Altering expression of cinnamic acid 4-hydroxylase in transgenic plants provides evidence for a feedback loop at the entry point into the phenylpropanoid pathway

Abstract

Pharmacological evidence implicates trans-cinnamic acid as a feedback modulator of the expression and enzymatic activity of the first enzyme in the phenylpropanoid pathway, L-phenylalanine ammonia-lyase (PAL). To test this hypothesis independently of methods that utilize potentially non-specific inhibitors, we generated transgenic tobacco lines with altered activity levels of the second enzyme of the pathway, cinnamic acid 4-hydroxylase (C4H), by sense or antisense expression of an alfalfa C4H cDNA. PAL activity and levels of phenylpropanoid compounds were reduced in leaves and stems of plants in which C4H activity had been genetically down-regulated. However, C4H activity was not reduced in plants in which PAL activity had been down-regulated by gene silencing. In crosses between a tobacco line over-expressing PAL from a bean PAL transgene and a C4H antisense line, progeny populations harboring both the bean PAL sense and C4H antisense transgenes had significantly lower extractable PAL activity than progeny populations harboring the PAL transgene alone. Our data provide genetic evidence for a feedback loop at the entry point into the phenylpropanoid pathway that had previously been inferred from potentially artifactual pharmacological experiments.

Figure 1

A, Binary vector constructs used to alter expression levels of C4H in transgenic plants. Top, Empty vector control construct; middle, antisense construct; bottom, construct for over-expression. RB and LB, T-DNA right and left borders, respectively; NOS-Pro, NOS promoter sequence; NPT II, neomycin phosphotransferase sequence; NOS-ter, NOS terminator sequence; CaMV 35S Pro, CaMV 35S promoter from positions −800 to +1; E, EcoRI cleavage sites; CA4H, alfalfa cinnamic acid 4-hydroxylase cDNA sequence (1,740 bp). B, RNA gel-blot analysis of C4H transcript levels in transgenic tobacco harboring alfalfa C4H gene constructs. Plants shown were transformed with the empty vector construct (A designations), antisense construct (B and G designations), or sense construct (C designations). C, Extractable activities of C4H in young leaves from independent transgenic tobacco lines containing C4H sense or antisense constructs, or empty vector controls. Plants transformed with the sense construct but showing below control levels of C4H activity are classified as operationally co-suppressed.

Figure 2

Effects of modification of C4H expression on soluble and wall-bound phenolic compounds in leaves of tobacco primary transformants. A through C, HPLC traces (measured at 310 nm) of soluble phenolic compounds in leaf extracts from tobacco plants transformed with an empty vector construct (A, line 8A), a C4H antisense construct (B, line 13B), and a C4H sense construct leading to over-expression (C, line 201C). D through F, HPLC traces (measured at 270 nm) of solubilized wall-bound phenolic compounds from lines 8C, 13B, and 201C, respectively. G, Levels of CGA in leaves from a range of independent transgenic plants. Major soluble phenolic compounds eluting at around 13 and 16.5 min were identified as esters of caffeic acid. mAU, Milliabsorbance unit.

Figure 3

Relation between the activity of C4H and PAL or COMT activities in transgenic tobacco plants with C4H activity levels modified by expression of an alfalfa C4H transgene. A, C4H and PAL activities in young mature leaves; B, C4H and PAL activities in stems; C, C4H and COMT activities in stems. The lines analyzed were control (●) and C4H antisense or sense-suppressed plants (▪).

Figure 4

Levels of free cinnamic acid in stem tissue of transgenic plants with altered levels of C4H expression, plotted as a function of C4H activity. The lines analyzed were control (●) and C4H antisense or sense-suppressed plants (▪).

Figure 5

Segregation of PAL activity level in a cross between a PAL over-expressing line and a C4H antisense line. A, PAL activity in leaves from 76 F₁ progeny containing the bean PAL2 gene but no C4H antisense gene. B, PAL activity in leaves from 78 F₁ progeny harboring both the bean PAL2 gene and the alfalfa C4H antisense transgene. The plants designated by black bars and numbered in A and B were analyzed in a preliminary experiment, and the seeds of all the other plants were sown together at a later time. The three sets of error bars indicate the relative PAL activities, plus sd, in populations of independent progeny of the 10-6 PAL over-expressor line (n = 25), from vegetatively propagated 13B parent plants (n = 12), and from independent wild-type (Wt) plants (n = 9). Relative activities are given because of the need to analyze the individual plants in separate batches, each of which included a set of replicated samples to aid normalization (see “Materials and Methods”). The absolute values of PAL enzymatic activity in the two parent lines 10-6 and 13B in the second seed batch analyzed were 5.1 and 26.9 nmol h⁻¹ mg⁻¹ protein, respectively. C, Northern-blot analysis of alfalfa C4H and total PAL transcripts in parental lines (marked 10 and 13) and lines containing both PAL and C4H transgenes. Total RNA (15 μg/lane) was isolated from the parental lines and selected progeny from the 10-6/13B cross. Asterisks (*) indicate plant lines with high PAL activity. avg, Average.