. These data also suggest thatat least the p26 and p28 activities are RNases. The RNasesand nucleases observed in Figures 1 and 2 have beenobserved in the more than 30 traps that have been analyzedin a similar way during a 2-year period.

Days aftertrap opened:

p!7510 13 21 28 35 42

P20.5

p29

p23 •

p20.5 •

Days aftertrap opened:

p!754 7 10 13 21 28 35 42

p205

p20.5 •

Figure 1. Developmental regulation of secretory RNase activity in newly opened traps. Samples were taken at the timesindicated above each lane, after the opening of the trap, and assayed for RNases using RNase activity gels. Water was addedto the trap immediately upon opening (A and D), or withheld for 1 d (B and E) or 13 d (C and F) before samples werecollected. After the resolution of the RNase activities on 12% SDS-PACE gels containing 100 ng/ml_ RNA, the gels werestained with toluidine blue, which stains RNA a dark blue. Bands representing RNase activities appear white because of theabsence of intact RNA. Both nucleases and RNases were detected. The size of each activity is indicated at the left of the gel.A to C, Acidic RNases were revealed by processing the gel (see "Materials and Methods") at pH 5.5. D to F, Neutral RNaseswere revealed by processing the gel at pH 7.5. The results are representative of six independent experiments. www.plantphysiol.orgon May 3, 2020 - Published by Downloaded from

Copyright © 1997 American Society of Plant Biologists. All rights reserved.


1464 Gallie and Chang Plant Physiol. Vol. 115, 1997

Days aftertrap opened: 0

p!75 -1 4 7 10 13 21 28 35 42

A

B p!75-

p50-

p!75-

p50 -

"31^-

Figure 2. Developmental regulation of secretory nuclease activity innewly opened traps. Samples were taken at the times indicated aboveeach lane, after the opening of the trap, and assayed for nucleasesusing DNase activity gels. Water was added to the trap immediatelyupon opening (A), or withheld for 1 d (B) or 13 d (C) before sampleswere collected. After the resolution of the nuclease activities on 12%SDS-PAGE gels containing 10 iJ.g/mL DNA, the gel was stained withethidium bromide, which causes DNA to fluoresce under UV light.Those bands representing nuclease activity appear dark (nonfluoresc-ing) because of the absence of intact DNA. With this assay, onlynucleases are detected. The size of each activity is indicated at theleft of the gel. The results are representative of six independentexperiments.

The first time point of this experiment was the removalof a 1-mL sample from the trap immediately after theaddition of 15 mL of sterile water. Because the trap interioris sterile before its opening, the fluid in the trap is initiallysterile. This was confirmed by plating the sample on Luriabroth medium. No bacterial contamination was detectedfor the first 2 weeks after the opening of a trap, regardlessof whether antibiotics were included in the trap fluid.These data suggest that the hydrolases characterized in theyoung S. purpurea traps were produced by the plant in theabsence of microbial contamination. During the subsequent4 weeks, the bacterial count rose from 1 to 16 CFTJ/mL.When the same experiment was performed with antibioticsadded to the trap fluid, no more than 1 CFU/mL wasdetected. These data suggest that under our control condi-tions the trap fluid was not conducive to bacterial growthwhen only water was present.

To determine whether the presence of water in the trapwas necessary to induce secretion of the RNase and nucle-

ase activities, water was withheld from two additionaltraps that had opened at the time of the first trap (Figs. 1,A and D, and 2A) for 1 d (Figs. 1, B and E, and 2B) and 10 d(Figs. 1, C and F, and 2C). In the trap in which water waswithheld for 1 d, RNase and nuclease activities were al-ready present at the first time point (Figs. IB and 2B) andtheir subsequent expression pattern was similar to thatobserved in the trap in which water was added immedi-ately upon its opening. In the trap in which water waswithheld for 10 d (Figs. 1C and 2C), RNase and nucleaseactivities were as low and remained at that level at thesubsequent time points as was observed for the same de-velopmental stages of the trap in which water was addedimmediately after opening. In addition to the p29, p30, andp!75 nucleases observed in the DNase gel, an additionalnuclease of 31.5 kD was observed faintly (Fig. 2, A and B),and a nuclease of 50 kD was observed in two of the traps(Fig. 2, B and C). Whether this p50 is a proteolytic productof the p!75 nuclease is not known. However, because thep50 expression pattern in the DNA activity gel differs fromthe similar-sized activity in the RNase activity gel (com-pare Fig. 2B with Fig. IE), it may be a distinct activity.These data suggest that the developmental regulation ofthe expression and secretion of the RNase and nucleaseactivities were unaffected by the water status of the trap.

The developmental control of protease activity was alsoinvestigated using samples collected after the opening of atrap as described above. A low level of protease activitywas detected even in the fluid of newly opened traps (Fig.3A, d 0) and increased to a maximum from 7 to 13 d, afterwhich the level of activity decreased, a pattern that wassimilar to the expression pattern of the RNases and some ofthe nucleases. Because reduced nitrogen and phosphate aretwo compounds that have been implicated as target com-pounds for uptake by pitcher plants (Christensen, 1976;Jaffe et al., 1992; Chapin and Pastor, 1995), phosphataseactivity might be expected to be secreted into the trap fluidto complement the expression of the nuclease activities.Phosphatase activity was measured over a wide pH range,and a high level of activity, with a basic pH optima (pH8.0), was detected in fluid from young traps. This activitywas present upon the opening of a trap and steadily in-creased to reach a maximum between 7 and 21 d, afterwhich, in the absence of prey, the activity decreased (Fig.3B). This pattern of expression paralleled that observed forthe RNase and protease activities (compare Fig. 3B withFigs. 1A and 3A). The expression pattern of a phosphataseactive at an acidic pH differed from the basic phosphataseactivity in that its activity increased up to 42 d after trapopening, the latest developmental stage examined (com-pare Fig. 3, C with B). The pH of the trap fluid was alsoexamined during the period of hydrolase expression. Theinitial pH was approximately 7.4 and decreased to a min-imum of 6.5 during the first 7 d after the opening of the trapbefore increasing gradually again to pH 7.5 in the absenceof prey (Fig. 3D). The drop in pH correlated with theobserved increase in RNase, nuclease, protease, and basicphosphatase activities, whereas its subsequent rise corre-lated with the decrease in these activities. The developmen-tal control of the pH of the trap fluid was reproducibly www.plantphysiol.orgon May 3, 2020 - Published by Downloaded from




c 1500

5 l.MM-

1.250

2 0.750 - 3

0.500 -

0.250 -

navsafter n - " uapopenmg. O 1 4 7 10 13 21 28 35 42

B 1.250

2 0.750 9 -f 0.500

0.250

Davs after o uapopening: O 1 4 7 10 13 21 28 35 42

0.250

Davs after o uapopening: O 1 4 7 10 13 21 28 35 42

6.50

Days uapopening: after 6.00 6.25 1 O 1 4 7 10 13 21 28 35 42

Figure 3. Developmental regulation of pH and secretory protease and phosphatase activities in the fluid of newly opened traps. Sam- ples were taken at the times indicated below each histogram from traps, to which water was added immediately after their opening, and assayed in duplicate for protease (A), neutra1 phosphatase (B), or acidic phosphatase (C) activity and the average reported. The pH (D) of the fluid was also measured during the same developmental period. The results are representative of four independent experiments.

observed in newly opened traps. These results document that both the pH of the trap fluid and the expression of several secreted hydrolase activities are subject to developmental control.

Because insects make up the bulk of the prey found in pitcher traps, chitinase activity might be expected to be present in the trap fluid. To examine this possibility the samples used for the protease assay were used in a chitinase assay. In contrast to the observations with the other hydrolases examined, no chitinase activity was detected when chitin azure was used in the assay, or even when the more sensitive substrate 4-methylumbelliferyl P-D-N,N',NII-

triacetylchitotriose was used (data not shown). The absence of detectable chitinase activity may explain why the ex- oskeletons of prey typically remain intact in S. puvpurea traps even after a prolonged period.

Hydrolase Expression 1s lnduced by the Presence of Nutrients in a Trap

The reduction in the expression of RNase, nuclease, phosphatase, and protease activities in the first 6 weeks after the opening of a trap may be a regulated response to the absence of prey, or may represent the normal period of time that a trap engages in carnivory, after which it may develop solely as a photosynthetic organ. To test whether traps can perceive and respond to the presence of nutrients in the trap fluid, 4-month-old traps that had greatly reduced or nondetectable levels of RNase, nuclease, phosphatase, or protease activities in the trap fluid were used for stimulus-response studies. Fifteen milliliters of sterile water was added to a freshly washed trap that had been tested for a low level of hydrolase expression. After 1 week, to allow any residual secretory activity from the trap to reach a steady-state level, 600 pg of purified wheat germ RNA, 1 mg of herring sperm DNA, or 1 mg of purified BSA protein was added to the fluid of the trap. Samples of the fluid were removed at 2-d intervals up to 8 d after the initiation of the feeding and assayed first using M a s e activity gels.

Only a low level of activity was observed just before the initiation of feeding with BSA (Fig. 4A, lane l) , which represents the low level of hydrolase expression in response to the absence of nutrients in the trap fluid for 4 months. However, within 2 d after the addition of the BSA, induction of the p20.5, p21.5, p26, and p30 activities was observed (Fig. 4A, lane 2), and by 4 d the induction of the p29 activity was also observed (Fig. 4A, lane 3). The p20.5 and p21.5 activities are identified as RNases because they are observed in RNase activity gels (Fig. 4) but not in DNase activity gels (see Fig. 5). The pH of the trap fluid also decreased from an initial value of 7.05 to 6.2 during the experiment (Fig. 4D). Induction of several RNase activities was also observed when either RNA or DNA was introduced into the traps (Fig. 4, B and C, respectively), although some differences in expression were observed com- pared with the results obtained when BSA was used. The addition of RNA resulted in the induction of the p20.5 and p21.5 activities by 2 d (Fig. 48, lane 2) and the induction of the p30 nuclease by 4 d (Fig. 4B, lane 3). Curiously, the p30 activity decreased at 6 d and increased again at 8 d, although the level of the p20.5 and p21.5 activities remained constant during the same period and therefore they serve as interna1 controls for the p30 activity. This apparent cycling of the p30 activity was reproducibly observed in several repetitions of the RNA feeding experiment using several independent traps. Induction of the p20.5, p21.5, and p26 activities were observed after the introduction of DNA to a trap (Fig. 4C, lanes 2-5), similar to the observations after feeding with BSA (Fig. 4A, lanes 2-5). However, the induction of the p29 activity that was observed after feeding with BSA (Fig. 4A, lane 3), or the induction of the




A1 2 3 4 5

Days after addition: 0

Bp21.5p20.5

Days after addition: 0 2 4 6 8


DPH

7.25-

7.00-

6.75-

6.50-

6.25-

6.00Days after addition: 0 2 4 6 8

Figure 4. Induction of RNase activity by nucleic acids or proteinadded to traps. Samples were taken at the times indicated below eachlane after the feeding. RNase expression from traps fed with BSA (A),RNA (B), or DNA (C) was determined using RNase activity gels. Bothnucleases and RNases were detected. The size of each activity isindicated at the left of the gel. The pH (D) of the fluid of the trap fedwith BSA was also measured during the same period. The results arerepresentative of five independent experiments.

p30 activity after feeding with RNA (Fig. 4B, lane 4), wasnot observed after feeding with DNA (Fig. 4C). No induc-tion of hydrolase activity was observed in control (wateronly) traps of the same plants that were used in thesestudies (data not shown), which suggests that each trapresponds independently to the appearance of nutrients.

The extent to which the other hydrolytic activities wereinduced after the feeding with DNA, RNA, or BSA was alsoexamined. Only a low level of the p29 and p30 nucleasescould be detected in the trap fluid before the feeding (Fig.5A, lane 1). Addition of DNA to the trap resulted in theinduction of the p31.5, p36, p47, and p50 nucleases to alevel that was just detectable after 2 d (Fig. 5A, lane 2) and

reached a maximum level by 6 d (Fig. 5A, lane 4). The p30nuclease activity was also induced but the activity of thep29 nuclease increased only marginally (Fig. 5A, lanes 3-5).In the trap to which RNA was added, activity of the p30nuclease was greater than that of the p29 nuclease (Fig. 5B,lane 1), and induction of the p47 nuclease was first ob-served at 2 d (Fig. 5B, lane 2), of the p36 and p!75 nucleasesat 4 d (Fig. 5B, lane 3), and of the p31.5 nuclease at 8 d (Fig.5B, lane 5). The addition of BSA to a trap resulted in therapid induction of all nucleases (Fig. 5, C and D, lanes 2-5).The induction of the pi75 nuclease can be seen more clearlyon the 8% gel in Figure 5D, and a low level of anothernuclease (p70) was also detected faintly (Fig. 5D, lanes 4and 5). Whether this might be a proteolytic product of thepi 75 nuclease is not known.

Protease activity was induced in traps in which RNA,DNA, or protein had been added, with BSA having thegreatest inducing effect on protease expression (Fig. 6C),followed by DNA (Fig. 6B), and finally RNA (Fig. 6A). Incontrast to the results demonstrating that the protease,RNase, and nuclease activities are induced by the additionof RNA, DNA, or protein to a trap, phosphatase activitywas induced only after feeding with BSA (compare Fig. 6,F to D and E). Induction of phosphatase activity was ob-served 4 d after the addition of protein to a trap (Fig. 6F, d4), but the highest level of activity did not occur until 6 d(Fig. 6F, d 6-8). These data demonstrate that pitcher trapsare competent to respond to the presence of signals presentin the trap fluid, even after the initial expression of thehydrolases had substantially decreased in the absence ofprey. Nevertheless, there was some difference in the induc-

Days afteraddition: 0 2 4 6

1 I f I H B

Days afteraddition:f~\ p!75

1 2 3 4 5 2 3 4 5Figure 5. Induction of nuclease activity by nucleic acids or proteinadded to traps. Samples were taken at the times indicated below eachlane after the feeding. Nuclease expression from traps fed with DNA(A), RNA (B), or BSA (C and D) was determined using DNase activitygels. The nucleases were resolved using 12% (A-C) or 8% (D)SDS-PAGE gels. The size of each activity is indicated at the left of thegel. The results are representative of three independent experiments. www.plantphysiol.orgon May 3, 2020 - Published by Downloaded from




2 0.600 -

2 0.400 - 2

4 0.200 - 3

$ 0.150

Q1 0.100 2.

0.050

O

0

Days after addition: O 2 4 6 8

0.150 3 E - 0.100

4 0.050

O Days after addition: O 2 4 6 8

2 0.250 4

n U

. h 2 0.200 m 5 0.150

0.100

0.050

Days after addition: O 2 4 6 8

0 1 ’ ” Days after addition: O 2 4 6 8

E 0.800

E, 0.600

2 0.400 3 4 0.200


1.400 1

p 1.000

3 0.800

0.400

0.200

0.600

O

r

b,

Days afteraddition: O 2 4 6 8

Figure 6. lnduction of protease and phosphatase activities after the addition of nucleic acids or protein to traps. Samples were taken at the times indicated below each histogram after the feeding. Each sample was assayed in duplicate and the average reported. Protease (A-C) and phosphatase (D-F) expression from traps fed with RNA (A and D), D N A (B and E), or BSA (C and F) was determined. The results are representative of four independent experiments.

tion of the individual hydrolases; some activities were coordinately induced by the addition of nucleic acids or protein, whereas the induction of phosphatase activity was observed only after the addition of protein to a trap. More- over, severa1 of the RNases or nucleases were induced to different extents, depending on whether RNA, DNA, or protein was used as the induction signal. No chitinase was detected in the traps after the introduction of nucleic acids or protein (data not shown). For these nutrient-response studies, the range of bacteria present varied from O to 3 CFU/mL. Even though significant differences in hydrolase expression were observed in traps, depending on whether protein, RNA, DNA, or water alone was introduced, no significant difference in the CFU was observed. These observations suggest that the observed regulation of hydrolase expression resulted from S. purpureu alone.

Reduced Nitrogen Serves as an Effective Signal for the lnduction of Hydrolytic Enzyme Expression

A necessary prerequisite for the hydrolase induction response by a pitcher plant is the ability to perceive the presence of prey. To identify compounds that might serve as signals to induce hydrolytic enzyme expression, NH,+ and phosphate, compounds previously found limiting in S. purpureu (Chapin and Pastor, 1995), were used in stimulus- response studies with traps that had nondetectable levels

of hydrolase activities in the trap fluid. As before, 15 mL of sterile water was added to a trap for 1 week to allow any secretory activity from the trap to reach a steady-state level, at which point NH,Cl, NaC1, KC1, or sodium phosphate (pH 6.8), to a final concentration of 50 mM, was added individually to 4-month-old traps, and samples were removed at 2-d intervals up to 8 d after the initiation of the feeding.

We first examined the samples using RNase activity gels to determine whether any of the compounds added to the trap fluid could serve as inducers (Fig. 7, A-E). Expression of the p30, p26, p21.5, and p20.5 activities increased from nondetectable levels within 2 d after feeding with NH,C1 (Fig. 7A, compare d 2 to d O), and continued at a high level for the duration of the 8-d period of the experiment (Fig. 7A, d 2-8). In contrast, the addition of sodium phosphate resulted in only a moderate stimulation of the expression of the p30 activity (Fig. 7B, compare d 2 to d O), with little increase in the expression of the p21.5 or p20.5 RNases. The expression of the p30 activity was slightly stimulated in response to the addition of KC1 (Fig. 7C, compare d 2 to d O) but NaCl or water had little effect on the expression of any activity (Fig. 7, D and E, respectively). Note that the trap containing only water was from the same plant used for the NH,C1 feeding to control for any whole-plant effect. The lack of any hydrolase expression from the water-






p21.5p20.5


Days after addition: 0 2

Days after addition: 0

E

8

0.200 -I

0.160 -

0.120 -

0.080

0.040 •

KpH

7.50 -i

6.75 •

6.00

5.25-

4.50

3.75Days after addition: 0 2 4 6 8 Days after addition: 0 2 4 6 8

G

i

0.200 •

0.160

0.120

0.080

0.040

LpH

7.50 i

6.75

6.00-

5.25-

4.50-

3.75Days after addition: 0 2 4 6 8 Days after addition: 0 2 4 6 8

H 0200

0160

0120

0.080

0.040

MpH

7.50-

6.75 -

6.00-

5.25-

4.50

Days after addition: 0 2 4 6 83.75-

Days after addition: 0 2 4 6

I!_<

gvi1/1<

0200 -

0.160

0.120 '

0.080 -

0040 '

o -™Days after addition: 0

ji-

"̂i"inO*>

<

0.200

0 160

0 120 •

0080 '

0.040 '

n - ̂

N

pH

7.50-

6.75-

6.00-

5.25-

4.50-

3.75-Days after addition: 0 2 4 6

OPH

7.50-

6.75-

600-

5.25-

4.50

Days after addition: 0 2 4 63.75-

Days after addition: 0 2 4 6

Figure 7. Reduced nitrogen serves as an efficient inducer of RNase and protease activities. Four months after the openingof the traps, when hydrolytic expression was determined to be low, the induction of RNase (A-E), protease (F-J), or thechange in the pH of the trap fluid (K-O) was determined after the addition of NH4CI (A, F, and K), sodium phosphate (pH6.8) (B, C, and L), KCI (C, H, and M), or NaCI (D, I, and N) to a final concentration of 50 mM. A water-only control (E, J,and O) was included. Samples were taken at the times indicated below each lane after the feeding. RNase expression fromtraps was determined using RNase activity gels. The size of each activity is indicated at the left of the gel. The results arerepresentative of three independent experiments.




5 0.600 -

2 0.400 - 3

J? 0.400 3 4 0.200

O Days afteraddition: O 2 4 6 8

0.600

2 0.400

B 2

4 0.200 s

O Days after addition: 0 2 4 6 8

c 0.800

VI - 0.400 3

E 0.800 ',Oo0]

J? 0.400 s 4 0.200


F 0.800 '.Oo0]

VI 2 0.400 .j

4 0.200 L , II O

Daysafter addition: 0 2 4 6 8

G 0.800 '.Oo0]

2 0.400 "v

Days after addition: 0 2 4 6 8 Days after addition: 0 2 4 6 8

D 0.800 '.Oo0] H 0.800 7

Figure 8. lnduction of neutral and acidic phosphatase activities in response to chemical inducers. Four months after the opening of the traps, when phosphatase expression was determined to be low, the induction of neutral (A-D) and acidic (E-H) phosphatase activities was determined after the addition of NH,CI (A and E, respectively), KCI (B and F, respectively), or NaCl (C and C, respectively), to a final concentration of 50 mM. A water-only control was included (D and H, respectively). Samples were taken at the times indicated below each histogram after the feeding, assayed in duplicate, and the average reported. The results are representative of three independent experiments.

-

0 o.200 O L o.200 L"m Days after addition: 0 2 4 6 8 Days after addition: 0 2 4 6 8

control trap again demonstrates the independence of each trap to respond to inducers.

The expression of protease activity in response to the chemical inducers was similar to RNase and nuclease expression. Protease expression was substantially induced by NH4C1 (Fig. 7F), to a lesser extent by sodium phosphate (Fig. 7G), and not significantly by KC1, NaC1, or the water control (Fig. 7, H-J, respectively). The pH of the trap fluid was also regulated by the type of compound added to a trap. Addition of NH,C1 resulted in a rapid and marked reduction of pH in the trap fluid from an initial neutral pH to <4.5 (Fig. 7K). A reduction in the pH was also observed after the addition of either KC1 (Fig. 7M) or NaCl (Fig. 7N) that was greater than that observed for the water control (Fig. 70). Little change in pH was observed in trap fluid in which sodium phosphate (pH 6.8) had been added (Fig. 7L), as would be expected for this buffer. NH4C1 also stimulated the expression of both the acidic and neutral phosphatases (Fig. 8, A and E), whereas KC1 was substantially less effective (Fig. 8, B and F), and NaCl (Fig. 8, C and

G) and water (Fig. 8, D and H) had no detectable effect at all. Phosphatase induction by the addition of sodium phosphate to the trap fluid could not be assayed because the presence of the phosphate in the trap fluid repressed phosphatase activity in the assay. The observation that NH,C1 had a substantially greater effect on hydrolase expression than did NaCl suggests that NH4+, and not C1-, serves as the most effective signal in the induction of hydrolase enzyme expression of those compounds tested.

DISCUSSION

Carnivory has independently evolved in numerous species within the plant kingdom as a means to augment the acquisition of resources from the otherwise marginal environments that these species usually inhabit. In contrast to the usual mechanisms that plants employ to obtain minerals and nitrogen from the soil, carnivory in plants requires an investment of resources to capture and digest prey and, consequently, represents a risk that is particularly acute for



1470 Gallie and Chang Plant Physiol. Vol. 1 1 5, 1997

a plant growing in a poor environment. The cost of carnivory could be reduced by regulating the expression of those genes involved in the carnivorous process so that they are produced only in response to the presence of prey. For example, the Venus’s-flytrap (Dionaea muscipula) has evolved an active trapping mechanism that links the secretion of hydrolases to the capture of prey (Scala et al., 1969; Robins, 1976). However, the passive trapping mechanism through drowning, which is employed by the common pitcher plant, would require that this species either consti- tutively produce the hydrolases needed for digestion, re- gardless of whether prey is present, or employ a mechanism that perceives and responds to the presence of prey. The present study reveals that this carnivorous species uses a combination of both strategies that may be designed to maximize the benefit from carnivory, but to limit the costs associated with this evolutionary adaptation.

Several secreted hydrolase activities, including RNases, nucleases, phosphatases, and proteases, were used as marker enzymes to follow the regulation of carnivory during the development of traps or in response to the presence of resources in trap fluid. Hydrolase secretion commenced upon the opening of a trap and continued to accumulate for approximately 1 week. The induction of hydrolase expression was independent of whether water was present in a trap. The lack of an effect by the exogenous addition of water to a trap is not surprising in that Sarracenia spp. produce a liquid secretion that contains the hydrolase activities described in this study upon trap opening even in the absence of water (data not shown). In the absence of prey, however, hydrolase expression eventually declines and reaches a low leve1 of expression within a few weeks after the opening of a trap. Nevertheless, nonactive traps that express low to nondetectable levels of hydrolases remain competent to resume carnivory through the induction of hydrolase expression in response to the presence of the appropriate chemical signals within the fluid of the trap.

The regulation of hydrolase expression through a combination of developmental control and signal transduction reveals two evolutionary aspects of carnivory. First, the commensurate production of hydrolases upon opening of a trap represents a gene-based commitment to carnivory before the capture of prey. Second, the reduction in hydrolase expression in the absence of prey, and the subsequent induction of the same activities after the perception of the appropriate signals, suggests that the developmental commitment to carnivory in young traps is replaced by a signal transduction mechanism upon the maturation of a trap. This mechanistic switch provides a means to limit the cost of carnivory when a trap is not successful in capturing prey. The observation that more than 50% of Sarracenia spp. traps fail to capture prey (Cresswell, 1991) suggests that the ability to reduce the cost of carnivory in the absence of prey would indeed benefit this species in its natural environment. Moreover, the signal transduction mechanism would ensure that a trap could respond to a future capture event to benefit from the availability of resources.

Induction of hydrolase expression in response to prey requires. a mechanism to perceive its presence. Perception of prey might involve mechanical stimulus, as in the

Venus’s-flytrap, in which trap closure and secretion of hydrolases occur through the mechanical stimulation of specialized trigger hairs by the prey (Robins, 1976). No induction of hydrolase expression was observed by mechanical stimulation of the traps of Sarracenia spp. (data not shown). However, the addition of nucleic acids or protein was sufficient to induce the expression of severa1 hydrolases and the amount of the nucleic acids or protein used was equivalent to approximately three to five Drosopkila flies, based on nitrogen content (Dixon et al., 1980).

Hydrolase induction was trap specific in that, after the addition of an inducer to one trap, the induction of hydrolase activities was never observed in the other traps of the same pitcher plant. This observation demonstrates that each trap can respond independently to the appearance of nutrients in its trap fluid, which is another means by which the expenditure of the resources needed for carnivory can be minimized. RNases and nucleases could be induced not only by the presence of nucleic acids, but also by protein. Likewise, protease activity was induced by the presence of either protein or nucleic acids in the trap fluid. In contrast, phosphatase activity was induced by protein in the trap fluid but not by nucleic acid, indicating that the expression of the hydrolases is not completely coordinately regulated by the availability of nutrients. This suggests additional complexity to the signal transduction mechanism involved in responding to changes in the trap fluid.

The pH of freshly opened traps was in the neutral range but decreased by 1 pH unit within 4 d. However, in the absence of prey the pH of the trap fluid gradually increased again during the ensuing 5 weeks. A similar reduction of approximately 2 pH units in the fluid within pitchers of the unrelated carnivorous genus Nepenthes has been observed after trap opening (Morrissey, 1955; Nakayama and Ama- gase, 1968; Tokes et al., 1974; Heslop-Harrison, 1975). Rapid acidification of the trap fluid was also observed after the addition of nucleic acids, protein, or ions such as NH,+, K+, or Na+ to a trap. A similar decrease in the pH of the trap fluid in Nepenthes spp. after the capture of prey has been observed, suggesting that this may be a characteristic component of passive carnivory (Tokes et al., 1974). The decrease in pH after the opening of a trap or after feeding correlated with the observed increase in RNase, nuclease, phosphatase, and protease activities, and the subsequent increase in pH in the absence of prey correlates with the reduction in these secreted hydrolase activities. Several of these hydrolytic activities have pH optima that are mod- erately acidic. It is possible that the reduction of the pH is developmentally controlled to create optimal conditions for the secreted hydrolytic activities, or itself may serve as a signal to induce hydrolase gene expression. The induction of some hydrolase activities in response to sodium phosphate (pH 6.8), which maintained the pH of the trap fluid at a neutral pH (Fig. 7B), suggests that acidification of the trap fluid may not be a necessary prerequisite to the induction of hydrolase expression.

Reduced nitrogen in the form of NH4+ was identified as a compound that could serve as an effective signal for the induction of hydrolase expression. Phosphate and potassium were also stimulatory but much less so than NH4+. In




contrast to the well-documented examples of the induction of RNase expression in response to phosphate starvation in plant species (Nurnberger et al., 1990; Jost et al., 1991; Bariola et al., 1993; Loffler et al., 1993), the phosphate- induced expression of RNase activity in Sarracenia spp. would be an example of induction in response to available phosphate. Both nitrogen and phosphate have been implicated as the goals of the carnivory in Sarracenia spp. in that nitrogen was identified as the limiting nutrient followed by phosphate (Christensen, 1976; Chapin and Pastor, 1995). It has also been suggested that Sarracenia spp. rely on carnivory to acquire K+ (Plummer and Kethley, 1964), although direct feeding experiments have demonstrated that the uptake of nitrogen was greatest, followed by phosphate and potassium (Jaffe et al., 1992). Darwin observed that traps of the carnivorous plant Nepenthes could distinguish between ”useful” and ”nonuseful” matter and secrete digestive fluid only in response to the former, and suggested that nitrogenous compounds functioned as the stimulators (Darwin, 1875). However, Pfeffer suggested that the plants sought both nitrogen and phosphorus from their prey (Pfeffer, 1877). Our observations support both of these early hypotheses in that reduced nitrogen served as the most effective inducer, but was not the sole inducing compound. Such inducers presumably are released from the insect through bodily excretions during its struggle to es- cape, or from the release of hemolymph from the insect after its death.

Our observations also suggest that Sarraceniu spp. traps are developmentally prepared upon opening, or soon thereafter, to engage in carnivory, and that this is soon replaced by a mechanism that allows the plant to regulate its investment in carnivory in response to the presence or absence of prey. Those signals that can serve as inducers of hydrolase expression represent the compounds that are limiting in the habitats of carnivorous plants. Conse- quently, Sarracenia spp. appear to have evolved a highly regulated carnivorous mechanism, combining signal transduction and developmental controls, to minimize the risks and costs associated with carnivory as a means to maximize the benefits of extracting nutrients from their prey.

Received June 11, 1997; accepted August 28, 1997. Copyright Clearance Center: 0032-0889/97/115/1461/11.

LITERATURE ClTED

Albert VA, Williams SE, Chase MW (1992) Carnivorous plants: phylogeny and structural evolution. Science 257: 1491-1495

Bariola PA, Howard CJ, Taylor CB, Verburg MT, Jaglan VD, Green PJ (1993) The Arabidopsis ribonuclease gene RNSl is tightly controlled in response to phosphate limitation. Plant J 6:

Blank A, McKeon TA (1991) Three RNases in senescent and nonsenescent wheat leaves. Plant Physiol 97: 1402-1408

Blank A, Sugiyama H, Dekker CA (1982) Activity staining of nu- cleolytic enzymes after sodium dodecyl sulfate-polyacrylamide gel electrophoresis: use of aqueous isopropanol to remove de- tergent from gels. Ana1 Biochem 120: 267-275

Chapin CT, Pastor J (1995) Nutrient limitations in the northern pitcher plant Sarracenia purpurea. Can J Bot 73: 728-734

673-685

Christensen NL (1976) The role of carnivory in Sarracenia flava L. with regard to specific nutrient deficiencies. J Elisha Mitchell Sci SOC 92: 144-147

Compton MM, Cidlowski JA (1987) Identification of a glucocorticoid-induced nuclease in thymocytes: a potential “ly- sis gene” product. J Biol Chem 262: 8288-8292

Cresswell JE (1991) Capture rates and composition of insect prey of the pitcher plant Sarracenia purpurea. Am Midl Nat 125 1-9

Darwin C (1875) Insectivorous Plants. John Murray, London Danvin F (1878) Experiments on the nutrition of Drosera rotundi-

folia. J Linn SOC Bot 17: 17-32 Dixon KW, Pate JS, Bailey WJ (1980) Nitrogen nutrition of the

tuberous sundew Drosera erythrorhiza Lindl. with special refer- ente to catch of arthropod fauna by its glandular leaves. Aust J

Dore Swamy R, Mohan Ram HY (1969) Studies on growth and flowering in axenic cultures of insectivorous plants. I. Seed germination and establishment in cultures of Utricularia inflexa. Phytomorphology 19: 363-371

Fish D, Hall DW (1978) Succession and stratification of aquatic insects inhabiting the leaves of the insectivorous pitcher plant Sarracenia purpurea. Am Midl Nat 112 294-304

Gibson TC (1983) Competition, disturbance and the carnivorous plant community in south eastern U.S. PhD thesis. University of Utah, Salt Lake City

Heslop-Harrison Y (1975) Enzyme release in camivorous plants. In JT Dingle, RT Dean, eds, Lysozymes in Biology and Pathol- ogy, Vol 4. North Holland Publishers, Amsterdam, pp 525-578

Jaffe K, Michelangeli F, Gonzalez JM, Miras B, Ruiz MC (1992) Carnivory in pitcher plants of the genus Heliamphora (Sarrace- niaceae). New Phytol 122 733-744

Jost W, Bak H, Glund K, Terpstra P, Beintema JJ (1991) Amino acid sequence of an extracellular, phosphate-starvation-induced ribonuclease from cultured tomato (Lycopersicon esculentum) cells. Eur J Biochem 198: 1-6

Loffler A, Glund K, Irie M (1993) Amino acid sequence of an intracellular, phosphate-starvation-induced ribonuclease from cultured tomato (Lycopersicon esculentum) cells. Eur J Biochem

Morrissey S (1955) Chloride ions in the secretion of the pitcher plant. Nature 176: 1220-1221

Nakayama S, Amagase S (1968) Acid protease in Nepenthes: partia1 purification and properties of the enzyme. Proc Jpn Acad 44

Nap J-P, Bisseling T (1990) Developmental biology of a plant- prokaryote symbiosis: the legume root nodule. Science 250:

Nurnberger T, Abel S, Jost W, Glund K (1990) Induction of an extracellular ribonuclease in cultured tomato cells upon phosphate starvation. Plant Physiol 92: 970-976

Pfeffer W (1877) Ueber fleischfressende Pflanzen und iiber die Ernahrung durch Aufnahme organischer Stoffe iiberhaupt. Landwirtsch Jahrb 6: 969-998

Plummer GL (1963) Soils of the pitcher-plant habitats in the Geor- gia coastal plain. Ecology 44: 727-734

Plummer GL, Kethley JB (1964) Foliar absorption of amino acids, peptides, and other nutrients by the pitcher plant Sarracenia flava. Bot Gaz 125: 24.5-260

Robins RJ (1976) The nature of the stimuli causing digestive juice secretion in Dionaea muscipula Ellis (Venus’s flytrap). Planta 128:

Scala J, Iott K, Schwab DW, Semersky FE (1969) Digestive secretion of Dionaea muscipula (Venus’s-flytrap). Plant Physiol 44

Tokes ZA, Woon WC, Chambers SM (1974) Digestive enzymes secreted by the carnivorous plant Nepenthes macfarlanei L. Planta 119 39-46

Verma DPS (1992) Signals in root nodule organogenesis and en- docytosis of Rhizobium. Plant Cell 4: 373-382

Bot 28: 283-297

214: 627-633

358-362

948-954

263-265

367-371



Documents

Signal Transduction in the Carnivorous Plant Sa rra cenia purpurea’ · Plant Physiol. (1997) 115: 1461-1471 Signal Transduction in the Carnivorous Plant Sa rra cenia purpurea’