Molecular interactions between parasite and mosquito during midgut invasion as targets to block malaria transmission

Despite considerable effort, malaria remains a major public health burden. Malaria is caused by five Plasmodium species and is transmitted to humans via the female Anopheles mosquito. The development of malaria vaccines against the liver and blood stages has been challenging. Therefore, malaria elimination strategies advocate integrated measures, including transmission-blocking approaches. Designing an effective transmission-blocking strategy relies on a sophisticated understanding of the molecular mechanisms governing the interactions between the mosquito midgut molecules and the malaria parasite. Here we review recent advances in the biology of malaria transmission, focusing on molecular interactions between Plasmodium and Anopheles mosquito midgut proteins. We provide an overview of parasite and mosquito proteins that are either targets for drugs currently in clinical trials or candidates of promising transmission-blocking vaccines.


INTRODUCTION
According to the World Health Organization (WHO), there were approximately 229 million malaria cases worldwide in 2019, resulting in 409,000 deaths. Africa remains the heartland of malaria transmission, accounting for 94% of cases and deaths globally. Five Plasmodium species, P. falciparum, P. vivax, P. malaria, P. ovale, and P. knowlesi, cause malaria in humans, of which P. falciparum accounts for approximately 99.7% of malaria cases in Africa 1-3 . P. falciparum is also responsible for most malaria-related deaths, especially in children under five years old and in pregnant women 2,4 . Malaria is transmitted to humans by the bite of a female Anopheles mosquito. More than 30 species of Anopheles mosquitoes in various regions of the globe can transmit malaria. Anopheles gambiae is the major vector for the deadliest parasite, P. falciparum, in Africa 5,6 . Current malaria control practices rely heavily on clinical case management through a timely diagnosis and effective drug treatment in addition to vector-based measures such as insecticide-treated bed nets and indoor residual sprays of insecticides. However, the emergence of drug-resistant parasites and insecticide-resistant mosquitoes pose increasing challenges to malaria control and elimination 7,8 , demanding the development of innovative technologies.
Vaccine development against malaria has been on the research agenda for malaria control and elimination for decades. Malaria vaccines are generally divided into three categories according to the target stage of the parasite's life cycle-pre-erythrocytic, blood-stage, and transmission-blocking vaccines (TBV). However, vaccine development against such a complicated eukaryotic parasite is challenging. Currently, the most advanced vaccine, RTS,S, which targets the P. falciparum circumsporozoite surface protein (PfCSP), showed <37% protection against P. falciparum malaria in phase 3 clinical trials [9][10][11] . Similarly, the first field trial of a Pfs25-based TBV in Malian adults revealed significant limitations 12 . Given such partial effectiveness of vaccines against single targets at individual stages, it is anticipated that a more efficacious approach would involve the combination of subunit vaccines targeting multiple antigens and multiple stages 13 . This principle is illustrated in a study showing that combining a partially effective pre-erythrocytic vaccine and TBV could synergistically reduce the prevalence of mosquito infections in laboratory mice 14 . With the growing recognition of effective transmission interruption as one key measure for malaria elimination 3,15 , TBVs have received increasing attention. Currently, there are only a few parasite sexual-stage antigens as the "priority" TBV candidates (Table 1), which emphasizes the need for strenuous efforts in antigen discovery to broaden the antigen repertoire 16,17 . These efforts will benefit from a better understanding of the malaria parasite-vector interactions during the midgut invasion process. Here we review recent advances in the identification of key parasite and mosquito molecules that play roles in midgut invasion, and we evaluate their potential for TBV development.

THE MALARIA PARASITE DEVELOPMENT IN THE HUMAN HOST
The life cycle of Plasmodium parasites alternates between a female Anopheles mosquito and a vertebrate host, involving multiple unique developmental stages that target diverse host cell types 18,19 . In humans, the cycle begins after the injection of sporozoites into the bloodstream by an infected female Anopheles mosquito during its blood-feeding. Within 30 min, the sporozoites migrate from the bloodstream to the liver and invade hepatocytes. Inside the liver cells, the now intracellular parasite divides mitotically over several days and eventually develops into a schizont that contains thousands of merozoites 1,20 . Rupture of a schizont releases merozoites into the bloodstream to invade erythrocytes, which initiate the asexual erythrocytic stage of the replication cycle. Inside the red blood cells (RBCs), the parasites undergo schizogony, lasting over 48 h for P. falciparum, progressing through the ring, trophozoite, and schizont stages. The schizont subsequently ruptures, releasing 8-36 merozoites to invade other RBCs. This is the stage when the affected patient typically manifests clinical symptoms such as fever and chill in recurring episodes [20][21][22] .
During the asexual replication cycle within the RBCs, some parasites undergo sexual development and differentiate into male and female gametocytes, a process termed gametocytogenesis 4,23,24 . For P. falciparum, the gametocyte commitment generally starts before schizogony, wherein an individual schizont produces a progeny of merozoites that develops into either all asexual or all sexual forms [25][26][27] . Gametocytogenesis is influenced by a combination of host and environmental stress factors, including anemia [28][29][30][31] , drug treatment 29-32 , host immune response 33,34 , and high parasitemia 26 . In vitro, gametocytogenesis can be stimulated by manipulation of the parasite culture, for example by adding red blood cell lysate and lymphocytes together with serum. It was also shown that the nucleic acid synthesis inhibitor Berenil stimulates gametocytogenesis [35][36][37] .
Recent studies have provided new insights into the mechanisms of sexual commitment and differentiation. For example, sexual commitment in P. falciparum has been shown to be governed by the AP2-domain transcription factor PfAP2-G functioning as a master regulator [38][39][40] . PfAP2-G also regulates early gametocyte development, as it binds to the promoters of many early gametocyte genes such as Pfs16, etramp10.3, gexp05, Pfg14.744, and Pfg14.748 38 . The P. falciparum gametocyte development 1 (Pfgdv1) is an upstream activator of sexual commitment, antagonizing the heterochromatin protein 1-mediated gene silencing of Pfap2-g [41][42][43] . In addition, targeted gene disruption and complementation showed that P. falciparum gene implicated in gametocytogenesis (pfgig) also participates in gametocyte commitment 44 . Pfs16, a gametocyte-specific parasitophorous vacuole membrane (PVM) protein, is required for gametocyte development 45,46 . A comprehensive list of genes that are essential for gametocyte development has been reported 24 .
P. falciparum gametocytes take 7-10 days to traverse five morphologically distinct stages to reach maturity. Gametocytes in stages I-IV become sequestered in the extravascular spaces of the hematopoietic system such as bone marrow, spleen, brain, heart, and gut, whereas mature gametocytes are released back into the blood circulation 4,47 where they can be taken up by an anopheline mosquito.

THE ANOPHELES MIDGUT AS A TRIGGER FOR PARASITE TRANSFORMATION
When a mosquito bites a human, the ingested blood meal regulates mosquito gene expression 48 . Specifically, when the human blood reaches the mosquito midgut, it stimulates the midgut epithelium to secrete factors including 3-13% chitin microfibrils and a variety of proteins 49,50 , which forms a peritrophic matrix (PM) 51 that surrounds the ingested blood and separates the blood bolus from midgut epithelial cells. The newly formed PM provides another physical barrier that limits the infection of the mosquito by pathogens in the blood meal 52 .
Inside the mosquito midgut, the co-ingested mature gametocytes undergo gametogenesis, gamete fertilization, and ookinete development before penetrating the midgut epithelium and landing at the destination between the midgut endothelium and basal lamina 53,54 . Several mechanistic details of this sequence of events are now understood. Thus, immediately after ingestion, the gametocytes are activated collectively by environmental signals such as a sharp drop in temperature of approximately 5°C, a rise in pH from 7.4 to 8.0, and a mosquito-derived molecule xanthurenic acid (XA) 55 . Each female extracellular gametocyte forms a single immotile macrogamete, while a male extracellular gametocyte generates up to eight flagella-like microgametes in a process called "exflagellation". Gametocyte egress from the RBC occurs via an inside-out fashion with the rupture of the PVM first and the erythrocyte membrane later 56 . Recent studies revealed some of the molecular details of the gametogenesis process. XA activates membrane guanyl cyclases in gametocytes to synthesize cyclic guanosine monophosphate (cGMP), consecutively stimulating the cGMP-dependent protein kinase G (PKG) pathway and significantly increasing the intracellular calcium 57 . A recent CRISPR/Cas9-based functional study demonstrated that an intracellular membrane protein, the gametogenesis essential protein 1 (GEP1), is essential for XAstimulated gametogenesis regulation in P. yoelii 58 . Other signaling molecules identified to be essential for gametogenesis include the male gametocyte-specific kinase PfMAP-2 59 . Cytosolic Ca 2+ triggers the release of a perforin-like protein PfPLP2, which in turn permeabilizes the erythrocyte membrane, leading to P. falciparum gametocyte egress from erythrocytes 60 . P. falciparum patatin-like phospholipase 1 (PfPATPL1) with phospholipase A2 (PLA2) activity was found to play a crucial  46 . In a more recent study, antibodies to Pfs16 resulted in a significant reduction in the number of oocysts. In the same study, five additional parasite proteins including PF3D7_0303900, PF3D7_1204400 (Pfs37), PF3D7_1214800, PF3D7_1239400, and PF3D7_1472800 were found to interact with the mosquito midgut 62 . It has been shown that knocking out Pbg37 (PBANKA_060330) in P. berghei, the ortholog of Pfs37, led to a significant reduction in the formation of oocysts 63 . These proteins, along with the previously discussed Plasmodium proteins, may serve as novel targets for blocking malaria transmission. Several Plasmodium proteins involved in fertilization and zygote development have been reported. Pfs48/45 is expressed in both male and female gametes, and a gene disruption study of Pfs48/ 45 revealed that only the male gamete fertility was altered and failed to fertilize the macrogamete 64 . Pfs230 is another critical protein for male fertility and exflagellation 65,66 . A male-specific protein, a disulfide isomerase (PDI-Trans/PBANKA_0820300), is reportedly crucial for fertilization 67 . The female-specific protein P47 is required for the fertility of the macrogametes in P. berghei 66 , but not in P. falciparum 68 . However, Pfs47 in P. falciparum does play a fundamental role in parasitic survival by mediating the evasion of the immune system of the vector 69 . Similarly, the P. berghei P47 is also important during ookinete-to-oocyst transition by protecting ookinetes from the complement-like response of mosquitoes 70 . Fertilization starts when a microgamete attaches to a macrogamete and undergoes membrane fusion. HAP2, a conserved protein of protozoan parasites, plants, and algae, has been reported to be essential for P. berghei gamete fusion 71 .
Following fertilization, the zygotes undergo meiosis and transform into motile and invasive ookinetes. Expression of over 500 genes in ookinetes is regulated by the AP2 family Plasmodium transcription factor AP2-O 72 , which affects ookinete development, motility, midgut penetration, mosquito immunity evasion, and oocyst development initiation 73 . Further studies demonstrated the function of these genes (Table 2), including ookinete surfaceassociated proteins like P25 and P28; secretory proteins like chitinase, perforins, PPLP3-5, and PSOPs 74 ; adhesive proteins like the secreted ookinete adhesive protein (SOAP) 75 , the von Willebrand factor A domain-related protein (WARP) 76 , circumsporozoite-and TRAP-related protein (CTRP) 77 ; mobile proteins like glideosome-associated proteins (GAP) and cell traversal protein for ookinetes and sporozoites (CelTOS); and heat shock protein (HSP) 20, 40, 70, and 90 73 . P25 and P28 share multiple functions and are crucial for ookinete development, midgut traversal, and ookinete to oocyst transformation. P25 and P28 are among the top candidates for TBV development 78,79 , a topic that is discussed in greater detail later in this review. The secreted enzymes like chitinase digest the physical barrier of the PM, assisting the ookinete to invade the mosquito midgut 80 . In terms of mobility, P. yoelii encoded guanylate cyclase β (GCβ) is expressed on the ookinetes and localized polarly at the ookinete extrados site (OES) and found to be essential for ookinete gliding. GCβ contains the N-terminal P4-ATPase-like domain (ALD) and the C-terminal guanylate cyclase domain (GCD), which are required for its polymerization and subsequent ookinete gliding. During ookinete development, CDC50A, as a co-factor of P4-ATPase, stabilizes GCβ. IMC sub-compartment protein 1 (ISP1) was found to be a crucial molecule for anchoring the GCβ/CDC50A complex at the OES of mature ookinetes 81 .

P. FALCIPARUM TBV CANDIDATE PROTEINS Pfs25 and Pfs28
The expression of Pfs25 and Pfs28 begins in the extracellular gametocytes within the mosquito vector 82 . Pfs25 is a 25 kDa sexual stage protein mostly expressed on the surface of macrogametes, zygotes, and ookinetes of P. falciparum inside the mosquito midgut 83 . Pfs25 was the first protein to progress to a clinical trial; however, the phase 1 trial of this protein, formulated in the adjuvant Montanide ISA51, was found to be reactogenic 84 .
Ensuing research focused on combining Pfs25 with other proteins. Fusion of Pfs25 antigen to IMX313, a protein multimerization technology, formulated a nanoparticle with enhanced immunogenicity 85 . A similar study demonstrated stimulation of functional antibody response against malaria infection and transmission in mice by combining Pfs25-IMX313 with RTS, S/AS01 86 . Several studies involving other recombinant products have moved into phase 1 clinical trial, including one involving the conjugation product of Pfs25 linked to a detoxified form of Pseudomonas aeruginosa exoprotein A (EPA) 87 , Pfs25-EPA conjugates formulated with Alhydrogel ® 88 , and a chimeric virus-like particle (VLP) containing Pfs25 fused to the alfalfa mosaic virus coat protein 89 . However, the antibody titers in these trials and the effectiveness of transmission-blocking (TB) potential were insufficient. Hence there is a need for an alternative adjuvant. Similarly, Pfs28 is a 28 kDa protein mainly expressed on the surface of zygotes and ookinetes of P. falciparum 90 . Chemical conjugation of Pfs28 to a mutant EPA significantly enhanced immunogenicity in mice immunized with conjugated Pfs28 91 . A 39-kDa chimeric recombinant protein produced by the fusion of Pfs25 and Pfs28 proteins secreted by Saccharomyces cerevisiae was found to be more potent than the two candidate proteins alone 92 .

Pfs230 and Pfs48/45
The expression of the gametocyte antigens Pfs230 and Pfs48/ 45 starts intracellularly within the human host 82 . Pfs230 and Pfs48/ 45 are 6-cys family proteins expressed on the surface of gametocytes and gametes of P. falciparum 93,94 . Pfs48/45 is expressed by both male and female gametes; however, it is required only for male fertility 64 . Despite their important role in malaria transmission, the development of these TBV candidate antigens has been hampered due to a lack of properly folded recombinant proteins 95 . Although the expression of a full-length recombinant Pfs230 has been hindered by its large size and a large number of disulfide bonds 96,97 , different domains at the N-terminal region of Pfs230 showed TB activities 98 . The Pfs230 N-terminal prodomain, Pfs230C (amino acids 443 to 1132), synthesized using a cell-free system, showed a sufficient complement-dependent malaria TB activity 98 . Antibodies generated against Pfs230C1, another region of the Pfs230 N-terminal (amino acids 443 to 731), significantly reduced the number of oocysts 99 . Pfs230D1+ , a modified form of Pfs230C1 constructed by eliminating the glycosylation property, was found to be homogeneous and more stable than Pfs230C1. Pfs230D1+ also demonstrated higher expression yield and transmission-reducing activity similar to Pfs230C1 100 . Another construct of Pfs230 domain C, 230CMB, produced using a plant-based expression system, showed strong TB activity 101 . In another study, conjugation of Pfs230 with an outer membrane protein complex of Neisseria meningitidis was found to have enhanced immunogenicity and functional activity of the Pfs230 protein 102 .
Similar to Pfs230, expression of a full-length recombinant protein Pfs48/45 has been difficult due to its size and complexity, which forced researchers to focus on protein domains. A C-terminal Pfs48/45 expressed in baculovirus demonstrated significant transmission-reducing activity and was found to be homogeneous with respect to its size, conformation, glycosylation, and folding 103 . Several studies showed that chimeras of the 6 C subunit of Pfs48/45 produced in Lactococcus lactis are a strong candidate for TBV [104][105][106] . In another study, monoclonal antibody 85RF45.1 against Pfs48/45 demonstrated a strong reduction of parasite transmission 107 . A recent study reported that a chimeric protein construct of Pfs230 and Pfs48/45, formed by the fusion of the Prodomain of Pfs230 and 6 C fragment of Pfs48/45, showed a higher TB potency than the single proteins alone 108 . This finding bolsters the previous findings with the Pfs25 and Pfs28 fusion protein. All together we can assume that protein chimeras of two or more different proteins can be used to generate antibodies with higher TB activity.

Pfs47 and HAP2
Pfs47 and HAP2 are other recently identified TBV target proteins. Pfs47 is a member of the 6-cys family of proteins expressed only in macrogametocytes, macrogametes, and ookinetes 94 . Although Pfs47 was found to be dispensable for female gamete fertility, it mediates parasite evasion of the mosquito immune system by suppressing the effect of midgut nitration, a crucial reaction to activate the complement-like system 68,69,109 . While monoclonal antibodies against the full-length Pfs47 protein failed to show TB activity 68 , antibodies targeting a specific region of Pfs47 revealed a significant TB activity 110 . More importantly, conjugation of Pfs47 with AP205 VLP enhanced its immunogenicity and TB activity 111 . Likewise, an in vivo study identified a specific region of P. berghei P47 (Pbs47) that offers protection against mosquito immunity. This particular domain has similar immunogenicity to that of Pfs47. Conjugation of the protective antigenic region to a bacteriophage AP205-VLP enhanced immunogenicity and TB activity 112 . A conjugated form of different vaccine candidates has shown higher immunogenicity and better TB activity. Therefore, a conjugated vaccine form of candidate antigens may help resolve the issue of poor immunogenicity during the development of an effective malaria TBV.
On the other hand, PfHAP2, a family member of HAP2 family protein, is expressed only in male gametocytes and activated male gametes 113 . The PfHAP2 recombinant protein-induced IgG antibodies have a significant TB activity 114 . In P. berghei, PbHAP2 induced antibodies also showed potent in vitro and in vivo TB activity 113 . A recent study also showed that the anti-Plasmodium vivax HAP2 (PvHAP2) antisera significantly reduced the oocyst numbers in mosquito feeding assays using clinical P. vivax isolates 115 . Collectively, these findings emphasize that HAP2 is a promising TBV candidate that warrants further investigation and trials.

MIDGUT PROTEIN INTERACTIONS AFFECTING PLASMODIUM TRANSMISSION
Much of TBV research effort has focused on discovering parasite proteins involved in mosquito invasion. However, recently a concerted effort has been made to elucidate mosquito midgut proteins involved in the parasite invasion process. Essential molecular interactions between the mosquito and the parasite have been discovered, revealing prime TBV candidates such as the AnAPN1, FREP1, and AgPfs47Rec (Fig. 1) as promising TBV targets. Besides these major players, in this section, we will also describe newly discovered mosquito midgut proteins that play a role in the P. falciparum transmission process.

AnAPN1
Anopheles alanyl aminopeptidase N (AnAPN1) is considered an important and promising mosquito-based TBV candidate. It is found in the midgut of A. gambiae and acts as a ligand for P. falciparum and P. berghei ookinete invasion 116,117 . The midgut microvilli glycol-conjugates interact with the ookinete Y. Keleta et al. and enhances infection of the mosquito 117 . Mosquito midgut glycoproteins are extensively glycosylated with a high proportion of N-linked N-acetylglucosamine (GlcNAc) and N-acetylgalactosamine (GalNAc) terminal oligosaccharides. Glycans, such as GalNAc, and lectins like jacalin, reduce parasite invasion inside the mosquito midgut 118,119 . AnAPN1 at the A. gambiae midgut luminal surface is the principal jacalin target and plays an important role in ookinete invasion. The conserved role of AnAPN1 in ookinete invasion of the mosquito midgut explained by the ability of anti-AnAPN1 IgG to strongly inhibit both P. berghei and P. falciparum development in various mosquito species 117 . Identifying the specific domain of the ligand that interacts with the parasite is important for designing an appropriate TBV.
The AnAPN1 monoclonal antibody, 4H5B7 mAb, showed a potent TB activity at low concentrations 116 . The recombinant AnAPN1 demonstrated acceptable vaccine potency and immunogenicity without immunization-related histopathologies in mice. The AnAPN1 protein has four domains (I to IV), and antibodies against the N-terminal domain-I were found to interfere with the sexual cycle of P. falciparum and P. berghei 117,120 . Although more research is still required to identify the critical interactions, the resolution of the structure of AnAPN1 is a major advance, which will guide structure-based vaccine design. One such attempt involved using only the relevant domain I of AnAPN1 as the immunogen to reduce unspecific immune stimulation. The domain I construct elicited adequate immune responses in mice, and the IgG antibodies showed potent TB activity 121 .

FREP1
FREP1 presents another leading candidate for a promising mosquito-based TBV. FREP1 belongs to the fibrinogen-related protein (FREP/FBN) family. In mammals, FREPs play a fundamental role in blood coagulation, while in mosquitoes, they generally act as pattern recognition receptors 122 . Several FREPs activate the mosquito immune system by attaching to external pathogens such as parasites, bacteria, and fungi to activate downstream defense signaling pathways 123,124 . FBN9 and FBN30 were found to inhibit Plasmodium infection in midgut epithelial cells. Their functions were illustrated by the increased Plasmodium infection of the midgut when FBN9 or FBN30 was knocked down 125,126 . FBN30 is a secreted octameric protein and specifically interacts with the P. berghei blood stages and ookinetes, but not sporozoites 127 . Intriguingly, FBN30 was found to interact with the clinically circulating P. falciparum, but no interaction occurred with the NF54 P. falciparum laboratory strain 127 . Understanding the underlying mechanism for this difference may be helpful for TBV development.
Anopheles FREP1 was discovered through an association study using clinically circulating P. falciparum and wild-derived A. gambiae 126 . Unlike FBN9 and FBN30, which hinder Plasmodium infection, FREP1 serves as a molecular anchor in the PM, assisting P. falciparum ookinete invasion of the mosquito midgut 128 . The Plasmodium α-tubulin-1 protein was found to strongly interact with FREP1 during in vitro studies, and antibodies against α-tubulin-1 blocked malaria transmission inside the mosquito 129 . An increased understanding of the molecular interaction of FREP1 with the Fig. 1 Key proteins responsible for the development of P. falciparum parasites inside the Anopheles mosquito midgut. 1 The male and female gametocytes ingested during a blood meal egress from the RBCs and become extracellular gametocytes which then develop into male and female gametes via numerous factors such as an increase in pH, a decrease in temperature, and the presence of XA in the midgut environment. This leads to a rise in intracellular Ca 2+ , IP3, and cGMP signaling and to the development of gametes. 2, 3 The microgametes (male) and the macrogametes (female) increase expression of certain key proteins such as Pfs48/45 and Pfs47 and generate a complete zygote after fertilization. 4 Zygotes develop into motile ookinetes that can penetrate the PM by breaking its chitin structure through chitinases (CHT1), perforins (PPLP3-5), and many other enzymes. 5 Several midgut proteins such as AnAPN1 and FREP1 also help parasites penetrate the peritrophic membrane and midgut epithelium to develop into oocysts at the basal lamina side.
α-tubulin-1 may offer another promising TBV candidate. Lastly, it was shown that FREP1 CRISPR-Cas9 knockout mosquitoes exhibited a profound suppression of parasite infection at both the oocyst and sporozoite stages. Yet, inactivation of the FREP1 gene inflicted substantial fitness costs on blood-feeding propensity, longevity following a blood meal, fecundity, as well as the rate of egg hatching 53 . A study identified direct interactions of the fibrinogenlike (FBG) domain of FREP1 with P. falciparum gametocytes and ookinetes. FBG is highly conserved in different species of Anopheles mosquitoes and presents as an effective TBV target against infection from a variety of Plasmodium species 124 . The FBG region between amino acids 463 and 677 of A. gambiae FREP1 shows more than 90% identity in protein sequence with the FREP1 FBG in 13 anopheline species, suggesting that anti-FBG IgG may potentially block malaria transmission in these 13 mosquito species 5 . Notably, anti-FBG serum showed greater than 81% TB efficiency against P. falciparum in A. gambiae and~67% TB efficiency against P. vivax in A. dirus 23 . This finding further supports that FREP1, specifically the FBG domain, is an ideal TBV target.

AgPfs47Rec (AGAP006398)
A recent study revealed that Pfs47 mediates immune evasion in mosquitoes through interaction with the Anopheles midgut receptor protein AgPf47Rec (AGAP006398) 130 . The interaction allows the parasite to evade the mosquito immune system by disrupting the c-Jun-N-terminal kinase (JNK) signaling pathway and eventually suppressing the effect of midgut nitration, a crucial reaction to activate the mosquito complement-like system 68,69,109,131,132 . Further studies are needed to determine if antibodies against AgPfs47Rec possess TB activity. In addition, the identification of mosquito species-specific AgPfs47Rec and its interaction with various Pfs47 haplotypes provide increasing evidence for a highly specific lock and key model between the ligand and receptor 130 , demonstrating a natural selection of Pfs47 haplotypes in the P. falciparum adaptation to different mosquito vectors.
Other midgut proteins An increasing number of midgut proteins interacting with the parasite and mediating parasite transmission to mosquitoes have been discovered (Table 3). In a recent study, the Anopheles midgut protein AGAP008138 was found to interact with sexual-stage Plasmodium parasites and, upon knockdown, increased mosquito susceptibility to P. falciparum 133 . Likewise, the evolutionarily conserved heme peroxidase gene HPX15 134 was found to interact with the P. falciparum sexual stages 133 , and knocking down its expression in the midgut after a bloodmeal decreased the oocyst load in A. gambiae due to the reduced integrity of the mucin barrier and subsequent activation of midgut immunity pathways 135 . In addition, some immunoglobulin-like secreted proteins such as AGAP002848 and AGAP002851 are pattern-recognition molecules and possibly potent inhibitors of parasite infection 133 . Some structure-related proteins such as AGAP006972 and AGAP006268 are needed to maintain the integrity of the midgut PM and suppress parasitic invasion 133,136 . These newly identified midgut proteins are all worthwhile of further investigation to determine their TB potential and suitability.

CONCLUSION
Here we reviewed recent advances in the study of malaria transmission, focusing specifically on molecular mechanisms of interactions between P. falciparum and A. gambiae mosquito midgut proteins. Overall, TBVs can halt the transmission cycle of the malaria parasite between the mosquito and its vertebrate host. Although TBV development has met many challenges, recent studies have identified novel protein candidates from both the parasite and vector sides. Further investigation of the molecular mechanisms of interaction between the mosquito midgut proteins and the malaria parasite will provide fundamental insights into malaria transmission and accelerate TBV development, ultimately bringing us one step closer to eradicating this devastating disease from this planet.

DATA AVAILABILITY
All data are provided in the main text, tables and the figure.