A comprehensive gene expression atlas of sex- and tissue-specificity in the malaria vector, Anopheles gambiae

Abstract

Background: The mosquito, Anopheles gambiae, is the primary vector of human malaria, a disease responsible for millions of deaths each year. To improve strategies for controlling transmission of the causative parasite, Plasmodium falciparum, we require a thorough understanding of the developmental mechanisms, physiological processes and evolutionary pressures affecting life-history traits in the mosquito. Identifying genes expressed in particular tissues or involved in specific biological processes is an essential part of this process.

Results: In this study, we present transcription profiles for ~82% of annotated Anopheles genes in dissected adult male and female tissues. The sensitivity afforded by examining dissected tissues found gene activity in an additional 20% of the genome that is undetected when using whole-animal samples. The somatic and reproductive tissues we examined each displayed patterns of sexually dimorphic and tissue-specific expression. By comparing expression profiles with Drosophila melanogaster we also assessed which genes are well conserved within the Diptera versus those that are more recently evolved.

Conclusions: Our expression atlas and associated publicly available database, the MozAtlas (http://www.tissue-atlas.org), provides information on the relative strength and specificity of gene expression in several somatic and reproductive tissues, isolated from a single strain grown under uniform conditions. The data will serve as a reference for other mosquito researchers by providing a simple method for identifying where genes are expressed in the adult, however, in addition our resource will also provide insights into the evolutionary diversity associated with gene expression levels among species.

Figure 1

Global expression coverage. (A) The proportion of probes giving at least 3 out of 4 mismatch calls in either male or female samples for tissue in the MozAtlas and FlyAtlas. (B) Tissue breadth. "House-keeping" genes were identified to have a tau-statistic under 0.15 (n = 909), and narrow expression a tau-statistic above 0.85 (n = 3446). Overall, only a third of genes were detected in all tissues.

Figure 2

Sexually dimorphic expression. (A) The proportion of sexually dimorphic expression in each tissue versus total sexual dimorphism. The ratio of female- to male-biased expression is provided (female:male). All ratios deviate significantly from equality (Chi-squared test; P < 0.05). (B) Hierarchical clustering of probes among tissues and sex with Euclidean Distance; female (red), male (blue). Branch support was estimated with 10,000 bootstrapped replicates. Expression enrichment against carcass for the (C) ovary and (D) testis. Significant gonad enrichment is highlighted in dark grey (ANOVA; M > 2; Q < 0.05). Inset figures show the proportion of gonad enrichment which is also sexually dimorphic in whole-body samples, i.e. 51% female-biased (dark red); 7% male-biased expression (dark blue).

Figure 3

Tissue-specific expression patterns. (A) Tissue specific transcription was investigated on the basis of probe detection (3 out of 4 mismatch calls). Instances where probes were detected in a single tissue and a single sex are also highlighted: Female (red), male (blue). (B) SNP A/S ratio for genes with tissue-specific expression. 95% C.I. was estimated with 10,000 bootstrapped replicates.

Figure 4

Gene expression for major chromosome arms. (A) Germline only expression; testis (blue), ovary (red). (B) Somatic only expression; male (blue), female (red). (C) Germline only X vs autosomal SNP A/S ratio. (D) Somatic only X vs autosomal SNP A/S ratio. (Chi-squared Test; *P < 0.05). Grey bars represent expected proportions. 95% C.I. was estimated with 10,000 bootstrapped replicates.

Figure 5

Orthology Relationships. (A) The oldest common ancestor in gene-families to either a Dipteran, Coleopteran, Hymenopteran or Metazoan ancestor. (B) Expression divergence of tissues for one-to-one orthology pairs (n = 4234). Euclidean distance was used to calculate similarity among tissues within and between species. Branch support was estimated with 10,000 bootstrapped replicates. Drosophila (grey); Anopheles (black). (C) Mean expression of Anopheles orthologue clusters and (D) mean expression of Drosophila orthologues clusters. Mean relative expression (RA) level for each cluster according to grayscale. (E) The number of overlapping orthologous genes between Anopheles and Drosophila expression clusters calculated with a hypergeometric probability distribution after multiple correction. Light grey (P < 0.05); Dark grey (P < 0.01).

Figure 6

Gene copies, family origins and tissue expression. (A)Anopheles gene expansions with restricted expression patterns (n = 325; tau-statistic = 1). (B) Single-copy gene-families with narrow spatial expression profiles in both Drosophila and Anopheles tissues (n = 143; tau-statistic > 0.85).