Serine protease homolog pairs CLIPA4-A6, A4-A7Δ, and A4-A12 act as cofactors for proteolytic activation of prophenoloxidase-2 and -7 in Anopheles gambiae

Abstract

Phenoloxidase (PO) catalyzed melanization and other insect immune responses are mediated by serine proteases (SPs) and their noncatalytic homologs (SPHs). Many of these SP-like proteins have a regulatory clip domain and are called CLIPs. In most insects studied so far, PO precursors are activated by a PAP (i.e., PPO activating protease) and its cofactor of clip-domain SPHs. Although melanotic encapsulation is a well-known refractory mechanism of mosquitoes against malaria parasites, it is unclear if a cofactor is required for PPO activation. In Anopheles gambiae, CLIPA4 is 1:1 orthologous to Manduca sexta SPH2; CLIPs A5–7, A12–14, A26, A31, A32, E6, and E7 are 11:4 orthologous to M. sexta SPH1a, 1b, 4, and 101, SPH2 partners in the cofactors. Here we produced proCLIPs A4, A6, A7Δ, A12, and activated them with CLIPB9 or M. sexta PAP3. A. gambiae PPO2 and PPO7 were expressed in Escherichia coli for use as PAP substrates. CLIPB9 was mutated to CLIPB9_Xa by including a Factor X_a cleavage site. CLIPA7Δ was a deletion mutant with a low complexity region removed. After PAP3 or CLIPB9_Xa processing, CLIPA4 formed a high M_r complex with CLIPA6, A7Δ or A12, which assisted PPO2 and PPO7 activation. High levels of specific PO activity (55–85 U/μg for PO2 and 1,131–1,630 U/μg for PO7) were detected in vitro, indicating that cofactor-assisted PPO activation also occurs in this species. The cleavage sites and mechanisms for complex formation and cofactor function are like those reported in M. sexta and Drosophila melanogaster. In conclusion, these data suggest that the three (and perhaps more) SPHI-II pairs may form cofactors for CLIPB9-mediated activation of PPOs for melanotic encapsulation in A. gambiae.

Graphical Abstract

1. Introduction

Melanization is a crucial defense response to pathogen or parasite infection in insects (Nappi and Christensen, 2005, Kanost and Gorman, 2008; Marieshwari et al., 2023). In this process, phenoloxidase (PO) catalyzes multiple steps of a chemical reaction series which converts monophenols to diphenols, quinones, and other reactive intermediates until stable eumelanin forms (Zhao et al., 2007). Since many of these compounds are toxic to viruses, bacteria, fungi, parasites, parasitoids, and host cells (Zhao et al., 2011), POs are produced as inactive precursors (PPOs) and activated by PAPs (i.e., PPO activating proteases) when needed, to maximize killing of invading organisms and minimize damage of the reactive chemicals to host tissues. A system of pattern recognition receptors, serine proteases (SPs), and noncatalytic serine protease homologs (SPHs) has evolved in each insect to detect pathogens, form protein complexes, and activate PPOs and other immune responses at the site of wounding or infection (Kanost and Jiang, 2015; Veillard et al., 2016). In most species examined to date, PPO activation occurs in the presence of a high M_r complex of clip-domain SPHs to ensure a potent melanization response against non-self (Kwon et al., 2000; Yu et al., 2003; Kan et al., 2008; Wang et al., 2020; Jin et al., 2022 and 2023). But it is unclear how the cofactor, PAP, and PPO generate PO with much higher specific activity than PO produced in the control mixture of PPO and PAP.

Most members of the immune SP-SPH system in insect hemolymph contain a regulatory clip domain and a protease or protease-like domain and, therefore, are also called CLIPs. CLIPs in subfamily A (CLIPAs) are catalytically inactive (i.e., SPHs) whereas most CLIP Bs, Cs, and Ds are SPs (Cao and Jiang, 2018). Pathogen recognition leads to autoactivation of a modular serine protease, which activates a cascade of CLIPs (Wang and Jiang, 2006; Kim et al., 2008; Buchon et al., 2009; Takahashi et al., 2015; Wang et al., 2022). In many cases, CLIPCs activate CLIPBs while CLIPBs activate CLIPAs and effectors (e.g., POs) (Zhang et al., 2021; An et al., 2011; Wang and Jiang, 2006; Wang et al., 2014). Substitution of the catalytic Ser with Gly in CLIPAs abolishes the amidase activity but, after limited proteolysis, they may actively regulate immune responses such as melanization.

CLIPs mediate PPO, TEP1 and Toll pathway activation in mosquitoes (Volz et al., 2005; Paskewitz et al., 2006; Yassine et al., 2014; Nakhleh et al., 2017; Sousa et al., 2020; Zakhia and Osta, 2022). While some CLIPBs and CLIPCs participate in melanization, exact roles of CLIPAs have not been defined in most cases. CLIPA14 showed cofactor activity in PPO3 activation by CLIPB9 in Aedes aegypti (Ji et al, 2022). Among the 22 A. gambiae CLIPAs, A8, A28 and A30 (i.e., SPCLIP1) may form a “cSPH pathway” for TEP1-opsonized melanization of E. coli and Plasmodium berghei (Povelones et al., 2013; El Moussawi et al., 2019; Sousa et al., 2020). RNA interference of A. gambiae CLIPs A2, A5, A7, A8, and A14 suggested their involvement in melanization (Volz et al., 2006; Yassine et al., 2014; Nakhleh et al., 2017). Human coagulation factor X_a-activated A. gambiae CLIPB9_Xa and CLIPB10_Xa cleaved Manduca sexta PPO1 and PPO2 but yielded low levels of PO activity (An et al., 2011; Zhang et al., 2021). Silencing CLIPs C9, B1, B3, B4, B8–10, B14, B15, and B17 negatively impacted melanization of P. berghei ookinetes and oocysts (Volz et al., 2005 and 2006; Paskewitz et al., 2006; Zhang et al., 2016; Sousa et al., 2020). However, it is unclear how much functional redundancy of the 110 CLIPs has affected results of the RNAi screenings in A. gambiae. Neither is it known whether PPO activation by a PAP (e.g., CLIPB9) needs one or more CLIPAs as a cofactor to generate fully active PO. Since there are nine A. gambiae PPOs (Christophides et al., 2002; Kwon and Smith, 2019), it is unknown if a few CLIPs play distinct roles in the activation of individual PPOs.

Beginning to fill these knowledge gaps, we re-examined the findings of our phylogenetic analysis of the 247 CLIPs from five holometabolous insects and identified orthologs of M. sexta PAPs and their cofactors in A. gambiae (Cao and Jiang, 2018; Jin et al., 2022). Based on the transcriptome and proteome data (Cao et al., 2017; He et al., 2017), we selected four CLIPAs as candidates of CLIPB9 cofactors for expression in insect cells. The purified proCLIPAs were characterized and activated by M. sexta PAP3 or Factor X_a-treated A. gambiae proCLIPB9_Xa to test a possible cofactor role. A. gambiae PPO2 and PPO7 were produced in Escherichia coli for use as PAP substrates in the activation assay. Positive results were obtained, which supported the cofactor hypothesis. Since only CLIPA7 was identified in the previous RNAi screenings, the successful generation of PAP cofactors in vitro clearly showed that well-designed biochemical analysis investigations can directly reveal molecular mechanisms of melanization in mosquitoes.

2. Materials and methods

2.1. cDNA cloning and recombinant expression of A. gambiae proCLIPAs and proCLIPB9x_a

CLIPA4, A6, and A12 fragments were amplified from an adult A. gambiae cDNA sample using primer pairs j784-j785, j786-j787, and j792-j793 (Table S1), respectively. PCR products at the expected sizes were cloned into pGEM-T vector (Promega). After sequence validation, the NdeI-XhoI fragments were retrieved and inserted to the same sites in pMFFMH6 (Fig. S1), a vector derived from pMFH6 (Lu and Jiang, 2008). Two CLIPA7 fragments were amplified from the same cDNA pool using primer pairs j788-j789 and j790-j791 (Table S1). After T/A cloning and sequence confirmation, the NdeI-BspEI and BspEI-XhoI fragments were ligated and inserted to the NdeI and XhoI sites of pMFFMH6 to yield proCLIPA7Δ/pMFFMH6. This cloning strategy kept the clip domain (K²⁸C…CCP⁷²) and serine protease-like domain (R⁵⁵⁶ITGD…FRGWI⁸⁰⁴) but deleted a low complexity region (G⁷⁶EEDD…QNPLD⁵³⁶) including 152 Gly, 64 Pro, 40 Ala, and 68 Thr/Ser. The shortened proCLIPA7Δ (GIHDYKDDDDKHME²²DEEVIK…CCPYPE⁷⁵ K⁵³⁷TVSV…DSFYL⁸¹⁹EQKLISEEDLHHHHHH, 41,203 Da, pI: 5.25) comprises a cDNA-coded part (underlined) flanked by FLAG and myc-H₆ tags (Fig. S1). Each of the four plasmids was used to generate a baculovirus stock (1–2×10⁸ pfu/ml) to express proCLIPs A4, A6, A7Δ or A12 in Sf9 cells (Sumathipala and Jiang, 2010). The proCLIPs A4, A6 and A12 have the sequences of GIHDYKDDDDKHMQ²¹QPID…SSLFA⁴²²LEQKLISEEDLHHHHHH (47,539 Da, pI: 5.37), GIHDYKDDDDKHMD²⁵DLSL…DSYTP⁴²⁷LEQKLISEEDLHHHHHH (46,905 Da, pI: 5.52), and GIHDYKDDDDKHMQ²⁶TCEG…YYTPA⁴³⁹LEQKLISEEDLHHHHHH (42,146 Da, pI: 5.14), respectively. A CLIPB9 fragment was first amplified from the cDNA pool using primer pair j1254-j1255 (Table S1). After cloning and sequence validation, the proteolytic activation site (I¹⁴⁴GMR*IYG G¹⁵¹) was mutated to IDGR*IYGG (Factor X_a cleavage site) using primer pairs j1254-j199 and j198-j1255. The two PCR products were combined, denatured, annealed, and extended for use as a template in another round of PCR amplification using primer pair j1254-j1255. The product was T/A cloned, sequence confirmed, and inserted into NdeI-XhoI digested pMFFMH6. The resulting plasmid, proCLIPB9x_a/pMFFMH6, was used to prepare a baculovirus stock (1–2×10⁸ pfu/ml) to express proClipB9x_a (GIHDYKDDDDKHMQ²⁶QQQC…I¹⁴⁴DGR*IYGG¹⁵¹…RSNIK⁴⁰⁰LEQKLISEEDLHHHHHH, 44,641 Da, pI: 5.98).

2.2. Expression and purification of the recombinant proCLIPs A4, A6, A7Δ, A12, and B9_Xa

Sf9 cells (2.4×10⁶ per ml) in 300 ml of Sf-900^™ III serum-free medium (Thermo Fisher Scientific) were separately infected with a baculovirus stock at a multiplicity of infection of 5–10 and grown at 27 °C for 96 h with gentle agitation at 150 rpm. After centrifugation at 5,000×g for 20 min to remove cells, the supernatant was diluted with equal volume of 1 mM benzamidine in distilled water and pH of the mixture was adjusted to 6.4 using HCl. After mixing at 4 °C for 10 min, fine particles were removed by centrifugation at 8,000×g for 30 min. The supernatant was applied to a dextran sulfate-Sepharose column (20 ml) equilibrated with buffer A (10 mM potassium phosphate, pH 6.4). Following a washing step with 100 ml buffer A, bound proteins were eluted from the column with a linear gradient of 0–1.0 M NaCl in 250 ml of buffer A at a flow rate of 1.5 ml/min. Based on the result of SDS-PAGE and immunoblot blot analysis, the proCLIP fractions were pooled and loaded onto a 1 ml Ni²⁺-nitrilotriacetic acid agarose column. After washing with 15 ml of 50 mM sodium phosphate (pH 7.5), bound proteins were eluted with a gradient of 0–0.25 M imidazole in 25 ml of the same buffer. Fractions containing the proCLIPs were pooled, dialyzed against 20 mM Tris-HCl (pH 7.5), and concentrated on Amicon Ultra-30 centrifugal filter devices (Millipore). Aliquots of the proteins were rapidly frozen in liquid nitrogen prior to storage at −80 °C.

2.3. cDNA cloning and expression of A. gambiae PPO2 and PPO7 in E. coli

A colony of A. gambiae G3 strain was maintained in the laboratory as described before (He et al., 2017). Total RNA samples were prepared from mosquito eggs, larvae, pupae, and adults using TRIzol reagent (Thermo Fisher Scientific). cDNA was synthesized from the total RNA (1 μg each) in 1× iScript Reverse Transcription Supermix (Bio-Rad) at 42 °C for 60 min in a 10 μl reaction. PPO2 cDNA fragments were amplified from the cDNA samples using primer pairs j1204-j1205 and j1206-j1207, whereas PPO7 fragments were amplified using primer pairs j1216-j1217 and j1218-j1219 (Table S1). After T/A cloning and sequence validation, the NdeI-XhoI and XhoI-HindIII fragments of PPO2 and the NdeI-ClaI and ClaI-HindIII fragments of PPO7 were separately ligated with NdeI-HindIII digested pSFM, a modified pET28b (Chen et al., 2014). The resulting plasmids, PPO2/pSFM and PPO7/pSFM, were used to transform E. coli BL21 gold (DE3) for expression. Soluble PPO2 (M_r: 80,996 Da, pI: 5.86, HIDYKDDDDKH M¹TD…ART⁶⁸⁸KLLEQKLISEEDL) and PPO7 (M_r: 82,536 Da, pI: 6.43, HIDYKDDDDKH M¹AT…ERT⁶⁹⁶KLLEQKLISEEDL) were produced at 16 °C for 16 h in the presence of 0.5 mM isopropyl-β-D-thiogalactopyranoside and purified by following the double Ni-nitrilotriacetic acid procedure (Hu et al., 2016).

2.4. Activation and IEARase activity measurement of CLIPB9_Xa and PAP3

To produce active CLIPB9_Xa, 2.5 μg proCLIPB9_Xa was incubated for 3 h at 37°C with 0.8 μg Factor X_a (P8010, New England Biolabs) in 25 μl of cleavage buffer (20 mM Tris-HCl, pH 7.5, with 100 mM NaCl and 2 mM CaCl₂). As described previously (Jiang et al., 2003a), 2.5 μl of the mixture and controls (250 ng proCLIPB9_Xa and 80 ng Factor X_a) were separately reacted with 150 μl 25 μM acetyl-Ile-Glu-Ala-Arg-p-nitroanilide (A0180; Sigma) in 0.1 mM CaCl₂, 100 mM NaCl, 100 mM Tris-HCl, pH 7.5, at room temperature. One unit of the IEARpNa hydrolytic activity is defined as the enzyme amount leading to 0.001 ΔA₄₀₅/min. To produce active PAP3, 40 ng of PAP3 purified from M. sexta hemolymph (Jiang et al., 2003b) was incubated for 1 h at room temperature with proPAP3 (0.4 μg) (He et al., 2018) in 12 μl of buffer B (20 mM Tris-HCl, 5 mM CaCl₂, 0.001% Tween-20, pH 7.5). One tenth of the mixture and control (200 ng proPAP3) were used to test IEARpNa hydrolysis as described above. The same amount of the samples (250 ng proCLIPB9_Xa with buffer or 80 ng Factor X_a, 200 ng proPAP3 with buffer or 40 ng PAP3) were analyzed by 10% SDS-PAGE and immunoblotting to examine extents of the cleavages.

2.5. Cleavage of A. gambiae PPO2 and PPO7 by CLIPB9_Xa or PAP3 in the absence of CLIPAs

To detect possible cleavage activation, 0.5 μg of the purified PPOs were incubated with CLIPB9_Xa (100 ng) or PAP3 (40 ng) in buffer B for 1 h at 37 °C. The reaction mixtures and controls were treated with SDS sample buffer, separated by 7.5% SDS-PAGE, transferred to nitrocellulose membrane, and subjected to immunoblot analysis using diluted antisera to A. aegypti PPO1 and PPO3 as primary antibody, kind gifts from Dr. Zhen Zou at Institute of Zoology, Chinese Academy of Sciences. Immunoblot images were scanned at 300 pixels per inch with an HP Scanjet G4010 using HP Photosmart Premier 6.5 and converted to 8-bit format for rolling ball background subtraction under optimal conditions. Bands of the 78–79 kDa PPO and its 73–74 kDa cleavage product (PO) were quantified using ImageJ (Schneider et al., 2012) to calculate cleavage efficiency (CE) as PO / (PPO+PO).

2.6. Proteolytic activation and electrophoretic mobility changes of proCLIPs A4, A6, A7Δ and A12 after PAP3 treatment

To detect mobility changes of the CLIPAs caused by PAP3 cleavage, aliquots of a proCLIPA (200–300 ng) were incubated with M. sexta PAP3 (40 ng) in 20 μl buffer B for 1 h at 37 °C. The reaction mixtures, controls, and M_r markers were separated by 10% SDS or native PAGE, transferred onto membranes, and detected using diluted M. sexta SPH1 (for CLIPs A6, A7Δ and A12), M. sexta SPH2 (for CLIPA4), or hexahistidine antiserum as primary antibody and goat-anti-rabbit/mouse IgG conjugated to alkaline phosphatase (Bio-Rad) as secondary antibody, and a BCIP-NBT substrate kit (Bio-Rad) for color development. To follow the time courses of cleavage events, aliquots of each proCLIPA (200 ng) were incubated with different amounts of PAP3 (100, 50, 25, 12.5, 6.3, 3.1, 0 ng) for 1 h at 37 °C, followed by 10% SDS-PAGE and immunoblot analysis. Duplicate reactions and controls were run on 10% native PAGE for detecting the formation of high M_r complexes of the four CLIPAs using the antibodies as described above.

2.7. Cleavage activation and electrophoretic mobility changes of proCLIPs A4, A6, A7Δ, and A12 after CLIPB9_Xa treatment

Aliquots of the proCLIPs (220 ng A4, 300 ng A6, 300 ng A7Δ, and 200 ng A12) were also separately incubated with A. gambiae CLIPB9_Xa (100 ng) in 20 μl buffer B for 1 h at 37 °C. The reaction mixtures and controls were subjected to 10% SDS-PAGE and native PAGE to detect cleavage and high M_r complex formation by immunoblot analysis, respectively. Concentration-dependent processing of the four proCLIPAs by different amounts of CLIPB9_Xa was examined by SDS- and native-PAGE analysis, as described in Section 2.6.

2.8. Activation of A. gambiae PPO2 and PPO7 by CLIPB9_Xa and PAP3 in the presence of CLIPAs

PPO7 (0.5 μg) and PAP3 (40 ng) were incubated with two to four of proCLIPs A4, A6, A7Δ, and A12 (0.3 μg each) in buffer B on ice for 1 h. Similarly, PPO2 (0.5 μg) and CLIPB9_Xa (100 ng) were reacted with 2 to 4 of the proCLIPAs under identical conditions. The reaction and control mixtures were subjected to 7.5% SDS-PAGE and immunoblot analysis using A. aegypti PPO3 or PPO1 antibody as indicated. Densitometric analysis of the PPO cleavage efficiency was performed as described in Section 2.5. PO activity in another set of reactions was determined by adding 2.0 mM dopamine solution to the samples and monitoring A_470nm on a microplate reader (Jiang et al., 2003a). PO specific activity was calculated as: PO activity / (PPO added × CE).

2.9. Identification of cleavage sites in the CLIPAs using LC/MS/MS

The purified proCLIPs A4, A6, A7Δ, and A12 (5–10 μg) were separately incubated with 1.0 μg CLIPB9x_a or PAP3 for 1 h at 37 °C. The reaction mixtures and proCLIPAs were denatured in urea and digested with chymotrypsin, LysC, and V8 proteinase (Zhang et al., 2014). The resulting peptides were desalted using C18 affinity media, dried, and redissolved in mobile phase A (0.1% HCOOH in H₂O). The samples were loaded onto an Acclaim PepMap RSLC C18 column (Thermo Fisher) for data-dependent LC-MS/MS analysis as described previously (Cao et al., 2020). Each sample was subjected to the Acclaim column via a gradient of 0–35% mobile phase B (0.1% HCOOH, 80% AcCN, 20% H₂O) developed over 120 min as described before (Jin et al., 2022). The survey scans were followed by both HCD and CID collisional MS/MS events triggered by target parent ions, with scanning of collisional fragment at 15,000 resolutions in the Orbitrap sector.

To identify cleavage sites in CLIPAs and quantify the resulting peptides, a database combined A. gambiae protein database (PEST Peptides_AgamP4.2) and background proteins from M. sexta, human, and insect cell were constructed for search, peptide spectrum matches were reviewed in Byonic to detect peptides not cut by nonspecific processing enzymes, and the details of peak area calculation were described previously (Jin et al, 2022).

To test the necessity of a cofactor for PPO activation in A. gambiae, we examined the result of phylogenetic analysis of the 247 CLIPs from five holometabolous insects (Cao and Jiang, 2018) and identified orthologs of M. sexta PAPs and their cofactors (Jin et al., 2022). A. gambiae CLIPs B9 and B10 are orthologs of M. sexta PAP1; CLIPs A5–7, A12–14, A26, A31, A32, E6 and E7 are 11:4 orthologous to M. sexta SPHIs (i.e., 1a, 1b, 4, 101); CLIPA4 is the ortholog of M. sexta SPHII (i.e., 2). High M_r complexes of SPHI-II (1a-2, 1b-2, 4–2, and 101–2) are cofactors of PPO activation by PAP3 in M. sexta. Likely due to positive selection, newly evolved SPH1b and SPH101 are more favorable substrates of PAP3 and better SPH2 partners than SPH1a or SPH4 (Jin et al., 2022). Based on the transcriptome and proteome data of A. gambiae (Cao et al., 2017; He et al., 2017), we selected 4 A. gambiae CLIPs (A4, A6, A7 and A12) as candidates of CLIPB9 cofactors for expression in insect cells. After processing with Factor X_a-treated CLIPB9_Xa or M. sexta PAP3, the purified CLIPAs formed complexes with different association states. CLIPB9_Xa or PAP3 alone cleaved A. gambiae PPO2 and PPO7 but yielded low PO activities. However, when one of the three CLIP pairs (CLIPs A4-A6, A4-A7Δ, and A4-A12) was present, much higher PO activities were generated in the reaction, suggesting that an SPHI-II complex is needed for PPO activation by a PAP in A. gambiae. Such auxiliary factors are now available for exploring mechanisms of melanization in vitro. Except for the conflicting data on CLIPA7, roles of CLIPA6 and CLIPA12 as partners of CLIPA4 were not discovered in the previous RNAi screenings in A. gambiae. Therefore, we demonstrated in this study that direct biochemical investigations can sometimes provide insights into molecular mechanisms of melanization in mosquitoes.

3. Results

3.1. Rationale for testing A. gambiae CLIPs A4, A6, A7Δ, and A12 as cofactor components for PPO activation

Analysis of the gene sequences, mRNA levels, and protein amounts provided justification for the selection of CLIPAs to investigate a possible role as PAP cofactor. We first re-examined the result of phylogenetic analysis of the 247 CLIPs from five holometabolous insects (Cao and Jiang, 2018) and identified orthologs of M. sexta PAPs and their cofactors (Fig. 1). A. gambiae CLIPs B8–10 are orthologous to M. sexta PAPs; CLIPs A5–7, A12–14, A26, A31, A32, E6, and E7 (E6 and E7: clade A members) are orthologous to M. sexta SPHIs (i.e., SPHs 1a, 1b, 4, and 101); CLIPA4 is orthologous to M. sexta SPH2, an SPHII group member (Jin et al., 2022). The CLIPs A4, A6, A7, A12, and A14 mRNA levels in larval, pupal, and adult stages were relatively high, with average/maximal FPKM values of 215/722, 340/1139, 200/1041, 83/227, and 215/722, respectively (Cao et al., 2017). In comparison, FPKM values of CLIPs A5, A13, A26, A31, A32, E6, and E7 mRNA levels were much lower, at 2.5/30.4, 18.5/11.7, 3.1/29.2, 2.6/65.0, 0.3/4.0, 9.4/65, and 0.41/4.0, respectively. In agreement with the transcriptome data, relative abundances of CLIPs A4, A6, A7, A12, and A14 proteins were 1140, 4565, 1548, 157, and 2296 in hemolymph of naïve larvae, and became 1127, 4961, 944, 117, 163, and 2343 after larvae were infected with E. coli (He et al., 2017). The other seven clip-domain SPHs were not detected in the LC-MS/MS analysis. Examination of the hemolymph samples from naïve and infected adult mosquitoes showed a similar pattern in the SPHI-II group (Jin et al., unpublished data). When we initiated this study, knockdown of A. gambiae CLIPA14 expression led to a potent melanization response against Plasmodium berghei in a TEP1-dependent manner (Nakhleh et al., 2017), suggesting that CLIPA14 is negative regulator of PPO activation. Therefore, we chose to investigate CLIPs A4, A6, A7, and A12 as candidates of cofactors for A. gambiae PPO activation. A more recent report of the cofactor role of A. aegypti CLIPA14 (Ji et al, 2022) intrigued us to explore the apparent difference in the two mosquito species and report results of the in vitro experiments in the near future. Besides, CLIPA7 contains a low complexity region between its clip and protease-like domains. After considering its proteolysis and glycosylation that may complicate the expression and purification, we decided to produce a mutant form of proCLIPA7 (i.e., A7Δ). The deleted region is 422 residues long, rich in Gly (150), Pro (63), Ala (38), Thr (35), Gln (33), Ser (26), and Val (26).

Fig. 1.

Phylogenetic relationships among the SPHIs and SPHIIs from five important insects. This monophyletic branch is directly taken from a phylogenetic tree of CLIPAs from A. gambiae (Ag, cyan), Apis mellifera (Am, pink), D. melanogaster (Dm, red), M. sexta (Ms, green), and Tribolium castaneum (Tc, blue) (Cao and Jiang, 2018). Minor modifications are made to better distinguish the two SPH subgroups. As marked by maroon asterisks, functional data are available for supporting the pairing of SPHI and SPHII in D. melanogaster (cSPH35-cSPH242) and M. sexta (SPHs 1a-2, 1b-2, 4–2, 101–4). Transcriptome and proteome data has led us to test and confirm the formation of cofactors (CLIPs A6-A4, A7Δ-A4, and A12-A4) in this study. As indicated by “?”, CLIPs A14-A4 and more SPHI-SPHII pairs are to be tested in a separate study.

3.2. Recombinant production of A. gambiae proCLIPs A4, A6, A7Δ, A12, B9x_a, PPO2, and PPO7

We cloned cDNA of the four CLIPAs, verified their sequences, and subcloned them into pMFFMH6 for in vivo transposition. Similarly, A. gambiae CLIPB9 cDNA was cloned and, for in vitro activation by bovine coagulation factor X_a, the cleavage site of B9 (IGMR*IYGG) was mutated to IDGR*IYGG. The bacmid DNA of CLIPs A4, A6, A7Δ, A12, and B9x_a were isolated for transfecting Sf9 cells to produce high titer viral stocks through serial infections. Led by the honeybee melittin signal peptide, the precursors were secreted into media with a hexahistidine tag fused to the carboxyl-terminus. The proCLIPs A4 (0.5 mg), A6 (1.0 mg), A7Δ (0.3 mg), A12 (0.2 mg), and B9x_a (0.5 mg) were purified from 100 ml of the conditioned media. They migrated to 53, 60, 42, 53, and 53 kDa positions on a 10% SDS-PAGE gel (Fig. 2), larger than their calculated M_r’s (47,539, 46,905, 41,203, 42,146, 44,641 Da), respectively. The proCLIPs A4, A6, A12, and B9 are likely glycosylated. As judged based on their staining patterns, these proteins were near homogeneous. Minute amounts of CLIPA7Δ (34 kDa) and CLIPA12 (42 and 31 kDa) fragments were detected by antibody against the hexahistidine tag.

Fig. 2.

SDS-PAGE and immunoblot analysis of recombinant A. gambiae proCLIPs A4, A6, A7Δ, A12 and B9x_a, PPO2, and PPO7. The purified proteins were resolved by 10% SDS-PAGE followed by Coomassie brilliant blue staining (A, 1 μg CLIPs, 2 μg PPOs) or immunoblot analysis (B, 1 μg CLIPs, 0.5 μg PPOs) using 1:1000 diluted antisera against hexahistidine tag, A. aegypti PPO1 and PPO3. A. gambiae PPO2 and PPO7 were cross-reacted by the PPO antibodies. Positions and sizes (in kDa) of the prestained M_r standards are indicated on the left, with the 75 kDa marker highlighted red.

We amplified A. gambiae PPO2 and PPO7 cDNA fragments, confirmed their sequences in pGEM-T, and then subcloned them into pSFM, a modified pET28b encoding H₆-SUMO, FLAG, and myc) (Chen et al., 2014). At 16 °C, the two proenzymes were produced in E. coli as soluble proteins and purified by affinity chromatography on Ni-NTA agarose columns. We isolated 2.2 mg PPO2 and 3.3 mg PPO7 from one liter of bacterial culture. They migrated as single band to the 79 and 78 kDa positions on SDS-PAGE (Fig. 2), which were close to the theoretical M_r’s of 80,996 and 82,536 Da, respectively.

3.3. Proteolytic cleavage of A. gambiae PPO2 and PPO7 by CLIPB9x_a and M. sexta PAP3

A. gambiae CLIPB9 is a PPO activating protease, which cleaved M. sexta PPOs at Arg⁵¹ but yielded low PO activity (An et al., 2011). To test if it can activate recombinant PPO2 or PPO7, we first produced proCLIPB9_Xa in the baculovirus insect cell expression system and incubated the purified proenzyme with bovine Factor Xa at 37 °C to generate CLIPB9_Xa. Three hours later, only half of the 53 kDa protein was cleaved and shifted to a 35 kDa position on reducing SDS-PAGE gel (Fig. 3A, left), corresponding to the catalytic domain recognized by antibody to 6×His tag at the C-terminus. Consistent with the cleavage activation, a significant increase in amidase activity (3.2 U) was detected using a chromogenic substrate, IEARpNA or acetyl-Ile-Glu-Ala-Arg-p-nitroanilide (Fig. 3A, right). We then tested if CLIPB9_Xa could activate the purified PPOs. After 1 h at 37 °C, the IEARase almost completely cleaved PPO2, likely between Arg⁵⁰ and Phe⁵¹, to form 74 kDa PO2 (Fig. 3C, left). Under the same conditions, CLIPB9_Xa and Factor X_a together cleaved about 1/5 of PPO7 to form 73 kDa PO7 and Factor X_a alone cut PPO7 likely at the same location (i.e., Arg⁵²-Phe⁵³) (Fig. 3D. left).

Fig. 3.

Proteolytic activation A. gambiae CLIPB9x_a, M. sexta PAP3, and their putative substrates PPO2 and PPO7. (A) ProCLIPB9xa (2.5 μg) was incubated with Factor Xa (0.8 μg) in 25 μl cleavage buffer at 37 °C for 3 h. One tenth of the mixture and control of proCLIPB9x_a only were subjected to 10% SDS-PAGE and immunoblot analysis (left) using 1:1000 diluted antiserum against the hexahistidine tag as primary antibody. Positions of proCLIPB9x_a (●) and its catalytic domain (○) are indicated on the right. The mixture (2.5 μl) and controls (250 ng proCLIPB9x_a and 80 ng Factor X_a) were incubated with 150 μl 25 μM IEARpNA in 0.1 mM CaCl₂, 100 mM NaCl, 100 mM Tris-HCl, pH 7.5, at room temperature for 5 min. Absorbance change at 405 nm was monitored in a kinetic mode to measure the amidase activity. One unit of activity is defined as the enzyme amount causing 0.001 OD increase per minute. The activities (mean ± SEM, n = 3) were plotted in a bar graph (right). (B) Similarly, proPAP3 (2.0 μg) was incubated with PAP3 (0.4 μg) in 100 μl buffer B at 37 °C for 1 h. One tenth of the mixture and control of proPAP3 (0.2 μg) only were subjected to 10% SDS-PAGE and immunoblot analysis (left) as described above in panel A. IEARpNA hydrolysis was assayed using the mixture (10 μl) or 0.2 μg proPAP3 and activity data (mean ± SEM, n = 3) were plotted (right). Significance of the activity differences was analyzed by Student’s t-test, with p < 0.01 and 0.001 labelled by ** and ***, respectively. Limited proteolysis of A. gambiae PPO2 (C) and PPO7 (D) by CLIPB9x_a and PAP3. The purified PPO2 (500 ng/μl, 1 μl) and PPO7 (500 ng/μl, 1 μl) from E. coli were separately incubated with A. gambiae CLIPB9x_a (100 ng/μl, 1 μl) or M. sexta PAP3 (40 ng/μl, 1 μl) in 10 μl buffer B at 37 °C for 1 h. As described in Section 2.4, the mosquito protease was produced by preincubating proCLIPB9x_a with Factor X_a, and PAP3 was activated from its proenzymes. The reaction mixtures and controls were heated at 95 °C in 1× SDS sample buffer, separated by 7.5% or 10% SDS-PAGE, and subjected to immunoblot analysis using 1:1000 diluted antisera against A. aegypti PPO1 and PPO3 (Wang et al., 2017), which crossreacted with A. gambiae PPO2 and PPO7, respectively. In panels C and D, asterisk (*) marks the PPOs; red arrowhead indicates the cleavage product around 70 kDa. Numbers and short bars on the left indicate sizes and positions of the prestained M_r standards, with the 75 kDa marker highlighted red.

In M. sexta, PAP3 activated proPAP3, proSPHI-II, and PPOs (Wang et al., 2014; Jin et al., 2022). PAP3 also generated an active cofactor of cSPH35 and cSPH242 for PPO1 activation by MP2 in Drosophila (Jin et al., 2023). Thus, we incubated M. sexta PAP3 with the recombinant proPAP3 to generate more active PAP3 (Fig. 3B), which was tested for a possible role as an activating protease of A. gambiae PPOs and proCLIPAs. PAP3 did not cleave PPO2 (Fig. 3C, right) but cut PPO7 at a low level (Fig. 3D, right). While PPO2 and PPO7 were cleaved by CLIPB9_Xa or PAP3 to different extents probably at the predicted activation site, no significant increase in PO activity was observed in the reactions (see below), suggesting that a cofactor is required in the PPO activation reactions.

3.4. Proteolysis of proCLIPA4, A6, A7Δ, and A12 by A. gambiae CLIPB9x_a or M. sexta PAP3

We first tested whether the activated CLIPB9_Xa could cleave proCLIPs A4, A6, A7Δ, and A12 to form high M_r complexes. As shown in Fig. 4, a minor part of proCLIPA4 became 10 kDa smaller on reducing SDS-PAGE, and the mobility decrease on native PAGE suggested a level of association of the cut form. All of the proCLIPA6 was cleaved by CLIPB9_Xa, yielding 42 and 31 kDa C-terminal fragments on SDS-PAGE and a high M_r smear extending from the stacking gel on native PAGE. Nearly 4/5 of proCLIPA7Δ was processed by CLIPB9x_a, generating 36 and 31 kDa bands on SDS gel and a large smear on native gel. The cleavage of proCLIPA12 was complete, producing a 31 kDa band on SDS-PAGE but association was less than CLAPA6 or CLAPA7Δ (Fig. 4B). These results were confirmed by PAGE and immunoblot analysis using diluted antisera to M. sexta SPH1a and SPH2 (Fig. S2). M. sexta SPH2 antibody cross-reacted with A. gambiae CLIPA4 (SPHII); SPH1a antibody recognized CLIPs A6, A7Δ, and A12 (SPHIs).

Fig. 4.

Immunoblot analysis of the CLIPB9x_a cleavage products from A. gambiae CLIPs A4, A6, A7Δ and A12 precursors after reducing SDS-PAGE (A) and native PAGE (B). As described in Fig. 3 legend, the proCLIPAs were separately incubated with CLIPB9x_a (100 ng/μl, 1 μl) and 18 μl buffer B. Following 10% PAGE, the reaction mixture and controls were analyzed using 1:1000 diluted antibody against the hexahistidine tag as primary antibody. Positions of proCLIPB9xa (●), its catalytic domain (○), proCLIPAs (*), and their C-terminal cleavage products (red arrowheads) are indicated in panel A. In panel B, the dashed line divides the stacking and separating gels. The smeared bands of CLIPB9x_a, proCLIPAs, and their cleavage products are marked by blue, green, and red vertical bars, respectively.

In general, similar changes were observed in the cleavage reactions containing M. sexta PAP3 (Fig. 5). Unlike CLIPB9x_a, PAP3 cleaved all proCLIPA4 to form a major band 10 kDa smaller and two minor bands at 42 and 36 kDa. The 60 kDa proCLIPA6 was all converted to 36 and 31 kDa C-terminal fragments. Similarly, a minor 36 kDa band and a major 31 kDa band were detected in the reactions of PAP3 with proCLIPA7Δ (42 kDa) and proCLIPA12 (53 kDa). The same results were obtained using M. sexta SPH1a and SPH2 antibodies (Fig. S3). On native PAGE (Fig. 5B), the cleavage products displayed association patterns similar to those after CLIPB9x_a treatment (Fig. 4B).

Fig. 5.

Immunoblot analysis of the PAP3 cleavage products from A. gambiae proCLIPs A4, A6, A7Δ and A12 after reducing SDS-PAGE (A) and native PAGE (B). The purified proCLIPs A4 (220 ng ng/μl, 1 μl), A6 (300 ng/μl, 1 μl), A7Δ (300 ng/μl, 1 μl), and A12 (200 ng/μl, 1 μl) were incubated with M. sexta PAP3 (40 ng/μl, 1 μl) and 18 μl buffer B at 37 °C for 1 h. The reaction mixtures were treated with 1× SDS sample buffer at 95 °C for 5 min or the same buffer lacking SDS and DTT at 25 °C for 5 min prior to electrophoresis on 10% gels. Immunoblot analysis was performed using 1:1000 diluted antibody against the hexahistidine tag. Positions of proPAP3 (●), its catalytic domain (○), proCLIPAs (*), and their C-terminal cleavage products (red arrowheads) are indicated in panel A. Positions and sizes of the prestained M_r standards are indicated on the left, with the 75 kDa marker highlighted red. In panel B, the dashed line divides the stacking and separating gels. The smeared bands of PAP3, proCLIPAs, and their cleavage products are marked by blue, green, and red vertical bars, respectively.

Concentration-dependent activation of the CLIPAs by the PAPs provided insights into orders of the cleavage events. The 10 kDa decrease in CLIPA4 suggested that a partial cleavage by CLIPB9x_a occurred first in the clip domain (Fig. 6A) and that a complete cleavage at the first site by PAP3 was followed by cleavage at the second site yielding the 42 kDa band (Fig. 7A). The distinction between the two cleavage events was less obvious in CLIPA6. The 42 kDa CLIPA6 was first generated by CLIPB9x_a, followed by the appearance of 31 kDa band at the next higher level of B9x_a (Fig. 6A). At the two lowest levels, PAP3 completely converted proCLIPA6 to the 42 kDa band, most of which was then cleaved at site-2 to form the 31 kDa band (Fig. 7A). A small part of proCLIPA7Δ was also cleaved by CLIPB9x_a at two sites and, since the intermediate cut at site-1 can be further cleaved at site-2 (Fig. 6A), appearance of the 31 kDa weaker band at the next higher level of B9x_a indicated that the upstream site-1 is more accessible than site-2. Complete activation of proCLIPA7Δ by PAP3 followed the same order of cleavage events (Fig. 7A). Like the other three CLIPAs, proCLIPA12 was cleaved by CLIPB9x_a and PAP3, displaying time courses of the partial and complete processing, respectively. The B9x_a cleavage induced CLIPA association in the order of A6 > A7Δ > A12 > A4 (Fig. 6B). More complete cleavage of the CLIPAs by PAP3 may have also led to smears into stacking gel in the same order (Fig. 7B). Due to PAP3 self-association, signals of the CLIPs A4 and A12 assemblies (lane 3) cannot be clearly distinguished from the high M_r smear of PAP3 (lane 1).

Fig. 6.

Concentration-dependent CLIPB9x_a processing of A. gambiae proCLIPs A4, A6, A7Δ and A12 analyzed by 10% SDS (A) and native (B) PAGE followed by immunoblot analysis. Aliquots of the purified proCLIPAs (200 ng) were incubated with 100 (lane 3), 50 (lane 4), 25 (lane 5), 12.5 (lane 6), 6.25 (lane 7), 3.1 (lane 8), and 0 (lane 2) ng of CLIPB9x_a (100 ng/μl, 1 μl) in 10 μl buffer B at 37 °C for 1 h. The mixtures and CLIPB9x_a control (100 ng, lane 1) were subjected to 10% PAGE and immunoblot analysis, as described in the legend to Fig. 3. Positions of proCLIPB9xa (●), its catalytic domain (○), proCLIPAs (*), and their C-terminal cleavage products (red arrowheads) are indicated in panel A. In panel B, the smeared bands of CLIPB9x_a, proCLIPAs, and their products are marked by blue, green, and red vertical bars, respectively.

Fig. 7.

Concentration-dependent PAP3 processing of A. gambiae proCLIPs A4, A6, A7Δ and A12 analyzed by 10% SDS (A) and native (B) PAGE followed by immunoblot analysis. Aliquots of the purified proCLIPAs (200 ng) were incubated with 100 (lane 3), 50 (lane 4), 25 (lane 5), 12.5 (lane 6), 6.25 (lane 7), 3.1 (lane 8), and 0 (lane 2) ng of PAP3 in 10 μl buffer B at 37 °C for 1 h. The mixtures and PAP3 control (100 ng, lane 1) were subjected to 10% PAGE and immunoblot analysis, as described in the legend to Fig. 3. Positions of proPAP3 (●), its catalytic domain (○), proCLIPAs (*), and their C-terminal cleavage products (red arrowheads) are indicated in panel A. In panel B, the smeared bands of PAP3, proCLIPAs, and their cleavage products are marked by blue, green, and red vertical bars, respectively.

3.5. Identification of the cleavage sites in A. gambiae CLIPs A4, A6, A7Δ, and A12

Since M. sexta PAP3 cut proSPH1b first at R⁸²*F⁸³ and then at R¹³³*T¹³⁴ and proSPH2 only at R⁷⁷*F⁷⁸ and since A. gambiae CLIPB9 cut M. sexta PPO at R⁵¹*F⁵² (Yu et al., 2003; Wang and Jiang, 2004; Jin et al., 2022; An et al., 2011), we predicted the PAPs cleave proCLIPAs next to R or K. To ensure peptides released from a second cut by chymotrypsin or Lys-C can be discerned from the trypsin-like PAPs, we preincubated the proCLIPAs with CLIPB9x_a or PAP3, treated the mixtures with chymotrypsin or Lys-C, and then performed LC-MS/MS analysis. As shown in Table 1, both PAPs cut proCLIPA4 once at K⁹⁷*F⁹⁸ and released F⁸⁴TCQPPPEFAEQNK⁹⁷ after chymotrypsin treatment. The peak area of the peptide from PAP3 complete proteolysis was 10 times as high as that from B9x_a incomplete cleavage (Table 1). The PAP cleavage released a 10,123 Da N-terminal fragment equal to the size difference between CLIPA4 and its precursor (Figs. 4–7). CLIPB9xa and PAP3 both cleaved proCLIPA6 first at R⁹⁴*F⁹⁵ and then at R¹⁶²*I¹⁶³, and chymotrypsin released peptides F⁹⁵SDDNPCVDY¹⁰⁴ and I¹⁶³TGSKNSEAEYGEFPW¹⁷⁸ (Fig. S4). While the peptides were detected in the controls of proCLIPA6-chymotrypsin, their levels were 1/10² to 1/10⁴ of those in the proCLIPA6-B9x_a/PAP3-chymotrypsin reactions (Table 1). PAP3 cleaved proCLIPA7Δ first at R⁵⁶*F⁵⁷ and then at R⁵⁵⁶*I⁵⁵⁷, and Lys-C treatment released the two peptides following K³⁶ and K⁵⁴⁴ (Fig. S4). Similarly, PAP3 cleaved proCLIPA12 first at R⁵⁹*I⁶⁰ and then at R¹¹³*I¹¹⁴. In summary, the first cleavage sites are R*F (2), K*F and R*I, all between Cys-3 and Cys-4 of the clip domains. The second cutting sites are all R*I, located at the junction right before the protease-like domains. These two sites perfectly corresponded with the proteolytic activation sites in their orthologs from M. sexta and D. melanogaster (Jin et al., 2022 and 2023), indicating that CLIPB9x_a- or PAP3-treated proCLIPAs may act as cofactors for PPO2 and PPO7 activation in A. gambiae.

Table 1.

Determination of the PAP3 and CLIPB9x_a cleavage sites in A. gambiae CLIPAs by LC-MS/MS analysis

sample	1st cleavage site	m/z	time	peak area	2nd cleavage site	m/z	time	peak area
proCLIPA4	(Y)FTCQPPPEFAEQNK*(F)	846.9	64	n.d.	-	-	-	n.d.
proCLIPA4 + CLIPB9xa	(Y)FTCQPPPEFAEQNK*(F)	846.9	64	1.50×108	-	-	-	n.d.
proCLIPA4 + PAP3	(Y)FTCQPPPEFAEQNK*(F)	846.9	64	1.55×109	-	-	-	n.d.
proCLIPA6	(R)*FSDDNPCVDY (L)	616.2	76	6.99×105	(R)*ITGSKNSEAEYGEFPW(M)	923.9	86	1.65×106
proCLIPA6 + CLIPB9xa	(R)*FSDDNPCVDY (L)	616.2	76	2.71×108	(R)*ITGSKNSEAEYGEFPW(M)	923.9	86	1.84×108
proCLIPA6	(R)*FSDDNPCVDY (L)	616.2	76	1.40×104	(R)*ITGSKNSEAEYGEFPW(M)	923.9	86	3.92×106
proCLIPA6 + PAP3	(R)*FSDDNPCVDY (L)	616.2	76	1.85×108	(R)*ITGSKNSEAEYGEFPW(M)	923.9	86	1.17×109
proCLIPA7Δ	(K)LHLCPNGELNTDGANIIDIR* (F)	745.7	89	1.93×108	(K)CGLRNVDGV GFR* (I)	450.6	54	3.77×108
proCLIPA7Δ + PAP3	(K)LHLCPNGELNTDGANIIDIR* (F)	745.7	89	1.25×1010	(K)CGLRNVDGV GFR* (I)	450.6	54	8.13×1010
proCLIPA12	(L)TAEGEDDDAPAPEVDLR* (I)	900.4	63	7.00×108	(R)*IGAGKVEEAEFGEFPW(S)	883.4	106	7.80×108
proCLIPA12 + PAP3	(L)TAEGEDDDAPAPEVDLR* (I)	900.4	63	1.95×109	(R)*IGAGKVEEAEFGEFPW(S)	883.4	106	3.71×109

A list of peptides identified as products of PAP3 or CLIPB9x_a processing and further cleavage with chymotrypsin (next to F, Y, W) or Lys-C (K). n.d., not detected; “-”, no 2^nd cleavage site.

3.6. Enhancement of A. gambiae PPO activation by CLIPB9 and PAP3 in the presence of CLIPAs

A. gambiae CLIPB9x_a cleaved 0.31 μg PPO2 and generated 7.8 U of PO2 (Fig. 8, lane 8, 25 U/μg), inclusion of proCLIPA4 produced slightly lower PO2 at 7.6 U, and adding a mixture of proCLIPs A6, A7Δ, and A12 (i.e., SPHIs) did not enhance PPO2 cleavage and yielded lower PO2 (5.0 U, lane 10, 15 U/μg). However, further addition of proCLIPA4 increased PPO2 cleavage to 0.45 μg and that yielded 46 U of PO2 (lane 11) or a 4.1–6.9-fold increase in specific activity to 102 U/μg. If only increased PPO cleavage and PO activity (lane 11 minus lane 8) were considered, the specific PO2 activity is estimated to be 269 U/μg. Apparently, the SPHII formed complexes with one or more of the SPHIs to greatly enhance PPO2 activation. Further tests of the A4-A6, A4-A7Δ, and A4-A12 pairs (Fig. 8C, lanes 12–14) yielded 34, 25, and 31 U of 0.40, 0.46, and 0.39 μg PO2, respectively, and the decreases were statistically insignificant when compared with A4-A6-A7Δ-A12 (lane 11). The calculated specific PO2 activity were 85, 55, 79 U/μg and, if only added PPO2 cleavage and PO2 activity were considered, 289, 116, and 281 U/μg when A4-A6, A-A7, and A4-A12 were present, respectively, much higher than 25 and 15 U/μg of the controls.

Fig. 8.

Immunoblot analysis of A. gambiae PPO2 activation by CLIPB9x_a in the presence of CLIPs A4, A6, A7Δ, and/or A12. (A) A. gambiae proCLIPB9x_a (1 μl, 100 ng/μl) was incubated with Factor Xa (1 μl, 40 ng/μl) at 37 °C for 3 h. The purified PPO2 (1 μl, 0.47 μg/μl) was added to the reaction and incubated on ice for 60 min with proCLIPAs (220 ng of A4, 300 ng of A6, A7Δ, and/or A12) in 12 μl buffer B and subjected to 7.5% SDS-PAGE and immunoblot analysis using 1:1000 diluted antiserum against A. aegypti PPO1. Same amounts of the proteins were mixed, incubated in 20 μl buffer B on ice for 60 min prior to PO activity assay. Asterisk (*) marks the PPO2; red arrowheads show the cleavage products around 75 kDa. The proCLIPs A4, A6, A7Δ, and A12 bands somehow recognized by the PPO1 antibody are indicated by blue, black, brown, and green dots (●), respectively. In panels B and C, PO activities (mean ± SEM, n = 3) of three biological replicates were plotted as bar graphs. Statistical significance of the activity difference was analyzed by one-way ANOVA. Bars with p <0.05 are labeled with different letters.

We then tested if PAP3 cleavage of A. gambiae PPO7 along with the cofactors can lead to PO7 activity boosts (Fig. 9, A and B). PO activity was 3.4 U for PPO7 only (lane 6) and 3.9 U for PPO7-PAP3 (lane 8). Compared with the controls (lanes 6–8, 3.4–6.2 U), there was a small but significant increase in PO7 activity (lane 9, 7.9 U), although PPO7 cleavage was below the limit of detection. When all the components were present (lane 10), about 0.07 μg of PPO7 was converted to PO7 at an estimated specific activity of 488 U/μg (33.2 U PO). After A4-A6-A7Δ-A12 (lane 10’, 30.7 U) was substituted by the A4-A6, A4-A7Δ, and A4-A12 pairs, PAP3 generated 30.8, 32.7, and 12.3 U of PO7 (lanes 11–13). Consistent with the lower PO activity (lane 13), we detected lower levels of the 73 kDa cleavage product (18.9. 24.4, 10.9 ng) by densitometry and, since the PO7 amounts were low, the corresponding PO7 specific activities were calculated to be 1,630, 1,339 and 1,131 U/μg in the reactions containing A4-A6, A4-A7Δ and A4-A12, respectively. These data clearly confirmed the cofactor role of A4-A6, A4-A7Δ, and A4-A12 in PPO7 activation.

Fig. 9.

Immunoblot analysis of A. gambiae PPO7 activation by M. sexta PAP3 in the presence of CLIPs A4, A6, A7Δ and/or A12. (A) A. gambiae PPO7 (1 μl, 0.49 μg/μl) and M. sexta PAP3 (1 μl, 40 ng/μl) were incubated on ice for 60 min with proCLIPAs (220 ng of A4, 300 ng of A6, A7Δ, and/or A12) in 12 μl buffer B and subjected to 7.5% SDS-PAGE and immunoblot analysis using 1:1000 diluted antiserum against A. aegypti PPO3. Same amounts of the proteins were mixed, incubated in 20 μl buffer B on ice for 60 min before PO activities were assayed by adding 150 μl of 2.0 mM dopamine in 50 mM sodium phosphate buffer (pH 6.5) to the microplate wells (Jiang et al., 2003a). Asterisk (*) marks the PPO7; “○” indicates the cleavage product around 70 kDa. Numbers and short bars on the left indicate sizes and positions of the prestained M_r standards, with the 75 kDa marker highlighted red. In panels B and C, PO activities (mean ± SEM, n = 3) of the three biological replicates were plotted as bar graphs. Statistical significance of the activity difference was analyzed by one-way ANOVA. Bars with p <0.05 are labeled with different letters.

4. Discussion

4.1. In vitro production of cofactors for PPO activation reactions, a breakthrough in the research on melanization in mosquitoes and other insects

Melanization is a mosquito resistance mechanism against pathogens and parasites (Collins et al., 1986; Molina-Cruz et al., 2016; Nakhleh et al., 2017). While early research was focused on host-parasite interactions between A. gambiae and P. berghei, for instance, much progress has been made lately regarding melanization of pathogens (Povelones et al., 2013; El Moussawi et al., 2019; Sousa et al., 2020; Dekmak et al., 2021; Zakhia and Osta, 2022). RNAi-based gene knockdowns indicated that a set of CLIP genes, including A2, A8, A14, A28, A30, B4, B8–10, B17, and C9, were involved in TEP1-opsonized melanization. While the serial RNAi screenings yielded indirect evidence on relative orders of CLIPs (e.g., A8···>A28···>A30) in the cascade pathways for TEP1, Toll, and/or PPO activation, A. gambiae CLIPs B4, B8, B9, and B10 were produced in insect cells as zymogens mutated for Factor X_a cleavage activation for biochemical analysis (An et al., 2011; Zhang et al., 2016, 2021, 2023). CLIPB4x_a directly activated proCLIPB8 and CLIPB8 led to proCLIPB9 activation. CLIPs B4x_a, B9x_a, and B10x_a cleaved M. sexta PPOs likely at the correct site (R⁵¹*F⁵²) but generated low PO activities.

In this study, we demonstrated that A. gambiae CLIPB9x_a cleaved proCLIPs A4, A6, A7Δ, and A12 to form cofactors for PPO2 activation by CLIPB9x_a (Fig. 8) and that M. sexta PAP3 more efficiently processed the four proCLIPAs, yielding the cofactors for PPO7 activation by PAP3 (Fig. 9). While CLIPB8 or CLIPB10 could be a better CLIPA activator than CLIPB9 or PAP3, M. sexta PAP3 played a key role in the elucidation of PPO activation mechanisms (Jin et al., 2022; this work). Replacing M. sexta PPOs purified from larval hemolymph, A. gambiae PPO2 and PPO7 expressed in E. coli were used effectively as substrates of the two recombinant PAPs to generate highly active POs (85, 55, 79 U/μg for PO2 and 1,630, 1,339, 1,131 U//μg for PO7) in the presence of high M_r complexes of SPHI-II. Because the proCLIPAs are available in substantial amounts for M. sexta PAP3 activation, it is now feasible to produce the cofactors and study A. gambiae PPO activation mechanism in vitro. Not long ago, we had to rely on the SPHI-II complex purified from hemolymph of M. sexta bar-stage pharate pupae to achieve a high level of PPO activation (Wang et al., 2020). The successful in vitro production of eight SPHI-II complexes (i.e., M. sexta SPHs 1a-2, 1b-2, 4–2, and 101–2; Drosophila cSPH35-cSPH242; A. gambiae CLIPs A4-A6, A4-A7Δ, A4-A12) (Jin et al., 2022 and 2023; this work) clearly showed that the biochemical approach is now practical for studying a similar process in small insects (e.g., mosquitoes).

4.2. Regulation of PPO activation by these and other SPHI-II complexes in A. gambiae

It is evident that sophisticated control mechanisms exist in A. gambiae to affect melanization positively or negatively depending on host-parasite combinations (Barillas-Mury, 2007). Like the difference in PPO gene numbers between mosquitoes (9–10 per species) and most other insects (1–3) (Lu et al., 2014), the size change in the SPHI group, 11 in A. gambiae and 1–4 in most lepidopterans (Jin et al., 2022), is also phenomenal, highlighting the role of melanization and its regulation in the mosquito. Ten of the eleven SPHI genes reside in a cluster of 15 CLIPA genes on chromosome (Chr) 3L (Cao et al., 2017). CLIPA4, the first one in this gene cluster, is located in the phylogenetic tree next to the junction of two SPHI branches, A12-A13-(A5/A14)/(E6/E7) and A6-A26 (on Chr 2R)-A7-(A31/A32) (Cao and Jiang., 2018), which suggests that CLIPA4 (SPHII) is ancestral to all the SPHI genes. The same ancestral relationship was supported by the data from Manduca and Drosophila (Cao and Jiang., 2018; Jin et al., 2022). Likely due to positive selection, recently evolved M. sexta SPHs 1b and 101 are more favorable PAP3 substrates and better SPH2 partners in M. sexta. Though poorly expressed in the 45 tissues samples (Cao et al., 2017), it would still be interesting to produce A. gambiae CLIPs A13, A5/A14, E6/E7, or A31/A32, pair with A4, test for possible cofactor activity, and compare with those of A4-A6, A4-A7Δ, and A4-A12 (Fig. 8) to see if the newly evolved SPHIs are positively selected for activating specific PPOs in other tissues or developmental stages.

RNAi is widely used in mosquitoes to reduce expression of target genes, observe phenotypes, and thus test their functional roles. However, this technique has its limitations, such as transient gene knockdown, off-target effect, and ineffective for proteins with long half-lives. Knockdown of A. gambiae CLIPA7 increased melanization of P. berghei ookinetes (Volz et al., 2006), which puzzled us after the cofactor activity of A4-A7Δ was detected (Fig. 8 and Fig. 9). Did deletion of the low complexity region (G⁷⁶EEDD…QNPLD⁵³⁶) cause an artifact in our experiment? We were partly relieved after the positive effect of CLIPA7 was reported for E. coli melanization (Zakhia and Osta, 2022). An off-target effect of RNAi may have caused the increased ookinete melanization. To be cautious, we need to use full-length CLIPA7 and its natural splicing variant (Wang et al., unpublished results) to further test the functional data of CLIPA7Δ. As mentioned in Section 3.1, we would also like to examine the role of A. gambiae CLIPA14 by testing whether CLIPA4-A14 forms a cofactor or interferes with PPO activation (Nakhleh et al., 2017). It would be interesting to determine which one, A. aegypti CLIPs A4, A14, or A4-A14, causes more A. aegypti PPO3 activation by A. aegypti CLIPB9x_a (Ji et al., 2022). Functional tests of A4 and A14 orthologs in the two mosquitoes may inform us how conserved the cofactor role, if any, is after divergence of the mosquito lineages 145–200 million years ago (Krzywinski et al., 2006).

4.3. Mechanism of PPO activation: insights from the A. gambiae system

After the initial reports of the cofactor requirement for PPO activation (Lee et al., 1998; Jiang et al., 1998), major progress was made on the protein isolation, cDNA cloning, action process, and structural determination of the clip-domain SPHs (Kwon et al., 2000; Lee et al., 2002; Kim et al., 2002; Yu et al., 2003; Wang and Jiang, 2004; Gupta et al., 2005; Piao et al., 2005; Lu and Jiang, 2008; Kan et al., 2008; Ji et al., 2022). While mechanisms for the auxiliary effect were explored in some of these studies, we carefully examined the biochemical process of the cofactor formation in Manduca and Drosophila (Jin et al., 2022 and 2023). In this study, we showed that A. gambiae proSPHIs were first cleaved at site-1 (R*F/I) between Cys-3 and Cys-4 of the clip domain and then at site-2 (R*I), the junction between the linker and protease-like domain (Figs. 6, 7 and 10). Like the M. sexta and D. melanogaster SPHIIs, A. gambiae CLIPA4 is proteolytically activated at site-1 where the conserved K*F bond resides in an exposed loop of the clip domains (Huang et al., 2007; Cao et al., 2015).

Fig. 10.

A model for the PPO activation and melanization in A. gambiae. Active CLIPB9 (in red) not only cleaves PPO2 or PPO7 but also generates its own cofactor, a high M_r complex of SPHI and SPHII (in purple). Precursors (in black) of the SPHIs, including proCLIPs A6, A7, A12, and perhaps other paralogs (Fig. 1), are cleaved at site-1 and then site-2 by CLIPB9. The proSPHII (i.e., proCLIPA4) is cleaved by CLIPB9 at site-1 only. M. sexta PAP3 (in green) can perform similar functions, sometimes more efficient than the endogenous PAP. Events of the proteolytic cleavage of PPOs and proSPHs lead to assembling of the SPHI, SPHII, PAP, and PO2/7 into a super-complex in the vicinity of invading organisms or damaged tissues to kill and encapsulate the pathogens or parasites. Since diffusing reactive compounds generated by POs are toxic to host cells, evolution of free POs with low activity may have minimized the adverse effect, while evolution of the cofactors, which may link pathogens, recognition proteins, and other immune factors including highly active POs, allows insects to aim the cytotoxic compounds at the invaders.

Native PAGE of the SPHI and II protein complexes provided insights into their association process in different insects. In M. sexta, cleaved SPHIs (especially 1b and 101) formed high M_r polymers mostly migrating in the stacking gel whereas SPHII existed as monomer, dimer, or multimers that may attach to the SPHI chain (Jin et al., 2022). In Drosophila, polymerization of SPH35 and SPH242 may have facilitated their further assembling into a duplex of SPHI and II chains (Jin et al., 2023). Self-association of A. gambiae CLIPs A6, A7Δ, A12, and A4 occurred in the decreasing order (Figs. 6B and 7B), and the lowest PO7 activity was detected in the presence of A12-A4 (Fig. 9C). As reported previously (Jin et al., 2022), there seems to be a correlation between the cofactor activity and the easiness of SPH self-association and SPHI-II chain coupling.

Determination of PPO cleavage extents and PO activities produced by PAPs in the presence of the cofactors allowed us to estimate the PO specific activities generated under physiological conditions. Based on the published data, we found M. sexta PAP1 generated PO at 520–587 U/μg in the reaction containing the purified PPOs and cofactor (Gupta et al., 2005). Drosophila MP2x_a activated PPO1 at 260 U/μg (Jin et al., 2023). In this study, PO2 specific activities after CLIPB9x_a treatment were 85, 55, and 79 U/μg (Fig. 8) while PO7 specific activities produced by PAP3 were 1,630, 1,339 and 1,131 U/μg (Fig. 9) in the presence of A4-A6, A4-AΔ7, and A4-A12, respectively. The low specific activities of PO2 may be caused by instability of PO2 active conformation since PPO2 treated with 0.002% cetylpyridinium chloride (CPC) yielded much higher PO2 specific activity (573 U/μg) (Hu et al., unpublished data). On the other hand, as the CPC treatment of PPO7 produced PO7 at 618 U/μg, we suspect that the specific activity of PO7 (1,131–1,630 U/μg) may have been skewed by the low conversion of PPO7 to PO7 (Fig. 9). Based on all these results, we predict the PO specific activity generated under more favorable in vivo conditions may be around 1,000 U/μg.

Highlights.

1. A. gambiae CLIPB9x_a cleaved proCLIPA4, 6, 7Δ, and 12 to form its own cofactors.

2. M. sexta PAP3 more efficiently produced the cofactors for A. gambiae PPO activation.

3. CLIPs A6, A7Δ, or A12 pairs with CLIPA4 for PPO2/7 activation by CLIPB9 or PAP3.

4. Cleavages of pro-cSPHIs at site-1 and -2 and pro-cSPHII at site-1 are conserved.

5. Direct biochemical analysis based on orthology is useful for functional elucidation.

The abbreviations used are:

PO and PPO

phenoloxidase and its precursor

PAP

PPO activating protease

SP and SPH

serine protease and its noncatalytic homolog

CLIP

clip-domain SP(H)

PGRP

peptidoglycan recognition protein

SDS-PAGE

sodium dodecyl sulfate-polyacrylamide gel electrophoresis