Anopheles fibrinogen-related proteins provide expanded pattern recognition capacity against bacteria and malaria parasites

Abstract

The fibrinogen-related protein family (FREP, also known as FBN) is an evolutionarily conserved immune gene family found in mammals and invertebrates. It is the largest pattern recognition receptor gene family in Anopheles gambiae, with as many as 59 putative members, while the Drosophila melanogaster genome has only 14 known FREP members. Our sequence and phylogenetic analysis suggest that this remarkable gene expansion in the mosquito is the result of tandem duplication of the fibrinogen domain. We found that the majority of the FREP genes displayed immune-responsive transcription after challenge with bacteria, fungi, or Plasmodium, and these expression patterns correlated strongly with gene phylogeny and chromosomal location. Using RNAi-mediated gene-silencing assays, we further demonstrated that some FREP members are essential factors of the mosquito innate immune system that are required for maintaining immune homeostasis, and members of this family have complementary and synergistic functions. One of the most potent anti-Plasmodium FREP proteins, FBN9, was found to interact with both Gram-negative and Gram-positive bacteria and strongly co-localized with both rodent and human malaria parasites in the mosquito midgut epithelium, suggesting that its defensive activity involves direct interaction with the pathogen. Interestingly, FBN9 formed dimers that bound to the bacterial surfaces with different affinities. Our findings indicate that the A. gambiae FREP gene family plays a central role in the mosquito innate immune system and provides an expanded pattern recognition and anti-microbial defense repertoire.

FIGURE 1.

Phylogenetic relationships between the FBG domains and the chromosomal location of FREPs. FREP genes are grouped into 4 large clusters, and the chromosomal locations (Chr.) of each FREP have been manually drawn, with red, green, purple, and blue lines indicating ClusterR, ClusterG, ClusterP, and ClusterR, respectively. FREP genes scattered throughout all the chromosomes are indicated by gray lines and grouped into ClusterX. The FREP and FBN are indicated, and the accession numbers are listed in supplemental Table S1. The tree was constructed with ClustalX and TreeView using NJ-joining methods.

FIGURE 2.

The sex, tissue-specific, and immune-responsive expression of FREP genes. At the left side are the FREP clusters, and the right side gives the corresponding FBN. M indicates the gene expression of that FREP in male mosquitoes compared with female mosquitoes and is followed by FREP expression within different tissues (Tissue exp.), compared with whole female mosquitoes. Th, Gut, Ab, 5B, 4A indicates thorax, gut, abdomen, cell line Sua5B, cell line Sua4A, respectively. The expression profiles of individual FREPs upon immune challenge of E. coli (Ec), S. aureus (Sa), B. bassiana (Bb), and Plasmodium-(P. berghei (Pb) and P. falciparum (Pf)) are given by comparing to naïve samples. Red and green color indicates higher and lower expression, respectively; black color indicates no difference within expression levels and gray indicates the value was not available. The expression profiles within each cluster are clustered based on the tissue expression analysis.

FIGURE 3.

The involvement of FBN22 and FBN39 in the defense against opportunistic bacterial infections. A, survival rates of the mosquitoes treated with dsRNA of FBN22-, FBN39-, and the pool of dsRNA-treated (pool) mosquitoes, dsGFP-injected (GFP) mosquitoes were served as control. The experiments shown represent at least three replicates; standard errors are not shown here to allow for clearer visualization (2-way analysis of variance; *, p < 0.01). B, CFUs of opportunistic bacteria isolated from FBN39 gene-silenced mosquito hemolymph at 4 days post dsRNA injection, non-treated mosquitoes (naive), and dsGFP-injected mosquitoes (GFP) were served as control. The isolated bacterial species are Serratia sp. (Serratia), Asaia bogorensis (A.b.), Pseudomonas veronii (P.v.), and Sphingomonas sp. (S.sp). Total Bac indicates the total number of bacteria been isolated.

FIGURE 4.

P. berhgei and P. falciparum oocysts infection levels of the pools or individual FREPs gene-silenced mosquitoes. A, median P. berghei oocysts infection level of mosquitoes been gene-silenced of a pool of FREPs from ClusterB (FBNb) and ClusterP (FBNp), dsGFP-injected (GFP) mosquitoes were served as control. Tep1 dsRNA-treated mosquitoes (Tep) served as the positive control. The standard errors for three biological replicates are shown; * indicates p < 0.05 and ** indicates p < 0.01. B, median P. berghei oocysts infection level of mosquitoes been gene-silenced of individual FREPs (FBN5, FBN6, FBN9) and a pool of these three FBNs and FBN26 (pool). C, P. falciparum oocysts counts from the groups of mosquitoes with same treatments as from A. D, P. falciparum oocysts counts from the groups of mosquitoes with same treatments as from B.

FIGURE 5.

Co-localization of FBN9 with bacteria in the Sua5B cell line, and with Plasmodium in the mosquito midgut. A, FITC-labeled E. coli (green) that had been co-incubated with Sua5B cells co-localized with FBN9 (red). Co-localization is indicated in white; cell nuclei were stained with DAPI (blue). Scale bars:10 μm. B, co-localization of FBN9 with P. berghei ookinetes in the mosquito midgut epithelium 24–30-h post-infection (B2-B4). The ookinetes in the midgut stained with preimmune of anti-FBN9 antibody showed no co-localization, which served as negative control (B1). Blue indicates DAPI-stained cell nuclei, green indicates P. berghei ookinetes, and red indicates anti-FBN9 staining. Scale bars, B1-B2: 10 μm; B3: 1 μm; B4: 2 μm. C and D, FBN9 co-localizes with P. falciparum ookinetes in the mosquito midgut epithelium. Double staining of midgut tissues of mosquito 24–30-h post-infection with P. falciparum using the anti-FBN9 (red) and anti-Pfs25 (green) antibodies. Scale bars, C1-C4: 10 μm; D1-D4: 5 μm.

FIGURE 6.

FBN9 forms dimers when binding to either Gram-negative or Gram-positive bacteria. Western analysis by using anti-FBN9 antibody shows the results from in vitro bacterial binding assays. In the absence of additional DTT (100 mm final concentration) in the loading buffer, a 66-kDa protein complex recovered from the surface of E. coli, P. veronii, and B. subtilis (lanes 2, 4, and 6, respectively). The Sua5B cell supernatant used in the binding assay showed the same size of protein (66 kDa), as shown in lane 1. Addition of 100 mm of DTT to the loading buffer resulted in a 33-kDa band instead of the 66-kDa band (lanes 3, 5, 7, for E. coli, P. veronii, and B. subtilis, respectively). + or – indicates the supplement or the absence of DTT, respectively.