Genome sequence of the ornamental plant Digitalis purpurea reveals the molecular basis of flower color and morphology variation

This study presents a high-quality long-read genome sequence of *Digitalis purpurea* that elucidates the molecular mechanisms underlying its flower color variation, caused by a disruptive insertion in the anthocyanidin synthase gene, and its terminal flower morphology, driven by a large insertion in the *DpTFL1/CEN* gene.

The Gist

Imagine a plant called Foxglove (Digitalis purpurea). You might know it as the tall, spiky flower with bell-shaped blooms that look like they're wearing purple dresses. For centuries, humans have used this plant as medicine for heart problems, but it's also famous for its beauty.

However, nature loves variety. Some Foxgloves are a deep, vibrant magenta with dark spots, while others are stark white. Some grow tall spikes of flowers, but occasionally, a single giant flower pops up right at the very top, stopping the spike from growing any higher.

This paper is like a detective story where scientists finally got the "instruction manual" (the genome) for the Foxglove plant to figure out exactly why these differences happen.

Here is the breakdown of their discovery, using some simple analogies:

1. The Big Instruction Manual (The Genome)

Think of the Foxglove's DNA as a massive library containing every instruction needed to build the plant. Until now, this library was messy and hard to read. The scientists used a new, high-tech "scanner" (called Nanopore sequencing) to read the long, continuous pages of this library without losing any pieces.

• The Result: They built a clean, complete map of the Foxglove genome. It's like turning a pile of shredded puzzle pieces into a clear, high-definition picture. This map is so good that it's 96% complete, meaning they found almost every single instruction book the plant needs.

2. The Mystery of the White Flowers (The Broken Paint Mixer)

Foxgloves get their purple color from a special paint called anthocyanin. To make this paint, the plant has a factory line with several workers (enzymes) passing a bucket of ingredients down the line.

• The Magenta Plant: Has a fully functional factory. The workers pass the bucket all the way to the end, and the final product is purple paint.
• The White Plant: The scientists found that in white flowers, one specific worker is missing. This worker is called ANS (Anthocyanidin Synthase).

The Twist: In the white flowers, a giant piece of junk DNA (a "transposable element," or a genetic parasite) had jumped right into the middle of the ANS worker's station.

• The Analogy: Imagine a construction crew building a wall. Suddenly, a massive boulder falls right in the middle of the path. The workers can't get past it, so the wall (the purple pigment) never gets finished. The result? A white flower.
• The Evidence: The scientists checked 89 different plants. Almost all the white ones had this "boulder" blocking the path, and almost all the purple ones had a clear path. It's the smoking gun!

3. The Mystery of the Giant Top Flower (The Broken Stop Sign)

Normally, a Foxglove spike grows upward, adding small flowers one by one, like beads on a string. It keeps growing until the season ends. This is called an "indeterminate" spike.

But sometimes, the plant stops growing and puts one giant flower right at the tip. This is called a "terminal flower."

• The Culprit: The scientists found a gene called TFL1/CEN. Think of this gene as a Stop Sign or a Traffic Light at the top of the plant.
  • Normal Plant: The Stop Sign is working. It tells the top of the plant, "Keep growing, don't make a flower yet." So, the plant keeps adding more small flowers.
  • Giant Flower Plant: A giant boulder (another piece of jumping DNA) crashed into the Stop Sign gene, breaking it.
  • The Result: Without the Stop Sign, the plant's "traffic light" turns green for flowers immediately. The top of the spike thinks, "Oh, I'm done growing! Let's make a flower!" and turns into a giant bloom.

4. The Spots on the Petals

If you look closely at a magenta Foxglove, you'll see dark purple spots on the bottom petals. These act like landing strips for bees and other pollinators, guiding them to the nectar.

The scientists found that the plant uses a special "spotlight" system. One set of genes turns on the purple paint in the spots, while another set keeps the rest of the petal a lighter color. It's like a stage manager using spotlights to highlight specific actors (the spots) while keeping the background dim.

Why Does This Matter?

• For Gardeners: Now we know exactly what genetic "switches" control flower color and shape. Breeders could potentially use this knowledge to create new, unique Foxgloves.
• For Science: This is the first time we've had a complete map of the Foxglove genome. It's a foundation for studying how plants make medicines (since Foxgloves are used for heart drugs) and how they evolve.
• The Big Picture: It shows how a single random event—like a piece of DNA jumping into the wrong place—can completely change how a plant looks, turning a purple flower white or a tall spike into a giant top flower.

In short: The scientists opened the Foxglove's instruction manual, found a broken "paint mixer" that causes white flowers, and a broken "stop sign" that causes giant top flowers. It's a perfect example of how tiny changes in our genetic code can lead to big changes in the world around us.

Technical Summary

1. Problem Statement

Digitalis purpurea (foxglove) is a widely distributed ornamental and medicinal plant known for its cardiac glycosides (e.g., digoxin). Despite its economic and ecological importance, the plant lacks a high-quality reference genome. Furthermore, the molecular mechanisms underlying two distinct phenotypic variations in D. purpurea remain uncharacterized:

1. Flower Pigmentation: While wild-type plants display magenta flowers with dark purple spots, white-flowering variants exist. The genetic basis for the complete loss of anthocyanin pigmentation in white flowers is unknown.
2. Inflorescence Architecture: Most D. purpurea plants have indeterminate spikes, but rare individuals develop large terminal flowers (determinate architecture). The genetic cause of this morphological shift is unexplored.

2. Methodology

The study employed a combination of long-read sequencing, de novo assembly, comparative genomics, and population genotyping.

• Plant Material & Sequencing:

  • Samples included a magenta-flowering plant (DR1, used for the reference), a second magenta plant (DR2), and two white-flowering plants (DW1, DW2).
  • High-molecular-weight DNA was extracted and sequenced using Oxford Nanopore Technologies (ONT) MinION (R9.4.1 flow cells).
  • RNA-seq data from various tissues (petals, spots, roots) were retrieved from the SRA to assist in gene annotation.

• Genome Assembly & Annotation:

  • Three assemblers (NextDenovo, Shasta, Flye) were tested. NextDenovo was selected as the best assembly based on continuity (N50) and completeness.
  • The assembly was polished using NextPolish.
  • Structural Annotation: Gene models were predicted using GeMoMa, leveraging RNA-seq hints and homologous proteins from related species (Salvia, Rehmannia, Penstemon).
  • Functional Annotation: Orthologs were identified via BLAST against Arabidopsis thaliana. Specific tools (KIPEs, MYB_annotator, bHLH_annotator) were used to annotate flavonoid biosynthesis genes and transcription factors.
  • Repeat Analysis: Transposable elements (TEs) were annotated using RepeatModeler2, RepeatMasker, and EDTA.

• Comparative Analysis & Variant Detection:

  • Reads from white and magenta plants were mapped to the reference genome to identify structural variants (SVs).
  • IGV (Integrative Genomics Viewer) was used for manual inspection of candidate genes in the anthocyanin pathway (CHS, CHI, F3H, DFR, ANS, arGST, 3GT) and the flowering regulator TFL1/CEN.
  • Ks Analysis: Synonymous substitution rates were calculated to detect Whole Genome Duplication (WGD) events.

• Validation:

  • A population of 89 plants was genotyped via PCR targeting the identified mutations in the ANS and TFL1/CEN loci to correlate genotype with phenotype.

3. Key Contributions

• First High-Quality Reference Genome: The study presents the first long-read-based de novo genome assembly for Digitalis purpurea (DR1_v1).
  • Stats: 940 Mbp assembly size, N50 of 4.3 Mbp, and 96.3% BUSCO completeness.
  • Annotation: 35,916 protein-coding genes identified.
• Discovery of Pigmentation Mechanism: Identification of a large transposable element (TE) insertion in the Anthocyanidin Synthase (ANS) gene as the cause of white flowers.
• Discovery of Morphology Mechanism: Identification of a large TE insertion in the DpTFL1/CEN gene (ortholog of Arabidopsis TFL1) responsible for the determinate (terminal flower) phenotype.
• Regulatory Insights: Characterization of the DpMYB75 gene family, revealing distinct paralogs (DpMYB75a vs. DpMYB75b) that likely regulate background pigmentation versus spot pigmentation, respectively.

4. Key Results

A. Genome Characteristics

• The genome contains ~73.8% repetitive sequences, dominated by LTR retrotransposons (62.4%).
• Heterozygosity is estimated at 0.7%.
• Ks analysis indicates a recent Whole Genome Duplication (WGD) event (peak at Ks = 0.21), explaining the high proportion of duplicated BUSCO genes (55.8%).

B. Molecular Basis of White Flowers

• The Mutation: A ~14.2 kb insertion was found in the ANS gene of white-flowering plants.
• Nature of Insertion: The insertion contains motifs of a retrotransposon (Reverse Transcriptase, GAG, Zinc Knuckle domains) but lacks functional RNase H and Protease domains, suggesting it is a degenerated, non-functional TE.
• Mechanism: The insertion disrupts the ANS open reading frame (ORF), rendering the enzyme non-functional. Since ANS is essential for converting dihydroflavonols to anthocyanidins, its loss blocks the pathway, resulting in white flowers.
• Evolutionary Timing: LTR divergence suggests the insertion occurred approximately 880,000 years ago, predating human horticultural efforts.
• Genotyping: In a population of 61 flowering plants, 93.4% showed a perfect genotype-phenotype correlation (homozygous mutant = white; heterozygous/wildtype = magenta).

C. Molecular Basis of Terminal Flowers

• The Mutation: A ~14.3 kb insertion was identified in the DpTFL1/CEN gene, located just 4 bp downstream of the start codon.
• Mechanism: This insertion causes a premature stop codon, leading to a loss-of-function mutation. Since TFL1/CEN acts as a flowering repressor to maintain indeterminate growth, its loss results in the conversion of the inflorescence apex into a flower (determinate growth).
• Genotyping: In 24 characterized plants, 23 showed consistent genotype-phenotype correlation. Most terminal-flower plants were homozygous for the mutation.

D. Pigmentation Regulation

• Gene expression analysis revealed that DpMYB75b is highly active specifically in the dark purple spots, while DpMYB75a is active in the surrounding petal tissue. This suggests a dual regulatory system controlling spot formation vs. general petal coloration.

5. Significance

• Resource for Plant Biotechnology: The high-quality genome provides a foundational resource for breeding programs aimed at manipulating flower color and architecture in Digitalis and related Plantaginaceae species.
• Insight into Specialized Metabolism: The study elucidates the specific genetic lesion (ANS disruption) responsible for anthocyanin loss, a common but often complex trait in ornamental plants.
• Evolutionary Conservation: The findings confirm the conservation of the TFL1/CEN mechanism in controlling inflorescence determinacy across diverse plant families (from Arabidopsis to Digitalis).
• Transposable Elements: The study highlights the role of ancient TE insertions as drivers of phenotypic diversity and domestication traits, even when the TEs themselves are degenerated.

Limitations: The study was limited by a small number of sequenced individuals (n=4) for the initial discovery, and minor genotype-phenotype mismatches in the genotyping population suggest potential environmental influences or additional genetic modifiers.