top of page

How Flowering Works – With A Little Dash of Biochemistry

Why is flowering important?

As many of you are probably aware, flowering is essential for the survival and propagation of a variety of plant species. This process happens within the division of Angiosperms (meaning “flowering plants”), which make up the majority of plants on Earth. This is likely because flowering brings a number of advantages to the plant:

1. Cross-pollination

It allows plants to be pollinated by other plants that are of the same species (cross-pollination). As a result, this lets them develop a more diversified genetic code within their population, making them more resistant to pathogens.

2. Endozoochory – fruit (seed) dispersion

Pollination allows the development of seeds, which form within fruits in flowering plants. These fruits are often appealing to animals and in certain cases, the ingestion of the fruits by animals allows its seeds to be dispersed to more distant places, which therefore helps the plants’ propagation. This strategy of dispersing seeds is called endozoochory.

Of course, flowering is also very important to animals who depend on the fruit or other parts of the plant as their diet. In fact, understanding this process is important for humans as well - as it allows us to predict or even control when fruits and seeds are produced. Meanwhile, this knowledge is not only important for agriculture, but also for industries such as floriculture and apiculture.

To develop a better understanding of flowering, we are going to analyse it in a bit more detail through some basic biochemistry lens. With that in mind, this article will focus on the following topics:

  1. What are the different components of a flower?

  2. How do different types of flowers exist?

  3. Timing - what determines when flowering occurs?

  4. Why is research in flowering important?

The components of a flower

Angiosperms, which is the group containing all flowering plants, is very diverse and contains around 300,000 species. As such, there are a lot of varieties in flower structures, though there are some organs that tend to be present in most species. For instance, male and female reproductive organs are typically both present in a flower.

The male part is called stamen, which is shown in blue. As you can see, it consists of the anther (which contains the pollen grains) and the filament (which supports the anther so that it is easily reachable by pollinators).

Meanwhile, the female part is called the pistil or carpel, shown in pink in the diagram below, and it has three components. The stigma is the tip of the pistil and it catches the pollen. Then, the pollen travels through a long tube called the style until it reaches the ovary, which contains female reproductive cells, known as the ovules. Besides that, the elongated structure of the style also allows the flower to catch more pollen.

In the meantime, if you want to learn more about the different floral organs, do visit this website.

Figure 1: A diagram showing the cross section of a flower and highlighting the reproductive organs. This diagram was modified from a figure created by Dianaperezval.

In essence, the structure of flowers allows both the processes of self-pollination and cross-pollination. As suggested by the name, self-pollination occurs when a plant’s pollen fertilises a flower from the same plant, whereas cross-pollination occurs when the pollen fertilises a flower from a different plant of the same species.

How do different types of flowers exist?

Introducing the ABC model

To explain the different structures of various flowers, two researchers, Enrico Coen and Elliot Meyerowitz, came up with the ABC gene model in 1991.

To keep things simple, the ABC model is a way to explain how the differential expression of certain genes led to the formation of certain flowering components. Essentially, they broadly divided the genes into three different groups (or classes as others may prefer to call it) called A, B, and C, each of which has different functions and produced different phenotypes upon expression.

The diagram below shows the classical ABC model:

Figure 2: The ABC model shows which class of genes contribute to sepals, petals, stamens, and carpels, all of which are different components of a plant. The sole expression of A genes results in the formation of sepals, whilst the sole expression of C genes leads to the generation of carpels. Meanwhile, the production of certain components of the plant requires the co-expression (i.e. the expression of two genes simultaneously) of two genes. For instance, petals need the co-expression of A and B genes, whilst the expression of both B and C genes are required to establish stamens. This diagram is modified from a figure created by Laura Aškelovičiūtė.

Some flower structures are determined by genes that belong to a single group, such as sepal and carpel, which are produced by A and C genes, respectively. However, this is not the case for petals and stamens, whereby petals are determined by A and B genes, and stamens are determined by B and C genes.

One great way of finding out what certain genes do is to mutate them - because then you can compare the mutated plant with the original and see what has changed. When genes from classes A, B, or C get mutated, it is possible to produce mutant flowers that lack one or more of the structures that are normally present in flowers. Such examples are shown by the following figure:

Figure 3: Mutated versions of the Arabidopsis thaliana plant in the class A, B and C genes, respectively. Mutants in class A genes show expression of stamen and carpels but a lack of sepals and petals. Meanwhile, mutants in class B genes express sepals and a carpel, but no stamens or petals. Furthermore, mutants in class C genes express petals and sepals, but not any carpel nor stamens.

As you may have noticed, all the flowers above either lack one or more of the four structures shown in the diagram or are positioned oddly.

Besides having mutations, the great diversity of flowers can be explained by the variations of the classical ABC model. One good example of this is the case of tulips, where instead of petals and sepals, it has a structure known as tepals:

Figure 4: A comparison between the appearance and ABC models of A. thaliana (left) and tulip (right) flowers. As you may have noticed, unlike A. thaliana, tulips lack sepals. This can be explained by an outward shift of the B class genes in the tulip’s ABC model, resulting in petal-like organs known as tepals (with no obvious sepals).

Introducing the CYCLOIDEA gene

Beyond the numerous variations of the classical ABC model, there is also a gene that is responsible for the symmetry of the flower.

Ancestral flowers tend to have radial symmetry, meaning that they have a symmetry about a central axis (e.g. tulips). But over time, flowers with bilateral symmetry have emerged, which means the plant is shaped in a way that it has mirror images along a midplane (e.g. orchids and peas).

Take a look at Figure 5 to get an idea.

Figure 5: Two examples of flowers showing radial symmetry (left) and bilateral symmetry (right), respectively.

How did this happen?

It turns out that the purple/yellow pansy flower shown on the right (which has bilateral symmetry) has an additional gene called CYCLOIDEA (abbreviated as CYC). In particular, CYC is expressed in the dorsal petals and makes them different from the ones at the bottom. And not to mention, when researchers mutated the CYC gene (and two other genes known as DICH and RAD), the flower’s symmetry reverted to radial!

Timing - what determines when flowering occurs?

Flowering is a complex process that involves at least 70 genes (based on what we know for now), with many of them connected by several defined response pathways.

For this article, we will focus on four molecules: FT, CO, FD, and TFL-1 in long-day flowering plants (i.e. plants that flower in summer).

The length of the day matters!

Photoperiodism is defined as the plants’ developmental responses to the length of a day (termed as photoperiod).

As showcased by the pioneering works of Garner and Allard in the 1920s, who studied tobacco and soybeans - many plants adjust their flowering period in response to variations in daylength. In addition, they also demonstrated that plants could be categorised into short-day (SD) and long-day (LD) plants, whereby short-day plants accelerate flowering when the day length drops below a critical value (while long-day plants do the opposite).

How does this even happen?

We are going to focus on long-day plants as these are by far the most intensively studied and understood, and it turns out that one of the secret components to this process would be the Flowering locus T gene.

Flowering locus T

The Flowering locus T (or FT) gene is a highly conserved floral integrator gene that encodes for florigen (i.e. flowering hormone). This protein, in turn, is needed in order for a plant to possess a normal flowering time as part of maximising reproductive success.

In fact, when the FT gene was overexpressed in experiments, this led to the production of excess FT proteins and consequently, a very early flowering. Conversely, flowering happens very late when the FT gene is knocked-out (or silenced).

However, to what extent does this gene control the flowering timing exactly?

To study the dynamics of FT, researchers measured the amount of FT in plants under simulated conditions over short and long days (click here to find out more). They used the model plant Arabidopsis thaliana (or A. thaliana), which flowers only on long days. Interestingly, FT was only differentially produced in plants that were experiencing long days and initiated the flowering process, fluctuating within each 24-hour period but peaking in the morning each day.

These experimental results show that there is an increase in FT levels during long days, which is correlated with flowering in A. thaliana, but how can the plant detect longer days and start producing more FT in the morning?


It turns out that the CONSTANS (CO) protein is needed for this process.

CO proteins are constitutively active (i.e. consistently expressed in an active form), and they trigger the production of FT. Nonetheless, FT is shown to only peak during long days. Why is this so?

This is where things get a little more complex.

At a cellular level, plants have photoreceptor proteins that can perceive light and act upon this information as part of the circadian clock and also upon the flowering cycle via CO. These photoreceptors are Phytochrome A and B (or PhyA and PhyB) and Cryptochrome (or Cry).

On one hand, PhyB promotes the degradation (i.e. destruction) of CO proteins. The activity of PhyA and Cry during the day inhibit (i.e. prevent) the action of PhyB. As a result, this creates an accumulation of CO at the end of the day.

On the other hand, as PhyA and Cry become inactive during the night, PhyB can promote the degradation of CO proteins that will be wiped out by the proteasome, a protein degradation machinery, before levels of FT can be significantly high for flowering in the morning.

For FT to be expressed, CO levels need to be sufficiently high in the morning. This is shown in Figure 6 below.

Figure 6: Regulatory process of flowering in long-day plants. Light induces activity of PhyA and Cry, which in turn inhibit PhyB activity, thereby promoting maintenance of CO. In absence of light, PhyA and Cry activity decays, thus permitting activity of PhyB, which induces CO degradation by the proteasome pathway. If CO levels are low (short days), all of it will be wiped out at night and FT cannot be differentially expressed by the next morning to induce flowering. If CO levels are high (long days), not all of it will be degraded at night, which will lead to FT being expressed by the next morning to induce flowering.

In short, the amount of CO proteins tends to increase during the day and decreases during the night. This means that only in long days, the amount of CO proteins can reach a higher enough quantity to trigger the production of FT proteins!

For short-day plants such as rice, CO represses flowering instead of promoting it, which ensures that the plant only flowers in short days where CO expression is sufficiently low. This is seen in Figure 7.

Figure 7: Regulatory process of flowering in short-day plants. Light induces activity of PhyA and Cry which inhibit PhyB activity, thereby promoting maintenance of CO. In the absence of light, PhyA and Cry activity decays permitting activity of PhyB, which induces CO degradation by the proteasome pathway. If CO levels are high (long days), not all of it will be degraded at night, which will lead to FT being repressed by the next morning to prevent flowering. If CO levels are low (short days), all of it will be wiped out at night and FT can be differentially expressed by the next morning to induce flowering.

Focusing back on the FT protein, there are experiments that have proved that it moves within the plant. In particular, when researchers tagged it with the green fluorescent protein (GFP), they observed that it moved through the phloem (from where it is produced) to the shoot (where it is required for flowering).

But what happens when FT arrives at its destination?


When FT arrives at the meristem (i.e. the tip of the plant), it interacts with a transcription factor called FD through a bridging protein called 14-3-3 (which works like an adaptor protein, which binds to FT and FD to induce their contact and pass on the signal to flower by FD).

Figure 8: A picture of a meristem on a plant species known as Crassula Ovata. This photograph was taken by Daniel Levine.

When FT, 14-3-3, and FD come together, they form a hetero-hexamer that is basically known as the “W” complex. When this protein complex is established, a phosphoryl group is added to the C-terminus end of FD. This allows FD to activate the genes involved in flowering, thereby leading to the production of the relevant proteins to stimulate flowering.

The peculiar nature of Terminal Flower 1

Finally, there is Terminal Flower 1 (or TFL1). This is a very puzzling molecule - as although it shares a similar structure to FT as well as interacts with the same proteins as FT (i.e. 14-3-3 and FD), its interaction has the opposite effect to the interactions involving FT!

In fact, while FT promotes flowering, TFL1 promotes a vegetative state in the meristem. This essentially means that the plant focuses most of its energy in carrying out photosynthesis, and is characterised by indeterminate growth and the repression of genes involved in flowering.

How is this possible?

Turns out that the TFL1 gene was only present in the meristem and absent in leaves unlike the FT gene, which originates from the phloem of leaf veins and needs to be transported to the shoot apical meristem (SAM) in order to promote flowering. Since that discovery, it was then proposed that the TFL1 gene could be active during the early stages of the plant’s development. And when it is the right time for flowering, FT is transported to the shoot apex and this will effectively replace TFL1.

Why is flowering research important?

As mentioned previously, flowering is not only essential for the plant involved (to boost the genetic diversity of the population and to produce seeds encapsulated in fruit), but also for the animals which depend on these plants for their diet. This includes humans as well. Most of our food relies on flowering and pollination (e.g. fruits, nuts, and cereal), and we ought to not forget that flowering underlies the foundation of various industries ranging from agriculture to floriculture to apiculture.

And on that note, there are still many challenges that the world of agriculture is currently facing, such as but not limited to pest control and navigating through the adverse climate conditions. In addition, the United Nations (UN) reported that with the global population estimated to reach 9.7 billion in 2050, this essentially equates to an additional two billion people needing food.

Therefore, improving our understanding of processes involved in flowering and what the best conditions are, such as temperature or humidity, would allow us to improve the outcomes in agriculture (e.g. crop yield, crop quality). Meanwhile, this knowledge would enable us to predict and manipulate the flowering timings with more precision, so that we can achieve a certain harvesting target, for instance. Additionally, it would potentially allow us to maximise the production of seeds for the next generation of crops, thereby generating more food to feed our world. In short, it comes with no surprise that research in this area could offer new solutions to resolve the agricultural challenges that we are facing now and in the future.


Yuning Shi

BSc Biochemistry with a Year in Industry

Imperial College London

bottom of page