Mosquito vitellogenin genes: Comparative sequence analysis, gene duplication, and the role of rare synonymous codon usage in regulating expression

Abstract

Comparative sequence analysis of mosquito vitellogenin (Vg) genes was carried out to gain a better understanding of their evolution. The genomic clones of vitellogenin genes were isolated and sequenced from all three subfamilies of the family Culicidae including Culicinae (Aedes aegypti, Ochlerotatus atropalpus, Ae. polynesiensis, Ae. albopictus, Ochlerotatus triseriatus and Culex quinquefasciatus), Toxorhynchitinae (Toxorhynchites amboinensis), and Anophelinae (Anopheles albimanus). Genomic clones of vitellogenin genes Vg-B and Vg-C were isolated from Ae. aegypti and sequenced. A comparison of Vg-B and Vg-C, with the previously characterized vitellogenin gene, Vg-A1, suggests that Vg-A1 and Vg-B probably arose by a recent gene duplication, and Vg-C apparently diverged from the two other members of the gene family in an earlier gene duplication event. Two vitellogenin genes orthologous to Vg-C were cloned from a Cx. quinquefasciatus DNA library, one of which is truncated at the N-terminal end. Single vitellogenin genes, orthologous to Vg-C, were cloned from the An. albimanus and Tx. amboinensis libraries. Incomplete sequences orthologous to Vg-B and Vg-C were isolated from the Oc. atropalpus library. Only partial sequences were isolated from Ae. polynesiensis, Ae. albopictus and Oc. triseriatus. Inferred phylogenetic relationships based on analysis of these sequences suggest that Vg-C was the ancestral gene and that a recent gene duplication gave rise to Vg-A1 and Vg-B after the separation of the genus Aedes. The deduced amino acid composition of mosquito vitellogenin proteins exhibits higher tyrosine and phenylalanine composition than other mosquito proteins except for the hexamerin storage proteins. Analysis of vitellogenin coding sequences showed that a majority of amino acid substitutions were due to conserved and moderately conserved changes suggesting that the vitellogenins are under moderately selective constrains to maintain tertiary structure. The vitellogenin genes of the three anautogenous mosquitoes, that require a blood meal to develop eggs, had very high synonymous codon usage biases similar to highly expressed genes of other organisms. On the other hand, the vitellogenin genes of autogenous mosquitoes, that develop at least one batch of eggs without a blood meal, exhibited low synonymous codon usage bias. An unusual pattern of synonymous codon usage was observed in the first 15 amino acid residues encoding the signal peptide in the vitellogenin genes, where a high number of rarely used synonymous codons are present. It is hypothesized that rare synonymous codons have selectively accumulated in the signal peptide region to down-regulate the rate of translation initiation in the absence of a blood meal. Real-time PCR gene expression experiments showed that all three Ae. aegypti vitellogenin genes were highly expressed after a blood meal, and expressed in non-blood-fed females, males, larvae and pupae at trace levels. Sequences were deposited in GenBank (accession numbers: Ae. aegypti Vg-B, AY380797, Vg-C, AY373377; Oc. atropalpus Vg-B, AY691321, Vg-C, AY691322; Ae. polynesiensis Vg-A1, AY691318, Vg-B, AY691319, Vg-C, AY691320; Ae. albopictus Vg-A1, AY691316, Vg-C, AY691317; Oc. triseriatus Vg-C, AY691323; Cx. quinquefasciatus Vg-C1, AY691324, Vg-C2, AY691325; Tx. amboinensis Vg-C, AY691326; An. albimanus Vg-C, AY691327).

Figure 1.

Generalized genomic organization of Aedes aegypti vitellogenin genes. The number in parenthesis indicates the number of amino acid residues.

Figure 2.

A multiple deduced amino acid sequence alignment of three members of vitellogenin family in Aedes aegypti. The conserved amino acid residues are capital letters. The conserved cysteine residues are marked with bold letters. An intron splice site is marked by Ï. Polyserine regions are underlined by dots (…). The symbol (ˆ) indicates a region where no nonsynonymous substitutions occur between Vg-A1 and Vg-B. The first set of arrows indicate the end of the signal peptide and the beginning of the small subunit. The second set of arrows indicate the position of the cleavage site where the ‘small’ subunit ends and the ‘large’ subunit begins.‡‡‡‡ indicate the cleavage sequence (RXRR) between the large and small subunits.

Continued

Figure 3.

Distribution of the compositional percentage of phenylalanine and tyrosine aromatic amino acid residues of Aedes aegypti proteins. The location of the vitellogenins and hexamerin proteins of Ae. aegypti are shown (Gordadze et al. 1999).

Figure 4 (Part 1).

Degree of nonsynonymous substitutions in pairwise comparison of the Aedes aegypti vitellogenin genes; (a) Vg-A1 and Vg-B, (b) Vg-A1 and Vg-B, and (c) Vg-B and Vg-C. The meaning of the bars is indicated above the figure.

(Part 2)

Figure 5.

Observed frequency of the first 50 codons in the Aedes aegypti vitellogenin genes, excluding the first initiator methionine. Green bars show the codons of the signal peptide, the first 11–14 amino acids, excluding the first methionine. Short bars indicate rare codons.

Figure 6.

Correlation of synonymous codon usage in 150 Aedes aegypti genes. ENC = effective number of codons, GC3 = GC content in the third position. Shown in red are the vitellogenin genes. r² = 0.5235.

Figure 7.

The correlation of synonymous codon usage bias and GC3 of Anopheles gambiae ribosomal genes. r² = 0.4606.

Figure 8.

The correlation of synonymous codon usage bias and GC3 of anautogenous and autogenous mosquito vitellogenin genes.

Figure 9.

Distribution of Aedes aegypti genes based on synonymous codon usage bias. ENC = effective number of codons. Vg-A1, 39.7, Vg-B 39.8. Vg-C 40.8.

Figure 10.

Expression of Aedes aegypti vitellogenin genes. mRNA levels were measured by real-time PCR.

Appendix 1

Alignment of the deduced amino acid sequences of mosquito vitellogenin genes. Conserved cysteine residues are highlighted. (…) indicate gaps inserted in the sequence. (∼∼∼∼) indicate missing sequences. Only partial sequences were obtained for Ae. albopictus, Oc. triseriatus, and Ae. polynesiensis.

Continued

Continued

Continued

Continued

Continued

Continued

Continued

Continued

Continued

Continued

Continued

Continued

Continued

Continued