Fertilization in flowering plants is a two-for-one event: a single pollen grain delivers two sperm cells, one of which fuses with the egg to create an embryo while the other fuses with a separate cell to create the nutritive tissue that feeds the developing seed. This “double fertilization” is unique to flowering plants and was first documented in 1898 in lilies. But the sperm cells do not swim freely through the air or water. Getting them to their target involves a surprisingly elaborate relay of chemical signals, mechanical growth, and molecular gatekeeping that begins the moment a pollen grain touches down on the stigma and does not end until the fruit begins to swell.
Landing and Hydration on the Stigma
A mature pollen grain is partially dehydrated when it leaves the anther, which helps it survive transit through wind or on a pollinator’s body. Once it lands on a receptive stigma, it has to rehydrate before anything else can happen. The stigma’s surface cells, called papillae, actively regulate this step. Research in Arabidopsis shows that pollen grains can hydrate and germinate on stigmas only during specific developmental windows, roughly from stage 9 through stage 13 of flower development. Pollen placed on leaf surfaces, stems, or even immature pistils does not hydrate under normal humidity, which means the stigma is not just a passive landing pad but actively supplies the conditions pollen needs.
Under artificially high humidity, pollen can soak up water on almost any surface, but germination still only occurs on stigmas and young floral organs. That distinction tells researchers that hydration and germination are controlled by separate factors. The stigma provides molecules for both, whereas other flower surfaces may carry the germination factor but lack the hydration cue.
Hydration itself is tightly quality-controlled. Genetic studies have found that mutations affecting a specific regulatory subunit in pollen can cause roughly a quarter of pollen grains to fail hydration on the stigma, even though the same mutant pollen hydrates normally in a wet laboratory dish. The implication is that pollen has to actively communicate with the stigma to take up water, and defective pollen gets screened out at this very first checkpoint.
Pollen Tube Growth Through the Style
Once hydrated, the pollen grain germinates by pushing out a tube through one of its apertures. This tube has a single mission: to deliver the two sperm cells from the stigma’s surface, down through the style, and into the ovary where the ovules are waiting. In some species, this journey is only fractions of a millimeter. In others, like maize, the tube can grow more than 20 centimeters to traverse the silk.
Pollen tubes grow exclusively at their tip, a mode of elongation called “tip growth.” New cell-wall material and membrane are delivered to the very end of the tube by waves of tiny secretory vesicles, a process coordinated by a signaling molecule called ROP GTPase. This molecule is activated in bursts at the tube’s apex, each burst preceding a spurt of elongation. The result is a pulsing, oscillatory pattern of growth rather than a smooth, continuous extension.
The tube also has to physically push through the pistil’s tissues. It uses what researchers describe as invasive growth, actively penetrating between or through cells of the transmitting tract in the style. Feedback from mechanosensitive channels in the tube’s membrane helps it adjust its internal pressure and wall stiffness as it encounters varying resistance, somewhat like a root tip navigating through soil.
Chemical Guidance to the Ovule
A pollen tube does not wander randomly through the pistil. It is guided by chemical signals at multiple scales. Early in its journey, cues from the transmitting tract tissue steer it generally downward through the style. Once the tube exits the transmitting tract and enters the ovary cavity, a more precise guidance system takes over, directed by the ovule itself.
The key attractants are small peptide signals produced by cells inside the ovule called synergid cells. In Arabidopsis, a family of these attractant peptides, known as LUREs, diffuse outward from the ovule and create a concentration gradient the pollen tube can follow. The tube detects these peptides through a receptor called PRK6, which sits on the membrane at the tube’s tip. When PRK6 binds the attractant, it activates the same ROP GTPase signaling system that drives tip growth, but does so asymmetrically. PRK6 accumulates on the side of the tube tip facing the signal source before the tube changes direction, essentially steering growth toward the ovule.
PRK6 does not work alone. It cooperates with related receptors (PRK1, PRK3, and PRK8) and recruits downstream signaling proteins that link attractant sensing to the core growth machinery. This receptor system is species-specific enough that transferring PRK6 from Arabidopsis into pollen of a related species, Capsella rubella, was sufficient to make that pollen respond to Arabidopsis LURE peptides. That specificity helps explain why pollen from one species generally cannot fertilize another.
Inside the Ovule
The target the pollen tube is heading toward is the female gametophyte, also called the embryo sac, buried inside the ovule. In about 70% of flowering plant species, the embryo sac follows a common layout known as the Polygonum type. It contains seven cells organized with striking polarity: at the end nearest the micropyle (the opening where the pollen tube enters) sit the egg cell flanked by two synergid cells; at the opposite end are three antipodal cells; and in the middle is a large central cell containing two polar nuclei.
Each cell type has a distinct job. The synergid cells are the ones producing the LURE attractant peptides and are packed with dense cytoplasm and abundant organelles, consistent with their active signaling role. The egg cell, by contrast, tends to have relatively sparse cytoplasm with large vacuoles. The central cell, with its two polar nuclei, will become one of the two fertilization targets. The antipodal cells at the far end often degenerate by the time fertilization occurs, and they are thought to have served a nutritional support function earlier in development.
Pollen Tube Reception and Sperm Release
When the pollen tube finally reaches the ovule, it enters through the micropyle and grows between or into one of the two synergid cells. What happens next is dramatic and precisely coordinated: receptor-like kinases on both the synergid cell and the pollen tube tip trigger a signaling cascade that causes the synergid cell to die and the pollen tube to burst open, releasing its two sperm cells into the embryo sac.
This mutual destruction is not an accident. The synergid’s death clears space and releases its cellular contents, which may help the sperm cells reach their targets. Recent two-photon live imaging in Arabidopsis pistils has revealed additional layers of control at this stage. The ovule and surrounding tissues deploy signals to prevent more than one pollen tube from entering the same ovule, a safeguard against delivering extra sperm. One receptor involved in this “polytubey block” is FERONIA, which functions in both the septum tissue and the ovule itself. However, these blocks are not absolute in the first 45 minutes or so after fertilization, which may explain occasional reports of extra sperm delivery.
Double Fertilization
With both sperm cells now inside the embryo sac, the two fertilization events happen nearly simultaneously. One sperm fuses with the egg cell, producing a diploid zygote that will develop into the embryo. The other sperm fuses with the central cell, specifically with its two polar nuclei, producing a triploid cell that will develop into the endosperm, the tissue that nourishes the growing embryo and, in cereal grains, is the part we eat.
This double fertilization was first described by Sergei Nawaschin in 1898, working with lily and fritillary flowers. It remains one of the defining features of angiosperms and distinguishes their reproduction from that of conifers and other non-flowering seed plants, where only the egg is fertilized and the nutritive tissue develops from the mother’s cells alone.
Even the prevention of errors at this step is carefully managed. When researchers experimentally delivered four sperm cells to a single ovule in Arabidopsis, the egg cell never fused with more than one sperm, maintaining a strict one-to-one rule. The central cell was more permissive: about half of central cells in these experiments fused with more than one sperm. So the egg cell has a robust polyspermy block, while the central cell’s block is present but leakier.
After Fertilization: Seeds and Fruit
Fertilization does not just create a new organism. It also triggers the transformation of the ovule into a seed and the surrounding ovary into a fruit. The signal that initiates this transformation is the plant hormone auxin. Within hours of fertilization, auxin levels rise sharply in the fertilized ovule. This auxin surge then activates synthesis of a second hormone, gibberellin, specifically inside the ovule. The gibberellins produced in the ovule are transported outward to the ovary wall (the valves, in Arabidopsis), where they promote cell expansion and the coordinated growth of the fruit.
The endosperm plays a central role in these hormonal cascades. A gene called AGL62, activated in the newly formed endosperm, is required for turning on auxin production. Mobile auxin from the endosperm subsequently stimulates development of both the seed coat and the surrounding fruit tissue. This means the endosperm is not just future food storage for the embryo; it acts as a hormonal command center that tells the mother plant to invest resources in building a fruit around the developing seed.
Imprinting adds another twist. In the endosperm, certain genes are expressed differently depending on whether they came from the mother or the father. Researchers have identified dozens of imprinted genes in Arabidopsis endosperm, regulated by DNA methylation differences between the parental genomes and by Polycomb group proteins. This parent-of-origin gene expression is thought to mediate a tug-of-war over resource allocation: the paternal genome may push for more endosperm growth (benefiting its offspring), while the maternal genome tempers that demand (preserving resources for future seeds).
Self-Incompatibility Systems
Many flowering plants have evolved mechanisms to reject their own pollen, forcing outcrossing and maintaining genetic diversity. One of the best-studied systems uses a molecule called S-RNase, which is present in the pistil. During cross-pollination, the pollen tube produces proteins called SLFs that recognize the S-RNase as foreign, tag it for destruction, and allow the tube to keep growing. During self-pollination, the SLF proteins cannot neutralize the matching S-RNase. The unchecked S-RNase then disrupts the pollen tube’s internal structure and triggers its death.
Recent work has clarified how this killing happens. The self-matched S-RNase molecules that escape degradation undergo phase separation, forming dense clusters in the pollen tube’s cytoplasm. These clusters wreck the cytoskeleton and throw off the cell’s chemical balance, leading to programmed cell death. The system is governed by a single genetic region with many variants, called the S-locus, and has been found across distantly related plant families including the Solanaceae (tomatoes, peppers), Rosaceae (apples, cherries), and Plantaginaceae.
When Fertilization Fails
Heat is one of the most common environmental threats to flower fertilization. Developing anthers and pollen grains are particularly sensitive to high temperatures at certain stages, and the failure to set fruit during heat waves has mainly been attributed to pollen damage rather than problems on the female side. As climate variability increases, this vulnerability has become a major concern in crop breeding programs, which are actively selecting for heat-tolerant pollen performance.
Interference from the wrong pollen is another common problem. In natural plant communities where multiple species bloom simultaneously and share pollinators, pollen from one species frequently lands on stigmas of another. This heterospecific pollen can reduce seed production in the receiving plant, not just by taking up space on the stigma but by actively interfering with the growth of the correct pollen tubes inside the style. In experiments with the monkeyflower Mimulus guttatus, pollen from a co-flowering species reduced seed set, and the effect depended on the composition and timing of the foreign pollen load.
Invasive plant species can make this worse. When pollen from the invasive Brassica nigra was applied to the native Penstemon parryi at the same time as the plant’s own pollen, seed production dropped substantially. Fewer of the native plant’s pollen tubes reached the base of the style, suggesting the invasive pollen was physically or chemically blocking their progress. The damage was worst when the foreign pollen arrived simultaneously or before the plant’s own pollen.
Bypassing Fertilization Entirely
Some plants have found a way around the entire process. Apomixis is a form of asexual seed production in which a plant generates seeds without meiosis or fertilization, producing offspring that are genetic clones of the mother. Apomixis occurs naturally in hundreds of species, including dandelions and some grasses, and has long been a target for agricultural biotechnology. If apomixis could be reliably engineered into crop plants, farmers could save hybrid seed and replant it without losing the yield advantage that comes from hybrid vigor, since every seed would be genetically identical to the high-performing parent.
Progress toward synthetic apomixis has accelerated. Researchers have shown that combining targeted modifications to both the meiosis and fertilization pathways can produce clonal seeds at high frequencies in laboratory settings. The approach typically involves knocking out genes required for normal meiotic recombination and chromosome reduction, while simultaneously triggering embryo development from the egg cell without sperm fusion. Translating this from model species like rice into a broad range of crops remains a major engineering challenge, but the foundational proof of concept exists.
Double Fertilization Beyond Flowering Plants
Although double fertilization is the hallmark of angiosperms, something resembling it has been observed in non-flowering seed plants as well. In some conifers, cycads, and Gnetales, two sperm can fertilize separate nuclei within the same archegonium, the structure that houses the egg in these plants. This “intra-archegonial double fertilization” requires that a second female nucleus persist long enough to be available when the sperm arrive, and that a second functional sperm enter the chamber. In Ephedra, the second fertilization product can even begin dividing, though it does not produce a functional endosperm equivalent. Whether this process is a precursor to angiosperm double fertilization or an independent evolutionary experiment remains an open question, but it shows that the raw ingredients for double fertilization existed before flowering plants appeared on the scene.