Authors: David S. Guttery, Mohammad Zeeshan, Anthony A. Holder, Rita Tewari
Categories: Review Articles, DNA replication, cell division, closed mitosis, crystalloids, reversible protein phosphorylation, Cell Cycle, Growth & Proliferation, Microbiology, Parasitology
Source: Biochemical Society Transactions
Doi: 10.1042/BST20230403
Malaria, a vector borne disease, is a major global health and socioeconomic problem caused by the apicomplexan protozoan parasite Plasmodium. The parasite alternates between mosquito vector and vertebrate host, with meiosis in the mosquito and proliferative mitotic cell division in both hosts. In the canonical eukaryotic model, cell division is either by open or closed mitosis and karyokinesis is followed by cytokinesis; whereas in Plasmodium closed mitosis is not directly accompanied by concomitant cell division. Key molecular players and regulatory mechanisms of this process have been identified, but the pivotal role of certain protein complexes and the post-translational modifications that modulate their actions are still to be deciphered. Here, we discuss recent evidence for the function of known proteins in Plasmodium cell division and processes that are potential novel targets for therapeutic intervention. We also identify key questions to open new and exciting research to understand divergent Plasmodium cell division.
Keywords: cell division, closed mitosis, crystalloids, DNA replication, Plasmodium, reversible protein phosphorylation
Cell division occurs through a tightly co-ordinated set of processes. Studies in several model eukaryotes including humans and yeast have highlighted highly conserved mechanisms of cell division [1]; however, variation in this process occurs, for example in the divergent branch of eukaryotes that includes the unicellular apicomplexan parasites. This group includes pathogens of socioeconomic importance such as Toxoplasma and Plasmodium, the causative agents of toxoplasmosis and malaria, respectively. These organisms have some conserved aspects of cellular development [2], but metabolic regulatory pathways, membrane biogenesis and cytoskeletal dynamics are highly divergent and subject to continuous intense research [3–7]. Plasmodium is one of the most devastating global causes of vector-borne disease, killing over 600 000 people per year [8]. In this review, we discuss recent developments in our understanding of Plasmodium cell division, focusing on closed mitosis during proliferation, development and transmission throughout the life-cycle in both mammalian host and mosquito vector.
The Plasmodium life cycle is characterised by a series of morphologically distinct stages. Infection in mammals is initiated by injection of motile haploid sporozoites into the dermis during a mosquito blood-meal. Sporozoites eventually invade host hepatocytes and begin a proliferative stage in which there is DNA replication and karyokinesis (nuclear division) but no cytokinesis (cell division) until the end of the intracellular stage — exoerythrocytic schizogony. This produces thousands of haploid merozoites, which egress from the host cell and invade erythrocytes to begin cyclic asexual proliferation by intra-erythrocytic schizogony (Figure 1A(i)) and invasion of further erythrocytes. Schizogony produces a multinucleate cell (a coenocyte) and subsequent cytokinesis forms new merozoites. During this replicative phase, some parasites differentiate into male and female gametocytes that are arrested at the G0 phase of the cell cycle and are responsible for initiation of the sexual stage and transmission to the mosquito. Ingestion in the blood meal by a mosquito activates the gametocytes, with male(micro) gametocytes undergoing three rounds of rapid DNA replication and mitosis without karyokinesis (termed endomitosis) within 15 min, followed by karyokinesis and cytokinesis leading to the budding of eight haploid gametes. These motile gametes find and fertilise the mature female(macro) gamete (Figure 1A(ii)) to produce a diploid zygote [7], which over a period of 24 h elongates and differentiates into a crescent (or ‘banana’)-shaped motile ookinete, accompanied by DNA replication (to 4N) and the first stages of meiosis. The ookinete traverses the mosquito midgut wall and develops into an oocyst over a period of up to 21 days. Within the oocyst, sporogony proceeds with multiple rounds of DNA replication and karyokinesis, similar to schizogony, followed by cytokinesis to produce hundreds of motile haploid sporozoites (Figure 1A(iii)) that migrate to the mosquito salivary glands for transmission during the next bloodmeal.
Figure 1. Plasmodium cell division and the malaria parasite life-cycle.(A) Life cycle of Plasmodium spp. highlighting schizogony in red blood cells (i), and male gametogenesis (ii) and sporogony (iii) in the mosquito midgut. (B) Modes of asexual cell division in Plasmodium (endomitosis) compared with model organisms (binary fission). In mammalian and yeast systems, open mitosis results in dissolution of the nuclear membrane at the G2/M transition and reassembly post-DNA segregation, with concomitant cytokinesis resulting in two individual daughter cells while in budding yeast the nuclear membrane remains intact and the chromosomes are separated on a spindle assembled within the nucleus. In Plasmodium, DNA replication is closed with replication of the genome and DNA segregation occurring within a single nucleus, with karyokinesis and cytokinesis at the end of this process resulting in eight haploid flagellate male gametes.
In the classical model system of mammalian cell division, mitosis is open with dissolution of the nuclear membrane. The replicated chromosomes condense, and the duplicated sister chromatids are separated on a microtubule-based bipolar spindle. In closed mitosis as found in budding yeast the nuclear membrane remains intact and the chromosomes are separated on a spindle assembled within the nucleus [9] (Figure 1B). In both cases karyokinesis and cytokinesis produce two daughter cells.
Mitotic cell division during Plasmodium schizogony is characterised by several rounds of DNA replication and asynchronous nuclear division (a process defined by DNA replication in the absence of cell division [10]), to produce a multinucleated single cell or coenocyte, and termed a schizont (reviewed extensively in [11,12]) (Figure 1B), followed by one round of cytokinesis at the end of schizogony to form merozoites. In both exo- and intra-erythrocytic schizogony mitosis is accompanied by rapid development and replication of organelles including the mitochondrion, the apicoplast (a plastid found in the Apicomplexa [13]), the inner-membrane complex (IMC), the basal membrane and organelles involved in host cell invasion such as the micronemes and rhoptries. The main difference between liver- and blood stage schizogony is that tens of thousands of merozoites are produced in hepatocytes [14]; whereas blood stage schizogony produces 8–32 merozoites each cycle [11] (Figure 1B).
In schizogony, nuclear and cell division differ from the canonical process. The model of eukaryotic cell cycle progression is of tightly regulated a growth or gap phase (G1), DNA synthesis (S-phase), another gap phase preparing for mitosis (G2), and nuclear division and segmentation (mitosis — M phase). Mitosis is followed by cytokinesis when organelles and the cytoplasm are physically distributed between the two new daughter cells. In Plasmodium, following host cell invasion, the initial haploid ring and trophozoite stages resemble G1 phase cells, then the trophozoite enters schizogony characterised by multiple rounds of asynchronous DNA synthesis and nuclear division (S-phase) to produce multiple nuclei without immediate cytokinesis after each round [11,12]. Hence, schizogony appears to lack a G2 phase due its continual and cyclical use of S-phase [15]. The divergent nature of Plasmodium cell cycle progression is exemplified by a lack of canonical cell cycle checkpoint proteins including p53, ataxia telangiectasia and Rad3-related (ATR), ataxia telangiectasia mutated (ATM) and retinoblastoma protein (Rb) among others [15,16]. In addition, Plasmodium (and other Apicomplexa) has an atypical repertoire of cyclins, cyclin-dependent kinases (CDKs) and CDK-related kinases (CRKs). Orthologues of three cyclins (Cyc1, Cyc3 and Cyc4 [17]), and seven CDK or CDK-related kinases, protein kinase (PK) 5, PK6, the serine/threonine kinase MRK1, CRK1, CRK3, CDK4 and CRK5 have been identified [16]. However, while CRK5 is likely not essential for Plasmodium blood stage development, it has been shown to be key for male gametogenesis and interacts with the *Plasmodium-*specific cyclin SOC2 [18].
Studies using ultrastructural expansion microscopy (U-ExM) and Stimulated Emission Depletion Microscopy (STED) have generated an atlas of the three-dimensional structures and organelles that drive Plasmodium schizogony and their temporal profiles [19–21] (for extensive review on use of U-ExM to analyse Plasmodium, see [22]). The first cellular feature of schizogony is production of the hemispindle, defined as an array of microtubules that radiate into the nucleus from a single microtubule organising centre (MTOC) embedded in the nuclear membrane, and also referred to as the centriolar plaque [15,23]. Once DNA replication is complete, driven by phosphorylation of origin of replication (ORC) and minichromosome maintenance (MCM) complexes by cyclin-dependent kinases such as cdc2-related kinase 4 (CRK4) [24–26], the MTOC organises the microtubules into a mitotic spindle that segregates the replicated chromosomes. The centriolar plaque in Plasmodium asexual stages is a protein-dense structure devoid of chromatin and centrioles, with an intra- and extranuclear compartment [12,23]. Here, centrin-4 is located at the MTOC (Figure 2A) and forms a complex with centrin-1 and -3, but it is not essential for asexual stage development and proliferation, suggesting redundancy with, and possible replacement by, other centrins [27]. In addition, phosphorylation of centrins and an Sfi1-like centrin-interacting centriolar plaque protein (PfSlp — which has a key role in Plasmodium nuclear microtubule homeostasis [28]) may play an important role in kinetochore assembly and centrin localisation during schizogony. For example, in human cells, phosphorylation of centrins 1 and 2 by casein kinase 2 (CK2) regulates their binding to Sfi1 [29], and phosphorylation of Sfi1 by CDK1 is essential for yeast spindle pole body separation in G1/S phase [30]. The Plasmodium homologue of CDK1 is PK5 but is not essential at any stage of the life-cycle [31]; whereas the catalytic subunit of PfCK2 (PfCK2α) may play a role during late gametocytogenesis [32]. Finally, protein dephosphorylation is also important, with depletion of the protein phosphatase (PP) 1 that is located at kinetochores resulting in decreased DNA replication in Plasmodium falciparum asexual blood stages [33].
Figure 2. Live-cell imaging of cell division markers during two stages of proliferation in Plasmodium.(A) Live cell imaging using NDC80-mCherry (kinetochore marker) and Centrin-4-GFP (MTOC marker) showing spatio-temporal dynamics of the kinetochore and MTOC during blood stage schizogony. (B) Live cell imaging using NDC80-mCherry (kinetochore marker) and SAS4-GFP (MTOC/basal body marker) showing spatio-temporal dynamics of the kinetochore and basal body during male gametogenesis. Scale bar = 5 µm.
Throughout schizogony, several organelles develop in preparation for daughter cell (merozoite) formation. For example, biogenesis of apical organelles including the rhoptries and micronemes occurs de novo. The IMC, a double layered membranous structure located beneath the plasma membrane (PM) [16] is pulled around the newly-forming daughter cells by the basal complex (BC), whose integrity is maintained by the PPP-type pseudophosphatase PPP8 [34] and acts as a contractile ring to constrain and define daughter cell boundaries [35]. Several proteins are known to be essential for IMC and BC biogenesis, including IMC sub-compartment proteins (ISP)1 and 3, merozoite organising protein and CINCH [36–38], and others [39]. Transition from a multinucleated schizont to multiple, individual merozoites containing a single haploid nucleus occurs via PM invagination, redistribution of the BC to the PM and reorganisation of the IMC [40]. The PM and IMC subsequently fully enclose the newly forming merozoites, which are ultimately pinched off following the final segmentation of the genome and cytokinesis facilitated by schizont egress antigen-1 [41] (see also [12] for essential players in DNA segmentation). In addition, throughout schizogony the mitochondrion and apicoplast expand and are divided such that one of each is distributed to every merozoite [42,43]. Key components of this process include the dynamins DYN1 and DYN2 and autophagy-related protein 8 (for in depth review, see [11]).
Rapid cell division occurs in male gametocytes in the mosquito midgut. In the mosquito's blood meal, a temperature drop to ∼20–25°C, a rise in pH and the presence of the metabolic intermediate xanthurenic acid (XA) activate gametocytes to begin gametogenesis [44]. Three rounds of rapid, closed mitosis occur in the male gametocyte over 8–12 min, increasing DNA content from 1N to 8N, followed by karyokinesis, cytokinesis and the budding of eight microgametes in a process termed exflagellation (Figure 1B). In P. falciparum the falciform (or sickle-shaped) microgametocyte rounds-up upon activation, accompanied by formation of a single MTOC that develops into a tetrad of basal bodies attached to the spindle pole. Subsequent formation of eight basal bodies results in nucleation of eight axonemes, which elongate and segregate with each endomitotic division [44]. Each MTOC is closely aligned with a nuclear pore, and at exflagellation each newly formed gamete contains a single haploid genome in a nucleus that has been pulled through the basal body wrapped around an axoneme [45].
Male gametogenesis is complex and highly co-ordinated. Key phosphoregulators and other well-established components have been reviewed extensively elsewhere [44,46], with calcium-dependent PK 4 (CDPK4), serine-arginine PK 1 (SRPK1) and mitogen-activated PK 2 (MAPK2), as well as the metallo-dependent PP, PPM1, known to be essential drivers in Plasmodium berghei (Pb) [31,47,48]. The essential roles of CDPK4, SRPK1 and MAPK2 for P. falciparum male gametogenesis have been confirmed [49–51]; but an essential role for PfPPM1 is yet to be determined. Global phosphoproteome and interactome studies have begun to identify CDPK4, SRPK1 and MAPK2 substrates [47,52]. CDPK4 has three key roles during male gametogenesis mediated by three SOC1, SOC2 and SOC3. Interaction of CDPK4 with SOC1 within the first 10 s of activation facilitates the loading of the MCM2-7/Cdt1 complex onto the ORC1-5/Cdc6 complex, thereby assembling the pre-replicative complex and initiating DNA replication [52]. In the next 30 s, CDPK4 phosphorylation of SOC2, a microtubule-associated protein facilitates elongation of tubulin. Finally, 9 min after activation CDPK4 phosphorylates SOC3 to initiate axoneme motility and complete cytokinesis [52]. Additionally, a divergent cyclin/cyclin-dependent kinase (CRK5) was recently proposed to be in a signalling cascade with CDPK4 and SOC2 that drives origin of replication firing in P. berghei [18] and P. falciparum [53]. Deletion of SRPK1 results in a disruption of phosphosites that do not overlap with those of CDPK4 deletion mutants such as proteins involved in microtubule motor activity, actin binding and origin of replication recognition, suggesting that SRPK1 has different substrates to CDPK4 in addition to being regulated itself by CDPK4 [47]. However, some altered phosphosites were common to both deletion mutants, including those on proteins involved in microtubule-based movement and ATP binding. In P. falciparum, SRPK1 gene deletion using CRISPR/Cas9 technology affected transcript splicing and abundance for genes coding many proteins involved in microtubule morphogenesis, cyclic nucleotide metabolism, lipid metabolism, osmophilic body formation and crystalloid components [49]. The substrates of MAPK2 are less known; however, PbMAPK2 is essential for chromosome condensation prior to exflagellation [54], potentially driving axoneme motility, and in P. falciparum it has been shown to co-ordinate axoneme beating [51].
Recently, new components that may have vital functions in male gametogenesis and exflagellation have been identified. Gametogenesis essential protein 1 (GEP1) has recently been identified as a master trigger for gametogenesis, essential for XA-stimulated activation. Disruption of GEP1 gene abolished XA-stimulated cyclic GMP synthesis and completely ablated downstream cellular and signalling events [55]. Deletion of a gene encoding a P. falciparum protein with an AT-rich DNA interaction domain (PfARID) resulted in down-regulation of genes important for gametogenesis, with complete ablation of exflagellation [56]. Other players include Pb22, APC3 and radial spoke protein 9 (RSP9) in P. berghei [57–59], p25α in Plasmodium yoelii [60] and the patatin-like phospholipase, PLP2 in P. falciparum [61]. For a complete overview of essential regulators of gametogenesis, see [44].
The rapid rounds of mitosis during male gametogenesis are driven by formation of microtubular spindles to facilitate segregation of sister chromatids to the spindle poles. Within the first few seconds following microgametocyte activation, clusters of genes coding proteins involved in microtubule activity, microtubule binding and microtubule-based movement are activated [47], along with a tetrad of kinetosomes that form the MTOCs and kinetochores [62,63]. These proteins anchor mitotic spindles in the nucleoplasm and facilitate growth of axonemes in the cytoplasm [64]. A combination of live-cell imaging of tagged kinesin-8B, NDC80, kinesin-5 and spindle-assembly abnormal (SAS) 4 with transmission electron microscopy have revealed key aspects of the highly dynamic kinetosome and MTOC formation [62,63,65,66]. NDC80 is a highly conserved component of the kinetochore that mediates chromosome attachment to spindle microtubules and can be used as a marker for chromosome segregation [67], Following completion of each round of male gametocyte chromosome segregation, NDC80 is located at distinct foci in the nucleus, but extends to form a bridge across the entire nuclear body during segregation, while the basal body marker, SAS4 remains associated with the MTOC, making focal points in the cytoplasm [65] (Figure 2B). Kinesin-5, a molecular motor that is structurally and functionally conserved across eukaryotes and involved in spindle pole separation during mitosis [68], is also found located on mitotic spindles, colocalised with NDC80 during male gametogenesis. This interaction may be regulated by PP1, since PP1 co-localises with NDC80 at the kinetochores [69].
Four recent studies have highlighted the importance of microtubule end-binding proteins (EBs) for spindle-kinetochore attachment, chromosome segregation and partitioning of nuclei during male gametogenesis. EB1, found only in Plasmodium, has been found to decorate the full-length of spindle microtubules during male gametogenesis [64,70,71]. Deletion of the EB1 gene results in anucleate P. falciparum male gametes [72], and complete ablation of oocyst development in both P. berghei and P. falciparum [70–72]. Regulation of EB1 activity is co-ordinated by the divergent Aurora Kinase 2, which associates with microtubule proteins near the spindle-kinetochore interface, and potentially forms a microtubule-anchoring complex comprised of EB1, MISFIT and myosin K (MyoK) [71].
In the oocyst there is a lengthy period of growth over 1–3 weeks with several rounds of DNA replication and karyokinesis followed by a final step of cytokinesis to generate hundreds of haploid sporozoites in a process termed sporogony. The density of oocysts on the midgut wall is heavily determined by the gametocyte density in the bloodmeal [73] and, despite the lengthy process (especially when compared with male gametogenesis), oocyst development is highly dynamic.
Differentiation of the oocyst from the ookinete following attachment to the basal lamina of the midgut wall is thought to be triggered by interaction of P25/28, circumsporozoite- and TRAP-related protein and secreted ookinete adhesive protein with laminin [74]. This initiates sporogony with an increase in the size of the oocyst to a diameter of 50–60 µm, accompanied by retraction of the parasite PM around the developing sporoblast forming deep invaginations and cytoplasmic lobes. Ultrastructural microscopy and live-cell imaging have so far not identified when cytokinesis occurs during this process, although the kinetochore marker NDC80 is revealed at multiple foci adjacent to the nuclear DNA [63]. Several proteins are known to be essential for completion of sporogony, for de novo fatty acid synthesis (FAS) and for scavenging components from the host (in particular the FASII pathway [75]), DNA replication and metabolism [76] and evasion of the mosquito immune system [77]. Reverse genetic studies have shown that deletion of genes encoding 3-hydroxyacyl-CoA dehydratase (DEH — a component of FAS), the DNA repair protein meiotic recombination 11 (MRE11), P-type cyclin, cyclin 3 (CYC3) and several of the Limulus clotting factor C, Cochlin and Lgl1 (LCCL) lectin domain adhesive-like proteins (LAPs — essential for crystalloid biogenesis [78], see below) results in complete ablation of sporogony and parasite transmission [44,79]. Reversible protein phosphorylation also appears to facilitate sporogony, since deletion of the genes for PKs cyclin-G associated kinase (GAK) and PK7, along with the metallo-dependent PP, PPM5 also completely ablates sporogony [31,48]. Recent studies have identified further regulators of Plasmodium Infection of the Mosquito Midgut Screen (PIMMS)-01, -57, and -22 [80], and aquaporin 2 [81].
An organelle that may be crucial for oocyst development and sporogony is the crystalloid. These tightly packed, small spherical structures are found only in ookinetes and young oocysts but their origin and function remained elusive until recently (reviewed extensively in [82]). Namely, deletion of the six LCCL LAPs results in failure of the oocyst to undergo sporogony [78,79,83,84], as do deletion mutants of the recently discovered CRystalloid Oocyst Not Evolving gene [85], among others [82].
Answering these questions will require a combination of reverse genetics, state-of-the-art microscopy, cell biology, proteomics, phosphoproteomics and advanced bioinformatics, which will uncover exciting lines of research to shed light on the divergent aspects of Plasmodium cell division.
The authors declare that there are no competing interests associated with the manuscript.
D.S.G. and R.T. are supported by a BBSRC grant [BB/X014681/1 and BB/X014452/1] and ERC advance grant funded by UKRI Frontier Science [EP/X024776/1] to R.T., M.Z. and R.T. are supported by the BBSRC [BB/N017609/1]. A.A.H. is supported by The Francis Crick Institute [FC001097], which receives its core funding from Cancer Research UK [FC001097], the UK Medical Research Council [FC001097], and the Wellcome Trust [FC001097].
Open access for this article was enabled by the participation of University of Nottingham in an all-inclusive Read & Publish agreement with Portland Press and the Biochemical Society under a transformative agreement with JISC.
Conceptualisation: D.S.G., M.Z., and R.T.; Investigation: D.S.G., M.Z., and R.T.; Supervision: D.S.G., M.Z., A.A.H., and R.T.; Visualisation: M.Z.; Writing — original D.S.G., M.Z., A.A.H., and R.T.; Writing — review and D.S.G., M.Z., A.A.H., and R.T.