The monarch butterfly genome yields insights into long-distance migration

Abstract

We present the draft 273 Mb genome of the migratory monarch butterfly (Danaus plexippus) and a set of 16,866 protein-coding genes. Orthology properties suggest that the Lepidoptera are the fastest evolving insect order yet examined. Compared to the silkmoth Bombyx mori, the monarch genome shares prominent similarity in orthology content, microsynteny, and protein family sizes. The monarch genome reveals a vertebrate-like opsin whose existence in insects is widespread; a full repertoire of molecular components for the monarch circadian clockwork; all members of the juvenile hormone biosynthetic pathway whose regulation shows unexpected sexual dimorphism; additional molecular signatures of oriented flight behavior; microRNAs that are differentially expressed between summer and migratory butterflies; monarch-specific expansions of chemoreceptors potentially important for long-distance migration; and a variant of the sodium/potassium pump that underlies a valuable chemical defense mechanism. The monarch genome enhances our ability to better understand the genetic and molecular basis of long-distance migration.

Figure 1. Natural history of the monarch butterfly

A. Migration south. The eastern North American monarch butterfly undergoes a long-distance fall migration to a restricted site in central Mexico (yellow oval). The population of monarchs west of the Rocky Mountains undergo a truncated fall migration. Red arrows, flight paths. From Reppert et al., 2010. B. Journey north. Eastern migrants remain at the overwintering areas until spring, when the same butterflies reproduce and migrate northward to lay fertilized eggs on newly emerged milkweed in the southern United States (red arrows). Successive generations of spring and summer monarchs re-populate the home range (black arrows). From Reppert et al., 2010. C. Life cycle. Complete metamorphosis from egg to larva (5 instars) to pupa (chrysalis) to adult. The male butterfly (upper right) has visible black spots on their hind wings that are missing in females (lower left, underwing view). The larvae feed on milkweed (plants of the genus Asclepias). Photograph of lithograph, John Abbot, 1797. See also Figure S4 and Table S9.

Figure 2. Lepidopteran orthology and evolution

A. Orthology assignment of 12 insect and two mammal genomes. Bars are subdivided to represent different types of orthology relationships: 1:1:1 indicates universal single-copy genes, but absence and/or duplication in a single genome is tolerated; N:N:N indicates other universal genes, but absence in a single genome or two genomes within the different orders is tolerated; Diptera indicates dipteran-specific genes, presence in at least one mosquito and one fly genome; Lepidoptera indicates lepidopteran-specific genes, presence in both the monarch and Bombyx genomes; Hymenoptera indicates hymenopteran-specific genes, presence in at least one bee or wasp genome, and one ant genome; Insect indicates all other insect-specific orthologs; Mammal indicated mammalian-specific orthologs; Patchy indicates orthologs that are present in at least one insect and one mammal genome; Homolog indicates partial homology detected with E<10⁻⁵ but no orthology grouped; SD, species-specific duplicated genes; ND, species-specific genes. The phylogeny on the left was calculated using maximum likelihood analyses of a concatenated alignment of 1,642 single-copy proteins from the 1:1:1 subgroup. The tree was rooted using mammals as outgroup. Bootstrap values based on 1,000 replicates are equal to 1,000 for each node. B. The distribution of pairwise amino acid identity. Histogram shows the distribution of sequence identity of 8,221 1:1 orthologs between the monarch and Bombvx (diverged ~ 65 million years ago; (Grimaldi and Engel, 2005)). To highlight the similar level of molecular divergence, 8,897 orthologs between two ants (Linepithema humile and Pogonomyrmex barbatus, which diverged ~100–150 million years ago; (Moreau et al., 2006)), 6,875 orthologs between two mosquitoes (Anopheles gambiae and Aedes aegypti, which diverged ~150 million years ago; (Krzywinski et al., 2006)), and 6,520 orthologs between bee and wasp (Apis mellifera and Nasonia vitripennis, which diverged ~ 180 million years ago; (Werren et al., 2010)) were plotted in black, green, and red, respectively. C. Microsynteny between monarch and Bombyx genomes. Alignment of monarch scaffolds and silkworm chromosomes is shown by pairwise dot plots based on gene homology. 1,802 monarch scaffolds (>10 kb) were anchored to the corresponding position based on the consensus order of gene homology. The arrow denotes the position of a scaffold between the monarch and Bombyx that showed particularly strong microsynteny, which is magnified below. See also Figure S2.

Figure 3. Sun compass components focusing on the circadian clock

A. Model delineating the components used for sun-compass navigation. The compass mechanism involves the monarch eye sensing of skylight cues, including color gradient or the sun itself (violet, blue and green circles) and the polarization pattern of ultraviolet (UV) light (violet circle crosshatched), and the brain integration of skylight-cue stimulated neural response in the central complex (CC; grey dashed lines). In addition, time-compensation is provided by circadian clocks located in the antenna. The integrated time-compensated sun compass information is relayed to the motor system to induce oriented flight. The brain circadian clocks are located in the pars lateralis (PL) and communicate with the pars intercerebralis (PI). The PL may also communicate with the central complex. Modified from Reppert et al., 2010. B. Maximum-likelihood phylogenetic tree of insect vertebrate-like opsins (pteropsins). The tree was rooted using the monarch UV, blue and long-wavelength opsins. C. Schematic of the proposed clockwork mechanism in the monarch butterfly, including the core transcriptional/translational feedback loop (thick arrows) and the modulatory feedback loop (dashed arrows), both incorporating monarch orthologs of all described Drosophila clock genes (Dubruille and Emery, 2008). CLOCK (CLK) and CYCLE (CYC) heterodimers drive the transcription of period (per), timeless (tim), and type-2 cryptochrome (cry2), which upon translation form complexes and 24 hours later cycle back into the nucleus where CRY2 inhibits CLK:CYC-mediated transcription. Light entrainment is mediated by type-1 cryptochrome (CRY1), which promotes TIM degradation. Casein kinase II (CKII), doubletime (DBT) and the protein phosphatase 2A (PP2A) are involved in the posttranslational modifications of PER and TIM, and supernumerary limbs (SLIMB) and jetlag (JET) signal their degradation. The gene(s) involved in CRY2 degradation are unknown (red question mark). The modulatory feedback loop regulates the expression of clock through VRILLE (VRI) and PDP1. Monarch vri has five consensus CACGTG E-box elements in its promoter and pdp1 has five in its first intron, through which CLK and CYC could drive their transcription; each transcription factor also contains PAR DNA binding domains that could modulate CLK transcription by binding to PAR-like binding sites present in the monarch clk promoter. Clockworkorange (CWO) modulates the amplitude of the clock. D. Pdh expression in monarch brain. Top, The primers used for RT-PCR amplification are shown (red arrows) on a schematic of the pdh locus identified from the genome assembly. (Right) Agarose gel showing the pdh amplicon migrating at ~250bp (red arrow). Bottom, Alignment of monarch PDH peptide sequence with those previously described in insects. E. Insights into the evolution of the arthropod circadian clock. The presence/absence of the core clock components period (Per), type-2 cryptochrome (Cry2), type-1 cryptochrome (Cry1), timeless (Tim) and timeout (Tim2) has been assessed in all the published arthropod genomes, including 17 insect species. The presence of a given gene is represented by a colored box and the numbers represent the number of copies found in the genome. The question mark represents an absence that may be due to a degree of incompleteness in the genome given the presence of this gene in all others ant genomes. See also Figure S3, Table S2, Table S3, Table S4, Table S5 and Table S7.

Figure 4. Juvenile Hormone regulatory pathway

A. Proposed endocrine regulation of reproductive arrest and longevity in migratory monarch butterflies. Decreasing daylength (decrease in sun size) is sensed by circadian clocks in the pars lateralis (PL). This information is relayed to the pars intercerebralis (PI) in which the production and/or secretion of insulin-like peptides is decreased resulting in a decrease juvenile hormone (JH) biosynthesis in the corpora cardiac-corpora allata (CC-CA) complex. JH deficiency affects target tissues, leading to reproductive quiescence and increased longevity. B. JH biosynthetic and degradation pathways (Belles et al., 2005). Enzymes in the monarch genome are in blue. Gray denotes enzymes proposed to catalyze Farnesyl-PP to farnesoic acid, but their actual existence has not yet been verified. C. Sexually dimorphic pattern of JH biosynthesis. Bars indicate relative gene expression levels of enzymes, as annotated in B, in migrants relative to summer counterparts (data calculated from GSE14041 of GEOdatabase); n=5 per sex per enzyme). Significance was estimated by two-sample t-tests. *, p<0.05; **, p<0.01; ***, p<0.001. See also Table S6.

Fig. 5. miRNAs in summer and migratory monarchs

A. Expression of monarch miRNAs. The relationship between the expression values in reads per million (RPM) of summer butterflies and migrants plotted on a natural logarithmic (log) scale. Each of the 116 identified miRNAs is represented as a colored dot (blue, conserved; green, lepidopteran specific; red, monarch specific). Inset is a pie chart showing the classification of non-coding small RNA sequencing reads from the merged summer and migratory samples. B. miRNAs expressed differentially between a pool of 10 summer butterflies and a pool of 10 migratory monarchs. Top, bars show the mean miRNA levels that were up-regulated ≥ 1.5-fold in migrants compared to summer butterflies. The normalized expression value (in RPM) of migrants is displayed above each bar. Bottom, mean miRNA levels up-regulated in summer butterflies compared to migrants. See also Table S8.

Figure 6. Insights into chemosensory function in the monarch butterfly

A. Three-dimentional reconstruction of right antennal lobe (AL) of a female migrant monarch. Each glomerulus is highlighted with a unique color without physiological significance. B. Unrooted tree of candidate monarch (Dp, red lines) and Bombyx mori (Bm, blue lines) olfactory receptors (Ors). Green boxes, monarch-specific expansions. Purple box, pheromone receptor candidates. BmOR2 was renamed as BmORCO. ORl, OR-like; ORc, OR candidate. C. Phylogenetic relationship of the monarch antennal ionotropic receptor (IR) candidates with Bombyx and Drosophila antennal IRs. Red lines, monarch; blue, Bombyx; green, Drosophila. The monarch IR names in red represent genes present in the monarch genome but not in Bombyx. D. Phylogenetic relationship of the monarch gustatory receptor (Gr) candidates with Bombyx and Drosophila Grs. Red, monarch; blue, Bombyx; green, Drosophila.