Plant organelle division orchestrated by the mosaic machinery of endosymbiotic relics and eukaryotic host factors

Abstract

Plastids retain the imprint of their cyanobacterial ancestry by dividing primarily through binary fission. This process is executed by elaborate division machinery that has continued to evolve since endosymbiosis. Early identification and characterization of the Arabidopsis accumulation and replication of chloroplasts (arc) mutants paved the way for the molecular dissection of the plastid division apparatus, whose components have both prokaryotic ancestral and eukaryotic host origins. Molecular and cell biological studies revealed that plastid division operates through interdependent interactions among the Min-system positioning factors, the stromal FtsZ ring, envelope-associated ARC6, PARC6, and PDV proteins, and cytosolic ARC5/DRP5B proteins, which act in concert to position the division site and drive plastid separation under developmental and environmental regulation. Comparisons with mitochondrial and peroxisomal proliferation unveil shared division factors among these metabolically connected organelles. Insights into plant organelle division may inform strategies for the manipulation of organelle abundance to improve crop quality and resilience.

Introduction

During cell expansion and division, subcellular metabolic compartments (organelles) scale with the changing cellular space, maintaining their number, size, and distribution to ensure proper cellular functions (Sheahan et al. 2007). Plastids are organelles that originated from a free-living cyanobacterium that became an endosymbiont within a eukaryotic host cell approximately 1.5 billion years ago (Keeling 2010). Extensive research has been performed to elucidate mechanisms underlying the division of plastids, focusing on chloroplasts, the paramount photosynthetic organelles. This review focuses primarily on plastid division, using chloroplast division as the central model. We first highlight 4 publications in Plant Physiology in the 1990s that laid the genetic foundation for chloroplast division research (Pyke and Leech 1992, 1994; Pyke et al. 1994; Robertson et al. 1996). We then summarize the suite of follow-up molecular biology studies that led to the elaborate model of the highly coordinated chloroplast division machinery, which consists of components arising from both the cyanobacterial ancestor and the eukaryotic host during evolution. We also provide an overview of the division of mitochondria and peroxisomes—organelles that are physically and metabolically associated with chloroplasts and share some division factors. Finally, we discuss the potential use of organelle abundance manipulation to improve crop performance.

Identification and characterization of chloroplast division mutants

Before the molecular genetics era, how chloroplasts divide during leaf cell expansion had already been described through cell biological studies. As semiautonomous organelles containing their own genomes, plastids cannot be produced de novo but instead arise from pre-existing plastids, mainly through binary fission (Leech et al. 1981). During fission, the chloroplast narrows at its midpoint, forming an isthmus where tori can be observed on both sides of the envelopes. Then, the 2 halves of the dividing chloroplast twist in opposite directions around the central isthmus. As constriction proceeds, the chloroplast is eventually separated into 2 (Boasson et al. 1972; Platt-Aloia and Thomson 1977; Leech et al. 1981; Hashimoto 1986; Oross and Possingham 1989). However, the molecular genetic basis of this process remained unclear.

The emergence of Arabidopsis thaliana as a formidable genetic system empowered the use of forward genetics to identify genes controlling chloroplast division. The Leech group developed a rapid image screening procedure to identify ethyl methanesulfonate-mutagenized Arabidopsis lines with altered chloroplast accumulation patterns, in which the candidates were confirmed visually by counting chloroplasts using differential interference contrast microscopy (Pyke and Leech 1991), resulting in the identification of accumulation and replication of chloroplasts (arc) mutants.

Among the first characterized arc mutants, arc1 accumulates more but smaller chloroplasts, whereas arc2 and arc3 contain fewer, larger chloroplasts (Fig. 1a) (Pyke and Leech 1992). Genetic analysis showed these phenotypes to be caused by mutations at single recessive nuclear loci, providing direct evidence that chloroplast number is regulated by nuclear genes. A compensatory relationship between chloroplast number and individual chloroplast size during mesophyll growth was also revealed, in which the total projected chloroplast area (ie chloroplast plan area) retains a similar percentage relative to mesophyll cell size across wild type and the arc mutants despite differences in chloroplast number. This study suggested that chloroplast expansion and division are largely independent processes but are also coordinated during leaf cell development (Pyke and Leech 1992).

Figure 1.

Phenotypes of plastid division mutants. (a) Isolated mesophyll cells from fully expanded first leaves of Arabidopsis wild type (WT) and 5 arc mutants viewed with differential interference contrast (DIC) optics. WT, arc1, arc2, arc3, and arc5 are in the A. thaliana (L.) Heynh. var. Landsberg erecta background, which contains approximately 120 chloroplasts per mesophyll cell, while arc6 is in the A. thaliana ecotype Wassilewskij background, which contains approximately 80 chloroplasts per mesophyll cell. Bars = 25 µm. Adapted from Pyke and Leech (1992, 1994) and Pyke et al. (1994); copyright © American Society of Plant Biologists. (b) FtsZ2-1 localization in mesophyll cells of A. thaliana WT, arc3, and arc6 mutants in the Col-0 background, observed by FtsZ2-1 immunofluorescence. Merged images show FtsZ2-1 signal in green and chlorophyll autofluorescence in magenta. Bars = 10 µm. Adapted from Du et al. (2025).

Mature mesophyll cells of the arc6 mutant contain only 1 to 3 greatly enlarged chloroplasts (Fig. 1a), which is even fewer than that in young postmitotic mesophyll cells, implicating the role of ARC6 in proplastid division (Pyke et al. 1994). The mutant also has curled and twisted leaves, reduced biomass, but retains fertility. The similarly low chloroplast number in arc6 guard cells further indicated that ARC6 acts broadly in chloroplast accumulation and that the arc6 lesion likely appears early in leaf primordia or apical development (Pyke et al. 1994).

The arc5 mutant contains enlarged and centrally constricted chloroplasts in mature mesophyll cells (Fig. 1a) (Pyke and Leech 1994). The greatly reduced chloroplast number is similar to the estimated number of wild-type young plastids in new postmitotic mesophyll cells, while far fewer than chloroplasts in the wild type at the same growth stage (Pyke and Leech 1994). Further observations of the arc5 plastids throughout development (Robertson et al. 1996) revealed that, unlike arc6, arc5 shows no obvious defect in proplastid morphology in shoot apical meristem and leaf primordia. Ultrastructural analysis of mesophyll chloroplasts further showed extremely narrow isthmi, visible torus, and distorted thylakoid membrane organization in the mutant, suggesting that arc5 chloroplasts can initiate division and proceed through constriction, but fail to complete plastid separation, thus placing ARC5 function at very late chloroplast division (Robertson et al. 1996).

Together, these early reports established a genetic framework for plastid division. Many arc mutants remain viable and fertile when grown under controlled experimental conditions despite dramatic changes in chloroplast number and size, indicating some level of flexibility in chloroplast abundance within a plant cell. The distinct phenotypes of the arc mutants suggested that ARC proteins play differential roles in plastid division.

Molecular dissection of the chloroplast division apparatus

To elucidate the molecular mechanisms of chloroplast division, 2 complementary strategies have been employed. Besides forward genetics, reverse genetics studies of plant homologs of prokaryotic cell division genes were also fruitful in identifying plastid division components retained from the cyanobacterial progenitor (Lutkenhaus et al. 1980; Osteryoung and Vierling 1995). These studies not only dissected the highly complex division machinery, but also elucidated how the components are recruited, stabilized, remodeled, and coupled at the division site.

Filamentous temperature-sensitive Z (FtsZ) proteins form the stromal FtsZ ring (Z ring)

The Z ring is composed of FtsZ proteins, tubulin-like GTPases that polymerize into protofilaments to form the ring on the stromal surface of the inner envelope membrane (IEM), marking the future division site and providing an internal scaffold for the assembly of the division apparatus (TerBush et al. 2013). Green-lineage FtsZs originated from a single cyanobacterial-type FtsZ gene that was transferred to the nuclear genome after endosymbiosis and underwent duplication into 2 conserved but functionally distinct FtsZ families, FtsZ1 and FtsZ2 (Osteryoung et al. 1998; Vitha et al. 2001). A Z-ring system was established, in which the dynamics- and stability-promoting functions of a single cyanobacterial FtsZ were partially separated between FtsZ1 and FtsZ2 (TerBush and Osteryoung 2012; Porter et al. 2023). Additional diversification occurred in different plant lineages: the land plant FtsZ2 family further duplicated into 2 largely functionally redundant genes, FtsZ2-1 and FtsZ2-2, while FtsZ3 is present in charophytes, a group of green algae closely related to land plants, as well as in bryophytes and lycophytes, but was lost in ferns and seed plants (Martin et al. 2009; Schmitz et al. 2009; Karamoko et al. 2011; Grosche and Rensing 2017). A recent study in the moss Physcomitrium patens suggests that FtsZ3 regulates both Z-ring assembly and chloroplast constriction (Lang et al. 2026). Depletion or elevated expression of FtsZ1 or FtsZ2 severely impairs chloroplast division, often leading to the formation of a few and enlarged chloroplasts in mesophyll cells (Osteryoung et al. 1998; Stokes et al. 2000; McAndrew et al. 2001; Vitha et al. 2001). Consistent with their distinct roles, FtsZ1 and FtsZ2 cannot substitute for each other even when total FtsZ abundance is restored, and their molar ratio remains close to 1:2 during leaf development, highlighting the importance of FtsZ identity and stoichiometry in Z-ring function (McAndrew et al. 2008; Schmitz et al. 2009).

Z-ring dynamics rely on the GTPase-dependent turnover of the FtsZ protofilaments. GTP binding promotes the assembly of soluble FtsZ subunits into protofilaments that form a composite GTPase active site, whereas GTP hydrolysis weakens the association between adjacent FtsZ subunits to allow GDP-bound subunits to dissociate from the ends of the protofilaments. This cycle maintains subunit turnover between protofilaments and the soluble FtsZ pool (Chen et al. 2018a). These dynamics suggest that chloroplast FtsZ protofilaments may undergo treadmilling like the bacterial system (Bisson-Filho et al. 2017; Yang et al. 2017), although this has not been confirmed. FtsZ filament dynamics are further shaped by the distinct properties of FtsZ1 and FtsZ2. In situ immunofluorescence labeling showed that antisense depletion of each paralog produced different filament patterns; that is, FtsZ2 formed disorganized, jagged filaments, rings, and spirals around enlarged chloroplasts in the FtsZ1-1 antisense plants, whereas FtsZ1 appeared mainly as short filaments and bright foci in the FtsZ2-1 antisense plants (Vitha et al. 2001). Consistent with these in situ patterns, heterologous yeast assays and in vitro biochemical studies support a model in which FtsZ2 forms relatively stable filaments, while the more dynamic FtsZ1 promotes FtsZ2 turnover and limits excessive FtsZ2 self-assembly during Z-ring remodeling (TerBush and Osteryoung 2012; Porter et al. 2021).

Differences in the C-terminus affect how FtsZ1 and FtsZ2 polymers associate with the chloroplast IEM. Higher plant FtsZ1 contains a short C-terminal motif that confers weak direct binding to the IEM but is not involved in interactions with IEM transmembrane proteins (Liu et al. 2022). By contrast, the FtsZ2 C-terminal peptide interacts with IEM transmembrane proteins ARC6 and PARALOG OF ARC6 (PARC6) but lacks direct membrane-binding activity (An et al. 2024). These complementary properties in filament dynamics and membrane association help explain how FtsZ1-FtsZ2 heteropolymers form a Z ring that is anchored to the IEM division site while still undergoing constant turnover and remodeling.

The Min system positions and remodels the Z ring

Because FtsZ proteins can self-assemble, their polymerization must be spatially restricted. The chloroplast Min proteins are derived from the cyanobacterial Min system involved in cell division, but their functions have been adapted to and further diversified in plants to affect Z-ring positioning and dynamics. In Escherichia coli, Z-ring placement is mediated by the MinCDE system, named after the minicell phenotype of min mutants. MinC inhibits FtsZ polymerization, and MinD recruits and activates MinC to form the inhibitory MinCD complex. MinE activates the ATPase activity of MinD, causing the release of MinD and MinC from the membrane and thereby driving the pole-to-pole oscillation of MinCD. As a result, MinC concentration is enriched near the cell poles while reduced at mid-cell, where the Z ring assembles (Ramm et al. 2019). Cyanobacteria also show MinCDE oscillation, but MinC additionally concentrates at mid-cell through a mechanism dependent on MinD and Cdv3, a feature not observed in E. coli (MacCready et al. 2017).

Plants have retained MinD and MinE, known as MinD1 and MinE1, respectively. MinC has been lost, and its inhibitory role is believed to be fulfilled by ARC3, a proposed functional analog of MinC with little sequence similarity (Shimada et al. 2004; Maple et al. 2007). ARC3 is a stromal protein that interacts with FtsZ2, serving as a negative regulator of Z-ring formation (Zhang et al. 2013). Active ARC3 enhances GTPase activity-dependent FtsZ2 debundling and disassembly of existing filaments, thus presumably inhibiting Z-ring assembly at non-division sites (Shaik et al. 2018). ARC3-mediated regulation is coordinated with MinD1 (ARC11) and MinE1 (ARC12) (Zhang et al. 2013), and with additional Min-related factors acquired during plant evolution, such as MCD1 (MULTIPLE CHLOROPLAST DIVISION SITE1) (Nakanishi et al. 2009). In Arabidopsis chloroplasts, ARC3 distribution appears more similar to the cyanobacterial pattern, but its oscillation has not been detected (Chen et al. 2019). Consistent with a role for MinD1 in preventing ectopic Z-ring assembly, Arabidopsis minD1/arc11 mutant produces chloroplasts with variable size and number and asymmetric constrictions. AtMinE1 overexpression phenocopies minD1, consistent with a role for MinE1 in counteracting MinD1-dependent inhibition at the proper division site (Colletti et al. 2000; Itoh et al. 2001; Fujiwara et al. 2008).

MCD1 is an IEM protein localizing to both a mid-chloroplast ring and puncta dispersed on the envelope membrane, recruiting MinD1 to regulate Z-ring positioning (Nakanishi et al. 2009). MCD1 is also required for ARC3- and MinE1-dependent regulation of Z-ring positioning, suggesting that it helps to integrate Min-related spatial control with membrane-associated FtsZ organization during division-site placement (Chen et al. 2018b).

The envelope-associated division complexes are coordinated by ARC6 and PARC6 together with PLASTID DIVISION1 (PDV1) and PDV2

ARC6 and PARC6 (CDP1/ARC6H) are bitopic IEM proteins that arose at different times during evolution. ARC6 is derived from the cyanobacterial cell-division protein Ftn2, whereas PARC6 appears to have originated in Charophytes, likely through duplication of an ancestral ARC6-like gene followed by divergence (Vitha et al. 2003; Glynn et al. 2009; Zhang et al. 2009; Ottesen et al. 2010; Zhang et al. 2016; Liu et al. 2024). The stromal region of ARC6 binds the FtsZ2 filaments, tethers the Z ring to the IEM, and inhibits the disassembly of GDP-bound FtsZ2 (Maple et al. 2005; Sung et al. 2018). PARC6 also directly interacts with FtsZ2 and promotes the turnover of FtsZ2 filaments when heterologously expressed in yeast (Chen et al. 2019).

Both ARC6 and PARC6 directly interact with the C-terminal MORN domain of ARC3 to release ARC3 autoinhibition, allowing ARC3 to access FtsZ and inhibit excessive FtsZ assembly at the division site (Chen et al. 2019; Du et al. 2025) (Fig. 1b). ARC6 therefore couples 2 opposing requirements of Z-ring regulation—by activating ARC3-mediated remodeling while stabilizing the FtsZ2 filaments. The J-like domain of ARC6 helps balance these activities by limiting ARC6-ARC3 interaction while promoting ARC6-FtsZ2 interaction. In heterologous yeast reconstitution assays, this regulation prevents ARC3 overactivation while maintaining Z-ring formation and enhancing Z-ring dynamics and constriction (Du et al. 2025). Similarly, PARC6-ARC3 interaction at the mid-plastid division site may tune ARC3 activity rather than simply blocking Z-ring assembly (Chen et al. 2019). Beyond this ARC3-centered regulatory module, ARC6 also provides a link to Min-dependent division-site positioning through its direct interaction with AtMinD1 (Zhang et al. 2021). ARC6 also interacts with MCD1, which binds FtsZ2 in an ARC6-dependent manner (Chen et al. 2018b).

In the intermembrane space (IS), outer envelope membrane (OEM) proteins PDV1 and PDV2, which also originated in charophytes, interact with PARC6 and ARC6 respectively, linking the inner-envelope division machinery to cytosolic fission factors (Chen et al. 2018a; Liu et al. 2024). These interactions recruit PDV1 and PDV2 to the chloroplast division site and transmit positional information from the internal Z-ring system to the outer envelope (Glynn et al. 2008, 2009). In turn, PDV proteins may also help to organize inner-envelope division complexes. In pdv1, pdv2, and pdv1 pdv2 mutants, FtsZ appears as multiple rings or spirals at the constricted regions of enlarged, dumbbell-shaped chloroplasts (Miyagishima et al. 2006). Structural models further suggest that an ARC6 dimer could bind 2 adjacent GDP-bound FtsZ2 subunits through its stromal region, thereby stabilizing the FtsZ protofilaments (Sung et al. 2018). The PDV2 dimer can promote ARC6 dimerization before forming an ARC6-PDV2 heterotetramer and may therefore help condense inner-envelope division proteins at the division site (Wang et al. 2017). PDV1 can also induce PARC6 dimerization (Sun et al. 2023). In the pdv1 mutant the abundance of PARC6 is strongly reduced, indicating that PDV1 is important for PARC6 stability (Sun et al. 2023). Thus, PDV1 and PDV2 may provide bidirectional coordination between stromal FtsZ-based assembly, inner-envelope division complexes, and the cytosolic fission machinery.

ARC5 (DRP5B) and the outer PD ring execute chloroplast fission and separation

At the outer envelope, the discontinuous DRP5B (DYNAMIN-RELATED PROTEIN 5B) ring and the outer plastid-dividing (PD) ring are involved in late constriction. DRP5B, mapped from the arc5 mutant (Gao et al. 2003), is a member of the dynamin family GTPases—membrane-remodeling proteins mediating endocytosis as well as organelle fission and fusion in eukaryotes (Ferguson and De Camilli 2012). DRP5B appears to have arisen early in the green lineage and is retained across plants (Liu et al. 2024).

The outer PD ring is an electron-dense structure observed at the plastid division site in algae (Kuroiwa 1989) and land plants (Hashimoto 1986; Oross and Possingham 1989). Current understanding of this structure comes mainly from studies of the red alga (Cyanidioschyzon merolae) chloroplasts, whose outer PD ring is composed of polyglucan filaments and requires the glycosyltransferase PLASTID-DIVIDING RING 1 (PDR1) for its formation (Yoshida et al. 2010).

The red algal ortholog of AtDRP5B, CmDnm2, acts together with the outer PD ring during the late stage of plastid division, where CmPDR1 is recruited to the chloroplast division site to fulfill its role in outer PD-ring formation while CmDnm2 is recruited at a later stage (Yoshida et al. 2010). In isolated red algal chloroplasts, dynamin is required for the division machinery to constrict after stretching (Yoshida et al. 2006). Immunoelectron microscopy further detected CmDnm2 between PD-ring filaments, supporting its role in PD-ring filament sliding (Yoshida et al. 2006). These observations, together with the membrane-fission activity of dynamin proteins, led to the hypothesis that CmDnm2 first promotes PD-ring filament sliding and later acts as a pinchase to mediate membrane scission at the bridge between daughter chloroplasts (Yoshida et al. 2006). Consistent with this model, the outer PD ring thickens as constriction proceeds (Miyagishima et al. 1999), and CmDnm2 moves from the outside to the inside of the constricted PD ring (Miyagishima et al. 1999; Yoshida et al. 2006).

Whether a similar mechanism operates for the outer PD-ring in land plants is unclear. AtDRP5B is also recruited from discrete cytosolic patches to the outer chloroplast surface at the division plane, where it assembles into a discontinuous ring structure, similar to CmDnm2 (Miyagishima et al. 2006). This punctate pattern resembles the localization of PDV1 and differs from the continuous localization pattern of Z ring and ARC6 (Miyagishima et al. 2006). DRP5B localization at the OEM surface is dynamically regulated by PDV1 and PDV2, where PDV2 mainly helps with DRP5B targeting to the chloroplast surface and assembly into a ring structure while PDV1 assists in DRP5B dissociation from the outer envelope (Sun et al. 2020a).

Recent work showed a key role for the endoplasmic reticulum (ER) in recruiting DRP5B to the division site and in plastid membrane constriction. Before localizing to the mid-plastid, DRP5B is found to be associated with the ER membrane. This association later forms a band around the plastid division site, and as it tightens, plastid-ER membrane contact sites coordinate isthmus narrowing with plastid movement and rotation, ultimately facilitating daughter plastid separation (Ghosh et al. 2025).

Other chloroplast division factors

Besides the outer PD ring, an electron-dense stromal ring has also been observed at the plastid division site in algae and plants (Hashimoto 1986; Kuroiwa 1989; Oross and Possingham 1989), whose structure is distinct from the Z ring, but the molecular composition remains unknown (Miyagishima et al. 2001). Thus, the chloroplast division apparatus contains additional, still-unidentified components beyond the currently characterized core machinery.

Besides the ARC loci that encode core plastid division components, ARC1 encodes FtsHi1, 1 of 5 Arabidopsis chloroplast-envelope FtsHi proteins related to FtsH metalloproteases. The seedling of the weak ftsHi1-1 (arc1) allele mutant is pale and contains increased numbers of small chloroplasts, as well as disrupted thylakoid development and reduced de-etiolation efficiency, whereas the plant carrying the strong ftsHi1-2 allele is embryo lethal (Kadirjan-Kalbach et al. 2012). ARC2 encodes the cyanobacterium-derived plastid chaperonin 60α1 (ptCpn60α1), and its mutation causes fewer and larger chloroplasts and reduced chlorophyll accumulation in the weak mutant allele and albino seedlings with severely impaired plastid greening in strong alleles (Suzuki et al. 2009). Related ptCpn60β proteins also contribute to plastid division, and reduced ptCpn60 activity perturbs normal Z-ring formation (Suzuki et al. 2009). However, how FtsHi proteins and plastid chaperonins connect plastid biogenesis, protein folding, and chloroplast division is still elusive.

Several additional non-core chloroplast division factors are involved in membrane organization and the separation of divided plastids, but the underlying biochemical mechanisms are not well characterized. Examples include the FUZZY ONIONS-like protein FZL (Gao et al. 2006), CLUMPED CHLOROPLASTS 1 (CLMP1) (Yang et al. 2011), and CRUMPLED LEAF (CRL) (Asano et al. 2004).

Thylakoid division

As the sites where the major photosynthetic complexes are embedded, thylakoid membranes are organized into grana stacks and stroma lamellae, which are interconnected and form a continuous luminal space (Gu et al. 2022). During chloroplast division, this internal membrane network must also be partitioned between the 2 daughter chloroplasts, yet how this process is coordinated with envelope constriction remains poorly understood. In arc6, pdv1, or pdv2, enlarged chloroplasts also contain enlarged thylakoids, indicating that thylakoid partitioning likely depends on both the inner- and outer-envelope division systems (Ho et al. 2023). Arabidopsis FtsZ1 and FtsZ2 proteins localize not only to the stroma but also to thylakoid membranes, and ftsZ1 and ftsZ2-2 chloroplasts have perturbed thylakoid membrane networks (El-Kafafi et al. 2008; Karamoko et al. 2011). Conversely, the CURVATURE THYLAKOID1 (CURT1)/CurT family, originally characterized as conserved thylakoid membrane proteins that shape thylakoid architecture by inducing membrane curvature (Armbruster et al. 2013; Heinz et al. 2016), is linked to cell or chloroplast division. The cyanobacterial CurT protein supports symmetric cell division and thylakoid fission, likely through its interaction with FtsZ, and the Chlamydomonas reinhardtii CURT1 proteins appear to support symmetric chloroplast division, likely through CURT1-FtsZ1 interaction (Dann et al. 2025). Therefore, chloroplast division factors may also contribute to thylakoid partitioning in dividing chloroplasts and the maintenance of thylakoid membrane organization.

Division of specialized plastids

Most plastid division research has focused on chloroplasts, and other plastid types are generally thought to share similar division machinery (Sun et al. 2020b). Increased expression of potato FtsZ1 results in fewer but larger starch granules in potato tubers, although amyloplast size was not examined (de Pater et al. 2006), and wheat parc6 endosperm cells contain enlarged amyloplasts with an increased number of starch granules (Esch et al. 2023). Another example is the Orange (Or) gene, which encodes a DnaJ cysteine-rich protein localized to mid-leucoplasts and regulates the activity of phytoene synthase in carotenoid biosynthesis (Lu et al. 2006; Zhou et al. 2015). The melon (Cucumis melo) Or^His allele confers orange flesh by promoting chromoplast differentiation, but cells of the Or^His fruits often contain only 1 or 2 enlarged chromoplasts caused by the interference of ARC3-PARC6 interaction by Or^His (Tzuri et al. 2015; Sun et al. 2020b). However, the requirement for individual division components can differ among plastid and cell types. For example, ARC6 and PARC6 have broad effects on plastid division (Pyke et al. 1994; Fujiwara et al. 2018; Ishikawa et al. 2020), whereas proplastid and non-mesophyll plastids in arc5 are largely normal or only mildly affected (Robertson et al. 1996; Fujiwara et al. 2018). Finally, budding-like proliferation has been observed in some plastid types, indicating an additional mode of plastid division besides binary fission (Kulandaivelu and Gnanam 1985; Forth and Pyke 2006; Yun and Kawagoe 2009; Du et al. 2013).

Chloroplast division is regulated at multiple levels

Plastid proliferation is tuned to growth and development through the regulation of division timing, protein abundance/quality, complex assembly, transcription, and hormone signaling.

In unicellular algae with 1 or a few chloroplasts, chloroplast division is tightly coordinated with the cell cycle (Nishida et al. 2005; Adams et al. 2008; Gillard et al. 2008). In C. merolae, this coupling involves reciprocal control between the nucleus and chloroplast, whereby nucleus-encoded division proteins are expressed in S phase and progression of chloroplast division releases prophase arrest to allow nuclear division to proceed (Sumiya et al. 2016). In land plants with multiple chloroplasts per cell, proplastid division keeps pace with mitotic cell division (Marrison et al. 1999), and the prereplication factor CDT1 provides a molecular link between cell-cycle regulation and plastid division (Raynaud et al. 2005). Arabidopsis CDT1a localizes to plastids and interacts with ARC6, and AtCDT1a/b knockdown severely impairs plastid division while slowing cell-cycle progression and increasing endoreduplication (Raynaud et al. 2005). After cells divide, plant chloroplast division continues as young mesophyll cells expand, with chloroplast number positively correlated with cell size (Pyke and Leech 1992; Marrison et al. 1999). During leaf expansion, PDV2 level shows the most pronounced decrease among the division components examined, in parallel to reduced chloroplast division rates, whereas FtsZ2-1, ARC6, and DRP5B levels are relatively stable (Okazaki et al. 2009). Unlike most plastid division mutants, which are recessive, pdv2 heterozygotes show a semi-dominant phenotype, and its rescue depends on the dosage of PDV2, further substantiating PDV2 level as a key control point for chloroplast division (Chang et al. 2017).

At the post-translational level, ARC3 is controlled by both cytosolic and chloroplast protein quality control pathways (Yuan et al. 2026). In the cytosol, the E3 ubiquitin ligase PUB52 promotes degradation of ARC3 precursors through direct ubiquitination. Within chloroplasts, the chaperone-protease CLPC1 promotes ARC3 degradation in the stroma, whereas ARC2 partially suppresses CLPC1-mediated ARC3 degradation and protects ARC3 from excessive proteolysis. The relative abundance of reduced and oxidized ARC3 forms changes during leaf development, suggesting another potential regulatory layer of ARC3 activity (Yuan et al. 2026).

The formation of division complexes is also regulated by light and redox state. In the dark or under oxidizing conditions, an intramolecular disulfide bond in PARC6 prevents PDV1 binding. Light reduces this bond, allowing PDV1-induced PARC6 dimerization, and light-induced elevation of chloroplast Mg²⁺ levels further promotes PARC6 dimerization and chloroplast division (Sun et al. 2023).

At the transcription level, FAR-RED ELONGATED HYPOCOTYL3/CHLOROPLAST DIVISION 45 (FHY3/CPD45), a key regulator of far-red-light signaling, promotes chloroplast division by activating ARC5/DRP5B expression independently of its role in phytochrome signaling (Chang et al. 2015). This activation requires cooperation with FAR1-RELATED SEQUENCE 4/CHLOROPLAST DIVISION 25 (FRS4/CPD25), a FHY3 homolog that binds the DRP5B promoter more efficiently but lacks transcriptional activation activity on its own (Gao et al. 2013). Several broader developmental regulators also affect chloroplast division, such as the GATA transcription factors GATA NITRATE-INDUCIBLE CARBON-METABOLISM-INVOLVED (GNC) and CYTOKININ-RESPONSIVE GATA1 (CGA1) that positively affect chloroplast development and division (Chiang et al. 2012), and GROWTH-REGULATING FACTOR5 (GRF5) that promotes chloroplast division as part of the leaf-development and photosynthetic programs (Vercruyssen et al. 2015).

Phytohormone signaling impacts plastid division as well. Exogenous cytokinin treatment or overexpression of the transcription factor CYTOKININ RESPONSE FACTOR2 (CRF2) increases the chloroplast division rate and selectively upregulates the level of PDV proteins (Okazaki et al. 2009), further supporting PDV abundance as a key regulatory point in chloroplast division. Cytokinins also positively regulate the transcription of GNC and CGA1 (Chiang et al. 2012). Additionally, exogenous auxin increases FtsZ and PDV2 transcript and protein levels and induces disorganized division rings, some of which form in small chloroplasts, resulting in pronounced chloroplast size heterogeneity within individual cells (Wang et al. 2024). Further, gibberellins promote chloroplast division, probably indirectly, through mesophyll cell expansion (Jiang et al. 2012). By contrast, abscisic acid (ABA) negatively regulates plastid division, as the tomato ABA-deficient mutant high-pigment 3 (hp3) shows increased FtsZ transcript levels and more plastids per fruit cell than the wild type, although the sizes of cell and individual plastids are unchanged. Increased plastid number and elevated pigmentation in the fruit are also observed in several other ABA-deficient tomato mutants, further confirming the negative role of ABA in plastid division in developing fruit (Galpaz et al. 2008).

Taken together, decades of genetics, cell biological, and molecular studies have established an elegant model of chloroplast division. Briefly, at the stromal side of the inner envelope, FtsZ proteins assemble into the Z ring, whose position and dynamics define the future division site, while ARC6, PARC6, PDV1, and PDV2 proteins connect the internal division machinery with the outer envelope and cytosolic factors. At the cytosolic surface of the plastid, ARC5 (DRP5B) assembles at the division site to execute the final stage of constriction and fission (Table 1, Fig. 2). Division components are subjected to multiple levels of regulation, from gene transcription to complex stability.

Table 1.

Major division factors in plant chloroplasts, mitochondria, and peroxisomes.

Protein	AGI locus ID	Localization	Origin	Main functions and phenotypes	References
FtsZ1	AT5G55280	Stromal side of chloroplast IEM	C	Promotes Z-ring turnover. FtsZ1 RNAi plants show jagged FtsZ2 filament structures. Either depletion or elevated expression of FtsZ1 results in 1 or a few enlarged chloroplasts.	Osteryoung and Vierling (1995); Osteryoung et al. (1998); Stokes et al. (2000); Vitha et al. (2001)
FtsZ2-1	AT2G36250	Stromal side of chloroplast IEM	C	Supports Z-ring stability. FtsZ2-1 RNAi plants show short FtsZ1 filaments and bright foci. Either depletion or elevated expression of FtsZ2 results in 1 or a few enlarged chloroplasts.	Osteryoung et al. (1998); Stokes et al. (2000); McAndrew et al. (2001); Vitha et al. (2001)
FtsZ2-2	AT3G52750	Stromal side of chloroplast IEM	C	Functionally largely overlaps with FtsZ2-1. Reciprocal complementation suggests partial substitution between the 2 isoforms. Reduced total FtsZ2 abundance results in a few enlarged chloroplasts.	Schmitz et al. (2009)
FtsZ3	N.A.	Stromal side of chloroplast IEM	C*	Restrains excessive FtsZ2 assembly and promotes FtsZ-ring constriction through interacting with ARC6. ftsZ3 loss-of-function or GTPase-defective mutants contain enlarged or elongated chloroplasts with long FtsZ filaments forming multiple rings or spirals; C-terminal-truncated ftsZ3 mutants contain mildly enlarged, dumbbell-shaped chloroplasts.	Martin et al. (2009); Grosche and Rensing (2017); Lang et al. (2026)
ACCUMULATION AND REPLICATION OF CHLOROPLASTS 3 (ARC3)	AT1G75010	Stromal side of chloroplast IEM	H	Restricts ectopic Z-ring formation. arc3 shows multiple Z rings within a heterogeneous chloroplast population characterized by reduced chloroplast numbers and variable sizes and shapes.	Pyke and Leech (1992); Shimada et al. (2004); Maple et al. (2007); Zhang et al. (2013)
MinD1/ACCUMULATION AND REPLICATION OF CHLOROPLASTS 11 (ARC11)	AT5G24020	Stromal side of chloroplast IEM	C	Contributes to ARC3-dependent restriction of ectopic Z-ring formation. minD1 shows phenotypes similar to those of arc3.	Marrison et al. (1999); Colletti et al. (2000); Fujiwara et al. (2004); Zhang et al. (2013)
MinE1/ACCUMULATION AND REPLICATION OF CHLOROPLASTS 12 (ARC12)	AT1G69390	Stromal side of chloroplast IEM	C	Counteracts MinD1-dependent inhibition to regulate Z-ring placement. minE1 contains a few enlarged chloroplasts with short FtsZ filaments and dots. MinE1 overexpression phenocopies arc3 and minD1.	Itoh et al. (2001); Glynn et al. (2007); Fujiwara et al. (2008); Zhang et al. (2013)
MULTIPLE CHLOROPLAST DIVISION SITE 1 (MCD1)	AT1G20830	Chloroplast IEM	H	ARC3- and MinE1-dependent regulation of Z-ring positioning. mcd1 shows phenotypes similar to those of arc3.	Nakanishi et al. (2009); Chen et al. (2018b)
ACCUMULATION AND REPLICATION OF CHLOROPLASTS 6 (ARC6)	AT5G42480	Chloroplast IEM	C	Tethers and stabilizes FtsZ2 at the IEM, and coordinates IEM and OEM division machinery through PDV2. arc6 contains one to a few enlarged chloroplasts with numerous short, disorganized FtsZ filaments.	Pyke et al. (1994); Vitha et al. (2003); Glynn et al. (2008); Sung et al. (2018)
PARALOG OF ARC6 (PARC6)/CHLOROPLAST DIVISION SITE POSITIONING 1 (CDP1)/ARC6 HOMOLOG (ARC6H)	AT3G19180	Chloroplast IEM	H	Promotes FtsZ2 filament turnover and coordinates IEM and OEM division machinery through PDV1. parc6 contains chloroplasts of heterogeneous size, including enlarged dumbbell-shaped chloroplasts with multiple Z rings or spirals at the constriction site.	Glynn et al. (2009); Zhang et al. (2009); Ottesen et al. (2010); Zhang et al. (2016); Sun et al. (2023)
PLASTID DIVISION 1 (PDV1)	AT5G53280	Chloroplast OEM	H	Links PARC6 to the OEM division machinery and promotes DRP5B-OEM dissociation. pdv1 contains chloroplasts of heterogeneous size, including enlarged dumbbell-shaped chloroplasts with multiple Z rings or spirals at the constriction site.	Miyagishima et al. (2006); Sun et al. (2020a, 2023)
PLASTID DIVISION 2 (PDV2)	AT2G16070	Chloroplast OEM	H	Links ARC6 to the OEM division machinery and helps recruit DRP5B/ARC5 to the OEM for ring assembly. pdv2 contains chloroplasts of heterogeneous size, including enlarged dumbbell-shaped chloroplasts with multiple Z rings or spirals at the constriction site.	Miyagishima et al. (2006); Glynn et al. (2008); Wang et al. (2017); Sun et al. (2020a)
DYNAMIN-RELATED PROTEIN 5B (DRP5B)/ACCUMULATION AND REPLICATION OF CHLOROPLASTS 5 (ARC5)	AT3G19720	Cytosolic surface of the OEM	H	Mediates chloroplast constriction and peroxisome division. drp5B shows reduced chloroplast number, enlarged dumbbell-shaped chloroplasts with multiple Z rings or spirals at constriction sites, and clustered peroxisomes with impaired function. drp5B exhibits mitochondrial-division defects despite DRP5B's lack of association with mitochondria.	Pyke and Leech (1994); Robertson et al. (1996); Gao et al. (2003); Zhang and Hu (2010); Aung and Hu (2012)
DYNAMIN-RELATED PROTEIN 3A (DRP3A)	AT4G33650	Mitochondrial OEM and peroxisome membrane	H	DRP3A and DRP3B function redundantly in mitochondrial fission. DRP3A plays a predominant role in peroxisomal fission. drp3A and drp3B contain fewer, mildly elongated mitochondria, whereas drp3A drp3B contains massively interconnected mitochondria. drp3A contains larger, fewer peroxisomes, while the peroxisome phenotype in drp3B is much weaker.	Fujimoto et al. (2009); Zhang and Hu (2009); Aung and Hu (2012)
DYNAMIN-RELATED PROTEIN 3B (DRP3B)	AT2G14120	Mitochondrial OEM and peroxisome membrane	H
FISSION 1A (FIS1A)	AT3G57090	chloroplast OEM, mitochondrial OEM, and peroxisome membrane	H	Required for mitochondrial and peroxisomal division. fis1A and FIS1B RNAi plants contain enlarged mitochondria and peroxisomes with reduced numbers.	Scott et al. (2006); Zhang and Hu (2009); Ruberti et al. (2014)
FISSION 1B (FIS1B)	AT5G12390	Mitochondrial OEM and peroxisome membrane	H
ELONGATED MITOCHONDRIA 1 (ELM1)	AT5G22350	Mitochondrial OEM	H	Recruits DRP3 to mitochondrial division sites. elm1 contains elongated mitochondria.	Arimura et al. (2008)
PEROXIN 11A (PEX11A)	AT1G47750	Peroxisomal membrane	H	PEX11 isoforms have partially overlapping roles in peroxisome division. pex11abcd shows severe physiological defects, and pex11abcde is nonviable under normal germination conditions.	Lingard and Trelease (2006); Nito et al. (2007); Orth et al. (2007); Zhu et al. (2026); Tharp et al. (2026)
PEROXIN 11B (PEX11B)	AT3G47430	Peroxisomal membrane	H
PEROXIN 11C (PEX11C)	AT1G01820	Peroxisomal membrane	H
PEROXIN 11D (PEX11D)	AT2G45740	Peroxisomal membrane	H
PEROXIN 11E (PEX11E)	AT3G61070	Peroxisomal membrane	H
PEROXISOMAL AND MITOCHONDRIAL DIVISION FACTOR 1 (PMD1)	AT3G58840	Mitochondrial OEM, peroxisome membrane	H	PMD1 is required for mitochondrial and peroxisomal division, whereas PMD2 is specific to mitochondria. pmd1 contains enlarged peroxisomes and elongated mitochondria. PMD2 RNAi lines contain elongated mitochondria with wild-type-like peroxisomes.	Aung and Hu (2011)
PEROXISOMAL AND MITOCHONDRIAL DIVISION FACTOR 2 (PMD2)	AT1G06530	Mitochondrial OEM	H

Host origin factors represent broadly eukaryotic, green-lineage, and plant-specific components. Asterisk indicates factors absent from Arabidopsis and present only in specific green-lineage groups.

IEM, inner envelope membrane; OEM, outer envelope membrane; C, cyanobacterial origin; H, host origin.

Figure 2.

Chloroplast division model. The chloroplast division machinery is organized from the stroma through the inner envelope membrane (IEM), intermembrane space (IS), and outer envelope membrane (OEM) to the cytosolic side, and includes proteins from cyanobacterial and green-lineage origins. FtsZ1 and FtsZ2 assemble into the stromal Z ring, whereas Min system proteins, including ARC3, MinD1, MinE1, and MCD1, regulate Z-ring placement and remodeling. Inner-envelope proteins ARC6 and PARC6 link the stromal division machinery across the IS to outer-envelope PDV2 and PDV1, respectively, promoting recruitment of the cytosolic dynamin-related protein DRP5B/ARC5 to the division site. The outer PD ring is on the cytosolic side. An inner PD ring (not shown) has also been observed in several green-lineage species, but its molecular composition remains unknown.

Conserved and common factors in plant organelle division

Fluorescence microscopy-facilitated forward genetic screens and reverse genetic strategies have also been employed to elucidate the division mechanisms of other plant organelles, most notably mitochondria and peroxisomes. In addition to performing unique functions of their own, chloroplasts, mitochondria and peroxisomes often coordinate in key metabolic pathways such as photorespiration, which occurs sequentially in these organelles to recycle the toxic by-products of Rubisco's oxygenase activity and support central metabolism (Eisenhut et al. 2019).

Mitochondria arose long before plastids, after the engulfment of an α-proteobacterium by an archaea-related host cell with primitive eukaryotic features (Zaremba-Niedzwiedzka et al. 2017). However, the currently identified plant mitochondrial division machinery contains fewer known components than chloroplasts and is dominated by eukaryotic host-derived factors. Consistent with the emergence of mitochondria during early eukaryotic evolution, mitochondrial division shares key common factors between plants and other eukaryotic species, including dynamin-related proteins (DRPs) and the tail-anchored FISSION1 (FIS1) proteins (Arimura 2018). Mammalian Drp1 is recruited to mitochondria by outer-membrane proteins including Mff, MiD49/MIEF2, MiD51/MIEF1, and likely Fis1. In budding yeast, the dynamin-related protein DNM1 is recruited to mitochondria by FIS1 with the assistance of the adaptor protein Mdv1 or Caf4 (reviewed in Kamerkar et al. (2025)). Except for FIS1, these mammalian or yeast OEM proteins do not seem to have apparent homologs in plants. Instead, the plant-specific mitochondrial outer-membrane protein ELM1, identified from an Arabidopsis mutant with elongated mitochondria (elm1), recruits DRP3 to the division sites (Arimura et al. 2008). Similar to their homologs in other eukaryotes, Arabidopsis DRP3 (DRP3A and DRP3B) proteins localize to the tips and constriction sites of mitochondria, and mutations in DRP3 result in elongated or clustered mitochondria due to impaired fission (Arimura et al. 2008; Aung and Hu 2009; Fujimoto et al. 2009; Zhang and Hu 2009; Aung and Hu 2012) (Fig. 3a). Arabidopsis fis1 mutants contain enlarged and reduced number of mitochondria (Scott et al. 2006; Zhang and Hu 2009), although whether FIS1 (FIS1A and FIS1B) serves as an anchor for DRP3 proteins remains to be clarified. Arabidopsis DRP1C and DRP1E and rice DRP1E are involved in mitochondrial morphogenesis or fission (Jin et al. 2003; Xu et al. 2020), yet their mode of action, especially the relation to FIS1, has not been elucidated. Besides the dynamin-related protein CmDNM1, division of the single mitochondrion in C. merolae requires CmFtsZ1 and the ZapA-like protein (ZED), which assemble into matrix rings beneath the inner membrane (Takahara et al. 2000; Yoshida et al. 2009), and Mda1, which is structurally related to yeast Mdv1 and Caf4 (Nishida et al. 2007). In plants, these proteins either seem to have been lost or do not play a role in mitochondrial division, as in the case of CmFtsZ1 in C. merolae (Nishida et al. 2007).

Figure 3.

Mitochondrial and peroxisome morphology phenotypes of Arabidopsis drp3 mutants and PMD1 overexpressors. (a) Confocal images of mitochondria (ScCOX4-YFP) in leaf epidermal cells, showing mostly elongated mitochondria in the drp3 mutants. Scale bars = 10 μm. Adapted from Aung and Hu (2012). (b) Confocal images of leaf mesophyll cells in which peroxisomes are labeled by YFP-PTS1 (Peroxisome Targeting Signal type 1, SKL), showing stronger defects in peroxisome division in drp3A than in drp3B. Chloroplasts are in red. Scale bars = 10 μm. Adapted from Zhang and Hu (2009). (c) Confocal images of leaf epidermal cells showing the over-proliferation and aggregation of mitochondria (labeled by ScCOX4-YFP) and peroxisomes (labeled by CFP-PTS1) in the PMD1-overexpressing plant. Scale bar = 10 μm and applies to all 4 panels. Reproduced from Aung and Hu (2011). (d) TEM images of Arabidopsis leaf mesophyll cells showing the ultrastructure of peroxisomes (Perox), mitochondria (Mito), and chloroplasts (Chpt) in Col-0 and 35S_pro:PMD1 plants, highlighting the aggregation of both organelles. Bars = 1 µm. Adapted from Aung and Hu (2011).

The single-membraned peroxisomes can arise through de novo biogenesis from specialized ER subdomains (Hoepfner et al. 2005), in plants called ER nests (Wright et al. 2025). Peroxisomes also multiply very frequently by growth/elongation followed by fission, using the same mitochondrial division proteins—DRPs and their putative anchor, FIS1, across eukaryotic species, as well as Mff in mammals (Tábara et al. 2025). In Arabidopsis, although both DRP3A and DRP3B contribute significantly to mitochondrial fission, DRP3A appears to play a much stronger role than DRP3B in peroxisome division (Fujimoto et al. 2009; Zhang and Hu 2009) (Fig. 3b). The peroxisome morphology phenotype of the Atfis1 mutant is also relatively weak (Zhang and Hu 2009).

Prior to fission, the conserved peroxisome integral membrane protein PEX11 uses its N-terminal amphipathic helix to mediate membrane curvature and elongation (Opaliński et al. 2011). Human PEX11β was shown to induce peroxisome division in a FIS1- and DRP1-dependent manner independent of Mff, directly linking PEX11 function to Fis1 and Drp1 (Schrader et al. 2022), although such a connection has not been made in plants. Earlier studies of the 5-member Arabidopsis PEX11 gene family suggest partial functional redundancy among the isoforms (Lingard and Trelease 2006; Nito et al. 2007; Orth et al. 2007), which was recently elaborated by analyses of higher-order pex11 mutants that found strong physiological defects in pex11cde triple and pex11abcd quadruple mutants and seedling lethality in pex11abcde quintuple mutant (Tharp et al. 2026; Zhu et al. 2026). A recent study discovered the role of Arabidopsis PEX11 proteins beyond peroxisome division, by serving as mediators of peroxisomal intralumenal vesicle formation (Tharp et al. 2026).

The plant-specific Peroxisomal and Mitochondrial Division proteins PMD1 (peroxisomal and mitochondrial) and PMD2 (mitochondrial) are also involved in the proliferation of peroxisomes and mitochondria. Loss of function of the dual-localized PMD1 protein results in enlarged and elongated mitochondria and peroxisomes, while PMD1 overexpression induces the proliferation and aggregation of these 2 organelles (Aung and Hu 2011) (Fig. 3c, 3d). PMD1 function seems to be independent of the DRP3–FIS1 machinery (Aung and Hu 2011), and PMD1 can interact with actin (Frick and Strader 2018), indicating its action in a separate process involved in organelle proliferation.

Interestingly, Arabidopsis DRP5B/ARC5, the major plastid division DRP, also associates with peroxisomes and impairs peroxisome division when mutated (Zhang and Hu 2010) (Fig. 4a, 4b). Furthermore, FIS1A localizes to all 3 organelles (Ruberti et al. 2014), raising the possibility that this membrane protein may coordinate the dynamics of these 3 metabolically linked organelles. It is likely that during eukaryotic evolution, FIS1 and several DRP proteins were picked up by multiple organelles for their dynamics, which may lead to some level of collaboration among the machinery that governs the dynamics of these functionally associated organelles (Fig. 4c).

Figure 4.

Dual role of DRP5B in Arabidopsis chloroplast and peroxisome division and shared division factors among mitochondria, peroxisomes and chloroplasts. (a) Confocal images taken of mesophyll cells of plants expressing the 35S_pro:DsRed2-PTS1 and DRP5B_pro:GFP-DRP5B transgenes, showing the dual localization of DRP5B on chloroplasts and peroxisomes. GFP-DRP5B is shown in green, DsRed2-PTS1-labeled peroxisomes in magenta, and chlorophyll autofluorescence in red. Scale bar = 10 µm and applies to all 4 panels. Adapted from Zhang and Hu (2010). (b) Confocal images of mesophyll cells showing impaired division in both peroxisomes (green, labeled by YFP-PTS1) and chloroplasts (red). Scale bars = 10 µm. Adapted from Zhang and Hu (2010). (c) Schematics of the shared division components among chloroplasts, peroxisomes, and mitochondria that indicate potential coordination in organelle division. Note that whether FIS1A/FIS1B serves as an anchor for DRP3A/DRP3B has not been directly proven in plants, and ELM1 is mitochondrial-specific.

Perspectives

Building on the arc mutant studies first reported in Plant Physiology in the 1990s (Pyke and Leech 1992, 1994; Pyke et al. 1994; Robertson et al. 1996), subsequent cell biology and molecular genetic studies have established an elegant model depicting the highly complex and coordinated chloroplast division machinery, a mosaic of components descended from the cyanobacterial endosymbiont and new inventions during the evolution of green eukaryotes (Fig. 2).

Some of the key division factors are shared among chloroplasts, mitochondria and peroxisomes (Fig. 4c), suggesting the division machinery of the 3 organelles may operate within an integrated regulatory network that coordinates organelle abundance and dynamics to support plant growth and adaptation to environmental variation (Pan et al. 2020). In addition, the FHY3-FRS4-mediated activation of ARC5/DRP5B expression (Gao et al. 2013; Chang et al. 2015) seems to be analogous to some extent to the light induction of peroxisome proliferation, a phytochrome A-mediated pathway that acts through the transcription factor HYH to activate expression of the peroxisome elongation factor gene PEX11b (Desai and Hu 2008; Hu and Desai 2008), while another transcription factor, Forkhead-Associated Domain Protein 3 (FHA3), represses PEX11b expression (Desai et al. 2017). It will be interesting to investigate if there is crosstalk between the 2 far-red light signaling pathways that may coordinate the proliferation of the 2 organelles.

Alterations of general plastid division often lead to suboptimal outcomes. Plastid division mutants or overexpression lines with smaller and more numerous chloroplasts have a greater total chloroplast surface area, but reduced plant biomass (Dutta et al. 2017; Głowacka et al. 2023). Plants with fewer but enlarged chloroplasts show reduced mesophyll conductance, enhanced non-photochemical quenching, altered composition, structure, and capacity of the photosynthetic apparatus, and impaired chloroplast movement responses (Jeong et al. 2002; Austin Li and Webber 2005; Weise et al. 2015; Dutta et al. 2017; Głowacka et al. 2023). However, changes in chloroplast size can influence other plastid-derived processes, such as elevated JA and JA-Ile levels upon wounding in Arabidopsis arc mutants accumulating enlarged chloroplasts, suggesting an impact of chloroplast size on jasmonate biosynthesis (Baral et al. 2026).

Therefore, the value of manipulating plastid division may depend on the target trait, especially in specialized plastids where plastid-derived products directly determine crop quality. FtsZ1 overexpression in potato tubers and loss of PARC6 function in wheat grains both produce larger starch granules (de Pater et al. 2006; Esch et al. 2023), which could be advantageous to traits such as milling efficiency, functional starch properties, and nutritional quality (Li et al. 2023). In Arabidopsis OR^His lines that recapitulate the melon OR^His chromoplast-number-constraint phenotype, overexpression of PDV1 increased chromoplast number and further enhanced carotenoid accumulation (Sun et al. 2020b), pointing to the potential of plastid division pathways as targets for improving crop traits.

Manipulating organelle abundance may also be promising in enhancing plant stress tolerance and disease resistance by strengthening organelle-mediated stress responses. For instance, Arabidopsis PEX11a was shown to be involved in cadmium stress perception and fast cell response to oxidative stress, indicating that increased peroxisome abundance may contribute to stress adaptation through restoration of redox homeostasis (Rodríguez-Serrano et al. 2016). Overexpression of a rice PEX11 gene increased salt tolerance (Cui et al. 2016), raising the potential of elevating peroxisome abundance for improved crop stress response. Interestingly, a rice dynamin-related protein, OsDRP1E, plays a negative role in rice resistance to the rice blast fungus Magnaporthe oryzae, as its knockout shows mitochondrial elongation and enhanced resistance, whereas its overexpression causes mitochondrial fragmentation and increased pathogen susceptibility (Xu et al. 2020), highlighting an exciting opportunity to improve plant disease resistance through modulating organelle dynamics.

Advances box.

1. Early identification and characterization of the arc mutants established the genetic framework for the molecular dissection of the chloroplast division machinery.

2. Chloroplast division components consist of both cyanobacterium- and green lineage-derived proteins, acting through interdependent interaction modules, with bridging and reciprocal interactions that organize the chloroplast division machinery across stromal and envelope layers.

3. Chloroplast division is regulated through the control of Z-ring dynamics, envelope coordination, protein abundance, and transcription, as well as by hormones and developmental cues.

4. Mitochondria, peroxisomes, and chloroplasts share some common division components, indicating possible coordination in the division of these metabolically linked organelles.

5. Manipulation of organelle abundance may provide new opportunities for crop improvement.

Funding

U.S. National Science Foundation (MCB2148206) to JH; the Office of Science, Office of Basic Energy Sciences, U.S. Department of Energy to JH, under Award Number DE-FG02-91ER20021.

Data availability

No new data were generated or analysed in support of this research.