Cooperation between symbiotic partners through protein trafficking

Symbiosis can act as a major driver of evolution, as symbionts’ genetic resources can enhance host resistance to stress and expand their ecological niches (1). This is particularly true for endosymbionts, i.e., symbionts that live within the body or cells of their host, as the intimate contact between partners favors molecule exchanges. However, apart from metabolites, the extent of macromolecule trafficking between endosymbiotic partners remains underexplored. In the current issue of PNAS, Ling et al. (2) provide novel insights into this phenomenon. They provide evidence that not only endosymbionts can shuttle entire proteins across several membranes, but proteins from different organisms can cooperate in vivo to increase host adaptability to environmental changes.

Endosymbiosis is widespread in nature and particularly prevalent among insect species. Notably, insects thriving on nutritionally poor or unbalanced diets (e.g., plant sap, seeds, blood) have repeatedly established obligate endosymbioses with bacteria that complement their diet with nutrients lacking in their habitat (1). Endosymbionts are often housed within specialized cells of the host named bacteriocytes (3). Due to a relaxed evolutionary pressure within these cells, the bottleneck effect from vertical transmission, and the absence of recombination with free-living relatives, these obligate endosymbionts experience significant genome reduction, including the loss of virulence genes along with genes redundant with the host pathways or unnecessary in the new habitat (4). Genomic erosion is also promoted by the loss of the endosymbiotic DNA repair machinery, which accelerates the accumulation of mutations. In the textbook example of obligate endosymbiosis, the bacteriocyte-bound symbionts of aphids, Buchnera aphidicola, have lost most ancestral genes and have specialized in the production of amino acids required by aphids.

While reducing the endosymbiotic genome leads to a cost-effective metabolic interaction between host and bacteria, the massive genomic erosion can also lead to negative effects on both symbiotic partners, including the restriction of ecological niches and the ultimate risk of extinction in case of increased environmental pressure (4). Buchnera, along with other reduced-genome obligate endosymbionts, lack the plasticity to adapt to environmental changes, which makes them particularly vulnerable to biotic and abiotic stresses and can compromise the host’s survival. Heat tolerance, for instance, varies drastically among aphid species and was shown to be associated in part with Buchnera genetic features (5). Indeed, a single mutation in the homopolymer promoter of Inclusion body-associated protein A (IbpA), a prokaryotic small heat shock protein gene, impairs Buchnera ibpA expression and is linked to reduced aphid survival upon heat stress (5). Replacement of ibpA-deficient Buchnera with wild-type Buchnera rescues the host from heat-shock events (6).

Ling et al. showcase a remarkable cooperation between Buchnera and Serratia under heat stress and provide a mechanistic explanation for the Serratia-associated thermotolerance of infected aphids.

The deleterious effects of the irreversible and extensive endosymbiont gene loss can be countered by the replacement of the primary endosymbiont or the colonization by a facultative symbiont that further expands the capabilities of the symbiotic system (4). In some cases, in particular when genes involved in essential nutrient biosynthesis are lost, facultative symbionts can even become obligate, leading to co-obligate symbiosis (3). The interaction between insects and their obligate endosymbionts is often sensitive to heat stress, potentially explaining the widespread examples of facultative endosymbionts associated with heat tolerance (7). Serratia symbiotica has been shown to increase aphid thermotolerance and field studies have shown that the incidence of Serratia increases in warmer climates compared to colder ones, even at the same site of sampling, suggesting that Serratia protection to heat stress benefits the host and might therefore favor Serratia prevalence (8). In pea aphids, previous studies suggested that Serratia protects the host by releasing metabolites shielding Buchnera upon heat stress (9). Yet, aphids lacking Serratia have an increased prevalence of the Buchnera mutated ibpA gene both in the field and in laboratory strains (5, 10), suggesting further compensation mechanisms between the endosymbionts.

Ling et al. (2) showcase a remarkable cooperation between Buchnera and Serratia under heat stress and provide a mechanistic explanation for the Serratia-associated thermotolerance of infected aphids, whereby gene complementation between the two endosymbionts restores an incomplete DNA mismatch repair (MMR) system in Buchnera. Indeed, the MMR system safeguards the integrity of the genome, repairing DNA errors arising through replication, thanks to the recruitment of different proteins, including MutS, MutL, and MutH (11). MutS homodimer recognizes mismatched bases and recruits the MutL homodimer that activates the endonuclease MutH, which is absent in Buchnera and present in the Serratia genome. MutH nicks the DNA strand near the mismatch, allowing UvrD helicase and an exonuclease to degrade the DNA. Finally, the DNA gap is repaired by a DNA polymerase followed by DNA ligation (11). Ling et al. (2) provide microscopy and mass spectrometry evidence for the translocation of the Serratia MMR protein MutH outside of Serratia and the sheath cells in which they are located (i.e., small cells that surround bacteriocytes and can harbor facultative symbionts) into the aphid bacteriocytes and further into Buchnera cells (Fig. 1). Serratia MutH appears to complement Buchnera MutL and MutS proteins to form an active MMR. This is proposed to slow the accumulation of mutations, including the detrimental Buchnera ibpA allele prevalent in Serratia-free aphids.

Fig. 1.

Central biological functions impacted by protein trafficking from host to endosymbiont (ES) (in orange) in insects and unicellular protists and the first case of protein shuttling between aphid endosymbionts (in pink) as evidenced by Ling et al. (2).

While metabolic complementation (the exchange of metabolites) between species has been extensively studied (12), the exchange of entire proteins, particularly across multiple membranes, remains poorly understood. Existing literature highlights instances where hosts provide proteins missing from the genomes of endosymbionts (Fig. 1) [(13–17), reviewed in detail in ref. 18]; however, the novelty presented in Ling et al. (2) is that other endosymbionts can also be a source of protein exchange and adaptation. Moreover, Serratia MMR protein MutH was not the only Serratia-encoded protein found inside Buchnera cells, and other proteins related to nutrient metabolism, gene regulation, and stress response were also detected, highlighting the fact that the extent of protein exchange could be much larger than anticipated.

A key unanswered question is how these proteins are trafficked across multiple membrane layers and remain intact, especially when endosymbionts are physically compartmentalized in different cells, as seen with Buchnera in bacteriocytes and Serratia in sheath cells and/or other bacteriocytes (distinct from the ones harboring Buchnera) (Fig. 1). Outside the insect world, two model systems from unicellular protists, the amoeba Paulinella chromatophora and the kinetoplastid Angomonas deanei, were shown to bear targeting signals for the correct addressing of host nuclear proteins back to their endosymbionts, similarly to what is seen with organelles (15, 16). Even though the mechanism remains unclear, these proteins are thought to be transported either through endoplasmic reticulum vesicles or outer membrane channels. Within insects, host-encoded proteins were shown to be shuttled into endosymbionts in the aphid—Buchnera (14), the tripartite nested mealybug (13), the cereal weevil—Sodalis (19), and the red palm weevil–Nardonella (17) systems. In the case of aphids, each Buchnera cell within bacteriocytes is individually surrounded by a host-derived symbiosomal membrane, and the molecular basis of transport across such membrane remains largely unresolved. The mealybug (Planococcus citri) system is even more complex, as there are two endosymbionts, and one (Moranella endobia) resides in the cytoplasm of the other (Tremblaya princeps). Proteins from the host are transferred to the innermost endosymbiont, Moranella, and do not accumulate in the cytoplasm of Tremblaya, to create a peptidoglycan layer around Moranella cells (13). Although the mechanism of protein trafficking remains to be elucidated, the authors also pointed out that it could be the mRNA rather than the protein that is trafficked to the endosymbiont cells instead. Ling et al. (2) also suggest that aphid bacteriocytes have increased expression of genes involved in vesicle synthesis and transport compared to other aphid tissues, and Serratia bears genes associated with outer membrane vesicle formation, further supporting the hypothesis of vesicle trafficking. Whatever the mechanism, it is clear that organisms can exchange such macromolecules, directly affecting their adaptability traits.

Unfortunately, identifying the process that governs exchanges between symbiotic partners is hampered by the fact that the majority of insect endosymbionts, including Buchnera and most Serratia symbiotica strains, are not culturable in vitro due to extensive genome erosion, preventing functional genetic assays (20). Finally, Ling et al. (2) and others highlight how the beneficial effects brought on by facultative endosymbionts are contingent on the environment and might remain cryptic under normal conditions (4, 7), stressing the need for multicontext dependent studies.