Chemical Inactivation of the Cinnamate 4-Hydroxylase Allows for the Accumulation of Salicylic Acid in Elicited Cells

Abstract

The cinnamate (CA) 4-hydroxylase (C4H) is a cytochrome P450 that catalyzes the second step of the main phenylpropanoid pathway, leading to the synthesis of lignin, pigments, and many defense molecules. Salicylic acid (SA) is an essential trigger of plant disease resistance. Some plant species can synthesize SA from CA by a mechanism not yet understood. A set of specific inhibitors of the C4H, including competitive, tight-binding, mechanism-based irreversible, and quasi-irreversible inhibitors have been developed with the main objective to redirect cinnamic acid to the synthesis of SA. Competitive inhibitors such as 2-hydroxy-1-naphthoic acid and the heme-coordinating compound 3-(4-pyridyl)-acrylic acid allowed strong inhibition of C4H activity in a tobacco (Nicotiana tabacum cv Bright Yellow [BY]) cell suspension culture. This inhibition was however rapidly relieved either because of substrate accumulation or because of inhibitor metabolism. Substrate analogs bearing a methylenedioxo function such as piperonylic acid (PIP) or a terminal acetylene such as 4-propynyloxybenzoic acid (4PB), 3-propynyloxybenzoic acid, and 4-propynyloxymethylbenzoic acid are potent mechanism-based inactivators of the C4H. PIP and 4PB, the best inactivators in vitro, were also efficient inhibitors of the enzyme in BY cells. Inhibition was not reversed 46 h after cell treatment. Cotreatment of BY cells with the fungal elicitor β-megaspermin and PIP or 4PB led to a dramatic increase in SA accumulation. PIP and 4PB do not trigger SA accumulation in nonelicited cells in which the SA biosynthetic pathway is not activated. Mechanism-based C4H inactivators, thus, are promising tools for the elucidation of the CA-derived SA biosynthetic pathway and for the potentiation of plant defense.

Phenylpropanoids form a large family of plant-specific compounds implicated in a broad range of functions. Among the numerous chemical structures stemming from the pathway, lignin is a quantitatively major biopolymer that plays a key role in plant mechanical support and water transport and as physical barrier against pathogen infection. Other chemical classes of phenylpropanoids such as flavonoids, isoflavonoids, stilbenes, or coumarins have essential functions as antimicrobials, UV protectants, signaling molecules mediating interaction with insect or symbiotic bacteria, or pathogen response (Dixon and Paiva, 1995). An essential mediator of pathogen response and systemic acquired resistance is salicylic acid (SA; Dempsey et al., 1999).

Although much is known about the diversity and accumulation of the phenylpropanoid products, less is understood about networking and control of their biosynthesis. In particular, the diversity of enzymes catalyzing the same reactions, evolution of some biosynthetic branches believed to be resulting from speciation, and equilibration/compensation mechanisms between the different branches of the pathway remain elusive. In addition, some biosynthetic branches such as those of SA or coumarins are not yet elucidated either at the biochemical or at the molecular level.

The objective of this work was to develop new chemical effectors of the phenylpropanoid pathway. Chemicals allowing inactivation or enhancement of selected steps of the pathway are useful tools for both biochemical and molecular investigations, constituting alternatives or complements to mutation or transgenic techniques for gene up- or down-regulation. The main advantages of such chemical approaches are the simultaneous inhibition of all isoenzymes catalyzing the same reaction (provided that a reaction does not involve different families of proteins) and easy transposition to orthologous gene products.

The upstream part of the phenylpropanoid metabolism consists of three enzymatic steps leading to 4-coumaroyl CoA (Fig. 1). The cinnamate (CA) 4-hydroxylase (C4H) catalyzes the second step, i.e. the conversion of CA into p-coumarate. This enzyme is a member of the structural family of cytochrome P450 heme thiolate proteins, which catalyze monooxygenation of a broad range of substrates within all organisms (Werck-Reichhart and Feyereisen, 2000). In plant biosynthetic pathways, P450s catalyze critical hydroxylation steps and a variety of more complex reactions (Kahn and Durst, 2000). The C4H gene CYP73A1 was isolated from Jerusalem artichoke (Helianthus tuberosus; Teutsch et al., 1993). CYP73A1 orthologs have then been isolated from more than 20 plant species including tobacco (Nicotiana tabacum cv Bright Yellow [BY]; http://drnelson.utmem.edu/P450dbplant.html; Ralston et al., 2001). All of them belong to the CYP73A subfamily of P450 genes, and when the proteins were expressed in heterologous systems, they all displayed C4H activity.

Figure 1.

C4H and branching in the upper phenylpropanoid pathway. PAL, Phe ammonia-lyase; 4CL, 4-hydroxycinnamate CoA ligase; AOPP, amino-β-phenyl-propionic acid is an inhibitor of PAL (Amrhein et al., 1983); MDCA, methylene dioxocinnamic acid is an inhibitor of 4CL (Funk and Brodelius, 1990).

The substrate specificity and several inhibitors of the recombinant CYP73A1 expressed in yeast (Saccharomyces cerevisiae) have been described in detail (Pierrel et al., 1994; Batard et al., 1997). CYP73A1 substrates are planar, aromatic, molecules with an anionic site opposite (i.e. at about 8.5 Å) to the position of oxidative attack (Schalk et al., 1997, 1998).

The most efficient inhibitors of the C4H share the same structural characteristics as the substrates. Depending on the functions present in the molecule, they behave as simple competitors or mechanism-based irreversible or quasi-irreversible inactivators (Pierrel et al., 1994; Schalk et al., 1997, 1998). The most potent inhibitors reported so far are 2-hydroxy-1-naphthoate (2HN) and piperonylic acid (PIP). 2HN is a very potent competitive inhibitor that binds the active site with a 10-fold higher affinity than the substrate (K_s = 0.28 μm) but is not metabolized (Schalk et al., 1997). PIP is a natural compound isolated from the bark of Paracoto tree that behaves as a selective and potent mechanism-based quasi-irreversible inhibitor of the C4H (Schalk et al., 1998). Oxidative attack of the PIP methylenedioxo function leads to a carbene that forms a coordination with the P450 heme iron. The enzyme-inhibitor complex is fairly stable, but the carbene can be displaced at high-substrate concentration. In elicitor-treated BY2 cells, PIP also behaves as an effective C4H inactivator and inhibits the formation of p-coumarate and accumulation of the downstream metabolite scopoletin (Schalk et al., 1998).

A single branch point at CA is reported in the upstream phenylpropanoid pathway, which leads to benzoic acids such as the defense signal SA by a mechanism that is not yet elucidated (Chong et al., 2001). Preliminary experiments have suggested that PIP inhibition of C4H activity promotes the accumulation of SA in tobacco BY2 cells (Chong et al., 2001).

PIP and 2HN are both at least partially reversible inhibitors of C4H. Irreversibly binding inactivators of the enzyme should thus be more efficient for obtaining strong inhibition and redirection of metabolic fluxes in vivo. Compounds containing a terminal alkyne are known as potent irreversible inactivators of P450s (Ortiz de Montellano and Komives, 1985). 2,4-Dichlorophenoxypropyne was described as such an inactivator of C4H (Pierrel et al., 1994). This compound was however not designed to target C4H, and its inhibition and binding constants are far from optimal. In this paper, we report new competitive and mechanism-based irreversible inactivators of CYP73A1, and compare their relative efficiencies at inhibiting C4H in tobacco cell cultures and promoting accumulation of SA with those of PIP and 2HN.

RESULTS

Binding of Inhibitors to CYP73A1

The set of CA analogs tested as inhibitors of CYP73A1 is shown in Figure 2. New types of P450 inhibitors were added to PIP and 2HN in these experiments, including trans-3-(pyrid-4-yl)-acrylic acid (3PA), the sp₂ hybridized ring nitrogen of which is positioned to coordinate as the sixth ligand to the heme iron. Such a compound, anchored both on the protein and the heme, is expected to behave as a tight-binding but reversible inhibitor of the C4H (Ortiz de Montellano and Correia, 1995). Analogs bearing a terminal alkyne function and more closely mimicking the structure of cinnamic acid than 2,4-dichlorophenoxypropyne, 4-propynyloxybenzoic acid (4PB), 4-propynyloxymethylbenzoic acid (4PO), and 3-propynyloxybenzoic acid (3PB) were also synthesized. Oxidative attack of the triple bond of such compounds is likely to lead either to heme alkylation or to the formation of a ketene that can react with the protein or be hydrolyzed to a carboxylic acid metabolite (Ortiz de Montellano and Komives, 1985; Ortiz de Montellano and Correia, 1995; Regal et al., 2000). These terminal acetylenes are expected to behave as mechanism-dependent irreversible inhibitors of the enzyme.

Figure 2.

Molecules tested as inhibitors of CYP73A1.

The binding of ligands near the heme of the oxidized P450 influences the spin state and the redox potential of the heme iron, which triggers the reaction or makes it more difficult, and results in a change in visible absorbance of the heme protein (Jefcoate, 1978). The binding properties of compounds in Figure 2 were investigated with yeast microsomes expressing CYP73A1 (Table I). All three propynyl derivatives led to a shift of the maximum of absorbance of the oxidized P450 from 420 to 390 nm, similar to that observed with cinnamic acid or PIP (Schalk et al., 1998). The resulting “type I” difference spectrum reflects the displacement of a water sixth ligand to the heme iron, correlated to an increase in spin and redox potential that favors electron supply and activates the catalytic cycle (Raag and Poulos, 1989). The amplitude of the binding spectra obtained with 4PB, 3PB, and 4PO is larger than that previously reported for 2,4-dichlorophenoxypropyne (Pierrel et al., 1994). These new compounds are thus better positioned for oxidative attack in the active site and are therefore expected to behave as more efficient mechanism-based inactivators of the C4H. Displacement of water from the active site, however, is incomplete, and dissociation constants remain high compared with the K_s for CA (8 μm; Pierrel et al., 1994), indicating that these inhibitors are not yet fully optimized.

Table I.

Inhibitor binding and inhibition constants determined with recombinant C4H (CYP73A1)

Inhibitor	Binding		Suicide Inhibition		Competitive Inhibition
	Ks	High spina	Ki	Kinact	Ki
	μm	%	μm	min−1	μm
Mechanism-based irreversible inactivator
3PB	204	40	10.5	0.346
4PO	86	11	580	0.22
4PB	390	78	53	0.41
3-(2,4-Dichlorophenoxy)-1-propyne	Type I (low shiftb)		94	0.122
Mechanism-based quasi-irreversible inactivator
PIP	45	34	17	0.064
Reversible inhibition
3PA	1.7	Type IIc			2
2HN	0.28	58			0.048

Binding and inhibition constants were determined using H. tuberosus C4H (CYP73A1) in microsomes from yeast W(R). Binding constants were determined from the shift of heme maximum of absorbance detected upon binding of increasing concentrations of inhibitor. Inhibition constants were calculated from residual C4H activity.

^aA 100% conversion of CYP73A1 to high spin is obtained for saturating cinnamic acid concentration, with an ɛ_{type I} of 125 mm⁻¹ cm⁻¹ (Urban et al., 1994). 

^bA shift of the maximum of absorbance was observed, but ΔA_max was too small for determination of the binding parameters. 

^cA type II difference spectrum was recorded with an ɛ_{type II} of 95 mm ⁻¹ cm⁻¹. 

As expected for a compound forming a coordination to ferric heme iron, the binding of 3PA induces a shift of the maximum of absorbance of CYP73A1 from 414 to 428 nm, giving rise to a “type II” difference spectrum. The resulting high- to low-spin transition is accompanied by a change in redox potential that makes P450 reduction more difficult. The dissociation constant for 3PA is slightly lower than that measured for CA, and the maximum amplitude of the difference spectrum (ε_{type II} of 95 mm⁻¹ cm⁻¹) is high, which indicates that 3PA is a nearly optimized ligand of CYP73A1. 3PA should thus behave as a strong competitive inhibitor of C4H. In agreement with these data, no metabolism of 3PA is detected upon incubation with recombinant CYP73A1.

In Vitro Mechanism-Based Inactivation of CYP73A1

4PB, 3PB, and 4PO were tested as mechanism-based inactivators of CYP73A1. Recombinant yeast microsomes were incubated with NADPH and various concentrations of each of the molecules before determination of residual C4H activity. C4H inactivation was concentration- and time-dependent for all three compounds and did not occur in the absence of NADPH. A high concentration of CA in the assay for residual activity did not reverse the inhibition. Pseudo-first-order kinetics were obtained in the initial phase of the inactivation (Fig. 3). The time required for half-maximal inactivation (t_1/2) at each inactivator concentration was calculated and plotted versus the reciprocal of the inhibitor concentration. The plots were typical of an inactivation that proceeds with saturation (Silverman, 1996). Kinetic constants calculated from the plots are summarized in Table I and compared with those previously determined for 2,4-dichlorophenoxypropyne and PIP (Pierrel et al., 1994; Schalk et al., 1998). Higher K_inact observed for 4PB and 3PB were correlated to lower K_I and to large changes in heme iron spin state upon binding of the inhibitors. On the basis of these data, inactivation of CYP73A1 with 100 μm inhibitor occurs in 2.5 min with 4PB or 3PB, compared with 10 min with 2,4-dichlorophenoxypropyne or PIP and 21 min with 4PO. 4PB and 3PB thus appear to be the most potent irreversible inactivators of C4H enzymes.

Figure 3.

Time- and concentration-dependent inactivation of CYP73A1 by 4PB, 3PB, and 4PO. Recombinant CYP73A1 in W(R) yeast microsomes was preincubated in the presence of NADPH and different concentrations of the propynyl derivatives before assaying residual C4H activity. A, Time course of the C4H inactivation measured upon incubation with 4PB. Similar pseudo-first-order kinetics of inactivation were obtained with 3PB and 4PO. B, Plot of the half-inactivation times, estimated from linear regression analysis of curves shown in A, versus reciprocal of inhibitor concentration. C and D, Half-inactivation time plots, as in B, for 3PB and 4PO.

Attempts to detect carboxylic acid metabolites of 4PB upon incubation with CYP73A1 were unsuccessful, indicating either a further modification of the product of the reaction or a very low partition ratio. In favor of the second hypothesis, only a very small proportion of the inhibitor was processed after complete inactivation.

Competitive Inhibition of CYP73A1

3PA does not bear any function likely to lead to P450 inactivation but very closely mimics cinnamic acid. Combined with this strong structural homology to CA, coordination of its 4-nitrogen to the heme iron should lead to a very tight binding to the active site. No time-dependent inhibition was observed with 3PA that, as expected, behaved as a typical and strong competitive inhibitor (Fig. 4). The K_I determined from the kinetic analysis is 2 μm, which is in good agreement with the K_S measured in binding experiments.

Figure 4.

Inhibition of C4H activity by 3PA. Lineweaver-Burk representation of the inhibition of C4H activity of recombinant CYP73A1 (3 nm in the assay) by 3PA (0, 1, 3, or 10 μm). Data are means of duplicate experiments.

C4H Activity and Inhibition in Vivo

The best inhibitor of each type, i.e. 2HN (simple competitive), 3PA (iron-coordinating, competitive), PIP (quasi-irreversible, mechanism-dependent), and 4PB (irreversible, mechanism-dependent) was selected for comparison of C4H inhibition efficiency in planta.

To set up a test for the evaluation of in vivo C4H activity, a tobacco BY cell suspension culture was fed with 20 μm [¹⁴C]CA. Free phenolic acid content was extracted, and radiolabeled CA and p-coumarate were quantified by thin-layer chromatography (TLC) and HPLC. p-Coumarate was the only radiolabeled metabolite recovered in these assays. Its identity was confirmed by HPLC-diode array analysis. The time-dependent formation of p-coumarate relative to initial CA is represented in Figure 5. Approximately 25% of the initial [¹⁴C]CA is converted into p-coumarate in 30 min. p-Coumarate then declines, probably because of further conversion into non-ether-soluble esters or other metabolites. C4H inhibition in BY cells was therefore evaluated by measuring the formation of p-coumarate after 30 min of incubation with radiolabeled substrate.

Figure 5.

Time course of the conversion of CA into p-coumarate in a tobacco BY cell suspension culture. One-milliliter aliquots of 5-d-old cells were incubated with 20 μm [¹⁴C]CA in Eppendorf tubes at 30°C. CA to p-coumarate conversion was evaluated after TLC analysis of extracted free phenolics, and identity of radiolabeled spots was confirmed by HPLC-diode array analysis. Data are expressed as a proportion of initial CA recovered as p-coumarate at the end of the incubation.

The inhibitors were added to the BY cell suspension culture, and aliquots were collected for monitoring of residual activity (Fig. 6). Strong C4H inhibition was observed with all four molecules in aliquots collected immediately after addition of the inhibitors. After 4 h, 2HN, 3PA and PIP inhibition was more than 95%. After 22 h, competitive inhibition by 2HN and 3PA was largely reversed, either because of an accumulation of substrate and displacement of the inhibitor or because of an inhibitor conjugation and/or compartmentation. Both mechanism-based inactivators, 4PB and PIP, remained fully effective for at least 46 h. Slow displacement by CA of the inhibitory carbene formed upon PIP metabolism was previously reported (Schalk et al., 1998). The present data thus suggests that if CA concentrations reached in this experiment are high enough for reversion to occur, the enzyme inactivation by PIP in vivo was probably faster than the displacement of the derived carbene by CA. This would be in agreement with previous observations in a recombinant system in vitro (Schalk et al., 1998).

Figure 6.

Inhibition of C4H activity in tobacco BY cells. Inhibitors in DMSO were added to a 5-d-old suspension culture. DMSO was added to the control. Final concentrations in the culture were 0.06% (v/v) DMSO, 40 μm 2HN, 40 μm 3PA, 100 μm 4PB, and 10 μm PIP. Aliquots collected after 0.5, 4, 22, and 46 h of incubation were assayed for conversion of [¹⁴C]CA (20 μm). Data are means ± sd of four measurements.

The same cell treatment was performed after 30 min of elicitation with β-megaspermin, a proteinaceous fungal elicitor triggering hypersensitive response (Baillieul et al., 1995). C4H inhibition with all four compounds was similar to that observed with nonelicited cells at 0.5 h. Only 4PB provided effective inhibition (70%) after 22 h. After 46 h, none of the inhibitors influenced CA conversion anymore. Increased flux of metabolites and induced enzyme synthesis thus relieve the inhibition, probably because of displacement of reversible inhibitors and complete metabolism of mechanism-dependent inactivators. In agreement with this hypothesis, the irreversible mechanism-dependent and poorly metabolized inhibitor 4PB is the compound that provides the longer lasting inhibition.

Inhibition of the PAL

Cinnamic acid and structurally related molecules were reported to behave as PAL inhibitors (Sato et al., 1982). Because our aim was to develop C4H-specific inhibitors allowing redirection of the flux of metabolites from cinnamic acid toward that of SA, the effects of 2HN, 3PA, PIP, and 4PB on PAL activity were also investigated (Fig. 7). PAL activity was tested on a whole extract of tobacco BY cells pretreated with β-megaspermin to induce the enzyme expression. With PIP, 4PB, and 3PA, no PAL inhibition was observed at concentrations up to 500 μm. A concentration-dependent inhibition was however observed with 2HN, with PAL inhibition reaching 20% with 100 μm of the compound. The concentration of 2HN used for subsequent C4H inhibition in vivo (40 μm) thus should not very significantly interfere with PAL activity, and no PAL inhibition from 3PA, PIP, and 4PB is expected.

Figure 7.

Inhibition of PAL activity. Whole extracts of β-megaspermin pretreated BY cells were diluted 5-fold in 100 mm borate buffer, pH 8.8, and incubated with 85 μm [¹⁴C]Phe and various concentrations of inhibitors for 1 h at 37°C. Radiolabeled cinnamic acid was extracted and counted by liquid scintillation. Data are means ± sd of two experiments. 100% activity = 19 pmol h⁻¹μL⁻¹.

SA Accumulation in Inhibitor-Treated BY Cells

The expected impacts of an in vivo C4H inactivation are a decreased formation of downstream metabolites, such as scopoletin (Schalk et al., 1998), and the accumulation of CA and derived compounds such as benzoic and SAs (Chong et al., 2001). To compare the relative efficiencies of 4PB and PIP for triggering accumulation of SA, BY cell suspension cultures were treated with inhibitor concentrations of 10 μm. Low inhibitor concentrations were used in this experiment to ensure selectivity of the inhibition and to avoid interfering with the benzoate hydroxylase activity that was previously suggested to be a P450 enzyme (Leon et al., 1995). Total (free plus conjugated) SA accumulation was monitored with and without a prior treatment of the cells with β-megaspermin (Fig. 8). The basal content in control, dimethylsulfoxide (DMSO)-treated cultures was 300 ng total SA g⁻¹ wet cells. No significant variation of this basal SA level was detected over 30 h in the absence of β-megaspermin treatment. Upon treatment with elicitor, SA accumulation was induced to reach a maximum of 1,800 ng g⁻¹ at 24 h.

Figure 8.

Accumulation of total SA in inhibitor-treated cells. A 5-d-old tobacco BY cell culture was treated with 10 μm PIP, 10 μm 4PB, or DMSO. Identical treatment was performed 30 min after elicitation with β-megaspermin (β-meg). Free and conjugated SA were extracted from aliquots and hydrolyzed, and total SA was purified and quantified by fluorometry. The experiment was performed on two subcultures maintained independently to evaluate variations of cell response.

When the BY cells were treated with inhibitors, only very slight increases in SA, never exceeding 45% of the control, were detected in the absence of β-megaspermin. In the presence of the elicitor, SA accumulated with a time course similar to that observed in the absence of inhibitor, but levels of total SA at 24 h were about 8-fold that of the control for cells treated with PIP and 3.6-fold with 4PB. The highest levels of SA reached after treatment with PIP and β-megaspermin is close to 13 μg g⁻¹ wet cells. In addition to higher SA accumulation, PIP also triggers an earlier response compared with 4PB, already close to maximal at 12 h. Upon 4PB treatment, however, SA accumulation persist longer and is still 3-fold that of the control after 30 h.

DISCUSSION

New Reversible and Irreversible C4H Inhibitors

The main objective of this work was to develop new C4H inhibitors to control the flow of the metabolites in the main phenylpropanoid pathway and to investigate the impact of a redirection of this flow on the synthesis of SA and plant defense. Efficient C4H inhibitors may in particular constitute very useful tools to help characterization of the SA biosynthetic routes that remain elusive.

Different types of inhibitors have been described, based on specific properties of cytochromes P450. In addition to classical competitive inhibitors, the synthesis of structural analogs of the substrate bearing activatable functions allows the design of molecules selectively inactivating specific P450 isoforms (Ortiz de Montellano and Correia, 1995). Once metabolized, such compounds can behave as tight-binding or irreversible ligands of the enzyme. Another type of inhibitor, which simultaneously binds the protein and the prosthetic heme iron, is often also more effective than molecules that rely on only one of these interactions. The different types of P450 inhibitors were tested and compared in this study.

3PA was assayed for its potential ability to coordinate heme iron by the intermediate of the electron pair of its ring nitrogen and to simultaneously bind through its carboxylate the substrate docking region of the protein. It behaved as a good ligand and formed the expected heme coordination resulting in characteristic absorption spectra. As expected for such a type of direct heme ligand inducing a high- to low-spin shift of the iron, it was not metabolized but behaved as a competitive inhibitor. 3PA dissociation constant was however not significantly lower than that measured for the substrate and much higher than that of the classical competitive inhibitor 2HN. A possible explanation of this relatively weak interaction of 3PA can be found in NMR measurements of the distances of the substrate protons to the heme iron, which indicate that the initial positioning of the substrate and its analogs in the active site is approximately parallel with the heme plane (Schoch, 2001). Such a positioning does not favor stable 3PA coordination to the heme iron. In good agreement with measured binding constants, 2HN behaved as a better inhibitor of the C4H than 3PA. A stronger anchoring of 2HN to the C4H is probably provided by its additional phenolic function that does not compromise other interactions with the protein.

We showed previously that PIP is a potent quasi-irreversible mechanism-dependent inhibitor of the C4H (Schalk et al., 1998). Tight heme-binding of the inhibitory carbene formed during PIP metabolism can however be slowly reversed by high concentrations of CA in vitro. An attempt was thus made at designing completely irreversible mechanism-dependent inhibitors. 4PB, 3PB, and 4PO were designed taking into account the C4H substrate pharmacophore deduced from substrate specificity experiments (Pierrel et al., 1994; Schalk et al., 1998), including size, aromatic ring, and carboxylate group for anchoring in the active site. An activatable terminal acetylene was grafted in different positions onto oxybenzoic or oxytoluic acids. The resulting molecules partially mimic cinnamic and naphthoic acids, which are the best substrates of the C4H, and the acetylenic function should be positioned to favor its oxidative attack in the active site of the CAH. A limitation for the optimization of such molecules and to adapt them to the C4H substrate pocket was however the linear structure of the alkyne carbons.

As expected, 4PB, 3PB, and 4PO all behaved as mechanism-dependent inactivators of recombinant C4H in yeast microsomes. 4PB and 3PB, being better positioned in the active site, were more effective inhibitors of C4H than 4PO and than previously described compounds (Pierrel et al., 1994). 2-Ethynylnaphthalene, an effective P450 inactivator of several mammalian enzymes (Hopkins et al., 1992; Roberts et al., 1993; Strobel et al., 1999), which was included in our assays, did not bind C4H nor inhibit its activity, confirming the need for an appropriate design to achieve efficient and selective enzyme inactivation.

Compared Inhibition Efficiency in Vivo

All types of inhibitors efficiently blocked the C4H activity in cell culture experiments. Efficiency of 3PA and 2HN was correlated to their respective enzyme affinities. Inhibition by such competitive inhibitors was however rapidly reversed. This was possibly attributable to elevated CA concentrations in the treated cells or resulted from their inactivation by detoxification systems (conjugation and compartmentation). Fast reversion of C4H inhibition is a strong limitation to the use of 3PA and 2HN for in vivo experiments.

Very similar responses to the quasi-irreversible and irreversible mechanism-dependent inactivators, PIP and 4PB, were obtained when C4H inhibition was assayed in cell cultures. PIP was still effective 46 h after cell treatment, when the experiment was carried on in the absence of elicitor. This suggests that the CA concentrations in the PIP-treated cells are not sufficient to dissociate the enzyme-inhibitor complex or that the dissociation rate of the complex is slower than the further inactivation of the enzyme. PIP thus remains a useful C4H inactivator, effective at a low concentration, and is therefore likely to be specific. Its efficiency is further proved by the SA accumulation triggered by PIP concentrations as low as 10 μm.

C4H inhibition by 3PA and 2HN being very transient, only PIP and 4PB were tested for their ability to redirect CA to the synthesis of SA. In cells cultured under normal axenic conditions, low SA concentrations were detected and these concentrations were only slightly increased upon treatment with C4H inhibitors (Fig. 8). Under such conditions, the SA biosynthetic pathway thus seemed almost inactive, possibly because of low precursor supply (Chong et al., 2001). Upon treatment with β-megaspermin, the SA pathway was activated, and the activation was transient, reaching a maximum around 24 h after treatment. Addition of both PIP and 4PB to the elicited cells led to a dramatic increase in SA accumulation and to an effective reorientation of CA metabolism. At a 10 μm concentration, PIP allowed for an almost complete C4H inhibition and a very strong (8-fold) but also early (within less than 6 h) increase in SA accumulation compared with untreated cells. At 30 h, SA accumulation then dropped very sharply. Because PIP-inhibition of C4H activity is already reversed after 22 h in elicited cells (data not shown), this is more likely to reflect fast C4H recovery resulting from enzyme-inhibitor dissociation or PIP inactivation than a decrease in CA to SA conversion, in PAL activity, or in inhibition of an upstream step in the shikimate pathway. Used at the same low concentration, 4PB leads to a smaller increase in SA than PIP, most likely as a result of its lower affinity for C4H and only partial enzyme inhibition (Fig. 3). The low metabolism of this inhibitor and its small partition ratio (i.e. high inactivation efficiency) however ensure a longer lasting effect.

CONCLUSIONS

A set of C4H inhibitors has been developed, allowing short to long term modulation of the enzyme activity in plant tissues. Such inhibitors should be useful tools to help elucidate the molecular mechanisms of the conversion of CA to SA and for understanding regulation of the SA and general phenylpropanoid pathways. Their reported effect on SA accumulation in elicited tobacco cell cultures (this work; Chong et al., 2001) strongly suggests that, in this plant material, SA biosynthesis branches on CA and not chorismate as recently reported for Arabidopsis (Wildermuth et al., 2002). If such inhibitors cannot trigger SA accumulation per se, they might be very useful to potentiate and/or accelerate the onset of defense systems in elicited or pathogen-infected plants.

MATERIALS AND METHODS

Chemicals

CA, PIP, 2-HN, NADPH, SA, and DMSO were from Sigma-Aldrich (L'Isle d'Abeau Chesnes, France), 3PA was from Maybridge (Tintagel, UK), trans-[3′-¹⁴C]CA was from Isotopchim (Ganagobie, France), [1′-¹⁴C]SA and l-[¹⁴C(U)]-Phe were from NEN (Zaventem, Belgium). 4-Hydroxybenzoic acid, 3-hydroxybenzoic acid, 4-(hydroxymethyl) benzoic acid, 4-hydroxycinnamic acid, 3-hydroxycinnamic acid, propargyl bromide, and acetyl chloride were from Aldrich (Milwaukee). β-Megaspermin purified from culture filtrates of Phytophthora megasperma (Baillieul et al., 1995) was kindly provided by Dr. S. Kauffmann (Institut de Biologie Moléculaire des Plantes, Strasbourg, France).

Synthesis of the Propynyl Derivatives (Fig. 9)

Figure 9.

General synthetic route for the synthesis of propynyl inhibitors of CYP73A1.

Flash column chromatography techniques were performed by using silica gel (mesh-230–400). The gas chromatography/mass spectroscopy (GC/MS) analyses of the synthesized products were carried out on a gas chromatograph/mass spectrometer (Chem Station 5997, Hewlett Packard, Palo Alto, CA). NMR spectra were recorded on an Omega 400 MHz FT-NMR spectrometer (General Electric, Fairfield, CT).

4-Hydroxybenzoate

Acetyl chloride (3.6 g, 46 mmol) and methanol (15 mL) were mixed together at 0°C and then 4-hydroxybenzoic acid (5 g, 30.5 mmol) was added. After the mixture was stirred at room temperature overnight, the methanol was removed, and the residue was dissolved in CH₂Cl₂. The solution was washed with a saturated aqueous solution of NaHCO₃ and water and then dried over magnesium sulfate. The solvent was removed under reduced pressure to give 4-hydroxybenzoic acid methyl ester (4.8 g, 88% yield). ¹H NMR δ 7.99 (d, 2H, Ar-H), 7 (d, 2H, Ar-H), 4.6 (s, 1H, OH), and 4.3 (s, 3H, CH₃); GC/MS m/z 152.

4-Propynyloxybenzoate

Solution of powdered KOH (6 g, 108 mmol) in DMSO (12 mL) was prepared. After stirring the mixture for 5 min, 4-hydroxybenzoic acid methyl ester (4.8 g, 27 mmol) was added followed immediately by propargyl bromide (7 g, 48 mmol). Stirring continued for 1 h at room temperature. The reaction mixture was poured into water (40 mL) and extracted with CH₂Cl₂ (2 × 30 mL). The combined organic extracts were washed with water, dried over magnesium sulfate and reduced to dryness under reduced pressure. Chromatography on silica gel with eluent hexane:ethyl acetate (3:1, v/v) gave 4PB methyl ester (3 g, yield 59%). ¹H NMR δ 7.99 (d, 2H, Ar-H), 7 (d, 2H, Ar-H), 4.9 (s, 2H, CH₂), 4.3 (s, 3H, CH₃), and 3.1 (s, 1H, CH); GC/MS m/z 190.

4PB

4PB methyl ester (3 g, 16 mmol) was added to a solution of NaOH (1.6 g, 40 mmol) in ethanol (25 mL) and water (8 mL). After stirring overnight at 25°C the alcohol was removed under reduced pressure, and the reaction mixture was extracted with methylene chloride (2 × 30 mL). The combined organic extracts were washed with water and acidified with 2 n HCl. The product precipitated as white crystals, which were recrystallized from ethanol-water to give 4PB (2.4 g, yield 84%). Melting point (mp) 195°C, ¹H NMR δ 7.99 (d, 2H, Ar-H), 7 (d, 2H, Ar-H), 4.9 (s, 2H, CH₂), 3.1 (s, 1H, CH). GC/MS m/z 176.

3PB

This compound was prepared from 3-hydroxybenzoic acid following the chemical procedure above. mp 130°C, ¹H NMR δ 7.65–7.40 (m, 4H, Ar-H), 4.8 (s, 2H, CH₂), and 3.1 (s, 1H, CH); GC/MS m/z 176.

4PO

The compound was prepared from 4-(hydroxymethyl) benzoic acid following the chemical procedure above. mp 135°C, ¹H NMR δ 8 (d, 2H, Ar-H), 7.5 (d, 2H, Ar-H), 4.65 (s, 2H, CH₂), 4.2 (s, 2H, CH₂), and 3 (s, 1H, CH); GC/MS m/z 190.

Yeast (Saccharomyces cerevisiae) Expression and Microsome Preparation

In vitro inhibition experiments were performed using microsomal preparations from yeast W(R) transformed with the expression plasmid pCA4H/V60, allowing co-expression of the yeast P450 reductase and CYP73A1 (Urban et al., 1994). Microsomal fractions were prepared as described by Pompon et al. (1996).

Spectrometric Measurements and Enzyme Assays

P450 content was calculated from reduced P450-CO versus reduced difference UV-visible spectra (Omura and Sato, 1964). Low- to high-spin conversion and dissociation constants of enzyme-ligand complexes were evaluated from type I ligand binding spectra as described by Schalk et al. (1997). Efficiency of ligand-induced high- to low-spin conversion was deduced from type II difference spectra (Jefcoate, 1978).

trans-Cinnamic acid hydroxylation was assayed using radiolabeled trans-[3′-¹⁴C]cinnamic acid as described by Pierrel et al. (1994), except that the TLC analysis was performed using ether:petroleum ether:formic acid (50:50:1, v/v) as a solvent. 4-Coumaric (R_F 0.3) and cinnamic acids (R_F 0.6) were scraped, and radioactivity was counted by liquid scintillation. For determination of the kinetic constants, data were fitted using the nonlinear regression program DNRPEASY derived from DNRP53 (Duggleby, 1984).

The time-dependent inactivation of the C4H in yeast microsomes was evaluated using a general dilution procedure (Silverman, 1996). CYP73A1 (400 nm) was preincubated at 28°C in 100 mm sodium phosphate pH 7.4, in the presence of 500 μm NADPH, 1 mm Glc-6-phosphate, 0.2 unit of Glc-6-phosphate dehydrogenase, and various concentrations of inactivator. For each preincubation time, an aliquot of the inactivation medium was diluted 20-fold into a second incubation mixture containing 150 μm of [¹⁴C]CA to assay residual activity.

Treatment of Tobacco (Nicotiana tabacum) Cell Suspension Culture

The tobacco BY cell suspension was maintained as described for BY-2 lines (Nagata et al., 1992). For measurement of SA accumulation or Phe ammonia-lyase (PAL) activity, 50 nm β-megaspermin was added to the culture 30 min or 12 h before the experiment, respectively. Inhibitors were added solubilized in DMSO. The final DMSO concentration in the culture was 0.06%.

In Vivo Enzyme Assays

In vivo conversion of trans-CA into p-coumarate was evaluated using 1 mL of cell culture (about 0.3 g wet cells) incubated with 20 μm of [¹⁴C]CA (140 Bq/nmol) at 30°C under agitation. The reaction was stopped by addition of 500 mm HCl from a 4 n solution. Free phenolic acids were extracted three times with 2 volumes of ether:petroleum ether (50:50, v/v), concentrated under nitrogen and analyzed by TLC. Radiolabeled spots were scraped, eluted with ether, and further analyzed by HPLC on a LiChrosorb RP-18 column (Merck 4 × 125 mm, 5 μm) at a flow rate of 1 mL min⁻¹, with elution 5 min isocratic, followed by a 20-min linear gradient from 10% to 52% (v/v) acetonitrile, in water containing 0.2% (v/v) acetic acid. trans-CA and 4-coumarate were identified from their retention time, UV absorbance, and radioactivity, with reference to commercial standards.

For C4H inhibition experiments, 15 mL of 5-d-old-BY cell suspension were subcultured 46 h in the presence of the inhibitor. At time intervals, 1-mL aliquots were assayed for residual C4H activity. For PAL inhibition measurements, filtered BY cells were extracted and enzyme activity was monitored, as described previously (Pellegrini et al., 1994).

SA Quantification

For evaluation of SA accumulation, 40 mL of 5-d-old BY cell suspension culture was treated with inhibitors simultaneously in the presence or absence of elicitor. At time intervals, 3-mL aliquots were filtered, weighted, and frozen in liquid nitrogen. After addition of 2 nCi of [¹⁴C]SA as internal standard, total phenolics were extracted successively with 2 mL g⁻¹ cells of 90% and 100% (v/v) methanol and vacuum dried. Conjugated phenols were hydrolyzed 1 h at 80°C after the addition of 1 mL of 4 n HCl. Free acids were then extracted three times with 2.5 volumes of ether:petroleum ether (50:50, v/v). Organic phases were pooled, dried under nitrogen, and resuspended in solvent A (97% 20 mm sodium acetate, pH 5, and 3% [v/v] acetonitrile).

Samples were injected on a Nova Pak RP-C18 column (3.9 × 150 mm, 5 μm; Waters, Milford, MA) equilibrated in solvent A, at a flow rate of 1 mL min⁻¹, with elution 10 min isocratic and then a 2-min linear gradient to 80% (v/v) acetonitrile. SA was detected by fluorometry (λ_ex315 nm, λ_em 405 nm). To obtain a consistent signal and to avoid saturation of the detector, fractions from 0.1 to 0.005 of each sample were injected until a suitable response was obtained. Quantification of the peak areas was performed relative to commercial standards. An aliquot (0.25) of each fraction injected was counted by liquid scintillation to determine recovery of the internal standard.