Interactions between bacteria and eukaryotes are widespread, and have attracted research interest and fascinated generations of scientists involved in multiple biological and biomedical fields.

The members of alphaproteobacterial order Rickettsiales engage in complex interactions with diverse eukaryotic hosts. All experimentally characterised Rickettsiales are metabolically-defective, thus dependent on their host for their long-term survival and multiplication. They include disease agents in humans and other vertebrates, such as the eponym Rickettsia, as well as Orientia, Anaplasma and Ehlrichia, all vectored by haematophagous arthropods such as ticks, mites and lice. Another renowned member of this order is Wolbachia, which was defined as master manipulator of invertebrate biology for its ability to influence the reproduction of multiple arthropod hosts to ensure its own transmission, while being a obligate mutualist in other host species.

In the last decades, thanks to the effort of multiple research groups worldwide, including ourselves, it has become evident that most of the phylogenetic diversity of Rickettsiales live in association with a wide range of aquatic unicellular eukaryotes (e.g., ciliates, amoebae, flagellates, algae), which have been deemed as most probable ancestral hosts. Thus, it is implied that convergent (or parallel) evolution is a common and repeated instance for lineages in this order of bacteria, which have been (and still are) able to shift multiple independent times from being associated with unicellular and/or aquatic hosts, to multicellular and/or terrestrial ones, including the multiple evolutionary origins of pathogens.

Rickettsiales are a very ancient lineage, their origin being estimated to be almost coincident with the eukaryotes that happen to be their hosts. The recent discovery of metagenome-derived early-diverging Rickettsiales lineages that are metabolically-richer and may not be host-associated could even backdate such estimates. Given their long history, it seems not surprising that several features of the evolution of Rickettsiales are still elusive, in particular concerning their early adaptations to colonise eukaryotic cells, live in association with their hosts and exploit them for their own proliferation. This lack of understanding also stems from the poor knowledge on their most numerous aquatic representatives, still neglected from a genomic and experimental point of view.

In the latest years, we collaboratively joined our research efforts in order to address such fundamental evolutionary questions on the roots and dynamics of the interactions between the Rickettsiales and their hosts. Few years ago (2019), we published on the ISME Journal the discovery of Deianiraea vastatrix, a peculiar Rickettsiales representative, taking its name from Greek myth of Deianira, who killed her husband Hercules with a poisoned shirt. Indeed, differently from all characterised Rickettsiales (still to date), this bacterium did not invade the cells of its host, the unicellular ciliate protist Paramecium, but was attached to its extracellular surface, multiplying there, and resulting lethal for the host. Deianiraea is phylogenetically derived, but presents putatively ancestral genomic traits, namely biosynthetic pathways for amino acids.

Taken together, these findings led us to wonder for the first time about a possible alternative to the conventional and parsimonious views on the evolution of Rickettsiales, which assumed ancestor already presenting derived traits such as obligatorily intracellular condition within some eukaryotic host. Conversely, we conjectured an alternative “intracellularity late” hypothesis, implying that the Rickettsiales ancestor could have been already host-associated, but not yet obligatorily intracellular. Intracellularity would have evolved independently in multiple (but not all) sublineages. At the time, we judged the two views as potentially equivalent.

The herein presented study was fostered by the purpose to test which between the “early” and “late” hypothesis for the evolution of intracellularity and of host-dependence in Rickettsiales could find more support.

Herein, we selected comparative genomics as the most suitable approach. In the sake of understanding the deep evolution of this broad and diversified lineage, we reasoned with an integrated phylogenetic perspective, namely aiming at including the most possible genus-level lineages of Rickettsiales. To this purpose, we performed targeted de novo sequencing of neglected lineages, carefully selected published genomes, and mined published metagenomic datasets. Eventually, we assembled a large and phylogenetically-representative genomic dataset of 113 Rickettsiales. Thus, we remarkably extended the available diversity of known lineages, as well as identifying three novel families within the Rickettsiales. Leveraging this to our knowledge unparalleled dataset, we could get a view on the evolution of Rickettsiales and the interaction with eukaryotic hosts with an unprecedented detail.

Accordingly, we collected multiple lines of evidence supporting the host-dependence and intracellularity “late” hypotheses. Specifically, we found strong indications that multiple metabolic pathways, likely present in the last common ancestor of Rickettsiales and absent in many but not all host-associated representatives, making them host-dependent, were not lost once at the root of the Rickettsiales tree, but in multiple independent and more recent instances along their phylogeny. They pertain to several biosynthetic pathways, such as for nucleotides, amino acids, as well as energy metabolism. For nucleotides, we even could tentatively track the losses of biosynthetic capabilities, as a consequence of single horizontal acquisitions of suitable membrane transporters, which enable to steal such compounds from the hosts, thus making synthesis dispensable.

In parallel we focused on the evolution of genomic features involved in modulating interactions with the hosts. For what concerns some of them, in particular secretion, motility and attachment systems, we evidenced significant parallelism with biosynthetic pathways, indicating a relatively recent specialisation of the host-interactions in different Rickettsiales sublineages. We found that emergence and expansion of some families of proteins involved in adhesion and invasion of host cells appeared to be lineage-specific, thus further corroborating an independent and differential adaptation to the interactions with host cells.

Such findings have the potential to provide novel perspectives on the evolution of other bacteria engaging in stable interactions with eukaryotes, in particular representatives of ancient and diverse lineages, such as Chlamydiae and Legionellales.

At the same time, we strongly encourage direct experimental investigations on extant representatives, which, besides genomics, could provide further invaluable resources to understand their past evolution.

References

Castelli M, et al. (2019) Deianiraea, an extracellular bacterium associated with the ciliate Paramecium, suggests an alternative scenario for the evolution of Rickettsiales. ISME Journal 9:2280-2294. https://doi.org/10.1038/s41396-019-0433-9

Castelli M, et al. (2024) Host association and intracellularity evolved multiple times independently in the Rickettsiales. Nat. Comm.15, 1093. https://doi.org/10.1038/s41467-024-45351-7

Salje J (2021) Cells within cells: Rickettsiales and the obligate intracellular bacterial lifestyle. Nat. Rev. Microbiol. 19, 375–390. https://doi.org/10.1038/s41579-020-00507-2

Schön ME, et al. (2022) The evolutionary origin of host association in the Rickettsiales. Nat. Microbiol. https://doi.org/10.1038/s41564-022-01169-x

Wang S, Luo H (2021) Dating Alphaproteobacteria evolution with eukaryotic fossils. Nat. Commun. 12, 3324. https://doi.org/10.1038/s41467-021-23645-4