Evolution of germ plasm assembly and function among the insects

Abstract

Germ plasm is a substance capable of driving naive cells toward the germ cell fate. Germ plasm has had multiple independent origins, and takes on diverse forms and functions throughout animals, including in insects. We describe here recent advances in the understanding of the evolution of germ plasm in insects. A major theme that has emerged is the complex and convoluted interactions of germ plasm with symbiotic bacteria within the germline, including at the very origin of oskar, the gene required for assembling germ plasm in insects. Major advancements have also been made in understanding the basic molecular arrangement of germ plasm in insects. These advances demonstrate that further analysis of insect germ plasm will be fruitful in illuminating diverse aspects of evolutionary and developmental biology.

Introduction:

In animals, the specification of primordial germ cells is crucial in establishing the germline and occurs primarily through two mechanisms zygotic induction (cell-cell signaling) and maternal provision (inheritance of a maternally deposited germ plasm) [1]. Zygotic induction is suspected to be the ancestral mode, with repeated convergent evolution of the derived maternal provision mode within several phyla [1,2]. Both modes of specification are found among insects, although maternal provision has only been conclusively demonstrated within the Holometabola (insects that undergo complete metamorphosis), where the shift to maternal provision appears to be dependent upon co-option of the gene oskar for germline function [3] (Figure 1).

Figure 1.

Phylogenetic tree (based on [20]) showing major events in the evolution of insect germ plasm and the oskar gene.

The gene oskar is the primary determinant of the germline progenitors in Drosophila melanogaster [4,5]. It was discovered in a screen for maternal determinants of embryonic patterning and is a member of a set of genes whose mutants have defective abdominal patterning and a grandchild-less phenotype [5,6]. Among this class of genes, only oskar is both necessary and sufficient for establishing the germline, and is the critical nucleator of the fly germ plasm [4]. The step-wise assembly of germ plasm has been extensively described in D. melanogaster and it begins with the targeting and localization of oskar mRNA to the posterior cortex of the oocyte. Localization to the posterior pole releases transcripts from the translational repression of bulk cytoplasmic oskar mRNA, producing localized Oskar protein at the posterior pole. Oskar protein directly interacts with several mRNAs and proteins, resulting in their recruitment and anchoring to the posterior pole [7,8]. This function of Oskar leads to the nucleation of small ribonucleoprotein complexes, termed polar granules, that mediate packaging of mRNAs essential for downstream germline development into pole cells [9]. While oskar mRNA is initially required for proper germ plasm assembly through the action of its protein product, the message itself has been found to be toxic to pole cells [10]. To protect against this effect, oskar mRNA is restricted to a distinct subset of germ plasm granules, termed founder granules, that are organized around Staufen, rather than Oskar protein [10,11]. Segregation into these founder granules appears to mediate the compartmentalized degradation of oskar mRNA within the germ plasm prior to pole cell formation [10].

The importance of the germline to the survival of the species, and the apparent diversity in modes of specifying this crucial cell type, make the evolution of germline specification an important topic in evolutionary and developmental biology. Here we discuss recent results that shed light on the molecular evolution of Oskar, changes in composition of germ plasm, and novel interactions with endosymbiotes that influence germ plasm form and function.

The origin and evolution of Oskar among insects

oskar, unlike most of the other germline factors, initially had no obvious homologs in other model organisms, and no evident functional domains. The molecular origins of Oskar became clearer as more insect genome data became available. For example, the N-terminal region of the Oskar ortholog in the wasp Nasonia vitripennis showed strong similarity to what became known as the LOTUS (named after Limkain-Oskar-TUdor domain containing proteinS) domain [3,12]. A second region toward the C-terminus, termed the OSK domain, showed significant homology to bacterially derived proteins and did not appear in eukaryotic sequences [7]. These observations raised the possibility that oskar may have been partially derived from a horizontal gene transfer event [3].

This idea was recently rigorously tested and elaborated using in-depth and wide-ranging bioinformatic approaches. Using sensitive similarity searches, the authors identified domains with similarity to either the LOTUS or OSK domain throughout the tree of life [13]. When all of the LOTUS -like domains were aligned and subjected to phylogenetic analysis, it was revealed that LOTUS domains from Oskar orthologs clustered exclusively among Tudor domain-containing genes only found in eukaryotes [13]. This provides strong evidence that the LOTUS domain of Oskar was derived from the nuclear genome of the organism where Oskar first originated. In contrast, the OSK domain returned mostly prokaryotic sequences, and the Oskar derived domains clustered firmly among bacterial sequences in the phylogenetic analysis [13]. Additional stringent analyses concluded that the closest relatives of the OSK domains outside of Oskar proteins themselves are bacterial GDSL hydrolases, and that there is very little chance that an overlooked insect protein could have been the progenitor of the OSK domain [13]. This strongly suggests that the OSK domain entered an insect genome by horizontal transfer. Furthermore, many of the bacterial hydrolases related to Oskar come from prokaryotic lineages known to infect insect germlines, providing a clear pathway for a potential horizontal transfer [13].

While both domains of the Oskar protein have critical roles in binding proteins and RNA during the assembly of germ plasm particles in D. melanogaster and likely other holometabolous insects [7,8], the ancestral role of Oskar was almost certainly not in establishing the germline. An Oskar ortholog was discovered in the cricket Gryllus bimaculatus [14], which lacks a maternal germ plasm [15]. Lack of maternal germ plasm is ancestral to insects, and germ plasm is only widespread among the Holometabola [3]. Functional analysis showed that cricket Oskar is important for patterning the embryonic nervous system, and has no effect on germline determination [16]. Other studies have also found roles for Oskar, as well as other germline components, in the D. melanogaster nervous system [17–19]. This evidence suggests that the ancestral role for Oskar among insects may have been in the nervous system, and it was only relatively recently (300 million years ago [20]) co-opted into a role in maternal germline determination [3,14].

A recent study found that Oskar orthologs are present in the Ametabola, the sister group of the winged insects, suggesting that the oskar gene originated about 420 million years ago [20] in an insect lacking maternal germline determinants [21] (Figure 1). Thus, it appears this novel gene was preserved for around 65 million years prior to its recruitment to a germline role within the Holometabola approximately 345 million years ago [20]. This study also focused on tracing the evolution of Oskar presence and function throughout insects and found that oskar has been lost from several major insect lineages. Most notably, oskar was lost in the Hemiptera, where many sequenced genomes have been completed, as well as in the crown group Lepidoptera, encompassing most of the diversity in this highly speciose order, including the model species Bombyx mori and Danaus plexippus (Figure 1). They also identified several lineages where oskar had duplicated, and the extra copy was maintained [21]. The functional significance of these duplications will be an area of high interest in the future.

Differential patterns of molecular evolution in the LOTUS and OSK domains were also observed [21]. The LOTUS domain appears to be evolving at a similar rate between hemimetabolous and holometabolous insects, while the OSK domain appears to be more rapidly evolving among the Holometabola [21]. The nature of the selective pressure driving this pattern is yet unknown, but the pattern suggests that recruitment for a germline role in the ancestor of the Holometabola had a major impact at least on the RNA binding function of Oskar, since the OSK domain is thought to be the prime mediator of this function [22]. Finally, they also found that the novel Long-OSK isoform is found outside of the Drosophila genus, but is restricted to a specific clade of closely related Diptera [21] (Figure 1).

Evolution of germ plasm form and function

While oskar appears to be the crucial nucleator and organizer of germ plasm in Holometabolous insects [3], the identities of the other molecules incorporated into this organelle are less well known outside of D. melanogaster. A recent study used RNAseq of bisected embryos to identify the mRNA localized to the germ plasm of Nasonia vitripennis [23]. Only a few transcripts are shared between the N. vitripennis and D. melanogaster germ plasm, indicating that the mRNA composition of germ plasm may evolve rapidly and extensively among the Holometabola. One proposed reason for the extensive divergence in N. vitripennis is the unique form of its germ plasm. Unlike the collection of small granules tightly associated with the posterior embryonic cortex found in D. melanogaster, the functional equivalent of germ plasm in N. vitripennis consists of a single, extremely large particle, called the oosome, that migrates through the interior of the embryo before budding into the pole cells [24]. Consistent with this, many of the unique transcripts localized to the oosome have predicted functions that are not clearly related to specifying the unique features of germ cells. Rather, many are predicted to code for cytoskeletal components and even transmembrane proteins [23]. Several of the tested putative cytoskeleton/membrane protein coding transcripts were shown to be important for the maintenance of oosome integrity, indicating that they are important for the novel structure of the N. vitripennis oosome [23]. One of these codes for an ankyrin domain-containing protein (Nv-ooCLANK) and is part of a wasp-specific family of proteins that originated through horizontal transfer, likely from an intracellular symbiont [25]. This suggests that genetic interactions between intracellular prokaryotes and the germline are ongoing and contributing to the evolution of germ plasm.

An extreme, and evolutionarily informative, modification of germ plasm was recently described in ants [26]. There is an obligate symbiosis between the intracellular bacteria Blochmannia and the ant species Camponotus floridanus, where Blochmannia are essential for embryonic development and can be found concentrated at the posterior pole near the germ plasm [26]. While initially the C. floridanus germ plasm is reminiscent of other insects, i.e., tightly localized and densely packed with oskar mRNA, it later separates into distinct zones early in embryogenesis, which has not been previously observed. Some of these new zones correlate with specific fates later in development, such as cells that facilitate transmission of Blochmannia to the next generation and cells that transport Blochmannia to the gut, where they aid in extracting nutrition [26]. Elimination of Blochmannia through antibiotic treatment affected the mRNA and protein composition of all four zones, suggesting Blochmannia is integrated into a network regulating the germline gene expression in all novel germ plasm zones during embryogenesis.

Further evidence of integration of Blochmannia into a novel complex regulatory network comes from the Hox genes abdominal-A (abd-A) and Ultrabithorax (Ubx). These genes have broadly conserved roles in demarcating positional fates in animals, but until now have not been reported as germ plasm components in other species. In C. floridanus, one or both of them is found in each of the novel germplasm zones. abd-A and Ubx are lost in the novel germ plasm zones when Blochmannia are absent, suggesting they are downstream of Blochmannia. When the Hox genes are knocked down with RNAi, germline genes are disrupted in the germ plasm zones, suggesting that they are upstream of germline genes [26]. Thus, a complex web of interactions among germ plasm components and an intracellular prokaryote is required for development in C. floridanus[26].

A recurring theme of bacterial influence on insect germ plasm is becoming apparent, from the origin of oskar, the incorporation of prokaryotic genes into the Nasonia oosome, and the integration of Blochmannia into the novel network of germ plasm regulation in C. floridanus. Further examples come from the intracellular bacteria Wolbachia, which is one of the most widespread bacterial endosymbionts among insects [27]. For example, it has been shown that Wolbachia employ mechanisms that interact with germ plasm targeting machinery, which increase its vertical transmission in D. melanogaster [28]. Complementarily, Wolbachia are highly enriched around the germ plasm in wasps in the Nasonia genus [29]. The levels of accumulating Wolbachia were shown to be regulated differentially between species by a maternal factor under strong selection pressure [29]. Combined, these findings raise the possibility that interactions between the germ plasm and intracellular bacteria (symbionts, or potentially pathogens) play an important role in driving the form, function and composition of the germ plasm in insects.

Conclusions:

Insects provide an ideal model system for the study of germ plasm evolution because of increasingly available genomic information, high level of species diversity, and the presence of both maternal provision and zygotic induction in several holometabolous orders. The presence of both modes suggests multiple “reversions” from maternal provision to zygotic induction in the Holometabola, which can be used as natural experiments to understand the mechanisms and evolutionary pressures driving changes in germline specification modes.

There are also indications that there may be independent derivation of maternal provision modes. For example, molecular evidence indicates that the pea aphid Acyrthosiphon pisum has derived a form of germ plasm, despite a lack of oskar in its genome and a general absence of maternal provision in other Hemiptera [30–32] (Figure 1). Morphological analyses have suggested the presence of germ plasm in the hemipteran orders Thysanoptera [33] and Psocoptera [34]. Finally, a reemergence of maternal provision has been suggested morphologically and molecularly in the silkmoth B. mori [35] (Figure 1). Further analyses are required to determine whether these observations represent true germ plasm, and whether, or to what degree, cases of independently derived germ plasm exist among insects.

The many transitions between maternal provision and zygotic induction modes [36], suggests a high level of plasticity in insect germline specification that is not found in other animal groups. A potential source of germ plasm diversity discussed in this review is the interplay between host germ plasm and bacterial endosymbionts. A large number of insect species contain bacterial endosymbionts, which can be obligate, such as those that provide essential nutrients to their host, or facultative, including one of the most widespread endosymbionts Wolbachia [37,38]. Many of these endosymbionts closely associate with the insect germline, contributing to their own vertical transmission. In some cases, this close association can influence germline development, and provide an adaptive advantage, such as the Camponotus-Blochmannia interaction. It will be interesting to see how adaptive, and potentially maladaptive, interactions between insects and their bacterial endosymbionts have driven the evolution of the germ plasm form, function, and presence across the insect tree of life.