Introduction
Social interactions among conspecific individuals create a scenario in
which interacting phenotypes influence trait variation (Farine,
Montiglio, & Spiegel, 2015; Moore, Brodie, & Wolf, 1997). When the
phenotype of an individual is influenced by environmental variation
attributable to the genotypes of its social partners, this is termed an
indirect genetic effect (Moore et al., 1997). Because individuals rely
on social care to develop into adults in eusocial species, indirect
genetic effects can be a heritable, evolutionarily-responsive, source of
trait variation (Linksvayer, 2015; Linksvayer & Wade, 2005; Wolf,
Brodie, Cheverud, Moore, & Wade, 1998). Variation in aggression (Avalos
et al., 2020), size, and reproductive allocation (Linksvayer, 2006;
Vojvodic et al., 2015) have each been attributed to indirect genetic
effects in studies of eusocial insects.
Recent studies have uncovered many instances of supergenes (i.e., tight
clusters of two or more loci each affecting a different developmental or
behavioral characteristic), which provide integrated control of complex
adaptive phenotypes segregating within species (Schwander, Libbrecht, &
Keller, 2014; Thompson & Jiggins, 2014 ; Wellenreuther & Bernatchez,
2018). In eusocial insects, supergenes are frequently responsible for
intraspecific variation in social organization (Kay et al., 2022b).
Understanding the direct and indirect genetic effects of supergenes is
of particular interest in social organisms (Kay et al., 2022b) or in
situations where supergenes induce variation in mating systems (Mank,
2022).
In the fire ant Solenopsis invicta , indirect genetic effects of a
supergene underpin the behavioral regulation of colony queen number by
workers (Ross & Keller, 2002; Wang et al., 2013). In monogyne colonies,
a single reproductive queen resides with her worker offspring, while in
polygyne colonies multiple reproductive queens reside with workers
composed of multiple family groups (Keller, 1993). This variation in
queen number is associated with a suite of important individual- and
colony-level phenotypic differences (DeHeer, 2002; DeHeer, Goodisman, &
Ross, 1999; Gotzek & Ross, 2007; Ross & Keller, 1995; Ross & Keller,
1998). The supergene controlling this fundamental social polymorphism is
formed by three overlapping chromosomal inversions that result in high
levels of linkage disequilibrium between >500 protein
coding genes (Helleu, Roux, Ross, & Keller, 2022; Huang, Dang, Chang,
& Wang, 2018; Wang et al., 2013; Yan et al., 2020). The fire ant
supergene thus provides a system with prominent roles for direct and
indirect genetic effects in shaping variation in social organization.
Reproductive queens in monogyne colonies of S. invicta are all
homozygous for the Social B (SB ) supergene haplotype,
which has largely retained ancestral synteny, whereas reproductive
queens in polygyne colonies in the U.S. are all heterozygous forSB and the inverted supergene haplotype Social b(Sb ) (Ross, 1997; Wang et al., 2013; Yan et al., 2020).
Importantly, the presence of workers with the Sb haplotype
affects the behavior of SB/SB workers. When the queen is removed
from colonies containing only SB/SB workers, the workers accept
only a single replacement queen with the genotype SB/SB(SB/Sb and supernumerary queens are executed), but SB/SBworkers that coinhabit colonies with Sb -bearing workers will
accept multiple SB/Sb replacement queens (Gotzek & Ross, 2008;
Ross & Keller, 2002). Supergenes mediating polymorphic social
organization have now been discovered in at least four additional ant
lineages and it is likely that they play an important role underlying
variation in social organization in many other eusocial insects (Kay et
al., 2022b).
So far only a few studies have investigated the consequences of indirect
genetic effects on gene expression in social insects. In the honey beeApis mellifera , the mixing of workers with different genotypes
has been shown to have indirect effects on individual worker behavior
and transcription of genes in the brain (Gempe, Stach, Bienefeld, &
Beye, 2012). By contrast, no detectable indirect genetic effects were
detected in the clonal raider ant Ooceraea biroi when gene
expression was compared between workers exposed to social partners from
different clonal lineages (in this species thousands of genes are
differentially expressed between individuals of the different lineages)
(Kay et al., 2022a). In S. invicta , indirect genetic effects of
the social supergene have previously been investigated on several
occasions (Arsenault et al., 2020; Dang, Cohanim, Fontana, Privman, &
Wang, 2019; Wang, Ross, & Keller, 2008). In a first study using
microarrays of pooled whole bodies of workers (Wang et al., 2008), a
greater number of genes were influenced by indirect effects associated
with the Gp-9 genotypic composition of their colony than by the
direct effect of their own Gp-9 genotype (Gp-9 is a an
odorant binding gene located in the social supergene (Wang et al.,
2013)). In another study comparing gene expression of pools of antennae
from multiple workers there were similar numbers of genes influenced by
direct and indirect genetic effects (Dang et al., 2019). Finally, a
comparison of gene expression of two tissues of individual
pre-reproductive queens revealed substantially fewer genes
differentially expressed because of indirect genetic effects than direct
genetic effects within each tissue type (Arsenault et al., 2020). The
contribution of direct and indirect effects of the supergene on
transcriptional variation seems to vary according to biological context
and scale in S. invicta . Thus, an assessment of general trends
across castes and tissues is warranted to test for general properties of
direct versus indirect genetic effects in the system.
Here we present the first report of direct and indirect effects of the
fire ant supergene in brains and gasters of the worker caste of S.
invicta . We integrated the analyses of our data with previously
reported data for pre-reproductive queens (Arsenault et al., 2020) to
provide a comprehensive view of the direct and indirect genetic effects
of the supergene across castes (Figure 1). We examined direct genetic
effects using transcriptomes from brains and abdominal tissues ofSB/SB and SB/Sb workers and unmated winged queens (gynes)
from polygyne colonies. We examined indirect genetic effects of the
supergene using transcriptomes from adult SB/SB workers and gynes
that developed in a monogyne social environment (where all nestmate
workers have the genotype SB/SB ) and a polygyne social
environment (where nestmate workers are mix of genotypes SB/SBand SB/Sb ; Figure 1). Our results suggest that direct genetic
effects of the fire ant supergene result in larger gene expression
differences than indirect genetic effects and that direct genetic
effects are more consistent than indirect genetic effects across tissue
types and castes. We also conducted cross-fostering experiments of
late-stage pupae to investigate how the social environment affects
patterns of gene expression in adults. Previous studies revealed thatSB/SB workers shift from rejecting all SB/Sb queens to
accepting multiple SB/Sb queens when their colony contains more
than 10-15% Sb- bearing adult workers (Gotzek & Ross, 2008; Ross
& Keller, 2002). We raised these SB/Sb individuals until two
weeks post-eclosion to investigate the role of social environment during
the first weeks of adulthood. These cross-fostering experiments revealed
no significant indirect effects of the supergene on transcription when
the type of social environment experienced by sampled individuals
differed only during adulthood.