|Home | About | Journals | Submit | Contact Us | Français|
Serpins (serine protease inhibitors) regulate some innate immune responses of insects by inhibiting endogenous proteases. In this study, we characterized the serpin (SRPN) gene family in the mosquito Anopheles gambiae, the major malaria vector in Sub-Saharan Africa. We identified 18 A. gambiae SRPN genes, all on chromosomes 2 and 3, through searches of genomic DNA and EST databases. In addition to SRPN10, previously documented to exhibit alternative splicing, we found three splicing isoforms of SRPN4. We completed sequencing of cDNAs for the A. gambiae serpins to obtain complete coding sequence information and to verify or improve gene predictions. The predicted SRPN9 and 15 in the initial genome annotation were determined to be a single gene (SRPN9). Sixteen of the serpins contained putative secretion signal sequences. Multiple sequence alignments showing conserved residues important in structural conformation, including the consensus pattern within the hinge region, indicated that most of the A. gambiae serpins may be inhibitory. Phylogenetic analyses confirmed that SRPN1, 2, 3, 8, 9 and 10 formed phylogenetic clusters with known inhibitory serpins from Drosophila melanogaster and Manduca sexta. Many of the A. gambiae serpins were expressed during all life stages. However, SRPN7, 8, 12, and 19 were expressed at very low levels in the adult stage. SRPN13 was expressed mostly in eggs and young larvae, whereas SRPN5 and 14 were expressed mostly in adults. Such differences in expression pattern suggest that the serpins are involved in multiple physiological processes. Determining the biological functions of the mosquito serpins will require future work to identify the proteases they inhibit in vivo.
Serpins (Serine Protease Inhibitors) are a superfamily of proteins with a unique structure comprising three β-sheets and seven to nine α-helices, folding into a conserved tertiary structure with a reactive center loop (RCL) near the carboxyl-terminal (Pearce et al., 2007). The RCL is exposed at the surface of the serpin and acts as a bait for a target protease. After a protease cleaves the RCL at the scissile bond between amino acid residues designated P1 and P1′, the serpin undergoes a large conformational change, and the protease is trapped with its active site distorted, resulting in its inactivation (Gettins, 2007). The metastable structural conformation of serpins is required for their inhibitory activity, and thus facilitates the serpins to act as suicide-substrate inhibitors (Silverman et al., 2001). Most serpins inhibit serine proteases, although some can inhibit proteases from other classes. For example, squamous-cell carcinoma antigen-1 inhibits cysteine proteases (Suminami et al., 1998), and Arabidopsis thaliana serpin-1 inhibits metacaspase-9, an A. thaliana cysteine protease (Vercammen et al., 2006). However, some other serpins lack protease inhibitory properties and carry out other functions, such as, hormone transport or acting as molecular chaperones or storage proteins (Dafforn et al., 2001; Huntington and Stein, 2001; Irving et al., 2000).
Serpins are found in an array of eukaryotes, including animals, plants, and some viruses (Gettins, et al., 2002; Irving et al., 2000). Serpins also occur in bacteria and archea (Cabrita et al., 2007; Kang et al., 2006; Irving et al., 2002; Irving et al., 2003). In insects, serpins have been identified in several species, but most studies of insect serpins are from the fruit fly Drosophila melanogaster and the tobacco hornworm Manduca sexta. So far, six M. sexta serpins have been characterized (Gan et al., 2001; Jiang and Kanost, 1997, Tong and Kanost, 2005; Wang and Jiang, 2004; Zhu et al., 2003) and shown to be inhibitory. With the exception of Manduca serpin-2, they all are involved in regulation of the prophenoloxidase (PPO) activation cascade, an innate immune response, but they are likely to have additional functions not yet discovered. The D. melanogaster genome contains at least 29 serpin genes (Reichhart, 2005). However, only a handful of the Drosophila serpins have been characterized through genetics or biochemistry and shown to be inhibitory and involved in regulating the Toll and PPO signaling pathways (Gubb et al., 2007).
Research on insect innate immune mechanisms is essential for understanding interactions between insect disease vectors and the pathogens they transmit. Of particular importance is the mosquito Anopheles gambiae, the major malaria vector of Plasmodium falciparum in Sub-Saharan Africa. Human hosts are infected with the Plasmodium parasites through bites of female Anopheles mosquitoes. The A. gambiae genome was completed (Holt et al., 2002), and 242 genes were identified to be immunity-related, including 14 serpin genes (SRPNs) (Christophides et al, 2002). Another serpin gene, SRPN15, was annotated as a putative haplotype of SRPN9. More recently, the serpin gene family in the A. gambiae genome was reexamined and compared with those in the genome of another mosquito, Aedes aegypti (Waterhouse et al., 2007), and four additional A. gambiae serpin genes (in some cases partial genes) were identified. In order to study functions of the serpins and conduct further bioinformatics analyses, knowing their accurate and complete coding sequences will be crucial. Problems with physical mapping and assembly of the A. gambiae genome persist, including unmapped scaffolds, physical gaps between mapped scaffolds, polymorphism, bacterial DNA contamination, and incomplete Y chromosome sequence (Sharakhova et al., 2007). Furthermore, gene predictions may be incomplete or inaccurate due to incorrect intron-exon prediction or regions of poor genomic sequence quality. A pilot study using a full-length enriched cDNA library from adult female A. gambiae mosquitoes discovered a number of genes previously unannotated (Gomez et al., 2005). It is likely that there are genes still missing from the genome annotation, particularly genes expressed predominantly in immature life stages, for which EST data are scant. Thus, information obtained from cDNA sequences is essential for fully characterizing gene structure, including intron-exon arrangement, alternative splicing, and complete protein-coding regions. The sequences of transcripts also serve as an important tool to study gene expression and to produce recombinant proteins from cloned cDNAs, which will be vital reagents for biochemical studies of the mosquito serpins.
In this study, we cloned and sequenced cDNAs for A. gambiae serpins (with the exception of SRPN10, which was previously characterized (Danielli et al., 2003)). We identified 18 serpin genes, with three splicing isoforms of SRPN4. The cDNA sequences were used to characterize the serpin gene family of A. gambiae through bioinformatics and phylogenetic analyses and to examine expression profiles of the serpins among different developmental stages.
Anopheles gambiae G3 strain was originally obtained from the Malaria Research and Reference Reagent Resource Center (MR4, http://www.mr4.org/). The colony was maintained according to Benedict (1997) at 27°C, 80% relative humidity, 16:8h light:dark cycle, with adult females feeding on equine blood through a membrane feeder. Larvae were provided ground VitaPro plus staple power flakes and baker’s yeast. Adults were provided cotton balls soaked in 10% sucrose as food.
Nucleotide sequences of initially predicted A. gambiae serpin genes were gathered from Genbank. InterPro domain entry IPR000215 (Serpin) was also used to screen the A. gambiae genome database in Ensembl for possible serpin genes that had not been identified previously. Predicted amino acid sequences associated with each gene were used for BLASTP queries against the NCBI database to confirm the presence of conserved serpin domains. The nucleotide sequences of serpin gene candidates obtained from both methods were used for EST database searches through GenBank BLASTN. The closest matches from the EST search results were selected, and the EST clones were ordered from the Malaria Research and Reference Reagent Resource Center (MR4) if available (see supplemental Table S1). These EST clones were from A. gambiae Giles adult cDNA libraries, and provided as DNA constructs in the pSPORT1 vector in Escherichia coli DH10B. SRPN10 had already been extensively studied (Danielli et al., 2003; 2005; Lycett et al., 2004), and thus it was excluded from this search.
For serpins not obtained from MR4, cDNA clones were produced by RT-PCR. Based on expression profiles of each serpin at different developmental stages, RNA from mosquitoes at a stage that showed a high level of expression was used to make cDNA, which was subsequently used as a template for PCR. cDNAs from adult males were used for SRPN1, 16, 17, 18; adult females for SRPN4A; day-9 larvae for SRPN7; and egg for SRPN13 (see section 2.6 for detail on cDNA preparation). PCR reactions were performed at 30-32 cycles using high fidelity Platinum Taq DNA polymerase (Invitrogen) and specific forward and reverse primers, corresponding to regions containing the start and stop codons to amplify cDNAs containing the entire open reading frame for each serpin. PCR products were analyzed by agarose gel electrophoresis and detected by ethidium bromide staining. The gel bands were excised, and the cDNAs were then extracted from the gel using QIAquick gel extraction kit (Qiagen). The cDNAs were cloned into pCR4-TOPO plasmid vector (Invitrogen) according to the manufacturer’s protocol. The cDNA constructs were used to transform One Shot TOP10 Chemically Competent E. coli (Invitrogen), which were subsequently allowed to grow on LB agar plates with ampicillin as a selective agent. Bacterial colonies were harvested individually, and plasmids were purified using QIAprep Spin Miniprep kit (Qiagen). To verify that plasmids contained cDNAs for the genes of interest, they were screened by restriction digestion using EcoRI (New England Biolabs).
The nucleotide sequence of a gene, named SRPN19 in this study, was incomplete at both ends in the gene prediction. Oligo-capping rapid amplification of cDNA ends (RACE) methods were used to obtain full-length 5′ and 3′ ends of the SRPN19 transcript. 5′- and 3′-RACE reactions were carried out using a GeneRacer kit (Invitrogen) according to the manufacturer’s protocol. Day-9 larval cDNA was used as a template, and high fidelity Platinum Taq DNA polymerase was used for amplification. Both RACE products were then cloned and sequenced. Finally, the entire cDNA from 5′ to 3′ ends was amplified using RACE-ready cDNA as a template. The product was cloned into pCR4-TOPO plasmid vector. One Shot TOP10 Chemically Competent E. coli was used for transformation, and the cDNA insert of this plasmid was sequenced.
Sequencing of all plasmids and PCR products in this study was carried out at either the Iowa State University Sequencing Facility (Ames, IA) or the DNA Sequencing Facility at the Department of Plant Pathology, Kansas State University (Manhattan, KS). The sequence data and chromatograms were thoroughly checked and assembled. Deduced amino acid sequences were produced by using the Translate tool via the ExPASy website (http://www.expasy.org/tools/dna.html) (Gasteiger et al., 2003). The final nucleotide and amino acid sequences were deposited in the GenBank database (see Table 1 for GenBank accession numbers).
When cDNAs of the A. gambiae serpins from MR4 were incomplete at the 5′-end, corresponding 5′-end cDNA sequences available from GenBank (with overlapping regions containing >95% nucleotide sequence identity) were used to assemble complete coding sequences. The presence of secretion signal sequences and location of their cleavage sites were predicted using the web program SignalP 3.0 (http://www.cbs.dtu.dk/services/SignalP/) (Bendtsen et al., 2004). Theoretical molecular weights and isoelectric points of the mature serpin proteins were calculated using the ProtParam tool via the ExPASy website (http://www.expasy.org/tools/protparam.html) (Gasteiger et al., 2003).
A multiple sequence alignment of amino acid sequences of A. gambiae serpins, human alpha-1 antitrypsin (HsA1AT), and Manduca sexta serpin-1K (MsSPN1K) was made using TCOFFEE (http://www.tcoffee.org/) with default parameters (Notredame et al., 2000). Conserved residues involving structural conformation of serpins were identified based on the amino acid residues of HsA1AT according to Irving et al. (2000). Secondary structures were predicted based on alignment with those identified in the crystal structure of MsSPN1K (Li et al., 1999).
Multiple sequence alignments of A. gambiae serpins, D. melanogaster serpins (DmSpn4A, 6, 27A, Nec), and M. sexta serpins (1K, 2, 3, 4A, 5A, 6) were created using TCOFFEE (default parameters) and MUSCLE (default parameters) (Edgar, 2004). The two alignments were compared, and a region of low similarity at the amino termini of the sequences was manually removed. The remaining portion was re-aligned using CLUSTAL embedded in the Mega 3.1 program (Kumar et al., 2004) to obtain an alignment file, which then was used for creating a neighbor-joining phylogenetic tree and bootstrap test. Gaps were excluded from the analyses, and bootstrapping with 1000 replications was performed.
Groups of mosquitoes at different developmental stages were homogenized, and total RNA was isolated using the Ultraspec RNA reagent (Biotecx Laboratories), followed by treatment with deoxyribonuclease I (Invitrogen) to remove genomic DNA. Five μg of total RNA was used as template for cDNA synthesis using the SuperScript First-Strand Synthesis System for RT-PCR (Invitrogen), and one μl of the resulting cDNA was used as a template for PCR. PCR reactions were performed using specific primers for each serpin (see Supplemental Table S2). The primers were designed using the web program Primer3 (http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi) and searched against the GenBank database of the A. gambiae sequences using BLASTN to verify specificity of each primer. Ribosomal protein S7 (RPS7) was used as an RT-PCR internal control. PCR products were analyzed by agarose gel electrophoresis and detected by ethidium bromide staining.
Additional RT-PCR experiments were done to confirm the existence of transcripts containing the alternative exons of SRPN4 isoforms. A forward primer (5′-AACGAGGACGAAGAAGACGA-3′) was located in exon 2, whereas reverse primers (5′-GCAGCACGACCATTACCTTT-3′ for SRPN4A; 5′-CAGGATGTTGGGAATCTTGC-3′ for SRPN4B; and 5′-CGACAAAGTCTAGCGGCTTC-3′ for SRPN4C) were located in the corresponding alternative exon3. The primers spanned at least one intron, depending on the location of exon3 of each isoform. PCR products were amplified with 25 cycles of 94 °C for 30 sec, 56 °C for 30 sec, and 72 °C for 40 sec. Adult female cDNA was used as a template for the PCR reactions.
In this study, we identified 18 serpin genes through BLAST searches of genome and EST databases. They are SRPN1-14, 16, 17, 18 and 19 (Table 1). We confirmed that previously described SRPN9 and SRPN15 are actually the same gene (explained below). Twelve of the serpin genes are present in the genome in four clusters of three genes each (Fig. 1). Two clusters, (SRPN7, 14, 18) and (SRPN1, 2, 3), are on chromosome arm 2L, one (SRPN11, 12, 17) on chromosome arm 2R, and one (SRPN5, 6, 16) on chromosome arm 3R. No serpin genes were found on 3L or the X chromosome (see Supplemental Table S2 for sequences GenBank accession numbers of the cDNAs and notes on annotation of the genes). Our results are generally consistent with the set of serpin genes identified by Waterhouse et al. (2007), but cDNA sequences we generated have improved gene models and predicted amino acid sequences for some members of the A. gambiae serpin family, described below.
We identified possible ESTs for A. gambiae serpins in the NCBI database and obtained clones for those available from MR4. However, the sequences of some of those cDNA clones were incomplete. No ESTs were found for SRPN4A, 7, 13 or 16-19. Some of these were serpins expressed mostly in the egg, larval, and/or pupal stages (see below), and thus were absent from the EST database, which included mainly cDNA clones from adult A. gambiae. For our cDNA collection, EST clones for 10 serpin genes were obtained from the MR4 repository. These cDNAs were from libraries made from adult A. gambiae Giles mosquitoes. The cDNAs for those serpins not available as ESTs from MR4 were cloned by RT-PCR, using specific primers designed from the gene predictions (see Supplemental Table S1 for details).
Alternative exon splicing has been identified in several insect species (Brandt et al., 2004; Hegedus et al., 2008; Jiang et al., 1996; Krüger et al., 2002;). The alternative splicing leads to serpin isoforms that differ in the carboxyl-terminal region that includes the reactive site loop. In A. gambiae, SRPN10 was previously determined to exhibit alternative splicing with the presence of four isoforms (Danielli et al., 2003). In our study, cDNA and genomic sequence analyses indicated that SRPN4 consists of three exons with three alternative forms of exon3 (Fig. 2A). This leads to three splicing isoforms of SRPN4, namely SRPN4A, 4B and 4C. In order to confirm the occurrence of alternative splicing of exon3, we conducted RT-PCR using a forward primer located in exon2 and reverse primers located in each exon3, with an expectation of different sizes of PCR products for each isoform. The results clearly verified the presence of the three isoforms (Fig. 2B). The initially predicted SRPN4 is SRPN4C. ESTs for SRPN4B and 4C, but not 4A were found in sequence databases. Additionally, detection of the SRPN4A transcript required a higher number of amplification cycles than the other two isoforms to reach comparable amplification levels (Fig. 5). These data suggest that transcripts for the SRPN4A isoform are rare in comparison with the 4B and 4C isoforms. In Ae. aegypti, four comparable splicing isoforms of the gene orthologous to SRPN4 were identified (Waterhouse et al., 2007). Functional differences among these isoforms remain unknown.
The initial gene set from the genome sequence included SRPN9 and SRPN15 (Christophides et al., 2002), whose predicted transcripts were nearly identical except for a 90 bp region present in SRPN15, but not in SRPN9. SRPN15 was annotated as a putative haplotype of SRPN9, and Waterhouse et al. (2007) eliminated SRPN15 from the serpin gene set. In our study, this extra sequence region found in the original SRPN15 was clearly present in the experimental cDNA sequence. In addition, we used PCR to amplify SRPN9/15 cDNA and A. gambiae genomic DNA templates, using primers that flanked the region in question. Both template samples produced one single product (data not shown), and the sequence of the amplified genomic DNA was identical to that of the cDNA. It appeared that a stretch of bases called as “N” in the genomic sequence data led to incorrect exon-intron boundary prediction at the 3′ end of the third exon, resulting in an incorrect sequence for the previously annotated SRPN9. Our findings confirm that SRPN9 and 15 are the same gene, and that the sequence of the predicted SRPN15 transcript contains the correct exon boundaries. We suggest that this gene should now be named SRPN9, and that the name SRPN15 should not be used further, to avoid future confusion.
SRPN13 was present in the original gene set (ENSANG00000013014) but was not found in the current genome assembly. There were no ESTs for SRPN13 (other than our cDNA clone). This might be because it was expressed primarily in eggs and very young larvae (Fig. 5), for which there were few ESTs. SRPN14 and 16 were novel among the A. gambiae serpin genes in being composed of a single exon. Comparisons of our cDNA sequences with gene models led to identification of incorrect predicted exon borders in SRPN7, SRPN9, and SRPN19 (Supplemental Table S3). In SRPN19, only one of the five exons was correctly predicted in the gene model.
General features of the A. gambiae serpin amino acid sequences are summarized in Table 1. Sixteen of the 18 serpins were predicted to be secreted proteins, based on detection of a putative secretion signal peptide. SRPN10 and SRPN12 lack a secretion signal sequence. SRPN10 was known from experimental results to be intracellular (Danielli et al., 2003; 2005). The putative secretion signal sequences ranged from 16 to 35 residues, with little sequence similarity among different serpin genes. For nine of the serpins, the amino-terminal residue of the mature protein was predicted to be glutamine. Serpins are generally 350-500 amino acids long, with molecular masses of 40-60 kDa (Gettins, 2002; Silverman et al., 2001). Tthe length of most mature serpin proteins in A. gambiae varies from 373 to 598 amino acid residues, with theoretical molecular masses of about 42-66 kDa with one exception, SRPN4A, which is unusually large. The length of the SRPN4A mature protein is 808 residues with a molecular mass of 90.8 kDa. This is probably the largest serpin known to date.
Most of the A. gambiae serpins have calculated pI between 5 and 6. However, SRPN4A, 4B, 5, 6, 13 and 17 are notably more basic, with calculated pIs of 8.4-9.7. The ranges in size and pI in the A. gambiae serpins are wider than those of serpins identified in M. sexta and D. melanogaster. It is interesting to speculate that the differences of serpins among these insects may be due to the variations of their feeding behavior (blood vs. non-blood feeding), ecology (aquatic vs. terrestrial habitats of specific life stages), and the types of pathogenic microbes that these insects encounter in their environment.
Serpins are usually formed from three β sheets (named A, B, C) and nine α-helices (also named alphabetically). Serpins are metastable proteins that undergo a large conformational change as they inhibit a protease. The protease cleaves a peptide bond in the exposed reactive center loop (RCL), and then the RCL inserts as a new strand into the A β-sheet (Silverman et al., 2001). Structural regions essential for conformational changes of serpins are the hinge, at which the chain bends to permit RCL insertion, the breach at the top of the A β-sheet, and the shutter, near the center of the A β-sheet, both of which must open to allow RCL insertion (Irving et al., 2000; Whisstock et al., 2000). An additional region known as the gate is involved in a structural transition (latency) in which RCL insertion occurs in the absence of RCL cleavage (Stein and Carrell, 1995). In order to examine these conserved regions and potential structural variations, we aligned the A. gambiae serpin sequences along with M. sexta serpin-1K (Protein Database Bank (PDB) file 1SEK) and human alpha-1 antitrypsin (HsA1AT, PDB file 1QLP_A), which are inhibitory serpins whose structures have been determined by X-ray crystallography (Elliott et al., 1998; Li et al., 1999). As expected, most amino acid residues known to form important structural regions in HsA1AT were also conserved in the mosquito serpins, with the exceptions of SRPN13 and SRPN19, in which these positions were notably less conserved (Supplemental Fig. S1). Residues at several positions were identical across species, but variations were also present, including some particularly large insertions between predicted secondary structural elements in some of the sequences. An insertion of 30-90 residues between helix D and strand 2A occurs exists in SRPN4, 5, 6, 16 and 19. This region includes a very acidic stretch of 6-9 uninterrupted Asp and Glu residues in SRPN4, 5, 6 and 16, which perhaps could be important as a binding site. SRPN12 and 19 have an insertion (36 and 31 residues, respectively) between helix F and strand 3A. The SRPN4 isoforms each have a greatly extended sequence in the loop between helix I and strand 5A (367 residues in SRPN4A). This region, within the alternatively spliced variable region of the SRPN4 variants, may eventually provide a clue to the functions of the SRPN4 isoforms. A much shorter insertion at the same position also occurs in SRPN16. SRPN13 is unique in containing a 39 residue insertion in the RCL region, and thus must have an unusual RCL compared with most other serpins.
The consensus pattern of specific residues in the hinge region can be used as a theoretical guideline to identify serpins that are inhibitory (Hopkins et al., 1993; Irving et al., 2000). We created an alignment of the hinge region of all A. gambiae serpins with known inhibitory serpins from M. sexta and D. melanogaster. HsA1AT was used as a reference. Indeed, the alignment showed that most of the hinge residues were conserved among the serpins (Fig. 3). SRPN13 and 19 have the least number of conserved residues in the hinge region, supporting a prediction that they may not function as inhibitors. We also predicted the position of the scissile bond of the A. gambiae serpins (Fig. 3), based on the alignment with insect serpins with known inhibitory properties. The putative P1 residues in the A. gambiae serpins were: Lys (SRPN1, 2, 8, 12, 14,19); Phe (SRPN3, 10C, 13, 18); Arg (SRPN4A, 4B, 4C, 5, 6, 7, 9, 10A, 10B, 16); and Leu (SRPN11, SRPN17). The P1 residue in SRPN10A was originally predicted to be Lys (Danielli et al., 2003). However, the alignment of the hinge and RSL region showed that SRPN10A exhibited similarity with DmSpn4A, which was experimentally verified to have Arg-Ala as P1-P1′ residues (Oley et al., 2004). Therefore, it is likely that the P1 residue in SRPN10A is also Arg. Determining which A. gambiae serpins are actually inhibitory and identification of the scissile bond in their RCLs will require further biochemical analysis, which can be accomplished by producing recombinant proteins from the cDNA templates of the mosquito serpins reported here.
Phylogenetic trees of subsets of the A. gambiae serpins were published previously (Christophides et al., 2002; Michel et al., 2005). In this study, we used all 18 of the A. gambiae mosquito serpins for a phylogenetic analysis. A neighbor-joining phylogenetic tree of the A. serpins was constructed, and known inhibitory serpins from M. sexta and D. melanogaster were included in the analysis. The tree showed seven phylogenetic clusters among the serpins, with SRPN12, 13 and 19 not grouped with any serpins (Fig. 4). SRPN1, 2 and 3 were grouped with MsSPN3A and DmSpn27A, which are inhibitors of prophenoloxidase (PPO)-activating proteases (PAPs) (De Gregorio et al., 2002; Zhu et al., 2003). Our previous experimental data demonstrated that recombinant SRPN1 and 2 can function as inhibitors of M. sexta PAP-3 (Michel et al., 2006). SRPN8 was grouped with MsSPN4 and 5, which are inhibitors of PPO activation (Tong and Kanost, 2005). SRPN9 was grouped with MsSPN6, which is also an inhibitor of M. sexta PAP-3 (Wang and Jiang, 2004; Zou and Jiang, 2005). SRPN10A was grouped with DmSpn4A and DmSpn6, along with MsSPN1, MsSPN2 and DmNec, all known to be functional serine protease inhibitors (Gan et al., 2001; Han et al., 2000; Jiang and Kanost, 1997; Levashina et al., 1999; Richer et al., 2004). Other phylogenetic clusters were: SRPN11 and 17; SRPN4, 5, 6 and 16; and SRPN7, 14 and 18. With the exception of SRPN12, the mosquito serpins that shared phylogenetic clusters are also located in gene clusters on the chromosomes (Fig. 1), as expected if they are products of fairly recent gene duplications. SRPN8 and 9 are of particular interest because they may be orthologous with Manduca serpins of known functions that have been characterized biochemically. Future characterization of recombinant A. gambiae serpins will be important for assessing their biochemical properties and biological functions in mosquitoes.
We examined the developmental expression profiles of A. gambiae serpins using semi-quantitative RT-PCR. This technique allowed us to determine if the serpins were expressed and at what life stages. We found that most of these serpins were expressed in all developmental stages (Fig. 5A). However, SRPN13 was expressed predominantly in eggs (developing embryos) and very young larvae. SRPN13 is similar to the ovalbumin family of clade-B serpins (data not shown), and its high expression in the eggs and very young larvae may point to a role during early development. SRPN7 and 12 transcripts were highly abundant in egg, larvae and tan pupae, whereas SRPN5 transcript was more abundant in black pupae (which are technically pharate adults) and adults. SRPN8 expression was high in eggs and larvae. SRPN14 transcript was more abundant in adults, whereas SRPN19 expression was not detected in adults. We reexamined the expression of serpins whose transcripts were not detected in adults by carrying out RT-PCR analysis of adult and pupal RNA samples at a higher number of amplification cycles. The results revealed that SRPN7 and 12 transcipts were present at a low level in black pupae and adults, whereas SRPN13 expression was not detected at all in these two stages (Fig. 5B). SRPN19 transcript was detected in pupae but not in adults.
In the A. gambiae adults, the expression level of SRPN10B transcript has been shown to be the most abundant of the four SRPN10 isoforms (Danielli et al., 2003). All SRPN10 isoform transcripts were also found to be abundant in the dissected adult midguts, in comparison with the thorax and the gut-free abdomen. The expression levels of SRPN10 proteins were very low in embryos and early instar larvae, very high in last instar larvae, and then decreased in pupae and adults, with the expression level in females higher than in males (Danielli et al., 2003). The differences of expression levels of serpins among life stages may imply different physiological roles in the mosquito.
Eighteen serpins have been identified in the mosquito A. gambiae. In addition to SRPN10 gene previously found to exhibit alternative splicing, SRPN4 with three isoforms as a result of alternative splicing of the last exon has been identified in this study. Some of the A. gambiae serpins are expressed fairly uniformly at all developmental stages, whereas others have unique patterns of stage-specific expression. cDNAs of the serpins were cloned, sequenced, and complete coding sequences were generated. The amino acid sequences, were aligned with serpins from other species, showing the conserved residues important in structural conformation required for inhibitory activity. The consensus patterns of the hinge region, along with the phylogenetic analysis data, suggest that most of the A. gambiae serpins are inhibitory. However, additional biochemical and biological studies will be required to determine their functions. The available cDNA information will permit production of recombinant proteins, allowing exploration of their functions in vitro and in vivo. The information will also be useful in further molecular genetic analyses in this medically important mosquito.
We thank Dr. Maureen Gorman for providing for providing some of the RNAs and cDNA pools that were used in this study and for technical advice, and we thank Ms. Sandi Yungeberg for maintaining the mosquito colony. We also thank MR4 for providing us with Anopheles gambiae cDNA clones originally contributed by Robert A. Holt. This work was supported by NIH grants AI31084 and GM41247, and contribution 07-124-J from the Kansas Agricultural Experiment Station.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.