Unraveling the regulation of floral fragrance biosynthesis

Abstract

Floral volatiles are biologically and economically important plant derived chemical compounds. In petunia, floral volatile benzenoid/phenylpropanoid (FVBP) biosynthesis is controlled spatially, developmentally, and hormonally at molecular, metabolic and biochemical levels. Over the last years, numerous genes have been shown to encode proteins that either directly catalyze a biochemical reaction yielding FVBP compounds, or are involved in metabolite flux prior to the formation of FVBP compounds. This FVBP gene network is specifically and coordinately transcribed. Multiple R2R3-MYB transcription factors are involved in the regulation of genes in the core metabolic pathways leading to a very unique mixture of emitted floral volatiles. The molecular puzzle is not complete, since the functions of the few FVBP transcription factors identified to date do not fully explain the transcriptional regulation of the entire gene network.

Introduction

Floral fragrance has gained attention over the last decade with attempts to understand the biosynthesis, emission, regulation and ecological impacts of emitted floral volatile compounds. Petunia x hybrida ‘Mitchell Diploid’ (MD) has functioned as a model system for floral volatile benzenoid/phenylpropanoid (FVBP) research. Our current understanding of FVBP biosynthesis consists of a complex model where phenylpropanoid production is controlled at genetic, molecular, biochemical and metabolic levels from primary metabolism to the production of individual FVBP compounds.

MD flowers afford investigators a specialized tissue enriched for very specific aspects of the phenylpropanoid pathway with low molecular background. In MD flowers, many derived products of the phenylpropanoid pathway are volatile compounds,^1,2 which can be used to assay numerous features of the pathway. Initial experimentation with petunia focused on genes and gene products that were responsible for the direct formation of volatile compounds emitted by the MD flower such as, S-ADENOSYL-L-METHIONINE:BENZOIC ACID/SALICYLIC ACID CARBOXYL METHYLTRANSFERASE (PhBSMT).³ Over time, the biosynthesis of seven FVBP compounds had been described and attributed to specific genes belonging to divergent biochemical families (e.g., methyltransferase, decarboxylase and reductase).^3–5 During this time a single transcription factor was isolated and implicated in the regulation of core biosynthetic genes involved in the shikimate pathway. The shikimate pathway ultimately provides the aromatic amino acid phenylalanine, which “feeds” the production of FVBP compounds.^4,6 This transcription factor is an R2R3-MYB named PhODORANT1 (PhODO1) because when the endogenous transcript was severely reduced, a concomitant reduction in emitted FVBPs was observed.⁷ Individual characterizations of all the aforementioned genes suggested a network or cluster of similarly regulated genes may comprise a specific FVBP gene syndicate.

FVBPs are derived from the aromatic amino acid phenylalanine, and the phenylpropanoid pathway (Fig. 1), which produces the majority of phenolic compounds found in the environment. Over 7,000 phenylpropanoid compounds have been documented,⁸ and phenylpropanoid compounds are thought to constitute approximately 20% of the total carbon in the terrestrial biosphere.⁹ The biological function and/or importance of these compounds can range from cell wall composition to environmental interaction.^10,11 Many of these phenolic compounds are important or useful to humans such as: dietary purposes, antioxidants and phytoestrogens; pharmaceuticals; and bioplastic materials. Taken together, the understanding of core regulation and biomechanics of the phenylpropanoid pathway is essential for numerous aspects of plant biology and modern society.

Figure 1.

A current model for the floral volatile benzenoid/phenylpropanoid (FVBP) pathway in petunia. Yellow lines, lipid bilayers; green lines, plastid envelope. Single red arrows depict biochemical reactions, while multiple red arrows indicate multiple steps. Major pathways shown are glycolysis (FVBP substrate, PEP; black box denotes possible separation), the pentose phosphate pathway (FVBP substrate, E-4-P; black box denotes separation), the shikimate pathway (FVBP substrate, chorismate), the phenylpropanoid pathway (FVBP substrates; phenylalanine, t-cinnamic acid and ferulic acid). FVBPs are boxed in purple. Enzymes are bold red and transcription factors are in bold gray, capitalized characters. A blue oval denotes mitochondrial location. The individual FVBP pathway branches are numbered consecutively 1–3. Biochemical and molecular steps that have been directly identified in petunia have the genus and species prefix to the proteins.

FVBP Gene Network

The emission of FVBP compounds from the MD flower is organized and coordinated.^1,12,13 Therefore, it is reasonable to hypothesize the expression of genes responsible for the biosynthesis of emitted FVBP compounds is also organized and coordinated. When the transcript accumulation of seven genes involved in FVBP production were examined simultaneously through various dimensions of gene expression, a specific and coordinate gene regulation was obvious.¹³ The highest relative level of transcript accumulation for all the FVBP genes examined is detected in the corolla limb tissue of an open MD flower. The vast majority of emitted FVBP compounds are detected from the corolla limb tissue of an open MD flower.^12,13 FVBP emission is greatly reduced by an endogenous ethylene signal upon a successful pollination and fertilization event.¹² Likewise, transcript accumulation from FVBP genes is greatly reduced by an exogenous ethylene signal.^12,13

FVBP biosynthesis and emission is specific to a tissue type (corolla limb), a developmental phase (receptive to fertilization, open flower), and can be molecularly and metabolically “turned off” by one phytohormone (ethylene). The commonality of regulation imparted upon the FVBP gene network enables the isolation of additional FVBP genes by screening candidates based on similarity of transcriptional regulation. This concept has been reinforced as FVBP research migrated from biochemical steps directly producing volatile compounds (e.g., PHENYLACETALDEHYDE SYNTHASE (PhPAAS)),⁴ to the identification of more structural biochemical steps located within the shikimate, phenylpropanoid and FVBP branch pathways (Fig. 1). One example of this is the recent isolation and characterization of two MD transcript sequences, CHORISMATE MUTASE 1 and 2 (PhCM1 and PhCM2). PhCM1 and PhCM2 were isolated based on the hypothesis that key biosynthetic enzymes of the shikimate and/or the phenylpropanoid pathways are central players in the regulation of the overall metabolic flux of aromatic substrates available for direct FVBP synthesis.¹⁴ PhCM1 and PhCM2 are both functional chorismate mutase enzymes, but PhCM1 is differentially regulated at the transcript level compared to PhCM2. PhCM1 transcript accumulation is very comparable to authenticated FVBP genes, while PhCM2 transcript accumulation resembles a constitutively expressed gene. In support of the FVBP gene network paradigm, reverse genetic approaches demonstrated that of the CM genes, PhCM1 is pointedly involved in the production of FVBP compounds in MD flowers.¹⁴

Other recent advancements include: a 3-KETOACYL-COA THIOLASE (PhKAT1) that was identified and characterized to be involved in the core benzenoid pathway,¹⁵ which is one of the three main FVBP pathway branches off the phenylpropanoid pathway itself (Fig. 1). Also, AROGENATE DEYDRATASE 1 (PhADT1) another core biosynthetic gene, “downstream” of CM and “upstream” of PHENYLALANINE AMMONIA-LYASE (PhPAL), was isolated and characterized as specifically involved with FVBP production.¹⁶ Individually, PhADT1 and PhKAT1 are a part of gene families with multiple members, but their respective transcript accumulation follows the established FVBP gene network profile.¹³

FVBP Metabolic Network

Both PhCM1 and PhADT1 experimentation provided insights into the metabolic aspects of the core phenylpropanoid pathway. In petunia, when either PhCM1 or PhADT1 transcripts are reduced in transgenic plants, resulting in reduced levels of phenylalanine accumulation, the most severely reduced emitted FVBP compounds are derived directly from phenylalanine, the first FVBP branch (Fig. 1).^14,16 This observation has been postulated to be a difference in enzymatic rates of the PhPALs (lower Km for phenylalanine) and PhPAAS (higher Km for Phenylalanine),¹⁶ which is conceptually straight-forward. The flux must be greater through the core phenylpropanoid pathway than an axillary pathway to produce the relatively high levels of volatile compounds ultimately derived from t-cinnamic acid.¹³ In the same transgenic backgrounds, floral volatiles derived from t-cinnamic acid are reduced less than from the first FVBP branch and phenylpropanoid volatiles derived from the third metabolic branch, post p-coumaric acid, are affected the least of the three metabolic branches of FVBP biosynthesis. Therefore, under substrate limiting conditions, the most distal biochemical step, CINNAMATE-4-HYDROXYLASE (C4H), in the core phenylpropanoid pathway would have the largest “demand” for substrates.

FVBP Transcription Factors

FVBP emission is a mixture of volatile compounds at drastically different concentrations, which confers a very specific fragrance in the MD flower. If individual compounds are reduced in transgenic plants, human subjects can perceive a difference in floral fragrance compared to a MD flower.¹⁷ Using reverse genetics approaches and removing a biochemical step in the FVBP pathway (PhPAAS),⁴ or reducing total substrate flux to the entire pathway (PhODO1),⁷ has elucidated many biological factors contributing to the production of floral fragrance. However, different species, cultivars, and/or varieties of flowers may emit a comparable volatile profile in regard to identity of the chemical composition, but the levels of individual compounds can be drastically different.^13,18 Other than through selective mutations,¹⁹ how does the flower achieve such a precise floral volatile array?

In MD, the FVBP gene network is transcriptionally regulated in a concerted manner,¹³ resulting in a specific floral volatile array. These genes are not merely “turned on” at anthesis, at least for one of the core phenylpropanoid pathway genes. This gene is transcriptionally regulated in a non-absolute way resulting in a specific level of transcript accumulation conferring a specific level of downstream product. MYB4 is an R2R3-MYB transcription factor that negatively regulates C4H in plants. The Arabidopsis MYB4 indirectly controls the accumulation of UV protectants in leaf tissue.²⁰ In petunia, MYB4 indirectly controls the precise emission of the phenylpropanoids, eugenol and isoeugenol.²¹ PhMYB4 and PhC4H transcripts accumulate in a similar manner, and both resemble that of verified FVBP genes with the highest relative accumulation in corolla limb tissue of an open flower. It appears that PhMYB4 impedes the already elevated level of PhC4H transcription to achieve a tightly controlled number of products, and thus an indirectly controlled level of derived volatile products (Fig. 1). The negative contribution of PhMYB4 is necessary for the exact mixture of MD floral volatiles, yet it is not necessary for the production of a single or all of FVBP compounds. Therefore, PhMYB4 functions to “fine-tune” the bouquet of floral fragrance, while sharing in the strict transcriptional regulation imparted upon the rest of the FVBP gene network.

PhODO1 is speculated to act on multiple promoters of genes in the core biosynthetic pathways leading to the upregulation of substrate availability for FVBP production.⁷ More recently, another R2R3-MYB transcription factor was identified from P. hybrida line 720, and named EMISSION OF BENZENOID II (PhEOBII), because much like PhODO1 when PhEOBII transcripts were reduced a concomitant reduction in total emitted FVBPs was recorded.¹⁸ Of particular interest, empirical data indicates that PhODO1 can activate the 5-ENOL-PYRUVYLSHIKIMATE-3-PHOSPHATE SYNTHASE (EPSP) promoter, while PhEOBII can activate PhPAL and ISOEUGENOL SYNTHASSE 1 (PhIGS1) promoters, and PhMYB4 represses the PhC4H promoters (Fig. 1). In short, the FVBP model has been populated with a fair number of biosynthetic proteins over the last decade, but the identification and placement of regulatory proteins has been minimal. It seems a single transcription factor regulating the entire FVBP transcriptional status may not exist. In contrast, the research and evidence suggests a score of transcriptional regulators function collectively to generate the level of regulation observed within the petunia FVBP gene network.

Conclusions and Perspectives

Genes involved in the production of FVBP compounds are a part of a gene network based on the shared transcriptional regulation in MD. Metabolic regulation of the FVBP pathway appears to be shared with individual enzymatic steps and an overall elevation in pathway activity. Additionally, the level of pathway activity is also regulated. The field of FVBP research has yet to elucidate all molecular and metabolic features of the primary regulation imparted upon this system. It seems there are still fundamental aspects of the FVBP gene network that remain to be elucidated.