Ex vivo maturation of the malaria parasite egress protease SERA6 aids pathway dissection and inhibitor development

Release (egress) of malaria parasites from host red blood cells (RBC) is a protease-dependent process involving breakdown of the RBC cytoskeleton by a parasite cysteine protease-like protein called SERA6. In the penultimate step of the egress cascade, SERA6 undergoes autoproteolytic maturation triggered upon cleavage by a serine protease called SUB1 and requiring interactions between SERA6 and fragments of another parasite protein called MSA180. Egress can be blocked by treatment of intraerythrocytic parasites with small molecules that prevent the autocatalytic SERA6 maturation step, suggesting that SERA6 is a druggable target. Here we describe the development of a cell-free in vitro system that recapitulates SERA6 maturation. We use the assay to confirm the strict requirement for MSA180 in SERA6 maturation by SUB1. Using a synthetic peptide substrate based on a predicted autocatalytic cleavage site we demonstrate that the fully mature SERA6 is an active proteolytic enzyme and we validate improved small molecule inhibitors of SERA6. Our best inhibitory compound efficiently blocks egress of asexual blood-stage parasites, confirming SERA6 as a new potential antimalarial drug target.

Malaria continues to exact extensive morbidity and mortality across the globe, with children below the age of 5 years and pregnant women at particular risk of severe disease (World Health Organization World Malaria Report 2025). Despite recent promising progress in the development of protective vaccines (Verma et al., 2025), chemotherapy remains crucial for malaria control. There is therefore widespread concern over the continuing spread of resistance of the causative agents, protozoan parasites of the genus Plasmodium, to mainstay artemisinin-based antimalarial drugs (Balmer et al., 2025; Boni et al., 2025). An improved understanding of fundamental parasite biology, especially that of the species causing the majority of severe disease, Plasmodium falciparum, is essential for the development of new target-based drugs. Such understanding is also invaluable for elucidating the mode of action of novel drug-like antimalarial compounds discovered through phenotypic screens (Hooft van Huijsduijnen and Wells, 2018; Siqueira-Neto et al., 2023).

Clinical malaria results from asexual replication of the parasite in circulating red blood cells (RBC). Parasite growth within the RBC, enabled through catabolism of RBC hemoglobin in a specialized parasite digestive vacuole, culminates in the formation of a mature multinucleated form called a schizont. Multiple daughter merozoites bud from the intracellular schizont, then rupture of the infected cell allows egress of the merozoites which rapidly invade fresh RBC to initiate a new replicative cycle. Egress involves several characteristic morphological and structural transitions. The first of these is thought to be permeabilization (‘poration’) of the intraerythrocytic parasitophorous vacuole (PV) within which the schizont resides (Hale et al., 2017), followed by rounding up of the PV which then ruptures with a concomitant microscopically discernible increase in mobility of the intraerythrocytic merozoites. PV rupture is rapidly followed by poration of the RBC membrane, then final breakage of the RBC membrane to allow merozoite dissemination in an explosive manner (Glushakova et al., 2005, 2009, 2010, 2018; Gilson and Crabb, 2009; Abkarian et al., 2011; Thomas et al., 2018).

Several lines of genetic and biochemical evidence have shown that egress is controlled through a proteolytic cascade triggered by the activation of a cGMP-regulated parasite kinase called PKG (Baker et al., 2020). PKG activity is itself dependent upon a guanylyl cyclase (GCα) (Nofal et al., 2021) as well as a serine-threonine phosphatase called PP1 (Seveno et al., 2025). PKG appears to function in coordination with a multi-pass membrane protein called ICM1 and a calcium-dependent protein kinase, CDPK5 (Balestra et al., 2021; Dvorin et al., 2010; Absalon et al., 2018). PKG activation leads to a parasite subtilisin-like serine protease called SUB1 being discharged from specialized merozoite subcellular organelles called exonemes into the lumen of the PV (Yeoh et al., 2007; Collins et al., 2013). There, SUB1 encounters a cysteine protease-like parasite molecule called SERA6 (Ruecker et al., 2012). Precise SUB1-mediated cleavage of the SERA6 precursor at two defined sites releases a central SERA6 module called p65, comprising an N-terminal prodomain fused to the papain-like putative catalytic domain. A rapid autocatalytic cleavage event then releases the SERA6 prodomain (referred to as p40) from p65, generating the terminal papain-like SERA6 domain, SERA6 PAP. This interacts with at least 2 fragments of a second PV-resident parasite protein called malarial SERA activator 180 (MSA180) which is also cleaved by SUB1 (Tan et al., 2021). Formation of the SERA6 PAP-MSA180 complex results in rapid degradation of the RBC cytoskeletal component β-spectrin, in turn leading to destabilization and rupture of the RBC membrane to allow merozoite egress (Thomas et al., 2018; Tan et al., 2021). Egress also involves at least one PV-resident phospholipase activity called LCAT (Ramaprasad et al., 2023), although the precise role of this enzyme in the pathway is unclear. Genetic ablation and/or pharmacological inhibition of any of these egress mediators, or of an aspartic protease called plasmepsin X (PMX) required for SUB1 maturation (Nasamu et al., 2017; Pino et al., 2017; Mukherjee et al., 2023; Withers-Martinez et al., 2024), modulates, delays or prevents egress (Figure 1). The key role of SERA6 maturation in the final stages of egress is also strongly supported by evidence that small molecules that inhibit the autocatalytic p65-to-PAP maturation step selectively block β-spectrin cleavage as well as RBC membrane rupture and egress, but do not prevent rounding up or rupture of the PV, nor the temporally associated poration of the RBC membrane (all of which are SERA6 independent) (Dans et al., 2020; Glushakova et al., 2009; Tan et al., 2021). These observations suggest that SERA6 could be a target for a new generation of protease-inhibitor based antimalarial drugs.

Despite this detailed picture of SERA6 maturation and its role in egress, it is unclear whether parasite components additional to SUB1 and MSA180 are required for SERA6 maturation. Furthermore, a direct demonstration that mature SERA6 PAP is an active protease capable of cleaving peptide-based substrates in an enzymatic manner in trans is lacking. Here, using parasite-derived SERA6 and MSA180 together with recombinant SUB1 we describe the development of a cell-free in vitro system that recapitulates SUB1-dependent SERA6 maturation. We use the assay to: (1) define the minimal requirements for SERA6 maturation; (2) to show for the first time that the fully processed SERA6 PAP is an enzymatically active protease; and (3) to validate improved small molecule inhibitors of SERA6, confirming SERA6 as a new potential antimalarial drug target.

Transgenic P. falciparum lines and maintenance

P. falciparum 3D7 and the transgenic P. falciparum lines SERA6-mTAP:loxP and MSA180-HA3:loxP were generated and maintained in vitro as described previously (Tan et al., 2021). To generate saponin extracts, highly mature schizonts enriched from the synchronized lines by arresting egress with C2 (Collins et al., 2013) were pelleted and resuspended at room temperature in a buffer composed of Tris-buffered isotonic saline (TBS) supplemented with 2 mM DTT, 2 µM C2 and 0.15% (w/v) saponin. The resulting extracts, enriched in released PV proteins, were clarified by centrifugation, snap frozen and stored in aliquots at -70°C until required.

Compounds and peptides

The CS compounds were synthesized by Enamine Ltd (Kyiv, Ukraine). Compound MMV676881 (a kind gift of MMV), the PKG inhibitor C2 (a kind gift of Simon Osborne, LifeArc UK), the NCATS compounds and the CS compounds were stored at -20°C in dry 100% DMSO at a concentration of 2–10 mM. Fluorogenic peptide PEP3 (Dabcyl-GNQIKLKSPYF-EDANS) and the HA peptide used for elution of HA-tagged protein from anti-HA antibody beads were synthesized in-house by the Crick Chemical Biology Science Technology Platform and were stored as 1 mM stock solutions in DMSO at -20 °C.

Affinity purification of HA-tagged parasite SERA6 and MSA180

Saponin extracts from C2-treated schizonts of the P. falciparum SERA6-mTAP:loxP and MSA180-HA3:loxP transgenic lines were processed in parallel using anti-HA antibody magnetic beads (Pierce) according to the manufacturer’s instructions (Thermo Fisher Scientific). Briefly, 50 µl aliquots of a suspension of magnetic beads were washed twice in TBS, 25 mM CHAPS, 2 mM DTT while aliquots (400–800 µl) of frozen saponin extracts containing HA-tagged SERA6 and MSA180 in TBS buffer, 2 mM DTT, 2 μM C2 and 0.15% (w/v) saponin with no additional protease inhibitors were thawed on ice. The extracts were supplemented with CHAPS to 25 mM final before being added to the washed magnetic beads. The resulting suspension was incubated on a rotating wheel for 2 h at 4 °C to allow binding. Unbound material was then removed and the beads washed five times in TBS, 25 mM CHAPS, 2 mM DTT before eluting bound protein by incubation with 200 µl of a solution of HA peptide (2 mg/ml in TBS) for 2 h at 37 °C. The eluted purified SERA6 and MSA180 were usually immediately used for reconstitution/processing and activity assays, or alternatively were stored at 4 °C for a maximum of 24 h before use. The starting saponin extracts, the unbound material, the first wash, the eluate samples and the beads (before and after elution) were retained for analysis by SDS PAGE and western blotting using nitrocellulose membranes (GE Healthcare). Membranes were probed with the primary anti-HA rat monoclonal antibody 3F10 (diluted 1:2,000), or a previously-described rabbit polyclonal anti-SERA6 antibody (called anti-S6C1: Ruecker et al., 2012) (1:1,000), or a polyclonal rabbit anti-MSA180 antibody (1:500) (Tan et al., 2021). HRP-conjugated polyclonal goat anti-rat (1:10,000) or goat anti-rabbit (1:3,000) antibody were used as secondary antibodies, followed by detection using Immobilon Western Chemiluminescent HRP Substrate (Millipore) according to the manufacturer’s instructions. Blots were visualized and documented using a ChemiDoc Imager (Bio-Rad) with Image Lab software (Bio-Rad).

Saponin extract, SERA6 and MSA180 reconstitution assays and SERA6 peptidase activity

Whole schizont saponin extracts or eluted, affinity-purified FL SERA6 and FL MSA180 proteins prepared as above were combined as required (usually 1:1 for the purified proteins) and made to 25 mM CHAPS, 15 mM CaCl 2 (a requirement for SUB1 activity) before adding 1.71 U/µl of rSUB1 in 50 mM Tris-HCl, pH 8.0, 150 mM NaCl. Processing was allowed to occur overnight at room temperature or 37°C, then assessed by western blotting analysis.

SERA6 peptidase activity was assessed in white 96-well opaque flat-bottom plates (Nunc) by monitoring fluorescence increases upon cleavage of the fluorogenic peptide PEP3 (Dabcyl-GNQIKLKSPYF-EDANS), used at final concentration of 10 µM in the assay. For each condition, a final volume of 200 µl/well was used with the following reaction buffer: 1xTBS, 25 mM CHAPS, 15 mM CaCl 2, 6 mM DTT and 1% DMSO. Fluorescence changes were monitored overnight at room temperature in a SpectraMax M5e Multi-Mode Microplate Reader (Molecular Devices LLC) (Ex 340 nm, Em 492 nm), taking readings at intervals of 10 min and blanking against reaction buffer only. Assays were carried out in duplicate and GraphPad Prism software was used to analyze the data. At the end of each run, the protein solutions from the flat-bottomed plates were extracted directly into SDS and stored at -20 °C prior to SDS-PAGE and western blot analysis.

CRK1_1 selectivity assays

Compound CRK1_1 was assessed for inhibitory activity against rSUB1, PMX and papain. Purified rSUB1 was diluted 1:500 from a stock solution (228 U/ml stock in 20 mM Tris–HCl pH 8.2, 150 mM NaCl, 10% glycerol) into SUB1 reaction buffer (20 mM Tris–HCl pH 8.2, 150 mM NaCl, 12 mM CaCl 2, 25 mM CHAPS). This was supplemented with CRK1_1 from a 100% DMSO stock solution (1 mM stock, 10 μM final). Following a 5 min incubation, the mixture was supplemented with the fluorogenic peptide SUB1 substrate SERA4st1F-6R12 (Withers-Martinez et al., 2012) (Ac-CKITAQDDEESC-OH labelled on both Cys side-chains with tetramethylrhodamine; 0.1 μM final). Recombinant PMX (Withers-Martinez et al., 2025) was similarly used at a final concentration of 0.36 μM in 25 mM sodium acetate pH 5.5, 0.005% Tween 20 reaction buffer, in this case using fluorogenic substrate Dabcyl-HSFIQEGKEE-EDANS (Ex 340 nm, Em 492) at a final concentration of 1.6 μM in the reaction buffer. The potent PMX inhibitor 7k (Kovada et al., 2023; Withers-Martinez et al., 2025) was used as a control (10 μM final). Commercially-obtained papain from papaya latex (Sigma, P4762) was diluted to 0.1 mg/ml in 25 mM Bis-Tris pH 6.5, 1 mM EDTA, 5 mM DTT and incubated for 30 min in the reaction buffer to allow activation before use at 0.01 mg/ml. In this case fluorogenic substrate Nα-Benzoyl-DLR-AMC (1mM stock in 100% DMSO) (Ex 380 nm, Em 460 nm) was used, at 10 μM final. The inhibitor control for papain was E64 (10 μM final). All assays were performed in white 96-well opaque flat-bottom plates (Nunc) at room temperature in a SpectraMax M5e Multi-Mode Microplate Reader (Molecular Devices LLC), taking readings at 3 min intervals and blanking with the reaction buffer. All assays were performed in duplicate and GraphPad Prism software used to analyze the data.

PEP3 cleavage mapping by reversed phase chromatography and mass spectrometry

Reversed-phase high pressure liquid chromatography (RP-HPLC) was used to monitor PEP3 cleavage. Samples (50 ml) from assays containing purified FL SERA6 and FL MSA180 treated with rSUB1 in the presence of fluorogenic peptide PEP3 as described were injected onto a Vydac C18 RP-HPLC column (4.6 mm x 25 cm) using a 100 μl loop and eluted with a 0-90% (v/v) gradient of acetonitrile in 0.1% (v/v) trifluoroacetic acid over a 30 min period, monitoring both absorbance (at 215 nm) and fluorescence (Ex 340 nm, Em 492 nm). Peaks of interest were collected manually and subjected to LC-MS analysis to identify their makeup. Briefly, a M-Class LC (Waters, UK) delivered a flow of approximately 50 µl/min. A C4 UPLC BEH 1.7µm, 2.1 x 5 mm pre-column (Waters, UK), trapped the injected peptides prior to separation on a C4 BEH 1.7 µm, 1.0 x 100 mm UPLC column, eluting with a 20-min gradient of acetonitrile (2% to 80%). The analytical column outlet was directly interfaced via an electrospray ionization source, with a hybrid quadrupole time-of-flight (Q-TOF) mass spectrometer (SynaptG2Si, Waters, UK). Data was acquired over an m/z range of 350–4000, in positive ion mode with a cone voltage of 40 v. Scans were summed together manually and deconvoluted using MaxEnt1 (Masslynx, Waters, UK).

SERA6 homology modelling, covalent and noncovalent docking of nitrile compounds and peptides

SERA6 homology modelling was performed in ICM-Pro (version3.9-1c/MacOSX, Molsoft LLC, San Diego, CA) and based on the SERA5 proenzyme template (PDB: 6X44, 2.17Å), and using the AlphaFold2 model (Q9TY96), the ‘catalytic pocket’ of which was in an active induced binding mode to accommodate a portion of the inhibitory prodomain. The ICM Full Model Builder procedure was used, making use of the ICM force field and resulting in a fully refined model (including side chains and loops) with a good stereochemistry validation (1.04% outliers). Flexible covalent docking of the nitrile compounds into the active site of the SERA6 model was also performed in ICM-Pro. Protonation of the predicted catalytic histidine H810 (Nδ1) was enabled to favor inhibitor stabilization via hydrogen-bonding with the thiolate form of the catalytic cysteine C644. The SERA6 prodomain binding region (N485-P492) was initially used to define the boundaries within the enzyme active site pocket for the docking procedure, then the entire prodomain region (D371-N611) was removed prior to docking. Potential energy maps of the SERA6 receptor and docking preferences were set up using the program default parameters. Energy terms were based on the all-atom vacuum force field ECEPP/3 and conformational sampling was based on the biased probability Monte Carlo (BPMC) procedure. The nitrile-based CS compounds were derived from the original purine and triazine nitrile actives and were rationally designed using the ICM-Pro ligedit option. Three independent docking runs were performed, with a thoroughness (length of simulation) varied from 3 to 10 and the selection of 3 docking poses. Ligands were ranked according to their ICM energetics (ICM score, unitless), which weighs the internal force-field energy of the ligand combined with other ligand-receptor energy parameters.

Flexible non-covalent docking of the SERA6-derived peptide (I602-F609 of the SERA6 primary amino acid sequence) was also performed using ICM-Pro. As before, the SERA6 prodomain segment (N485-P492) was used to define the active site SERA6 receptor pocket and then removed from the receptor. Potential energy maps of the SERA6 receptor pocket and docking preferences were set up using the program default parameters. Three independent docking runs were performed, with a thoroughness (length of simulation) varying from 5 to 10 and the selection of 2 docking poses.

Structural modelling of MSA180

MSA180 was modelled using the AlphaFold3 server (https://alphafoldserver.com) using the MSA180_PF3D7_1014100 amino acid sequence as entry. The MSA180 AlphaFold3 model 4 was used for structural investigation with the DALI protein structure comparison server (http://ekhidna2.biocenter.helsinki.fi/dali/) to search for similar folds. This identified the ubiquitin-like modifier FAT10 (PDB: 6GF1) as a match for the C-terminal region of the MSA180 p45 domain.

Ex vivo maturation of SERA6 in cell-free parasite extracts using recombinant SUB1

Maturation of SERA6 in the minutes leading up to egress is initiated by SUB1-catalysed cleavage of the SERA6 precursor within the PV to release the central p65 module, which comprises an N-terminal prodomain fused to the papain-like putative catalytic domain (Tan et al., 2021; Smith et al., 2020). An autocatalytic cleavage event then releases the SERA6 prodomain p40, generating the free terminal SERA6 PAP, thought to be the active form of SERA6. Formation of SERA6 PAP and its recruitment of fragments of MSA180 is thought to be important for the β-spectrin cleavage that facilitates lysis of the host RBC cytoskeleton and membrane to allow merozoite egress. Recombinant, enzymatically-active SUB1 has been available for some time (Withers-Martinez et al., 2014; Sajid et al., 2000). In contrast, extensive previous attempts by us and others to generate correctly-folded recombinant full-length (FL) SERA6, FL MSA180 or SERA6 PAP have failed, hindering efforts to further define the mechanism of and requirements for SERA6 maturation. To overcome this hurdle, we set out to develop an assay that exploits native SERA6 and MSA180 isolated from parasites cultured in vitro in human RBC.

The non-ionic glycoside detergent saponin readily permeabilizes RBC membranes and the PV membrane, but does not rupture the blood-stage malaria parasite plasma membrane (Burghaus and Lingelbach, 2001). On this basis, saponin has been widely used for the preparation of parasite extracts enriched in soluble PV proteins, including FL SERA6 (e.g. Ruecker et al., 2012). We reasoned that a saponin extract of mature intraerythrocytic schizonts arrested just prior to egress with the PKG inhibitor 4-[7-[(dimethylamino)methyl]-2-(4-fluorphenyl)imidazo[1,2-α]pyridine-3-yl]pyrimidin-2 amine (compound 2, C2) should contain FL SERA6 as well as all the components required for SERA6 maturation except for SUB1, which remains sequestered in exonemes in C2-arrested parasites (Collins et al., 2013). We predicted that addition of exogenous recombinant SUB1 (rSUB1) to such ex vivo extracts should trigger maturation of the endogenous FL SERA6. To test this notion, we generated extracts of mature, C2-arrested schizonts using a saponin-containing buffer. To enable downstream analysis, for this work we used the previously-described transgenic SERA6-mTAP:loxP P. falciparum line (Tan et al., 2021), which expresses fully functional SERA6 possessing an internal haemagglutinin (HA) epitope tag. The schizont extracts were then supplemented in vitro with rSUB1 and the fate of the endogenous FL SERA6 monitored. As shown in Figure 2A, this resulted in the expected time-dependent conversion of FL SERA6 to its mature PAP form, via the p65 intermediate. Importantly, conversion of p65 to PAP in the rSUB1-treated saponin extracts was blocked by the presence of the epoxide E64 (Figure 2B), a potent inhibitor of the autocatalytic step of SERA6 maturation and egress when used to treat intraerythrocytic parasites (Thomas et al., 2018; Tan et al., 2021). These results indicated that it is possible to reconstitute the entire SUB1-dependent SERA6 maturation pathway in a cell-free in vitro system and that the autocatalytic p65-to-PAP step of SERA6 maturation under such conditions remains susceptible to small molecule inhibitors.

Both SUB1 and MSA180 are necessary for complete ex vivo maturation of SERA6

The assay described above exploited as a source of FL SERA6 unfractionated saponin extracts, which contain a broad repertoire of PV proteins. Whilst in previous work we have shown through gene disruption experiments that both SUB1 and MSA180 are required for SERA6 maturation (Tan et al., 2021; Thomas et al., 2018), it has not been possible to address whether these are sufficient or whether additional components are required. We have previously described generation of the transgenic MSA180-HA3:loxP P. falciparum line (Tan et al., 2021) which expresses an epitope-tagged form of MSA180. We reasoned that we could take advantage of this line as well as the SERA6-mTAP:loxP parasite line to independently isolate the respective full-length MSA180 and SERA6 proteins from saponin extracts using affinity-purification, then recombine the isolated proteins in the presence of rSUB1 to test whether these three components alone are sufficient for SERA6 maturation. As shown in Figures 3A, B, the endogenous tagged SERA6 and MSA180 proteins could indeed be successfully isolated by adsorption onto anti-HA antibody beads followed by gentle elution from the beads using HA peptide. In the absence of added MSA180, purified FL SERA6 was converted by rSUB1 only to the p65 intermediate form, with no detectable formation of PAP (Figure 3C). In contrast, supplementing the eluted, reconstituted proteins with rSUB1 resulted in complete conversion of FL SERA6 to PAP (Figure 3D). FL MSA180 was also converted by rSUB1 in these experiments to a smaller product p45 (Figure 3D; Supplementary Figure 1), consistent with the presence of at least one SUB1 cleavage site within MSA180 as previously shown (Tan et al., 2021). Importantly, in the absence of added rSUB1 both FL SERA6 and FL MSA180 were predominantly stable over the period of incubation, with no detectable conversion to smaller products. The observed rSUB1-dependent maturation of SERA6 and MSA180 in the ex vivo system therefore mimicked the in vivo process. Taken together with our previous genetic data, these results show that SUB1 and MSA180 are both necessary for complete SERA6 maturation to PAP, with no apparent requirement for any other parasite component(s).

In previous work we showed that the MSA180 p40 and p45 fragments that interact with SERA6 PAP are derived from segments of MSA180 that are non-contiguous in the primary sequence and that formation of these fragments results from at least one internal cleavage mediated by SUB1 (Nagaoka et al., 2019; Tan et al., 2021). However, there is no experimental structural information on MSA180. To examine this issue in more detail we performed structural modelling of MSA180 using AlphaFold3. Intriguingly, this revealed that p40 and p45 are predicted to exist as folded domains that interact closely within the intact FL MSA180 (Figure 1; Supplementary Figure 2). Moreover, the C-terminal region of MSA180 p45 is predicted to possess a fold belonging to the ubiquitin-like (UBL) family of proteins (Winget and Mayor, 2010) that is solvent-exposed and conserved across orthologues of MSA180 in other Plasmodium species. The significance of this was not further explored in this study, but this modelling suggests the possibility that the MSA180 p45 UBL domain could play a role in its interactions with SERA6 PAP and/or component(s) of the RBC cytoskeleton.

Mature SERA6 is an enzymatically active protease

SERA6 maturation (conversion to PAP) in intraerythrocytic parasites is temporally associated with cleavage of host RBC β-spectrin; this is thought to lead to RBC membrane rupture through breakdown of the RBC cytoskeleton (Thomas et al., 2018; Tan et al., 2021). The major cleavage site within β-spectrin was mapped to a site close to its N-terminus, between the tandem calponin homology (CH) domains of the actin-binding region of β-spectrin. No amino acid sequence resembling this β-spectrin cleavage site is identifiable within the SERA6 sequence, but since prodomain cleavage during conversion of SERA6 p65 to PAP appears to be autocatalytic (as is common for many papain-like cysteine proteases; Verma et al., 2016), we reasoned that small peptides based on the internal cleavage site(s) within SERA6 p65 might act as substrates for PAP in trans. However, the internal cleavage site within SERA6 p65 has not been experimentally identified.

In cysteine proteases of the papain-like family, the S2 pocket of the active site cleft (Schechter and Berger nomenclature: Schechter and Berger, 1967) generally primarily defines enzyme specificity, with a common preference for hydrophobic and aromatic residues at the interacting P2 position of substrate (Rawlings et al., 2018). In the AlphaFold2 model of SERA6, a Tyr residue (Y783) is predicted to form the floor of the hydrophobic S2 pocket, restricting its size to favor interaction with Leu over larger hydrophobic residues. Similar to proprotease forms of two of the most closely related human cysteine proteases, cathepsin B (CatB) (Podobnik et al., 1997) and cathepsin L (CatL) (Coulombe et al., 1996) (Chowdhury et al., 2002), the SERA6 AlphaFold2 model predicts a portion of the prodomain interacting with and obscuring the enzyme active site in a reverse orientation to substrate (thus preventing cleavage), with a Leu sidechain (L489) filling the S2 pocket in SERA6. Autocatalytic activation of the cathepsins involves both unimolecular and bimolecular (or intermolecular) processes (Rozman et al., 1999; Pungercar et al., 2009; Ménard et al., 1998). In each case prodomain cleavage to generate the mature enzyme occurs at a position downstream from the site that initially fills the active site. At the primary sequence level these cleavage sites in CatB and CatL show little similarity to SERA6 (Figure 4A). To attempt to predict the putative internal SERA6 autocatalytic cleavage site, we therefore compared the AlphaFold2 model of P. falciparum SERA6 with the x-ray crystal structures of CatB and CatL (Figure 4B), focusing on the sites at which prodomain cleavage occurs in the cathepsins. Together with multiple sequence alignment of the SERA6 orthologues from several malaria parasite species, this analysis allowed us to tentatively identify a putative prodomain cleavage site within P. falciparum SERA6 (Figure 4A) that is conserved across SERA6 orthologues and includes Leu (L604 in P. falciparum SERA6) at the P2 position.

To test this prediction, we first took an 11-residue peptide sequence (GNQIKLKSPYF) corresponding to that spanning the putative prodomain cleavage site in SERA6 and used in silico docking into the AlphaFold2 model of the SERA6 papain-like catalytic domain to assess the capacity of the peptide to be accommodated in a substrate-like manner in the SERA6 active site. As shown in Figure 4C, the peptide was predicted to dock well into the SERA6 model, with the conserved L604 sidechain interacting with Tyr783 at the bottom of the SERA6 S2 pocket. Encouraged by this, we next produced a quenched fluorogenic peptide (Dabcyl-GNQIKLKSPYF-EDANS, called PEP3) based on the 11-mer peptide and designed such that proteolytic cleavage anywhere within the peptide backbone should produce a fluorescence increase. To experimentally assess the capacity of PEP3 to be cleaved by active SERA6, we added PEP3 to assays similar to those described in Figure 3D, containing reconstituted FL SERA6 and FL MSA180 affinity-purified from parasite saponin extracts. Increases in fluorescence corresponding to cleavage of PEP3 occurred following addition of rSUB1 (Figure 4D). Interestingly, the fluorescence increase exhibited a kinetic profile that encompassed a time delay or lag phase that approximately correlated with the period before the detectable appearance of SERA6 PAP, consistent with the notion that conversion to PAP was required for the initiation of PEP3 cleavage. Analysis of the assay supernatants by reversed-phase HPLC (RP-HPLC) and electrospray mass spectrometry to detect the fluorescent (EDANS-containing) product(s) of PEP3 cleavage revealed that PEP3 was cleaved in these assays primarily or exclusively at the Lys-Ser bond corresponding to the predicted autocatalytic cleavage site within SERA6 at the IKLK-SPYF bond (hyphen indicating the scissile bond) (Figure 4A; Supplementary Figure 3). Importantly, no PEP3 cleavage was detected in control incubations containing rSUB1 only, in accord with the well-established substrate specificity of SUB1 (Withers-Martinez et al., 2012, 2014) which is not predicted to be capable of cleaving PEP3 anywhere within its sequence. It was concluded that maturation of SERA6 to PAP resulted in the generation of a proteolytic activity capable of cleaving PEP3 at a preferred internal site corresponding precisely to the predicted autocatalytic site within SERA6.

Development of improved drug-like inhibitors of SERA6

We have previously described (Tan et al., 2021) two structurally-related nitrile compounds that, similar to the epoxide E64 and its more membrane-permeable analogue E64-d, inhibit both the final autocatalytic step of SERA6 maturation and SERA6-mediated cleavage of β-spectrin in P. falciparum schizonts undergoing egress. Both these nitrile compounds, a purine called MMV676881 and the triazine compound 31, also block parasite egress with a phenotype typical of inhibition or genetic ablation of SERA6 (Tan et al., 2021; Dans et al., 2020). This is characterized by normal rounding and rupture of the PVM with the characteristic increased mobility of the intraerythrocytic merozoites, concomitant with poration of the RBC membrane but failure of the RBC membrane to break (Thomas et al., 2018). The two nitrile compounds were similarly potent (EC 50 ~0.4 μM for inhibition of SERA6 maturation in intraerythrocytic parasites) but were predicted to be poorly soluble. To seek analogues with improved potency and/or solubility, we screened a small set of 30 related nitrile compounds (a kind gift of Min Shen and Matthew Hall, National Center for Advancing Translational Sciences [NCATS], NIH, USA) for their capacity to inhibit egress, using a previously-described assay based on spectrometric detection of hemoglobin released from live schizonts upon egress (Tan et al., 2021). This screen identified two further active (defined as an egress-inhibitory EC 50 <2 μM) purine nitrile compounds called compounds 33 and 37 (Figure 5A). Further examination showed that the newly identified hits were similarly potent as the previously-identified MMV676881 and compound 31 (which were also originally derived from the same NCATS set) in inhibition of SERA6 p65-to-PAP conversion in intact SERA6-mTAP:loxP schizonts (Figure 5B), confirming the new compounds as inhibitors of SERA6 maturation.

Notably, all four active compounds shared a 3,5-difluorophenyl moiety. Comparison with the inactive nitriles in the NCATS screening set suggested that replacement or modification of this 3,5-difluorophenyl group results in loss of inhibitory activity (Supplementary Figure 4). We also noted that cyclisation of the triazine compound 31 in two different vectors could rationalize the different architecture of the three purine nitrile actives (Supplementary Figure 4). All the NCATS nitrile compounds were originally described in an earlier study by Mott and colleagues (Mott et al., 2010) in which MMV676881 (referred in that work as compound 32) was co-crystallized with the Trypanosoma cruzi cysteine protease cruzain. This x-ray crystal structure showed that the cyano group of MMV676881 formed the expected covalent adduct with the nucleophilic active-site Cys of cruzain and the 3,5-difluorophenyl substituent extended into the largely hydrophobic specificity-determining S2 pocket of cruzain, whilst the purine core of MMV676881 orientated the vectors of these two key interacting groups (Figure 5C; Supplementary Figure 5). This binding pose is remarkably similar to that adopted by a number of similar purine nitriles bound to the active site of the papain-like cysteine protease cathepsin K (Altmann et al., 2004). To examine the possibility that MMV676881 and the other actives might interact in a similar manner with SERA6, we performed in silico covalent docking of MMV676881 into the ICM-Pro molecular model of the SERA6 papain-like catalytic domain, comparing this model with the cruzain-MMV676881 and cathepsin K-purine nitrile structures (Figure 5D). This modelling showed that indeed MMV676881 and the three other active compounds described here might be predicted to interact with SERA6 in a very similar manner to the MMV676881-cruzain and the cathepsin K-nitrile compound interactions.

Based on our structural predictions of the MMV676881-SERA6 interaction and pharmacophore analysis of the active compounds, we designed an additional series of eight nitrile analogues (all named with the prefix CS) that we predicted to be more soluble and drug-like than the existing four hits, whilst retaining the structural characteristics associated with SERA6 inhibition (see Supplementary Figures 6, 7 for their structures and predicted chemical properties). The new CS compounds were experimentally assessed for their capacity to prevent SERA6-mediated β-spectrin cleavage in treated parasites, to inhibit SERA6 maturation in the saponin extract assay developed above, and to prevent parasite replication in in vitro culture. As shown in Figures 6A–C, five of the CS compounds showed inhibitory activity in all three assays, with CS01–03 displaying the highest potency in the quantitative cell-based replication assay (which measures the transition from mature schizonts to newly invaded ‘ring’ stage parasites, a surrogate measure of successful egress of invasive merozoites). Notably, in this last assay CS01–03 was ~4.4-fold more potent than the original hit compound MMV676881. In silico covalent docking analysis of the active CS compounds was in agreement with the experimental findings, predicting that the active CS compounds have the capacity to form additional main-chain H-bond interactions with the SERA6 S2 pocket compared to MMV676881 (Figure 6D). In keeping with its relative potency in the other assays, compound CS01–03 consistently gave the best docking scores in this in silico analysis (Supplementary Figure 8).

In view of these promising results, we next focused on CS01-03. This compound possesses a chiral carbon atom and further docking studies suggested that the R -enantiomer likely binds to SERA6 better than the S -enantiomer (Supplementary Figure 9). To examine this experimentally, the CS01–03 racemate was subjected to chiral fractionation, resulting in the isolation of the two enantiomers CRK1_1 and CRK1_2. Of these, CRK1_1 was the more potent in the schizont-to-ring transition assay, with an estimated EC 50 ~100 nM (Supplementary Figure 10). In vitro drug metabolism and pharmacokinetic (DMPK) analysis of CRK1_1 and CRK1_2 indicated good solubility and stability for both enantiomers (Supplementary Figure 11). To determine whether the antimalarial action of CRK1_1 was through selective inhibition of SERA6 maturation as anticipated, CRK1_1 was evaluated in a range of further assays to establish its selectivity and mode of action. In the first of these assays, synchronous cultures of ring-stage parasites treated with CRK1_1 (500 nM, ~5-fold EC 50) were monitored by microscopic examination at intervals over the course of an entire ~48 h erythrocytic cycle. As shown in Figure 7A, some mild swelling of the digestive vacuole was apparent in treated parasites at the early trophozoite (31 h post-invasion) developmental stage, possibly due to inhibition of the non-essential papain-like cysteine protease hemoglobinase falcipain 2 (Sijwali et al., 2006). Crucially, however, CRK1_1 treatment did not block parasite development until the mature schizont stage where complete arrest of egress was observed. This indicated a good degree of selectivity for CRK1_1. In vitro enzyme assays supported this; whilst at high concentrations (10 μM) CRK1_1 showed some activity against mammalian papain, the compound showed no inhibition of rSUB1 or recombinant PMX (Supplementary Figure 12). This confirmed that the effects of CRK1_1 on egress were not due to inhibition of these crucial parasite proteolytic enzymes.

CRK1_1 was finally evaluated in the newly developed in vitro reconstitution assay and SERA6 peptidase activity assay described above. As shown in Figure 7B, SUB1-dependent maturation of SERA6 in the presence of CRK1_1 was arrested at the p65 intermediate form, as expected, indicating direct inhibition of the autocatalytic p65-to-PAP step of maturation. Furthermore, CRK1_1 prevented cleavage of the fluorogenic substrate PEP3 in the SERA6 activity assay (Figure 7C; Supplementary Figure 13), showing a direct on-target effect of CRK1_1 on the proteolytic activity of SERA6 PAP. It was concluded that CRK1_1 kills the parasite primarily through prevention of egress as a result of inhibition of SERA6 maturation and activity. Importantly, CRK1_1 displays improved potency, physicochemical properties and drug-likeness compared to the original hit MMV676881. CRK1_1 represents the most potent and selective drug-like inhibitor of P. falciparum SERA6 activity described to date.

Egress is a crucial step in the lifecycle of the blood-stage malaria parasite, being essential for parasite replication and directly causing much of the pathophysiology of the periodic fevers associated with the disease due to the release into the circulation of large amounts of parasite antigen upon schizont rupture. Whilst no antimalarial drugs currently in clinical use specifically target egress, it has become increasingly clear that egress involves several parasite enzymes that have high potential as targets for new types of chemotherapeutic approaches (for comprehensive recent reviews see Tan and Blackman, 2021; Dvorin and Goldberg, 2022). The present study significantly advances our understanding of the maturation of SERA6 and its role in malaria parasite egress. Using unfractionated saponin extracts derived from mature schizonts arrested with an inhibitor of SUB1 discharge (the PKG inhibitor C2), we first showed that simple supplementation of the extracts with rSUB1 was sufficient to obtain complete maturation of the endogenous FL SERA6 to its terminal PAP form. The kinetics of SERA6 maturation in these in vitro experiments were relatively slow (requiring several hours compared to the very rapid process in the parasite which takes just minutes; Tan et al., 2021). This clearly represents a distinct difference between the in vivo and ex vivo maturation pathways. We do not know the reason for this difference but we rationalize that it may be a result of suboptimal conditions of salt, pH or reactant concentrations in our assays. Further refinement of these experiments by affinity purification of FL SERA and FL MSA180 and reconstitution of the isolated proteins in the presence of rSUB1, combined with the development of a fluorogenic substrate based on a putative internal SERA6 cleavage site, allowed us to show that SUB1 and MSA180 are necessary for maturation of SERA6 to an active protease, or more accurately, a peptidase. SUB1 and MSA180 may even be sufficient for SERA6 maturation, but is important to note that our experimental approaches cannot formally rule out the possibility that other essential parasite components were co-purified with either FL SERA6 or FL MSA180 in our affinity purification procedures. We consider this unlikely given the lack of any evidence that the precursor proteins form heterocomplexes in the parasite PV. Future work should focus on determining whether in vitro activated SERA6 has the capacity to cleave known endogenous substrates, in particular β-spectrin.

Our study reinforces the essentiality of MSA180 for SERA6 maturation, but we have not determined the molecular function of MSA180, which remains unknown. We have previously speculated (Tan et al., 2021) that MSA180 may provide a scaffolding or chaperone-like role during SERA6 maturation and/or the presumed RBC cytoskeletal interaction required for β-spectrin cleavage. Intriguingly, our bioinformatic analysis using AlphaFold3 has shown that the p45 region of P. falciparum MSA180 (one of the two fragments of MSA180 that bind to SERA6 PAP) contains a solvent-exposed UBL domain that is predicted to be conserved across MSA180 orthologues in other Plasmodium species. UBL domains are often involved in protein-protein interactions (Grabbe and Dikic, 2009; Winget and Mayor, 2010). The MSA180 UBL domain is predicted to be less solvent-exposed within the fold of the FL MSA180 prior to its cleavage to generate the p40 and p45 domains, so it is tempting to speculate that MSA180 processing may potentially function to expose the UBL element to enable it to partake in binding interactions with SERA6 PAP or indeed with β-spectrin or other associated components of the RBC cytoskeleton. Remarkably, the AlphaFold3 predictions also support a stable pre-existing non-covalent interaction between the p45 and p40 segments within the intact FL MSA180, consistent with our evidence that both fragments interact with SERA6 PAP (Tan et al., 2021). However, these structural predictions remain entirely speculative, and further experimental work will be required to establish the precise role(s) of MSA180 and its putative UBL domain in SERA6 maturation and activity.

Our demonstration of the appearance of peptidase activity associated with maturation of FL SERA6 or p65 to PAP constitutes the first experimental evidence to our knowledge that SERA6 PAP (but not the precursor FL SERA6) is an active proteolytic enzyme. Due to the technical challenges presented by the SERA6 activity assays used in this study, only two independent activity assays were performed, demanding caution in interpretation of our findings. However, our results are consistent with the previously established temporal association between maturation of SERA6 and cleavage of host RBC β-spectrin at egress. Our use of saponin extracts as a source of materials for analyzing SERA6 maturation provides a simple but significant technical advance. Previously, screening for inhibitors of maturation has required work with continuous cultures of live P. falciparum asexual blood-stage parasites, with all the associated need for high containment culture facilities, laborious parasite manipulation approaches and ready access to human blood. Because saponin extracts can be generated in bulk and frozen in aliquots, it is now feasible to disseminate extracts to multiple laboratories if required to enable their use for multiple downstream assays under non-containment conditions. Our development of a simple, cell-free assay that measures SERA6 PAP activity using a synthetic fluorogenic peptide substrate will greatly accelerate the development and optimization of new selective small molecule inhibitors of SERA6. Our development of improved small molecule inhibitors of SERA6 activity and egress in this study validates the SERA6 activity assay. It is worth noting that although PEP3 clearly functions as a SERA6 substrate, the PEP3 sequence was based on a predicted internal cleavage site and we have not experimentally determined the site at which autocatalytic SERA6 cleavage takes place. This must await N-terminal amino acid sequencing of mature SERA6 PAP, a challenging aim given the low abundance of SERA6 in the parasite.

CRK1_1, the more active enantiomer of CS01-03, is the most potent and selective small molecule inhibitor of SERA6 maturation identified to date. It is generally difficult to definitively assign any drug-induced phenotype to engagement of a specific molecular target, and so we cannot unambiguously determine the mode of action of CRK1_1. This is particularly relevant here as our selectivity profiling of CRK1_1 was rather limited, and broader profiling of the compound against parasite and host cysteine proteases will be needed in future studies. However, parasites treated with CRK1_1-treated underwent complete arrest at the mature schizont stage, phenocopying the SERA6-null phenotype and so consistent with the lethal impact of CRK1_1 being primarily through inhibition of SERA6 activity. CRK1_1 is not a drug candidate, but orally available carbonitrile compounds have been used in the clinic for some time (e.g. as inhibitors of dipeptidyl peptidase-IV and the SARS-CoV-2 main protease; Citarella et al., 2023; Aloke et al., 2025) so further development of CRK1_1 analogues as a drug lead or experimental tool is warranted. In addition to the lethal effect of CRK1_1 on egress, our detailed phenotypic analysis of parasites treated for an entire erythrocytic lifecycle with CRK1_1 revealed swelling of the parasite food vacuole, typical of compounds that inhibit the hemoglobin digestion pathway. This was not unexpected, since hemoglobin catabolism involves the activity of parasite falcipains (Rosenthal, 2020), a class of four P. falciparum papain-like cysteine proteases, one or more of which is likely susceptible to inhibition by CRK1_1 due to their structural relatedness to SERA6. Our observation that this cross-inhibition was not sufficiently deleterious to arrest parasite growth was unsurprising, as extensive work by others has concluded that proteases within the hemoglobin digestion pathway are highly redundant, to the extent that several are no longer considered good targets for antimalarial drug development (Rosenthal, 2020; Nasamu et al., 2020). On the other hand, at least one falcipain appears to play crucial role(s) during the asexual blood-stage life cycle (Rosenthal, 2020). Compounds that kill the parasite through the inhibition of multiple targets are particularly attractive for drug development as they are generally expected to be less likely to select for parasite resistance than drugs that target a single pathway. To this end it may be possible to identify CRK1_1 analogues that potently inhibit SERA6 maturation as well as important falcipain functions, placing additional pressure on parasite survival and reducing opportunities for parasite resistance.