Trichomonas vaginalis is the causative agent of trichomoniasis, the most common non-viral STD in the world, with an estimated 276 million new cases annually (Menezes et al., 2016). Despite being typically asymptomatic in men, the parasite is still associated with urethral inflammation in addition to a heightened risk of prostate cancer (Krieger, 1995, Sutcliffe et al., 2006). However, while 50% of infected women are asymptomatic (Fouts and Kraus, 1980), infection with this protozoan parasite is associated with pregnancy complications such as low birth weight and premature delivery, in addition to a 1.5-fold increase in the risk of acquiring HIV-1 (Cotch et al., 1997, Røttingen et al., 2001, McClelland et al., 2007).

Trichomonas vaginalis is part of the eukaryotic supergroup Metamonada, specifically within the phylum Parabasalia (Adl et al., 2019). Parabasalia are a clade of anaerobic flagellated organisms, typically defined by mitochondria-related organelles known as hydrogenosomes and a ubiquitously expressed cytoskeletal arrangement known as the parabasal apparatus (Lindmark and Muller, 1973; Brugerolle, 1991, Čepička et al., 2017). These organisms typically live as endobionts, i.e., commensals or mutualists, in the anaerobic guts of metazoans. However, members of two parabasalid orders have independently made the transition to parasitism and migrated to other areas of animal bodies such as the oral cavity, or in the case of T. vaginalis, the vaginal epithelium.

Particularly important to the pathogenic mechanism of many parasites is the membrane trafficking system, which is responsible for intracellular transport of cargo, protein, and other materials in and around cells. The functionality of this system is critical in mobilizing virulence factors, promoting pathogenicity, and ensuring parasite survival within a hostile host environment (Klinger et al., 2016 inter alia). Membrane trafficking in T. vaginalis aids in the mediation of its pathogenesis by enabling the secretion of lytic molecules that cause haemolysis in vaginal epithelial cells (Fiori et al., 1996).

The process of membrane trafficking is regulated in part by vesicle formation machinery such as heterotetrameric adaptor protein complexes (HTAC), also known as adaptins or the AP complexes. The first complexes, adaptins 1 and 2 (i.e., AP-1 and AP-2) were discovered and characterized in 1984 (Pearse and Robinson, 1984). Third, fourth and fifth adaptin complexes (i.e. AP-3, AP-4, AP-5) were subsequently described, bringing the count to where it stands today (Newman et al., 1995, Panek et al., 1997; Dell'Angelica et al., 1999; Hirst et al., 2011). Coatomer protein coat-complex I (COPI), which functions in retrograde transport within the Golgi and from the Golgi to the endoplasmic reticulum (ER), consists of seven subunits and is related to the adaptins (Duden et al., 1991, Schledzewski et al., 1999). Most recently, the TSET complex was identified as a sixth pan-eukaryotic homologous protein complex (Hirst et al., 2014).

The primary function of adaptins is their involvement in the selection and sorting of vesicular cargo at the trans-Golgi network (TGN), endosomes, and the plasma membrane (Boehm and Bonifacino, 2001, Bonifacino and Glick, 2004). AP-1 and AP-2 are firmly associated with the formation of clathrin-coated vesicles, and despite not being enriched in clathrin-coated vesicles, AP-3 also forms an interaction with clathrin. However, clathrin-independent function of the complex is also known (Dell’ Angelica et al., 1998; Peden et al., 2002). Adaptin complexes AP-4 and AP-5 do not form any sort of interaction with clathrin and are involved in autophagy and late endosome to Golgi retrieval, respectively (Hirst et al., 1999, Hirst et al., 2018; Davies et al., 2017; Sanger et al., 2019). The majority of the work characterizing these complexes has been completed in animals and fungi. Nonetheless, where examined in other modern eukaryotes, e.g., plants and trypanosomes, homologous and comparable functions have been observed, albeit reflecting the unique biology of those organisms (Klinger et al., 2016, inter alia). Notably, the TSET complex is absent from animals and fungi, with the exception of a single remnant subunit. It was first described and extensively characterized in plants (Gadeyne et al., 2014; Johnson et al., 2021; Wang et al., 2021) and to a lesser extent in Dictyostelium (Hirst et al., 2014).

Each of these complexes has a homologous heterotetrameric core, comprised of a small “sigma (S)” a medium “mu (M)”, and a large “beta (B)” subunit. The other large subunit is named differently in each complex (e.g., gamma (G) for AP-1, alpha (A) for AP-2, delta (D) for AP-3, epsilon (E) for AP4, and zeta (Z) for AP-5 (see Robinson (2015) for a schematic representation), but this homologous set is referred to as gamma subunits more generally. COPI subunits follow the overall naming scheme, while the TSET complex uses complex-specific names (Hirst et al., 2014). The additional membrane deformation or accessory proteins to these complexes may also be homologous, as they share a common structure of beta-propeller fused with alpha-solenoid domains, and in the cases of the COPI and TSET proteins do retrieve one another by homology searching (Devos et al., 2004). Various studies have shown that each of these subunits is the products of gene duplications that predate the Last Eukaryotic Common Ancestor (LECA). The exception is the beta subunits of the AP-1 and AP-2 complexes which are each other’s closest relatives, and for which the shared beta subunit had not yet duplicated prior to the LECA (Dacks et al., 2008). Therefore, these subunits are referred to as AP1/2B compared with the beta subunits of other complexes (e.g., AP-3B). However, parallel duplications have taken place in multiple eukaryotic lineages of this subunit, resulting in independent beta subunits for the AP-1 and AP-2 complexes in animals, fungi, and more (Dacks et al., 2008). Nonetheless, the LECA complement sets a null hypothesis for the complement expected to be found in any given eukaryotic genome, unless shaped by downstream loss or expansion events that have been shown to have impacted various eukaryotic lineages (Dacks and Robinson, 2017, Larson et al., 2019, Richardson and Dacks, 2022).

The T. vaginalis genome, for example, shows substantial expansions in the complement of adaptin protein coding genes, 73, compared with the human count of 23 (Carlton et al., 2007). The additional two T. vaginalis AP-5 subunits described in 2011 brings the count to 75 (Hirst et al., 2011). This is consistent with the genome sizes typically seen in parabasalids, which are large relative to other eukaryotic microbial parasites (Zubáčová et al., 2008), but in contrast with the more typical genome streamlining observed in other parasitic lineages (Adams et al., 2020). Nonetheless, parasite-specific expansions and gene duplications tied to components that increase the survival and pathogenicity of parasites in their hosts have been noted in nematodes and pathogenic fungi (Soanes et al., 2008, Baskaran et al., 2017, Adams et al., 2020). Even Giardia lamblia, a parasitic metamonad with an extremely simplified and condensed genome (Morrison et al., 2007), has duplications in 40% of its genes (Sun et al., 2010). This includes those for membrane-trafficking components, for example, the recent functional characterization of Giardia’s expanded machinery in the endosomal organelle-associated ESCRT complexes (Pipaliya et al., 2021a).

It is well understood that adaptins serve a critical function in cells, which is exemplified by AP-1 mu knockout being embryonicly lethal in mouse models, in addition to deficiencies in AP-4 and AP-5 being associated with hereditary spastic paraplegia in humans (Meyer et al., 2000, Abou Jamra et al., 2011, Hirst et al., 2013b). Therefore, the presence of the many adaptin subunits may imply selection for an expanded complement, but the points at which these duplications took place, and their correlation with evolutionary divergence points within the Parabasalia, are currently unclear. Are such expansions T. vaginalis-specific, or did they instead occur at more ancient points in evolutionary history, correlating with shifts to endobiosis, parasitism, or specific (vaginal epithelial) parasite environments?

To resolve this, adaptin complements spanning multiple parabasalid lineages are required. High quality genomes are available, but at this time are limited to parasitic taxa only. Parabasalid transcriptomes can be used to supplement these data; however, the available data are limited to members of the parabasalid clades in which parasitism has taken root. The organisms included in this study represent the taxonomic diversity of the two best studied parabasalid orders, Trichomonadida and Tritrichomonadida. In addition to T. vaginalis, Trichomonas tenax, a parasite of the oral cavity (Riberio et al., 2015), and Trichomonas gallinae, a parasite found in the upper gastrointestinal tract of birds (Stabler, 1954) are included. Pentatrichomonas hominis is included to provide a sampling point outside the Trichomonas genus but within the Trichomonadida. This organism inhabits the gastrointestinal systems of vertebrates and has been linked to diarrhea and other gastrointestinal issues in humans and domesticated animals (Meloni et al., 2011; Li et al., 2015; Bastos et al., 2018). The other side of the ancient parabasalid split, illustrated in the phylogeny presented in Noda et al. (2012), is represented by three members of the parabasalid clade Tritrichomonadida: T. foetus, Histomonas meleagridis, and Dientamoeba fragilis. Tritrichomonas foetus infects the vaginal epithelium of cattle, much like T. vaginalis in humans (Yule et al., 1989). Histomonas meleagridis and D. fragilis are closely related parasitic parabasalids, infecting primarily poultry and the human bowel, respectively (Stark et al., 2016; Palmieri et al., 202). Excitingly, the recent discovery of a parabasalid sister lineage, the amoeboid genus Anaeramoeba (Táborský et al., 2017, Stairs et al., 2021), has provided a new opportunity to trace the evolution of the vesicle formation machinery to a point that predates the emergence of the phylum and provides a free-living relative to the exclusively parasitic dataset.

Here we explore the complements, evolutionary dynamics and history of the heterotetrameric vesicle coat complexes, adaptins and related complexes across the Parabasalia.