The rich somatic life of Wolbachia

Abstract Wolbachia is an intracellular endosymbiont infecting most arthropod and some filarial nematode species that is vertically transmitted through the maternal lineage. Due to this primary mechanism of transmission, most studies have focused on Wolbachia interactions with the host germline. However, over the last decade many studies have emerged highlighting the prominence of Wolbachia in somatic tissues, implicating somatic tissue tropism as an important aspect of the life history of this endosymbiont. Here, we review our current understanding of Wolbachia–host interactions at both the cellular and organismal level, with a focus on Wolbachia in somatic tissues.

Perhaps because CI operates with the combined effects of Wolbachia in the male and female germlines, and Wolbachia is transmitted through the latter, much of the work on Wolbachia has focused on its interaction with the host germline. However, work over the past decade has reinforced original observations that, in addition to localization to the germline, a conserved feature of Wolbachia infection is localization to somatic tissue. Equally significant, these studies have begun to shed light on the functional importance of tissue-specific somatic localization of Wolbachia. Here, we review the current state of knowledge regarding the somatic aspects of Wolbachia infection and the functional consequences for both host and endosymbiont.

| AN OVERVIEW OF DISTRIBUTION IN SOMATIC TISSUES
In the original description of Wolbachia, the authors describe not only a concentration of "rodlike organisms in the reproductive tissue but also in the somatic tissue" (Hertig & Wolbach, 1924). Since these initial descriptions, numerous researchers have documented the presence of Wolbachia in a variety of somatic tissues (Table 1). In fact, the few examples in the literature where Wolbachia is restricted to the reproductive tissues, such as certain strains of the mosquito Aedes albopictus and female Glossina morsitans tsetse flies (Dobson et al., 1999), appear to be the exception rather than the rule. PCR and fluorescent cytological approaches have been used to assay for the presence of Wolbachia, with both techniques revealing a broad distribution in specific somatic cells and tissues (Table 1) In brief, Wolbachia is prevalent in tissues of the nervous system in Drosophila and other flies (Albertson et al., 2013;Casper-Lindley et al., 2011;Dobson et al., 1999;Mitsuhashi, Saiki, Wei, Kawakita, & Sato, 2002;Moreira et al., 2009;Osborne, Leong, O'Neill, & Johnson, 2009;Strunov & Kiseleva, 2014). In Drosophila, the distribution of the pathogenic Wolbachia strain, wMelPop, in the nervous system of adults is temperature dependent, with increased temperature favoring the expansion of Wolbachia from the central brain to peripheral areas such as the optic lobe and retina (Strunov, Kiseleva, & Gottlieb, 2013).
The repeated observation of Wolbachia in specific somatic tissues suggests that somatic tissue tropism is not incidental, but rather a key aspect of Wolbachia biology. For instance, somatic localization of Wolbachia may be evolutionarily maintained because it aids horizontal transmission within and between species, thus serving as a mechanism to increase the genetic diversity of Wolbachia. Additionally, somatic Wolbachia may confer advantageous phenotypes in the host that enhance its germline transmission. Below, we further explore the mechanisms and functional significance of the somatic localization patterns of Wolbachia.

| Arthropods
As in many insect species, the Drosophila egg chamber consists of a syncytium of 15 nurse cells and an oocyte, all connected through cytoplasmic bridges (Spradling, 1993). During maturation, nurse cell cytoplasm is pumped into the oocyte. Importantly, specific determinants essential for anterior-posterior (AP) axis formation are also transported from the nurse cells to the specific regions of the maturing oocyte. Localization of these AP axis and germline determinants requires microtubules, microtubulebased motor proteins and association with posterior cortical cytoskeletal elements (Chang et al., 2011). Meanwhile, efficient transmission of most Wolbachia strains from one generation to the next requires that the bacteria concentrate at the posterior pole of the mature oocyte, as this is the future site of the germline (Kose & Karr, 1995). Thus, Wolbachia must migrate from the nurse cells to the posterior pole, navigating the constantly changing and tumultuous environment of the developing oocyte due to cytoplasmic streaming (Monteith et al., 2016). However, some strains, such as Wolbachia Riverside (wRi) of D. simulans F I G U R E 1 Wolbachia distribution in somatic tissues. Wolbachia has been detected by PCR and fluorescent cytology in various somatic tissues of numerous (A) fly, (B) mosquito, and (C) filarial nematode species, as indicated in green incorporate into the pole cells independently of posterior concentration by maintaining a high titer throughout the entire oocyte (Serbus & Sullivan, 2007;Veneti, Clark, Karr, Savakis, & Bourtzis, 2004), whereas others (wNo, wMa, wKi) maintain a predominantly anterior localization (Veneti et al. 2004). These differences may ultimately contribute to differential somatic localization in adult flies.
Functional studies in Drosophila demonstrate that Wolbachia movement through the nurse cells to the anterior pole of the oocyte relies on the minus-end directed motor protein dynein (Ferree et al., 2005). At this point in oogenesis the oocyte microtubules switch orientations such that transport to the posterior pole requires plus-end directed microtubule movement. It has been difficult to attribute a functional significance of this dramatic switch in microtubule orientation, as well as cytoplasmic streaming. It may be that these are defense mechanisms preventing germline transmission of microbial invaders. Accordingly, Wolbachia rely on the plus-end directed motor protein for transport and concentration at the posterior pole (Serbus & Sullivan, 2007). Finally, stable association with the posterior cortex requires key germ plasm and AP axis components such as Staufen and Oskar (Serbus & Sullivan, 2007). Thus, germline transmission of Wolbachia requires a sophisticated developmentally controlled association with dynein, kinesin, and finally conserved posterior determinants. Phylogenetic analyses of Wolbachia that vary in their niche tropism demonstrate that Wolbachia-encoded factors are required for the posterior concentration (Toomey, Panaram, Fast, Beatty, & Frydman, 2013). One possibility is that Wolbachia expresses a developmentally programmed set of surface proteins that facilitates sequential engagement with host dynein, kinesin, and finally pole plasm determinants.
In all insect species examined, there is also a significant fraction of Wolbachia that is not associated with the posterior cortex but remains dispersed throughout the oocyte, as shown in Figure 2 (Veneti et al., 2004). During the syncytial divisions following fertilization, these bacteria concentrate at the centrosomes and undergo cell-cycle regulated movements along the spindle and astral microtubules associated with the dividing syncytial nuclei (Albertson, Casper-Lindley, Jian, Tram, & Sullivan, 2009;Kose & Karr, 1995). As in the oocyte, it is likely this movement relies on the microtubule-based motor proteins dynein and kinesin (Ferree et al., 2005). The functional significance of these movements is unclear. One possibility is that it serves to distribute Wolbachia throughout the embryo such that they will fate map to numerous developmental lineages. Thus, as with the oocyte, the final distribution of the Wolbachia throughout the cellularized embryo prior to gastrulation is determined by a combination of host and Wolbachia factors.
Examination of Drosophila larva reveals that, as embryonic development progresses, Wolbachia also concentrates in the embryonic and larval epithelial-derived neuroblast stem cells (Albertson et al., 2009). In contrast to the symmetric segregation of Wolbachia in the syncytial divisions, Wolbachia in the neuronal lineage exhibits a highly asymmetric segregation pattern (Albertson et al., 2009). The dividing neuroblast produces a self-renewing neuroblast daughter cell and a daughter cell that will differentiate into larval neurons. Wolbachia almost exclusively segregates with the neuroblasts with only a few bacteria localizing to the cells that will differentiate into larval neurons. This asymmetric localization and segregation is largely dependent on the robust astral microtubules associated with the selfrenewing neuroblast cell. Larval neuroblast cells undergo a period of quiescence and ultimately divide and differentiate into the cells that will become the adult central nervous system (Homem & Knoblich, 2012). Thus, the asymmetric neuroblast localization during the larval divisions ensures their eventual localization to the adult brain (Albertson et al., 2013).
Unfortunately, we know little about Wolbachia localization during the pupal stages. However, numerous studies that have examined its cellular and tissue distribution in the adult stages. These are described in section 2 (an overview of distribution in somatic tissues) as well as Table 1 and Figure 1.

| Filarial nematodes
As with arthropods, Wolbachia is inherited primarily through the female germline in filarial nematodes (Kozek, 1977). In insects, axis determination and the site of germline formation is established during oogenesis. In filarial nematodes an asymmetric MTOC is also present before fertilization, in contrast to the model nematode

Caenorhabditis elegans. Posterior localization of Wolbachia in both
Drosophila and filarial nematodes relies on microtubules and motor proteins. During the establishment of polarity in the filarial nematode Brugia malayi, Wolbachia is associated with high levels of dynein, and dynein is required for their posterior localization (Landmann et al., 2014). Therefore, in both insects and filarial nematodes, microtubules and motor proteins are required for suggesting that it relies on conserved signaling factors associated with these species and perhaps others (Landmann et al., 2012).
During the subsequent L3 and L4 larval stages, the hypodermal chords become syncytial through a process of cell fusion. Following this, Wolbachia proliferate extensively and spread anteriorly to fill the chord. In order to infect the germline, Wolbachia then migrate from the chord into the germline, crossing multiple plasma membranes. Images demonstrate that Wolbachia achieves this in female worms through the depolymerization of actin-based microfilaments at the point of somatic-germline cell contact (Landmann et al., 2012). Surprisingly, Wolbachia does not invade the germline in male nematodes, indicating Wolbachia is responding to signaling molecules specific to the female germline. Thus, in the late larva and adult males, Wolbachia is exclusively localized in the hypodermal lineage, whereas in females, Wolbachia resides in the hypodermal and germline lineages (Fischer et al., 2011;Landmann et al., 2012).

| MOLECULAR MECHANISMS OF MIGRATION AND INVASION OF SOMATIC CELLS
The studies described above indicate that in insects and nematodes the adult somatic distribution of Wolbachia is largely determined by a combination of symmetric and asymmetric segregation patterns during the mitotic divisions and cell-to-cell migration (Albertson et al., 2009;Landmann et al., 2010Landmann et al., , 2012. With respect to the segregation patterns in both systems, it is clear that microtubules play a key role. Live imaging of the syncytial cortical divisions in Drosophila reveal Wolbachia maintains a tight association with the centrosome during interphase, but once the cell enters mitosis, Wolbachia undergoes extensive movement along pole to pole and astral microtubules (Albertson et al., 2009;Kose & Karr, 1995).
Based on studies in the oocyte, this is likely to be driven by the microtubule-based motor proteins dynein and kinesin (Ferree et al., 2005). This movement facilitates the even distribution of Wolbachia to daughter nuclei and serves to distribute them throughout the embryo. This is similar to what occurs in B. malayi, where Wolbachia moves along the astral and spindle microtubules during mitosis, facilitating their migration (Landmann et al., 2010). Wolbachia also relies on cortical microtubules and dynein to localize to the posterior cortex in B. malayi (Landmann et al., 2014).
The structural mechanisms by which Wolbachia engages host motor proteins and how this is regulated remain unknown. Sequence analysis reveals the Wolbachia genome contains several outer membrane proteins (WSPs, Wolbachia surface proteins) and these are likely to play a role in interacting with host cytoskeleton (Wu et al., 2004). However, electron microscopy has revealed that Wolbachia is encompassed by a host membrane (Callaini, Riparbelli, & Dallai, 1994;Fischer, Beatty, Weil, & Fischer, 2014) (Cho, Kim, & Lee, 2011). In further support of an endocytosis hypothesis, free Wolbachia is able to invade uninfected germline tissues of Anopheles mosquitoes when the two are cocultured ex vivo (Hughes, Pike, Xue, & Rasgon, 2012). In these experiments, Wolbachia more efficiently invades tissues of their native hosts as opposed to those of more divergent ones. This suggests that Wolbachia enter cells through a receptor-mediated mechanism that can be affected by polymorphisms in specific proteins that arise during speciation. These potential mechanisms are of particular importance to the finding that Wolbachia localizes to the somatic niche cells of the female germline in many Drosophila species (Fast et al., 2011;Toomey et al., 2013). Studies in which Wolbachia bacteria are injected into the adult abdomen demonstrate that Wolbachia can hone to these regions through migration (Frydman et al., 2006). How they achieve this remains unclear, as they must traverse a number of membrane and extracellular matrix barriers. However, receptor-mediated endocytosis into specific cell types after movement through the hemolymph is a plausible route.
Despite these intriguing lines of evidence, the role of the endocytic pathway in Wolbachia infection remains largely unexplored.
Such phylogenetic analyses provide clues to the most plausible routes of horizontal transmission. Horizontal transmission appears to take place within and between species through both direct and indirect interactions. For example, intraspecies horizontal transmission in organisms such as fruit flies and spiders likely happens through direct contact or the environment, given the ecological roles of these organisms do not allow for a vectored mechanism Haine, Pickup, & Cook, 2005). Likewise, interspecies horizontal transfer in intertidal amphipod crustaceans (Cordaux et al., 2001) and butterflies sharing the same habitat probably occurs through the environment (Dyson, Kamath, & Hurst, 2002). In plant-feeding pumpkin arthropods, Wolbachia transfer appears to be linked to feeding on particular leaf substrates (Sintupachee, Milne, Poonchaisri, Baimai, & Kittayapong, 2006) whereas in mycophagous Diptera, the mushroom habitat appears to play a role in horizontal transmission (Stahlhut et al., 2010).
Whether horizontal transmission is a common occurrence on shorter time scales remains uncertain, though studies tracing Wolbachia movement among bee populations suggest it is an infrequent event (Gerth, Rothe, & Bleidorn, 2013 Drosophila by feeding on infected corpses and subsequently being ingested by uninfected flies (Brown & Lloyd, 2015). In colonies of Cubitermes termites, the exchange of salivary secretions, also known as trophallaxys, appears to facilitate intraspecies transfer of Wolbachia between individuals of different castes (Roy, Girondot, & Harry, 2015).
Thus, a similar route may be involved in other social insects. For example, in Acromyrmex ants, Wolbachia is present in the fat body, hemolymph, and feces, suggesting the potential for fecal-oral transmission (Frost, Pollock, Smith, & Hughes, 2014 (Grenier et al., 1998;Kageyama, Narita, & Noda, 2008;Pigeault et al., 2014;Van Meer & Stouthamer, 1999). Though infection intensity appears to decline over time, in some cases stable germline infection can be achieved through injection (Grenier et al., 1998;Van Meer & Stouthamer, 1999). For example, Wolbachia injected into the abdomen of Drosophila can migrate to the germline (Frydman et al., 2006). Thus, one possible mechanism for natural horizontal transmission is through contact of an uninfected wounded individual with infected hemolymph from a wounded Wolbachia host, as has been demonstrated in woodlice (Rigaud & Juchault, 1995 that a species barrier to horizontal transmission appears to exist, but that this can in some cases be overcome both in nature as described above, but also in the laboratory. For instance, the establishment of

| EXTRACELLULAR SURVIVAL AND ROUTES OF TRANSMISSION
The ability of Wolbachia to transfer horizontally between organisms suggests that the bacterium is capable of surviving in an extracellular environment, though this idea is somewhat controversial. In the laboratory, Wolbachia has been isolated from both, infected cell cultures and tissues (Gamston & Rasgon, 2007;Rasgon, Gamston, & Ren, 2006). While Wolbachia obtained in this manner can be maintained in cell-free medium and retain viability for at least a week, no replication is apparent. Nonetheless, these results indicate that Wolbachia is able to survive at least for a limited time outside of host cells. However, the fact that Wolbachia lack the ability to synthesize many essential lipids (Wu et al., 2004) (Rigaud & Juchault, 1995). Meanwhile, in the nematode B. malayi, extracellular Wolbachia are found in the pseuodocoelom, indicating that perhaps pseudoscoelomic fluid serves as a route for Wolbachia transfer between germline and somatic tissues (Fischer et al., 2014), similar to hemolymph in insects.
In addition to surviving in the hemolymph, Wolbachia has been observed extracellularly in various other important host tissues where it can exert both beneficial and harmful effects with respect to the host. For instance, while Wolbachia has been shown to concentrate in the central brain and optic lobe with little detriment (Albertson et al., 2013), studies show that some virulent Wolbachia strains can exit these cells, perhaps through cell lysis, and invade the extracellular space in the brain, causing pathogenesis (Min & Benzer, 1997;Strunov & Kiseleva, 2014).  (Espino et al., 2009). These gut bacteria may provide essential metabolic pathways lacking from the insects, thereby controlling various aspects of host physiology and life history, and perhaps contributing to pathogen resistance.
Furthermore, in C. lectularius, Wolbachia resides within a highly specialized organ called the bacteriome (Hosokawa et al., 2010). The bacteriome is composed of bacteriocytes, a cell type similar to fat cells.
These are maternally transmitted and serve primarily to protect endosymbiotic bacteria in exchange for nutrients. In this case, it appears Wolbachia may also be acting as a nutritional mutualist. Indeed, removal of endogenous Wolbachia from these bedbugs reduced host growth and reproductive fitness through a mechanism dependent on biotin synthesis (Nikoh et al., 2014).
Infection in extracellular compartments and the tissues discussed above may not only be important for horizontal transmission, but may also explain the various effects of Wolbachia on host physiology that appear to be independent of the germline. Across diverse taxa, the gut is a key tissue for regulating immunity, metabolism, and longevity.
Likewise, the brain regulates these and other central processes while also controlling behavior. Thus, it is possible that the digestive tract is not only a route for Wolbachia transfer between hosts, but also, along with the brain, involved in the functional consequences of Wolbachia infection that are discussed below.

| THE FUNCTIONS OF SOMATIC INFECTION
In the mature oocyte, Wolbachia concentrates at the posterior pole facilitating its incorporation into the germline of the developing host embryo. In Drosophila and other insects, however, a large fraction of Wolbachia is also positioned anteriorly resulting in a distribution throughout the length of the embryos (Ferree et al., 2005;Serbus & Sullivan, 2007;Veneti et al., 2004). This Wolbachia fraction is not incorporated into the germline and fate maps to the somatic cells of the developing insect. In filarial nematodes,

| Effects on host behavior
There are many examples illustrating that vertically transmitted endosymbionts influence host behavior (Goodacre & Martin, 2012).
Presumably these behavior modifications have evolved to enhance transmission of the endosymbiont. Over the past decade, a number of publications demonstrate that Wolbachia also has profound effects on insect behavior. This is likely a consequence of Wolbachia localization in the central nervous system and fat bodies, as they are hormone sources and influence physiology and behavior (Albertson et al., 2013;Arrese & Soulages, 2010;Nassel, 1993).
In addition to mating behavior, feeding patterns appear to change during infection, as blood feeding success is reduced in Wolbachiainfected mosquitoes (Turley, Moreira, O'Neill, & McGraw, 2009). While in this particular case, reduced feeding is not associated with reduced olfaction, other studies have found that Wolbachia can reduce host responsiveness to olfactory food cues (Peng, Nielsen, Cunningham, & McGraw, 2008). Changes in locomotor activity, also induced by Wolbachia infection in Drosophila, may contribute to apparent behavioral alterations (Caragata et al., 2011;Evans et al., 2009). While the mechanisms that underlie the phenomenon of behavioral change are undetermined, Wolbachia likely gain from altering essential host behaviors.
Most prominently, changes in reproductive behavior may drive the spread of infection through populations by favoring the production of infected females. Similarly, changes in feeding behavior could confer a fitness advantage for infected individuals. For instance, in mosquitoes blood feeding is a costly behavior that can reduce fitness (Murdock, Moller-Jacobs, & Thomas, 2013).
Many conclusions on the effects of Wolbachia on insect behavior must be treated with caution because the unaffected control insects are often obtained through antibiotic-based curing of Wolbachia.
Antibiotic treatment is certain to have profound effects on the composition of the gut and other host microbe populations (Broderick & Lemaitre, 2012). In addition, antibiotic treatment of Drosophila not infected with Wolbachia has dramatic long-term effects on behavior and physiology, including mitochondrial function and lifespan (Albertson et al., 2013;Ballard & Melvin, 2007). Significantly, these effects persist many generations after the exposure to antibiotics (Albertson et al., 2013). Therefore, it is difficult to attribute the changes in behavior specifically to the loss of Wolbachia, despite the fact that most researchers attempt to control for this by curing several generations in advance of experimental manipulation. Given these issues, multiple generations of backcrossing is the preferred method of creating uninfected controls from infected insect lines when possible.

| Effects on host metabolism
Wolbachia localization to the fat body, a key endocrine tissue in insects (Arrese & Soulages, 2010), has been observed on numerous occasions. The Wolbachia genome encodes an array of proteins that may be involved in regulating metabolism (Darby et al., 2012).
This includes several facilitators of cation membrane transport that provide essential cofactors for enzymes in the respiratory chain.
Furthermore, in filarial nematodes, Wolbachia can directly influence the expression of host enzymes involved in glucose and glycogen metabolism . Therefore, it is unsurprising that Wolbachia increases the basal metabolic rate of infected mosquitoes as measured by the production of carbon dioxide (Evans et al., 2009). In Drosophila, Wolbachia also influence host iron-utilization, whereas in C. lectularius Wolbachia appear to play a role in the synthesis of B vitamins (Brownlie et al., 2009;Hosokawa et al., 2010). These experiments suggest that Wolbachia not only affects macronutrient metabolism, but also the provisioning of mineral micronutrients and cofactors. In addition, some behavioral effects of Wolbachia in Drosophila may be explained by alterations in hormone biosynthesis pathways. For example, wMelPop may increase aggressive male behavior through control of octopamine synthesis (Rohrscheib et al., 2015). While these interesting effects on metabolism have not yet been explained, an increase in insulin signaling is one possible source of Wolbachia's effects on host metabolism (Ikeya, Broughton, Alic, Grandison, & Partridge, 2009). Another possibility is that Wolbachia may affect mitochondrial mass or | 931 activity directly (Ballard & Melvin, 2007). Intriguingly, Wolbachiamediated metabolic alterations are suggestive of gainful manipulation of host physiology. Host diet in Drosophila, perhaps acting through the insulin signaling pathway, has been shown to regulate Wolbachia titer (Serbus et al., 2015). Therefore, it would not be surprising to discover that Wolbachia, like many other invasive bacteria, has the ability to modulate the metabolism of its host to increase its own transmission.

| Cell autonomous and non-autonomous effects on pathogen resistance
Wolbachia in infected flies and mosquitoes has the ability to confer resistance against a wide array of viral, bacterial, parasitic, and fungal pathogens (Eleftherianos et al., 2013). This property allows pathogen-infected hosts to survive and continue to reproduce in a situation where uninfected hosts would not survive, thus providing a great evolutionary advantage for Wolbachia and its host. In mosquitoes, Wolbachia provides resistance against the malaria parasite Plasmodium (Kambris et al., 2010) and the filarial nematode B. pahangi (Kambris, Cook, Phuc, & Sinkins, 2009) as well as protection from the bacterium Erwinia caratova (Kambris et al., 2009) and the dengue and chikungunya viruses ).
In Drosophila, Wolbachia infection imparts resistance against various positive-sense single-stranded RNA viruses such as: Drosophila C virus, noravirus, and cricket paralysis virus (Hedges, Brownlie, O'Neill, & Johnson, 2008;Rainey et al., 2016;Teixeira, Ferreira, & Ashburner, 2008) and against the entomopathogenic fungus Beauveria bassiana (Panteleev et al., 2007). However, Wolbachia protection does not include all infections. For instance, the titer of the intracellular bacteria Salmonella typhimurium and Listeria monocytogenes is not affected by Wolbachia in Drosophila, though it should be noted that these pathogens do not naturally infect flies (Rottschaefer & Lazzaro, 2012). Therefore, it is possible that Wolbachia may confer protection against intracellular bacteria that can naturally colozine arthropods.
Pathogen resistance imparted on the host by Wolbachia has been observed on numerous occasions and has been reviewed elsewhere (Eleftherianos et al., 2013). However, information regarding the conditions necessary for this phenotype, as well as mechanistic insight is still lacking (Rainey, Shah, Kohl, & Dietrich, 2014). One proposed mechanism is the priming of the immune response by Wolbachia that subsequently hastens pathogen removal upon infection. However, there is conflicting evidence for this claim and establishing a concrete link between Wolbachia and host immunity will greatly further understanding of the pathogen resistance phenotype (Bourtzis, Pettigrew, & O'Neill, 2000;Moreira et al., 2009;Rances et al., 2013;Wong, Hedges, Brownlie, & Johnson, 2011;Ye, Woolfit, Rances, O' Neill, & McGraw, 2013). Alternatively, some have suggested that the synthesis of reactive oxygen/nitrogen species and cholesterol is involved (Caragata et al., 2013;Pan et al., 2011;Wong, Brownlie, & Johnson, 2015). There is also some evidence that increased host cell autophagy driven by Wolbachia infection plays a role in viral resistance (Le Clec'h et al., 2012). Each of these mechanisms would require Wolbachiamediated effects on somatic tissues and cells that regulate the host response to infection, such as the gut, fat body, and hemocytes. The particular cells and tissues involved in each case are not fully known.
In Drosophila, Wolbachia titer in the head, gut, and malpighian tubules is correlated with antiviral protection (Osborne et al., 2009). Furthermore, the emergence of fluorescence-based assays for the detection of both Wolbachia and viruses have recently allowed for experiments that map their distribution and localization in whole insects (Kliot & Ghanim, 2015). In several tissues, such as the midgut and salivary

| Effects on stress resistance and longevity
As most mutualists and parasites, Wolbachia undoubtedly benefits from the health and longevity of its host. Therefore, it is not surprising that Wolbachia influences host responses to cellular stress and damage as well as lifespan. In insects, Wolbachia induces the production of host reactive oxygen species (ROS) (Pan et al. 2011;Wong et al., 2015). Perhaps because Wolbachia must persist in this oxidative intracellular environment without causing damage to the host, infection also upregulates host antioxidant genes (Brennan, Haukedal, Earle, Keddie, & Harris, 2012;Brennan, Keddie, Braig, & Harris, 2008). Wolbachia also reduces oxidative stress by regulating host iron homeostasis. Iron is a highly toxic precursor to ROS and the expression of Wolbachia bacterioferretin reduces labile iron concentrations, which in turn prevents toxicity (Kremer et al., 2009). Intriguingly, while Wolbachia protects against iron toxicity, resistance to lead is decreased during infection (Wang et al., 2012), suggesting that protection from heavy metals is restricted.
Reduced iron toxicity is associated with the inhibition of apoptosis in the wasp Asobara tabida (Kremer et al., 2009). In this organism, Wolbachia is required for proper oogenesis, and oocytes fail to mature when it is removed due to extensive apoptosis (Miller et al., 2010;Pannebakker et al., 2007). As mitochondria-derived ROS are also involved in modulating apoptosis, the ability of Wolbachia to regulate responses to these stressors may have far reaching consequences for host lifespan and reproduction.
Whether Wolbachia modulates apoptosis from host germline or somatic tissues is unclear. Apoptosis in the wasp oocyte is likely due to Wolbachia in the same tissues. On the other hand, the loss of Wolbachia in filarial nematodes through antibiotic therapy also induces apoptosis in both the adult germline and somatic cells of the embryo (Landmann, Voronin, Sullivan, & Taylor, 2011

| Somatic routes of germline infection
The discordance between Wolbachia and host insect phylogenies strongly argues for multiple horizontal transmission events over evolutionary timescales. Insight into possible mechanisms and routes of transmission have come from experiments in which Wolbachia injected into the abdomen is able to reach the germline through the somatic stem cells (Frydman et al., 2006), suggesting that this localization during natural infection serves to facilitate reaching of the germline for vertical transmission. Indeed, from the somatic stem cell niche, Wolbachia is supplied to the somatic stem cell, which can then divide and transmit Wolbachia to follicle cells (Toomey et al., 2013). From infected follicle cells, Wolbachia may then transfer to the developing oocyte (Toomey et al., 2013).

| CONCLUSIONS
While Wolbachia are most prevalent in the host germline and primarily studied for their effects on these tissues, the studies described in this review demonstrate that Wolbachia is consistently found both intra and extracellularly in important somatic tissues such as the nervous system, fat body, and gut of their arthropod hosts, and in hypodermal chords in the nematode hosts. Wolbachia distribution to these somatic tissues is primarily regulated by segregation patterns during embryonic development. However, active invasion of somatic tissues during development and adulthood is also involved. This mechanism not only regulates somatic distribution, but may be involved in the horizontal spread of infection, which appears to play an important ecological role in the transmission and diversification of Wolbachia. The presence of Wolbachia in somatic tissues may also explain many phenotypic alterations observed in infected hosts, such as: behavioral change, resistance to pathogenic infection, shifts in metabolism, and changes in longevity.
The effects that somatic Wolbachia has on the host germline suggest that invasion of the soma and somatic localization may have evolved as an altruistic mechanism to facilitate vertical transmission.
That is, by not entering the germline, somatic Wolbachia are essentially sacrificed, as they will not be inherited by the next generation. However, in doing so, they can produce many of the phenotypes described above that increase the transmission of their sister Wolbachia, thus benefiting the species as a whole. Whether Wolbachia originated as a germline endosymbiont that invaded the soma resulting in these advantageous phenotypes or as a somatic endosymbiont that invaded the germline for vertical transmission remains unresolved. There are cases of Wolbachia existing exclusively in the germline (tsetse fly), but also exclusively in somatic tissues (male nematodes). In addition, invasion of both somatic and germline tissues has been documented, further obscuring the origins of Wolbachia. studies focusing on transfer of Wolbachia between hosts under a variety of conditions will be helpful to fully determining the prevalent routes of horizontal transmission in nature. More importantly, studies directly mapping host phenotypes to Wolbachia in somatic tissues would greatly aid efforts to use this extraordinary endosymbiont for the public good.

FUNDING INFORMATION
This study was funded by the National Institute of General Medical Sciences, (Grant / Award Number: 'GM104486').

CONFLICTS OF INTEREST
The authors declare no conflicts of interest.