Overview and detectability of the genetic modifications in ornamental plants

The market of ornamental plants is extremely competitive, and for many species genetic engineering can be used to introduce original traits of high commercial interest. However, very few genetically modified (GM) ornamental varieties have reached the market so far. Indeed, the authorization process required for such plants has a strong impact on the profitability of the development of such products. Considering the numerous scientific studies using genetic modification on ornamental species of interest, a lot of transformed material has been produced, could be of commercial interest and could therefore be unintentionally released on the market. The unintentional use of GM petunia in breeding programs has indeed recently been observed. This review lists scientific publications using GM ornamental plants and tries to identify whether these plants could be detected by molecular biology tools commonly used by control laboratories.


Introduction
Ornamental plants are economically important for the horticultural industry 1 . They are sold all over the world, and used in gardening, landscaping, and floristry as cut flowers 2 . Genetic modification of ornamental plants has been used to introduce original traits of high commercial interest for the producers and/or consumers, such as improved floral anatomy and morphology, new floral color, induced early flowering, enhanced fragrance or longevity, stress tolerance or disease resistance [1][2][3][4][5][6] . At least 50 ornamental plants have been transformed, the main species including rose (Rosa hybrida), chrysanthemum (Chrysanthemum morifolium), petunia (Petunia hybrida), and carnation (Dianthus caryophyllus) 4 . However, very few genetically modified (GM) ornamental varieties have obtained the regulatory approval and have reached the market so far. Indeed, the authorization process required for such plants has a strong impact on the profitability of the development of such products. In the database of biotech/GM crops approval of the International Service for the Acquisition of Agri-biotech Applications (ISAAA), only three ornamental species are listed: petunia, rose, and carnation. Nineteen GM events have been registered in the ISAAA database for carnation, one GM event for petunia, and two GM events for rose. So far, only flower color-modified varieties of carnation and rose have been released on the market of some countries, depending on their regulation concerning production and/or commercialization of GMO 4,6 . In Europe, only two GM carnation varieties may currently be marketed as cut flowers. Genetic modification of ornamental plants has also commonly been used by research groups to study the genetics behind metabolic pathways, physiology, development, interactions with pathogens [7][8][9][10][11] . Numerous genetically transformed material of potential commercial interest have been produced. If by any mean, the traceability of such material was lost, it could be considered as non-GM and introduced in breeding programs.
In 2017, transgenic petunia plants were detected on the market in Europe and in the United States 6 . The detected GM events contained the A1 gene from Zea mays L. encoding the enzyme dihydroflavonol reductase, which was first introduced into a mutant petunia defective for this gene 12 . The first transformed plants had been obtained to study the biochemistry and genetics of the flavonoid metabolic pathway. These plants also had the particularity of producing flowers of an orange color not previously seen in the genus 12 . Further studies showed that differences in the methylation of the promoters occurring in the progeny of these plants conducted to the attenuation of the orange phenotype 13,14 . Such plants have never been through the authorization process required in the European Union and should not have been commercialized. Petunia plants containing the A1 gene have nevertheless been found on the market. Furthermore, the detected plants showed several phenotypes (unpublished data), suggesting that the transgenic material has been used in breeding programs for several years. In respect to the absence of approval for marketing or cultivation in Europe, numerous GM petunia plants had to be withdrawn from the European market in 2017.
The construct inserted in the genome of the detected GM petunia contained the A1 gene but also a 35S promoter (p35S) derived from the Cauliflower mosaic virus (CaMV). This promoter, with the nopaline synthase terminator from Agrobacterium tumefaciens (tNos), and to a lesser extent sequences such as the phosphinothricin acetyltransferase gene from Streptomyces hygroscopicus (bar) or the 5-enolpyruvylshikimate-3-phosphate synthase gene from Agrobacterium tumefaciens sp. strain CP4 (ctp2-cp4epsps), have been recurrently introduced in constructs used for genetic modification of plants. To test whether a sample may or not contain GM plant material, laboratories use these sequences as screening elements 15 . Diagnostic is mostly based on real-time PCR detection targeting the sequences of interest (p35s, tNos…) 16 . The recent detection of transformed unauthorized petunia developed for research purpose in the 80's demonstrates the risk that represents the potential nonintentional use of research plant material in breeding programs, especially when the transgene provides a trait of commercial interest. The aim of the present review was to assess the kind of GM ornamentals, which have been developed for research purpose and could therefore be found on the market if used involuntarily in breeding programs. Scientific publications related to research programs involving the transformation of ornamental species and for which the introduced trait could confer a commercial advantage compared with varieties on the market were listed. In addition, we investigated if such unauthorized material would be detectable by laboratories testing for the presence of GM plants.

Biotic stress resistance
Biotic stress occurs in plants due to damage done by plant pathogens or pests (bacteria, fungi, virus, insects, parasites, nematodes, aphids) during production, storage, distribution, and end-consumer use. Plant growth and yield are strongly influenced by biotic stress. Several ornamental species have been transformed to introduce or enhance resistance to pathogens or pests. Various genes coding for toxins, protease inhibitors, trypsin inhibitors, lectins, etc. have been employed to create pest resistance in plants 5 . As an example, increased resistance to aphids was observed in transgenic chrysanthemum overexpressing transcription factor CmWRKY48 17 or the protease inhibitor Sea Anemone Equistatin (SAE) 18 . Expression of a cystatin transgene in Lilium longiflorum conferred resistance to the root lesion nematode Pratylenchus penetrans 19 . Resistance against bacterial and fungal diseases has been created by transferring various genes coding for chitinase, glucanase, osmotin, defensin, etc. into various ornamental plants 5 . Transgenic gladiolus expressing a synthetic antimicrobial peptide, D4E1, had an increased resistance to the soilborne fungus Fusarium oxysporum f. sp. gladioli 20 . The antimicrobial peptide tachyplesin I (tpnI) expression reduced bacterial proliferation, colonization, and disease symptoms (reduced by 95-100%) in transgenic ornithogalum plant tissues 21 . Virus resistance has been generally introduced by transferring coat protein gene of the virus in the plant. As an example, resistance to cucumber mosaic virus (CMV) was enhanced when a defective CMV replicase gene was transferred to the lilium 8 . Disease resistance in ornamental plants is greatly expected as disease symptoms on plants or cut flowers make the product unmarketable or unacceptable for export 4 .

Abiotic stress resistance
Production of ornamental plants can be affected by abiotic stresses, such as drought, salinity, cold, heat, herbicides, etc. These stresses can affect plant growth and productivity as they generally damage the cellular machinery by inducing physiological, biochemical, and molecular changes in the plants 5 . Research on GM ornamentals has been conducted to improve abiotic stress resistance. Genes targeted for genetic transformation code for stress-protecting compounds have been classified into three categories: (i) genes coding the synthesis of various osmolytes such as mannitol, glycine betain, proline, and heat shock proteins, (ii) genes responsible for ion and water uptake and transport like aquaporins and ion transporter, etc., and (iii) genes regulating transcriptional controls and signal transduction mechanism, such as MAPK, DREBI, etc. 5 . As an example, overexpression of AtDREB1A gene increased drought and salt stress tolerance 22 , as does the overexpression of ClCBF1 23 in transgenic chrysanthemum. The overexpression of the Chrysanthemum crassum plasma membrane Na + /H + antiporter gene CcSOS1 improved the salinity tolerance of chrysanthemum plants 7 . A construct carrying both CcSOS1 (from Chrysanthemum crassum) and CdICE1 (from Chrysanthemum dichrum) was constitutively expressed in the chrysanthemum variety "Jinba" 24 . Improved sensitivity to low temperature, drought, and salinity was observed in transgenic plants as measured by visible damage and plant survival 24 . The MtDREB1C gene, isolated from Medicago truncatula, enhanced freezing tolerance in transgenic China rose without morphological or developmental abnormality 25 . Resistance to herbicide was obtained in transgenic Lilium longiflorum expressing the bialaphos-resistance gene PAT under the constitutive CaMV35S promoter 26 .

Flower color modification
Flower color is one of the most important traits of ornamental plants influencing its commercial value. Flower color can also attract pollinators and protect floral organs 27 . Several ornamental plants have been engineered for flower color modifications by targeting flavonoids, anthocyanins, carotenoids, and betalains 28 . To create new flower color, biosynthetic pathways can be modified through the introduction of new genes, overexpression, or silencing of target genes 5 . The first ornamental plant genetically modified for flower color was an orange pelargonidin-producing petunia variety 12 . This was achieved by the expression of the A1 gene from Zea mays L. encoding the enzyme dihydroflavonol reductase in a petunia plant defective for this gene 12 . Other examples concerned roses, carnations, and chrysanthemum which lack violet/blue varieties, due to the absence of delphinidin-based anthocyanins 29 . This is attributed to their deficiency of flavonoid 3′,5′-hydroxylase (F3′5′H), a key enzyme in the synthesis of delphinidin 29 . Accumulation of delphinidin was achieved in roses or chrysanthemum by introducing the F3′5′H gene, turning the flower color purple or violet [29][30][31] . Blue chrysanthemum was produced by introducing the A3′5′GT gene encoding anthocyanin 3′,5′-O-glucosyltransferase, in addition to F3′5′H, into the host plant 31 .

Early flowering
Flowering time is also an important trait of ornamental plants which can been engineered. Several reports have described successful gene introduction to produce flowers in comparatively short time, allowing the production of flowers at a lower cost 32 . MADS box genes constitute an example as they can control flowering time and floral organ development 3 . Overexpression of AP1-like genes, member of MADS box gene family, from Asteraceae induced early-flowering in transgenic chrysanthemum plants 32 . Transgenic Dendrobium orchids overexpressing DOAP1, an AP1 ortholog, displayed earlier flowering and earlier termination of inflorescence meristems into floral meristems than wild-type orchids 33 . Transgenic lisianthus plants, transformed with the MADS box gene OMADS1 from orchid flowered significantly earlier than nontransgenic plants 34 . Similarly, overexpression of MADS box genes, like DOSOC1 or OMADS1, promoted early flowering in transgenic orchids 35,36 .

Morphology modification
Plant morphology can be engineered to provide economic advantages during commercial production of ornamental plants. Indeed, the manipulation of plant size and development can produce plants with uniform morphology, and reduce the use of chemical growth regulators thereby reducing production costs 37 . Genetic engineering of homeotic, Agrobacterium, phytochrome, and gibberellin genes has provided potential targets to modify and control plant growth and development 38 . A dwarf chrysanthemum was produced through heterologous expression of the mutant Arabidopsis gai (gibberellic acid insensitive) gene driven from its own promoter by decreasing gibberellin response 39 . Ectopic expression of a tobacco (Nicotiana tabacum L.) phytochrome B1 gene in chrysanthenum modified plant architecture; GM plants were shorter with larger branch angles than wild-type plants 37 . A miniature chrysanthemum was produced through RNAi expression vector targeting both DmCPD and DmGA20ox, which are related to brassinosteroids (BR) and gibberellins (GA) biosynthesis, respectively 40 . Production of compact plants of Kalanchoe was achieved by overexpression of homeotic genes KxhKN4 and KxhKN5 41 . The development of non-branching plants can reduce manual labor requirements during cultivation, decreasing the cost of flower production 42 . Transgenic gypsophila, carnation, or rosa exhibiting improved rooting were generated by expressing Agrobacterium rhizogenes ROL genes 38,43,44 . Enhancement of branching phenotype was observed in chrysanthemum by expression of the cytokinin biosynthesis gene ipt (isopentenyl transferase) 45 .

Flower anatomy modification
Ornamental plants have been engineered to create new flower shapes that can increase their ornamental value. Flower formation involves the development of sepals, petals, stamens, and pistils, regulated by several genes involved in flower organ identity 34 . The overexpression of CmTCP20, a member of teosinte branched1/cycloidea/proliferating cell factors (TCPs) gene family, led to larger flower inflorescences and longer petals in chrysanthemum 46 . Overexpression of CmCYC2c in Chrysanthemum lavandulifolium led to significant increase in flower numbers and petal ligule length of ray florets 47 . Alteration of flower transition and formation was observed in transgenic Eustoma grandiflorum plants ectopically expressing the MADS box gene LMADS1-M from lily (Lilium longiflorum) 34 . Other studies have reported the modification of flower traits in transgenic torenia using chimeric repressors of Arabidopsis AGSRDX 48 or TCP3 49 .

Perfume modification
Floral scent volatiles are implicated in the evolutionary success of many plants as they attract pollinators and seed dispersers. Most fragrance compounds belong to three major groups: terpenoid, phenylpropanoid/benzenoid, and aromatic amino acid 3 . Most modern cut-flowers such as roses and carnations lack fragrance, probably because of the selection for other traits by breeders. Genetic engineering of ornamental plants could allow to transfer fragrance from one species to another, to induce new fragrance or enhance poor fragrance 4 . Fragrance in petals of transgenic lisianthus (Eustoma grandiflorum) was induced by overexpressing the Clarkia breweri BEAT gene (benzyl alcohol acetyl transferase) and fedding the transformed plants with alcoholic substrates 50 . Carnation plants were transformed with the linalool synthase (lis) gene from Clarkia breweri leading to the production of linalool and its derivatives in the transgenic plants 51 . In another study, an increase in the production of volatile phenylpropanoid/benzenoid compounds was observed in transgenic petunia flowers transformed with the Production of Anthocyanin Pigment1 (Pap1) Myb transcription factor from Arabidopsis thaliana 43 . Similarly, Pap1transgenic rosa flowers showed increased levels of volatile phenylpropanoid and terpenoid compounds when compared with control flowers 52 .

Longevity
Yellowing of leaves in ornamental plants, associated with senescence, have a negative attribute decreasing plant attractiveness, quality, and vase life 3 . In cut flowers, long vase life is a critical trait as they must survive several weeks before reaching the market 4 . Ethylene is a plant hormone, which plays a major role in the senescence process of plants 53 . Resistance to ethylene or inhibition of ethylene biosynthesis genes can increase shelf life 3 . Several ornamental plants have been transformed using genes encoding key enzymes of the ethylene pathway in order to inhibit senescence 53 . Expression of a mutated ethylene receptor gene etr1-1 from A. thaliana reduced ethylene sensibility in several ornamental plants, including orchid species 54,55 , Campanula carpatica 56,57 , chrysanthemum 58 , carnation 53 , Pelargonium zonale 59 , and Kalanchoe blossfeldiana 56,60 . Similarly, expression of a mutated ethylene receptor gene (mDG-ERS1s) reduced ethylene sensitivity in modified chrysanthemum 61 . Transgenic carnation plants containing an antisense ACC oxidase gene exhibited low ethylene production and delayed petal senescence 62 . An alternative approach to delay senescence in plants is to promote an increase in cytokinin levels 59 . Overproduction of cytokinins in petunia flowers transformed with PSAG12-IPT, enzyme which catalyzes the first step of the cytokinins biosynthesis pathway, delayed flower senescence, and decreased sensitivity to ethylene 9 . Delayed leaf senescence and enhanced resistance to exogenous ethylene were observed in miniature rose transformed with the same gene 63 .

Detection of GM ornamentals
When available, information on the transgene has been taken into account to provide an indication of the detectability of the events using common molecular biology tools (Table 1). Among the constructs, at least 83.0% present one of the screening sequences p35S or tNos commonly targeted in GMO detection strategies and 35.8% present both. If tested, these plants would therefore be identified as genetically modified.

Conclusion
The aim of this review was to list scientific publications related to GM ornamental plants developed by research groups and potentially presenting a commercial interest and to investigate if such unauthorized material would be detectable by laboratories testing for the presence of GM plants if unintentionally released on the marketplace. The goal was not to be exhaustive but to be as representative as possible of the constructs developed for research purpose. With 166 publications considered, it provides an overview of the constructs which could unintentionally reach the market without preliminary approval. The first conclusion of this review concerned the number of transformed ornamental species and the number of publications found per species. If many species have been transformed, two species of strong interest, chrysanthemum, and petunia, represent >40% of the collected data. With no consideration of any health, or environmental risk, of trade or commercial importance, these two species represent higher risks of unintentional use simply by the number of constructs available. The second outcome of the review is the ability of laboratories to detect at least 83.0% of the identified constructs.
This review also outlined the diversity of GM traits of potential commercial interest for the producers and/or consumers studied. Color modification dominates the GM ornamental research, however, since the first transgenic color-modified carnation variety released 20 years ago, very few GM ornamental varieties have been field tested and have obtained the regulatory approval. Indeed, with the exception of color-modified varieties of carnation and rose, no transgenic varieties of flowers have reached the market so far. The cost and time required to obtain the regulatory approval are the major factors limiting the commercialization of such plants and the profitability of such developments.
Whole-genome sequencing of ornamental plants and the development of new molecular markers allow the identification of new genes of interest and related pathways 3 . Considering the ease of use of the new genome editing tools 64 and especially of the CRISPR/Cas9 system 65 , more and more transformed material will be available. To our knowledge, there are five scientific publications related to successful genome editing of ornamental plants with trait of commercial interest, including ipomoea [66][67][68] , petunia 69 , and torenia 70 . CRISPR/Cas9-mediated mutagenesis of the EPHEMERAL1 locus delayed petal senescence in transformed ipomoea 66 . Flower color changes was induced using CRISPR/Cas9 technology 67,68,70 . As an example, ipomoea with pale-yellow petals was produced for the first time by successful knockout of InCCD4, involved in carotenoid degradation, using CRISP/Cas9 system 68 . Recently, the Court of Justice of the European Union (CJEU judgment of 25 July 2018, C-528/16) clarified that organisms obtained by new breeding techniques (NBT) are GMOs and fall within the scope of the EU GMO legislation. Meanwhile, several countries exporting crops to the EU (e.g., USA and Argentina) decided to not regulate such organisms as GMOs. Soon, it is probable that numerous varieties obtained by NBT will be available on the market outside of the EU, without any notification procedures in most countries. Accidental imports might occur between countries having different regulatory practices and policies regarding NBTs and environmental risk contamination should be considered.
Furthermore, groups will probably reduce the use of constructs including easily detectable promoters and terminators. The detectability of the produced transformed events will probably decrease in the near future; however, these considerations are not specific to ornamental plants.