ABSTRACT
Cell division is a fundamental process shared across diverse life forms, from yeast to humans and plants. Multicellular organisms reproduce through the formation of specialized types of cells, the gametes, which at maturity enter a quiescent state that can last decades. At the point of fertilization, signalling lifts the quiescent state and triggers cell cycle reactivation. Studying how the cell cycle is regulated during plant gamete development and fertilization is challenging, and decades of research have provided valuable, yet sometimes contradictory, insights. This Review summarizes the current understanding of plant cell cycle regulation, gamete development, quiescence, and fertilization-triggered reactivation.
Introduction
The cell is the basic structural and functional unit of life. Through cell division, new cells are produced, ensuring the continuity of life and supporting the development and function of all living beings. Cells typically divide following a unidirectional order of four cell cycle stages: the S (synthesis) phase, when cells replicate their nuclear DNA; the M (mitosis) phase, when sister chromatids separate and are allocated to the nascent daughter cells; and two gap phases, G1 and G2, that space the M and S phases (Fig. 1). Following these rules, two types of cell division can occur: mitosis and meiosis. During mitosis, diploid cells replicate their DNA content and, through one round of division, divide to produce two diploid daughter cells. Mitosis is typical of somatic cells and contributes to the growth, development and repair of tissues and organs. In contrast, meiosis is a process where cells replicate their genome, but then the DNA content is distributed among four cells through two consecutive chromosome separations. The resulting cells are therefore haploid and contain half of the genetic material of the mother cell. Meiosis is a type of cell division restricted to the cells of the germline, through which female and male gametes – eggs and sperm – are produced. Multicellular organisms that reproduce through sexual reproduction, such as animals and plants, have evolved sophisticated mechanisms to regulate cell cycle progression during the development of female and male gametes. Once formed, gametes must enter a quiescent state to prevent cell division in the absence of fertilization. In some species, quiescence can last for months or even years. At fertilization, quiescency must be lifted, to allow the newly formed embryo to divide and grow. This raises important fundamental questions: What are the molecular mechanisms that establish the quiescent state in the gametes? How do gametes recognize that fertilization has occurred and that the cell cycle must be resumed? How does this fundamental aspect of sexual reproduction vary across different species?
The cell cycle in Arabidopsis. Graphical representation of the Arabidopsis cell cycle and of its regulators involved in each phase. The phases of the cell cycle (G1-S-G2-M) are represented with arrows, reflecting the unidirectionality of the process. High levels of the cell cycle inhibitors KRPs and RBR1 cause the cell cycle to arrest in the G1 phase by inhibiting CDKA activity and sequestering E2F, thus preventing E2F from activating the genes required for DNA replication. At the G1-S transition, levels of CYC D and A types increase, allowing the formation of CDKA–CYC complexes. With the activating phosphorylation from CAK, CDKA kinase activity is enhanced. CDKA then phosphorylates RBR1, leading to the proteasome-mediated degradation of RBR1. This releases E2F from inhibition, enabling the expression of genes necessary for DNA replication. At the S-G2 boundary, the cellular concentration of CYC A and B types increases, promoting their dimerization with CDKA-B kinases. The activity of the CDKA-B kinases is inhibited by KRPs, and by phosphorylation by WEE1. In the G2 phase, CDKA-B activity is further suppressed by the SIM/SMR inhibitors. However, as the cell prepares for mitosis, phosphorylation of CDKA-B by CAK activates the CDKA-B–CYCB complex, which in turn stimulates the activity of the APC/C complex. The APC/C mediates the ubiquitination of A- and B-type CYCs, leading to their degradation. The destruction of CYCB allows the cell to commit to mitosis. APC/C, ANAPHASE PROMOTING COMPLEX; CDKA, CDKB, CYCLIN-DEPENDENT KINASE A/B type; E2F, E2F TRANSCRIPTION FACTOR; KRP, KIP-RELARED PROTEIN; RBR1, RETINOBLASTOMA RELATED 1; SIM/SMR, SIAMESE/SIAMESE-RELATED; CAK, CDK-ACTIVATING KINASE; WEE1, WEE1 KINASE HOMOLOGUE; P, phosphorylation; Ub, ubiquitination.
The cell cycle in Arabidopsis. Graphical representation of the Arabidopsis cell cycle and of its regulators involved in each phase. The phases of the cell cycle (G1-S-G2-M) are represented with arrows, reflecting the unidirectionality of the process. High levels of the cell cycle inhibitors KRPs and RBR1 cause the cell cycle to arrest in the G1 phase by inhibiting CDKA activity and sequestering E2F, thus preventing E2F from activating the genes required for DNA replication. At the G1-S transition, levels of CYC D and A types increase, allowing the formation of CDKA–CYC complexes. With the activating phosphorylation from CAK, CDKA kinase activity is enhanced. CDKA then phosphorylates RBR1, leading to the proteasome-mediated degradation of RBR1. This releases E2F from inhibition, enabling the expression of genes necessary for DNA replication. At the S-G2 boundary, the cellular concentration of CYC A and B types increases, promoting their dimerization with CDKA-B kinases. The activity of the CDKA-B kinases is inhibited by KRPs, and by phosphorylation by WEE1. In the G2 phase, CDKA-B activity is further suppressed by the SIM/SMR inhibitors. However, as the cell prepares for mitosis, phosphorylation of CDKA-B by CAK activates the CDKA-B–CYCB complex, which in turn stimulates the activity of the APC/C complex. The APC/C mediates the ubiquitination of A- and B-type CYCs, leading to their degradation. The destruction of CYCB allows the cell to commit to mitosis. APC/C, ANAPHASE PROMOTING COMPLEX; CDKA, CDKB, CYCLIN-DEPENDENT KINASE A/B type; E2F, E2F TRANSCRIPTION FACTOR; KRP, KIP-RELARED PROTEIN; RBR1, RETINOBLASTOMA RELATED 1; SIM/SMR, SIAMESE/SIAMESE-RELATED; CAK, CDK-ACTIVATING KINASE; WEE1, WEE1 KINASE HOMOLOGUE; P, phosphorylation; Ub, ubiquitination.
Studying gamete development and quiescence poses significant technical challenges. In animals like frogs, eggs are easily collected and have indeed provided foundational insights into gamete biology, including cell cycle regulation. In contrast, mammalian gametes often develop during fetal stages, making their collection in sufficient numbers for downstream analyses complicated. Techniques like in vitro fertilization have been instrumental in overcoming these barriers, enabling detailed investigations of gamete development. Plants produce hundreds of male gametes, which are easy to collect, but female gametes are much harder to access as they develop deep within multiple layers of maternal tissue, making their collection both difficult and time-consuming. This has resulted in a historical gap in understanding cell cycle regulation and quiescence in plant female gametes, though recent advances are beginning to uncover the signals that govern these processes, which brings us closer to engineering fertilization in plants to guarantee seed production in crops.
In this Review, I summarize the molecular mechanisms regulating the cell cycle in eukaryotic cells and highlight how these processes govern the formation of plant gametes. I also bring together both classic and recent, sometimes controversial, discoveries about plant cell cycle regulation around the moment of fertilization.
The molecular basis of the cell cycle
The cell cycle has been extensively investigated, with research conducted across a broad range of organisms, including yeast, worms, insects, amphibians, mammals and plants. At the end of the 19th century, several scientists made groundbreaking contributions to our understanding of cell division. While observing stained cells of salamander embryos, the German biologist Walther Flemming described filamentous structures within the nucleus (Flemming, 1882). These structures were later termed ‘chromosomes’ by Wilhelm von Waldeyer-Hartz, from the Greek words for ‘colour’ (Khroma) and ‘body’ (Soma) (Waldeyer-Hartz, 1888). In the publication Zellsubstanz, Kern und Zelltheilung, Flemming drew the first detailed map of the sequential events happening during cell division, leading to chromosome separation (Flemming, 1882). Around the same time, Belgian embryologist Edouard Van Beneden was the first to observe and describe meiosis in animal egg cells (Van Beneden, 1883) and Polish botanist Eduard Strasburger described female gametophyte development and mitosis in plants (Strasburger, 1876).
Since these early discoveries, a collective research effort has yielded a comprehensive understanding of the cell cycle across the cellular, tissue and organism levels. We now know that the control of progression through the cell cycle phases relies on the irreversible modification of protein activity, mainly achieved through protein phosphorylation. One of the first major discoveries in the field dates back to the 1970s, when Yoshio Masui's pioneering work in frog oocytes led to the hypothesis of the existence of a factor, the Maturation Promoting Factor (MPF), that is sufficient and necessary to drive cell cycle progression through the consecutive phases (Masui and Markert, 1971). The MPF was revealed to be a protein complex, consisting of two protein kinases (with phosphorylation activity) and an activating partner. One of the kinases was initially described in budding yeast, where it was named Cdc2/Cdc28 (Hartwell et al., 1974; Nurse et al., 1976). Cdc2/Cdc28 is conserved across mammals and plants, which has led to the concept of a universal cell cycle engine (Hartwell et al., 1974; Nurse et al., 1976; Gutierrez, 2016, 2022).
Cdc2/Cdc28 is required to trigger the G1-S and G2-M transitions, and it is constitutively present in the cell (Hartwell et al., 1974; Nurse et al., 1976). To guarantee that Cdc2/Cdc28 is effective only under certain conditions and solely at the G1-S and the G2-M boundaries, its full activation relies on its dimerization with a partner protein. This partner protein was first identified by Tim Hunt's team, who termed it ‘cyclin’ because its expression fluctuates in a cyclical manner along the cell cycle phases (Draetta et al., 1989; Evans et al., 1983; Gautier et al., 1990; Labbé et al., 1989; Simanis and Nurse, 1986). The second kinase in the MPF complex, named Greatwall (in frog) or Mastl (microtubule-associated serine/threonine kinase-like, in mammals) (Castro and Lorca, 2018; García-Blanco et al., 2019; Gharbi-Ayachi et al., 2010; Mochida et al., 2010), was discovered later, and was identified as a key component necessary for mitosis progression (Gharbi-Ayachi et al., 2010; Mochida et al., 2010). Greatwall promotes correct timing and progression of mitosis by counteracting the activity of protein phosphatase 2A (PP2A), which acts to dephosphorylate mitotic substrates (Gharbi-Ayachi et al., 2010; Mochida et al., 2010). Plants lack homologs of the Greatwall kinase but possess MAST kinases, which feature a catalytic domain similar to that of Greatwall (Chudinova et al., 2017; Karpov et al., 2010). However, it remains to be determined whether these proteins also share functional homology with Greatwall (Chudinova et al., 2017; Karpov et al., 2010).
Cell cycle regulation in Arabidopsis
In plants, the homologues of yeast Cdc2/Cdc28 are known as CYCLIN-DEPENDENT KINASES (CDKs). CDKs are Serine-Threonine kinases that can phosphorylate a range of targets (Nigg, 1995). In this Review, unless otherwise stated, the data discussed come from Arabidopsis studies. The model plant Arabidopsis thaliana possesses at least 29 CDK and 15 CDK-like (CDKL) genes, which can be categorized into eight classes: CDKs from A- to G-type, and CDKLs (Joubès et al., 2000; Vandepoele et al., 2002). CDKA is the equivalent of the canonical Cdc2 in fission yeast and is directly involved in cell cycle progression at the G1-to-S and G2-to-M phase transitions, whereas the activity of the B-type CDKs (there are four B-type CDKs in Arabidopsis; Joubès et al., 2000) seems to be restricted to the late G2-M phase (Nowack et al., 2012).
During cell cycle progression, CDKs are activated by binding to a cyclin (CYCs, Fig. 1). In Arabidopsis, at least 50 CYCs have been identified and grouped into ten classes: A-, B-, C-, D-, H-, L-, SDS-, Q-, T- and P-types (Wang et al., 2004). The heterodimerization of CYCs with CDKs is required for the activation of CDK kinase activity (Arellano and Moreno, 1997). Some CYCs possess a destruction box (Dbox), which is recognized by ubiquitin-protein ligases and ultimately serves as a signal for their proteosomal-mediated degradation (Glotzer et al., 1991). The progression through the consecutive phases of the cell cycle relies on the regulated synthesis and waves of destruction of CYCs, which modulate the activation status of CDKs (Arellano and Moreno, 1997). These waves play a crucial role in driving the G1-S and G2-M transitions (Bähler, 2005). Various classes of CDKs and CYCs form diverse complexes in Arabidopsis, providing multi-layered regulation of cell cycle progression in response to external stimuli, developmental stage, and tissue and cell identity (Van Leene et al., 2010).
To ensure that CDKs are activated only when CYC concentrations reach certain thresholds, proteins belonging to the Kip-related proteins/Interactor-Inhibitor of Cyclin-Dependent Kinase (KRP/ICK) family (seven in Arabidopsis; De Veylder et al., 2001; Lui et al., 2000; Wang et al., 1997), as well as members of the SIAMESE/SIAMESE RELATED (SIM/SMR) family (17 in Arabidopsis; Churchman et al., 2006; Peres et al., 2007), inhibit CDK activity by binding to and physically blocking the catalytic site of CDKs (Fig. 1). Degradation of KRPs through ubiquitination and subsequent proteosome-mediated degradation lifts their inhibitory effect (Hengst, 2004; Verkest et al., 2005). Indeed, KRP degradation is essential to ensure mitotic progression (Liu et al., 2008; Ren et al., 2008).
The activity of CDKs is also regulated by phosphorylation. In yeast, animals and plants, CDK-activating kinase (CAK) phosphorylates a conserved threonine residue in the CDKs, thus enhancing accessibility of the CDK catalytic site (Dissmeyer et al., 2007; Harashima et al., 2007; Umeda et al., 2005). The first plant CAK orthologue was identified in rice and later named Oryza sativa CYCLIN DEPENDENT KINASE D;1 (Os;CDKD;1; Hata, 1991). In Arabidopsis, there are three CDKD genes: CDKD;1, CDKD;2 and CDKD;3 (Shimotohno et al., 2003). CDKDs are in turn phosphorylated and activated by the kinase CYCLIN-DEPENDENT KINASE F;1 (CDKF;1; Shimotohno et al., 2004; Takatsuka et al., 2009).
In animals and yeast, the kinase WEE1 negatively regulates CDK activity through phosphorylation, thereby inhibiting ATP binding and substrate recognition (McGowan and Russell, 1993; Parker and Piwnica-Worms, 1992). Plants possess WEE1 homologues (Gonzalez et al., 2004; Sorrell et al., 2002; Sun et al., 1999), which share some similarities with their animal counterparts but also exhibit distinct functions. In Arabidopsis, WEE1 phosphorylates CDKA;1 (De Schutter et al., 2007), similar to the WEE1-CDK interaction seen in animals. However, unlike in animals, this phosphorylation does not appear to impact CDKA;1 activity (Dissmeyer et al., 2010, 2009). In Arabidopsis, WEE1 can also inhibit cell cycle progression by directly interacting with and phosphorylating the E3 ubiquitin ligase F-BOX-LIKE 17 (FBL17) (Pan et al., 2021), which normally acts as an inhibitor of KRPs (Gusti et al., 2009; Kim et al., 2008; Zhao et al., 2012). Upon phosphorylation by WEE1, FBL17 becomes polyubiquitinated and is degraded, leading to the accumulation of KRPs and ultimately halting CDK activity and cell cycle progression (Pan et al., 2023, 2021).
These cell cycle components – CDKs, CYCs and inhibitors – operate in feedback loops so that each factor promotes the expression and stability of its activators, while repressing the activity of its inhibitors (Venta et al., 2012). This behaviour makes the cell cycle a robust, wired mechanism that generates two stable, opposite steady states: inactive and active (Harashima et al., 2013; Stallaert et al., 2019). The transition from one state to the other occurs as a biological switch (Ferrell, 2002; Pomerening et al., 2003). This property ensures the unidirectional progression of the cell cycle, making the system resistant to small fluctuations in protein concentrations because it takes higher levels of CDK activity (and CYC concentrations) to initiate mitosis than those required to maintain it (Harashima et al., 2013). This characteristic is defined as hysteresis, which is when a system requires more of a signal to switch states than to stay in the current state (Dragoi et al., 2024; Pomerening et al., 2003; Sha et al., 2003).
The G1-S transition in Arabidopsis
The G1-S transition in Arabidopsis is primarily controlled by A- and D-type CYCs (Oakenfull et al., 2002; Takahashi et al., 2010). At the G1-S transition, the main target of the CDKA–CYCD complex is the homologue of the animal Retinoblastoma protein (pRb), a tumour suppressor protein, the dysfunction of which is often associated with tumorigenesis and cancer progression (Friend et al., 1986; Fung et al., 1987; Lee et al., 1987). pRb homologues have been identified in plants, including RBR1 in Arabidopsis (Ach et al., 1997a; Grafi et al., 1996; Kong et al., 2000; Nakagami et al., 1999; Xie et al., 1996). RBR1 exhibits remarkable functional conservation with the animal pRb (Ach et al., 1997a; Grafi et al., 1996; Kong et al., 2000; Nakagami et al., 1999; Xie et al., 1996). The main function of both animal pRb and plant RBR1 is to inhibit the progression of cells into S-phase by directly binding to E2F transcription factors, thus preventing E2F from activating genes necessary for DNA synthesis and cell cycle progression (Fig. 1) (Hiebert et al., 1992; Zamanian and La Thangue, 1992). In addition, similar to pRb in animals, plant RBR1 recruits chromatin-remodelling complexes such as histone methyltransferases and deacetylases to locally repress transcription (Brehm and Kouzarides, 1999; Gu et al., 2011; Jullien et al., 2008; Ötvös et al., 2021; Rossi et al., 2003; Vandel et al., 2001). D-type CYCs directly interact with pRb/RBR1 via the conserved LxCxE motif, thus tethering CDK activity to pRb/RBR1 (Dowdy et al., 1993; Ewen et al., 1993; Kato et al., 1993; Soni et al., 1995; Xie et al., 1995). CDK-dependent phosphorylation inhibits pRb/RBR1 dimerization with E2F and leads to its proteosomal-mediated degradation, thus promoting progression into S-phase (Burke et al., 2010; Harbour et al., 1999; Mittnacht et al., 1994; Mittnacht and Weinberg, 1991; Rubin et al., 2005) (Fig. 1). Reducing RBR1 levels can suppress the phenotype of the cdka;1 mutant, suggesting that RBR1 is also the main target of CDKA;1 in Arabidopsis (Nowack et al., 2012; Ramírez-Parra et al., 1999; Sekine et al., 1999). RBR1 and pRb are also both involved in cell cycle arrest and DNA repair in response to DNA damage thanks to their ability to both halt DNA replication by blocking the recruitment or maintenance of crucial replication factors on chromatin, and to interact with DNA repair complexes (Knudsen et al., 1998, 2000; Angus et al., 2004; Biedermann et al., 2017; Horvath et al., 2017).
The concentration of G1-S inhibitors is also important for ensuring that cell division occurs appropriately according to the size of the cell. In large cells, cell cycle inhibitors become diluted, and this condition triggers the G1-S transition (D'Ario and Sablowski, 2019). Conversely, in small cells, the concentration of inhibitors remains high, prolonging the G1 phase to allow for cell growth before division (D'Ario and Sablowski, 2019). In human cells, the level of pRb correlates with the ability of a cell to divide: small cells contain a high concentration of pRb and thus their progression into the G1-S transition is inhibited, while large cells dilute pRb so that cell cycle progression is promoted (Zatulovskiy et al., 2020). A similar mechanism occurs in yeast, where the dilution of the pRb functional homologue Whi5 is dependent on cell growth (Schmoller et al., 2015). In plants, members of the KRP family bind to mitotic chromosomes, ensuring that daughter cells inherit equal amounts of the inhibitor, regardless of their size (D'Ario et al., 2021; Sablowski and Gutierrez, 2022). In smaller daughter cells, the higher concentration of KRPs delays entry into the S phase, allowing the cell to grow to the appropriate size during the G1 phase (D'Ario et al., 2021).
The G2-M transition in Arabidopsis
The G2-M transition in Arabidopsis is governed by B-type CYCs, which, in complex with CDKs, phosphorylate several substrates to regulate processes including nuclear envelope breakdown, chromosome condensation, spindle assembly, microtubule organization and chromosome segregation (Blethrow et al., 2008; Menges et al., 2005; Romeiro Motta et al., 2022; Wang et al., 2004). To complete mitosis and promote cytokinesis, CYCB must be degraded by the Anaphase Promoting Complex/Cyclosome (APC/C, Fig. 1), a multi-subunit ubiquitin-ligase complex that recognises the Dbox in B-type CYCs, targeting them for degradation (Townsley and Ruderman, 1998). APC/C activation specificity relies on the coactivators CDC20 and CCS52B (Pesin and Orr-Weaver, 2008), the mRNAs of which are retained in the nucleus until nuclear envelope breakdown, when they are released into the cytoplasm to be translated (Yang et al., 2017).
The expression of genes specific to the G2/M transitions is regulated by a multimeric, evolutionarily conserved complex known as the DREAM (DP, Rb-like, E2F and MuvB) complex (Guiley et al., 2015; Sadasivam and DeCaprio, 2013). In animals, the DREAM complex is composed of E2F and Myb-type transcription factors and Rb-related proteins (Guiley et al., 2015). This complex is conserved across various species, including nematodes, fruit flies, animals and plants (Harrison et al., 2007, 2006; Kobayashi et al., 2015; Korenjak et al., 2004; Lewis et al., 2004; Litovchick et al., 2007; Pilkinton et al., 2007; Schmit et al., 2007). In plants, there are at least two distinct flavours of the DREAM complex, depending on the transcriptional activity of the MYB-type transcription factors (Kobayashi et al., 2015; Latorre et al., 2015; Magyar et al., 2016). DREAM complexes containing activator MYB3Rs are involved in promoting expression of mitotic genes, particularly those genes that are essential for cytokinesis (Haga et al., 2007). In contrast, repressor MYB3Rs function outside the mitotic phase, playing a key role in silencing mitotic genes and imposing quiescence in mature organs (Kobayashi et al., 2015).
Cell cycle regulation during plant reproduction
Gametes – sperm in males and eggs in females – are specialized cells that are haploid at maturity. Once fully developed, gametes enter a state of quiescence, which is lifted by signals occurring at fertilization, enabling the zygote to initiate mitotic cycles (Tao and Nodine, 2019; Khanday et al., 2023). Precisely orchestrated cell cycle events govern gamete development, the acquisition of quiescence, and exit from quiescence at fertilization. Failures in any of these processes can lead to developmental defects, potentially causing abortion of the progeny (Chaudhury et al., 1997; Guitton and Berger, 2005).
Gametogenesis evolved independently in animals and plants (Barrett, 2002). In animals, functional gametes are the direct product of meiosis of primordial germ cells, which are specified during early embryogenesis (Saffman and Lasko, 1999). Spermatogenesis begins and proceeds continuously postnatally, whereas oogenesis initiates during foetal development and mature oocytes await fertilization at arrested meiosis II (Pei et al., 2023). In contrast, the life cycle of land plants alternates between two generations: the sporophyte and the gametophyte (Pandey et al., 2022). The sporophyte is the diploid phase, through which haploid spores are produced by meiosis (Hisanaga et al., 2019). During the gametophytic phase, these haploid spores undergo mitosis to produce the female and male gametes (Hisanaga et al., 2019). The structure that harbours the gametes is called gametophyte; pollen is the gametophyte for the male gametes (Twell, 2011), and the embryo sac is the gametophyte for the female gametes (Drews and Koltunow, 2011). The dominance of either the gametophytic or sporophytic phase varies by species: in mosses, the gametophyte is dominant, whereas in vascular plants, such as gymnosperms and angiosperms such as Arabidopsis, the sporophyte is predominant (Hisanaga et al., 2019). The transitions between these phases happen at two developmental points: fertilization, where two haploid gametes fuse to form a diploid sporophyte, and meiosis, where the diploid sporophyte generates haploid gametophytes. Thus, in flowering plants, gametes are not the direct products of meiosis. Instead, haploid spores formed after meiosis undergo additional rounds of mitosis to produce the gametes (Berger and Twell, 2011; Dresselhaus et al., 2016). The male gametes (sperm cells) arise from two rounds of mitosis (Twell, 2011), while the female gametes (comprising the egg cell and central cell) result from three rounds of mitosis (Drews and Koltunow, 2011).
In angiosperms, reproductive structure development and gametogenesis typically occur after the plant has undergone a vegetative phase during which only vegetative tissues, such as leaves, are produced, and no reproductive structures are formed (Araki, 2001). Upon receiving external signals, such as changes in light and temperature, as well as internal cues such as hormone levels, the vegetative meristem transitions into an inflorescence meristem (Chahtane et al., 2023). In plants like Arabidopsis, this identity change marks the plant's irreversible commitment to the reproductive phase. Flowers arise from the flanks of the apical meristem, starting as a small, undifferentiated cluster of cells (Smyth et al., 1990). During development, the floral organs are specified and differentiate into flowers. The mature Arabidopsis flower consists of four concentric whorls: from the outside in, these are four sepals, four petals, six stamens, and a central pistil made of two fused carpels (Alvarez-Buylla et al., 2010). At the tip of each stamen, the anthers are the site of male gametophyte (pollen) development. Inside the pistil, ovules house the female gametophyte (the embryo sac). Within these male and female gametophytes, the germline cells develop as sperm in pollen and as the egg cell and central cell in the embryo sac (Berger and Twell, 2011).
Male reproductive development
The mature male gametophyte (the pollen grain) contains the male gametes (the sperm cells) (Borg et al., 2009; McCormick, 2004). Pollen development begins when a sporogeneous cell undergoes meiosis to produce four haploid microspores (Fig. 2) (Borg et al., 2009; McCormick, 2004; Twell, 2011). The microspore then undergoes an asymmetric mitotic division to generate a large vegetative cell and a small, unique male generative cell (Fig. 2) (Borg et al., 2009; McCormick, 2004). This stage is defined as ‘bicellular pollen’. The generative cells then divide once more by mitosis to produce the two sperm cells, creating the final, mature male gametophyte (Borg et al., 2009; McCormick, 2004). A pollen with one vegetative cell and two sperm cells is defined as tricellular pollen (Fig. 2) (Borg et al., 2009; McCormick, 2004). Pollen grains from most angiosperms are held at the bicellular stage when mature, and the last mitosis event that produces the two sperm cells occurs after pollen germination within the pollen tube (Brewbaker, 1967). In other angiosperms such as Arabidopsis and rice, the mature pollen is tricellular because the last mitosis of the generative cell occurs before pollen germination (Berger and Twell, 2011; Brewbaker, 1967).
Male reproductive development in Arabidopsis. (A-D) Graphical summary of the main events during Arabidopsis male reproductive development, and of the cell cycle regulators involved in the meiosis and mitosis events. Pollen development begins in the anthers, where diploid pollen mother cells undergo meiosis to produce four haploid microspores. During meiosis (A), CDKA–CYC complexes phosphorylate a series of target proteins, such as ASY1 and TDM1, which control chromosome axis formation and exit from meiosis. During male gametogenesis, the microspores undergo an asymmetric mitotic division, producing a large vegetative cell and a smaller generative cell. Cell division in the vegetative cell is inhibited by RBR1 and KRPs (B). The generative cell undergoes another round of mitosis, resulting in two sperm cells. This process is mediated by DUO1, which activates the CDKA;1–CYCB complex (C). Both mitotic divisions are regulated by the CDKA;1–CYCD complex. FBL17 facilitates the degradation of KRPs, thereby enabling CDKA;1 activity during the second mitotic division. CDKA;1 activity, in turn, promotes the phosphorylation and degradation of RBR1, thus releasing E2F from inhibition and enabling the expression of genes necessary for DNA replication (D). ASY1, ASYNAPTIC 1; CDKA, CYCLIN-DEPENDENT KINASE A; CYCA/D, CYCLIN A/D-type; E2F, E2F TRANSCRIPTION FACTOR; KRP, KIP-RELATED PROTEIN; RBR1, RETINOBLASTOMA RELATED 1; DUO1, DUO POLLEN 1; FBL17, F-BOX LIKE 17; SCF, SCF ubiquitin ligase complex; TDM1, THREE DIVISION MUTANT 1; P, phosphorylation.
Male reproductive development in Arabidopsis. (A-D) Graphical summary of the main events during Arabidopsis male reproductive development, and of the cell cycle regulators involved in the meiosis and mitosis events. Pollen development begins in the anthers, where diploid pollen mother cells undergo meiosis to produce four haploid microspores. During meiosis (A), CDKA–CYC complexes phosphorylate a series of target proteins, such as ASY1 and TDM1, which control chromosome axis formation and exit from meiosis. During male gametogenesis, the microspores undergo an asymmetric mitotic division, producing a large vegetative cell and a smaller generative cell. Cell division in the vegetative cell is inhibited by RBR1 and KRPs (B). The generative cell undergoes another round of mitosis, resulting in two sperm cells. This process is mediated by DUO1, which activates the CDKA;1–CYCB complex (C). Both mitotic divisions are regulated by the CDKA;1–CYCD complex. FBL17 facilitates the degradation of KRPs, thereby enabling CDKA;1 activity during the second mitotic division. CDKA;1 activity, in turn, promotes the phosphorylation and degradation of RBR1, thus releasing E2F from inhibition and enabling the expression of genes necessary for DNA replication (D). ASY1, ASYNAPTIC 1; CDKA, CYCLIN-DEPENDENT KINASE A; CYCA/D, CYCLIN A/D-type; E2F, E2F TRANSCRIPTION FACTOR; KRP, KIP-RELATED PROTEIN; RBR1, RETINOBLASTOMA RELATED 1; DUO1, DUO POLLEN 1; FBL17, F-BOX LIKE 17; SCF, SCF ubiquitin ligase complex; TDM1, THREE DIVISION MUTANT 1; P, phosphorylation.
Below, I discuss the important roles that cycle components play in each step of pollen formation, from meiosis to asymmetric cell division, and mitosis.
Male sporogenesis
During meiosis, several types of CYCs are expressed in the microspore cells (Bulankova et al., 2013). A-type CYCs are involved in the correct segregation of meiotic chromosomes, and the timely entry and exit from meiosis (Bulankova et al., 2013; Cairo et al., 2022; Cifuentes et al., 2016; d'Erfurth et al., 2010) (Fig. 2A). For example, CYCA1;2 in complex with CDKA;1 phosphorylates THREE-DIVISION MUTANT 1 (TDM1; also known as MALE-STERILE 5), a tetratricopeptide repeat protein that controls meiotic exit by suppressing translation (Cairo et al., 2022; Cifuentes et al., 2016; d'Erfurth et al., 2010). Moreover, CDKA;1 directly phosphorylates the chromosome axis protein ASYNAPTIC 1 (ASY1), which is involved in the formation of the chromosome axis during meiosis (Yang et al., 2020).
The activity of CYCB3;1, a B-type CYC, is required during meiosis I, where it localizes to the meiotic spindles to repress the premature onset of cell wall formation (Bulankova et al., 2013; Sofroni et al., 2020). This function is shared by the cyclin-related factor SOLO DANCERS (SDS), a highly divergent cyclin, the expression of which is restricted to male and female meiotic cells, where it is required for meiotic recombination and chromosome pairing (Azumi et al., 2002). SDS also forms a complex with CDKA;1 to phosphorylate SWITCH1 (SWI1), leading to SWI1 degradation (Yang et al., 2019). SWI1 inhibits WINGS APART-LIKE (WAPL), so the degradation of SWI1 allows WAPL to perform its function of removing cohesin from chromosomes (Yang et al., 2019).
Male gametogenesis
After meiosis, each haploid microspore undergoes mitosis to generate the vegetative cell and the germ cell (Twell, 2011). In Arabidopsis, the acquisition of vegetative cell identity relies on the activity of RBR1, and rbr1 vegetative cells retain microspore features (such as the ability to divide symmetrically) (Chen et al., 2009). The vegetative cell must arrest its cell cycle to prevent further divisions, and this process is governed by the cell cycle inhibitors RBR1 and KRPs (likely KRP1,3,4,6,7) (Chen et al., 2009; Kim et al., 2008; Zhao et al., 2012) (Fig. 2B). Meanwhile, germ cell identity is regulated by the germline-specific MYB transcription factor DUO1 POLLEN1 (DUO1), which activates CYCB1;1 to initiate mitosis (Brownfield et al., 2009) (Fig. 2C). Mutations in DUO1 lead to germ cells that are unable to divide, resulting in a significant proportion of blocked divisions (Brownfield et al., 2009).
The germ cell undergoes two mitotic divisions, which in Arabidopsis pollen are primarily regulated by CDKA;1 and factors modulating its activity (Aw et al., 2010; Iwakawa et al., 2006; Nowack et al., 2006; Zhao et al., 2012) (Fig. 2D). Heterozygous mutants of CDKA;1 develop pollen grains arrested at the bicellular state, where the vegetative cell develops properly but the generative cell fails to undergo mitosis to produce the two sperm cells (Iwakawa et al., 2006; Nowack et al., 2006). Interestingly, in some cdka;1 mutant pollen grains, mitosis occurs after pollen germination and pollen tube elongation (Aw et al., 2010). Alongside CDKA;1, CDKB1;1 and CDKB1;2 also have a role in pollen development (Nowack et al., 2012). While pollen development in single and double cdkb1;1 cdkb1;2 mutants is phenotypically similar to wild-type Arabidopsis, the triple mutant lacking CDKA;1, CDKB1;1 and CDKB1;2 activity (CDKA;1/cdka;1 cdkb1;1/cdkb1;1 cdkb1;2/cdkb1;2) produces a small proportion of pollen grains that fail to undergo any mitotic division (Nowack et al., 2012). The second mitotic division is also controlled by FBL17, which targets KRP6 and KRP7 for proteasome-mediated degradation (Gusti et al., 2009; Kim et al., 2008; Noir et al., 2015; Zhao et al., 2012). Mutations in fbl17 lead to pollen arrest at the bicellular stage because the sustained levels of KRP6 and KRP7 inhibit CDKA;1 activity and thus cell cycle progression (Gusti et al., 2009; Kim et al., 2008; Zhao et al., 2012).
Female reproductive development
In flowering plants, the mature female gametophyte contains two female gametes known as the egg cell and the central cell (Yadegari and Drews, 2004). During the process of double fertilization, two sperm cells fertilize the egg and the central cell to generate the zygote and the endosperm, a triploid tissue that provides nutrients to the embryo during its growth (Berger, 1999).
In Arabidopsis, female gametophyte development begins when an archesporial cell enlarges and differentiates into a megaspore mother cell (Grossniklaus and Schneitz, 1998; Schneitz et al., 1995; Yadegari and Drews, 2004). This cell, identifiable as a single, large and elongated subepidermal cell, undergoes meiosis to yield four megaspores, of which three undergo programmed cell death (Fig. 3) (Schneitz et al., 1995; Yadegari and Drews, 2004). The surviving megaspore develops into the functional megaspore, which undergoes three rounds of mitosis without cytokinesis to generate eight haploid nuclei (Fig. 3) (Grossniklaus and Schneitz, 1998; Yadegari and Drews, 2004). Cellularization ensues, culminating in the maturation of the female gametophyte: one egg cell, two synergid cells, three antipodal cells, and a central cell formed by the fusion of two haploid polar nuclei (Fig. 3) (Grossniklaus and Schneitz, 1998; Yadegari and Drews, 2004).
Female reproductive development in Arabidopsis. (A,B) Graphical summary of key events during Arabidopsis female reproductive development, and of the cell cycle regulators involved in the meiosis and mitosis events. The megaspore mother cell undergoes meiosis (A), forming a tetrad of haploid spores. KRPs inhibit CDKA;1, allowing RBR1 to accumulate and halt cell cycle progression by directly repressing WUS expression; this absence of WUS in the megaspore mother cell is crucial for limiting mitotic divisions and initiating meiosis. After meiosis, three spores degenerate, while the remaining spore undergoes three rounds of mitosis without cytokinesis (B). The mitotic divisions are orchestrated by CDKA;1–CYCD complexes during G1-S transitions and CDKA-B–CYCB complexes during G2-M transitions. Cellularization then takes place to produce one egg cell (magenta), two synergids (green), three antipodal cells (grey), and a central cell (yellow). The central cell is formed by the fusion of two haploid polar nuclei. APC/C, ANAPHASE PROMOTING COMPLEX; CDKA, CYCLIN-DEPENDENT KINASE A; CYCD, CYCLIN D-type; E2F, E2F TRANSCRIPTION FACTOR; CAK, CDK-ACXTIVATING KINASE; KRP, KIP-RELARED PROTEIN; RBR1, RETINOBLASTOMA RELATED 1; RFCs, REPLICATION FACTORS; WUS, WUSCHEL; P, phosphorylation; Ub, ubiquitination.
Female reproductive development in Arabidopsis. (A,B) Graphical summary of key events during Arabidopsis female reproductive development, and of the cell cycle regulators involved in the meiosis and mitosis events. The megaspore mother cell undergoes meiosis (A), forming a tetrad of haploid spores. KRPs inhibit CDKA;1, allowing RBR1 to accumulate and halt cell cycle progression by directly repressing WUS expression; this absence of WUS in the megaspore mother cell is crucial for limiting mitotic divisions and initiating meiosis. After meiosis, three spores degenerate, while the remaining spore undergoes three rounds of mitosis without cytokinesis (B). The mitotic divisions are orchestrated by CDKA;1–CYCD complexes during G1-S transitions and CDKA-B–CYCB complexes during G2-M transitions. Cellularization then takes place to produce one egg cell (magenta), two synergids (green), three antipodal cells (grey), and a central cell (yellow). The central cell is formed by the fusion of two haploid polar nuclei. APC/C, ANAPHASE PROMOTING COMPLEX; CDKA, CYCLIN-DEPENDENT KINASE A; CYCD, CYCLIN D-type; E2F, E2F TRANSCRIPTION FACTOR; CAK, CDK-ACXTIVATING KINASE; KRP, KIP-RELARED PROTEIN; RBR1, RETINOBLASTOMA RELATED 1; RFCs, REPLICATION FACTORS; WUS, WUSCHEL; P, phosphorylation; Ub, ubiquitination.
Female sporogenesis
Multiple subepidermal cells in the early ovule primordium possess the potential to become a megaspore mother cell; however, as the ovule develops, the number of cells capable of adopting such a fate is progressively reduced until only one cell attains the megaspore mother cell status (Grossniklaus and Schneitz, 1998). Positional cues, influenced by genetic and epigenetic networks and gradients of phytohormones, appear to orchestrate the specification of both the archesporial cell and the megaspore mother cell (Ceccato et al., 2013; She et al., 2013; Hernandez-Lagana et al., 2021; Huang et al., 2022). KRP4, KRP6 and KRP7 redundantly act within the megaspore mother cell to inhibit CDKA;1 activity so that RBR1 accumulates and represses cell cycle progression through direct suppression of WUSCHEL (WUS) expression (Zhao et al., 2017; Fig. 3A). This absence of the transcription factor WUS in the designated megaspore mother cell is essential for limiting the number of mitotic divisions and triggering meiosis (Zhao et al., 2017). Indeed, female gametophytes lacking RBR1 or KRP function (such as in the rbr1 mutant or the krp1-7 septuple mutant) produce multiple megaspore mother cells, indicating that these megaspore mother cells are progressing into mitosis rather switching to the meiotic cycle (Cao et al., 2018; Zhao et al., 2017).
Female gametogenesis
Following megaspore mother cell specification, all the steps required to produce a functional female gametophyte are orchestrated by RBR1, CDKs and CYCs, and by the individual expression levels and protein turnover (by proteasome-mediated degradation) of these factors (Fig. 3B). In the triple mutant CDKA;1/cdka;1 cdkb1;1/cdkb1;1 cdkb1;2/cdkb1;2, megaspore mother cells can successfully undergo meiosis, but mitosis is severely compromised and ∼50% of embryo sacs arrest at various stages, meaning that they lack a cellularized egg apparatus and correct cellular identity (Nowack et al., 2012).
Lack of CDK activation by CYCs or CAKs also impairs female gametophyte development. Indeed, arrested development, tissue degeneration, and embryo sacs containing one, two, or four nuclei that fail to cellularize have been observed in multiple mutants for the D-type CYCs cycd2;1 cycd3;3 (Zhang et al., 2022), in the absence of three or four B-type CYCs (cycb1;1/cycb1;1 CYCB1;2/cycb1;2 cycb1;4/cycb1;4, and cycb1;1/cycb1;1 CYCB1;2/cycb1;2 CYCB1;3/cycb1;3 cycb1;4/cycb1;4; Romeiro Motta et al., 2022), and in the double mutant of the two CAKs CDKD;1 and CDKD;3 (cdkd;1-1 cdkd;3-1, Takatsuka et al., 2015).
Progression of the megaspore mother cells through the mitotic cycle requires DNA synthesis, so absence of subunits of the DNA replication machinery such as REPLICATION FACTORS (RFCs) also impairs the division of the megaspore mother cell (Liu et al., 2013; Wang et al., 2012a,b).
Finally, mutations in some of the APC/C subunits also trigger female gametogenesis defects, because B-type CYCs are not efficiently degraded and thus mitosis cannot proceed normally (Pines, 2011). Phenotypes such as embryo sacs arrested at the one-cell stage, mis-positioned nuclei, abnormal nuclear number and degradation of the nuclei or embryo sac, have been observed in mutants for APC1 (Wang et al., 2013), APC2 (Capron, 2003), CDC27A and CDC27B (HOBBIT) (Blilou et al., 2002; Pérez-Pérez et al., 2008), APC4 (Wang et al., 2012a,b), NOMEGA (APC6) (Kwee and Sundaresan, 2003), APC13 (Saze and Kakutani, 2007) and APC10 (Eloy et al., 2011).
Cell cycle stage in the mature gametes: awaiting fertilization
Regulation of cell cycle progression is also important to ensure that mature gametes are arrested at a cell cycle stage that will allow synchrony between male and female gametes when they meet at karyogamy (Friedman, 1999). Thus, attaining and maintaining quiescence is extremely important to prevent unwanted division in the gametes, a condition that can significantly impact an organism's fitness, waste energy and resources, and ultimately can lead to the abortion of progeny (Adhikari et al., 2010, 2009; Reddy et al., 2010).
In most animal species, mature sperm arrest in the G1 phase, whereas mature oocytes are arrested in metaphase of meiosis II (Griswold, 2016; Pei et al., 2023). In plants, the cell cycle phase at which gamete cells arrest varies by species, meaning that karyogamy in seed plants can take place in three different ways: G1 karyogamy, where the zygote DNA is replicated post-fusion; S-phase karyogamy, where the two nuclei go through S-phase before fusing; and G2 karyogamy, where nuclei fuse and then enter M-phase (Carmichael and Friedman, 1995).
Classical techniques to determine the cell cycle stage of the gametes involve direct DNA staining methods such as DAPI, Propidium Iodide or Feulgen staining (Box 1). More recent strategies, such as feeding plants with fluorescently-labelled nucleotide analogues or analysing the dynamics of cell cycle component expression (e.g. CDT1a, PCNA, LIG1 and ORCs for S-phase, CYC B-type for mitosis), have further enhanced our understanding of DNA synthesis and cell cycle stages (Desvoyes et al., 2020; Echevarría et al., 2021). Numerous studies have sought to elucidate the cell cycle stage of mature plant gametes, with a historical emphasis on sperm cells due to their relative ease of isolation and analysis compared to the more challenging egg and central cells. Different research groups have reported conflicting results regarding the cell cycle stages of plant gametes, particularly in Arabidopsis (Table 1). For example, Friedman and others have reported that Arabidopsis sperm nuclei begin DNA synthesis soon after the second mitotic division of the generative cell (Durbarry et al., 2005; Friedman, 1999; Kim et al., 2008; Rotman et al., 2005), and that DNA replication continues as the sperms travel in the pollen tube, reaching the G2 stage with a 2C DNA content just before karyogamy (Table 1; Fig. 4). This conclusion is supported by gene expression studies showing that genes required for active DNA synthesis are expressed during pollen germination and pollen tube growth (Borges et al., 2008; Pina et al., 2005). Conversely, recent studies by Liu et al. and Voichek et al. found no evidence of DNA replication in mature sperm cells through EdU incorporation and DNA sequencing, suggesting that they do not enter S-phase at maturation (Liu et al., 2020; Voichek et al., 2023) (Table 1; Fig. 4). These conflicting results might be explained by the different techniques used. DAPI staining, employed in Friedman's study, may not reliably measure DNA content in sperm nuclei because of the extreme condensation of chromatin in these cells, potentially leading to inaccuracies in determining the cell cycle phase (Flores-Tornero and Becker, 2023; Liu et al., 2020). On the other hand, both the EdU staining and DNA sequencing techniques adopted in Liu et al. (2020) and Voichek et al. (2023) were conducted on sperm cells collected from pollen tubes grown in vitro, while DAPI staining by Friedman was done on whole flowers at various developmental stages. Pollen tubes grown by in vitro methods often fail to fertilize ovules, suggesting that the growth of the pollen tube through the female tissue is necessary for successful fertilization (Desnoyer and Grossniklaus, 2023). Therefore, semi-in vitro techniques might miss the ideal conditions for proper pollen activation and possibly also cell cycle progression.
DAPI (4′,6-diamidino-2-phenylindole) staining: DAPI is a very bright blue fluorescent DNA stain that associates with the minor groove of double-stranded DNA, with a preference for AT-rich regions. Its relative brightness matches to the abundance of nuclear DNA, and thus is used to quantify nuclear ploidy, from which the cell cycle stage is inferred. DAPI binds specifically to DNA and not RNA, so generates very little background signal. An example of applying this technique to Arabidopsis pollen is Borg et al. (2019).
PI (Propidium Iodide) staining: PI is a red-coloured DNA stain, which binds to double-stranded DNA by intercalating between base pairs. Its brightness is relative to the DNA content of a nucleus. PI binds nucleic acid in general, including RNA, so the staining procedure includes extensive RNase treatment to degrade RNA. An example of applying this technique to Arabidopsis ovules is She et al. (2013).
Feulgen staining: Feulgen staining specifically marks DNA. It liberates aldehydes from the deoxy sugar, allowing them to react with fuchsin-sulphurous acid and yield a magenta colour. However, chemicals present in plant tissues such as oils, resins, or tannins interfere with Feulgen staining detection, potentially leading to an overestimation of DNA quantity. An example of applying this technique to maize ovules is Barrell and Grossniklaus (2005).
EdU (5-ethynyl-2′-deoxyuridine): EdU is a nucleoside analogue of thymidine and is incorporated into newly synthesized DNA during S-phase. Its detection uses click-chemistry, where the EdU alkyne group undergoes a copper catalyzed click reaction with the azide group on an Alexa Fluor probe. EdU incorporation is used to assess the G1-S transition and S-phase. An example of applying this technique to Arabidopsis ovules is Simonini et al. (2024).
DNA sequencing: The quantity of DNA doubles from the G1 to the G2 phase, resulting in replicated genomic regions generating double the number of sequencing reads compared to non-replicated regions. If some genomic regions are doubled, it indicates that the cell is in the S phase. However, for this technique to be informative, it requires a comparison between non-replicated and replicated regions. DNA sequencing can provide evidence of DNA synthesis but not of ploidy, meaning it cannot distinguish whether a cell is in the G1 or G2 phase. An example of applying this technique to Arabidopsis pollen is Voichek et al. (2023).
Marker lines: Cell division can be monitored in vivo by visualizing the dynamics of constitutive nuclear markers, such as histone variants (H2B or H3) fused to fluorescent proteins. Specific cell cycle phases can also be detected by observing the accumulation, dynamics (e.g. diffuse signal versus speckles), or degradation of proteins that are specific to those phases. For example, expression of CYCB1;1 indicates mitosis, whereas PCNA condensation in speckles indicates the presence of replication foci during S phase, and CDT1a degradation happening at the G1-S boundary marks the beginning of G1-S transition. Some examples are found at Desvoyes et al. (2020), Echevarría et al. (2021), Voichek et al. (2023) and Simonini et al. (2024).
Alternative models of the cell cycle stage of Arabidopsis gametes at maturity, fertilization and karyogamy. Schematics summarizing alternative models (proposed by Friedman, 1999; Liu et al., 2020; Voichek et al., 2023 and Simonini et al., 2024) of the cell cycle stage of Arabidopsis female and male gametes at maturity, at the moment of fertilization and at karyogamy. Conflicting evidence has led to different models of the state of mature gametes (left panel): central cells (yellow) have been proposed to be in G1 (according to Liu et al. and Voichek et al.) or arrested S (according to Friedman et al. and Simonini et al.); egg cells (red) have been proposed to be in G0 (Voichek et al.), G1 (Liu et al.) or G2 (Friedman et al. and Simonini et al.); sperm cells (green) have been proposed to be in G1 (Liu et al. and Voichek et al.) or G2 (Friedman et al. and Simonini et al.). The progression of events during fertilization and karyogamy (middle and right panels) is closely tied to the cell cycle stages of the fusing gametes. If both gametes are in G1 or G0, DNA synthesis is initiated immediately after fusion, as suggested by studies such as Liu et al. and Voichek et al. For G2-phase gamete fusion, the fertilized gamete transitions directly into mitosis, as proposed in findings by Simonini et al. and Friedman et al. In scenarios where one gamete is in S-phase and the other is in G2, the S-phase gamete must first complete DNA replication to reach G2 before mitosis can proceed, a mechanism described in the context of the central cell by Simonini et al.
Alternative models of the cell cycle stage of Arabidopsis gametes at maturity, fertilization and karyogamy. Schematics summarizing alternative models (proposed by Friedman, 1999; Liu et al., 2020; Voichek et al., 2023 and Simonini et al., 2024) of the cell cycle stage of Arabidopsis female and male gametes at maturity, at the moment of fertilization and at karyogamy. Conflicting evidence has led to different models of the state of mature gametes (left panel): central cells (yellow) have been proposed to be in G1 (according to Liu et al. and Voichek et al.) or arrested S (according to Friedman et al. and Simonini et al.); egg cells (red) have been proposed to be in G0 (Voichek et al.), G1 (Liu et al.) or G2 (Friedman et al. and Simonini et al.); sperm cells (green) have been proposed to be in G1 (Liu et al. and Voichek et al.) or G2 (Friedman et al. and Simonini et al.). The progression of events during fertilization and karyogamy (middle and right panels) is closely tied to the cell cycle stages of the fusing gametes. If both gametes are in G1 or G0, DNA synthesis is initiated immediately after fusion, as suggested by studies such as Liu et al. and Voichek et al. For G2-phase gamete fusion, the fertilized gamete transitions directly into mitosis, as proposed in findings by Simonini et al. and Friedman et al. In scenarios where one gamete is in S-phase and the other is in G2, the S-phase gamete must first complete DNA replication to reach G2 before mitosis can proceed, a mechanism described in the context of the central cell by Simonini et al.
Cell cycle stage of mature gametes before karyogamy
Plant species . | Sperm cell cycle stage . | Egg cell cycle stage . | Central cell cycle stage . | Reported in . |
---|---|---|---|---|
Brown algae | G1 | G1 | – | McCully, 1968 |
Green algae | G1 | G1 | – | Matsuda et al., 1990 |
Red algae | G1 | G1 | – | Goff and Coleman, 1984 |
Gnetum gnemon | G2 | G2 | – | Carmichael and Friedman, 1995 |
Chlorophytum elatum | S | – | – | Ermakov et al., 1980 |
Ligularia dentata | S | – | – | Ermakov et al., 1980 |
Crepis capillaris | G2 | – | – | Ermakov et al., 1980 |
Elytrigia elongata | G1 | – | – | Ermakov et al., 1980 |
Nuphar | G1 | – | – | Williams and Friedman, 2002 |
Illicium | G1 | – | – | Williams and Friedman 2004 |
Kadsura | G1 | – | – | Friedman et al., 2003 |
Tradescantia paludosa | G1 | G1 | – | Woodard, 1956 |
Lycium barbarum | G2 | G2 | – | Deng et al., 2012 |
Helleborus bocconei | G2 | G2 | G2 | Bartoli et al., 2017 |
Ginko biloba | – | G2? | G2? | Avanzi and Cionini, 1971 |
Vigna unguiculata | – | G1 | G2 | Gursanscky et al., 2020 preprint |
Ornithogalum caudatum | G2 | G2 | G1 | Morozova, 2002 |
Galanthus nivalis | – | G1 | – | Bannikova et al., 1985 |
Allium odorum | G2 | G2 | – | Morozova and Ermakov, 1993 |
Arabidopsis thaliana | G2 | – | – | Friedman, 1999 |
G1 | G0 | G1 | Voichek et al., 2023 | |
G1 | G1 | G1 | Liu et al., 2020 | |
G2 | G2 | Arrested S | Simonini et al., 2024 | |
– | G1 | G2 | Berger et al., 2008 | |
Zea mays | G1 | – | – | Bino et al., 1990 |
G1 | G1 | – | Mogensen et al., 1995 | |
G1 | G0 | – | Chen et al., 2017 | |
G1 | – | – | Swift, 1950 | |
Hordeum vulgare | G1 | G1 | – | Mogensen and Holm, 1995 |
– | G1 | G1 | Bennett and Smith, 1976 | |
G2 | G2 | – | D'Amato et al., 1965 | |
Torenia fournieri | G1 | – | – | Liu et al., 2020 |
Oryza sativa | G1 | – | – | Sukawa and Okamoto, 2018 |
Triticum aestivum | G1 | – | – | Pónya et al., 1999 |
– | G1 | – | Bannikova et al., 1985 | |
Nicotiana tabacum | G1 or S | G1 | – | Tian et al., 2005 |
G2 | – | – | D'Amato et al., 1965 | |
S | – | – | Hesemann, 1973 | |
– | G1 | – | Bannikova et al., 1985 | |
Lilium regale | – | G1 | – | Bannikova et al., 1985 |
Lilium longiflorum | G2 | – | – | Bino et al., 1990 |
Plant species . | Sperm cell cycle stage . | Egg cell cycle stage . | Central cell cycle stage . | Reported in . |
---|---|---|---|---|
Brown algae | G1 | G1 | – | McCully, 1968 |
Green algae | G1 | G1 | – | Matsuda et al., 1990 |
Red algae | G1 | G1 | – | Goff and Coleman, 1984 |
Gnetum gnemon | G2 | G2 | – | Carmichael and Friedman, 1995 |
Chlorophytum elatum | S | – | – | Ermakov et al., 1980 |
Ligularia dentata | S | – | – | Ermakov et al., 1980 |
Crepis capillaris | G2 | – | – | Ermakov et al., 1980 |
Elytrigia elongata | G1 | – | – | Ermakov et al., 1980 |
Nuphar | G1 | – | – | Williams and Friedman, 2002 |
Illicium | G1 | – | – | Williams and Friedman 2004 |
Kadsura | G1 | – | – | Friedman et al., 2003 |
Tradescantia paludosa | G1 | G1 | – | Woodard, 1956 |
Lycium barbarum | G2 | G2 | – | Deng et al., 2012 |
Helleborus bocconei | G2 | G2 | G2 | Bartoli et al., 2017 |
Ginko biloba | – | G2? | G2? | Avanzi and Cionini, 1971 |
Vigna unguiculata | – | G1 | G2 | Gursanscky et al., 2020 preprint |
Ornithogalum caudatum | G2 | G2 | G1 | Morozova, 2002 |
Galanthus nivalis | – | G1 | – | Bannikova et al., 1985 |
Allium odorum | G2 | G2 | – | Morozova and Ermakov, 1993 |
Arabidopsis thaliana | G2 | – | – | Friedman, 1999 |
G1 | G0 | G1 | Voichek et al., 2023 | |
G1 | G1 | G1 | Liu et al., 2020 | |
G2 | G2 | Arrested S | Simonini et al., 2024 | |
– | G1 | G2 | Berger et al., 2008 | |
Zea mays | G1 | – | – | Bino et al., 1990 |
G1 | G1 | – | Mogensen et al., 1995 | |
G1 | G0 | – | Chen et al., 2017 | |
G1 | – | – | Swift, 1950 | |
Hordeum vulgare | G1 | G1 | – | Mogensen and Holm, 1995 |
– | G1 | G1 | Bennett and Smith, 1976 | |
G2 | G2 | – | D'Amato et al., 1965 | |
Torenia fournieri | G1 | – | – | Liu et al., 2020 |
Oryza sativa | G1 | – | – | Sukawa and Okamoto, 2018 |
Triticum aestivum | G1 | – | – | Pónya et al., 1999 |
– | G1 | – | Bannikova et al., 1985 | |
Nicotiana tabacum | G1 or S | G1 | – | Tian et al., 2005 |
G2 | – | – | D'Amato et al., 1965 | |
S | – | – | Hesemann, 1973 | |
– | G1 | – | Bannikova et al., 1985 | |
Lilium regale | – | G1 | – | Bannikova et al., 1985 |
Lilium longiflorum | G2 | – | – | Bino et al., 1990 |
Discrepancies have also emerged in determining the cell cycle stage of egg and central cells, possibly due to the technical challenges associated with collecting these types of cells; female gametes develop inside the ovules, protected by layers of maternal tissues, and are difficult to isolate in sufficient numbers for standard ploidy analysis, such as flow cytometry. Consequently, the commonly used methods for assessing the ploidy and cell cycle stages of female gametes rely on DNA staining, observation of the expression dynamics of cell cycle markers and incorporation of nucleotide analogues (Barrell and Grossniklaus, 2005; Liu et al., 2020; She et al., 2013; Simonini et al., 2024; Voichek et al., 2023) (Table 1; Fig. 4). Two recent studies have explored the cell cycle stages at which Arabidopsis egg and central cells pause before fertilization. Voichek et al. (2023) propose that the mature egg cell and central cell are arrested in the G0 and G1 phases, respectively, without replicating their DNA before fertilization. This conclusion is supported by the absence of markers for the G1-phase, pre-replication, or mitosis, along with the diffuse fluorescence of S-phase markers such as PCNA1-GFP (which indicates a lack of DNA replication) in the mature egg cell. The authors observe a similar expression pattern in the central cell, with a general absence of most markers, except for weak expression of some G1-phase and pre-replication markers, from which they conclude that the central cell is in G1. However, Simonini et al. (2024) present an alternative view, proposing that the egg cell is in G2 phase, while the central cell is arrested in S-phase. These conclusions are supported by propidium iodide staining to quantify DNA content, EdU incorporation to detect DNA synthesis, and the use of various marker lines, in a strategy similar to that employed by Voichek and colleagues. The discrepancies between the two studies may stem from challenges in interpreting the expression profiles of the markers used. While these cell cycle markers perform well in tissues like roots, leaves and meristems (D'Ario et al., 2021; Desvoyes et al., 2020; Han et al., 2022), they often yield inconsistent or weak signals in the female gametophyte (Simonini et al., 2024; Voichek et al., 2023). This limitation highlights the need to complement marker-based analysis with additional techniques to accurately capture the dynamics in female gametes.
The molecular basis of cell cycle arrest in quiescent gametes has also been explored. Expression of the CYCB;1 marker is undetectable in mature female gametophytes awaiting fertilization, which is consistent with cell cycle arrest (Johnston et al., 2008). RBR1 plays a pivotal role in the establishment of the quiescent state by preventing divisions within the mature female gametophyte (Ebel et al., 2004; Johnston et al., 2008). Indeed, rbr1 mutant female gametophytes exhibit supernumerary nuclei at the micropylar end, likely originating from the overproliferation of the egg apparatus (Ebel et al., 2004; Johnston et al., 2008). Moreover, rbr1 mutant unfertilized central cells can initiate endosperm development autonomously, further indicating improper cell cycle arrest (Ebel et al., 2004; Ingouff et al., 2007; Johnston et al., 2008; Simonini et al., 2024). Thus, RBR1 functions as a brake that blocks cell cycle progression in the female gametes.
Autonomous division of unfertilized central cells can also be promoted by perturbations to factors beyond the core cell cycle components. Mutations in the four subunits of the Fertilization-Independent-Seed Polycomb Repressive Complex 2 (FIS-PRC2) [FERTILIZATION INDEPENDENT ENDOSPERM (FIE), FERTILIZATION INDEPENDENT SEED (FIS2), MULTICOPY SUPPRESSOR OF IRA1 (MSI1) and MEDEA (MEA) (Chaudhury et al., 1997; Grossniklaus et al., 1998; Kiyosue et al., 1999; Köhler et al., 2003; Ohad et al., 1996)] lead to autonomous division of the central cell and the formation of an endosperm-like structure. The absence of the PRC2 subunits MSI1 (in Arabidopsis) and FIE (in rice) triggers the formation of embryo-like structures, which originate from the autonomous division of the unfertilized egg cell (Guitton and Berger, 2005; Köhler et al., 2003; Wu et al., 2023). The molecular mechanism by which PRC2 regulates cell cycle arrest in female gametes remains largely unknown. It is hypothesized that PRC2 may directly regulate the gene expression of core cell cycle components, such as A- and D-type CYCs (Adhikari and Davie, 2020; Iovino et al., 2013; Johnston et al., 2008; Simonini et al., 2021). Additionally, PRC2 might interact directly with RBR1 through MSI1 (Ach et al., 1997b; Mosquna et al., 2004; Jullien et al., 2008), further influencing cell cycle control.
Despite these recent findings, much remains to be understood about cell cycle regulation in plant gametes. What mechanisms trigger cell cycle arrest in S-phase in the central cell? How do egg cells and central cells maintain their quiescent state over time? New technologies that require little input material, such single-cell technologies (Baysoy et al., 2023), could eventually help resolve some of these longstanding questions and offer a benchmark approach for studying cell cycle progression in plant gametes more comprehensively.
Cell cycle reactivation in the gametes at the moment of fertilization
In vertebrates, fully grown oocytes are arrested at the prophase of the first meiosis (prophase I). In the final stage of their maturation, they prepare for fertilization: meiosis is resumed but the egg arrests again at metaphase of the second meiosis (metaphase II). Quiescency in the egg is achieved by inactivation of the APC/C complex, so that destruction of CycB is prevented and cell cycle arrest is maintained (Hörmanseder et al., 2013). At fertilization, the sperm breaks this cell cycle arrest by delivering a phospholipase C protein, which generates calcium oscillations in the fertilized egg (Madgwick et al., 2005; Madgwick and Jones, 2007; Saunders et al., 2002). This Ca2+ surge activates the calmodulin-dependent protein kinase II (CamKII), which in turn activates the APC/C complex (Stricker, 1999). This results in the destruction of CycB1 and cell cycle progression (Nixon et al., 2002). Ca2+ oscillations in the egg are sufficient to initiate the entire signalling cascade required to reinstate cell division (Machaty, 2024). Indeed, parthenogenesis can be induced by adding calcium ionophores (molecules that cause the release of intracellular Ca2+) to sea urchin eggs (Steinhardt and Epel, 1974).
A calcium wave also occurs during fertilization in plants. Unfertilized egg and central cells have low calcium concentrations (Denninger et al., 2014; Hamamura et al., 2014). Upon pollen tube rupture and sperm discharge, a calcium spike pervades the entire embryo sac, including the egg and central cells (Denninger et al., 2014; Digonnet et al., 1997; Hamamura et al., 2014; Ngo et al., 2014). However, unlike in animals, calcium influx is not sufficient to induce parthenogenetic embryo development in plants, suggesting that additional factors, which might be delivered by the sperm, are required to activate cell division (Koltunow and Grossniklaus, 2003).
One such factor is the D-type cyclin CYCD7;1. The CYCD7;1 protein and its messenger RNA are carried in the sperm and delivered to the female gametes through karyogamy (Simonini et al., 2024). CYCD7;1 induces RBR1 degradation in the central cell, thus releasing the central cell from quiescence and allowing it to complete DNA replication (Simonini et al., 2024; Fig. 5). When a central cell receives pollen lacking CYCD7;1, RBR1 is not degraded, and so cell cycle arrest in the central cell persists (Simonini et al., 2024). This mechanism appears to be specific to the central cell, as the egg cell is not affected by absence of paternally-derived CYCD7;1, and RBR1 absence (either through mutation or induced degradation) does not stimulate egg cell cycle progression (Ebel et al., 2004; Ingouff et al., 2009; Johnston et al., 2008; Simonini et al., 2024). Thus, central cells and egg cells regulate their cell cycles differently.
Cell cycle reactivation in the central cell. Cell cycle progression in the central cell is halted by RBR1 activity (top panel). During karyogamy (bottom panel), the delivery of paternally produced CYCD7;1 protein and its messenger RNA forms an active CDKA;1–CYCD complex. This complex triggers RBR1 degradation, thereby enabling cell cycle progression.
Cell cycle reactivation in the central cell. Cell cycle progression in the central cell is halted by RBR1 activity (top panel). During karyogamy (bottom panel), the delivery of paternally produced CYCD7;1 protein and its messenger RNA forms an active CDKA;1–CYCD complex. This complex triggers RBR1 degradation, thereby enabling cell cycle progression.
Other paternally derived or paternally expressed factors have also been shown to regulate the cell cycle after fertilization. For example, paternal alleles of auxin biosynthetic genes of the YUCCA family are required for auxin production in the female gametes to induce endosperm proliferation (Figueiredo et al., 2016, 2015). Moreover, the transcription factor BABY BOOM (BBM) accumulates in the sperm, and its delivery is required to initiate embryogenesis (Khanday et al., 2023; Ren et al., 2024). Indeed, ectopic expression of BBM in plant egg cells induces autonomous proliferation of the egg and thus initiates embryogenesis (Boutilier et al., 2002; Chen et al., 2022; Conner et al., 2015; Khanday et al., 2023; Passarinho et al., 2008). The PARTHENOGENESIS (PAR) gene, identified in apomictic (asexually reproducing) dandelions, is sufficient to induce embryogenesis without fertilization when introduced into lettuce egg cells, a sexually reproducing species (Underwood et al., 2022). In sexually reproducing plants, PAR alleles are typically expressed in pollen, suggesting that the gene product acts at fertilization to lift a block on cell division (Underwood et al., 2022).
After karyogamy with the sperm nucleus, the fertilized egg cell and central cell initiate division at very different speeds. While the zygote takes about 16-20 h before committing its first division, the primary endosperm nucleus initiates mitotic divisions about 4-6 h after fertilization (Maruyama et al., 2015; Simonini et al., 2024). Mutations affecting subunits necessary for DNA replication lead to cell cycle arrest during early seed development. For example, mutations in subunits of the Origin Recognition Complex (ORC), in the Mini Chromosome Maintenance family members (MCMs) and in DNA replication licensing factors (RFCs), induce seed abortion, producing an embryo arrested at very early stages and an underdeveloped endosperm with extremely enlarged nuclei and disc-shaped or pitted nucleoli (Collinge et al., 2004; Herridge et al., 2014; Ni et al., 2009; Qian et al., 2018; Springer et al., 2000).
Cdk1 activity appears to be important for embryogenesis in animals; for example, Cdk1 knockout mice die before reaching the morula stage (Santamaría et al., 2007; Satyanarayana et al., 2008). Surprisingly, in plants, seeds that lack functional CDKA;1 can still develop fully, although they exhibit slower growth and fewer cell divisions compared to wild-type seeds (Nowack et al., 2012). Removing CDKB activity further enhances the cdka;1 mutant phenotype, with a small percentage of seeds from CDKA;1/cdka;1 cdkb;1/cdkb1;1 mutant plants containing embryos arrested at a range of different stages of seed development (Nowack et al., 2012).
Conclusions
For decades – if not centuries – biologists have investigated the processes that govern gamete development, reproduction, fertilization and embryogenesis. Central to the study of these fundamental aspects of developmental biology is the need to comprehend how cells function and divide. It is now clear that cell division is not merely a mechanism for increasing cell numbers within tissues or organs. Rather, it is intricately linked to tissue patterning, cellular identity acquisition, and the establishment of body axes and polarity. This regulatory role of cell division is especially prominent during plant development. Unlike animal cells, plant cells are encased in a rigid cell wall, rendering them immobile from the earliest stages of embryogenesis. As a result, plants rely heavily on spatially and temporally regulated cell divisions – often directional or asymmetric – to initiate identity acquisition, tissue and organ patterning, and the development of body axes throughout their entire life cycle.
In addition to understanding the molecular mechanisms that control cell division, it is equally important to explore the signals that trigger cell cycle arrest or reactivation. These processes are fundamental not only for reproduction but also in broader biological contexts, such as in the study of how certain types of diseases, such as cancer, arise. In the context of plant development, identifying the signals and key players that regulate cell division is crucial for enhancing reproductive strategies to secure global food production. One promising approach is apomixis, a form of asexual reproduction where seeds form without fertilization (Goeckeritz et al., 2024; Koltunow and Grossniklaus, 2003; Ozias-Akins and Conner, 2020). Engineering crops for apomixis could significantly boost yields and reduce reliance on traditional breeding methods. To accomplish this, it is necessary to bypass the natural quiescent state of plant gametes and activate the cell division pathways that typically govern embryo and endosperm development. Therefore, understanding the molecular mechanisms controlling cell cycle arrest and reactivation during plant reproduction brings us closer to achieving major breakthroughs in plant breeding and sustainable agriculture.
Over the past few decades, technological advancements have dramatically enhanced our ability to study cell division at both the molecular and cellular levels. Since the early yeast screenings of the 1970s, many of the core factors involved in cell cycle progression have been identified. However, in the context of reproduction, particularly in plants, much of the knowledge gained remains largely descriptive, and deep knowledge about the regulation of cell division at the molecular level during plant reproduction is still missing. For example, what are the signals that induce cell cycle arrest in mature plant gametes? What triggers the reactivation of the cell cycle? In the case of the central cell, we now know that its reactivation depends on the paternal delivery of core cell cycle components, but what about the egg cell? Understanding how these distinct gametes maintain their arrested states and what mechanisms control their reactivation remains an open question in plant reproductive and developmental biology. Technological innovation will be pivotal in answering these questions. Recently developed single-cell technologies allow us to achieve cellular resolution and understand the transcriptomes of individual cells, shedding light on their status during different stages of development. However, since cell cycle progression and regulation primarily occur at the protein level, there is still the need for methods that can detect protein species, their abundance from limited sample sizes and also their associated post-translational modification, such as phosphorylation and ubiquitination. The coming years could see the development and optimization of game-changing techniques, such as spatial proteomics. This would enable us to monitor protein accumulation, degradation and interactions between cell cycle components with cellular resolution – all without the need to isolate cells from their native environment. Plants present an ideal model system for applying these cutting-edge technologies, particularly for studying cell cycle dynamics during gamete development and reproduction. Unlike animals, where gametes are formed during embryonic development, plant gametes develop at much later stages, making plants an excellent system for studying the early steps of gametogenesis in real time. Additionally, plants produce gametes in abundance, providing a significant advantage when considering sample sizes and the reproducibility of experimental results. Another key benefit is that mutations affecting cell cycle components, which often induce abortion of the embryo in animals, are viable in plants, allowing us to study their effects on reproduction and development. Even in cases where mutations prove lethal in plants, it is still possible to analyze segregating populations to assess the impact of these mutations during gametogenesis. This flexibility offers unique opportunities to dissect the role of specific cell cycle regulators and their broader implications during development. The knowledge gained from such studies will bridge the gap between purely descriptive observations and a deeper, mechanistic understanding of how cell division is regulated during the development of multicellular organisms.
Acknowledgements
I thank Celia Baroux, Stefano Bencivenga, Nina Chumak and Edouard Tourdot for useful comments on the manuscript.
Footnotes
Funding
This manuscript is supported by a Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung PRIMA grant to S.S. (PR00P3_201653). Open Access funding provided by Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung. Deposited in PMC for immediate release.