The Purple Tomato Story; From Laboratory Bench to the Consumer

Abstract

A world apart from academic research, the path from developing a polyphenol-rich crop to a product available to consumers is not one taken by many research scientists. Here we review the steps taken to commercialize anthocyanin-enriched purple tomatoes in the USA. In describing some of the difficulties encountered and the work that was necessary for a successful commercial launch of a new biotech product, we hope to encourage others to believe that there is a viable route to market, and an appetite for polyphenol-enriched foods that can protect health.

Figure 1

The objective for creating purple tomatoes. In normal red tomatoes, the transcriptional complex that regulates anthocyanin biosynthesis is switched off; no anthocyanins are made in fruit which are red due to their high content of the carotenoid, lycopene. Introduction of two of transcription factors (Rosea1 and Delila) with expression specifically in fruit causes induction of anthocyanin biosynthesis (particularly delphinidin) and the development of purple fruit. Health benefits can be assessed by comparing health status of disease models on diets containing red and purple isogenic foods.

Figure 2

Schematic representation of binary vector pDEL.ROS used to produce purple tomatoes. a) Construction of pDEL.ROS in vector pRK290 which carries a tetracyclin resistance gene (tet) for selection in bacteria. b) Detail of gene arrangements in pDEL.ROS including promoters (NOSp and E8p) 3′sequences (Ocs 3, CMV), genes encoding transcription factors (DEL, ROS) and nptII (NPT) gene encoding kanamycin resistance. Plasmid construction: The 2175bp E8 promoter was amplified from the tomato genomic DNA by PCR. The promoter was cloned in pJIT60 to replace the CaMV 35S promoter and in pJAM1500, (pJIT60 containing a Gateway Destination Cassette; Invitrogen) such that E8 replaced the 35S promoter. This resulted in plasmids pE8.60 and pE8.1500 respectively. The region containing E8-Gateway-CaMV 3′ sequence from pE8.1500 was cloned in pSLJ7291 resulting in plasmid pSLJ.E8.1500. The full-length Delila cDNA was amplified by PCR and inserted in this plasmid using Gateway recombination technology, resulting in the binary construct pSLJ.E8.DEL. The full-length Rosea1 cDNA was amplified by PCR and inserted in the plasmid pE8.60, resulting in plasmid pE8.ROS. After the introduction of a double stranded oligonucleotide containing a SalI restriction site in this plasmid, the region containing E8-Rosea1 cDNA-CaMV3′ sequence was cloned as a SalI-XhoI fragment into the XhoI site of pSLJ.E8.DEL resulting in the binary construct pDEL.ROS.

Figure 3

Analysis of genome sequence of tomato plant carrying the purple tomato transformation event (a) Results of search of purple tomato genome sequence for sequences of 150 nucleotides or greater with identity to pRK290 sequence taken from CLD04541: nucleotides 5854–25385 deposited in NCBI as AF184978.1. (https://www.ncbi.nlm.nih.gov/nuccore/AF184978.1). The purple tomato genome sequence has been deposited in ENA accession number ERR3500875. No reads aligned to this sequence with 100% identity over 150 nucleotides as shown in the alignment deposited in ENA under accession number ERR3502329. (b) Alignment of reads from Illumina hi-seq over the sequence of the Rosea1 (Ros1) gene in the sequenced purple tomato plant. Aligned reads are shown as gray bars. Dark lines within gray bars show sequence mis-matches to the sequence of pDEL.ROS in individual reads. The dark blue profile shows the degree of read coverage for each part of the gene sequence. This alignment has been deposited in ENA as accession number ERR3502327. (c) Alignment of reads from Illumina hi-seq over the sequence of the Delila (Del) gene in the sequenced purple tomato plant. Aligned reads shown as gray bars. Dark lines within gray bars show sequence mis-matches to the sequence of pDEL.ROS in individual reads. The dark blue profile shows the degree of read coverage for each part of the gene sequence. This alignment has been deposited in ENA under accession number ERR3502327.

Figure 4

Fruit-specific phenotypes of purple tomatoes expressing both Del and Ros1 under the control of E8 promoter. (a) Purple and wild-type (red) tomatoes (b) Comparative analysis of phenylpropanoid content and composition in purple vs wild type tomato fruit. HPLC chromatogram of methanol extracts from the purple line and wild- type (red line) tomato fruit. HPLC analysis, recorded at 535 nm of extracts from peel or flesh of ripe fruit. Peaks marked with numbers represent anthocyanins. Classification and identification of methanol soluble compounds was performed based on PDA absorbance and ESI-Q-TOF mass spectrometry. The purified compounds were analyzed by HPLC and ESI-MS/MS. Spectral characteristics, molecular ions and fragments obtained are tabulated. Identification was confirmed by hydrolysis and HPLC analysis of the respective acyl and sugar moieties.

Figure 5

Purple fruit of different varieties. Fruit were photographed in the F2 generation from crosses between purple tomato in the Moneymaker background and Ailsa Craig, Ohio 8423, Goldkrone, Maglia Rosa, and Lucinda.

Figure 6

Purple fruit in the Goldkrone genetic background of tomato. Fruit from different lines segregating in the F3 from a cross between purple tomato in the Moneymaker genetic background and the yellow tomato variety Goldkrone. Lines producing smaller fruit were selected for home growers, those producing larger fruit were selected for sales as “snacking tomatoes” and “limited edition” products for supermarkets. The anthocyanin content of the different lines is shown below in mg cyanidin-3-glucoside (C-3-G) equivalents per 100 g fresh weight of fruit.

Figure 7

Workflow for RSR review of biotechnology traits in crops (courtesy of USDA/APHIS).