by Christoph
There was joy, even excitement among archaeologists – the microbiologists who study archaea, not the diggers – when the first images of Promethearchaeum syntrophicum MK‑D1 made the rounds (see part 1, and here in STC). This sense of joy deepend still when images of its cousin Lokiarchaeon ossiferum Loki-B35 were added a little later (Figure 5). The ancestors suddenly had faces, literally, and were no longer only vaguely recognizable from genome sequences!
(click to enlarge)
Figure 5. Complex, variable architecture of Loki‑B35 cells. a,b, SEM imaging of fixed cells showed small coccoid cells with extensive protrusions. Example micrographs from n = 2 independent cultures are shown. For a and b, scale bars, 500 nm. Frontispiece: cropped from a. Source
Cultivating Lokiarchaeum ossiferum Loki‑B35
Unlike MK-D1, which was enriched stepwise from a deep marine sediment drill core sample taken 80 km offshore the Shimokita Peninsula of Japan (see part 1), Rodrigues-Oliveira, Wollweber, Ponce-Toledo et al. (2022) enriched Lokiarchaeum ossiferum Loki‑B35 from sediments of a small estuarine canal connected to the Mediterranean near Piran, Slovenia. The rationale for investigating a more accessible sampling site was that in the meantime 16S rRNA gene sequences from 'Asgard' archaea had been detected in variety of anoxic and often marine environments worldwide.
For the choice of the most promising fraction of their drill core sample (the 13–16 cm fraction, see here) and throughout their numerous cultivation/enrichment efforts, the researchers monitored the presence and abundance of "Lokis" using quantitative PCR (qPCR) with 'Asgard'-specific 16S rRNA primers. Various initial enrichment cultures of the original sample in serum flasks containing sterile-filtered water from the source and supplemented with complex organics of different compositions and with various headspace conditions (anoxic!) either failed to grow or grew after 140 days at 20°C but did not "survive" two consecutive transfers. By using a modification of the medium reported for the cultivation of P. syntrophicum MK-D1 by Imachi, Nobu et al. (2020), the researchers observed cell growth, but "Loki" abundances never reached more than 2–8%. Only through developing a minimized lokiarchaeal medium (MLM) with casein hydrolysate as the sole carbon source, lokiarchaeal relative abundances reached between 25% and 80% in cultures after several transfers.
(click to enlarge)
Figure 6. Complex and variable architecture of Loki‑B35 cells. c–f, Slices through cryo-tomograms (c,e; thickness, 9.02 nm) and the corresponding neural-network-aided 3D volume segmentations (d,f) of two different Loki‑B35 cells. The insets in c and e show 2D overview images of the two different target cells. Cell bodies (c,d) and networks of protrusions (e,f) both contained ribosomes (grey arrowheads), cytoskeletal filaments (orange arrowheads) and complex surface densities (blue arrowheads). For c and e, scale bars, 100 nm (tomogram) and 1 μm (2D overview). g,h, Expanded views of slices from tomograms in c and e, showing ribosome chains, complex surface proteins and filaments (colour code as in c–f) in a junction of a cell bridge (g) and a constricted part of the protrusion network (h). For g and h, scale bars, 100 nm. (Click here for panels i–l) i–l, Slices through cryo-tomograms showing a putative chemoreceptor array (i; indicated by a white arrowhead) and different types of connections between cell bodies and protrusions ( j–l). The coloured arrowheads indicate filaments and surface structures as defined for c–f. The white arrowheads in j indicate weak densities at the neck of the junction. Slice thickness, 9.02 nm (j) or 10.71 nm (i and k–l). For i–l, scale bars, 100 nm. Source
Amplicon sequencing analyses of 16S rRNA genes revealed that the culture with the highest enrichment, Loki-B35, consisted of three dominant and two minor species: a single 'Asgard' sequence representing Lokiarchaeum ossiferum Loki‑B35 (79%), a sulfate-reducing bacterium of the Desulfovibrio genus (10%), a hydrogenotrophic methanogen of the Methanogenium lineage (6%), as well as a Halodesulfovibrio and a member of the Methanofastidiosales genus (both at around 2%). Both Halodesulfovibrio sp. and Methanogenium sp. were also syntrophic partners in the enrichment cultures of P. syntrophicum MK‑D1 (see part 1), which stems from a geographically different and deep-sea environment. It is thus likely that both "Lokis" rely on a similar metabolism, involving the fermentation of peptides to H2 and/or small organic acids such as formate. This assessment is supported by a coarse comparison of the genomes (better: chromosomes) of P. syntrophicum MK‑D1 (4.46 Mbp) and L. ossiferum LokiB-35 (6.04 Mbp). See here for the (groups of) orthologous genes shared between MK‑D1 and Loki‑B35, and here for the synteny of their genomes, which is readily recognizable despite the Loki‑B35 genome being larger by 1.6 Mbp.
In co-cultures, Loki-B35 grows without lag phase to cell densities of up to 5.0×107 cells/mL within 50 to 60 days when started with a 10% inoculum, but with extremely long lag phases of 90–120 days(!) when inoculated from stationary cultures. The generation time was 7–14 d, and compared with MK‑D1, Loki-B35 grows approximately twice as fast and to higher cell densities.
The "face" and the anatomy of Loki‑B35
In scanning electron micrographs (SEM), Loki-B35 cells appear as spheres of variable size (0.3–1.0 µm), which are distinctly smaller than the other co-cultured cells and therefore easily distinguished from them (Figure 5). The cells mostly appear as individuals but are also, albeit infrequently, found in aggregates with cells of the co‑cultured species. Rodrigues-Oliveira et al. (2022) assume that, although nutrient exchange with co-cultured species may be necessary for Loki-B35's growth, persistent cell–cell contact seems to not be obligate throughout its lifecycle. Their round-shaped cell bodies are associated with elaborate and heterogeneous protrusions, and their cell envelope features complex unordered densities protruding from a single membrane ("blebs"). In contrast to MK‑D1, the long protrusions appear more irregular, frequently branching or expanding into bulbous structures when looked at by cryo‑EM and cryo‑ET (Figure 6) (download here a movie for the images in Figure 6, d+f).
(click to enlarge)
Figure 7. Lokiactin is involved in cytoskeleton formation. a, Slice through a cryo-tomogram showing a cytoskeletal filament inside a protrusion at higher magnification. Slice thickness, 5.36 nm. Scale bar, 100 nm. b, Filament segments were extracted from cryo-tomograms for structural analysis. 2D classes that were obtained after helical reconstruction of 2D-projected filament particles are shown, indicating a twisted double-stranded architecture. Box size, 34.3 × 34.3 nm. c, Sub-tomogram average (24.5 Å resolution) of the cytoskeletal filament displaying helical parameters with a high similarity to eukaryotic F-actin and archaeal Crenactin. Structural docking shows that an F-actin-like filament is consistent with the reconstructed map. Scale bar, 50 Å. h, Immunofluorescence staining of Loki‑B35 cells with two different Lokiactin-specific antibodies analysed by stimulated emission depletion (STED). (h) imaging (representative images of n = 3 (g) or n = 2 (h) independent preparations). The distribution of fluorescent signal indicates the presence of Lokiactin-based cytoskeletal structures in cell bodies and protrusions, being consistent with observations from cryotomograms. The images in h show single slices of representative deconvolved STED images detecting Lokiactin (red/orange, abberior STAR 580-labelled secondary antibodies) and DNA (blue, SPY505-DNA). For g and h, scale bars, 1 μm. Source
The network of filaments was particularly well resolved by cryo-ET imaging of very thin (<100 nm) protrusions (Figure 6, e+f), which contained sometimes bundled filaments that connected different parts of the cell. Ribosomes were homogenously distributed throughout the cell body, cell bridges and protrusions (Figure 6, c–f), where they could sometimes be observed as membrane-associated chains (Figure 6, g), but except for two instances internal membrane‑ bound compartments were not found. The unique cell envelope was mostly continuous between the cell body and protrusions, even though the transition zones showed high variability. Some appeared as stable junctions (Figure 6, j) with potential densities appearing to stabilize the 'neck', whereas others formed very thin constrictions (Figure 2.2, c+d) or were only loosely attached, and single cytoskeletal filaments frequently traversed the junctions into the protrusions (Figure 6, l).
In the cell envelope, the single membrane was found not only decorated with a layer of small unordered densities, but also with a plethora of structures further protruding from the membrane (Figure 6, c; Figure 6, l (blue arrowheads)). Some densities connected different parts of the protrusion network (Figure 6, d+f), whereas others formed elaborate assemblies that localized to regions of high membrane curvature (Figure 6, f; Figure 6, h). In two instances, the researchers detected putative chemoreceptor arrays on the cytoplasmic side of the membrane (Figure 6, i). The genome of Loki‑B35 encodes a full set of chemotaxis proteins, which are absent from the MK‑D1 genome, but also present in a number of other 'Asgard' archaea. Together with the extensive repertoire of surface proteins, these may mediate cell–cell communication and interactions.
Actin and tubulin filaments in Loki‑B35 cells
Note Would you, maybe, need a primer on actin and tubulin filaments? Here is a two-minute explainer by eukaryotic cell biologist Alex Long from UCSF in San Francisco CA, USA.
The observations that filaments connected different parts of the protrusion network and often extended across constricted membrane tubes (Figure 6, f) indicated to the researchers that the cytoskeleton functions as a scaffold to maintain the elaborate cellular architecture of Loki‑B35. Could they be actin filaments? This was an obvious assumption, as archaea generally express one to several actins that are more closely related to eukaryotic actins than to bacterial actins (MreB, MamK). The Loki‑B35 genome encoded four actins, one of which, Lokiactin, is specific for 'Asgard' archaea (see here for a phylogeny).
In panel a of Figure 7, you see a slice through a cryo-tomogram with a cytoskeletal filament inside a protrusion at high magnification. Filament segments were extracted from the cryo-tomograms for the structural analysis (Figure 7, b), which proceeded by helical reconstruction of 2D-projected filament particles from sub-tomograms (Figure 7, c). The reconstructed filament particles show the typical twisted double-stranded architecture of actin filaments like F‑actin and Crenactin (Figure 7, c).
Rodrigues-Oliveira et al. (2022) could prove that the filaments found in Loki-B35 cells are indeed Lokactin filaments using STED microscopy (a technique explained here in STC) with specific anti-Lokactin antibodies. The cell body is "stained" blue for DNA (chromosome) in Figure 7, h, while the red/orange "staining" by labeled secondary antibodies against the Lokactin-specific primary antibodies extends from the cell body into and throughout the protrusions.
(click to enlarge)
Figure 8. Cryo-tomograms of Loki‑B35 show various cytoskeletal elements in addition to Lokiactin. (A–D) slice (A) and corresponding segmentation (B), showing a long tube-like structure with a diameter of 12–14 nm spanning the entire body of a Loki‑B35 cell. (click image to see panels A–H) (C) and (D) show magnified views of tube ends and potential structures that connect the tube with the cell envelope. (E and F) Cryo-tomogram (E) and segmentation (F) of a tubular structure (diameter of 16–18 nm) connected to the cell envelope. (G and H) An example of a thin (9 nm) tubular structure in membrane tubes of cellular protrusions. (A, C–E, and G) Orange arrowheads indicate Lokiactin filaments, green arrowheads indicate other tubular densities. Scale bars: 100 nm. Source
Loki‑B35 cells not only have a cytoskeleton of Lokiactin filaments, but also of tubulin filaments, which I only learned while writing this paragraph. Wollweber et al. (2025) could clearly identify them in cryo-tomograms (Figure 8). They localize in the cell body and also in the protrusions, are mostly short, but occasionally long, and then run through the cell body in its entire length with hints of membrane anchoring (Figure 8, b). Phylogenetic studies revealed that Loki‑B35 and a few but not all 'Asgard' archaea express tubulins related to eukaryotic α/β-tubulin and bacterial BtubA/B tubulins. Heterologous expression and polymerization assays in vitro of the Loki-B35 tubulins AtubA, AtubB, and AtubB2 combined with structural studies by electron microscopy allowed the authors to conclude that the Loki-B35 tubulins AtubA, AtubB, and AtubB2 assemble into canonical AtubA/B and non-canonical AtubA/B2 heterodimers that polymerize into microtubules of 5 or 7 protofilaments, respectively (click here for a diagram).
Proteome analysis revealed that Lokiactin and two of the tubulins, AtubA and AtubB, were among the 100 most highly expressed proteins in Loki-B35. However, the researchers employed ultrastructure‑expansion microscopy (U-ExM) on live cells and discovered that only Lokiactin filaments are found consistently throughout the entire lifecycle (cells fixed at 35, 41, 59, or 91 d after inoculation), while AtubA/B filaments were only detectable in a subset of young cells (cells fixed 28 or 35 d after inoculation). The functional role(s) of tubulin filaments in L. ossiferum Loki‑B35 therefore remains a big unknown.
In the forthcoming, final part of Cultivating the Ancestors..., I will introduce two more family members, Margulisarchaeum peptidophila HC1 and Flexarchaeum multiprotrusionis SC1, both distant cousins of Promethearchaeum syntrophicum MK‑D1 (parts 1+2) and Lokiarchaeon ossiferum Loki-B35. However, the family resemblance is convincing.
Do you want to comment on this post? We would be delighted to hear from you! Please send us an email, or comment on Mastodon or Bluesky. We'll be posting comment threads that arise below (with your permission). Feel free to subscribe if you'd like to get regular notifications about new blog posts.
Comments