Ant genomics: strength and diversity in numbers

Authors


  • The first two authors contributed equally to this manuscript.

  • The authors share a passion for the genetics of caste determination and social complexity, behavioural ecology, and evolutionary history of ants. They are also interested in the development of molecular and genomic tools to further the understanding of ants and their roles in Earth's ecosystem. C.D.S. has worked on several insect genome projects and is interested in using ants to study the genetic and epigenetic components of behavioural disorders. C.R.S. is interested in the proximate and ultimate factors underlying division of labour in social insects. U.M. is interested in the behaviour, evolution, ecology, microbiology, and natural history of social insects. J.G. is an evolutionary biologist interested in the genetic basis of species differences and speciation in Hymenoptera and the proximate and ultimate mechanisms of genetic variability in social insects.

  • MEETING REVIEW

Christopher Smith; E-mail: smithcd@sfsu.edu

Abstract

A recent workshop held at the Arizona State University Center for Social Dynamics and Complexity gathered over 50 prominent researchers from around the globe to discuss the development of genomic resources for several ant species. Ants play crucial roles in many ecological niches and the sequencing of several ant genomes promises to elucidate topics ranging from the genetic basis for social complexity, longevity and behaviour to systems biology and the identification of novel antimicrobial compounds. Unlike other species, most ant genomes are being generated by individual labs and small collaborations without the annotation and computational resources that support prominent model organism genome databases such those for the fruitfly and roundworm. Attendees summarized their current progress and future plans for several ant genomes and discussed how best to coordinate the analysis and annotation of ant sequences to benefit the broad research interests of the social insect community.

The genomics revolution has had a major impact on understanding the molecular basis of development and disease, but has focused on only a few model organisms and humans. There are approximately 1.6 million described plants, animals and fungi on Earth, but fewer than 10 genomes of these species are well annotated and described. Reduced sequencing cost and improvement in comparative analysis now allow genomic tools to be developed for the many non-model species that impact fundamental aspects of the biosphere and our daily lives. For instance, the budding field of sociogenomics (Robinson et al. 2005) aims to explain complex interactions and emergent social behaviours by understanding processes from the bottom-up (from genes to societies). The molecular mechanisms underlying complex societies are only beginning to be understood, largely as a result of genome resources that were developed first for the honeybee (Honeybee Genome Sequencing Consortium 2007). Ants are another ecologically and functionally diverse group of species whose genomes promise to answer key biological questions ranging from behaviour to biomedicine to ecosystem health. The diversity in social organization, behavioural and phenotypic plasticity, morphology, communication and ecology in the ants is unrivaled and make ants one of several emerging model taxa (Smith et al. 2009). To develop a coordinated research agenda, a workshop on ant genomics was held on 16–17 January 2009 at the Center for Social Dynamics and Complexity at Arizona State University. More than 60 scientists from the leading ant research laboratories in both North America and Europe met to coordinate the financing, labour, sequencing and annotation of a phylogenetically representative (Fig. 1, Table 1) set of ant genomes within the next 3–5 years. A list of workshop participants and/or laboratories involved in ant genomics discussions can be found at the ant genomics portal (http://www.antgenomics.org).

Figure 1.

 Ant forms and phylogeny. (A) Morphological diversity in the ants, from right to left, a queen (centre) and different sized workers of a carpenter ant; a minor and major worker of marauder ants (the major is 500× as massive as the minor); and a major and minor of army ants. (B) Ants funded for large-scale genomic or transcriptomic sequencing put into a phylogenetic framework (phylogeny modified from Brady et al. (2006)); the left tree is of the major ant lineages whereas the right tree is a more detailed version of the myrmicine branch which includes the most intensively studied taxa. The ants pictured are those funded, they are: 1 = Harpegnathos saltator (jumping ant), 2 = Linepithema humile (Argentine ant), 3 = Camponotus floridanus (carpenter ant), 4 = Pogonomyrmex barbatus (harvester ant), 5 = Solenopsis invicta (fire ant) and 6 = three species of attine fungus gardening ants from the genera Atta, Acromyrmex and Apterostigma (the fungi and other symbionts are also scheduled to be sequenced). All photos © Alex Wild, used with permission.

Table 1.   Summary of ant genome and transcriptome projects
SpeciesProjectResearch group(s)Sequencing technologyFunding source(s)Progress
  1. Numbers in parentheses correspond to numbers in Fig. 1.

  2. †Members include M. Goodisman, J. Heinze, K. Ingram, K. McCubbin, P. Pamilo, C.-J. Shih, W.-J. Wu, S. Y.

  3. ‡AR and J Leenards Foundation, Swiss National Science Foundation, Rub Foundation, Agassiz Foundation, Herbette Foundation, Chuard-Schmid Foundation, University of Lausanne, USDA-CSREES-AFRI.

  4. §National Institute of Mental Health, San Francisco State University, University of California Berkeley.

Solenopsis invicta (5)EST
Genome
Keller L
Shoemaker D
Fire Ant Genome Consortium†
Sanger Illumina 454Multiple Sources‡>10× genome coverage
Annotated EST library
Linepithema humile (2)Transcriptome
Genome
Smith CD
Tsutsui ND
454 Titanium
(some paired-end)
Multiple Sources§7× genome coverage
Mixed life stage transcriptomes
Harpegnathos saltator (1)
Camponotus floridanus
(3)
Transcriptome
EST Genome
Berger S
Liebig J
Reinberg D
Illumina, SangerHoward Hughes Medical Institute>90× genome coverage Annotated EST library
Pogonomymex barbatus (4)Transcriptome
Genome
Gadau J
Smith CR
454 Titanium (some paired-end) and SOLiDNational Science Foundation>10× genome coverage
Atta cephalotes
Apterostigma dentigerum

14 fungal & bacterial species (6)
Transcriptome
Genome
Meta-genome
Clifton S
Currie C
Gerardo N
Slater S
Suen G
Taylor J
Weinstock G
454 Paired-end TitaniumRoche PharmaceuticalsUnknown

Genomic information on ants promises to benefit diverse research fields well beyond those focusing on social insects. For example, large-scale transcriptome microarray (Wang et al. 2008) and candidate gene ‘macroarray’ (Goodisman et al. 2005; Graff et al. 2007) approaches with fewer markers have already shed new light on processes ranging from caste development to ageing. These studies have built on molecular studies of single genes and gene networks implicated in ant behaviour and caste differentiation (Tian et al. 2004), evolutionary development (Abouheif & Wray 2002), aggression (Lucas & Sokolowski 2009), reproduction (Khila & Abouheif 2008, Lu et al. 2009) and forageing behaviours (Ingram et al. 2005). Indeed, recent honeybee studies have shown that DNA methylation differences are associated with the developmental switch that drives caste development (Elango et al. 2009). Below we explore a short list of the most exciting questions that could be immediately addressed with multiple ant genomes in hand.

Why ants?

Ants are dominant components of all ecosystems, exceeding in biomass most other animal groups in tropical and temperate environments (Hölldobler & Wilson 1990). In addition to providing essential ecosystem services, ants are some of the most ecologically and economically costly invasive species. Ants are also an incredibly diverse group with 14 602 described species (http://www.antweb.org) representing diverse morphologies (Fig. 1), life histories and ecologies. Their evolutionary and ecological success is chiefly because of their social organization and complex division of labour. Thus, the development of genomic tools for ants will inform a diverse range of fields and major research areas (several are described below) ranging from medicine, to development, to behaviour and evolutionary ecology.

The predominant strength and weakness of ant genomics are one-in-the-same, diversity. While comparative genomics in ants can address many fundamental questions, developing genomic resources for many species is not only costly, but consensus over the most appropriate species is difficult to reach. With few funding agencies willing to back a single large project (10–20 genomes), a noncentralized strategy emerged with different groups independently seeking funding (Table 1). A common goal of sharing resources and facilitating research within the ant genomics community was agreed upon. Furthermore, meeting attendees agreed on the need for database and annotation standards similar to those adopted for the fruitfly (Drosophila 12 Genomes Consortium 2007), honeybee and wasp genome projects. As described below, progress is rapid and funding for several ant genomes and transcriptomes has already been acquired and is currently being put to use (Table 1).

What genomic resources are becoming available for ants?

As a nongenetic model system, ants lack many of the basic genomic tools available for other model organisms. Two reviews published in 2008 (Goodisman et al. 2008; Smith et al. 2008) discussed a relative paucity of ant genomic tools and active research. While many resources are not yet publicly available, sequence data are rapidly being generated and the number of genomic studies on ants has multiplied several fold (Table 1) (Wurm et al. 2009).

Fortunately, most ants have moderately sized genomes (200–600 Mb), making it feasible to sequence many species at relatively high coverage (i.e. at least 6×) (Tsutsui et al. 2008). EST libraries and genome sequencing is underway for seven species (Table 1, Fig. 1), with sequences more tentatively planned for other species. It is expected that by 2011 these resources will be made public, assembled and will have preliminary computational gene annotations. Most of the new genome and transcriptome sequences will be generated using ‘next generation’ sequencing, especially the platforms available from Roche (454), Illumina (Solexa) and ABI (SOLiD). While the obvious advantage to this approach is enough raw sequence to represent a high percentage of the genome, such assemblies suffer from fragmentation into numerous short contigs that are difficult to map to full length chromsomes. Another issue is that techniques to merge data sets from different sequencing platforms into a single assembly are just being developed. Furthermore, as nongenetic model systems, most ants lack markers that would aid the physical mapping of contigs onto chromosomes. Nonetheless, the planned assemblies will yield sequence of high quality that is sufficient for gene finding, promoter analyses, SNP detection and development of microarray platforms. Improvements to the assembly of ant genomes will be facilitated by the sequencing of haploid males and the generation of linkage maps from SNPs of brothers.

Current and potential insights from ant genomics and transcriptomics

Integration of behavioural genomics, developmental genomics and sociomplexity

Like the honeybee, ants have a complex division of labour with many behavioural castes (e.g. nurse ants, guards, foragers). Unlike the honeybee, however, many ant lineages have independently evolved worker subcastes that differ remarkably in morphology (Fig. 1A). Despite nearly identical genome sequences, the worker castes in such polymorphic societies differ in adult morphology as much as highly diverged species. Studying the differentiation of behavioural and morphological castes at the transcriptome level will provide insights into the mechanisms of diverse types of plasticity, from developmental to neurological systems. Moreover, ants have a functional DNA methylation system that, like the honeybee (Elango et al. 2009), is probably involved in caste differentiation (Kronforst et al. 2008; Kucharski et al. 2008) making them exciting systems to explore how epigenetic modifications affect the evolution of developmental and social processes.

Communication and systems biology

The ecological dominance of ants is based on efficient communication (largely chemical) between colony members. Understanding the genomics underlying kin recognition and territoriality should help devise control strategies of the most-successful ant invaders, whose success often depends on greater tolerance of conspecifics. Ant genomics will provide unique insights into the mechanistic basis of information flow, behavioural response thresholds and systems optimization of many interacting individuals. Furthermore, ant genomics will provide insights into the evolution of chemosensory and neuronal genes regulating social processes operating at multiple levels of organization (e.g. individual and colony).

Disease management in social environments

As social organisms living in dense colonies of related individuals, ants provide an opportunity to study epidemiological processes within a designed interaction network. The interactions of individuals (castes, work chains, etc.) as well as the architecture of the interaction space are presumably under selection to maximize colony efficiency while also minimizing the spread of diseases. Both of these can be studied at phenotypic and genomic levels. Furthermore, many ants are engaged in complex symbiotic interactions. The genomics of these interactions will give novel insights into the co-evolutionary dynamics of disease resistance, including the potential discovery of novel antifungal or antimicrobial peptides. Planned genome, transcriptome and metagenome sequences of leafcutter ants and their symbionts promised to shed considerable light on these interactions (Table 1).

Genetics of longevity, ageing and behaviour

Ant queens are among the longest-lived insects known (up to 30 years), often exceeding the average age of their workers or males by one or more orders of magnitude (Hölldobler & Wilson 1990). In contrast to most other species, ants show a positive correlation between reproduction and longevity (Schrempf et al. 2005). The adaptations permitting this astounding prolongation of life are not understood, but genomic comparison of the insulin signalling pathway and free radical metabolism genes between independent origins of sociality (e.g. honeybees and wasps) will shed light on the genetics and evolution of ageing and fertility.

Quantitative and reverse genetics

Although laboratory maintenance of many species has been achieved, controlled laboratory crosses have only been described for a few species. Genera such as Cardiocondyla, Harpegnathos and Monomorium, can be mated and propagated indefinitely in the laboratory, facilitating heritability and quantitative-genetics analyses through inbreeding, cross-breeding and backcrossing. Haplodiploid genetics will facilitate development of molecular maps and mapping polygenic traits, but will be limited until improved husbandry techniques are developed for most species. RNA interference has been successfully used to decrease gene expression in an ant (Lu et al. 2009) and multiple groups are working to develop transposable element (TE) or bacterial systems to overexpress or knock down gene targets. The creation of transgenic lines in other insect species via TEs (Pavlopoulos et al. 2004) suggests that these approaches should be well suited to ants and will help ameliorate the lack of traditional genetic tools.

Comparative genomics

Because of their ancient origin, well over 100 million years ago (Brady et al. 2006) and their remarkable subsequent diversification into more than 14 000 described species, genome-scale comparisons between ants offer a unique opportunity to observe what differences have driven the evolution of this hyperdiverse taxon. Comparison with the fruit fly, honeybee, pea aphid, flour beetle, silkworm moth and the jewel wasp genomes may aid in identifying DNA sequences associated with eusociality. Another outstanding question is whether the simple repeat and TE landscapes of ant genomes resemble eusocial honeybees (AT-rich sequence, few TEs) or solitary parasitoid wasps (AT-rich, diverse TEs). Such repetitive regions will further complicate genome assembly, but may also give insight into the life history of ants and suggest whether eusociality creates conditions that prune selfish DNA from the genome. Similarly, comparisons between judiciously chosen sister taxa may be particularly instructive, such as between an inquiline parasite (species existing exclusively as workerless queens that parasitize other ant species) and its unmodified sister lineage. Other ant species have lost the queen morph and form societies headed by reproducing workers. Such transitions to worker-less or queen-less ant species have evolved multiple times among ants, permitting identification of the common genomic footprints associated with these key changes in social organization.

Looking forward

Myrmecology has entered the genomic era, particularly comparative genomics, because of a long tradition of interpreting diversity within a comparative context. The scale of integrating genome knowledge across so many species is daunting and will require coordinated and standardized databases, new software tools, data-sharing policies and unprecedented cooperation. Undaunted, the ant genomics community has decided to follow the ancient wisdom, ‘look to the ant thou sluggard and consider her ways and be wise’ (Proverbs 6:6).

Acknowledgements

The authors are grateful to all workshop participants and hope to have captured the direction of the workshop and the collective wisdom of its participants through the text of this document. On behalf of all participants, the authors thank the Center for Social Dynamics and Complexity as well as the School of Life Science at ASU for hosting and financially supporting the workshop. Christopher D. Smith was supported by NIMH grant 5SC2MH086071. The authors would also like to thank the two anonymous referees for their comments on this manuscript. We also thank Shelley Berger, Nicole Gerardo, Laurent Keller, Juergen Liebig, Danny Reinberg, DeWayne Shoemaker, and Neil Tsutsui for sharing unpublished results.

Ancillary