Organic biomorphs may be better preserved than microorganisms in early Earth sediments

Reconstructing the early history of life on Earth relies on our ability to identify the remnants of microorganisms in the rock record. Traces of ancient microbial life can be chemical (e.g., biologically derived organic compounds or isotopic signals) or morphological (e.g., microfossils), but their correct identification is rife with challenges (Schopf et al., 2010; Javaux, 2019). One of the main issues is the potential for abiotic processes to generate microstructures mimicking prokaryotic cell morphologies, and thus create “false positive” evidence of early life in the geological record. Different types of microstructures emulating microbial cells have been described from laboratory-based (Fox and Yuyama, 1963; Barge et al., 2016; McMahon, 2019; García-Ruiz et al., 2020) and field-based (Brasier et al., 2002; Marshall et al., 2011; Wacey et al., 2018) studies. Recently, organic microstructures bearing both morphological and chemical resemblance to microbial filamentous and spherical cells were synthesized experimentally (Cosmidis and Templeton, 2016). These objects, called here organic biomorphs, self-assemble when sulfide is oxidized in the presence of organics. They form spontaneously in a wide range of geochemical conditions relevant to modern and ancient sulfidic environments. Furthermore, the organic biomorphs form from diverse organic compounds, including simple prebiotic organics (e.g., glycine) (Cosmidis et al., 2019), and so possibly assembled in Earth's early environments prior to the origin of life.

Figure 1 compares the morphologies of laboratory-synthesized organic biomorphs with putative organic microfossils from the Archean and Proterozoic record, namely the 3.5 Ga Dresser Formation (Baumgartner et al., 2019) (Fig. 1A), 3.4 Ga Strelley Pool Formation (Wacey et al., 2011) (subset in Fig. 1C), and 2.4–2.2 Ga Turee Creek Group (Schopf et al., 2015; Barlow and Van Kranendonk, 2018) (Figs. 1C, 1G, 1K, 1M, and 1O) of Western Australia, and the 1.9 Ga Gunflint Formation (Wacey et al., 2013) (Figs. 1E and 1I) in northeastern North America. Organic biomorphs shown for comparison include (1) spheroids, either isolated or clustered (Fig. 1D); (2) tubes or filaments, either rigid (Figs. 1J and 1L) or flexible (Figs. 1B and 1N), in some cases branching (Fig. 1L) or forming angles (Fig. 1H), isolated (Fig. 1H), clustered (Fig. 1J), or forming “rivers” (Fig. 1N) or tangled networks (Fig. 1P); and (3) rosettes of filaments (Fig. 1F).

These morphological similarities are insufficient for disproving the biogenicity of the putative microfossils exemplified here, which in some cases has been well established (e.g., the Gunflint microfossils; Lepot et al., 2017). Yet, this comparison clearly cautions against using morphological criteria for identifying microfossils, especially in rocks marked by sulfide-rich conditions.

Euxinic conditions were common in Proterozoic oceans (Lyons et al., 2009), and reduced sulfur species could have been locally concentrated in Archean sedimentary environments despite low global oceanic sulfate (e.g., Marin-Carbonne et al., 2018). Precambrian microfossils such as those shown in Figure 1 are commonly closely associated with pyrite (e.g., Wacey et al., 2011, 2013; Baumgartner et al., 2019), indicating localized sulfide enrichments and hence conditions expedient for the formation of organic biomorphs. It is currently unknown, however, whether organic biomorphs would be preserved as pseudofossils in the rock record.

Chert formations (i.e., silica-rich rocks) host a majority of the Precambrian putative microfossils (Javaux and Lepot, 2018; Javaux, 2019). We investigated the preservation potential of organic biomorphs in silica-rich depositional environments through experimental silicification. Freshly collected mats of the sulfur-oxidizing bacterium Thiothrix sp. were used as a microbial control. While most previous experimental silicification studies have focused on the preservation of cyanobacteria (e.g., Benning et al., 2004; Lalonde et al., 2005), many putative Precambrian microfossils have been interpreted as sulfur-cycling microorganisms (Wacey et al., 2011; Schopf et al., 2015; Barlow and Van Kranendonk, 2018). Furthermore, Thiothrix forms intracellular sulfur globules (Larkin and Strohl, 1983), making it a good biological analog for the sulfur-encapsulating organic biomorphs.

Organic biomorphs were obtained by reacting sulfide in solution with dissolved yeast extract (a complex organic mixture) under sterile conditions. Following earlier experimental silicification protocols (Toporski et al., 2002; Benning et al., 2004), the biomorphs were dispensed into a sterile supersaturated sodium-metasilicate solution and stored at room temperature for up to five months. The same procedure was applied to freshly collected mats of Thiothrix sp., a filamentous sulfur-oxidizing bacterium (Larkin and Strohl, 1983). Samples were collected at different times during silicification and characterized using scanning electron microscopy (SEM) and transmission electron microscopy (TEM) coupled with energy-dispersive X-ray spectroscopy (EDS), Raman spectro-microscopy, attenuated total reflectance–Fourier transform infrared spectroscopy (ATR-FTIR), X-ray absorption near-edge structure (XANES) at the sulfur K-edge, and scanning transmission X-ray microscopy (STXM) at the carbon K-edge, nitrogen K-edge, and sulfur L-edge. See the Supplemental Material¹ for full details.

The biomorphs used for silicification consisted of spheres (0.5–3 μm in diameter) interspersed with filaments (0.1–1 μm thick). As experimental silicification proceeded, the spherical biomorphs mostly retained their original morphology, while silicified Thiothrix cells rapidly evolved from cylindrical shapes into deformed, flattened ribbons (Fig. 2). Deflated cells are not artifacts of sample preparation or imaging mode, given that non-silicified Thiothrix cells retained their cylindrical shapes in SEM images. On the contrary, despite some fragmentation during silicification, many filamentous biomorphs remained intact four weeks after the beginning of the experiment (Fig. 2D). The biomorphs were morphologically preserved by rapid (within 24 h) precipitation of nano-colloidal silica on their surfaces (Fig. 2B), eventually forming a thin crust around the spheres and the filaments (Fig. 3). In contrast, a thick silica gel formed around Thiothrix cells (Fig. 2G). This difference in the extent of silica polymerization may be attributed to the presence of extracellular polymeric substances around Thiothrix (Benning et al., 2004; Lalonde et al., 2005), or to metabolic activity of the cells, which initially decreased the pH of the silicification medium, increasing silica saturation (Fig. S1 in the Supplemental Material). In contrast, the pH of biomorph silicification medium remained constant at pH 7 (±0.5 pH units) throughout the experiment.

Elemental sulfur, initially present as intracellular amorphous globules inside Thiothrix, progressively diffused from the cells during silicification (Fig. 3). Many sulfur-depleted Thiothrix filaments were observed one week into the experiment, while sulfur reprecipitated as the metastable allotrope β-S₈ in the extracellular medium (Fig. S2). Likewise, sulfur diffused from the biomorphs during silicification, leaving behind empty organic vesicles (Fig. 3), but it also accumulated along the envelope of the silicified biomorphs (Fig. 3F). Originally enclosed as γ-S₈, sulfur re-precipitated post-silicification as α-, β-, and γ-S₈ (Fig. S2). Sulfur K-edge XANES spectroscopy also identified oxidized sulfur (sulfate, thiosulfate, and/ or sulfones and ester sulfates) associated with silicified biomorphs (Fig. S3, and Supplemental Material text).

Sulfur diffusion from Thiothrix cells and biomorphs during silicification indicates solubilization, most likely as polysulfides (Kleinjan et al., 2005). Polysulfides are highly reactive toward organics, causing rapid organic-matter sulfurization (Raven et al., 2016). The sulfur-rich organic envelopes of the silicified biomorphs (Fig. 3F) may indicate sulfurization, and ATRFTIR analysis showed the formation of sulfur-bearing groups such as organic sulfates and sulfonic groups during silicification. However, intramolecular sulfur incorporation in silicified biomorphs could not be confirmed due to possible interference with siloxane groups (see Fig. S4, and Supplemental Material text). Early dia-genetic sulfurization favors the preservation of organic microstructures in ancient rocks (Lepot et al., 2009), so determining whether organic biomorphs can be sulfurized will aid in the evaluation of their preservation potential.

We note that the sulfur loss from the organic biomorphs during silicification would likely give rise to chert-hosted organic microstructures that do not contain sulfur-bearing minerals. Instead, in iron- and sulfide-rich environments, the diffusion of polysulfides from the biomorphs would result in pyrite precipitation in their close vicinity (Rickard and Luther, 2007), as is commonly observed in chert-hosted ancient organic microstructures (e.g., Wacey et al., 2011, 2013; Schopf et al., 2015; Baumgartner et al., 2019).

As determined using ATR-FTIR (Fig. S4) and STXM (Fig. 4), the organic composition of the biomorphs prior to silicification was dominated by carboxylic groups and unsatu-rated carbon, along with aliphatics, alcohols, and C = O groups. As silicification proceeded, the aliphatic, aromatic, and unsaturated carbon content of the biomorphs increased, while signals from other organic components decreased. In contrast, the spectroscopic signature of both native and silicifying Thiothrix cells was dominated by amide groups from proteins. Previous studies have shown that amides can be incorporated into organic biomorphs when peptides are present in their synthesis medium (Cosmidis and Templeton, 2016). Here, amides were not detected in the biomorphs, although they formed in the presence of yeast extract, which contains peptides. Nitrogen in the biomorph envelope was instead present in an unidentified inorganic or organic form (see the Supplemental Material). Thus, carbon and nitrogen speciations of the Thiothrix cells versus the biomorphs are notably different.

The initial nitrogen/carbon (N/C) ratio of the biomorphs, determined using STXM (Alleon et al., 2015), was 0.27 ± 0.02 before silicification and increased to 0.40 ± 0.02 after silicification (Fig. 4I). The initial N/C ratio of Thiothrix cells was 0.16 ± 0.02 (within the range of previously measured bacteria; Alleon et al., 2017). This ratio increased after one week of silicification to 0.42 ± 0.02, a value similar to that of silicified biomorphs, but higher than N/C values reported for Precambrian microfossils (0.09–0.25) (Alleon et al., 2016b, 2018). It remains to be determined how the organic composition and N/C ratio of the silicified biomorphs may be affected by diagenesis at high temperature and pressure (e.g., Alleon et al., 2016a). Differences in the molecular compositions of the biomorphs compared with microbial cells may lead to different chemical behaviors during diagenesis, allowing for fossil biomorphs and bacteria to be distinguished in rocks based on their thermally matured residues (Alleon et al., 2017).

Organic biomorphs that form via the abiogenic reaction of sulfide with organics are likely to be preserved as pseudofossils in cherts through rapid silica encrustation and, possibly, organic-matter sulfurization. In addition to their striking morphological resemblance to putative Precambrian microfossils, these pseudofossils would share chemical characteristics with actual fossil bacteria, such as complex organic compositions and presence of nitrogen atoms. Importantly, organic biomorphs demonstrate better morphological preservation compared with microbial cells during silicification.

While these findings do not refute previous interpretations of Precambrian putative micro-fossils, our results call for a circumspect approach when evaluating the biogenicity of chert-hosted organic microstructures. In particular, a bacterial morphology, even when associated with an organic composition, can no longer inform biogenicity, especially in geochemical contexts indicative of sulfidic environments. We argue for a concerted effort to better identify potential abiosignatures (i.e., signatures of abiotic processes) (Chan et al., 2019) in the geological record, particularly those that mimic genuine biosignatures. This work is fundamental to avoiding potential pitfalls of biological “false positives” and increasing confidence in our ability to detect life in ancient rocks and astrobiological explorations.

We thank Jennifer Williams (Pennsylvania State University [PSU]) for her help with the collection of Thiothrix samples; Brandi Cron and Jennifer Macalady (PSU) for assistance with their taxonomic identification; Maxwell Wetherington (PSU) for his help with Raman analyses; and Tawanda (TJ) Zimudzi (PSU) for his help with FTIR measurements. Sulfur K-edge XANES analyses were performed on beamline 4–3 of the Stanford Synchrotron Radiation Light-source (SSRL; Stanford, California, USA) with the assistance of Matthew Latimer and Erik Nelson. Use of the SSRL, SLAC National Accelerator Laboratory, is supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences under contract DE-AC02-76SF00515. The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of the U.S. National Institute of General Medical Sciences or the National Institutes of Health. We thank Jian Wang for his support on beamline SM of the Canadian Light Source (CLS; Saskatoon, Saskatchewan, Canada). CLS is supported by the Canada Foundation for Innovation, Natural Sciences and Engineering Research Council of Canada, the University of Saskatchewan, the Government of Saskatchewan, Western Economic Diversification Canada, the National Research Council Canada, and the Canadian Institutes of Health Research. This work was supported through startup funding provided by the PSU Department of Geosciences to Cosmidis, following on preliminary work supported by the NASA Astrobiology Institute Co-operative Agreement NN15BB02A. We thank Craig Marshall and an anonymous reviewer for their insightful comments.