Recruitment of Human C1 Esterase Inhibitor Controls Complement Activation on Blood Stage Plasmodium Falciparum Merozoites

Home , Factor H, MASP2 (protein)

Recruitment of Human C1 Esterase Inhibitor Controls Complement Activation on Blood Stage Plasmodium falciparum Merozoites

This information is current as Alexander T. Kennedy, Lakshmi C. Wijeyewickrema, Alisee of September 29, 2021. Huglo, Clara Lin, Robert Pike, Alan F. Cowman and Wai-Hong Tham J Immunol 2017; 198:4728-4737; Prepublished online 8 May 2017;

doi: 10.4049/jimmunol.1700067 Downloaded from http://www.jimmunol.org/content/198/12/4728

Supplementary http://www.jimmunol.org/content/suppl/2017/05/06/jimmunol.170006 Material 7.DCSupplemental http://www.jimmunol.org/ References This article cites 59 articles, 22 of which you can access for free at: http://www.jimmunol.org/content/198/12/4728.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on September 29, 2021

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2017 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Recruitment of Human C1 Esterase Inhibitor Controls Complement Activation on Blood Stage Plasmodium falciparum Merozoites

Alexander T. Kennedy,*,† Lakshmi C. Wijeyewickrema,‡ Alisee Huglo,*,† Clara Lin,*,† Robert Pike,‡ Alan F. Cowman,*,† and Wai-Hong Tham*,†

The complement system is a front-line defense system that opsonizes and lyses invading pathogens. To survive, microbes exposed to serum must evade the complement response. To achieve this, many pathogens recruit soluble human complement regulators to their surfaces and hijack their regulatory function for protection from complement activation. C1 esterase inhibitor (C1-INH) is a soluble regulator of complement activation that negatively regulates the classical and lectin pathways of complement to protect human

tissue from aberrant activation. In this article, we show that Plasmodium falciparum merozoites, the invasive form of blood stage Downloaded from malaria parasites, actively recruit C1-INH to their surfaces when exposed to human serum. We identiﬁed PfMSP3.1, a member of the merozoite surface protein 3 family of merozoite surface proteins, as the direct interaction partner. When bound to the merozoite surface, C1-INH retains its ability to complex with and inhibit C1s, MASP1, and MASP2, the activating proteases of the complement cascade. P. falciparum merozoites that lack PfMSP3.1 showed a marked reduction in C1-INH recruitment and increased C3b deposition on their surfaces. However, these DPfMSP3.1 merozoites exhibit enhanced invasion of RBCs in the presence of active complement. This study characterizes an immune-evasion strategy used by malaria parasites and highlights the http://www.jimmunol.org/ complex relationship between merozoites and the complement system. The Journal of Immunology, 2017, 198: 4728–4737.

n estimated 214 million cases of malaria occur annually, initial contact of the merozoite with the RBC (4). Prior to invasion resulting in 438,000 deaths worldwide (1). Plasmodium and being safely enclosed in the RBC, merozoites are directly ex- A falciparum, the etiological agent of the most lethal posed to the human immune system, including complement (2). malaria infections, establishes an asexual blood stage cycle that is Complement is a key frontline immune defense system against responsible for the clinical symptoms of malaria. The blood stage invading pathogens. The complement system is composed of a cycle commences when a parasite first enters a RBC, where it then cascade of sequentially activated zymogens that lead to opsono- grows as a trophozoite before dividing via schizogeny to produce phagocytosis or membrane lysis of the foreign particle. Complement by guest on September 29, 2021 16–32 merozoites, which are the invasive form for this stage of the can be activated via the classical pathway, the lectin pathway, or the life cycle. Merozoites egress from the infected RBC and find a alternative pathway (5). The classical pathway is triggered by Ag–Ab healthy RBC to invade to continue the next cycle of growth and complex recognition by C1q, activating proteases C1r and C1s (6). expansion. Upon egress, merozoites remain invasion competent Although the lectin pathway is activated by mannose binding lectin, for 10–30 min (2). The invasion process involves interactions ficolins, and collectins that recognize carbohydrate motifs on the between parasite invasion ligands and RBC surface proteins that pathogen causing activation of the MASP1 and MASP2 proteases initiate a cascade of molecular events that ultimately results in (7), the alternative pathway is activated by domain flexibility, penetration into RBCs (3). This process normally takes 2 min from causing low levels of constitutive zymogen activation, and it attacks any surface lacking negative regulators of complement activation (8). All three pathways induce C3 convertase formation. These *Infection and Immunity Division, The Walter and Eliza Hall Institute, Parkville, Victoria 3052, Australia; †Department of Medical Biology, University of Melbourne, convertases cleave C3 to C3b, exposing a concealed thioester Parkville, Victoria 3052, Australia; and ‡Department of Biochemistry and Genetics, domain that subsequently links covalently to hydroxyl and amine La Trobe Institute for Molecular Science, Bundoora, Victoria 3083, Australia groups on pathogen surfaces (9, 10). C3b is a target for opsono- ORCIDs: 0000-0001-6084-4887 (L.W.); 0000-0002-2832-1549 (A.H.); 0000-0002- phagocytosis and may also associate with factor B to produce the 2083-0269 (R.P.); 0000-0001-5145-9004 (A.F.C.); 0000-0001-7950-8699 (W.-H.T.). C3 convertase of the alternative pathway, leading to a positive Received for publication January 12, 2017. Accepted for publication April 10, 2017. feedback loop. As C3b density increases, it leads to creation of C5 This work was supported by the Victorian State Government Operational Infrastruc- convertases that ultimately drive formation of the membrane at- ture Support and the National Health and Medical Research Council. W.-H.T. holds a Future Fellowship from the Australia Research Council. A.T.K. was supported by an tack complex on pathogen surfaces causing direct lysis (6). Australian Postgraduate Award from the Australian Government. Complement regulatory proteins prevent aberrant activation of Address correspondence and reprint requests to Dr. Wai-Hong Tham, The Walter and complement on self-surfaces. These regulatory proteins can be di- Eliza Hall Institute of Medical Research, 1G Royal Parade, Parkville, VIC 3052, vided into membrane zone and fluid-phase regulators. The membrane Australia. E-mail address: [email protected] zone regulators, which are expressed by cells on their plasma The online version of this article contains supplemental material. membrane, include complement receptor (CR)1, CR2, CR3, and Abbreviations used in this article: AMA1, apical membrane Ag 1; C1-INH, C1 esterase inhibitor; CR, complement receptor; FH, factor H; FHL-1, FH-like 1; CR4, membrane cofactor protein, decay accelerating factor, and MFI, mean fluorescence intensity; MSP, merozoite surface protein; NHS, normal CD59 (11). The fluid-phase regulators, which circulate in plasma human serum; NHS-E, NHS inactivated by addition of EDTA; NHS-HI, heat- and are recruited onto self-cell surfaces to enhance their protection, inactivated NHS; PfGPI, P. falciparum GPI; zC1s, zymogen form of C1s. include factor H (FH), FH-like 1 (FHL-1), FH related protein 1, C4b Copyright Ó 2017 by The American Association of Immunologists, Inc. 0022-1767/17/$30.00 binding protein, vitronectin, and C1 esterase inhibitor (C1-INH) www.jimmunol.org/cgi/doi/10.4049/jimmunol.1700067 The Journal of Immunology 4729

(11). Complement regulatory proteins can control complement by through an interaction with P. falciparum GPI (PfGPI) (35). Supra- facilitating cleavage of C3b to less active iC3b that cannot partici- physiological levels of C1-INH restricted the ability of merozoites to pate in the alternative pathway convertase, accelerating the decay of invade; however, no challenge with complement was undertaken to convertases, directly inhibiting formation of the membrane attack test whether this recruitment provided protection from complement complex, or by inhibiting protease function (12). (35). In the mosquito stages, the parasite protein GAP50 recruits FH The fluid-phase regulator C1-INH controls the lectin and clas- and FHL-1 to the surface of gametocytes (the sexual forms of the sical pathways by inactivating the initiating proteases C1r, C1s, malaria parasite) to provide protection from active human comple- MASP1, and MASP2 (13–17). C1-INH has an N-terminal domain ment in the mosquito blood meal (36). Disruption of FH and FHL-1 with O- and N-linked glycans attached and a C-terminal serpin recruitment blocked gametocyte progression to downstream stages of domain (18). Protease inhibition occurs by proteases recognizing the life cycle in a complement-dependent manner, indicating that and cleaving the reactive center loop within the C1-INH serpin complement-evasion strategies are critical for parasite survival domain; this causes molecular rearrangements that trap the pro- throughout the life cycle (36). tease in a covalent linkage between C1-INH and the active site, In this article, we show that P. falciparum merozoites recruit inactivating the protease and inhibitor in the process (19). Rapid C1-INH from human serum through an interaction with one of the inhibition of these proteases by C1-INH limits downstream zy- MSPs. By disrupting this interaction, we demonstrate the complex mogen cleavage, thus preventing complement activation. The nature of the relationship among merozoites, the complement serpin domain also regulates the contact system of coagulation, system, and invasion. the fibrinolytic system, and the kallikrein–kinin system by binding and inhibiting the serine proteases FXIIa, plasma kallikrein, Materials and Methods Downloaded from plasmin, and tissue plasminogen activator (19). Parasite culturing Many pathogens are able to thrive in serum as a result of the evolution of complement-evasion strategies. One strategy involves Parasites were maintained by serial passage in RPMI 1640–HEPES containing 0.5% AlbuMAX II (Life Technologies) or 10% human O+ serum pathogens recruiting human complement regulators onto their (Australian Red Cross Blood Bank) with 2% hematocrit of O+ blood own surface for protection. Pathogens commonly use their surface (Australian Red Cross Blood Bank). We used the D10-PHG strain (37) for proteins to bind to the fluid-phase regulators FH, FHL-1, C4b the majority of our initial assays. We also used the 3D7 and FCR3 strains, http://www.jimmunol.org/ binding protein, and C1-INH (20). Neisseria meningitidis achieves as well as transgenic knockouts of PfMSP3.1 (DPfMSP3.1) and PfMSP3.4 D this through its FH binding protein that recruits FH and FHL-1 by ( PfMSP3.4), which were both made on a 3D7 background (29, 30). mimicry of self-carbohydrate motifs, whereas the flaviviruses Merozoite purification dengue virus, West Nile virus, and yellow fever virus use their Merozoites were obtained as previously described (2). Briefly, strains were nonstructural protein 1 to bind C4b binding protein (21–23). In tightly synchronized, and 36–40-h schizonts were magnet enriched addition, Bordetella pertussis, the causative agent of whooping (MACS, Miltenyi Biotec). Parasites were returned to culture in the pres- cough, recruits C1-INH with its surface protein Vag8 (24). Dis- ence of 10 mM E-64. After approximately 6 h of development, para- ruption of complement regulator recruitment can increase com- sitophorous vacuole–enclosed merozoites were washed once in PBS or

m by guest on September 29, 2021 plement deposition, leading to increased phagocytosis or lysis that RPMI 1640–HEPES prior to merozoite release via passage over a 1.2- m filter (Sartorius). Flow cytometry assays required hemozoin and digestive ultimately results in pathogen destruction; alternatively, a lack of vacuole removal, which was achieved by flowing the merozoite filtrate regulator recruitment may also result in an organism being avir- through a magnetic column twice (MACS, Miltenyi Biotec). ulent (20, 23–26). Hence, complement regulator–recruiting proteins are important virulence factors. Complement deposition of merozoites P. falciparum merozoites are covered in a densely packed pro- Purified merozoites were spun at 8000 3 g for 3 min, and supernatant was teinaceous surface coat made of merozoite surface proteins (MSPs) removed. Merozoites were resuspended in relevant buffer, spun again, and that are anchored to the merozoite surface by GPI or are periph- incubated in human serum treatments or purified complement components at 37˚C for between 1 and 10 min. If required, merozoites were opsonized erally associated with other MSPs (27). MSPs have been catego- with Abs for 10 min at 37˚C and washed once prior to incubation in human rized into multiple families that include the MSP1, MSP2, six-cysteine, serum or purified components. Complement deposition with human serum MSP7, serine repeat Ag, and MSP3 family. The MSP3 family is was stopped by the addition of protease inhibitor mixture with EDTA composed of eight members (PfMSP3.1 to PfMSP3.8) that as- (Roche) and incubation on ice, whereas all other depositions were stopped with incubation on ice alone. The suspension was centrifuged and washed sociate peripherally with the merozoite surface through protein– three times with PBS containing 1% BSA. These complement-deposited protein interactions (28, 29). All eight MSP3 members have a merozoite samples were used for Western blot analyses, immunofluores- conserved N-terminal NLRNA/G motif followed by a polymor- cence assay, and flow cytometry, as described below. phic N terminus that may contain heptad repeats, a proline-rich Western blotting analyses domain, or a Duffy binding-like domain, whereas six of the eight membersshareaconservedCterminus that contains a glutamic Complement-deposited merozoites were resuspended in reducing or non- acid–rich region followed by a leucine-like zipper region (28). reducing protein sample buffer and fractionated by SDS-PAGE before transfer onto nitrocellulose membranes (GE Healthcare Life Sciences). Members of the MSP3 family have been implicated in RBC After blocking in 0.1% v/v PBS–Tween 20 containing 10% g/v skim milk, binding, free heme binding, and recruitment of IgM to shield membranes were incubated with Abs against C1-INH, C3 (both from proteins from the IgG response (29–32). Complement Technologies), PfMSP3.1, Pf92, or apical membrane Ag 1 Recent studies have highlighted how P. falciparum cells protect (AMA1) (Walter and Eliza Hall Institute Antibody Facility) Abs at 1/1000 themselves from complement-mediated assault. In the blood stages, dilution in 0.1% v/v PBS–Tween 20 containing 1% g/v skim milk, washed with 0.1% v/v PBS–Tween 20, and incubated with 1/1000 dilution of schizonts and merozoites recruit the human complement regulators secondary Abs conjugated to HRP in 0.1% v/v PBS–Tween 20 containing FH and FHL-1 to their surface (33, 34). For merozoites, an MSP 1% skim milk. Development was with ECL (Amersham), as per the from the six-cysteine family, Pf92, binds to FH and FHL-1 (34). manufacturer’s instructions, and visualization was on film (FUJI) or Deletion of the Pf92 gene led to significantly increased complement- via a ChemiDoc One-Touch using Image Lab Touch v1.2.0.12 (Bio-Rad). Membranes were stripped, as required, using 25 mM Tris-glycine (pH 2) mediated destruction of merozoites, highlighting how important containing 1% SDS. For immunoprecipitation assays, replicate blots these complement-evasion strategies are for parasite survival (34). loaded with one eighth of the sample were used to detect immunopre- Merozoites have also been reported to bind C1-INH to their surface cipitation of the target Ag. 4730 C1-INH REGULATES MEROZOITE COMPLEMENT DEPOSITION

Immunofluorescence assay Complex formation assay Immunofluorescence assays were performed as previously described (38). For in-solution assays, 0.5 mM purified C1-INH (CSL) was incubated with Briefly, merozoites were dried onto coverslips prior to incubation with 1 mM recombinant C1s, zC1s, MASP1, or MASP2 proteases in GPBS++ 1/500 anti–C1-INH (Complement Technologies) and anti-AMA1 (Walter buffer for 10 and 30 min at 37˚C. For merozoite assays, isolated mero- and Eliza Hall Institute Antibody Facility) primary Abs. After three washes zoites were incubated with a 0.5 mM solution of purified C1-INH (CSL) in in PBS, coverslips were probed with 1/500 dilution of secondary Abs GPBS++ buffer prior to incubation with 1 mM solutions of recombinant (Alexa Fluor 555 and Alexa Fluor 647). After washing, coverslips (Zeiss) C1s, zC1s, MASP1, or MASP2 in GPBS++ for 10 and 30 min at 37˚C. were mounted on glass slides in VECTASHIELD (Vector Labs) containing Reactions were stopped on ice and washed three times with PBS con- 0.1 ng/ml DAPI (Sigma) and imaged using a DeltaVision Elite widefield taining 1% w/v BSA. In-solution assays and deposited merozoites were fluorescence microscope (DeltaVision Technologies) with a 1003 Olym- resuspended in reducing protein sample buffer, and proteins were separated pus oil objective (NA 1.40) on a CoolSNAP HQ2 camera (Photometrics) at by SDS-PAGE on 3–8% Tris-Acetate gels (Life Technologies), followed ambient temperature. Data were collected and underwent enhanced ratio by Western blotting. deconvolution using Softworx (v6.1; DeltaVision Technologies). Maxi- mum intensity projection, pseudocoloring, and image merging were per- Invasion assay formed with Fiji (v2.0; ImageJ). Isolated merozoites were added to 0.5% hematocrit O+ blood suspended in Flow cytometry of C1-INH binding 25% normal human serum (NHS), 25% heat-inactivated NHS (NHS-HI), or media in a final volume of 50 ml. Ninety-six–well plates were then Merozoites were purified as described above with hemozoin removal. incubated on a plate shaker (500 rpm; 37˚C) for 10 min and for an addi- Complement-deposited merozoites were incubated with 1/1000 dilution of tional 20 min at 37˚C in a gassed culture box. Plates were washed three anti–C1-INH Abs (Complement Technologies) for 30 min at room tem- times and resuspended with 50 ml of media. After 24 h, parasites were perature in the dark to prevent quenching of GFP. After washing in 1% fixed with 0.25% glutaraldehyde (Sigma) and stained with 103 SYBR BSA, they were incubated for 30 min at room temperature with 1/2000 Green (Life Technologies). Parasitemia was determined by collecting Downloaded from ethidium bromide (Bio-Rad) and 1/1000 dilution of anti-goat conjugated to .50,000 events with a FACSCalibur flow cytometer (BD) using CellQuest Alexa Fluor 633. Samples were washed, and .50,000 events were col- Pro v5.2.1. Analyses were performed with FlowJo v8.8.7 (TreeStar). lected using a FACSCalibur (BD) with CellQuest Pro v5.2.1. Analyses were performed with FlowJo v8.8.7 (TreeStar) to determine the mean Statistical analyses fluorescence intensity (MFI) of 633+ merozoites. Background subtraction was performed using the MFI of merozoites stained with secondary Abs Statistical analyses were conducted using Prism for Mac version 6.0f (GraphPad). For our flow cytometry analyses of C1-INH binding, we and ethidium bromide only. http://www.jimmunol.org/ performed one-way ANOVA with multiple comparisons from three inde- Immunoprecipitation assays pendent experiments to test whether a significant difference in anti–C1-INH binding was present (Fig. 1C). For our initial invasion assay (Fig. 6C, 6D), Saponin-treated parasitophorous vacuole–enclosed merozoites were solu- we used a paired repeated-measures two-way ANOVA to account for bilized at 4˚C overnight using TNET buffer (1% v/v Triton X-100, randomized block effects from five independent experiments. As expected, 150 mM NaCl, 10 mM EDTA, 50 mM Tris [pH 7.4]), and the resultant we observed a significant interaction effect (p = 0.0002 and 0.0033 for supernatant was precleared overnight at 4˚C with protein G Sepharose Fig. 6C and 6D, respectively), so we applied the Sidak multiple- beads. Purified C1-INH, recombinant serpin domain, or deglycosylated comparison test to examine whether the difference between NHS and C1-INH was added to a final concentration of 5 mg/ml or an equivalent NHS-HI invasion was significant within each individual additive condition. volume of PBS. Anti–C1-INH, anti-Pf92, or anti-MSP3 Abs were cross- To compare wild-type PfMSP3.1 with DPfMSP3.1 (Fig. 6E, 6F), we used a

linked to protein G Sepharose beads, as per the manufacturer’s instructions two-way ANOVA on data from at least three independent experiments to by guest on September 29, 2021 (Thermo Scientific), and were used in immunoprecipitation mixtures prior determine whether the difference in relative invasion was significant. We to washing with TNET and elution. For Western blotting, beads were did not observe any interaction effects. eluted with glycine (Thermo Scientific); however, for liquid chromatography–tandem mass spectrometry protein identification applications, we eluted the beads with 125 mM TCEP, 25% v/v TFE in 0.025% formic acid Results at 40˚C with shaking for 10 min. Mass spectrometry samples were neu- Merozoites recruit human complement regulator C1-INH to m tralized by addition of TEAB and then digested with 5 g/ml trypsin their surfaces overnight prior to liquid chromatography–tandem mass spectrometry. Abs and lectins target merozoites upon their release from infected Deglycosylation of C1-INH RBCs, activating the classical and lectin pathways of complement. Deglycosylation was performed by incubating 100 mg/ml C1-INH at 37˚C We investigated whether C1-INH, a major regulator of these overnight with PBS or with 10,000 U/ml PNGase F, 800,000 U/ml pathways, is recruited to the merozoite surface as a mechanism of O-glycosidase, 1000 U/ml neuraminidase (all from NEB), or a combina- complement evasion. Purified merozoites were exposed to NHS tion of these glycosidases. SDS-PAGE of deglycosylated proteins was conducted on 4–12% Bis-Tris gels and visualized with Simply Blue stain and washed to remove nonspecific binders prior to analysis of (Life Technologies) or via Western blotting, as described above. tightly bound proteins in nonreducing protein sample buffer. C1-INH has a predicted molecular mass of 75 kDa; however, it Recombinant complement protease expression migrates atypically on many gel systems, resulting in an apparent Recombinant human C1s (G282-D688), MASP1(G289-N699), and MASP2 molecular mass of between 90 and 115 kDa. In our assays, we (G288-F686) were expressed in Escherichia coli as insoluble proteins in the observed that C1-INH present in NHS or purified from serum inclusion bodies, as described previously (39). The enzymes were dena- migrates at 100 kDa in nonreducing conditions (Fig. 1A, left tured and refolded, followed by purification using a combination of anion- exchange and gel-filtration chromatography, all as previously described panel). Furthermore, we found that free merozoites recruited (39). All proteins were obtained in good yield and purity. The zymogen C1-INH in NHS and in NHS inactivated by addition of EDTA form of C1s (zC1s) was activated using immobilized C1r, as previously (NHS-E) (Fig. 1A, right panel). We observed a marked reduction described (39). MASP1 and MASP2 were autoactivated during anion- in C1-INH recruitment when we used NHS-HI (Fig. 1A). This is exchange chromatography. The activated proteases were titrated using likely due to C1-INH polymerization occurring at elevated tem- human plasma–purified C1 inhibitor (CSL, Melbourne, Australia) to yield the final active concentrations of each enzyme form. peratures (41). These results show that C1-INH is recruited to Recombinant human C1-INH (T111-A500) was expressed in SG130009 merozoites, regardless of whether complement activation occurs. E. coli cells. Following sonication and centrifugation (18,000 rpm for 45 To further examine whether C1-INH was recruited to the mer- min), the supernatant was filtered and loaded onto a 5-ml HisTrap column. ozoite surface, an immunofluorescence assay was used. The C1-INH was eluted with 250 mM imidazole, 25 mM sodium phosphate (pH 8). The eluant was checked for inhibitory activity against C1s, pooled, merozoite surface was demarcated by staining with anti-AMA1 and gel filtered. Inhibitory-activity assays, including stoichiometry of in- Abs, whereas C1-INH was detected using anti–C1-INH Abs. In hibition and kass, were performed as previously described (40). NHS, NHS-HI, and NHS-E conditions, we observed colocaliza- The Journal of Immunology 4731

to the merozoite surface, as opposed to being present on residual digestive vacuoles or RBC membranes. To observe the recruitment of C1-INH in merozoite populations, we used a flow cytometry–based assay. C1-INH binding to merozoites was detected using an anti–C1-INH Ab. The results show that merozoites incubated in NHS or NHS-E have an MFI of 51 or 53, respectively, which is significantly elevated compared with buffer and NHS-HI, which display an MFI of 2 or 7, respectively (p , 0.0001, one-way ANOVA, Fig. 1C). No significant difference was observed between NHS-HI and buffer (Fig. 1C). Col- lectively, our results show that P. falciparum merozoites actively recruit the human complement regulator C1-INH to their surfaces. PfMSP3.1 binds C1-INH To identify the MSP that recruits C1-INH, immunoprecipitation assays were performed using anti–C1-INH Abs in the presence or absence of serum-purified C1-INH and solubilized merozoite lysate as a source of MSPs. Mass spectrometry analyses of eluates

identiﬁed PfMSP3.1 as a potential candidate that was only de- Downloaded from tected in the presence of C1-INH (data not shown). To conﬁrm these results, we used the same immunoprecipitation strategy as above and analyzed eluates by Western blotting using anti- PfMSP3.1 Abs (Fig. 2A). Anti–C1-INH Abs immunoprecipitated C1-INH in a complex with PfMSP3.1, which migrates at 40 kDa

when processed and on the merozoite surface (Fig. 2A). Fur- http://www.jimmunol.org/ thermore, immunoprecipitation of C1-INH did not pull down the known complement regulator, binding protein Pf92, which binds FH, or the related family member PfMSP3.2, indicating that PfMSP3.1’s interaction with C1-INH is specific (Fig. 2A, 2B) (34) Reciprocal immunoprecipitation assays were conducted using anti-PfMSP3.1 and anti-Pf92 Abs in the presence of purified C1-INH and solubilized merozoite lysate (Fig. 2C). Although PfMSP3.1 and Pf92 were successfully immunoprecipitated by their respective Abs, only anti-PfMSP3.1 immunoprecipitated C1-INH in a complex by guest on September 29, 2021 (Fig. 2C). PfMSP3.1 is essential for C1-INH recruitment to merozoites There are two major allelic variants of PfMSP3.1: the K1-like allele and the 3D7-like allele (42). To determine whether C1-INH recruitment was conserved between PfMSP3.1 allelic variants, merozoites from the 3D7-like 3D7 strain and the K1-like FCR3 strain were incubated in buffer, NHS, NHS-HI, or NHS-E to examine C1-INH recruitment (42). We observed that 3D7 and FCR3 merozoites were able to bind C1-INH in NHS, NHS-HI, and NHS-E, albeit with reduced signal in the NHS-HI condition (Fig. 3A, upper panels). FIGURE 1. C1-INH from human serum binds to merozoite surfaces. (A) To examine C1-INH deposition, merozoites deposited with NHS, NHS-HI, To determine whether PfMSP3.1 is solely responsible for NHS-E, or buffer only (B) for the indicated times were analyzed by Western C1-INH binding, we examined whether C1-INH recruitment was blot with anti–C1-INH Abs. Loading controls were probed with anti-AMA1 affected in transgenic PfMSP3.1-knockout parasites (DPfMSP3.1) Abs. Migration of C1-INH from NHS and purified human C1-INH (29). As expected, DPfMSP3.1 merozoites lacked PfMSP3.1 ex- (pC1-INH) is shown (left panel), and molecular mass marker (in kDa) is pression, whereas 3D7 merozoites showed PfMSP3.1 expression, indicated to the left. (B) Merozoites were deposited with buffer, NHS, NHS- represented by a band migrating at 40 kDa (Fig. 3B, bottom panel). HI, or NHS-E. C1-INH was detected using a polyclonal anti–C1-INH Ab, 3D7 and DPfMSP3.1 merozoites were incubated in buffer, NHS, followedbyananti-goatsecondaryAbconjugatedtoAlexaFluor647(ma- NHS-HI, or NHS-E to examine C1-INH recruitment. Although 3D7 genta). AMA1 was detected with polyclonal anti-AMA1 Abs, followed by merozoites recruited C1-INH in NHS, NHS-HI, and NHS-E, anti-rabbit secondary Ab conjugated to Alexa Fluor 555 (green). DAPI (cyan) DPfMSP3.1 merozoites were deficient in the ability to recruit and the differential interference contrast (DIC) channel are shown. Scale bars, 2 mm. (C) Binding of C1-INH to merozoites deposited with buffer, NHS, C1-INH in NHS and NHS-E (Fig. 3B, top panel). A faint band D NHS-HI, or NHS-E was measured by flow cytometry. Data represent the consistent with C1-INH was observed binding to PfMSP3.1 background subtracted mean MFI of three independent experiments 6 SEM. merozoites in the NHS-HI condition. However, given the lack of *p , 0.05 versus buffer, one-way ANOVA. n.s., not significant. recruitment in NHS and NHS-E, this product may represent binding of heat-denatured C1-INH rather than native C1-INH. tion of C1-INH and AMA1 on the surface of merozoites (Fig. 1B). Using anti-AMA1 Abs as a loading control, we detected slightly The staining was specific, because C1-INH signal was absent in more parasite protein in the DPfMSP3.1 lysates (Fig. 3B, middle buffer control. These results demonstrate that C1-INH is recruited panel). We also examined C1-INH recruitment in transgenic-knockout 4732 C1-INH REGULATES MEROZOITE COMPLEMENT DEPOSITION Downloaded from FIGURE 2. MSP PfMSP3.1 binds C1 inhibitor. (A) Western blots of anti–C1-INH coimmunoprecipitation eluates from parasite lysate containing PBS (2C1-INH) or purified C1-INH (+C1-INH). Membranes were probed with anti-PfMSP3.1 (top panel), anti-Pf92 (middle panel), and anti-C1-INH (bottom panel) Ab. Asterisk (*) indicates a cross-reactive band. (B) Western blots of anti–C1-INH coimmunoprecipitation eluates from parasite lysate containing PBS (2C1-INH) or purified C1-INH (+C1-INH). Membranes were probed with anti-PfMSP3.2 and anti-C1-INH Abs. Asterisk (*) indicates a cross-reactive band, and arrows indicate the expected migration of unprocessed (upper) and processed (lower) PfMSP3.2. (C) Western blots of anti-Pf92 (a-Pf92 IP) and anti- PfMSP3.1 (a-PfMSP3.1 IP) coimmunoprecipitation eluates of parasite lysate containing purified C1-INH. The membrane was probed with anti–C1-INH Abs (top panel), anti-PfMSP3.1 Abs (bottom panel), and polyclonal anti-Pf92 (middle panel) Ab. Molecular mass markers are indicated (in kDa) to the left. http://www.jimmunol.org/ parasites of another MSP3 family member, PfMSP3.4 (DPfMSP3.4), neuraminidase and O-glycosidase reduced the molecular mass to 70 and observed that C1-INH binding was not affected on merozoites kDa (Supplemental Fig. 1, upper panel). Removal of both N-and incubated with NHS, NHS-HI, and NHS-E (Fig. 3C). This result O-linked glycans using PNGase F, neuraminidase, and O-glycosidase indicates that the loss of C1-INH recruitment is specific to the altered the molecular mass to 63 kDa (Supplemental Fig. 1). DPfMSP3.1 strain (Fig. 3B). These differences are consistent with previous reports (43). Anti– C1-INH Abs were still able to detect deglycosylated C1-INH, PfMSP3.1 binds to the C1-INH serpin domain independently of albeit with reduced sensitivity after treatment with O-glyco- glycosylation sidase and neuraminidase (Supplemental Fig. 1, lower panel). by guest on September 29, 2021 To examine whether C1-INH glycosyl groups were required for Western blotting analyses of eluates from anti-PfMSP3.1 im- PfMSP3.1 binding, we treated C1-INH with a series of glycosi- munoprecipitation assays containing deglycosylated C1-INH and dases and monitored complex formation with PfMSP3.1. Removal solubilized parasite proteins showed that removal of N- and of N-linked glycans from purified C1-INH using PNGase F led to O-linked glycosyl groups did not affect the ability of C1-INH to a reduction to 72 kDa, whereas removal of O-linked glycans with complex with PfMSP3.1 (Fig. 4A). As expected, reciprocal im-

FIGURE 3. DPfMSP3.1 merozoites are unable to recruit C1-INH. (A) C1-INH deposition on 3D7 (left panel) and FCR3 (middle panel) merozoites deposited with buffer, NHS, NHS-HI, or NHS-E was examined by Western blot with anti–C1-INH Abs, and anti-AMA1 Abs as a loading control. The migration of puriﬁed C1-INH is shown (right panel). (B) Deposition of C1-INH on 3D7 (WT) and DPfMSP3.1 merozoites that were deposited with buffer, NHS, NHS-HI, or NHS-E was examined by Western blot with anti–C1-INH Abs, anti-PfMSP3.1 Abs, and with anti-AMA1 Abs as a loading control. Bands marked with an asterisk (*) represent cross-reactive protein species. (C) C1-INH deposition on DPfMSP3.4 merozoites deposited with buffer, NHS, NHS-HI, or NHS-E was examined by Western blot with anti–C1-INH Abs and anti-AMA1 Abs as a loading control. Molecular mass markers are indicated (in kDa) to the left. The Journal of Immunology 4733 munoprecipitation assays with anti–C1-INH Abs revealed that incubated in buffer alone or with the proteolytically inactive zC1s PfMSP3.1 is still able to complex with all forms of deglycosylated (Fig. 5A). This result is consistent with attachment of the 26-kDa C1-INH (Fig. 4B). These results suggest that PfMSP3.1 recruits recombinant protease to form the C1-INH–protease complex. C1-INH by recognition of the polypeptide chain and not its ex- MASP2 additionally produced a 110-kDa product that is likely tensive glycosyl groups. “cleaved” C1-INH–protease complex created by exposure of MASP2 To map whether the C1-INH N-terminal domain or serpin domain cleavage sites during conformational rearrangements upon complex contains the PfMSP3.1 interaction site, we used a recombinant formation (Fig. 5A). fragment of C1-INH that contains only the serpin domain, which Western blotting analyses of merozoites that were incubated with encompasses amino acid residues 111–500. Analyses of immuno- C1-INH, followed by C1s or zC1s, showed that C1-INH formed a precipitation eluates using anti-PfMSP3.1 Abs revealed PfMSP3.1 in C1-INH–protease complex only with C1s, and not with zC1s as a complex with the serpin domain (Fig. 4C). A reciprocal immu- expected (Fig. 5B). Furthermore, similar experiments using MASP1 noprecipitation assay with anti–C1-INH Abs precipitated C1-INH or MASP2 revealed the presence of C1-INH and the C1-INH–pro- serpin domain in a complex with PfMSP3.1(Fig.4D).Together, tease complex for these two proteases (Fig. 5C). In addition, after 30 these data illustrate that the PfMSP3.1 binding site resides within the min, cleaved C1-INH–protease complex was detected in the MASP2 C1-INH serpin domain and is not reliant on C1-INH glycosylation. condition (Fig. 5C). Furthermore, C1-INH–protease complexes could not be detected on merozoites that were incubated with C1-INH retains its protease inhibitory activity while bound to C1-INH, followed by buffer or zC1s, despite the presence of C1-INH PfMSP3.1 on these cells (Fig. 5B, 5C). Together, these results show that C1-INH

C1-INH inhibits C1s, C1r, MASP1, and MASP2 by forming a retains its protease-inhibitory functions against C1s, MASP1, and Downloaded from covalent ester bond with the proteases to create a C1-INH–protease MASP2 when bound to the merozoite surface by PfMSP3.1. complex. To determine whether C1-INH was still active when bound to the merozoite surface, we examined whether C1-INH– Deletion of PfMSP3.1 leads to increased complement protease complex was present on merozoites. Using an in-solution deposition on the merozoite surface and an enhancement of complex-formation assay, we showed that 100-kDa puriﬁed invasion in the presence of human serum ∼

C1-INH formed an 125-kDa complex when incubated with C1-INH regulates the lectin and classical pathways of complement http://www.jimmunol.org/ recombinant C1s, MASP1, and MASP2 but not when C1-INH was activation, and both lead to pathogen lysis. We ﬁrst determined

FIGURE 4. PfMSP3.1 binds to the C1-INH serpin domain and is not dependent on C1-INH glycosylation. (A) Western blots of anti- PfMSP3.1 coimmunoprecipitation eluates from parasite lysate containing C1-INH (+C1-INH), by guest on September 29, 2021 PNGase F (PNGF)-treated C1-INH, O-glycosidase (Oglc) and neuraminidase (Nm)-treated C1-INH, C1-INH treated with all three enzymes, or an equivalent volume of PBS (2C1-INH). Membranes were probed with anti–C1-INH Abs (upper panel) and anti-PfMSP3.1 Abs (lower panel). (B) Western blots of anti–C1-INH coimmunoprecipitation eluates from parasite lysate containing C1-INH (+C1-INH), PNGF-treated C1-INH, Oglc and Nm-treated C1-INH, C1-INH treated with all three enzymes, or an equivalent volume of PBS (2C1-INH). Membranes were probed with anti-PfMSP3.1 Abs (upper panel) and anti–C1-INH Abs (lower panel). Arrows indicate unprocessed (upper) and processed (lower) PfMSP3.1. (C) Western blots of anti-PfMSP3.1 coimmunoprecipitation eluates from parasite lysate containing puriﬁed C1-INH (+pC1-INH), recombinant C1-INH serpin domain (+serpin), or an equivalent volume of PBS (2C1-INH). Membranes were probed with anti–C1-INH Abs (upper panels) and anti-PfMSP3.1 Abs (lower panel). (D) Parasite lysate was incubated with 5 mg/ml puriﬁed C1-INH (+pC1-INH), recombinant C1-INH serpin domain (+serpin), or an equivalent volume of PBS (2C1-INH). Immuno- precipitation eluates anti–C1-INH Abs were analyzed by Western blotting with anti-PfMSP3.1 Abs (upper panel) and anti–C1-INH Abs (lower panel). Arrow indicates PfMSP3.1. Asterisk (*) indicates an unknown band. Molecular mass markers are indicated (in kDa) to the left. 4734 C1-INH REGULATES MEROZOITE COMPLEMENT DEPOSITION

FIGURE 5. C1-INH retains the ability to form complexes with proteases when bound to the merozoite surface. C1-INH–protease complex formation was determined by Western blotting with anti-C1-INH Abs (A–C; upper panels) and was probed with anti-AMA1 Abs for a loading control (B and C; lower panels). Arrows indicate C1-INH–protease complex (upper) and cleaved complex (lower). Molecular mass markers are indicated (in kDa) to the left. (A) Downloaded from Purified C1-INH (0.5 mM) was incubated with 1 mM recombinant C1s, zC1s, MASP1, MASP2, or an equivalent volume of buffer (GPBS++) for 10 and 30 min. (B) Wild-type merozoites were deposited with purified C1-INH for 10 min prior to incubation with C1s, zC1s, or buffer for 10 and 30 min. (C) Wild-type merozoites were deposited with purified C1-INH for 10 min prior to incubation with MASP1, MASP2, or buffer for 10 and 30 min. whether rabbit Abs to MSPs were able to activate the classical moderate, but significant, increase in invasion compared with wild-

pathway on the merozoite surface. Merozoites were incubated with type merozoites (p , 0.0001, two-way ANOVA, Fig. 6E, 6F). This http://www.jimmunol.org/ PBS or opsonized with anti-PfMSP3.4 Abs prior to incubation difference increased moderately after two cycles of parasite growth with NHS for 1, 2.5, 5, or 10 min or with NHS-HI for 10 min. and invasion in active serum, because the mean difference for anti- Within 1 min, DPfMSP3.1 (Fig. 6A) and wild-type (data not PfMSP3.2, anti-MSP1, and anti-PfMSP3.4 conditions increased shown) merozoites opsonized with anti-PfMSP3.4 had a higher from 6.62, 3.97 and 12.1 to 18.6, 28.8, and 22.7, respectively C3b density than did those incubated in PBS, which is consistent (Fig. 6E, 6F). with classical pathway activation. Minimal C3b deposition was observed in NHS-HI, indicating that C3b deposition in the NHS Discussion conditions was derived from complement activation (Fig. 6A). Pathogens exposed to serum require strategies to overcome the We next examined whether C1-INH could regulate lectin and onslaught of complement, which can lead to clearance through the by guest on September 29, 2021 classical pathway complement activation on the merozoite surface classical, lectin, or alternative pathway. In this article, we show by comparing C3b deposition on wild-type and DPfMSP3.1 that P. falciparum blood stage merozoites recruit the human merozoites that had been incubated with PBS (data not shown) or complement regulator C1-INH to their surfaces to downregulate opsonized with anti-PfMSP3.4 Abs (Fig. 6B). In both conditions, complement deposition. Merozoite-bound C1-INH retained the C3b density was higher on DPfMSP3.1 merozoites compared with ability to form protease inhibitor complexes with C1s, MASP1, wild-type merozoites (Fig. 6B). and MASP2. PfMSP3.1, a member of the MSP3 family of surface We conducted invasion assays to determine whether increased proteins, recruited C1-INH via an interaction with the C1-INH C3b density on DPfMSP3.1 merozoites led to greater complement- serpin domain and did not depend on C1-INH glycosylation for mediated destruction. These assays were conducted with alterna- binding. Genetic deletion of PfMSP3.1 abolished C1-INH re- tive and lectin pathway activation via the addition of PBS, BSA, or cruitment and increased deposition of the complement component prebleed IgG or with activation of the alternative, lectin, and C3b on the merozoite surface. This resulted in enhanced invasion classical pathways through the addition of anti-PfMSP3.2, anti- of DPfMSP3.1 merozoites in active complement, supporting a MSP1, or anti-PfMSP3.4 Abs. We found that wild-type merozoite complex relationship between P. falciparum and the complement invasion in the presence of NHS containing anti-MSP1 or anti- system. PfMSP3.4 Abs was 26 and 30% lower, respectively, than invasion PfMSP3.1 is a component of two vaccines currently in trials in the presence of NHS-HI containing the same Abs (Fig. 6C). against blood stage malaria (44–47). Previous studies show that DPfMSP3.1 merozoite invasion was not signiﬁcantly reduced in anti-PfMSP3.1 Abs inhibit parasite growth when cocultured with NHS containing anti-MSP1 Abs, and invasion was only reduced monocytes and that anti-PfMSP3.1 Abs are associated with pro- by 13.8% in the presence of NHS containing anti-PfMSP3.4 Abs tection from severe malaria (46–50). We found that PfMSP3.1 compared with in the presence of NHS-HI containing anti-PfMSP3.4 interacts with the immune system by recruiting C1-INH to the Abs (Fig. 6D). Together, these results indicate that, under certain merozoite surface (Figs. 1–3). PfMSP3.1 is a member of the conditions of complement activation, wild-type and, to a lesser MSP3 family and is composed of three major domains: N-terminal extent, DPfMSP3.1 merozoites undergo complement-mediated heptad repeats, a glutamic acid–rich region, and a C-terminal destruction during invasion. leucine-like zipper region. It lacks a GPI anchor and instead as- To further examine whether C1-INH recruitment affected sociates peripherally with the merozoite surface via interactions complement-mediated destruction, we compared wild-type and with the MSP1 complex (29). The C-terminal region of PfMSP3.1 DPfMSP3.1 merozoite invasion by determining the percentage of is conserved, whereas the N-terminal region is polymorphic, merozoite invasion in both cell lines for each NHS condition containing deletions and substitutions in sequences ﬂanking and normalized to merozoite invasion in the matched NHS-HI con- within nonrepeat regions of the heptad repeats that can be broken dition. Our results show that DPfMSP30.1 merozoites exhibited a into 3D7-like and K1-like alleles (42, 51). C1-INH binding is a The Journal of Immunology 4735 Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021

FIGURE 6. DPfMSP3.1 merozoite invasion is increased in the presence of active classical pathway complement activity. (A and B) C3b deposition was determined by Western blotting with anti-C3 Abs (upper panels) and with anti-AMA1 Abs (lower panels) as a loading control. Molecular mass markers are indicated (in kDa) to the left. (A) DPfMSP3.1 merozoites were incubated with PBS or opsonized with anti-PfMSP3.4 (a-MSP3.4) Abs and then incubated for the indicated times in 5% NHS or NHS-HI. Purified C3 as a b-chain migration control is shown (left panel). (B) 3D7 (WT) and DPfMSP3.1 merozoites were opsonized with anti-PfMSP3.4 Abs and then incubated for the indicated times in NHS or NHS-HI. (C) 3D7 (WT) merozoites were allowed to invade RBCs in the presence of active serum (NHS) or NHS-HI in the presence of PBS, BSA, rabbit preimmune IgG (Prebleed), or rabbit IgG targeting PfMSP3.2 (a-MSP3.2), MSP1 (a-MSP1), or PfMSP3.4 (a-MSP3.4). (D) DPfMSP3.1 merozoites were allowed to invade RBCs in the presence of active serum (NHS) or NHS-HI in the presence of PBS, BSA, rabbit preimmune IgG (Prebleed), or rabbit IgG targeting PfMSP3.2 (a-MSP3.2), MSP1 (a-MSP1), or PfMSP3.4 (a-MSP3.4). For (C) and (D), invasion events were normalized to invasion in media alone, and error bars indicate SEM from five independent experiments, **p , 0.01, ****p , 0.0001, Sidak multiple-comparison test. (E) Wild-type (3D7) and DPfMSP3.1 growth for one cycle of invasion in active serum, expressed as a percentage of one cycle of invasion in NHS-HI within each above condition. (F) Wild-type (3D7) and DPfMSP3.1 growth for two cycles of invasion in active serum, expressed as a percentage of two cycles of invasion in NHS-HI within each above condition. (E and F) Error bars indicate SEM for at least three independent experiments. The p values were determined using two-way ANOVA. conserved feature of PfMSP3.1 variants, with 3D7 (3D7-like) and suggesting no requirement for PfGPI in our experiments (Fig. 3B). FCR3 (K1-like) merozoites recruiting C1-INH (Fig. 3A). In ad- However, we did observe low-level C1-INH recruitment to dition, we showed that D10 strain merozoites, which represent a DPfMSP3.1 merozoites in NHS-HI conditions, suggesting that recombination between K1 and 3D7-like PfMSP3.1, also recruit another ligand, such as PfGPI, may bind C1-INH to merozoites C1-INH (Fig. 1). It will be interesting to determine which region under certain conditions (Fig. 3B). Because PfMSP3.1-independent of PfMSP3.1 recruits C1-INH and how this corresponds with re- binding of C1-INH was only observed in NHS-HI, we hypothesize gions targeted by protective and vaccine-induced Abs. that it may occur in response to temperature. At temperatures above C1-INH has been shown to bind to the surface of P. falciparum 55˚C, C1-INH undergoes conformational changes due to insertion merozoites through an interaction with PfGPI (35). This previous of the reactive center loop that produces polymeric C1-INH and a report used a far Western blotting and GPI microarray approach small amount of latent monomeric C1-INH (41). Thus, in our heat- with a plasma-purified form of C1-INH that is pasteurized and inactivated serum, there may be no native C1-INH available for lyophilized for use as a therapeutic (35). In contrast, our experi- PfMSP3.1 to bind, explaining the reduced signal. Meanwhile, an ments with donor serum and free merozoites show that deletion of alternate ligand may bind to the latent monomeric or polymeric PfMSP3.1 was sufficient to ablate C1-INH binding on merozoites, C1-INH present in NHS-HI, thus explaining the C1-INH signal on 4736 C1-INH REGULATES MEROZOITE COMPLEMENT DEPOSITION wild-type and DPfMSP3.1 merozoites incubated in that condition dates, elucidating their immune-evasion functions will be an important (Figs. 1, 3B). Clearly, further studies will be required to validate this consideration during vaccine development. hypothesis and characterize its significance. Because PfMSP3.1 also binds to free heme, future studies will also be needed to determine Acknowledgments whether PfMSP3.1 bound to C1-INH binds to heme as well (31). We thank the Bio21 Mass Spectrometry Facility at the University of Recruitment of C1-INH to the surface of B. pertussis by its Melbourne for assistance with mass spectrometry sample preparation, surface protein Vag8 is the only other reported example of a data collection, and analysis. pathogen hijacking C1-INH for complement protection (24). However, the Vag8 binding site within C1-INH has not been Disclosures identified. Pathogens often exploit self-binding regions within host The authors have no financial conflicts of interest. complement regulators for recruitment. C1-INH binds to the surface of macrophages, neutrophils, and endothelial cells using its glycosyl groups; however, our results demonstrate that C1-INH References glycosylation is not essential for binding to PfMSP3.1 (Fig. 4A, 1. World Health Organization, World Malaria Report 2015. Available at: http:// apps.who.int/iris/bitstream/10665/200018/1/9789241565158_eng.pdf. Accessed: 4B) (52–55). Our results show that PfMSP3.1 binds to the C1-INH April 24, 2017. serpin domain (Fig. 4C, 4D). Further studies will be required to 2.Boyle,M.J.,D.W.Wilson,J.S.Richards,D.T.Riglar,K.K.A.Tetteh, D. J. Conway, S. A. Ralph, J. Baum, and J. G. Beeson. 2010. Isolation of viable determine which region of the C1-INH serpin domain PfMSP3.1 Plasmodium falciparum merozoites to define erythrocyte invasion events and binds to and whether this overlaps with C1-INH’s extracellular advance vaccine and drug development. Proc. Natl. Acad. Sci. USA 107: matrix protein binding region (56). 14378–14383. Downloaded from 3. Cowman, A. F., and B. S. Crabb. 2006. Invasion of red blood cells by malaria Recruited C1-INH retained the ability to form protease inhibitor parasites. Cell 124: 755–766. complexes with C1s, MASP1, and MASP2 on the merozoite 4. Weiss, G. E., P. R. Gilson, T. Taechalertpaisarn, W.-H. Tham, N. W. de Jong, surface (Fig. 5). This allowed C1-INH to control complement K. L. Harvey, F. J. Fowkes, P. N. Barlow, J. C. Rayner, G. J. Wright, et al. 2015. Revealing the sequence and resulting cellular morphology of receptor-ligand deposition on the merozoite surface, with a loss of C1-INH re- interactions during Plasmodium falciparum invasion of erythrocytes. PLoS cruitment on DPfMSP3.1 coinciding with increased C3b deposi- Pathog. 11: e1004670. 5. Walport, M. J. 2001. Complement. First of two parts. N. Engl. J. Med. 344: tion (Fig. 6B). Surprisingly, the loss of complement regulation did 1058–1066. http://www.jimmunol.org/ not result in increased complement-mediated destruction. Instead, 6. Schatz-Jakobsen, J. A., D. V. Pedersen, and G. R. Andersen. 2016. Structural DPfMSP3.1 merozoites invaded RBCs more readily in active insight into proteolytic activation and regulation of the complement system. Immunol. Rev. 274: 59–73. complement, with the difference potentially compounding over 7. Garred, P., N. Genster, K. Pilely, R. Bayarri-Olmos, A. Rosbjerg, Y. J. Ma, and multiple cycles of parasite growth in active complement (Fig. 6E, M.-O. Skjoedt. 2016. A journey through the lectin pathway of complement-MBL 6F). Although surprising, this result is supported by the current and beyond. Immunol. Rev. 274: 74–97. 8. Zipfel, P. F., M. Mihlan, and C. Skerka. 2007. The alternative pathway of understanding that merozoites use one of their invasion ligands, complement: a pattern recognition system. Adv. Exp. Med. Biol. 598: 80–92. PfRh4, to engage CR1 as an entry portal (57). CR1 is a comple- 9. Ricklin, D., E. S. Reis, D. C. Mastellos, P. Gros, and J. D. Lambris. 2016. Complement component C3 - The “Swiss Army Knife” of innate immunity and ment regulator found on the RBC surface, which has C3b and host defense. Immunol. Rev. 274: 33–58. iC3b binding properties. It was recently shown that engagement of 10. Law, S. K., N. A. Lichtenberg, and R. P. Levine. 1979. Evidence for an ester by guest on September 29, 2021 CR1 by C3b and iC3b deposited on the merozoite surface leads to linkage between the labile binding site of C3b and receptive surfaces. J. Immunol. 123: 1388–1394. enhanced merozoite invasion (58). These findings suggest that 11. Zipfel, P. F., and C. Skerka. 2009. Complement regulators and inhibitory pro- DPfMSP3.1 merozoites are invading more readily than wild-type, teins. Nat. Rev. Immunol. 9: 729–740. because the lack of C1-INH recruitment results in increased C3b 12. Schmidt, C. Q., J. D. Lambris, and D. Ricklin. 2016. Protection of host cells by complement regulators. Immunol. Rev. 274: 152–171. deposition and, thus, increased CR1 engagement. In contrast, 13. Ziccardi, R. J. 1981. Activation of the early components of the classical com- disruption of other merozoite complement-evasion strategies in- plement pathway under physiologic conditions. J. Immunol. 126: 1769–1773. creased complement-mediated destruction of merozoites (34). We 14. Kerr, F. K., A. R. Thomas, L. C. Wijeyewickrema, J. C. Whisstock, S. E. Boyd, D. Kaiserman, A. Y. Matthews, P. I. Bird, N. M. Thielens, V. Rossi, and propose that DPfMSP3.1 merozoites do not undergo this signifi- R. N. Pike. 2008. Elucidation of the substrate specificity of the MASP-2 protease cant complement-mediated destruction, because they are still able of the lectin complement pathway and identification of the enzyme as a major physiological target of the serpin, C1-inhibitor. Mol. Immunol. 45: 670–677. to recruit FH and FHL-1 through Pf92 (34). FH and FHL-1 con- 15. Matsushita, M., S. Thiel, J. C. Jensenius, I. Terai, and T. Fujita. 2000. Proteolytic vert C3b to less active iC3b on the merozoite surface and accel- activities of two types of mannose-binding lectin-associated serine protease. erate decay of positive-feedback loop C3 convertases, which J. Immunol. 165: 2637–2642. 16. Petersen, S. V., S. Thiel, L. Jensen, T. Vorup-Jensen, C. Koch, and collectively prevent strong complement amplification from the J. C. Jensenius. 2000. Control of the classical and the MBL pathway of com- alternative pathway on the merozoite surface that may otherwise drive plement activation. Mol. Immunol. 37: 803–811. membrane attack complex formation and complement-mediated de- 17. Presanis, J. S., K. Hajela, G. Ambrus, P. Ga´l, and R. B. Sim. 2004. Differential substrate and inhibitor profiles for human MASP-1 and MASP-2. Mol. Immunol. struction (34, 59). Thus, merozoite complement-evasion strategies 40: 921–929. may provide overlapping protection. In addition, P. falciparum may 18. Bock, S. C., K. Skriver, E. Nielsen, H. C. Thøgersen, B. Wiman, V. H. Donaldson, R. L. Eddy, J. Marrinan, E. Radziejewska, R. Huber, et al. strike a fine balance of allowing complement deposition to en- 1986. Human C1 inhibitor: primary structure, cDNA cloning, and chromosomal hance RBC invasion while avoiding complement-mediated de- localization. Biochemistry 25: 4292–4301. struction. Future studies should examine whether merozoite 19. Davis III., A. E., P. Mejia, and F. Lu. 2008. Biological activities of C1 inhibitor. Mol. Immunol. 45: 4057–4063. C1-INH recruitment also influences the other systems that C1-INH 20. Blom, A. M., T. Hallstro¨m, and K. Riesbeck. 2009. Complement evasion strat- regulates, which include the coagulation, kallikrein–kinin, and fi- egies of pathogens-acquisition of inhibitors and beyond. Mol. Immunol. 46: brinolytic systems. 2808–2817. 21. Avirutnan, P., R. E. Hauhart, P. Somnuke, A. M. Blom, M. S. Diamond, and P. falciparum merozoites clearly exploit human complement J. P. Atkinson. 2011. Binding of flavivirus nonstructural protein NS1 to C4b regulators, actively recruiting them to their surfaces or engaging binding protein modulates complement activation. J. Immunol. 187: 424–433. 22. Schneider, M. C., B. E. Prosser, J. J. Caesar, E. Kugelberg, S. Li, Q. Zhang, them as cognate receptors during the RBC-invasion process (33, 34, S. Quoraishi, J. E. Lovett, J. E. Deane, R. B. Sim, et al. 2009. Neisseria men- 57, 60). We hypothesize that there are many more complement- ingitidis recruits factor H using protein mimicry of host carbohydrates. Nature evasion strategies involving host–pathogen interactions among the 458: 890–893. 23. Madico, G., J. A. Welsch, L. A. Lewis, A. McNaughton, D. H. Perlman, human immune system, parasite MSPs, and parasite invasion li- C. E. Costello, J. Ngampasutadol, U. Vogel, D. M. Granoff, and S. Ram. gands. Because some of these proteins are putative vaccine candi- 2006. The meningococcal vaccine candidate GNA1870 binds the complement The Journal of Immunology 4737

regulatory protein factor H and enhances serum resistance. J. Immunol. 177: 42. Huber, W., I. Felger, H. Matile, H. J. Lipps, S. Steiger, and H. P. Beck. 1997. 501–510. Limited sequence polymorphism in the Plasmodium falciparum merozoite sur- 24. Marr, N., N. R. Shah, R. Lee, E. J. Kim, and R. C. Fernandez. 2011. Bordetella face protein 3. Mol. Biochem. Parasitol. 87: 231–234. pertussis autotransporter Vag8 binds human C1 esterase inhibitor and confers 43. Liu, D., X. Gu, J. Scaﬁdi, and A. E. Davis, III. 2004. N-linked glycosylation is serum resistance. PLoS One 6: e20585. required for c1 inhibitor-mediated protection from endotoxin shock in mice. 25. Brooks, C. S., S. R. Vuppala, A. M. Jett, A. Alitalo, S. Meri, and D. R. Akins. Infect. Immun. 72: 1946–1955. 2005. Complement regulator-acquiring surface protein 1 imparts resistance to 44. Sirima, S. B., B. Mordmuller,€ P. Milligan, U. A. Ngoa, F. Kironde, F. Atuguba, human serum in Borrelia burgdorferi. J. Immunol. 175: 3299–3308. A. B. Tiono, S. Issifou, M. Kaddumukasa, O. Bangre, et al; GMZ2 Trial Study 26. McDowell, J. V., J. Wolfgang, E. Tran, M. S. Metts, D. Hamilton, and Group. 2016. A phase 2b randomized, controlled trial of the efﬁcacy of the R. T. Marconi. 2003. Comprehensive analysis of the factor h binding capabilities GMZ2 malaria vaccine in African children. Vaccine 34: 4536–4542. of Borrelia species associated with lyme disease: delineation of two distinct 45. Sirima, S. B., I. Ne´bie´,A.Oue´draogo, A. B. Tiono, A. T. Konate´, A. Gansane´, classes of factor h binding proteins. Infect. Immun. 71: 3597–3602. A. I. Derme´, A. Diarra, A. Oue´draogo, I. Soulama, et al. 2007. Safety and im- 27. Cowman, A. F., D. Berry, and J. Baum. 2012. The cellular and molecular basis munogenicity of the Plasmodium falciparum merozoite surface protein-3 long for malaria parasite invasion of the human red blood cell. J. Cell Biol. 198: synthetic peptide (MSP3-LSP) malaria vaccine in healthy, semi-immune adult 961–971. males in Burkina Faso, West Africa. Vaccine 25: 2723–2732. 28. Singh, S., S. Soe, S. Weisman, J. W. Barnwell, J. L. Pe´rignon, and P. Druilhe. 46. Audran, R., M. Cachat, F. Lurati, S. Soe, O. Leroy, G. Corradin, P. Druilhe, and 2009. A conserved multi-gene family induces cross-reactive antibodies effective F. Spertini. 2005. Phase I malaria vaccine trial with a long synthetic peptide de- in defense against Plasmodium falciparum. PLoS One 4: e5410. rived from the merozoite surface protein 3 antigen. Infect. Immun. 73: 8017–8026. 29. Lin, C. S., A. D. Uboldi, C. Epp, H. Bujard, T. Tsuboi, P. E. Czabotar, and 47. Esen, M., P. G. Kremsner, R. Schleucher, M. Ga¨ssler, E. B. Imoukhuede, A. F. Cowman. 2016. Multiple P. falciparum merozoite surface protein 1 com- N. Imbault, O. Leroy, S. Jepsen, B. W. Knudsen, M. Schumm, et al. 2009. Safety plexes mediate merozoite binding to human erythrocytes. J. Biol. Chem. 291: and immunogenicity of GMZ2 - a MSP3-GLURP fusion protein malaria vaccine 7703–7715. candidate. Vaccine 27: 6862–6868. 30. Lin, C. S., A. D. Uboldi, D. Marapana, P. E. Czabotar, C. Epp, H. Bujard, 48. Theisen, M., S. Soe, K. Brunstedt, F. Follmann, L. Bredmose, H. Israelsen, N. L. Taylor, M. A. Perugini, A. N. Hodder, and A. F. Cowman. 2014. The S. M. Madsen, and P. Druilhe. 2004. A Plasmodium falciparum GLURP-MSP3 merozoite surface protein 1 complex is a platform for binding to human eryth- chimeric protein; expression in Lactococcus lactis, immunogenicity and induc- rocytes by Plasmodium falciparum. J. Biol. Chem. 289: 25655–25669. tion of biologically active antibodies. Vaccine 22: 1188–1198. Downloaded from 31. Imam, M., S. Singh, N. K. Kaushik, and V. S. Chauhan. 2014. Plasmodium 49. Badell, E., C. Oeuvray, A. Moreno, S. Soe, N. van Rooijen, A. Bouzidi, and falciparum merozoite surface protein 3: oligomerization, self-assembly, and P. Druilhe. 2000. Human malaria in immunocompromised mice: an in vivo heme complex formation. J. Biol. Chem. 289: 3856–3868. model to study defense mechanisms against Plasmodium falciparum. J. Exp. 32. Crosnier, C., Z. Iqbal, E. Knuepfer, S. Maciuca, A. J. Perrin, G. Kamuyu, Med. 192: 1653–1660. D. Goulding, L. Y. Bustamante, A. Miles, S. C. Moore, et al. 2016. Binding of 50. Oeuvray, C., H. Bouharoun-Tayoun, H. Gras-Masse, E. Bottius, T. Kaidoh, Plasmodium falciparum merozoite surface proteins DBLMSP and DBLMSP2 to M. Aikawa, M. C. Filgueira, A. Tartar, and P. Druilhe. 1994. Merozoite surface human immunoglobulin M is conserved among broadly diverged sequence protein-3: a malaria protein inducing antibodies that promote Plasmodium fal-

variants. J. Biol. Chem. 291: 14285–14299. ciparum killing by cooperation with blood monocytes. Blood 84: 1594–1602. http://www.jimmunol.org/ 33. Rosa, T. F. A., A. Flammersfeld, C. J. Ngwa, M. Kiesow, R. Fischer, P. F. Zipfel, 51. Osier, F. H. A., S. D. Polley, T. Mwangi, B. Lowe, D. J. Conway, and K. Marsh. C. Skerka, and G. Pradel. 2016. The Plasmodium falciparum blood stages ac- 2007. Naturally acquired antibodies to polymorphic and conserved epitopes of quire factor H family proteins to evade destruction by human complement. Cell. Plasmodium falciparum merozoite surface protein 3. Parasite Immunol. 29: Microbiol. 18: 573–590. 387–394. 34. Kennedy, A. T., C. Q. Schmidt, J. K. Thompson, G. E. Weiss, 52. Chang, N. S., R. J. Boackle, and R. W. Leu. 1991. Characterization of C1 in- T. Taechalertpaisarn, P. R. Gilson, P. N. Barlow, B. S. Crabb, A. F. Cowman, and hibitor binding to neutrophils. Immunology 73: 95–101. W.-H. Tham. 2016. Recruitment of factor H as a novel complement evasion 53. Cai, S., and A. E. Davis, III. 2003. Complement regulatory protein C1 inhibitor strategy for blood-stage Plasmodium falciparum infection. J. Immunol. 196: binds to selectins and interferes with endothelial-leukocyte adhesion. J. Immunol. 1239–1248. 171: 4786–4791. 35. Mejia, P., M. Diez-Silva, F. Kamena, F. Lu, S. M. Fernandes, P. H. Seeberger, 54. Bergamaschini, L., G. Gobbo, S. Gatti, L. Caccamo, P. Prato, M. Maggioni, A. E. Davis, III., and J. R. Mitchell. 2016. Human C1-inhibitor suppresses P. Braidotti, R. Di Stefano, and L. R. Fassati. 2001. Endothelial targeting with

malaria parasite invasion and cytoadhesion via binding to parasite glyco- C1-inhibitor reduces complement activation in vitro and during ex vivo reper- by guest on September 29, 2021 sylphosphatidylinositol and host cell receptors. J. Infect. Dis. 213: 80–89. fusion of pig liver. Clin. Exp. Immunol. 126: 412–420. 36. Simon, N., E. Lasonder, M. Scheuermayer, A. Kuehn, S. Tews, R. Fischer, 55. Bergamaschini, L., S. Gatti, L. Caccamo, P. Prato, L. Latham, P. Trezza, P. F. Zipfel, C. Skerka, and G. Pradel. 2013. Malaria parasites co-opt human M. Maggioni, G. Gobbo, and L. R. Fassati. 2001. C1 inhibitor potentiates the factor H to prevent complement-mediated lysis in the mosquito midgut. Cell protective effect of organ preservation solution on endothelial cells during cold Host Microbe 13: 29–41. storage. Transplant. Proc. 33: 939–941. 37. Wilson, D. W., B. S. Crabb, and J. G. Beeson. 2010. Development of fluorescent 56. Patston, P. A., and M. Schapira. 1997. Regulation of C1-inhibitor function by Plasmodium falciparum for in vitro growth inhibition assays. Malar. J. 9: 152. binding to type IV collagen and heparin. Biochem. Biophys. Res. Commun. 230: 38. Riglar, D. T., D. Richard, D. W. Wilson, M. J. Boyle, C. Dekiwadia, L. Turnbull, 597–601. F. Angrisano, D. S. Marapana, K. L. Rogers, C. B. Whitchurch, et al. 2011. 57. Tham, W.-H., D. W. Wilson, S. Lopaticki, C. Q. Schmidt, P. B. Tetteh-Quarcoo, Super-resolution dissection of coordinated events during malaria parasite inva- P. N. Barlow, D. Richard, J. E. Corbin, J. G. Beeson, and A. F. Cowman. 2010. sion of the human erythrocyte. Cell Host Microbe 9: 9–20. Complement receptor 1 is the host erythrocyte receptor for Plasmodium falci- 39. Duncan, R. C., F. Mohlin, D. Taleski, T. H. Coetzer, J. A. Huntington, parum PfRh4 invasion ligand. Proc. Natl. Acad. Sci. USA 107: 17327–17332. R. J. Payne, A. M. Blom, R. N. Pike, and L. C. Wijeyewickrema. 2012. Iden- 58. Biryukov, S., E. Angov, M. E. Landmesser, M. D. Spring, C. F. Ockenhouse, and tification of a catalytic exosite for complement component C4 on the serine J. A. Stoute. 2016. Complement and antibody-mediated enhancement of red protease domain of C1s. J. Immunol. 189: 2365–2373. blood cell invasion and growth of malaria parasites. EBioMedicine 9: 207–216. 40. Wijeyewickrema, L. C., E. Lameignere, L. Hor, R. C. Duncan, T. Shiba, 59. Makou, E., A. P. Herbert, and P. N. Barlow. 2013. Functional anatomy of R. J. Travers, P. R. Kapopara, V. Lei, S. A. Smith, H. Kim, et al. 2016. Poly- complement factor H. Biochemistry 52: 3949–3962. phosphate is a novel cofactor for regulation of complement by a serpin, C1 60. Egan, E. S., R. H. Jiang, M. A. Moechtar, N. S. Barteneva, M. P. Weekes, inhibitor. Blood 128: 1766–1776. L. V. Nobre, S. P. Gygi, J. A. Paulo, C. Frantzreb, Y. Tani, et al. 2015. Malaria. A 41. Patston, P. A., J. Hauert, M. Michaud, and M. Schapira. 1995. Formation and forward genetic screen identifies erythrocyte CD55 as essential for Plasmodium properties of C1-inhibitor polymers. FEBS Lett. 368: 401–404. falciparum invasion. Science 348: 711–714. Figure S1 Non-Reducing Reducing PNGF - + - - + PNGF - + - - + OGlc - - + - + OGlc - - + - + Nm - - + + + Nm - - + + + 150 150 100 100 75 Coomassie 75 Coomassie

50 50 150 150 100 100 α 75 α-C1-INH 75 -C1-INH

50 50

Figure S1. Deglycosylation of purified C1-INH. Purified C1-INH was treated with PNGase F (PNGF), O-glycosidase (Oglc), neuraminidase (Nm) or a combination prior to separation by SDS-PAGE. Proteins were visualised by Coomassie brilliant blue (upper panel) or via western blot (lower panel) with anti-C1-INH antibodies in non-reducing (left panel) and reducing (right panel) conditions. Molecular weight markers are indicated in kDa on the left hand side.