Transgene-Mediated and Elicitor-Induced Perturbation of Metabolic Channeling at the Entry Point into the Phenylpropanoid Pathway

Correspondence to: Richard A. Dixon, radixon{at}noble.org (E-mail), 580-221-7380 (fax)

	ABSTRACT

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

³H–L-Phenylalanine is incorporated into a range of phenylpropanoid compounds when fed to tobacco cell cultures. A significant proportion of ³H–trans-cinnamic acid formed from ³H–L-phenylalanine did not equilibrate with exogenous trans-cinnamic acid and therefore may be rapidly channeled through the cinnamate 4-hydroxylase (C4H) reaction to 4-coumaric acid. Such compartmentalization of trans-cinnamic acid was not observed after elicitation or in cell cultures constitutively expressing a bean phenylalanine ammonia–lyase (PAL) transgene. Channeling between PAL and C4H was confirmed in vitro in isolated microsomes from tobacco stems or cell suspension cultures. This channeling was strongly reduced in microsomes from stems or cell cultures of transgenic PAL-overexpressing plants or after elicitation of wild-type cell cultures. Protein gel blot analysis showed that tobacco PAL1 and bean PAL were localized in both soluble and microsomal fractions, whereas tobacco PAL2 was found only in the soluble fraction. We propose that metabolic channeling of trans-cinnamic acid requires the close association of specific forms of PAL with C4H on microsomal membranes.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

The phenylpropanoid pathway is involved in the biosynthesis of a wide variety of natural products from plants. Many of these products have important functions in plant development and in interactions of the plant with its environment (Hahlbrock and Grisebach 1979 ; Hahlbrock and Scheel 1989 ; Dixon and Paiva 1995 ). Many studies have addressed the transcriptional regulation of genes encoding enzymes of the phenylpropanoid pathway and subsequent changes in extractable enzyme activities in response to developmental and environmental cues (Cramer et al. 1985 ; Lawton and Lamb 1987 ; Hahlbrock and Scheel 1989 ). Far less attention has been paid to how the cell regulates flux into different end products of the pathway once all of the enzymatic machinery is assembled.

The first committed step in the biosynthesis of phenylpropanoid compounds is the conversion of L-phenylalanine (Phe) to trans-cinnamic acid by L-Phe ammonia–lyase (PAL; EC 4.3.1.5; Figure 1). PAL is a tetrameric enzyme whose subunits are encoded by a multigene family in most species that have been studied (Cramer et al. 1989 ; Nagai et al. 1994 ; Wanner et al. 1995 ; Fukasawa-Akada et al. 1996 ). PAL genes are transcriptionally activated after microbial infection or treatment of plant cells with microbial elicitors (Edwards et al. 1985 ; Lawton and Lamb 1987 ). The second step in the phenylpropanoid pathway, the hydroxylation of trans-cinnamic acid to 4-coumaric acid, is catalyzed by a cyto-chrome P450 monooxygenase, cinnamic acid 4-hydroxylase (C4H; EC 1.14.13.11; Russell and Conn 1967 ; Fahrendorf and Dixon 1993 ; Teutsch et al. 1993 ). C4H is induced by light, elicitors, and wounding (Fahrendorf and Dixon 1993 ; Buell and Somerville 1995 ; Batard et al. 1997 ; Bell-Lelong et al. 1997 ), and its induction often is closely coordinated with PAL induction (Mizutani et al. 1997 ).

View larger version (19K): [in this window] [in a new window]   Figure 1. Biosynthesis of Phenolic Compounds in Tobacco. L-Phe is deaminated by PAL to yield trans-cinnamic acid (t-CA), which is converted by C4H to 4-coumaric acid (4-CA). 4-Coumaric acid can be hydroxylated by coumarate 3-hydroxylase (C3H) to caffeic acid (CafA), which is acted upon by an O-methyltransferase (OMT) to yield ferulic acid (FerA). Salicylic acid (SA) and pHBA are formed via chain shortening of trans-cinnamic acid and 4-coumaric acid, respectively, although the specific mechanisms are not clear. Similarly, the exact biosynthetic origins of vanillin (Van), vanillic acid (VA), and coumarins (such as scopoletin in tobacco) have not been determined definitively; these pathways are indicated by dotted lines and question marks. Benzoic acid (BA) is most probably the precursor of salicylic acid, which is formed by the action of a benzoic acid 2-hydroxylase (BA2H). R1 and R2 are commonly substituted positions in coumarins. In scopoletin, R1 = OCH3 and R2 = OH.

Enzymes of complex metabolic pathways may be present in the cell in arrays of consecutive, physically associated enzymes that are assembled on membranes or other physical structures (Subramanian et al. 1973 ; Margna and Margna 1978 ; Cutler and Conn 1981 ; Cutler et al. 1981 ; Srere 1987 ). Such enzyme organization can result in the channeling of pathway intermediates without their release into general metabolic pools (Stafford 1981 ; Hrazdina 1992 ). Cytochrome P450 enzymes, such as C4H, are anchored to the external surface of the endoplasmic reticulum (Chapple 1998 ), but PAL has been regarded generally as an operationally soluble enzyme. However, studies performed with microsomes isolated from potato (Czichi and Kindl 1975 ) and cucumber cotyledons (Hrazdina and Wagner 1985 ) have suggested that PAL and C4H activities are colocalized on membranes of the endoplasmic reticulum. Furthermore, trans-cinnamic acid formed endogenously via the PAL reaction is a better substrate for C4H than externally added trans-cinnamic acid in in vitro assays. This finding has been interpreted as evidence for channeling of trans-cinnamic acid in the conversion of Phe to 4-coumaric acid (Czichi and Kindl 1975 , Czichi and Kindl 1977 ; Hrazdina and Wagner 1985 ; Hrazdina and Jensen 1992 ).

To gain further insight into the phenomenon of metabolic channeling, it is necessary to develop a model system that allows for comparison of metabolic compartmentalization/channeling in vivo and in vitro. This system must be amenable to manipulation to increase or decrease various components of the potential channel, and molecular information on the gene products involved in channeling must be available. Tobacco represents such a system. We have generated a series of transgenic tobacco plants containing a heterologous bean PAL2 gene, in which the levels of extractable PAL activity are increased or decreased when compared with wild-type plants (Elkind et al. 1990 ; Howles et al. 1996 ; Sewalt et al. 1997 ). These plants allow assessment of the effects of quantitative and qualitative changes in a component of a potential metabolic channel on the operation of that channel. Furthermore, it is now possible to design probes to distinguish between different members of the PAL gene family and therefore to address which forms of PAL may be associated with channeling.

Here, we present evidence, from in vivo and in vitro labeling experiments with ³H–L-Phe and ¹⁴C–trans-cinnamic acid, for metabolic channeling that involves coupling of PAL and C4H in tobacco stem tissue and cell suspension cultures. We demonstrate that specific forms of PAL are associated with tobacco microsomes and that microsomal association of a heterologous PAL enzyme in transgenic plants can perturb channeling. We also show that metabolic channeling is no longer measurable after activation of the phenylpropanoid pathway by elicitation.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

Metabolic Compartmentalization of trans-Cinnamic Acid in Tobacco Cell Suspension Cultures
If a biosynthetic intermediate is channeled, it will not equilibrate freely with an externally added intermediate, and the endogenously formed and externally supplied compound will exist in the cell in different pools. In vivo labeling experiments can show the existence of different metabolic pools, although these pools may or may not result from metabolic channeling. Conversely, demonstration of a single pool of a metabolite that fully equilibrates with the externally added compound would constitute evidence against channeling. Therefore, we used tobacco cell suspension cultures to confirm the existence of more than one pool of trans-cinnamic acid in tobacco as a precursor to further studies on channeling between PAL and C4H that make use of transgenic plants and in vitro assays. When fed ³H–L-Phe, such cultures accumulate a range of labeled phenolic compounds, including the conjugates of trans-cinnamic acid, 4-coumaric acid, caffeic acid, and ferulic acid, as shown in Figure 2A. These compounds are derived from Phe via trans-cinnamic acid and 4-coumaric acid, as shown in Figure 1.

View larger version (15K): [in this window] [in a new window]   Figure 2. HPLC Traces Showing Incorporation of Tritium from 3H–L-Phe into Phenylpropanoid Compounds in Tobacco Cell Suspension Cultures. (A) Wild-type tobacco cell suspension cultures. (B) Tobacco cell suspension cultures overexpressing PAL from the bean PAL2 gene (OX434). Cultures were treated simultaneously with 10-4 M 3H–L-Phe and unlabeled trans-cinnamic acid (10-4 M). Cells were harvested 24 hr after the addition of the precursors. Extracts of soluble phenolics were treated with esterase and separated by HPLC. Fractions were collected, and radioactivity in trans-cinnamic acid (t-CA), 4-coumaric acid (4-CA), caffeic acid (CafA), and ferulic acid (FerA) was determined by liquid scintillation counting. Au, absorbance units.

If suspension-cultured tobacco cells possessed only a single pool of trans-cinnamic acid with which ³H–trans-cinnamic acid formed by the PAL reaction from ³H-Phe readily equilibrated, simultaneous feeding of ³H-Phe and unlabeled trans-cinnamic acid would result in a lowering of the specific activity of tritium in the various phenylpropanoid compounds derived from trans-cinnamic acid when compared with the values obtained in the absence of unlabeled trans-cinnamic acid. At the end of the labeling period, the specific activity of reisolated trans-cinnamic acid would represent the minimum specific activity of the ³H–trans-cinnamic acid precursor pool, assuming no metabolic compartmentalization, and the final specific activity of the various phenylpropanoid compounds would be equal to or lower than this value, depending on the sizes of their pools and the pools of the intermediates of the pathway during the period of labeling. Thus, assuming complete equilibration of endogenously formed ³H–trans-cinnamic acid with endogenous unlabeled internal pools and externally applied unlabeled trans-cinnamic acid, the ratio of the specific activity of a particular product to the specific activity of the total extractable trans-cinnamic acid should be 1.0 or <1.0.

Figure 3 shows the results from a series of such isotope dilution experiments with tobacco cell cultures. In the absence of externally applied unlabeled trans-cinnamic acid, the product/precursor specific activity ratio was between 0.2 and 0.99 for 4-coumaric acid, caffeic acid, and ferulic acid (Figure 3A). Small but not statistically significant increases in the product/precursor specific activity ratios were observed for all compounds after simultaneous feeding of 10^-5 M unlabeled trans-cinnamic acid and ³H-Phe. However, much larger increases (up to fivefold for ferulic acid) were obtained after feeding of 10^-4 M unlabeled trans-cinnamic acid, due to a very large decrease in the specific activity of trans-cinnamic acid, whereas the specific activity of reisolated 4-coumaric acid remained in the same range as that of the downstream products. In a similar experiment in which unlabeled 4-coumaric acid was fed simultaneously with ³H-Phe, no increase in the ratios of product/4-coumaric acid precursor specific activities were observed (Figure 3B). The same results were obtained when incorporation of the coumarin scopoletin or ³H into p-hydroxybenzoic acid (pHBA) was measured in the presence or absence of unlabeled trans-cinnamic acid or 4-coumaric acid (data not shown). These results indicate that externally added unlabeled trans-cinnamic acid does not dilute isotope incorporation from ³H-Phe into phenylpropanoid compounds in tobacco cell cultures, whereas externally added unlabeled 4-coumaric acid does, suggesting a specific metabolic compartmentalization of trans-cinnamic acid. This does not directly prove channeling or necessarily indicate the presence of more than one endogenous pool of trans-cinnamic acid in the absence of an externally fed compound.

View larger version (21K): [in this window] [in a new window]   Figure 3. In Vivo Labeling Experiments Demonstrate Metabolic Compartmentalization of trans-Cinnamic Acid in Tobacco Cell Suspension Cultures. Results are expressed as the ratio of incorporation of 3H–L-Phe (determined as specific activity) in a particular phenolic product (Iproduct) to that in reextracted trans-cinnamic acid (It-CA) or 4-coumaric acid (I4-CA). Effects of the addition of unlabeled trans-cinnamic acid and 4-coumaric acid on the ratio of Iproduct to It-CA and to I4-CA are shown. Results shown are the means ±SD of two independent experiments. Compounds were extracted 24 hr after treatment. (A) Effects of the addition of unlabeled trans-cinnamic acid on the ratio of Iproduct to It-CA in wild-type tobacco cell suspension cultures. (B) Effects of the addition of unlabeled 4-coumaric acid on the ratio of Iproduct to I4-CA in wild-type tobacco cell suspension cultures. (C) Effects of the addition of unlabeled trans-cinnamic acid on the ratio of Iproduct to It-CA in PAL-overexpressing tobacco cell suspension cultures. (D) Effects of the addition of unlabeled 4-coumaric acid on the ratio of Iproduct to I4-CA in PAL-overexpressing tobacco cell suspension cultures. Open bars, no unlabeled trans-cinnamic acid or 4-coumaric acid; hatched bars, 10-5 M trans-cinnamic acid or 4-coumaric acid; filled bars, 10-4 M trans-cinnamic acid or 4-coumaric acid. Extracts of soluble phenolics were treated with esterase and separated by HPLC. Fractions were collected and radioactivity determined by liquid scintillation counting. CafA, caffeic acid; FerA, ferulic acid; 4-CA, 4-coumaric acid.

PAL-overexpressing tobacco lines, resulting from constitutive expression of a bean PAL2 gene, overproduce phenylpropanoid compounds (Howles et al. 1996 ), whereas underexpressing lines, resulting from epigenetic gene silencing (Elkind et al. 1990 ; Bate et al. 1994 ), have reduced levels of phenylpropanoid compounds. We initiated callus and then cell suspension cultures from PAL-overexpressing transgenic tobacco; constitutive PAL activities in the cell suspension cultures were at least threefold higher than in comparable cultures derived from wild-type plants. We then repeated the above-mentioned in vivo labeling experiments to determine whether upregulation of PAL activity affects metabolic compartmentalization of trans-cinnamic acid.

Figure 2B shows the pattern of phenylpropanoid compounds and the incorporation of tritium from ³H–L-Phe into these compounds in PAL-overexpressing cultures. As shown in Figure 3C and Figure 3D, expression of the bean PAL2 transgene results in a loss of metabolic compartmentalization of trans-cinnamic acid. Thus, in contrast to the situation with wild-type cultures (Figure 3A), the addition of unlabeled trans-cinnamic acid had no effect on the ratios of product/precursor specific activities for the various phenylpropanoid compounds analyzed (Figure 3C). This indicates full equilibration of endogenously formed ³H–trans-cinnamic acid with the pool of externally added compound in PAL-overexpressing cultures. As occurs with the wild-type cultures (Figure 3B), unlabeled 4-coumaric acid had no effect on the product/precursor specific activity ratio in PAL-overexpressing cultures (Figure 3D). Again, the same results were obtained when incorporation of ³H into scopoletin or pHBA was measured in the presence or absence of unlabeled trans-cinnamic acid or 4-coumaric acid (data not shown).

In PAL-overexpressing transgenic tobacco plants, constitutive PAL activity is increased without a corresponding increase in constitutive C4H activity (Howles et al. 1999 ). In contrast, elicitation induces a coordinated increase in both activities (Howles et al. 1999 ). We repeated the above-mentioned in vivo labeling experiments with suspension cultures of wild-type tobacco that had been exposed to a yeast elicitor for 6 hr before the addition of an isotopic label. The addition of unlabeled trans-cinnamic acid had little effect on the ratios of product/precursor specific activities in the various phenylpropanoid compounds analyzed (Figure 4C), in contrast to the situation with unelicited cultures shown in Figure 4A. This indicates full equilibration of endogenously formed ³H–trans-cinnamic acid with the pool of externally added compound in elicited cultures. As occurs with the unelicited cultures (Figure 4B), unlabeled 4-coumaric acid had no effect on the product/precursor specific activity ratio in elicited cultures (Figure 4D). These results indicate that coordinated activation of the phenylpropanoid pathway as a result of elicitation results in a loss of the metabolic compartmentalization of trans-cinnamic acid observed in unelicited cultures.

View larger version (21K): [in this window] [in a new window]   Figure 4. Effects of Elicitation on Metabolic Compartmentalization of trans-Cinnamic Acid in Tobacco Cell Suspension Cultures. Results show the ratio of the Iproduct to It-CA and to I4-CA (see legend to Figure 3) in a particular phenolic compound in cells with or without elicitation. The experiment was conducted in the same way as described in the legend to Figure 3, except that 6 hr before the addition of 3H–L-Phe and unlabeled trans-cinnamic acid or 4-coumaric acid, cells were treated either with water or the yeast elicitor, and cells were incubated with the precursors for only 6 hr before being harvested. (A) Effects of the addition of unlabeled trans-cinnamic acid on the ratio of Iproduct to It-CA in unelicited tobacco cell suspension cultures. (B) Effects of the addition of unlabeled 4-coumaric acid on the ratio of Iproduct to I4-CA in unelicited tobacco cell suspension cultures. (C) Effects of the addition of unlabeled trans-cinnamic acid on the ratio of Iproduct to It-CA in elicited tobacco cell suspension cultures. (D) Effects of the addition of unlabeled 4-coumaric acid on the ratio of Iproduct to I4-CA in elicited tobacco cell suspension cultures. Open bars, no unlabeled cinnamic acids; hatched bars, 10-5 M trans-cinnamic acid or 4-coumaric acid; filled bars, 10-4 M trans-cinnamic acid or 4-coumaric acid. Results shown are the mean ±SD of two independent experiments. CafA, caffeic acid; FerA, ferulic acid; 4-CA, 4-coumaric acid.

Changes in Phenolic Metabolites as a Result of Elicitation or PAL Overexpression
To determine whether metabolic compartmentalization of trans-cinnamic acid is associated with accumulation of specific products of phenylpropanoid biosynthesis, we compared the patterns and amounts of phenolic metabolites resolved by HPLC from extracts of wild-type and PAL-overexpressing cell suspension cultures by using the cell samples analyzed in Figure 2 and Figure 3. As shown in Table 1, there was a twofold increase in trans-cinnamic acid levels in PAL-overexpressing cultures compared with wild-type cell cultures, presumably due to the higher PAL activity in these cells. PAL overexpression led to a very strong accumulation (up to 10-fold) of a compound that we tentatively identified as a vanillin derivative on the basis of its UV spectrum and chromatographic properties (_max of 204, 228, 276, and 304 nm; _min of 216 and 246 nm; 97% spectral identity to vanillin; HPLC retention time of 31.4 min, eluting at 24% acetonitrile) as well as a twofold increase in scopoletin. No significant difference in the accumulation of 4-coumaric acid, ferulic acid, pHBA, salicylic acid, or vanillic acid could be observed in the two different cell lines, although caffeic acid levels were reduced by 50% in the PAL-overexpressing cells. The addition of trans-cinnamic acid resulted in higher levels of all of the above-mentioned metabolites, except for salicylic acid, in PAL-overexpressing cells when compared with wild-type cells. This indicates that the perturbation of channeling as a result of PAL overexpression leads to increased metabolism of exogenous trans-cinnamic acid.

Surprisingly, elicitation of wild-type cultures resulted in lower levels of most of the extractable phenolics, as shown in Table 2. The addition of 10^-4 M trans-cinnamic acid led to threefold and 5.5-fold increases in 4-coumaric acid levels and 20-fold and 10-fold increases in trans-cinnamic acid levels in unelicited and elicited cells, respectively, but the total amounts of 4-coumaric acid, caffeic acid, ferulic acid, and scopoletin were ~50% lower in the elicited cells. This may be due to more rapid metabolism of these compounds and their subsequent deposition in the insoluble cell wall fraction. No significant amounts of these compounds could be detected in the culture medium or in the soluble cell wall fraction (data not shown).

Channeling of trans-Cinnamic Acid in Tobacco Stem Microsomes
Plant microsomes can convert L-Phe to 4-coumaric acid, which is not metabolized further to any significant degree (Czichi and Kindl 1975 , Czichi and Kindl 1977 ; Hrazdina and Wagner 1985 ). The value of the coupling factor, as defined by Czichi and Kindl 1977 and Kindl 1979 , is a rigorous criterion for the coupling of enzymatic reactions in vitro. The coupling factor is used to compare the ratios of tritium to carbon-14 in the product (in this case, 4-coumaric acid) with the ratios of tritium to carbon-14 in the intermediate (in this case, trans-cinnamic acid) in dual labeling experiments. It thereby gives an estimation of the level of coupling between consecutive enzymes. The use of the ³H-labeled primary substrate (L-Phe) and ¹⁴C-labeled secondary substrate (trans-cinnamic acid) by enzymes located on isolated microsomes will result in the formation of labeled 4-coumaric acid with a specific tritium/carbon-14 ratio. If the value of the ratio in the product is higher than the respective value in the reisolated intermediate (i.e., a coupling factor >1.0), the bound or channeled form of the intermediate does not freely exchange with the external pool of the intermediate, indicating channeling between L-Phe and 4-coumaric acid.

We initially chose tobacco stem tissue for in vitro channeling assays because of the high levels of PAL and C4H activities in this tissue associated with lignification (Sewalt et al. 1997 ). We first confirmed the presence of PAL activity in microsomes isolated from tobacco stem tissue. The total activity of PAL in washed microsomes isolated from wild-type tobacco plants amounted to 5 to 10% of the total activity of PAL in the soluble enzyme fraction. As shown in Figure 5A, the specific activities of microsomal PAL were between 30 and 40% of the specific activities of soluble PAL. Microsomal PAL is relatively tightly associated with the microsomes and is not simply cytoplasmic contamination, as shown below.

View larger version (20K): [in this window] [in a new window]   Figure 5. Specific Activities of PAL in the Soluble and Microsomal Protein Fractions and Metabolic Channeling between PAL and C4H in Microsomes from Stems of Wild-Type and Transgenic Tobacco Plants. Plants were wild-type (WT), PAL overexpressers (OX434), or gene-silenced PAL underexpressers (C23). (A) Open bars, soluble (105,000g supernatant) PAL activity; filled bars, microsomal (105,000g pellet) PAL activity. Total activities of microsomal PAL were 5 to 10% of total soluble PAL activity. pkat, picokatals. (B) Washed and resuspended microsomes (250 to 500 µg of protein per assay) were incubated simultaneously with 500 nmol 3H–L-Phe (5 nCi/nmol) and 30 nmol 14C–trans-cinnamic acid (2 nCi/nmol). The coupling factor is defined as the tritium/carbon-14 ratio in the product, 4-coumaric acid, divided by the tritium/carbon-14 ratio in the reisolated intermediate, trans-cinnamic acid. Open bars, coupling factors in microsomes alone; filled bars, coupling factors in microsomes with 100 µL of the soluble PAL fraction added. Data shown are the means ±SD of six (wild-type) or three (OX434 and C23) independent preparations.

Microsomal fractions were incubated simultaneously with ³H–L-Phe and ¹⁴C–trans-cinnamic acid, and the ratios of tritium to carbon-14 in reisolated trans-cinnamic acid and in 4-coumaric acid were compared. Coupling factors were between 5 and 11 in microsomes from wild-type plants, suggesting significant channeling between PAL and C4H (Figure 5B). When soluble PAL from the cytoplasmic supernatant was added to the channeling assays, the ratio of tritium to carbon-14 in 4-coumaric acid remained almost the same, but the tritium/carbon-14 ratio in the reisolated trans-cinnamic acid intermediate was strongly increased, resulting in reduced coupling factors, as shown in Figure 5B. This is presumably due to an excess of ³H–trans-cinnamic acid formed from ³H–L-Phe, which is not converted to ³H–4-coumaric acid because of the preference of the channeled system for ³H–trans-cinnamic acid originating via the microsomal PAL reaction.

We next studied the effects of transgenic modification of PAL activity on channeling in tobacco stem microsomes. PAL-overexpressing plants (OX434) had a two- to threefold higher specific activity of PAL, and PAL sense-suppressed plants (C23) a four- to sixfold lower specific activity of PAL in both soluble and microsomal fractions, when compared with wild-type plants (Figure 5A). Total microsomal activities were five to 10% of that detected in the soluble fraction, as was the case for wild-type tobacco plants. In microsomes from PAL-suppressed plants, coupling factors were in the same range as previously demonstrated for wild-type plants (Figure 5B). In contrast, the coupling factors for microsomes from PAL-overexpressing plants were reduced significantly to values of 1 to 2. These values are similar to those obtained on adding soluble PAL to wild-type microsomes. These results indicate that either higher PAL activity associated with the microsomes from PAL-overexpressing plants or the presence of a heterologous PAL species in the microsomes (see below) leads to a reduction in the extent of coupling between microsomal PAL and C4H.

Channeling between PAL and C4H in Microsomes from Tobacco Cell Suspension Cultures
If the changes in apparent metabolic compartmentalization of trans-cinnamic acid observed in in vivo labeling experiments indeed reflect changes in channeling rather than some in vivo labeling artifact, channeling should be demonstrable in microsomes from unelicited tobacco cell cultures but not in microsomes from elicited or PAL-overexpressing cultures. Such correlations between in vivo and in vitro labeling results would strengthen the validity and physiological relevance of the in vitro channeling assays.

The specific activity of soluble PAL in unelicited cell suspension cultures overexpressing bean PAL2 was approximately sixfold higher than in wild-type cell suspension cultures, as shown by comparing time zero values in Figure 6A and Figure 6C. The coupling factor between PAL and C4H in microsomes from unelicited wild-type cells (0 hr after elicitation) was ~12 (Figure 6B) and is similar to that observed in microsomes from stem tissue (Figure 5B). In agreement with the previous results obtained using stem tissue, microsomes from cell cultures overexpressing bean PAL had much reduced coupling factors, with a coupling factor below 1.0 and therefore indicative of no channeling (Figure 6D).

View larger version (21K): [in this window] [in a new window]   Figure 6. Specific Activities of PAL in the Soluble and Microsomal Protein Fractions, and Metabolic Channeling between PAL and C4H in Microsomes of Elicited Wild-Type and PAL-Overexpressing Tobacco Cell Suspension Cultures. Isolation of microsomal proteins and determination of coupling factors were conducted as described in the legend to Figure 5. Wild-type and PAL-overexpressing cultures were treated with the yeast elicitor or an equal volume of water (unelicited) and then harvested at the times shown. pkat, picokatals. (A) Soluble PAL activity in the unelicited (open circles) and elicited (filled circles) wild-type line. (B) Microsomal PAL activity (filled circles) and coupling factors (open bars) in the elicited wild-type line. (C) Soluble PAL activity in the unelicited (open circles) and elicited (filled circles) PAL-overexpressing lines. (D) Microsomal PAL activity (filled circles) and coupling factors (open bars) in the elicited PAL-overexpressing lines. Data shown are the means ±SD of four (wild-type) or three (PAL-overexpressing) independent experiments.

We next determined the effects of elicitation on the in vitro microsomal coupling of PAL and C4H in the wild-type and PAL-overexpressing cell suspension cultures. Treatment with crude yeast elicitor led to an approximately three- to fivefold increase in soluble PAL activity in wild-type cultures by 12 hr after elicitation (Figure 6A) but to a much smaller fold increase in the PAL-overexpressing cultures (Figure 6C). Microsomal PAL activity in elicited wild-type cells reached its maximum, an approximately fivefold increase, 9 hr after elicitation, as shown in Figure 6B. However, the coupling factors in microsomes from elicited wild-type cells decreased significantly by 3 hr after elicitation, to reach minimum values of ~1.4 at 12 hr after elicitation. These results demonstrate that the elicitation of PAL activity results in the perturbation of coupling between PAL and C4H. Elicitation did not strongly enhance microsomal PAL activity in PAL-overexpressing tobacco cells, and it had no effect on the coupling factors in microsomes from these cells, with the values being in the same range as in unelicited cells—near or below 1.0 (Figure 6D). Therefore, these data indicate a correlation between metabolic compartmentalization of trans-cinnamic acid in vivo (Figure 3 and Figure 4) and coupling of the PAL and C4H reactions on microsomes in vitro.

The Nature of PAL Associated with Microsomal Membranes
The above-mentioned results suggest that PAL, or one or more specific forms of PAL, may be closely associated with C4H as an enzyme complex on microsomal membranes. Cytochrome P450s, such as C4H, are associated with the endoplasmic reticulum by way of a membrane anchor region at the N terminus (Chapple 1998 ). The catalytic region of cytochrome P450 is in the cytoplasm. To obtain more information on the association of PAL with microsomal membranes, we fractionated tobacco stem homogenates by ultracentrifugation into soluble and microsomal fractions. Both fractions were assayed for PAL and C4H activity with or without a 20-min pretreatment with trypsin. The results in Table 3 indicate that as expected, no PAL activity could be detected in the soluble fraction after trypsin treatment. In contrast, 19% of the PAL activity in the microsomal fraction was retained. All microsomal C4H activity was destroyed by trypsin treatment. Because microsomal PAL activity is partially protected from the hydrolytic action of trypsin, a small fraction probably is located in the lumen of the endoplasmic reticulum or, alternatively, inside membrane vesicles formed during the preparation of microsomes. In this respect, the addition of the detergent Triton X-100 to the microsomal preparation before trypsin treatment resulted in an almost complete loss of microsomal PAL activity. The twofold increase in soluble PAL activity after the addition of detergent may indicate activation of the enzyme.

Tobacco PAL is encoded by two gene families, each of which contains two very closely related members. The two families are represented by single-copy PAL genes in the two progenitor species, Nicotiana sylvestris and N. tomentosiformis (Fukasawa-Akada et al. 1996 ). In this study, PAL1 refers to the product encoded by the PAL gene of family I, described by Fukasawa-Akada et al. 1996 , and PAL2 refers to the product of the PAL gene from family II, as reported by Nagai et al. 1994 . To determine whether specific forms of PAL are associated with tobacco microsomes, we raised antibodies against synthetic peptide sequences specific for tobacco PAL1, tobacco PAL2, and bean PAL2 (Howles et al. 1996 ). Soluble and microsomal proteins were isolated from wild-type, PAL-overexpressing (from the bean PAL2 transgene), and PAL-suppressed tobacco stem tissues, subjected to SDS-PAGE, and probed with these antibodies. As shown in Figure 7A, antibodies generated against tobacco PAL1 cross-reacted with a protein band of just under 86 kD (the size of the native PAL subunit), which was present in both the soluble and microsomal protein fractions. By contrast, antibodies specific for tobacco PAL2 cross-reacted with a more diffuse protein band of similar size found only in the soluble protein fraction (Figure 7B). Note that the levels of tobacco PAL proteins were similar in wild-type, PAL-overexpressing, and PAL-suppressed plants, despite the different levels of PAL activity in these plants; this may be due to post-translational regulation.

View larger version (56K): [in this window] [in a new window]   Figure 7. Protein Gel Blot Analysis of PAL Protein Levels in the Soluble and Microsomal Fractions from Wild-Type and Transgenic Tobacco Plants. WT8 is a wild-type plant; OX434 is a PAL-overexpressing line; C23 is a PAL-suppressed line. Soluble PAL (sP) represents the PAL protein in the 130,000g supernatant, and microsomal PAL (mP) the PAL protein in the 130,000g pellet. Proteins (10 µg per lane) were separated by SDS-PAGE. (A) Tobacco PAL1 protein detected using the anti-tobacco PAL1 antiserum. (B) Tobacco PAL2 protein detected using the anti-tobacco PAL2 antiserum. (C) Bean PAL2 protein detected using the anti-bean PAL2 antiserum. Positions of the molecular mass markers are indicated at left in kilodaltons. The 80-kD PAL subunits are marked with arrows at right.

Probing the protein gel blots with antibodies specific for bean PAL2 protein (Figure 7C) revealed that a very high amount of bean PAL2 protein is present in the soluble protein fraction, but a substantial amount of the heterologous gene product also is located in the microsomes of PAL-overexpressing plants. The ~50-kD band in the microsomal fraction cross-reacting with all three antibodies is most probably the result of nonspecific binding. Essentially identical results were obtained by protein gel blot analysis of soluble and microsomal fractions from cell suspension cultures of wild-type and PAL-overexpressing tobacco (data not shown). Furthermore, we could not detect any significant change in the subcellular localization of the various PAL forms after elicitation of the cell cultures (data not shown).

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

Relation of Metabolic Compartmentalization of trans-Cinnamic Acid in Vivo to Metabolic Channeling as Revealed by Coupling Assays in Vitro
As discussed by Srere 1987 , objections have been raised to the idea that isotopic data can unequivocally indicate more than one pool of a particular intermediate. The major argument is that multiple pools arise due to heterogeneity of the cell population. De-differentiated tobacco cell suspension cultures are highly homogeneous with respect to cell type but somewhat heterogeneous with respect to cell size and degree of aggregation of small cell clusters. We cannot rule out totally the presence of different cell types containing different pools of intermediates that are differentially accessible to an endogenously formed or externally applied intermediate. However, this is unlikely because the isotopic labeling results are reproducible between independent cell culture batches, the isotope dilution experiments reveal at least two pools of trans-cinnamic acid but only a single pool of 4-coumaric acid, and more importantly, the effects on compartmentalization of transgenic or elicitor-mediated perturbation of the pathway can be reproduced by using in vitro channeling assays. The difference in isotope dilution behavior of trans-cinnamic acid compared with 4-coumaric acid potentially could result from uptake of trans-cinnamic acid into an inaccessible compartment, such as the cell wall, whereas all of the 4-coumaric acid may be taken into the cytoplasm. This appears not to be the case because if trans-cinnamic acid were taken up in this way, isotope dilution experiments with PAL-overexpressing transgenic cell cultures would artifactually reveal metabolic compartmentalization, which these do not.

This study provides a significant advance over previous work in this area by showing that the alteration of flux into the pathway by transgenic upregulation of the first enzyme, or coinduction of all the enzymes of the putative channel, leads to a loss of channeling, as revealed by both isotope dilution experiments in vivo and microsomal channeling assays in vitro. Thus, the in vitro assays appear to reflect the in vivo situation. This is an important observation in view of the often-cited criticism that in vitro enzyme complexes may not function at the presumed cellular pH and concentration conditions of the cell (Srere 1987 ).

Note that the variation we observed in the product/precursor specific activity ratios determined in vivo is quite small, considering the potential biological variation between different cell culture batches, and the values for variation in the in vitro coupling factors are similar. In the latter case, the variation is probably a composite of biological variability and variation caused by sensitivity to disruption of the microsomal association between PAL and C4H. In this respect, it previously has been suggested that reversible inhibition of protein and nucleotide biosynthesis in gently sonicated yeast cells results from disruption of loosely associated enzyme complexes (Burns 1964 ).

Previous studies with cucumber hypocotyls showed that hormonal or environmental stimuli could affect channeling between PAL and C4H in vitro. Thus, a coupling factor of ~5.0 was reduced to ~1.5 after exposure of green hypocotyls to ethylene, whereas it was increased to ~25 after irradiation with high-intensity UV light (Czichi and Kindl 1977 ). Paradoxically, the latter treatment was reported to induce soluble PAL activity but not to induce microsomal PAL (Czichi and Kindl 1977 ). No in vivo labeling data are reported in this study. In contrast, elicitation of tobacco cell cultures results in coordinated induction of both PAL (soluble and microsomal) and C4H activities (Howles et al. 1999 ) but in a loss of channeling between PAL and C4H, as assessed by both in vivo isotope dilution and in vitro microsomal channeling assays. Note that elicitation induces a different set of defense responses from those induced by UV irradiation (Hahlbrock and Scheel 1989 ). The loss of channeling in PAL-overexpressing plants and cell cultures could result from two possible mechanisms. Quantitatively, the overexpression of PAL without a corresponding increase in C4H could lead to a spillover of trans-cinnamic acid. Alternatively, the qualitative difference in the PAL proteins in the PAL-overexpressing suspension cell cultures may reduce channeling if the bean PAL2 protein cannot correctly couple with tobacco C4H.

Association of PAL with Microsomal Membranes
Metabolic channeling may have more than one cellular function. In its simplest form, it might provide for rapid turnover of low concentrations of labile intermediates that have no other metabolic functions (Srere 1987 ). However, the phenylpropanoid pathway presents a more complex case. PAL is encoded by small multigene families in all of the plants studied to date, including Arabidopsis (Wanner et al. 1995 ), and the pathway has several downstream branches leading to functionally distinct end products. It has long been proposed that different forms of PAL may be involved in the synthesis of different end products, primarily on the basis of differential product inhibition of different PAL forms (Alibert et al. 1972 ; Jones 1984 ). Localization of one or more specific PAL isoforms on the surface of the endoplasmic reticulum would provide the necessary structural basis for assembling complexes in which different PAL forms could channel metabolites into different pathways of phenylpropanoid metabolism.

Previous studies have used immunolocalization and biochemical fractionation techniques to show the association of PAL with endoplasmic reticulum membranes (Czichi and Kindl 1975 , Czichi and Kindl 1977 ; Wagner and Hrazdina 1984 ; Hrazdina and Wagner 1985 ). Our results now demonstrate microsomal association of specific forms of PAL. Thus, tobacco PAL1 is found in both soluble and microsomal fractions, whereas tobacco PAL2 is not found in microsomes. PAL activity measured in microsomal preparations from tobacco stem tissues therefore reflects the localization of specific PAL forms on microsomal membranes rather than an artificial entrapment of PAL proteins into microsomal vesicles formed during the isolation process. This suggests that the molecular basis for the channeling of trans-cinnamic acid is the coupling of specific PAL forms with C4H located together on the microsomal membranes.

Our data do not suggest a mechanism for the apparent loss of channeling after elicitation. Protein gel blot analysis did not reveal any major difference in the localization of PAL forms after elicitation. The mechanism could therefore be subtle, perhaps involving specific post-translational modifications to PAL or C4H, or might simply reflect a change in the in vivo PAL/C4H activity ratio, with a resultant spillover of trans-cinnamic acid.

Implications of Metabolic Channeling for Phenylpropanoid Pathway Regulation
The addition of trans-cinnamic acid to bean cell cultures inhibits PAL at the transcriptional level and induces the synthesis of a proteinaceous inactivator of PAL (Bolwell et al. 1986 , Bolwell et al. 1988 ). Furthermore, downregulation of C4H by antisense gene expression in transgenic tobacco leads to a corresponding decrease in PAL activity, suggesting that trans-cinnamic acid is sensed as a metabolic regulator of phenylpropanoid pathway flux in vivo (J. Blount and R.A. Dixon, unpublished results). Tobacco PAL is particularly sensitive to direct inhibition by trans-cinnamic acid in vitro (OaNeal and Keller 1970 ). Tight coupling between PAL and C4H therefore could maintain, in the microsomal "compartment," a low trans-cinnamic acid pool that would avoid feedback inactivation of PAL, as suggested by Noe et al. 1980 . Note, however, that the addition of 10^-4 M trans-cinnamic acid in our isotope dilution experiments did not appear to inhibit flux through the PAL reaction in vivo, as assessed by incorporation of tritium from ³H–L-Phe into the various phenylpropanoid compounds (data not shown).

PAL-overexpressing tobacco cells, in which channeling through the C4H reaction is no longer measurable, accumulate twofold higher levels of the potentially antifungal coumarin glycoside scopolin and the corresponding aglycone scopoletin (Ahl Goy et al. 1993 ; Gutierrez et al. 1995 ) than do wild-type cells. It is not possible to conclude that this results from a release from channeling of trans-cinnamic acid because of uncertainty as to the biosynthetic origin of scopoletin. The levels of scopoletin in tobacco cell cultures treated with a yeast elicitor are much lower than those in untreated cells, which could be due to degradation by induced peroxidases in elicited cells, as has been shown in other systems (Gutierrez et al. 1995 ; Breton et al. 1997 ; Edwards et al. 1997 ). A similar decline of constitutive phenolics associated with cell wall incorporation has been described for isoflavonoids in elicitor-treated cell suspension cultures of Pueraria lobata (Park et al. 1995 ), and we assume that the decline in the levels of most hydroxycinnamic acid derivatives in elicited tobacco cells is associated with such further metabolism rather than being a consequence of changes in metabolic channeling. In contrast, the benzoic acid derivatives pHBA, salicylic acid, vanillic acid, and the partially characterized vanillin derivative accumulated to the same level in elicited and unelicited tobacco cells. There appears to be independent regulation of the metabolic pathways involved in the biosynthesis of hydroxycinnamic acid derivatives, which include lignin precursors, and the biosynthesis of benzoic acid derivatives, which include salicylic acid (Figure 1).

Although the majority of phenylpropanoid compounds derived from the natural products of plants require both PAL and C4H for their synthesis, salicylic acid probably is derived directly from trans-cinnamic acid by chain shortening and ring hydroxylation (Lee et al. 1995 ). It is tempting to speculate that salicylic acid is synthesized via an uncoupled form of PAL, for example, nonmicrosomally associated PAL2 in tobacco. However, tobacco PAL1 could be equally involved in salicylic acid biosynthesis, because overexpression of bean PAL, which, like tobacco PAL1, is localized both cytoplasmically and microsomally in transgenic tobacco stem tissue, results in increased salicylic acid production and corresponding increases in disease resistance in intact tobacco plants (Felton et al. 1999 ).

Our results obtained in experiments with PAL-overexpressing tobacco cell cultures and elicited wild-type cultures do not support the idea that a loss of channeling leads directly to a higher accumulation of salicylic acid produced from unchanneled trans-cinnamic acid, because the levels of salicylic acid are the same as those in wild-type tobacco cells. However, a marked difference can be seen in the accumulation of the vanillin derivative, which is 10 times higher in PAL-overexpressing tobacco cells when compared with wild-type cells, although it is not significantly induced by elicitation. Currently, we cannot conclude whether the accumulation of this compound in PAL-overexpressing cells is due to a perturbation in trans-cinnamic acid channeling, because the biosynthetic pathway(s) involved in the formation of benzoic acid derivatives, such as vanillic acid and vanillin, still is to be unequivocally elucidated (Zenk 1965 ; Funk and Brodelius 1990 ; Yazaki et al. 1991 ; Schnitzler et al. 1992 ).

Two important predictions concerning phenylpropanoid pathway organization now can be tested in the model system of tobacco. The first is that the differential subcellular localization of tobacco PAL1 and PAL2 has functional consequences related to metabolic channeling. This prediction can be addressed by studying the metabolic consequences of differentially downregulating expression of PAL1 or PAL2. The second is that tobacco PAL1, but not PAL2, will be in close physical association with C4H. This prediction can be addressed by immunolocalization studies by using transgenic plants expressing epitope-tagged PAL and C4H species. These experiments are currently in progress.

	METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

Plant Material
Tobacco (Nicotiana tabacum cv Xanthi-nc) plants were either wild type or transformed with the bean phenylalanine (Phe) ammonia–lyase PAL2 gene (Elkind et al. 1990 ). The transformed plants displayed either increased PAL activity (Howles et al. 1996 ) or epigenetic gene silencing with reduced levels of activity. Used as controls were untransformed wild-type plants or plants from which the PAL2 transgene had segregated. All plants were grown under greenhouse conditions (18°C at night and 27°C by day) and harvested just before flowering. Stem samples were taken from internodes seven to 11, counting from the top, frozen in liquid N₂, and ground in a tissue grinder.

Callus cultures were initiated from leaf discs of wild-type and transformed tobacco plants, as described previously (Bate et al. 1994 ). Liquid cultures were initiated and maintained in a modified Schenk and Hildebrandt medium, as described previously (Dixon et al. 1981 ), and subcultured every 7 to 10 days.

Five days after subculturing, dark-grown tobacco cell suspension cultures (75-mL batches) were treated with a yeast elicitor (Schumacher et al. 1987 ; 75 µg mL^-1 glucose equivalents) and harvested at various times after elicitation. Control cells were treated with the same amount of distilled water.

Chemicals
³H–L-2,3,4,5,6-Phe (124 Ci/mmol) and uniformly labeled L–¹⁴C-Phe (453 mCi/mmol) were supplied by Amersham (Little Chalfont, United Kingdom). ¹⁴C–trans-Cinnamic acid was synthesized enzymatically from uniformly labeled L–¹⁴C-Phe by using PAL from Rhodotorula glutinis (14 units per mg of protein; Sigma) as described elsewhere (Edwards and Kessmann 1992 ).

Precursor Dilution Experiments
Tobacco cell suspension cultures (75-mL batches) were incubated 4 days after subculturing with 7.5 µmol of ³H–L-Phe (1 µCi/µmol) with or without unlabeled trans-cinnamic acid or 4-coumaric acid (0.75 or 7.5 µmol). After 24 hr, the cells were filtered through a nylon mesh and ground in liquid N₂. The soluble phenolics were extracted three times with 8 mL of ice-cold acetone at 4°C in the dark. The extracts were combined and concentrated to dryness under a stream of N₂, and the residue was dissolved in methanol (500 µL per gram fresh weight). Aliquots (20 µL) were separated by HPLC, as described below. Fractions of 500 µL were collected and counted in a liquid scintillation counter.

Enzymatic Hydrolysis of Phenolic Extracts
For the enzymatic hydrolysis of phenolic esters of caffeic acid, 4-coumaric acid, ferulic acid, and trans-cinnamic acid, 200 µL of the extracts was concentrated, dissolved in buffer (200 mM Tris-HCl, pH 8.0), and incubated overnight at 37°C with an esterase from rabbit liver (190 units; EC 3.1.1.1; Sigma). Phenolic glucosides of scopoletin, the vanillin derivative, p-hydroxybenzaldehyde (pHBA), and salicylic acid were hydrolyzed with almond ß-glucosidase (100 units; EC 3.2.1.21; Sigma). Extracts then were processed for analysis of phenolic aglycones as described above.

Separation of Phenolics by Reverse Phase HPLC
Organic extracts from enzyme assays and plant or cell suspension phenolic fractions were applied to an ODS reverse phase HPLC column (5-mm particle size, 4.6 x 250 mm; Metachem Technologies, Inc., Torrance, CA) and eluted in 1% phosphoric acid with an increasing acetonitrile concentration gradient (0 to 5 min, 5% [v/v] acetonitrile; 5 to 10 min, 5 to 10% acetonitrile; 10 to 25 min, 10 to 17% acetonitrile; 25 to 30 min, 17 to 23% acetonitrile; 30 to 65 min, 23 to 50% acetonitrile; 65 to 74 min,100% acetonitrile; and 74 to 85 min, 5% acetonitrile) at a constant flow rate of 1 mL min^-1. UV absorbance was monitored with a photodiode array detector (Hewlett Packard , Waldbronn, Germany). Quantification of phenolics was based on calibration curves achieved with authentic standards (Sigma) at 270 and 330 nm.

Preparation of Membrane Fractions
Frozen (-70°C) and ground stem material or suspension cells (4 to 4.5 g fresh weight) were homogenized for 3 x 10 sec in 8 mL of a Tris-HCl buffer (200 mM Tris, pH 8.0, 400 mM sucrose, 1 mM EDTA, 40 mM sodium ascorbate, and 5 mM 2-mercaptoethanol) by using an Ultraturrax blender (Brinkmann Instruments, Inc., Westbury, NY). The homogenate was centrifuged (10,000g for 30 min) and filtered through a syringe filled with glass wool. The filtrate was ultracentrifuged (130,000g for 1 hr), the supernatant was decanted, and the pellet was blot dried. After resuspending the pellet in 2.5 mL of Pi buffer (200 mM potassium phosphate, pH 8.0, and 3 mM 2-mercaptoethanol) with a rubber spatula, the suspension was subjected to a second ultracentrifugation (130,000g for 1 hr). The supernatant was decanted, and the microsomal pellet was blot dried. The microsomes were resuspended carefully in 300 µL of assay buffer (200 mM potassium phosphate, pH 8.0, 6 mM MgCl₂, and 3 mM 2-mercaptoethanol). In all cases, 2-mercaptoethanol was added fresh to the buffers, and all steps were conducted on ice or at 4°C.

PAL and Cinnamic Acid 4-Hydroxylase Assays
Soluble PAL activity was determined in the ultracentrifugation supernatants (desalted on a PD-10 column equilibrated with 200 mM boric acid, pH 8.8, and 13 mM 2-mercaptoethanol) by using ¹⁴C–L-Phe as a substrate, essentially as described by Legrand et al. 1976 . For determination of PAL activity in the microsomal fraction, 25 µL of the microsomal suspension was diluted with 25 µL of boric acid (200 mM, pH 8.8, containing 13 mM 2-mercaptoethanol) before adding the substrate. Cinnamic acid 4-hydroxylase (C4H) activity was determined in the microsomal fraction according to Edwards and Kessmann 1992 . Protein concentrations were determined according to Bradford 1976 , using BSA as a standard.

In Vitro Channeling Assays
Washed microsomes (200 µL) were preincubated with 245 µL of C4H assay buffer for 5 min, and the reactions were started with 500 nmol of ³H–L-Phe (5 mCi/mmol), 30 nmol of ¹⁴C–trans-cinnamic acid (2 mCi/mmol), and 1 mmol of NADPH in a total volume of 155 µL at 30°C. The reactions were stopped after 10 min with 50 µL of 6 N HCl.

Channeling assays in the presence of soluble PAL were conducted with 200 µL of the microsomal suspension, 100 µL of the soluble enzyme fraction, 145 µL of the C4H assay buffer, and the substrates, as described above. The assays were extracted with 3 x 700 µL of ethyl acetate, and the extracts were concentrated to dryness and dissolved in 30 µL methanol. Caffeic acid (1 µg) was added to the assays as an internal standard before extraction. Products were separated by reverse phase HPLC and monitored by UV absorbance at 270 nm, and the fractions (250 µL) were collected. The radioactivity in fractions containing 4-coumaric acid and trans-cinnamic acid was determined by liquid scintillation counting, by using the automatic quench compensation for tritium/carbon-14 dual label counting on an LS 1701 scintillation counter (Beckmann Instruments, Fullerton, CA).

Treatment of Plant Extracts with Trypsin
Soluble and microsomal enzyme fractions were prepared as described above, incubated with Triton X-100 (Sigma) (final concentration of 0.1 or 0.6%) or homogenization buffer (controls) for 30 min at 4°C, and then treated with 1 mg mL^-1 trypsin (10,200 units per mg of protein; Sigma) for 20 min at 30°C. PAL and C4H enzyme assays were started by the addition of substrates, incubated for 20, 40, and 60 min, and stopped with HCl as described above.

Generation of Antibodies Specific for Tobacco and Bean PAL Forms
For raising antibodies specific to tobacco PAL proteins, the following amino acid sequences, derived from the available tobacco PAL sequences (Nagai et al. 1994 ; Pellegrini et al. 1994 ; Fukasawa-Akada et al. 1996 ), were used: VRDKSANG (positions 69 to 76 from tobacco PAL1) and VAQNGHQEMDFCVKV (positions 4 to 18 from tobacco PAL2). These sequences represent the few stretches that are unique between the different tobacco PAL forms. Synthetic peptides were coupled to the carrier protein keyhole limpet hemocyanin, and antibodies were raised in rabbits (Genosys Biotechnologies, Inc., The Woodlands, TX). Antibodies specific for bean PAL2 have been described previously (Howles et al. 1996 ).

Protein Gel Blot Analysis
Soluble and microsomal proteins (10 µg) were subjected to denaturing SDS-PAGE on precast 8 to 16% Tris-glycine gels (Novex, San Diego, CA) and transferred to nitrocellulose membranes (Trans-Blot; Bio-Rad). The membranes were blocked and probed with the primary antibody in 5% fat-free milk powder (Carnation, Glendale, CA) dissolved in 0.2% Tween 20 (Sigma) in Tris-buffered saline (TBST; Ausubel et al. 1994 ). A goat anti-rabbit IgG–horseradish peroxidase conjugate (Bio-Rad) was used as secondary antibody at a 1:10,000 dilution in TBST. Bands were visualized on a film by using a chemiluminescence assay (ECL; Amersham). Kaleidoscope prestained standards (Bio-Rad) were used as molecular weight markers.

	ACKNOWLEDGMENTS

We thank Drs. Mitsuo Okazaki and John C. Watson for providing tobacco PAL1 and PAL2 cDNA clones, Drs. Kentaro Inoue and Nancy Paiva for critical reading of the manuscript, and Cuc Ly for artwork. This work was supported by the Samuel Roberts Noble Foundation.

Received March 9, 1999; accepted May 11, 1999.

	REFERENCES

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

Ahl Goy, P., Signer, H., Reist, R., Aichholz, R., Blum, W., Schmidt, E., and Kessmann, H. (1993) Accumulation of scopoletin is associated with the high disease resistance of the hybrid Nicotiana glutinosa x Nicotiana debneyi.. Planta 191:200-206.

Alibert, G., Ranjeva, R., and Boudet, A. (1972) Recherches sur les enzymes catalysant la formation des acides phenoliques chez Quercus pedunculata (Ehrh.). II. Localisation intracellulaire de la phenylalanine ammonia-lyase, de la cinnamate 4-hydroxylase, et de la "benzoate synthase.". Biochim. Biophys. Acta 279:282-288[Medline].

Ausubel, F.M., Brent, R., Kingston, R.I.E., Moore, D.D., Seidman, J.G., Smith, J.A., and Struhl, K. (1994) Current Protocols in Molecular Biology. New York, John Wiley and Sons.

Batard, Y., Schalk, M., Pierrel, M.-A., Zimmerlin, A., Durst, F., and Werck-Reichhart, D. (1997) Regulation of the cinnamate 4-hydroxylase (CYP73A1) in Jerusalem artichoke tubers in response to wounding and chemical treatments. Plant Physiol. 111:951-959.

Bate, N.J., Orr, J., Ni, W., Meromi, A., Nadler-Hassar, T., Doerner, P.W., Dixon, R.A., Lamb, C.J., and Elkind, Y. (1994) Quantitative relationship between phenylalanine ammonia-lyase levels and phenylpropanoid accumulation in transgenic tobacco identifies a rate-determining step in natural product synthesis. Proc. Natl. Acad. Sci. USA 91:7608-7612[Abstract/Free Full Text].

Bell-Lelong, D.A., Cusumano, J.C., Meyer, K., and Chapple, C. (1997) Cinnamate 4-hydroxylase expression in Arabidopsis.. Plant Physiol. 113:729-738[Abstract].

Bolwell, G.P., Cramer, C.L., Lamb, C.J., Schuch, W., and Dixon, R.A. (1986) L-Phenylalanine ammonia-lyase from Phaseolus vulgaris. Modulation of the levels of active enzyme by trans-cinnamic acid. Planta 169:97-107.

Bolwell, G.P., Mavandad, M., Millar, D.J., Edwards, K.J., Schuch, W., and Dixon, R.A. (1988) Inhibition of mRNA levels and activities by trans-cinnamic acid in elicitor-induced bean cells. Phytochemistry 27:2109-2117[CrossRef].

Bradford, M.M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein–dye binding. Anal. Biochem. 72:248-254[CrossRef][ISI][Medline].

Breton, F., Sanier, C., and D'Auzac, J. (1997) Scopoletin production and degradation in relation to resistance of Hevea brasiliensis to Corynespora cassiicola. J. Plant Physiol. 151:595-602.

Buell, C.R., and Somerville, S.C. (1995) Expression of defense-related and putative signaling genes during tolerant and susceptible interactions of Arabidopsis with Xanthomonas campestris pv. campestris. Mol. Plant-Microbe Interact. 8:435-443.

Burns, V.W. (1964) Reversible sonic inhibition of protein, purine, and pyrimidine biosynthesis in the living cell. Science 146:1056-1058[Abstract/Free Full Text].

Chapple, C. (1998) Molecular-genetic analysis of plant cytochrome P450-dependent monooxygenases. Annu. Rev. Plant Physiol. Plant Mol. Biol. 49:311-343[CrossRef][ISI].

Cramer, C.L., Bell, J.N., Ryder, T.B., Bailey, J.A., Schuch, W., Bolwell, G.P., Robbins, M.P., Dixon, R.A., and Lamb, C.J. (1985) Coordinated synthesis of phytoalexin biosynthetic enzymes in biologically-stressed cells of bean (Phaseolus vulgaris L.). EMBO J. 4:285-289[ISI][Medline].

Cramer, C.L., Edwards, K., Dron, M., Liang, X., Dildine, S.L., Bolwell, G.P., Dixon, R.A., Lamb, C.J., and Schuch, W. (1989) Phenylalanine ammonia-lyase gene organization and structure. Plant Mol. Biol. 12:367-383.

Cutler, A.J., and Conn, E.E. (1981) The biosynthesis of cyanogenic glucosides in Linum usitatissimum (linen flax) in vitro. Arch. Biochem. Biophys. 212:468-474[Medline].

Cutler, A.J., Hösel, W., Sternberg, M., and Conn, E.E. (1981) The in vitro biosynthesis of taxiphyllin and the channeling of intermediates in Triglochin maritima.. J. Biol. Chem. 256:4253-4258[Abstract/Free Full Text].

Czichi, U., and Kindl, H. (1975) Formation of p-coumaric acid and o-coumaric acid from L-phenylalanine by microsomal membrane fractions from potato: Evidence of membrane-bound enzyme complexes. Planta 125:115-125.

Czichi, U., and Kindl, H. (1977) Phenylalanine ammonia-lyase and cinnamic acid hydroxylases as assembled consecutive enzymes on microsomal membranes of cucumber cotyledons: Cooperation and subcellular distribution. Planta 134:133-143.

Dixon, R.A., and Paiva, N.L. (1995) Stress-induced phenylpropanoid metabolism. Plant Cell 7:1085-1097[CrossRef][ISI][Medline].

Dixon, R.A., Dey, P.M., Murphy, D.L., and Whitehead, I.M. (1981) Dose responses for Colletotrichum lindemuthianum elicitor-mediated enzyme induction in French bean cell suspension cultures. Planta 151:272-280.

Edwards, K., Cramer, C.L., Bolwell, G.P., Dixon, R.A., Schuch, W., and Lamb, C.J. (1985) Rapid transient induction of phenylalanine ammonia-lyase mRNA in elicitor-treated bean cells. Proc. Natl. Acad. Sci. USA 82:6731-6735[Abstract/Free Full Text].

Edwards, R., and Kessmann, H. (1992) Isoflavonoid phytoalexins and their biosynthetic enzymes. In Gurr S.J., McPherson M.J., Bowles D.J., eds. Molecular Plant Pathology—A Practical Approach. Oxford, UK, IRL Press. 45–62.pp.

Edwards, R., Stones, S.M., Gutierrez-Mellado, M.-C., and Jorrin, J. (1997) Characterization and inducibility of a scopoletin-degrading enzyme from sunflower. Phytochemistry 45:1109-1114