Moving on: How malaria parasites exit the liver

An essential step in the life cycle of malaria parasites is their egress from hepatocytes, which enables the transition from the asymptomatic liver stage to the pathogenic blood stage of infection. To exit the liver, Plasmodium parasites first disrupt the parasitophorous vacuole membrane that surrounds them during their intracellular replication. Subsequently, parasite‐filled structures called merosomes emerge from the infected cell. Shrouded by host plasma membrane, like in a Trojan horse, parasites enter the vasculature undetected by the host immune system and travel to the lung where merosomes rupture, parasites are released, and the blood infection stage begins. This complex, multi‐step process must be carefully orchestrated by the parasite and requires extensive manipulation of the infected host cell. This review aims to outline the known signaling pathways that trigger exit, highlight Plasmodium proteins that contribute to the release of liver‐stage merozoites, and summarize the accompanying changes to the hepatic host cell.


| INTRODUC TI ON
Parasites of the genus Plasmodium are intracellular pathogens and cause malaria.During their complex life cycle, they undergo essential phases of replication in both the insect vector and the vertebrate host.The symptoms of malaria occur while parasites are replicating in the blood of infected individuals, whereas the preceding multiplication of parasites in the liver is asymptomatic.Motile Plasmodium sporozoites are delivered into the skin of the vertebrate host by the bite of an infected mosquito and actively invade blood vessels to be transported to the liver.Here, they invade hepatocytes and develop inside a parasitophorous vacuole (PV) separated from the host cell cytoplasm by the PV membrane (PVM).Although replication in the liver is limited to one infection cycle, the parasite multiplication during this cycle is extensive and generates thousands of merozoites capable of infecting red blood cells.The efficient transition of these merozoites from the liver stage to the blood stage is a crucial step in the life cycle of Plasmodium parasites that enables subsequent mass proliferation of parasites in the blood and ultimately transmission to the mosquito vector.
After an initial developmental reprogramming and growth phase, liver-stage Plasmodium parasites grow as schizonts: the nuclei divide in the absence of cell division resulting in a large multinucleated cell.In the late stages of parasite development, the plasma membrane of the parasite begins to invaginate leading to the creation of a cytomere.Subsequently, individual merozoites form through cytokinesis (recently reviewed by Roques et al., 2023).In the rodentinfecting Plasmodium species, such as Plasmodium berghei and Plasmodium yoelii, that are often used for liver-stage studies, this developmental process takes approximately two and a half days (Sturm et al., 2006;Tarun et al., 2006).For the human-infecting species, this liver-stage development typically requires at least a week (Goswami et al., 2020;Mikolajczak et al., 2015;Warrell & Gilles, 2002).In vitro, the process of P. berghei egress was observed to begin roughly within 1 h of completed merozoite development (Burda et al., 2015).
The first observable step in egress is the rupture of the PVM, which is followed by the release of merozoites into the host cell cytoplasm (Figure 1a-c).In vitro, this process is associated with a detachment of the infected cells from the cell culture dish.Detachment of merozoite-filled cells from neighboring hepatocytes is likewise observed in liver tissue in vivo (Baer et al., 2007;Sturm et al., 2006).
Packages of infectious parasites bud from the plasma membrane of the infected cells (Graewe et al., 2011;Sturm et al., 2006; Figure 1d), and these host-cell-derived structures are called merosomes (Sturm et al., 2006).First evidence for merosome formation was provided by early electron microscopy analysis of fixed infected livers (Meis et al., 1985).This study showed that merozoites are released in groups into the bloodstream, but from this study, it was not evident that parasites are still surrounded by a host cell membrane.Longterm time-lapse microscopy as well as in vivo intravital imaging of P. berghei-and P. yoelii-infected hepatocytes revealed that up to several thousand merozoites are released from the infected hepatocyte into the sinusoidal lumen within host-cell-derived membrane (Baer et al., 2007;Graewe et al., 2011;Sturm et al., 2006).In a study using the rodent-infecting species P. yoelii, merosomes that emerge from infected liver cells into the vascular system were shown to travel to the lung where they rupture and release merozoites that can immediately infect red blood cells to initiate the pathogenic blood phase of infection (Baer et al., 2007;Figure 1e,f).Merosome formation is not a rare event: in multiple studies, merosomes were observed to emerge from more than half of the tissue-resident infected cells under observation (Baer et al., 2007;Sturm et al., 2006).Nor is merosome formation exclusive to the rodent-infecting species, since similar structures have been observed emerging from Plasmodium falciparum-infected human hepatocytes engrafted into immunocompromised mice (Vaughan et al., 2012).
In this review, we aim to summarize the known signaling pathways and proteins contributing to the release of liver-stage merozoites and discuss the parasite-induced changes to the host cell during merosome formation.We will highlight the advances that have been made to understand Plasmodium's exit from the liver stage but also indicate gaps that need to be filled to better understand this fascinating process by which hepatocyte-derived merozoites transit so successfully to the blood circulation.

| Cyclic nucleotide and calcium signaling in egressing merozoites
Cyclic nucleotide and calcium signaling have been shown to be shared features of blood-stage and liver-stage egress.To better contextualize the information on signaling pathways in liver-stage egress, it is worth outlining some main signaling events in blood-stage merozoites.Like in the liver stage, PVM rupture precedes bloodstage merozoite release from infected red blood cells (Wickham et al., 2003).In mature blood-stage schizonts prior to PVM rupture, an increase in the parasite cyclic guanosine monophosphate (cGMP) level is facilitated by the guanylyl cyclase GCα (Nofal et al., 2021) and is needed for activation of the central cGMP-dependent protein kinase (PKG).Asexual blood-stage parasites without functional PKG complete intracellular development until the schizont stage, but fail to egress from the PV (Koussis et al., 2020;Ressurreição et al., 2020).
Drug-induced increases in cGMP can even stimulate premature egress of non-infectious blood-stage merozoites via PKG (Collins F I G U R E 1 Timeline of liver-stage egress.After merozoites form at the end of the liver stage, the PVM ruptures (gray dotted line) and liver-stage merozoites are released into the host cytoplasm.The infected cell then detaches from the neighboring cells and merosomes bud from the infected cell plasma membrane directly into the liver sinusoids.Merosomes remain intact in the blood circulation until they reach the lung, where merosomes rupture and release merozoites that invade red blood cells.PVM, parasitophorous vacuole membrane.et al., 2013).PKG activation results in hydrolysis of phosphoinositides and increased levels of intracellular calcium (Brochet et al., 2014).
The calcium-dependent protein kinase 5 (CDPK5) was shown to be one downstream factor interacting with PKG and its depletion heavily impairs the egress of P. falciparum blood-stage parasites (Absalon et al., 2018;Dvorin et al., 2010).Further, a protein called 'important for calcium mobilization-1' (ICM1) is phosphorylated in a PKG-dependent manner and parasites lacking ICM1 are able to escape from the PV but only very rarely egress from the surrounding erythrocyte plasma membrane (Balestra et al., 2021).Notably, the related apicomplexan parasite Toxoplasma gondii also relies on calcium and cyclic nucleotide signaling for its egress from the host cell.Signaling is mediated, among other proteins, by forms of CDPK, PKG, and the cAMP-dependent protein kinase (PKA) (Jia et al., 2017;Lourido et al., 2012;Nofal et al., 2022).As opposed to Toxoplasma, signaling via PKA likely is dispensable for the egress of blood-stage Plasmodium (Patel et al., 2019) and some evidence suggests that this may also be the case for Plasmodium egress from the liver stage (Choudhary et al., 2019).
Plasmodium PKG is not only required for egress of merozoites from blood cells but also from liver cells.Because PKG is essential for blood-stage egress and at various other points of the Plasmodium life cycle (Brochet et al., 2014;McRobert et al., 2008;Taylor et al., 2010), conditional genetic targeting was required to study this enzyme in the liver stage (Falae et al., 2010).P. berghei parasites from which the pkg gene was conditionally deleted during the mosquito stage undergo normal intrahepatic development but fail to exit from the liver and establish blood-stage infection.In vitro, fewer merosomes are formed by cells infected with parasites lacking PKG expression.Further studies confirmed that PKG activity is necessary for efficient parasite egress from the liver (Govindasamy et al., 2016): The specific PKG inhibitor Compound 1 was used to treat parasites either harboring wild-type PKG or an enzymatically active, but Compound 1-insensitive version of PKG.While Compound 1 reduced in vitro merosome formation from cells infected with parasites expressing wild-type, Compound 1-sensitive PKG, this effect was less pronounced in infection with Compound 1-resistant PKG-expressing parasites.These results indicated that the enzymatic activity of PKG is important for efficient merosome formation.
In addition to PKG, CDPK5 was recently shown to play a role in liver-stage egress (Govindasamy & Bhanot, 2020).The calciumdependent protein kinase is needed for the efficient formation of merosomes.It is expressed in mature liver-stage parasites and localizes to the periphery of detached merosomes in vitro.Like with PKG, conditional targeting was required to study the role of CDPK5 in the late liver stage.Auxin-induced CDPK5-depletion reduced merosome formation but did not have the severe effect seen upon knockout of PKG.This result might be explained by incomplete degradation of CDPK5 in the specific experimental setup; some full-length CDPK5 as well as a partially degraded protein species remained in liver-stage schizonts after auxin treatment.Additionally, activity of other members of this protein family, like CDPK4, or other proteins involved in egress may compensate for the absence of CDPK5.The study further suggested that CDPK5 may be important for the final step of liver exit, the release of infectious merozoites from merosomes in the vasculature (Govindasamy & Bhanot, 2020).The authors raise the possibility that merosome rupture could be triggered by intramerosome signaling events occurring in the lung capillaries.

| Molecular events upstream of PKG activation
Little is known about the events upstream of the described signaling pathways for either liver-or blood-stage egress.Two parasite proteins that likely act upstream of GCα were recently shown to be involved in efficient blood-stage egress.The membrane protein cell division control protein 50B (CDC50B) was proposed to be required for efficient cGMP synthesis by GCα (Patel et al., 2022), and P. falciparum parasites lacking CDC50B expression took longer to egress.
Another protein that was recently shown to act early during bloodstage egress is the P. falciparum protein phosphatase 1 (PP1) (Paul et al., 2020).PfPP1 expression is essential for blood-stage development as well as for egress from the PV.The authors propose that the phosphatase directly dephosphorylates GCα and facilitates GCα interaction with phosphatidylcholine as a trigger for egress signaling (Paul et al., 2020).
All the aforementioned proteins that are known to play a role in blood-stage egress are also expressed by P. berghei during late liverstage development (Figure 2).The presence of GCα, ICM1, PP1, and CDC50B at the onset of liver egress may suggest that one or several of these factors are not only needed for blood-stage egress but are also involved in liver exit, as was shown for PKG and CDPK5.However, increased expression of these proteins during late liver-stage development and egress might as well point toward a function in erythrocyte invasion and establishment of early blood-stage infection.Accordingly, a potential function of these factors in liver egress remains to be investigated experimentally.

| Proteases are essential players in the exit process
In blood-stage Plasmodium parasites, the above-described signaling events trigger the discharge of specialized organelles known as exonemes (Agarwal et al., 2013;Collins et al., 2013).The Plasmodium subtilisin-like serine protease SUB1 localizes to these exonemes and is discharged into the PV prior to egress of blood-stage merozoites (Collins et al., 2013;Yeoh et al., 2007).Upon delivery into the blood-stage PV, SUB1 proteolytically modifies PV-resident proteins and merozoite surface proteins and is required for egress (Das et al., 2015;Koussis et al., 2009;Silmon de Monerri et al., 2011;Yeoh et al., 2007).SUB1 is also expressed in late liver-stage parasites and, like in the blood stages, it is proteolytically processed and localizes to punctate structures that are presumed to be exonemes (Suarez et al., 2013).
The involvement of SUB1 in the liver stage was studied in two independent studies through conditional deletion of the sub1 gene from P. berghei in the mosquito stage (Suarez et al., 2013;Tawk et al., 2013).The point of arrest was dependent on the conditional knockout strategy used, but in both studies, SUB1-deficient parasites failed to egress from the liver.Suarez and colleagues observed developmental arrest prior to cytomere formation, whereas Tawk and colleagues detected developed liver-stage merozoites, but failure of the PVM to rupture in cells infected with SUB1-deficient parasites.The latter results are comparable to the phenotype observed upon conditional knockout of SUB1 in P. falciparum blood stages that similarly led to an egress defect associated with a block in PVM rupture (Thomas et al., 2018).In spite of the differences between the two studies, SUB1 is clearly required for liver-stage Plasmodium egress, but whether the role of SUB1 in the liver is directly analogous to its role in the blood will require further investigation.SUB1 is synthesized as a catalytically inactive zymogen and itself requires proteolytic processing for its activity (Blackman et al., 1998).Although this processing is partially autocatalytic (Sajid et al., 2000), one step in this processing is mediated by the Plasmodium aspartic protease Plasmepsin X (PMX) in blood-stage parasites (Nasamu et al., 2017;Pino et al., 2017).Recombinant PfPMX can proteolytically process SUB1 among other substrates, and this processing is inhibited by the aspartic protease inhibitors 49c, WM4, and WM382 (Favuzza et al., 2020;Pino et al., 2017).49c treatment not only inhibits blood-stage egress but is also active against liver stages.Treatment of liver-stage P. berghei with 49c allows merozoite formation but disrupts egress both from cultured cells and from infected mouse livers (Pino et al., 2017).Treatment of infected cells with WM382 also reduced rates of liver-stage parasite egress in vitro (Favuzza et al., 2020).PMX is expressed late in the liver stage (Figure 2), and although 49c and WM382 can target other aspartic proteases, the data support the hypothesis that PMX activity is required for egress from the liver.If SUB1 processing is blocked in 49c-and WM382-treated liver-stage merozoites as it is in the blood and if this block is complete, processing of SUB1 is not required for liver-stage merozoite formation, which is notable given the differing phenotypes observed upon deletion of the gene encoding SUB1 described above (Suarez et al., 2013;Tawk et al., 2013).
Liver-stage egress is also inhibited by the cysteine protease inhibitor E64, which was shown to prevent complete PVM breakdown following liver-stage merozoite formation (Sturm et al., 2006;Sturm & Heussler, 2007).One family of parasite proteins that contain a papain-like cysteine protease domain is the serine-rich antigen (SERA) family.Members of the conserved SERA family are processed by SUB1 in blood-stage parasites and play a role in parasite egress (Aly & Matuschewski, 2005;Collins et al., 2017;Putrianti et al., 2020;Ruecker et al., 2012;Stallmach et al., 2015;Thomas et al., 2018;Yeoh et al., 2007).PfSERA5 is a SUB1 substrate and fulfills an important regulatory role during blood-stage egress by preventing the pre-mature release of parasites (Collins et al., 2017;Stallmach et al., 2015).PfSERA5 is one of the SERA proteins containing a catalytically inactive serine residue instead of a cysteine in the predicted protease domain (Arisue et al., 2007;Kiefer et al., 1996).
In P. berghei, PbSERA1 and PbSERA2 belong to this group of serinetype SERAs and have been shown to be expressed and processed in liver-stage parasites (Putrianti et al., 2010;Schmidt-Christensen et al., 2008).Both PbSERA1 and PbSERA2, however, are dispensable for the entire parasite life cycle, including liver-stage egress (Putrianti et al., 2010).PfSERA6, in contrast, is a proteolytically active cysteine protease that was shown to be essential for parasite egress from infected red blood cells (Thomas et al., 2018).PfSERA6 is activated by proteolytic processing by SUB1 and by autocatalytic processing that requires the SUB1-regulated cofactor MSA180 (Tan et al., 2021).The P. berghei ortholog of PfSERA6, PbSERA3, is also a substrate of SUB1 (Pace et al., 2019;Tawk et al., 2013), appears to be required for blood-stage growth (Putrianti et al., 2010), and is expressed and proteolytically processed during late liver-stage development of P. berghei parasites (Schmidt-Christensen et al., 2008).In infected erythrocytes, PfSERA6 was shown to enzymatically cleave β-spectrin to destabilize the red blood cell cytoskeleton after PVM breakdown (Thomas et al., 2018).Whether PfSERA6/PbSERA3 is similarly required for liver-stage egress or destabilizes the host cell cytoskeleton in this stage remains to be determined.
PbSERA4 is another cysteine-type SERA that was recently shown to be proteolytically processed in late liver stages and required for the efficient transition of P. berghei from the liver to the blood (Putrianti et al., 2020).P. berghei in which the PbSERA4 gene was disrupted replicate proficiently in the blood, infect mosquitos and undergo normal liver-stage development; however, they form fewer merosomes compared to wild-type parasites in infected cultured cells.Consequently, PbSERA4-deficient parasites need 2 days more than wild-type parasites to initiate detectable blood-stage infection after sporozoite injection into mice.Whether proteins are cleaved by PbSERA4 specifically prior to parasite egress from the liver remains unknown.
In short, multiple members of the SERA family are expressed at the end of both the liver and blood stages, but their essentiality in these stages varies.Whether the cofactor MSA180 plays a role in the function of SERA proteins in the liver remains to be determined.Deletion of SERA-encoding genes has frequently resulted in the upregulation of other SERA genes, suggesting that they may be able to functionally compensate for one another (McCoubrie et al., 2007;Putrianti et al., 2010Putrianti et al., , 2020)).Another possibility is that changes to the genomic region encoding the SERAs disrupt the regulation of this gene family.In either case, the value of having multiple proteins from this gene family expressed simultaneously in egressing parasites can only be determined by better defining their roles in this process.Protease cascades can facilitate the regulation of complex processes such as egress, and identifying the terminal substrates of the proteolytic cascades described above will be one key to understanding the mechanisms driving exit of liver-stage merozoites.

| Effector proteins involved in membrane rupture during liver-stage egress
The protein LISP1 is conserved across Plasmodium species and expressed only in the liver stage of development (Ishino et al., 2009).
LISP1 is localized to the PVM and plays a role in the breakdown of the PVM during liver-stage egress.Although LISP1 can be detected well before the cytomere stage in growing liver-stage schizonts, development of LISP1-deficient parasites is morphologically similar to wild-type parasites through the formation of merozoites, and these hepatocyte-derived LISP1-deficient merozoites are infectious to red blood cells when mechanically released from their host cells.
In spite of successful merozoite formation, the PVM surrounding these LISP1-deficient parasites in the liver remains intact and impedes merozoite release.When animals were infected with sporozoites in which the lisp1 gene was disrupted, lisp1-knockout parasites emerged into the blood from the liver, but in amounts notably lower than in animals infected with wild-type parasites.
While LISP1 is required for PVM breakdown, its primary amino acid sequence does not give any indication of an enzymatic activity it might have that would facilitate membrane rupture.Two classes of proteins that are expected to modify lipids or disrupt membrane integrity are phospholipases and pore-forming proteins.Plasmodium parasites express a small, conserved family of Plasmodium perforinlike proteins (PPLP) that are characterized by the presence of a membrane-attack complex/perforin domain (Kaiser et al., 2004).
These have been shown to be important for Plasmodium egress in particular stages of the life cycle (reviewed in Sassmannshausen et al., 2020).Although PPLP1 is needed for Plasmodium to exit hepatic cells and other tissues, this role occurs at the beginning of infection during sporozoite migration through cells in the liver (Ishino et al., 2005;Risco-Castillo et al., 2015;Yang et al., 2017).The function of the PPLPs in parasite egress after liver-stage development has not been analyzed so far.However, all of the PPLPs show relatively low expression in late liver stages based on RNAseq (Caldelari et al., 2019), which together with their absence from the P. berghei merosome proteome (Shears et al., 2019), does not suggest a major role in egress following the liver stage (Figure 2).
In contrast to the PPLPs, a function in liver-stage egress could already be shown for the P. berghei phospholipase (PbPL), which is required for PVM breakdown at the end of liver-stage development (Burda et al., 2015).PbPL shares some sequence similarity with the secreted human phospholipase lecithin: cholesterol acyltransferase, and recombinant PbPL shows phospholipase and membranolytic activity (Bhanot et al., 2005).PbPL is expressed on the surface of sporozoites and sporozoites lacking PbPL are impaired in their ability to traverse epithelial cell layers, suggesting a role for this phospholipase in modifying host cell membranes (Bhanot et al., 2005).During liver-stage development, PbPL localizes to the PVM in infected hepatocytes.Parasites devoid of PbPL undergo normal development in the liver until merozoites are formed, but the absence of PbPL leads to the formation of fewer detached cells in comparison with wild-type parasites, suggesting a potential role in egress.A quantitative live-cell imaging assay confirmed that PbPL-deficient parasites exhibit impaired PVM rupture, resulting in delayed parasite egress (Burda et al., 2015).Interestingly, PbPL-deficient sporozoites also have a defect in egressing from the oocyst (Burda et al., 2015), and a very recent study provided evidence that conditional knockout of the PbPL homolog in P. falciparum blood stages also leads to an egress phenotype there (Ramaprasad et al., 2023).Together, these findings highlight that PL is necessary for efficient parasite release during different egress processes of the malaria parasite occurring in the insect and the vertebrate host.
Apart from PbPL, another putative phospholipase has been implicated in liver-stage egress.The P. berghei phosphatidic acid preferring phospholipase A1 homolog PbPLA1 is expressed in all life cycle stages except sporozoites and localizes to the cytoplasm of the parasite.Disruption of PbPLA1 did not affect parasite development in the blood or the mosquito but similar to PbPL resulted in impaired PVM rupture and consequently reduced numbers of detached cells, implicating PbPLA1 in the egress process of liver-stage parasites (Srivastava & Mishra, 2022).The exact function of PbPLA1 still remains to be determined but the cytoplasmic localization of the enzyme argues against a direct lytic function on the PVM.

| Molecular basis of merosome formation
1.3.1 | An unusual kind of host cell death is associated with merosome formation Upon PVM rupture, substantial modifications of the host cell are induced by liver-stage parasites, culminating in the release of merosomes, in which hepatocyte-derived merozoites are released into the blood circulation.The formation of merosomes is linked to a special kind of host cell death that exhibits certain characteristics of typical apoptosis, such as nuclear condensation and mitochondrial disintegration but lacks others such as caspase activation.Importantly, the integrity of the host cell plasma membrane remains intact, and phosphatidylserine is not exposed to the outer leaflet of the plasma membrane, which would function as a typical "eat me" signal for phagocytotic cells (Sturm et al., 2006).Calcium accumulates in intracellular merozoites, suggesting that the parasite actively takes up calcium to prevent phosphatidylserine exposure on the surface of infected cells and merosomes (Sturm et al., 2006).How parasites activate or modulate this unique kind of host cell death on a molecular level is not understood, but an important protein for controlling host cell death could be the cysteine protease inhibitor PbICP, which has been shown to inhibit host cell apoptosis and to be released into the host cell upon PVM rupture (Rennenberg et al., 2010).

| Parasite-induced remodeling of the host cell membrane and cytoskeleton
Apart from host cell death, merosome formation is also associated with pronounced morphological changes in host hepatocytes.
In vivo, very flexible merosomes are formed that squeeze through the liver endothelium (Baer et al., 2007;Sturm et al., 2006).For merosomes to pass through the narrow gaps in the endothelium, their membrane must exhibit remarkable flexibility.Another essential requirement for merosome formation is the detachment of the infected cell from adjacent cells, which both necessitates the modulation of cell-cell connections and the host cell cytoskeleton.Indeed, live-cell imaging of P. berghei-infected hepatocytes expressing fluorescent reporter proteins to visualize the actin cytoskeleton revealed substantial parasite-induced changes to the cytoskeleton during egress.Upon PVM rupture and after subsequent cell detachment, the actin cytoskeleton completely loses its connection to the plasma membrane and was found as a condensed structure within host cells (Figure 3), thereby considerably destabilizing the host cell membrane (Burda et al., 2017).
Actin filaments are linked to the cytosolic face of the plasma membrane through several mechanisms (reviewed in Kapus & Janmey, 2013).Commonly, attachment occurs via a spectrin-based membrane cytoskeleton, consisting of spectrin tetramers that are connected to the plasma membrane and to actin by scaffolding proteins.Another type of connection between cortical actin and the plasma membrane is formed by ezrin, radixin, and moesin (collectively, the ERM proteins), which link actin to transmembrane proteins (Ponuwei, 2016).P. berghei liver-stage release is associated with pronounced changes in host plasma membrane proteins (Burda et al., 2017;Graewe et al., 2011).Live-cell imaging of fluorescently tagged host proteins revealed that proteins are rapidly lost from the host cell membrane during parasite egress, even when they contain multiple membrane-spanning domains (Figure 3).This included the single-pass membrane protein E-cadherin, which plays an important function in anchoring a cell in its surrounding tissue.
Proteins associated with the plasma membrane via palmitoylation, myristoylation, or by a GPI anchor were also lost from the plasma membrane upon egress (Burda et al., 2017).Notably, the elimination of membrane proteins appears to also take place during merosome formation in vivo, since the hepatocyte-specific single-pass transmembrane domain-containing protein Asialoglycoprotein receptor 1 (ASGR1) is absent from in vivo-generated merosome membranes (Baer et al., 2007).Given that proteins in the plasma membrane can serve as anchor sites for the cytoskeleton (Kapus & Janmey, 2013), the clearance of host plasma membrane proteins might contribute to the destabilization of the actin-plasma membrane linkage observed during liver-stage egress.
The connection of the spectrin-based membrane cytoskeleton to the plasma membrane as well as the binding of ERM proteins to the plasma membrane and to actin is mediated and regulated by phospholipids, with phosphatidylinositol 4,5-bisphosphate (PIP 2 ) being the most important lipid for this regulation.PIP 2 activates ERM proteins and this activation is necessary for their actin-plasma membrane linking activity (Ponuwei, 2016).Furthermore, spectrin and some spectrin-binding proteins directly bind to phosphoinositides, and spectrin can additionally strongly bind to the more abundant phospholipid, phosphatidylserine (Kapus & Janmey, 2013).Live-cell imaging of P. berghei-infected host cells expressing fluorescently tagged proteins that specifically bind to PIP 2 , phosphatidylinositol-3,4,5-trisphosphate and phosphatidylserine revealed that all of them lose their localization to the host cell membrane during Plasmodium egress-induced cell detachment (Figure 3).Likewise, the PIP 2dependent actin-plasma membrane linker protein ezrin also loses its host cell membrane association upon Plasmodium egress (Burda et al., 2017).These results suggest that parasites substantially modify the lipid composition of the host plasma membrane during their release from hepatocytes.Interestingly, PIP 2 not only has an effect on ERM protein function but it is also required for talin, a protein that links the cortical actin cytoskeleton to integrins, which are key proteins in mediating cell-extracellular matrix or cell-cell attachment (reviewed in Harburger & Calderwood, 2009;Mandal, 2020).Therefore, loss of PIP 2 could contribute to the observed disconnection of the actin cytoskeleton from the host cell membrane and to the detachment of the host cell from the surrounding tissue during liverstage egress, although this still needs to be proven experimentally.
The observed parasite-induced modification of host cell phospholipids during egress might also contribute to a weakening of the spectrin-based membrane cytoskeleton, whose fate during Plasmodium-induced cell detachment remains, thus far, unknown.
As mentioned above, β-spectrin is proteolytically cleaved within its actin-binding domain by PfSERA6 in asexual blood-stage parasites (Thomas et al., 2018).However, whether PfSERA6/PbSERA3 or another SERA protein can cleave the β-spectrin isoforms expressed in the liver and if this also contributes to the collapse of the host actin network remains to be tested.Alternatively, if the spectrin network remains intact in merosomes, it may confer a certain degree of stability for their passage through the endothelial barrier.
Considering the dramatic disruption of the plasma membrane composition and its link to the cytoskeleton, it is remarkable that the integrity of the host cell plasma membrane is maintained throughout liver egress.However, the key question remains, how all of these F I G U R E 3 Changes to the host cell upon breakdown of the parasitophorous vacuole membrane (PVM).Breakdown of the PVM is accompanied by the collapse of the host cell actin cytoskeleton (purple).At the same time, the composition of the host cell plasma membrane is dramatically modified.These modifications include the loss of specific lipids such as phosphatidylserine (PS), phosphatidylinositol 4,5-bisphosphate (PIP 2 ), and phosphatidylinositol-3,4,5-trisphosphate (PIP 3 ) from the plasma membrane.The PIP 2 -dependent actin-plasma membrane linker ezrin and various types of membrane-anchored proteins also lose their host cell membrane localization upon PVM rupture.
changes are induced on the molecular level by malaria liver stages.
Since the above-described host plasma membrane alterations only occurred after rupture of the PVM, an attractive hypothesis would be that they are induced by one or several PV/PVM-localized effector proteins that are released into the host cell upon PVM rupture.
Identification of these effectors in the future might be facilitated by looking at similar phenomena in other organisms.Noteworthy, for example, the pathogenic bacteria Vibrio parahaemolyticus and Shigella flexneri were both shown to employ inositol phosphate phosphatases to hydrolyze PIP 2 in the host cell membrane.These inositol phosphate phosphatases reduce the tethering force between the plasma membrane and PIP 2 -binding cytoskeletal anchoring proteins and induce significant cell blebbing.This disruption of plasma membrane integrity can lead to cell lysis by V. parahaemolyticus or facilitate the uptake of S. flexneri (reviewed in Ham et al., 2011).Whether Plasmodium liver stages similarly express PIP 2 -modifying enzymes that are released upon PVM rupture into the host cell, still needs to be investigated.
A possible alternative scenario could be that parasites trigger a signaling cascade in hepatocytes leading to the activation of host cell proteins that induce the observed modifications of the host cell membrane and thereby facilitate merosome formation.Remarkably, a host signaling cascade dependent on heterotrimeric G protein alpha-subunit GNAQ was identified in P. falciparum-infected erythrocytes and T. gondii-infected cells that is necessary for the cytolysis of infected cells and subsequent parasite release (Millholland et al., 2013).Knockdown or antibody-mediated depletion of GNAQ results in significant impairment in the egress of both P. falciparum and T. gondii.However, CRISPR/Cas9-based knockout of GNAQ in HeLa cells does not affect P. berghei liver-stage egress, while an egress defect for these cells infected with T. gondii was visible (Burda et al., 2020).This suggests that differences in host cell signaling may exist between the lytic egress mechanisms of P. falciparum blood stages and T. gondii, and the non-lytic merosome formation of Plasmodium liver stages.

| Merosome transport in vivo
In vivo, P. berghei-containing merosomes appear in the bloodstream from 46 h after sporozoite infection (Baer et al., 2007;Garnham et al., 1969;Sturm et al., 2006).Comparably, P. falciparum merosomes emerge from human hepatocytes engrafted into mice also upon merozoite formation, which for P. falciparum is 7 days after infection (Vaughan et al., 2012).The asynchronous formation of P. berghei and P. yoelii merosomes was observed over a period of at least 10 h (Baer et al., 2007;Sturm et al., 2006).Not only do parasites reach maturity asynchronously, but merosome release from an individual cell can last over several hours (Baer et al., 2007).In both the P. yoelii and P. falciparum infection models, some cells did release merozoites without the formation of merosomes (Baer et al., 2007;Vaughan et al., 2012).In P. yoelii-infected cells, this host cell lysis was observed sometimes even before merozoite formation, so it was assumed to be indicative of "abortive" parasite development (Baer et al., 2007).Whether or not merozoites released individually contribute to the initiation of the blood infection stage is undetermined.Some infected cells under observation failed to release their entire content into the liver sinusoids, and in contrast to merozoites in merosomes, merozoites that fail to exit the host after merosome budding can be labeled with propidium iodide, which indicates loss of plasma membrane integrity and parasite necrosis (Baer et al., 2007).
How the merosomes pass through the sinusoidal cell layer into the bloodstream remains, thus far, unknown.Departing parasites leave no vacant gap in the tissue upon exit, and it is tempting to speculate that physical pressure from neighboring hepatocytes following detachment of the infected cell aids the extrusion of merosomes into the sinusoids (Burda et al., 2017;Sturm et al., 2006).Merosomes initially vary in size when they emerge into the sinusoids, and their deformability allows them to travel within the hepatic vasculature in spite of their large size (Baer et al., 2007).However, they travel more slowly than the circulating blood cells and can remain in the liver vasculature for over 1 h (Sturm et al., 2006).The larger merosomes are reported to subdivide in the vasculature and become smaller and more uniform by the time they leave the liver, containing between 100 and 200 merozoites (Baer et al., 2007).
Blood leaving the liver passes through the right ventricle and the lungs before circulating in the rest of the body.Merosomes and individual parasites including freshly invaded red blood cells are found in the lung capillaries following liver infection, but merosomes have not been detected elsewhere in the circulation beyond the lungs (Baer et al., 2007).What triggers the release of merozoites from the merosomes in the lung capillaries is unknown.While capillary diameter very likely plays a role, merosomes exceed the size of both the liver and lung micro-vasculature.Baer and colleagues make the argument that larger merosomes are flexible enough to pass through the liver vasculature, so small merosomes can be expected to also traverse the lung microvasculature, and yet there is no evidence that merosomes exit the lungs intact.Whether or not other specific factors or interactions in the lung microvasculature stimulate merozoite release from merosomes is not known.An intriguing hypothesis is that invading well-oxygenated red blood cells in the lungs supports liverstage merozoites in establishing the first round of infection in the blood, but this idea remains to be experimentally tested.

| CON CLUDING REMARK S
In spite of the significant knowledge that has been gained in the last decades, many unanswered questions remain concerning how egress of liver-stage Plasmodium is triggered, how the PVM is breached, and how the parasites passage into the blood.More research is needed to unravel the molecular basis of this delicately orchestrated process.
One observation of many studies on liver-stage exit is that complete inhibition of egress is rarely seen.Most compounds or genetic modifications that impact liver-stage merozoite egress slow down the process or reduce the number of released parasites rather than block exit entirely.This phenomenon might suggest that the process of egress is well protected through redundancy and that parallel pathways exist to initiate and regulate this crucial step in the parasite life cycle.
Understanding the molecular details of the multiple steps in this process would not only expand our knowledge of parasite biology but also aid in targeting this stage with preventative treatment strategies.
Inhibiting egress from the liver would prevent blood-stage infection and disease.For chemical therapeutics, targeting parasite egress at both the liver and blood stages would be a particularly attractive strategy.The commonalities across these stages suggest that such an approach may be feasible, but the ability of parasites to escape blocks in egress necessitates a detailed understanding of the molecular players in both stages.Furthermore, the use of attenuated sporozoites is an attractive vaccine strategy to immunize against the pre-erythrocytic stages and prevent infection (Mwakingwe-Omari et al., 2021;Richie et al., 2015).Immunization with attenuated parasites that develop into mature liver stages produces a wider variety of antigens and confers better protection against reinfection than early-arresting parasites (Belnoue et al., 2004;Friesen & Matuschewski, 2011;Roestenberg et al., 2009).Therefore, reliably obstructing parasite exit from the liver, either chemically or through genetic modification, could support pre-erythrocytic vaccine development.
Low infection rates together with asynchronicity of liver-stage infection and egress present challenges to the study of merozoite exit from the liver.In blood cultures, merozoite egress can be reversibly blocked without impairing parasite development (Baker et al., 2017;Collins et al., 2013).A comparable technique has not yet been described for liver-stage parasites, but would likely enable a better understanding of egress dynamics and the impact of drugs or genetic modifications with subtle or fast-acting phenotypes.Hepatic culture models promise to advance the study of the liver stage across Plasmodium species (Chua et al., 2019;March et al., 2013;Ng et al., 2015;Pewkliang et al., 2018).The rates of sporozoite infection and maturation to merozoites are notably very low in these models (reviewed in Valenciano et al., 2022), but current progress in robust single-cell analysis (Afriat et al., 2022;Mancio-Silva et al., 2022) and advanced imaging techniques will also add to the toolbox of Plasmodium liver-stage research.The same is true for the increasing possibilities of reverse genetics in malaria parasites (Birnbaum et al., 2017;Knuepfer et al., 2017;Lee et al., 2019), in which conditional gene modification is now possible across the life cycle, not only in rodent-infecting species (Fernandes et al., 2021;Walker & Lindner, 2019) but also in P. falciparum (McConville et al., 2023;Tibúrcio et al., 2019).Together these advances will aid in understanding this fascinating process of egress of merozoites from the liver in more detail and could thereby lead to the identification of novel pathways for therapeutic intervention.

F
Gene expression of candidate egress proteins in the liver infection stage.Genes encoding proteins that are likely involved in (top) or likely dispensable for (middle) Plasmodium liver-stage egress are visualized with their role in asexual blood-stage egress.Candidate genes that were not yet tested for a role in liver-stage egress are shown at the bottom.Liver-stage gene expression is based on the data from Caldelari et al.The log of the TPM values of unique and non-unique peptides is visualized and colorized relative to all values displayed(Caldelari et al., 2019) (https://plasm odb.org/plasmo).Within each group, genes are arranged according to their expression levels at 60 h post-infection.Detection of candidate proteins in the merosome proteome was reported by Shears et al.(Shears et al., 2019).
AUTH O R CO NTR I B UTI O N S Alyssa Ingmundson: Conceptualization; writing -original draft; visualization.Mattea Scheiner: Conceptualization; writing -original draft; visualization.Paul-Christian Burda: Conceptualization; writing -original draft.ACK N OWLED G M ENTS We acknowledge funding by the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG) within the SPP2225: Exit strategies of intracellular pathogens (project number 446556156 to P.C.B. and project number 446465266 to A.I.).Open Access funding enabled and organized by Projekt DEAL.