Sparse and stereotyped encoding implicates a core glomerulus for ant alarm behavior.

Ants communicate via large arrays of pheromones and possess expanded, highly complex olfactory systems, with antennal lobes in the brain comprising up to ∼500 glomeruli. This expansion implies that odors could activate hundreds of glomeruli, which would pose challenges for higher-order processing. To study this problem, we generated transgenic ants expressing the genetically encoded calcium indicator GCaMP in olfactory sensory neurons. Using two-photon imaging, we mapped complete glomerular responses to four ant alarm pheromones. Alarm pheromones robustly activated ≤6 glomeruli, and activity maps for the three pheromones inducing panic alarm in our study species converged on a single glomerulus. These results demonstrate that, rather than using broadly tuned combinatorial encoding, ants employ precise, narrowly tuned, and stereotyped representations of alarm pheromones. The identification of a central sensory hub glomerulus for alarm behavior suggests that a simple neural architecture is sufficient to translate pheromone perception into behavioral outputs.

Simple inheritance, complex regulation: Supergene-mediated fire ant queen polymorphism.

The fire ant Solenopsis invicta exists in two alternate social forms: monogyne nests contain a single reproductive queen and polygyne nests contain multiple reproductive queens. This colony-level social polymorphism corresponds with individual differences in queen physiology, queen dispersal patterns and worker discrimination behaviours, all evidently regulated by an inversion-based supergene that spans more than 13 Mb of a "social chromosome," contains over 400 protein-coding genes and rarely undergoes recombination. The specific mechanisms by which this supergene influences expression of the many distinctive features that characterize the alternate forms remain almost wholly unknown. To advance our understanding of these mechanisms, we explore the effects of social chromosome genotype and natal colony social form on gene expression in queens sampled as they embarked on nuptial flights, using RNA-sequencing of brains and ovaries. We observe a large effect of natal social form, that is, of the social/developmental environment, on gene expression profiles, with similarly substantial effects of genotype, including: (a) supergene-associated gene upregulation, (b) allele-specific expression and (c) pronounced extra-supergene trans-regulatory effects. These findings, along with observed spatial variation in differential and allele-specific expression within the supergene region, highlight the complex gene regulatory landscape that emerged following divergence of the inversion-mediated Sb haplotype from its homologue, which presumably largely retained the ancestral gene order. The distinctive supergene-associated gene expression trajectories we document at the onset of a queen's reproductive life expand the known record of relevant molecular correlates of a complex social polymorphism and point to putative genetic factors underpinning the alternate social syndromes.

Refractive index measurement of the mouse crystalline lens using optical coherence tomography.

In recent years, there has been a growing interest for using mouse models in refractive development and myopia research. The crystalline lens is a critical optical component of the mouse eye that occupies greater than 50% of the ocular space, and significant increases in thickness with age. However, changes in refractive index of the mouse crystalline lens are less known. In this study, we examined the changes in thickness and refractive index of the mouse crystalline lens for two different strains, wild-type (WT) and a nyx mutant (nob) over the course of normal visual development or after form deprivation. Refractive index and lens thickness measurements were made on ex vivo lenses using spectral domain optical coherence tomography (SD-OCT). Comparison of refractive index measurements on 5 standard ball lenses using the SD-OCT and their known refractive indices (manufacturer provided) indicated good precision (intra-class correlation coefficient, 0.998 and Bland-Altman coefficient of repeatability, 0.116) of the SD-OCT to calculate mouse lens refractive index ex vivo. During normal visual development, lens thickness increased significantly with age for three different cohorts of mice, aged 4 (average thickness from both eyes; WT: 1.78 ± 0.03, nob: 1.79 ± 0.08 mm), 10 (WT: 2.02 ± 0.05, nob: 2.01 ± 0.04 mm) and 16 weeks (WT: 2.12 ± 0.06, nob: 2.09 ± 0.06 mm, p < 0.001). Lens thickness was not significantly different between the two strains at any age (p = 0.557). For mice with normal vision, refractive index for isolated crystalline lenses in nob mice was significantly greater than WT mice (mean for all ages; WT: 1.42 ± 0.01, nob: 1.44 ± 0.001, p < 0.001). After 4 weeks of form deprivation to the right eye using a skull-mounted goggling apparatus, a thinning of the crystalline lens was observed in both right and left eyes of goggled animals compared to their naïve controls (average from both the right and the left eye) for both strains (p = 0.052). In form deprived mice, lens refractive index was significantly different between the goggled animals and non-goggled naïve controls in nob mice, but not in WT mice (p = 0.009). Both eyes of goggled nob mice had significantly greater lens refractive index (goggled, 1.49 ± 0.01; opposite, 1.47 ± 0.03) compared to their naïve controls (1.45 ± 0.02, p < 0.05). The results presented here suggest that there are genetic differences in the crystalline lens refractive index of the mouse eye, and that the lens refractive index in mice significantly increase with form deprivation. Research applications requiring precise optical measurements of the mouse eye should take these lens refractive indices into account when interpreting SD-OCT data.

Co-inheritance of recombined chromatids maintains heterozygosity in a parthenogenetic ant.

According to Mendel's second law, chromosomes segregate randomly in meiosis. Non-random segregation is primarily known for cases of selfish meiotic drive in females, in which particular alleles bias their own transmission into the oocyte. Here we report a rare example of unselfish meiotic drive for crossover inheritance in the clonal raider ant, Ooceraea biroi, in which both alleles are co-inherited at all loci across the entire genome. This species produces diploid offspring parthenogenetically via fusion of two haploid nuclei from the same meiosis. This process should cause rapid genotypic degeneration due to loss of heterozygosity, which results if crossover recombination is followed by random (Mendelian) segregation of chromosomes. However, by comparing whole genomes of mothers and daughters, we show that loss of heterozygosity is exceedingly rare, raising the possibility that crossovers are infrequent or absent in O. biroi meiosis. Using a combination of cytology and whole-genome sequencing, we show that crossover recombination is, in fact, common but that loss of heterozygosity is avoided because crossover products are faithfully co-inherited. This results from a programmed violation of Mendel's law of segregation, such that crossover products segregate together rather than randomly. This discovery highlights an extreme example of cellular 'memory' of crossovers, which could be a common yet cryptic feature of chromosomal segregation.

Unselfish meiotic drive maintains heterozygosity in a parthenogenetic ant.

According to Mendel's second law, chromosomes segregate randomly in meiosis. Nonrandom segregation is primarily known for cases of selfish meiotic drive in females, in which particular alleles bias their own transmission into the oocyte1,2. Here, we report a rare example of unselfish meiotic drive for crossover inheritance in the clonal raider ant, Ooceraea biroi. This species produces diploid offspring parthenogenetically via fusion of two haploid nuclei from the same meiosis3. This process should cause rapid genotypic degeneration due to loss of heterozygosity, which results if crossover recombination is followed by random (Mendelian) segregation of chromosomes4,5. However, by comparing whole genomes of mothers and daughters, we show that loss of heterozygosity is exceedingly rare, raising the possibility that crossovers are infrequent or absent in O. biroi meiosis. Using a combination of cytology and whole genome sequencing, we show that crossover recombination is, in fact, common, but that loss of heterozygosity is avoided because crossover products are faithfully co-inherited. This results from a programmed violation of Mendel's law of segregation, such that crossover products segregate together rather than randomly. This discovery highlights an extreme example of cellular "memory" of crossovers, which could be a common yet cryptic feature of chromosomal segregation.

Ooceraeahainingensis sp. nov.: A new Chinese Ooceraea (Hymenoptera, Formicidae, Dorylinae) species with a dealate queen, closely allied to the queenless clonal raider ant O.biroi.

The clonal raider ant, Ooceraeabiroi, is a queenless species that reproduces asexually, and these traits make it an attractive model system for laboratory research. However, it is unclear where on the ant phylogeny these traits evolved, partly because few closely related species have been described and studied. Here, we describe a new raider ant species, Ooceraeahainingensis sp. nov., from Zhejiang, China. This species is closely related to O.biroi but can be distinguished by the following features: 1) workers of O.hainingensis sp. nov. have an obvious promesonotal suture and a metanotal groove, whereas these characters are ambiguous in O.biroi; and 2) the subpetiolar process of O.hainingensis is prominent and anteroventrally directed like a thumb with sublinear posteroventral margin, while in O.biroi, it is anteroventrally directed but slightly backward-bent. Molecular phylogenetic analyses confirm that O.hainingensis is genetically distinct from O.biroi. Importantly, unlike O.biroi, O.hainingensis has a queen caste with wings and well-developed eyes. This suggests that the loss of the queen caste and transition to asexual reproduction by workers is specific to O.biroi and occurred after that species diverged from closely related congeneric species.

Evolution: How sweat bees gained and lost eusociality.

Eusocial insects divide labor between reproductive and non-reproductive individuals. The molecular mechanisms underlying the evolution of these castes have remained mysterious. A comparative genomic study of sweat bees points to a familiar factor as a regulator of behavioral specialization: juvenile hormone.

A caste differentiation mutant elucidates the evolution of socially parasitic ants.

Most ant species have two distinct female castes-queens and workers-yet the developmental and genetic mechanisms that produce these alternative phenotypes remain poorly understood. Working with a clonal ant, we discovered a variant strain that expresses queen-like traits in individuals that would normally become workers. The variants show changes in morphology, behavior, and fitness that cause them to rely on workers in wild-type (WT) colonies for survival. Overall, they resemble the queens of many obligately parasitic ants that have evolutionarily lost the worker caste and live inside colonies of closely related hosts. The prevailing theory for the evolution of these workerless social parasites is that they evolve from reproductively isolated populations of facultative intermediates that acquire parasitic phenotypes in a stepwise fashion. However, empirical evidence for such facultative ancestors remains weak, and it is unclear how reproductive isolation could gradually arise in sympatry. In contrast, we isolated these variants just a few generations after they arose within their WT parent colony, implying that the complex phenotype reported here was induced in a single genetic step. This suggests that a single genetic module can decouple the coordinated mechanisms of caste development, allowing an obligately parasitic variant to arise directly from a free-living ancestor. Consistent with this hypothesis, the variants have lost one of the two alleles of a putative supergene that is heterozygous in WTs. These findings provide a plausible explanation for the evolution of ant social parasites and implicate new candidate molecular mechanisms for ant caste differentiation.

Joint Evolution of Asexuality and Queen Number in an Ant.

Ants exhibit a striking diversity of reproductive systems, varying in traits such as the number of reproductives per colony [1], the mode of daughter production (sexual or asexual) [2], and the mode of caste determination (genetic or environmental) [3]. Species employing mixed reproductive systems present a unique opportunity to explore the causes and consequences of alternative breeding strategies. Mixed reproductive systems in ants include social polymorphism in colony queen number, whereby single-queen (monogyne) and multiple-queen (polygyne) colonies co-occur within species [4-7], and facultative asexuality, in which female offspring may be produced sexually or asexually within colonies [8-13]. Here, we document a remarkable confluence of multiple mixed reproductive systems in the tropical fire ant, Solenopsis geminata, in a population with three important features: (1) polygyne colonies produce workers sexually but queens asexually, whereas monogyne colonies produce both castes sexually; (2) polygyne queens mate with monogyne males to produce workers, but monogyne queens do not mate with polygyne males; and (3) different asexual/polygyne lineages evidently were founded separately by genetically distinct founder queens, which appear to have originated from the same neighboring monogyne population. Multiple asexual/polygyne genomes are transmitted undiluted in this system, but sterile workers produced with sperm from a sexually-reproducing/monogyne population are necessary for the persistence of these lineages. The intersection of social polymorphism, facultative asexuality, and genetic caste determination marks this population of S. geminata as an embodiment of the diversity of ant reproductive systems and suggests previously unknown connections between these phenomena.