Skip to main content
NCBI home page
Nature Communications logoLink to Nature Communications
. 2026 Feb 21;17:1987. doi: 10.1038/s41467-026-69826-x

Survival of cyanobacteria and mitigation of Fe(II) toxicity effects in a silica-rich Archean ocean

Carolin L Dreher 1, Olaf A Cirpka 2, Manuel Schad 3,4, Kurt O Konhauser 3, Andreas Kappler 1,5,✉
PMCID: PMC12932812  PMID: 41723153

Abstract

Banded iron formations (BIF) were deposited abundantly between 2.7-2.4 Ga from iron- and silica-rich oceans, with cyanobacterial oxygen (O2) as a possible oxidant for Fe(II)(aq) oxidation and Fe(III) oxyhydroxide precipitation. However, toxic reactive oxygen species (ROS) from Fe(II)/O2 interactions might have inhibited cyanobacterial growth, contributing to the delay between cyanobacterial evolution (>3.0 Ga) and the Great Oxidation Event (2.5 Ga). Here, we explored the impact of Fe(II)(aq) and SiO2(aq) on Synechococcus sp. PCC 7002. High Fe(II)(aq) ( > 500 µM) increased ROS formation, but elevated SiO2(aq) (2200 µM) suppressed ROS formation, promoting growth and O2 production. Diel light cycles further reduced ROS formation compared to continuous illumination. Modelling O2 distribution based on experimental rates revealed oxygenated surface waters at relevant upwelling rates. Together, our results indicate that high SiO2(aq) and day-night-light cycles in Archean oceans mitigated ROS stress, enabling cyanobacterial proliferation and enhancing their role in Earth’s oxygenation and BIF deposition.

Subject terms: Ocean sciences, Biogeochemistry, Climate sciences

This study shows O2-producing cyanobacteria likely prospered in silica- and iron-rich ancient oceans prior to the emergence of free O2. Silica and day-night light cycles helped the microbes to overcome toxicity effects caused by oxygen radicals.

Introduction

Sometime between the evolution of oxygen (O2)-producing cyanobacteria, possibly as early as 3.4 billion years ago1,2 and the permanent rise of atmospheric oxygen during the Great Oxidation Event (GOE) at 2.5–2.3 Ga3, free O2 began to accumulate in seawater. This accumulation likely began in the form of localized and/or transient oxygen oases, with estimated O2 concentrations ranging from 5 to 100 µM4–7, before episodically expanding across broader regions of the continental shelf8–10. The reasons for the protracted delay between the emergence of cyanobacteria and the GOE remain unresolved but may, at least in part, reflect environmental constraints that inhibited cyanobacterial growth and dispersal11,12.

One of the most distinctive sedimentary archives of the Archean, BIFs, may hold clues to these limiting environmental conditions (see13 for review). These chemical sediments consist predominantly of iron- (15–40 wt.% Fe) and silica-rich (40–60 wt.% SiO2) layers that precipitated from seawater following the oxidation of Fe(II)(aq), with estimated concentrations ranging from 0.03 to 0.5 mM14,15, and possibly exceeding 1.0 mM in some settings16,17. Prior to the emergence of silica-precipitating microorganisms (e.g., diatoms), silica inputs from chemical weathering of early continental crust and hydrothermal vents led to elevated dissolved silica concentrations in the seawater. These concentrations were constrained chiefly by saturation with respect to cristobalite or amorphous silica, on the order of 0.67-2.2 mM18,19. Recent studies further suggest that under these ‘high silica’ concentrations, the primary precipitate may have been a gel-like composite of ferric oxyhydroxide and silica20–22.

While the deposition of the oldest BIF was probably linked to anoxygenic phototrophic Fe(II)-oxidizers23,24, photochemical Fe(II) oxidation25,26 or abiotic deposition processes such as chemical greenalite deposition27,28, the proliferation of cyanobacteria would have accelerated Fe(II) oxidation and thus BIF deposition13,29. Cyanobacteria fix and reduce CO2 using water as electron donor to form biomass, also yielding O2 as by-product (reaction 1). As such, they were key players in the oxygenation of Earth30,31.

6CO2+12H2O→hνC6H12O6+6O2+6H2O 1

In early Precambrian oceans, the O2 may have reacted with Fe(II)(aq) to form Fe(III)(aq), which hydrolyzed into Fe(III) oxyhydroxide minerals. The reaction between dissolved Fe(II)(aq) and O2 may also yield reactive oxygen species (ROS), that is, highly reactive oxygen-containing molecules (i.e., the Fenton reaction32):

O2+Fe2+→Fe3++O2−∙ 2
O2−∙+Fe2++2H+→Fe3++H2O2 3
Fe2++H2O2→Fe3++OH−+OH∙ 4

Briefly, superoxide (O2−•), generated by the reaction of O2 and Fe(II)(aq) (reaction 2; note we provide actual chemical species here versus the generic Fe(II)), further reacts with Fe(II)aq to produce Fe(III)(aq) and hydrogen peroxide (H2O2) (reaction 3). The H2O2 then oxidizes the remaining Fe(II)(aq) to Fe(III)(aq), along with the formation of hydroxide anions (OH−) and hydroxyl radicals (OH•) (reaction 4). While the Fe3+ drives the formation of Fe(III) minerals, the generated radicals can undergo further reactions, forming additional ROS like ozone (O3) and hydroperoxyl radicals (OOH•) (for a review see ref. 33).

ROS are harmful to cells; extracellular ROS can damage cell membranes through oxidative stress34, while intracellular ROS can oxidize critical biomolecules such as RNA, DNA, proteins, and lipids35. Therefore, ROS formation in Archean oceans may have profoundly influenced microbial activity and evolution. Previously, Swanner and colleagues investigated ROS production under early ocean analog conditions using phosphate-buffered saline (PBS) media and varying concentrations of Fe(II)(aq) (0.1 and 1 mM) in the presence of Synechococcus sp. PCC 7002 at a density of 2×107 cells/mL12. Under continuous light, they found a 4-fold increase in intracellular ROS levels in the presence of 1 mM Fe(II)(aq) compared to 0.1 mM Fe(II)(aq), suggesting potential toxicity to cyanobacteria at elevated Fe(II)(aq) concentrations. Additionally, in incubation experiments of cyanobacteria under initially anoxic conditions (in the presence of up to 300 µM Fe(II)aq) reduced autofluorescence of the cyanobacteria was reported, accompanied by significantly lower actinic yield, and decreased growth rates at Fe(II)(aq) concentrations >180 µM.

In a follow-up study, Swanner and colleagues conducted Fe(II) oxidation experiments with Synechococcus sp. PCC 7002 under continuous illumination and a range of Fe(II)(aq) concentrations (7.5, 29, 577, and 4805 µM)11. At lower Fe(II)aq concentrations, cultures changed from colorless to green, indicating a high density of cyanobacterial cells. By contrast, higher Fe(II)(aq) concentrations produced an orange coloration, consistent with the formation of Fe(III) oxyhydroxide minerals. These observations were supported by direct cell counts, which confirmed reduced growth at higher Fe(II)(aq) concentrations, again suggesting toxic effects. Microscopic analyses further revealed that cells grown under high Fe(II)(aq) concentrations were smaller and exhibited reduced pigment content, as indicated by lower carotenoid and chlorophyll concentrations inferred from optical density measurements. Although Fe(II) oxidation rates increased with rising Fe(II)(aq) concentrations (from 13 µM/day at 29 µM to 475 µM/day at 4805 µM) the rate of O2 production in the liquid phase decreased from 120 µM/day to 47 µM/day, indicating diminished cellular activity. Notably, O2 was already detectable during Fe(II) oxidation at 29 and 475 µM Fe(II)(aq) but absent at 4805 µM Fe(II)(aq). Taken together, these findings suggest that elevated Fe(II)(aq) concentrations in Archean seawater could have imposed oxidative stress, limiting cyanobacterial productivity and expansion. This could potentially explain the lag between the evolution of cyanobacteria and the GOE11,12.

However, the potential mitigating impact of high SiO2(aq) concentrations on Fe(II)(aq) toxicity has not been previously considered. Our working hypothesis is that silica binds Fe(II)aq, forming Fe-Si-aggregates that slow Fe(II) oxidation and consequently inhibit ROS formation. Previous studies have shown that reduced Fe(II) oxidation rates lead to lower ROS formation36. Consistent with this, Dreher and colleagues37 showed, using SEM/EDS, that Fe-Si aggregates precipitate under similar experimental conditions, further supporting our hypothesis. To test this hypothesis, we experimentally examined the combined effects of Fe(II)(aq) and SiO2(aq) on cyanobacterial O2 production, Fe(II) oxidation, and potential ROS-related toxicity. Specifically, we incubated Synechococcus sp. PCC 7002 under alternating day-night-cycles (16 h light; 8 h dark) in artificial seawater medium, in the absence of SiO2(aq) (0 µM, which we refer to as ‘no-silica’) and in the presence of SiO2(aq) (2200 µM, which we refer to as ‘high-silica’) and varying initial Fe(II)(aq) concentrations (0, 500, 2500, 5000 µM). Over the course of the experiments, we monitored O2 production, Fe(II)aq and SiO2aq concentrations, total iron (Fe(tot)), cell numbers, and ROS production to assess the interdependent effects of Fe(II)(aq) and SiO2(aq) on cyanobacterial physiology and oxidative stress.

Results

Effects of different silica concentrations on Fe(II) oxidation by cyanobacterial O2 under alternating day-night-cycles

We conducted Fe(II) oxidation experiments, under initially anoxic conditions, using Fe(II)(aq) concentrations ranging from 0.5 to 5 mM. Following inoculation with the cyanobacterial strain Synechococcus sp. PCC 7002, we monitored Fe(II), Fe(tot), cyanobacterially produced O2, dissolved Si, and cell counts over time (for sterile controls see Fig. S3). The results are plotted in Fig. 1 on the respective timescale of the active Fe(II) oxidation, as this sets the focus on the period when there was still Fe(II)(aq) in the system, which is the most relevant for determining toxicity effects and the latter modeling. In setups without amendment of additional Fe(II)(aq), all cultures exhibited a color change from clear to green within two days, regardless of SiO2(aq) concentration (Fig. 1C), indicating cyanobacterial growth. In the ‘high silica’ setups compared to the ‘no silica’ setups, cell growth was initially more rapid but reached a slightly lower maximum cell density (Fig. 1C, Table 1). Under both experimental conditions, O2 concentrations peaked at 500 µM after 60 days (Fig. 1B). In the setups without additional Fe(II) amendment, the Fe(II)(aq) concentration remained relatively stable throughout the experiments, averaging approximately 50 µM in with and without both silica setups (Fig. 1A), a level attributable to the Fe(II)(aq) added as nutrient in the medium. In the ‘high silica’ setups, the initial SiO2(aq) was ca. 1600 µM (Fig. S1A). However, after 30 days, solid silica precipitation started, ultimately reducing the SiO2(aq) concentration to several hundred µM by the end of the experiment.

Fig. 1. Geochemical data of Fe(II) oxidation experiments containing Synechococcus sp. PCC 7002.

Fig. 1

Open in a new tab

Panels A–C show the results of setups without amendment of Fe(II)(aq). Panels D–F show the results of setups with amendment of 500 µM Fe(II)(aq), panels G–I with 2500 µM Fe(II)(aq), and panels J–L with 5000 µM Fe(II)(aq). Filled symbols indicate setups with 2200 µM silica (‘high silica’), while empty symbols indicate setups without silica (‘no silica’). Fe(II) (gray), accumulated oxygen (blue), cell numbers (green) are plotted as average values from triplicates with the standard deviations shown as error bars. Please note that the time scale (x-axis) varies for the different plots because of the differences in the duration of active Fe(II) oxidation.

Table 1.

Geochemical data of the Fe(II) oxidation (FeOx) experiments

Silica [µM] Fe(II) [µM] FeOx rate µM/day Length of FeOx [days] Maximum oxygen conc. [µM] Average maximum cell count [cells/mL] Green liquid observed after x days
0 0 – – 450 after 60 days 2.5 × 108 after 87 days 2
2200 0 – – 450 after 60 days 1.7 × 108 after 87 days 2
0 500 101 ± 8 5 450 after 60 days 2.0 × 108 after 87 days 2
2200 500 96 ± 10 5 450 after 60 days 2.0 × 108 after 87 days 2
0 2500 129 ± 5 21 350 after 100 days 2.3 × 108 after 140 days -
2200 2500 174 ± 11 17 490 after 100 days 1.7 × 108 after 87 days 31
0 5000 77 ± 6 58 290 after 140 days 0.6 × 108 after 140 days –
2200 5000 72 ± 20 58 100 after 140 days 2 × 108 after 140 days 87
Open in a new tab

The FeOx rate [µM/day] of the triplicates with the standard deviation, the length of active FeOx [days], the average maximum accumulated oxygen concentration [µM], the average maximum cell count [cells/mL,] and the amount of days after which the liquid turned green of the setups containing 0 µM added silica (‘no silica’) and 2200 µM added silica (‘high silica’) with amendment of 0–5000 µM Fe(II)(aq).

In experiments with amendment of 500 µM of Fe(II)(aq), all bottles turned green within two days, regardless of the SiO2(aq) concentration (Fig. 1F). Growth curves were similar across setups, with no discernible effect of SiO2(aq) (Fig. 1F, Fig. 2,Table 1), and maximum cell densities were reached after 87 days (Fig. 2). In both ‘no silica’ and ‘high silica’ setups, O2 accumulated in the headspace following complete Fe(II) oxidation, peaking at 450 µM after 60 days (Fig. 1E).

Fig. 2. Geochemical data of Fe(II) oxidation experiments containing Synechococcus sp. PCC 7002 over the entire course of the experiment.

Fig. 2

Open in a new tab

Panels A–C show the results of setups without amendment of Fe(II)(aq). Panels D–F show the results of setups with amendment of 500 µM Fe(II)(aq), panels G–I with 2500 µM Fe(II)(aq), and panels J–L with 5000 µM Fe(II)(aq). Filled symbols indicate setups with 2200 µM silica (‘high silica’), while empty symbols indicate setups without silica (‘no silica’). Fe(II) (gray), accumulated oxygen (blue), cell numbers (green) are plotted as average values from triplicates with the standard deviations shown as error bars.

Within the first five days of the experiment, 80–90% of the added 500 µM Fe(II)(aq) was oxidized in both ‘no silica’ and ‘high silica’ setups (Fig. 1D). With and without silica, the Fe(II) oxidation rates were similar within the calculated error (Table 1). In the ‘no silica’ and ‘high silica’ setups, Fe(II) oxidation rates were similar at about 100 µM/day (Fig. 1D; Table 1). In the ‘high silica’ setups (Fig. 1B), silica precipitation began after 60 days, reducing SiO2(aq) to approximately 100 µM.

In experiments amended with 2500 µM of Fe(II)(aq), the bottles turned orange in the absence of added silica (Fig. 1I), contrasting with experiments with 0 and 500 µM added Fe(II)(aq). However, in the presence of silica, the 2500 µM Fe(II) bottles turned yellow by day 21 and green by day 31. The orange coloration in the ‘no silica’ setup suggests abiotic Fe(II) oxidation and the formation of Fe(III) oxyhydroxide minerals, likely caused by cyanobacterially produced O2, but with limited cell growth and metabolic activity. By contrast, the green coloration in the ‘high silica’ setups indicates extensive cyanobacterial proliferation. It is important to interpret these color changes cautiously, as higher Fe(II) concentrations can intensify the orange hue regardless of biological activity, potentially confounding visual comparisons between setups.

Cell counts revealed that in the ‘no silica’ setups, cyanobacterial growth was delayed, with cell numbers increasing only after 25 days (Fig. 1I). Conversely, cell proliferation in the ‘high silica’ setups commenced immediately after inoculation and ultimately reached similar densities by day 150 (Fig. 1I). Silica availability had a marked influence on O2 accumulation. In the ‘no silica’ setups (Fig. 1H), O2 accumulated only after day 20 (ca. 50 µM by day 31). In comparison, in the ‘high silica’ setups (Fig. 1H) O2 accumulated as early as day 10, reaching 150 µM by day 31.

In the ‘no silica’ setups, the Fe(II) oxidation rate was lower compared to the ‘high silica’ setups (Fig. 1G; Table 1). Initial SiO2(aq) concentration in the ‘high silica’ setups were 1500 µM (Fig. S1C, Fig. 1C) and decreased to 700 µM within the first 20 days. By day 90, variability emerged among the triplicate bottles: one replicate maintained a stable SiO2(aq) concentration of 700 µM, whereas the other two showed continued declines to 300 µM and 100 µM, respectively.

In experiments amended with 5000 µM of Fe(II)(aq), the ‘no silica’ setups remained orange until day 87 (Fig. 1L), indicating Fe(III) oxyhydroxide formation and limited cyanobacterial growth. By contrast, the ‘high silica’ setups turned yellow by day 21 and green by day 87, reflecting more substantial cyanobacterial growth. Consistent with these color changes, total cell numbers were higher in the ‘high silica’ setups. In the ‘no silica’ setups, cell growth was delayed until Fe(II) oxidation was complete, after which cell densities increased to 5 × 10⁷ cells/mL (Fig. 1L). In one replicate, cyanobacteria eventually recovered, reaching 1.5 × 10⁸ cells/mL by day 140 Fig. 2. In the ‘high silica’ setups, final cell densities varied among the three replicates, reaching 2.5 × 10⁸, 1.5 × 10⁸, and 5 × 10⁷ cells/mL, respectively (Fig. 1L).

O2 accumulated in the ‘no silica’ setups after 60 days, reaching ca. 80 µM by day 87 (Fig. 1K), although replicate variability was high: one bottle showed no detectable O2, while another reached 300 µM O2 by day 140. In contrast, O2 concentrations in the ‘high silica’ setup increased after 57 days (Fig. 1K) and reached 200 µM by day 87. Two out of three replicates further reached 400 µM O2 by day 100, whereas in the third, in which only 60% of the Fe(II) was oxidized, showed no detectable O2.

The Fe(II) oxidation rates were similar between the ‘no silica’ and ‘high silica’ setups (Fig. 1J; Table 1), but were notably lower than the rates observed in experiments with 2500 µM Fe(II)(aq). In the ‘high silica’ setups, the initial SiO2(aq) concentration of 1500 µM declined to 500 µM by day 31 (Fig. S1D), after which it either stabilized or continued to precipitate, reaching final concentrations between 100 and 200 µM.

Improvement of long-term cyanobacterial cell viability by silica

In addition to experiments focusing on cyanobacterial growth and activity i.e., cell numbers, O2 production, resulting Fe(II) oxidation and dissolved silica (see Fig. S1), we explored the role of silica in promoting long-term cell viability experiments by observing potential mitigation effects of the harmful effects of ROS generated in the presence of both Fe(II)(aq) and O2. In all setups, the presence of Fe(II)(aq) delayed both cell growth and O2 production (Fig. 2). In the 500 µM Fe(II)(aq) (Fig. 2 E), ‘high silica’ setups initially showed higher O2 accumulation during the first 70 days, with both setups converging at 350 µM by day 60. In the 2500 µM Fe(II) experiments (Fig. 2G–I), the presence of SiO2(aq) led to significantly enhanced cyanobacterial activity: by day 150, O2 concentrations reached 500 µM and cell densities ca. 2.4 × 108 cells/mL, compared to 350 µM O2 and 1.8 × 108 cells/mL in the absence of SiO2(aq). The beneficial effect of silica was even more pronounced in the 5000 µM Fe(II) setups. After 250 days, O2 accumulation in the ‘high silica’ setups reached 250 µM, with cell densities of 2 × 108 cells/mL, while the ‘no silica’ setups showed limited cell growth, reaching only 5 × 107 cells/mL and 100 µM O2.

Formation of ROS in incubations of strain PCC 7002 with Fe(II)

Based on the lower cell growth and Fe(II) oxidation rates at higher Fe(II)(aq) concentrations (500, 2500–5000 µM), we hypothesized that cyanobacterially produced O2 reacted with Fe(II)(aq) to generate ROS, potentially leading to cellular stress or toxicity. However, our results also indicated that SiO2(aq) mitigates these negative effects, as evidenced by higher cell densities, earlier O2 accumulation, and increased Fe(II) oxidation rates in silica-amended setups. To further evaluate the impact of varying Fe(II)(aq) and SiO2(aq) concentrations on ROS production, we subsequently constructed a series of experiments using Synechococcus sp. PCC 7002 cell suspensions (5 × 109 cells/mL) exposed to different Fe(II)(aq) concentrations (0, 500, 2500 or 5000 µM), both in the absence and presence of 2200 µM SiO2(aq).

Our results (Fig. 3, Fig. S2, Table S1) showed that in the absence of Fe(II)(aq), ROS fluorescence signals were generally very low, yet slightly elevated in the presence of SiO2(aq) compared to setups without SiO2(aq) (ca. 500 a.u. and 400 a.u., respectively) (Fig. 3). After the amendment of 500 µM Fe(II)(aq), the ROS fluorescence signal, with and without silica, was similar to the setups without Fe(II), suggesting no significant ROS formation at this Fe(II)(aq) concentration. However, in the presence of 2500 µM of Fe(II)(aq), we observed a pronounced increase in ROS fluorescence in the absence of silica (2000 a.u.), while the signal remained near baseline in silica-amended setups. A similar trend was observed at 5000 µM Fe(II)(aq), where ROS levels reached 1300 a.u. without silica, but remained negligible in the presence of silica (Fig. 3). These results suggest that high Fe(II) concentrations promote ROS generation, and that dissolved silica effectively suppresses ROS formation under these conditions.

Fig. 3. Impact of silica on reactive oxygen species (ROS) formation at different iron concentrations.

Fig. 3

Open in a new tab

Average ROS fluorescence signals relative to an anoxic, abiotic control containing 5000 µM Fe(II)(aq) with and without 2200 µM SiO2(aq). The experimental setups contained strain PCC 7002 cells, either 0, 500, 2500, or 5000 µM Fe(II)(aq) in the absence (blue) and presence of 2200 µM SiO2(aq) (orange). Please note that the normalization of all measurements to the abiotic control can result in negative values. Error bars show the standard deviation calculated from quadruplicate setups.

Numerical modeling of the experimental O2 data

In order to simulate the total O2 production, we incorporated our raw laboratory data into a numerical model to simulate the total O2 production. To account for the observed plateau of O2 concentration at approximately 450 µM (Fig. 4A–E), we differentiated between ‘active’ cells (O2-producing; green solid line) and ‘inactive’ cells (non–O2-producing; calculated as the difference between total cells and measured active cells). The ‘inactive’ group included both dead cells and those estimated to have received insufficient light due to increased cell density during the experiment. While cultures were initially colorless, they gradually developed an intense dark green coloration, increasing turbidity, and thereby reducing the average light intensity available to each bacterial cell.

Fig. 4. Modeled lab data of the experimental setups containing 500 µM iron without silica (top) and with 2200 µM silica (bottom).

Fig. 4

Open in a new tab

Panel A shows the modeled data of setups without amendment of silica or Fe(II)(aq), B shows data of setups without silica amendment but with 500 µM Fe(II)(aq). D Shows setups with 2200 µM silica without Fe(II)(aq), E shows setups with 2200 µM silica and 500 µM Fe(II)(aq). Panels C and F show the data zoomed into the first 15 days of panels B and E, respectively. Green (‘cell sim.total’), blue (‘DO sim.’) and black (‘Fe(II) sim.’) lines are the modeled total cell densities and oxygen concentrations, respectively, the green circles (‘cells meas.’), the blue crosses (‘DO meas.’) and the black crosses (‘FeII) meas.’) show the associated raw data points. The bright blue dashed line (‘cells sim. active’) shows the calculated concentration of active cells. The black solid lines (‘Fe(II) sim.’) shows the modeled Fe(II) curve.

We also calculated the total O2 production in our experiments, accounting for both the dissolved O2 in the liquid and free O2 present in the headspace. Based on the equilibrium relationship between the aqueous solution and headspace (see Supplementary Information, formula 9), we estimate that O2 production (including the headspace) is approximately 42 times greater than measurements based solely on dissolved O2. This correction is crucial for accurately quantifying true O2 production in our experimental setup.

Oxygen and Fe(II) distribution modeled for the ancient water column

To scale our laboratory-derived O2 production to early ocean conditions, we calculated the oxygen production per cell (rphoto) from the first two days of the experiment, a period when no detectable O2 accumulated in the liquid phase and no measurable cell growth occurred (see “Methods” for details). We calculated the oxygen and Fe(II)aq distribution for three upwelling rates: 4, 95, and 473 m/yr. 4 m/yr represents the average upwelling velocity of modern oceans, while 95 m/yr and 473 m/yr are modeled average and enhanced oceanic upwelling velocities in the mid-Cretaceous North Atlantic Ocean, respectively38. Using an upwelling rate of 4 m yr⁻¹ (Fig. 5A–D), our model predicted O2 concentrations of at least 400 µM between 50 and 100 m depth in the water column, with the potential for substantially higher values. Given that the maximum O2 solubility in an open systems at room temperature and atmospheric pressure is ~300 µM39,40 (higher in closed systems), these results indicate that such upwelling rates would saturate and fully oxidize the photic zone, leading to complete oxidation of Fe(II)(aq) and excess O2 outgassing to the atmosphere.

Fig. 5. 1D steady-state model of oxygen distribution in the early ocean.

Fig. 5

Open in a new tab

Modeled oxygen (left, blue) and Fe(II) (right, gray) concentrations in the ancient ocean water column for an upwelling velocity of 4 (panels A–D), 95 (panels E–H) and 473 m/yr (panels I–L). Panels A, C, E, G, I, and K are calculated for hypothesized 105 cells/mL in the early ocean, B, D, F, H, J, L for 106 cells/mL. Panels A, B, E, F, I, and J are modeled based on the photosynthetic rate (rphoto) calculated from the lowest cell count (6.5 × 105 cells/mL) we measured in our experiments at d0; C, D, G, H, K, and L are based on the highest cell count (1.6 × 107 cells/mL) we measured at d0.

By increasing the upwelling rate to 95 m yr⁻¹, our model predicts that, at a cyanobacterial abundance of 10⁵ cells mL⁻¹ (Fig. 5E, G), Fe(II)(aq) would have reached the ocean surface. Oxygen concentrations at the surface would have reached between 0.04 and 0.4 µM maximum concentration. At a higher theoretical abundance of 10⁶ cells mL⁻¹ (Fig. 5F, H), however, the photic zone remained Fe(II)aq-free down to depths of ~200 m or ~125 m, respectively. In these high-cell-density scenarios, O2 accumulated to much higher concentrations within the upper 0–50 m of the water column, with modeled values reaching (200 µM) or exceeding (800 µM) maximum O2 saturation.

At an upwelling rate of 473 m/yr, Fe(II)(aq) reached the water surface in most scenarios (Fig. 5I, K, L). Under these conditions, O2 accumulation at the top of the water column was limited, ranging from 0.03 µM (Fig. 5K) to 0.8 µM (Fig. 5L). In scenario J, however, Fe(II)(aq) was absent only from the upper 50 m, allowing O2 to accumulate to 80 µM near the water surface. Overall, the results highlight the strong control of upwelling rate on O2 and Fe(II)(aq) dynamics. At low upwelling rates (4 m/yr), the photic zone was fully oxidized, whereas at 96 m/yr, O2 accumulated only near the surface. In contrast, most scenarios with rapid upwelling (473 m/yr) resulted in Fe(II)(aq) reaching the surface, suppressing O2 buildup and preventing full saturation.

Discussion

ROS formation and the mechanistic function of Si for ROS prevention

Our results demonstrate a clear relationship between Fe(II)(aq) concentrations and the extent of ROS formation in the absence of SiO2(aq), generally supporting the findings of Swanner and colleagues11,12. Differences between our study and theirs can likely be attributed to variations in experimental setups, such as constant illumination versus day-night-cycles, and static versus agitated conditions (see above). Additionally, minor methodological variations, including the use of phosphate-buffered saline (PBS)12 versus TRIS buffer in our study, as well as differences in fluorometric detection systems, may have contributed to the observed differences. Nevertheless, despite the variations, our study demonstrates a protective effect of SiO2(aq) against ROS-induced stress in cyanobacteria under conditions relevant to early Precambrian oceans. Even at high Fe(II)(aq) concentrations (2500 and 5000 µM), ROS levels remained below those observed in anoxic, abiotic controls when SiO2(aq) was present (Fig. 3). Furthermore, we measured higher O2 immediately following Fe(II) oxidation in the presence of SiO2(aq) at all Fe(II)(aq) concentrations (Fig. 1), suggesting enhanced cyanobacterial activity. Notably, in the 5000 µM Fe(II)(aq) experiments, SiO2(aq) enabled long-term recovery and sustained growth of Synechococcus sp. PCC 7002, with cell densities reaching 2×108 cells/mL after 150 days, compared with only 5 × 107 cells/mL in the absence of SiO2(aq).

The protective effect of SiO2(aq) against ROS formation can be attributed to the ability of Si to bind Fe(II)(aq), forming Fe(II)-Si complexes28,37, that reduce Fe(II) reactivity with O2, thereby inhibiting or reducing Fenton-type reactions and lowering ROS production. The formation of amorphous Fe(II)-Si aggregates, as hypothesized in our experiments, correlates well with previous studies simulating Archean ocean conditions (1 mM SiO2aq, 1.1 mM Fe(II)(aq), pH of 6.5–7.541,42;). This pH range is consistent with both our experimental conditions (pH 7) and estimates for circumneutral pH in early Precambrian oceans43,44. While Fe(II) within these aggregates remained susceptible to oxidation by O2, the resulting products likely precipitated as Fe(III)-Si-aggregates on the seafloor rather than as Fe(II)-Si phases (37).

Day-night-cycles reduce ROS formation compared to continuous light and static conditions

Our results showed that Fe(II)(aq) oxidation rates under day-night-cycle incubations were higher than those reported in a previous study using the same setup—including the same cyanobacterial strain (Synechococcus sp. PCC 7002), initial cell densities (106–107 cells/mL), similar Fe(II)(aq) and SiO2(aq) concentrations, light intensities (300–500 lux), and identical glassware—but conducted under continuous illumination (37) (Table 2). In these continuous light experiments, the presence of 2200 µM silica resulted in remarkably higher cell counts and O2 concentrations compared to the silica-free equivalents37. The difference in Fe(II) oxidation rates between day-night cycles and continuous illumination was especially pronounced at the lowest Fe(II)(aq) concentration (500 µM) in the absence of SiO2(aq), where Fe(II) oxidation rates reached 113 µM/day under day-night cycles versus only 32 µM/day under continuous light. The only exception occurred in the setup containing 2500 µM Fe(II)(aq) and 2200 µM SiO2(aq), where Fe(II) oxidation rates of 129 and 168 µM/day were observed in day-night-cycles and under continuous illumination, respectively.

Table 2.

Fe(II) oxidation rates in day-night-cycles (left column, data from this paper); Fe(II) oxidation rates in continuous light right column, data from Dreher and colleagues37

day-night-cycles continuous light
Silica [µM] Fe(II) [µM] FeOx rate µM/day FeOx rate µM/day
0 500 101 ± 7.7 32 ± 1
2200 500 96 ± 9.7 30 ± 1
0 2500 129 ± 5.1 168 ± 2
2200 2500 174 ± 10.8 147 ± 15
0 5000 77 ± 6.4 63 ± 10
2200 5000 72 ± 20.2 40 ± 2
Open in a new tab

The effects of continuous light exposure on phototrophic organisms have been well documented45, with prolonged and continuous illumination linked to increased ROS formation and cellular damage. In the context of Fe(II) oxidation, Hegler and colleagues examined the influence of light intensity and wavelength under continuous illumination, reporting maximum Fe(II) oxidation rates of 4.5 mmol/L/day at 800 lux, with overall Fe(II) oxidation rates increasing with light intensity46. More recently, Nikeleit and colleagues demonstrated that day-night-cycle exposure enhanced Fe(II) oxidation rates in the anoxygenic phototroph Chlorobium ferrooxidans strain KoFox, suggesting a physiological benefit of diel variation47. Furthermore, day-night-cycles have been shown to stimulate biomass production, protein synthesis, and coenzyme Q10 production in anoxygenic phototrophic bacteria such as Erythrobacter sp. NAP1 and Rhodobacter sphaeroides48,49.

During the light periods of day-night-cycles, cyanobacteria produce O2 most efficiently at light intensities between 30 and 50 µmol m² s-1 50–52. In dark periods, their metabolic activity is severely limited, with neither O2 nor glucose produced; energy is instead derived from the degradation of internally stored glycogen53. Muhetaer and colleagues52 showed that the effects of continuous light depend on the strain’s specific circadian rhythm. For example, M. aeruginosa exhibited increased H2O2/ OD730 ratios and partial cell death after 8 days of continuous light (e.g., H2O2/ OD730 of 400 at 300 µmol m–2 s–1) compared to two days (H2O2/ OD730 of 230 at 300 µmol m–2 s–1), whereas the strain P. galeata showed opposite trends, suggesting significant inter-strain variability in stress responses to prolonged illumination52. It is thus plausible that day-night-cycles applied in our experiments were more ideal for the cyanobacteria’s circadian rhythm than continuous light exposure. This may explain the enhanced Fe(II) oxidation rates observed in our experiments compared to previous experiments conducted under continuous illumination, likely due to higher O2 production during the light periods.

In addition to day-night-cycles, shaking the incubations enhanced cyanobacterial growth. Multiple studies have shown that Synechococcus strains benefit from fluid movement during growth, as it promotes homogenization of the medium and improves access to nutrients, light, and CO254–56. For example, Kuan and colleagues reported optimal growth rates for Synechococcus elongatus PCC7942 at shaking speeds of 100–150 rpm57. Our data are consistent with these findings, confirming the fluid agitation—analogous to conditions in the phototrophic zones of ancient oceans—stimulates cyanobacterial activity and, consequently, O2 production.

Beyond shaking, the configuration of incubation bottles also appears to influence cyanobacterial growth. More specifically, experiments with open bottles generally show faster doubling rates and higher cell densities58 compared to setups using closed bottles11,12. This difference may result from continuous headspace ventilation in open bottles, which facilitates O2 removal and potentially reduce ROS formation. However, comparisons between these studies should be made cautiously, as the Fe(II)(aq) concentrations used by Hermann and Gehringer were relatively low (20-120 µM)58, a range in which Swanner and colleagues also observed minimal ROS-related toxicity11,12.

Implications for the oxygen distribution in the seawater

From our laboratory-derived per-cell O2 production rates, we estimate that under low upwelling conditions (4 m yr⁻¹) (modern value) and a total Fe(II)(aq) concentration of 529 µM, the entire photic zone should be oxidized at cyanobacterial cell densities of approximately 10⁵–10⁶ cells mL⁻¹. At substantially higher upwelling rates (95 and 473 m yr⁻¹)38, overall O2 concentrations throughout the water column would be lower; nevertheless, cyanobacterial activity would still support O2 accumulation at the surface, forming a distinct oxygenated layer.

Incorporating the upwelling rates and Fe(II)(aq) concentrations applied in the studies of Swanner and colleagues12 into our models (see Fig. S6–S8), and applying our most conservative O2 production rate (rphoto of 0.122 µM/day), yields maximum O2 concentrations at least an order of magnitude greater than previously estimated (33 µM of maximum oxygen compared to 2–4 µM of Swanner and colleagues12). Moreover, the modest toxicity effects observed at 529 µM Fe(II)(aq) would be even less pronounced at lower concentrations (25–125 µM), as also reported by Swanner et al.11,12, suggesting that our model predictions may still underestimate true O2 production.

When applying this approach to specific scenarios, it produces results that diverge markedly from earlier models12. For example, with 25 µM Fe(II)(aq) and an upwelling rate of 95 m yr⁻¹, our model predicts O2 concentrations up to 30 µM and a Fe(II)(aq)-free zone extending to more than 200 m depth, whereas the earlier model estimated a maximum of 1–2 µM O2 and a Fe(II)(aq)-free zone less than 50 m deep. Under 25 µM Fe(II)(aq) and rapid upwelling (473 m yr⁻¹), both models predict Fe(II)(aq) transport to the surface, but our simulations still yield 2.5 µM O2 at the surface, while the earlier study predicted no measurable O2 accumulation. Finally, under low upwelling (4 m yr⁻¹) and higher Fe(II)(aq) concentrations (120 µM), our model produces 500 µM O2 with complete Fe(II)(aq) oxidation down to 400 m, in contrast to ~80 µM O2 and oxidation to only 250 m in the earlier model. Collectively, these findings demonstrate that using experimentally derived cyanobacterial O2 production rates results in substantially higher predicted O2 concentrations and deeper Fe(II)(aq)-free zones than previous modeling efforts have suggested.

Oxygenic photosynthesis on early Earth proceeded without harmful ROS stress to cyanobacteria

Our experiments involving cyanobacterial incubations at high Fe(II)(aq) concentrations in the absence of SiO2(aq) are consistent with previous studies demonstrating that O2-producing cyanobacteria were exposed to ROS under early Earth conditions11,12. However, our results suggest that ROS were not necessarily a major physiological stressor for cyanobacteria during this time. Dreher and colleagues36 reported increasing Fe(II)(aq) toxicity with increasing Fe(II)(aq) concentrations, but observed markedly enhanced bacterial growth in the presence of silica. Based on these results, we hypothesize that silica reacts with Fe(II)(aq), thus distinctly lowering Fe(II) reactivity and preventing Fe(II)-induced ROS formation. First, in experiments with iron and silica concentrations relevant to Archean oceans (500 µM Fe(II)(aq) and 2200 µM SiO2(aq)18,19), we observed rapid cell growth, high O2 production, and complete Fe(II)(aq) oxidation under simulated day-night-cycles (Fig. 2D–F). Second, we found that SiO2(aq) effectively suppressed ROS formation at Fe(II)(aq) concentrations as high as 5000 µM (Fig. 2G–L and Fig. 3). These findings imply that cyanobacteria inhabiting coastal environments with episodically elevated Fe(II)(aq) concentrations (several mM8,59) could have remained viable and active—particularly in contrast to the lower Fe(II)(aq) concentrations generally inferred for bulk seawater (0.02–500 µM respectively14,35,60–62).

The observation that ROS formation was inhibited even at high concentrations of both SiO2(aq) (2200 µM) and Fe(II)(aq) (5000 µM) suggests that these chemical conditions imposed minimal stress to cyanobacteria inhabiting early Precambrian oceans. This finding broadly supports the findings of Swanner and colleagues11,12, although in our experiments, lower Fe(II)(aq) concentrations did not induce significant ROS formation or toxicity. Taken together, our lab experiments, supported by modeling, indicate that ROS toxicity was likely not a major constraint on cyanobacterial growth or proliferation in early marine environments. Consequently, ROS stress alone is unlikely to explain the temporal lag between the emergence of oxygenic cyanobacteria ( > 3 Ga) and the onset of the GOE (2.5–2.3 Ga)31,63–65.

Our results add to the growing body of evidence demonstrating the geomicrobiological significance of Si-Fe interactions in early Precambrian oceans. For instance, SiO2(aq) binds to reactive surface sites of Fe-bearing minerals such as pyrite (FeS2), stabilizing it against oxidative dissolution and inhibiting ROS formation, a process that might have been relevant in the Archean66,67. Indeed, silica coatings on pyrite have been shown to reduce surface-bound sulfate generation during oxidative weathering by up to 91%68, indicating an increased stability of the pyrite. Furthermore, Fe(III)-silica aggregates can serve as adsorption sites for Fe(II) via outer-sphere complexation69, promoting scavenging of Fe(II)(aq) and therefore suppressing ROS production. Our results expand this knowledge by demonstrating that SiO2(aq) can inhibit ROS formation directly in the water column by binding to Fe(II)(aq), thereby contributing to a more favorable chemical environment for cyanobacterial activity in early oceans.

Methods

Cultivation of microorganisms and growth conditions

The cyanobacterium Synechococcus sp. PCC 7002, provided by Gen Enomoto from the University of Freiburg, was cultivated in liquid oxic A+ medium. This medium was prepared using the following stock solutions: Stock 1 (2 mL) (1:2 dilution, containing 3.89 g FeCl₃·6H₂O, 34.3 g H₃BO₃, 4.3 g MgCl₂·4H₂O, 0.315 g ZnCl₂, 0.03 g MoO₃, 0.003 g CuSO₄·5H₂O, and 0.0122 g CoCl₂·6H₂O in Milli-Q water to 1 L), stock 2 (10 mL) (100 g Trizma base adjusted to pH 8.2 with HCl in Milli-Q water to 1 L), stock 3 (100 mL) (0.5 g KH₂PO₄, 10 g NaNO₃, 24.4 g MgSO₄, 180 g NaCl, and 6 g KCl in Milli-Q water to 1 L), stock 4 (10 mL) (0.3 g Na₂EDTA in Milli-Q water to 100 mL), stock 5 (10 mL) (2.8 g CaCl₂ in Milli-Q water to 100 mL). The PCC 7002 culture was grown on a shaker at 60 rpm to a cell density of approximately 10⁸ cells/mL at 25 °C in an Erlenmeyer flask, under a 40-watt halogen light bulb providing an intensity of 300–500 lux.

Setup of Fe(II) oxidation experiments

The Fe(II) oxidation experiments aimed to culture the planktonic cyanobacterium Synechococcus sp. PCC 7002 in the presence of 0, 500, 2500, and 5000 µM Fe(II), along with 0 and 2200 µM dissolved monomeric silica. These experiments were set up in triplicate bottles each to analyze cell growth and accumulated oxygen concentration after complete Fe(II) oxidation by the O2 produced by the cyanobacteria. The bottles were kept at room temperature on a shaker at 60 rpm with day-night-cycles of 8/16 h at intensities of 300–500 lux. Each experiment was conducted in 250 mL Schott bottles containing 100 mL of liquid medium. Anoxic artificial seawater medium (ASW) was prepared with the following composition: 17.3 g NaCl, 8.6 g MgCl₂·6H₂O, 0.025 g MgSO₄·7H₂O, 0.99 g CaCl₂·2H₂O, 0.39 g KCl, 0.059 g KBr, 0.05 g KH₂PO₄, 0.25 g NH₄Cl, and 2.5 g NaHCO₃ per 1 L of Milli-Q (MQ) water. Salts, the bicarbonate buffer (30 mM), and the monomeric silica stock solution were separately purged with N₂ (or N₂/CO₂ at 90:10 % for the buffer) and autoclaved. After autoclaving, the salt solution in a Widdel flask was flushed with N₂/CO₂ while still hot (~70–80 °C), and the sterile silica solution was immediately added, resulting in the formation of white flocs. The medium was cooled to 20 °C, the bicarbonate buffer was added, and the pH was adjusted to 7.0 using 1 M anoxic HCl. Due to precipitate formation, the medium was transferred to 2 L Schott bottles and kept at room temperature for two days. It was then filtered through a 0.22 µm PES bottle-top filter inside a glovebox. Vitamins and trace element solutions were added under sterile conditions, and the pH was re-adjusted to 7.0. The prepared medium was aliquoted into sterile 250 mL Schott bottles, with each bottle containing a total volume of 100 mL, which included 5 mL of washed cyanobacteria and the respective FeCl₂ solution. During this process, a constant stream of N₂/CO₂ was maintained. Oxygen concentrations were measured, and any bottles with more than 3 µM oxygen were flushed with N₂/CO₂ again. Subsequently, FeCl₂ was added from an anoxic and sterile 100 µM or 1 M stock solution. The pH values were rechecked and remained stable at 7.0. Cyanobacteria were washed three times with ASW before being added to the medium as a 5% inoculum to achieve a final cell density of 10⁶ cells/mL. The experiments were initially fully anoxic, so all oxygen detected during the course of the experiments was produced solely by the inoculated cyanobacteria.

Quantification of Fe(II) and Fe(III) by the spectrophotometric Ferrozine assay

The ferrozine assay is a spectrophotometric method used to quantify the concentration of Fe(II). Samples were first dissolved and diluted in 1 M HCl to achieve a concentration of up to 1 mM Fe. For Fe(II) analysis, 80 µL of 1 M HCl was pipetted into a 96-well plate, followed by the addition of 20 µL of the sample. After a 15-min incubation, 100 µL of ferrozine solution (0.1 w/v) was added. The mixture was allowed to complex for 5 min before being analyzed using a Multiskan™ GO Microplate Spectrophotometer (Thermo Fisher Scientific, USA). Fe(II) is forming a purple complex with ferrozine molecules, which absorbs light at 562 nm. For total iron (Fe(tot)) determination, the procedure was modified by replacing the 1 M HCl with 80 µL of 10% (v/v) hydroxylammonium chloride (HAHCl) before adding the sample. This step was followed by a longer incubation period of 30 min to ensure the complete reduction of Fe(III) to Fe(II). To prevent degradation, both the ferrozine and HAHCl solutions were stored and incubated in the dark, as they are highly light-sensitive. Different standards between 0 and 1000 µM were used to calculate the respective iron concentration.

Quantification of dissolved silica by the molybdenum blue method

Dissolved silica, including monomeric silica and colloidal silicic acid (Si-Si-colloids), was quantified by the Molybdenum Blue Method70. The samples were filtered (0.22 µm, PES) and diluted in MQ to 1 mL. Subsequently, 40 µL of an ammonium heptamolybdate tetrahydrate (6.33 g) together with 50 mL of 4.5 M sulfuric acid solution were added to form a blue complex with the silica. To avoid complexing with phosphate, 40 µL of a 10 g/100 mL MQ oxalic acid solution was added. Lastly, 20 µL of 0.16 M ascorbic acid was added as reductant. After 30–60 min of incubation, the blue silica-molybdenum complexes were quantified at 810 nm. Different standards between 1 and 100 µM were used to calculate the silica concentration. Due to the filtration of the samples (0.22 µm) and removal of Si precipitates, the measured remaining silica fraction is named ‘dissolved silica’.

Oxygen quantification by optode sensors

Free dissolved oxygen was quantified using a luminophore optode foil from PreSens. Small 3 by 3 mm pieces of foil were affixed with silicone glue within the lower 2 cm of the bottles to ensure direct contact with the liquid phase. The interaction between the optode foil (luminophore) and oxygen molecules (quencher) results in an energy transfer that reduces the luminescence signal of the foil. The oxygen concentration in the liquid phase was then calculated using the Stern-Volmer equation. A temperature sensor was placed in a co-incubated Schott bottle under identical experimental conditions to account for temperature variations. Each container and optode foil setup underwent a fresh two-point calibration before measurements. The calibration included one data point at full oxygen saturation and another one under oxygen-free conditions. For the first calibration point, 100 mL of pure ASW medium was added to a 250 mL Schott bottle containing the optode foil and stirred vigorously until complete O₂ saturation was achieved. For the second calibration point, the reductant sodium dithionite (Na₂S₂O₃) was added to the same bottle to eliminate all remaining dissolved oxygen.

Intracellular ROS measurements

The membrane-permeable dye CM-H2DCFDA (5-(and-6)-chloromethyl-2′,7′-dichlorodihydrofluorescein diacetate, acetyl ester; Life Technologies GmbH) was used to quantify intracellular reactive oxygen species. Log-phase cultures were harvested via centrifugation and washed twice with TRIS buffer (pH 7.0). Cells were adjusted to a concentration of 5 × 108 cells/mL in TRIS buffer and degassed in the dark using a 90:10 N2/CO2 gas mixture. FeCl2 was added from anoxic stock solutions to achieve final concentrations of 0, 500, 2500, and 5000 µM We then added pH-adjusted silica from a metasilicate stock solution, pre-heated to 90 °C in a water bath to monomerize the silica, leading to a final concentrations of 0 and 2200 µM in duplicates in two different experimental runs. The suspensions were shaken for one hour. After incubation, cells were resuspended in fresh buffer containing 5 µM CM-H2DCFDA and incubated in the dark for 30 min. Following a second resuspension in fresh buffer, fluorescence emission at 519 nm (excitation at 490 ± 5 nm) was recorded using the infinite 200 PRO Fluorometer from Tecan Lifesciences. Background fluorescence from untreated cells and Fe minerals was subtracted to ensure accuracy.

Cell numbers by hemocytometry

The cell density was visually counted under a Leica CTR 5500 microscope using a Neubauer-improved hemocytometer from Hirschmann. Particularly, 10 µL of sample was pipetted between sample holder and coverslip. At a magnification of 400x, the cells were counted in the smallest squares until a final count of 100 cells. 180 µL cells were fixed by the addition of 20 µL 21% PFA solution. The minerals were dissolved by the addition 200 µL of 100 mM Fe(II)-EDAS solution and 600 µL of ca. 0.8 M oxalic acid The cell count per small square was then multiplied by the dilution factor and by 2.5 × 107 as volume factor to get the real cell count in mL/L.

MATLAB model from lab experimental data

We modeled our lab experimental data to quantify total oxygen production in the system in MATLAB (version R2023b), considering both dissolved oxygen and the contribution from the headspace. Based on the equilibrium partitioning between the aqueous solution in the head space based on the Henry-Law (see SI formula 9), oxygen production including the headspace is approximately 42 times greater than what is measurable in the dissolved liquid phase alone. Additionally, increasing cell density caused greater turbidity in the liquid phase, which led to shading effects reducing the overall photosynthetic rate per cell. We accounted for this by modeling the proportion of active versus inactive cells in the system. Detailed methods and the corresponding MATLAB code are provided in the Supplementary Information (SI).

MATLAB model to mimic the oxygen and Fe(II) distribution in the ancient ocean

To simulate oxygen distribution in ancient ocean water columns, we used the photosynthetic rate (rphoto) derived from our lab experiments under the most realistic conditions—500 µM Fe(II) and 2200 µM silica. We divided the experiment into two phases: Phase 1, when dissolved Fe²⁺ was still present, mimicking early ocean conditions, and Phase 2, after Fe²⁺ oxidation, to analyze the long-term bacterial response following Fe²⁺ exposure. From Phase 1, we calculated the daily oxygen production per cell (rphoto). According to reaction (5), one molecule of O₂ can oxidize four Fe²⁺ ions.

O2+4Fe2++8OH−↔4FeOOH+2H2O 5

We determined rphoto from the first two days of the experiment, during which no free oxygen was detected, and cyanobacteria were still in the lag phase without measurable growth, allowing the simplification that all produced oxygen was used for Fe(II) oxidation and the cell count stayed the same. Starting with 529 µM Fe(II) at day 0, 150 µM was oxidized over two days, corresponding to a rate of 75 µM/day Fe(II) oxidation, which equates to 18.75 µmol O₂/day based on reaction (5). Due to uncertainty in manual cell counts, we calculated rphoto for two different cell densities observed in our experiments: 6.48 × 10⁶ cells/mL (experimental cells 1) and 1.6 × 10⁷ cells/mL (experimental cells 2). Reported early ocean cyanobacterial concentrations range between 10⁵cells/mL (target cells 1) and 10⁶ cells/mL (target cells 2), so we modeled four scenarios combining these experimental and target cell densities to calculate rphoto values (see Table 3) using formula 6. Based on the study of Swanner and colleagues12, we further estimated oxygen distribution in the early ocean water column for upwelling rates of 4, 94, and 473 m/year.

18.75μMday*targetcellsexperimentalcells=rphoto 6

Table 3.

Photosynthetic rate (oxygen production per cell/ day) of the lab experiments calculated from 6.48 × 10⁶ cells/mL (‘experimental cells 1’) and 1.6 × 10⁷ cells/mL (‘experimental cells 2’) for 10⁵ cells/mL (target cells 1) and 10⁶ cells/mL (target cells 2) in the early ocean

rphoto Experimental cells 1 Experimental cells 2
Target cells 1 0.289 µM/day 2.89 µM/day
Target cells 2 0.112 µM/day 1.12 µM/day
Open in a new tab

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Supplementary information

Supplementary information (787.2KB, pdf)
Peer Review file (978.5KB, pdf)
Reporting Summary (1.6MB, pdf)

Acknowledgements

The authors would like to thank M. Herzog for providing the data of her Bachelor Thesis. J. López Rivoldi and M. Mergenthaler are acknowledged for their experimental support. The authors would further like to thank F. Schädler and L. Grimm for maintenance and help in the laboratory. Dr. J. P. Duda, Dr. E. Runge, Dr. A. Illin, D. Gutierrez Rueda and Y. Li, are acknowledged for fruitful discussions on experimental design and data interpretation. We would like to thank the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) for supporting grants for this project to AK, project ID AOBJ: 669932.

Author contributions

C.D., A.K. and K.K designed the experiments. M.S., A.K. and K.K. acquired the funding. C.D. performed the experiments and data analysis. O.C. provided the Matlab codes and modeled the lab data. C.D., A.K. and K.K. wrote and revised the paper, O.C. and M.S. reviewed and edited the paper.

Peer review

Peer review information

: Nature Communications thanks Kelsey Moore and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Funding

Open Access funding enabled and organized by Projekt DEAL.

Data availability

The raw data generated in this study have been deposited in the Mendeley database under 10.17632/wp96jybytw.2.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Change history

4/9/2026

Since the version of the article initially published, the Mendeley DOI in the Data availability section has been updated to 10.17632/wp96jybytw.2 in the HTML and PDF versions of the article. Additionally, the Reporting Summary was originally missing and has now been uploaded online.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-026-69826-x.

References

  • 1.Boden, J. S., Konhauser, K. O., Robbins, L. J. & Sánchez-Baracaldo, P. Timing the evolution of antioxidant enzymes in cyanobacteria. Nat. Commun.12, 4742 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Fournier, G. P. et al. The Archean origin of oxygenic photosynthesis and extant cyanobacterial lineages. Proc. R. Soc. B Biol. Sci.288, 20210675 (2021). [Google Scholar]
  • 3.Poulton, S. W. et al. A 200-million-year delay in permanent atmospheric oxygenation. Nature592, 232–236 (2021). [DOI] [PubMed] [Google Scholar]
  • 4.Kasting, J. F., Schopf, J. W. & Klein, C. Models relating to Proterozoic atmospheric and oceanic chemistry. The Proterozoic Biosphere: A Multidisciplinary Study, 1185–1187 (1992).
  • 5.Anbar, A. D. et al. A whiff of oxygen before the Great Oxidation Event? Science317, 1903–1906 (2007). [DOI] [PubMed] [Google Scholar]
  • 6.Czaja, A. D. et al. Evidence for free oxygen in the Neoarchean ocean based on coupled iron–molybdenum isotope fractionation. Geochim. Cosmochim. Acta86, 118–137 (2012). [Google Scholar]
  • 7.Olson, S. L., Kump, L. R. & Kasting, J. F. Quantifying the areal extent and dissolved oxygen concentrations of Archean oxygen oases. Chem. Geol.362, 35–43 (2013). [Google Scholar]
  • 8.Kendall, B., Lyons, T. W., Kaufman, A. J., Poulton, S. W. & Anbar, A. D. Pervasive oxygenation along late Archaean ocean margins. Nat. Geosci.3, 647–652 (2010). [Google Scholar]
  • 9.Chen, X. et al. Transient marine bottom water oxygenation on continental shelves by 2.65 billion years ago. Nat. Geosci.18, 423–429 (2025). [Google Scholar]
  • 10.Liang, X., Stüeken, E. E., Alessi, D. S., Konhauser, K. O. & Li, L. A seawater oxygen oscillation recorded by iron formations prior to the Great Oxidation Event. Nat. Geosci.18, 417–422 (2025). [Google Scholar]
  • 11.Swanner, E. D. et al. Physiology, Fe (II) oxidation, and Fe mineral formation by a marine planktonic cyanobacterium grown under ferruginous conditions. Front. Earth Sci.3, 60 (2015).
  • 12.Swanner, E. D. et al. Modulation of oxygen production in Archean oceans by episodes of Fe(II) toxicity. Nat. Geosci.8, 126–130 (2015). [Google Scholar]
  • 13.Konhauser, K. O. et al. Iron formations: A global record of Neoarchaean to Palaeoproterozoic environmental history. Earth-Sci. Rev.172, 140–177 (2017). [Google Scholar]
  • 14.Holland, H. D. The oceans: a possible source of iron in iron-formations. Econ. Geol.68, 1169–1172 (1973). [Google Scholar]
  • 15.Morris, R. C. Genetic modelling for banded iron-formation of the Hamersley Group, Pilbara Craton, Western Australia. Precambrian Res.60, 243–286 (1993). [Google Scholar]
  • 16.Derry, L. A. Causes and consequences of mid-Proterozoic anoxia. Geophys. Res. Lett.42, 8538–8546 (2015). [Google Scholar]
  • 17.Jiang, C. Z. & Tosca, N. J. Fe(II)-carbonate precipitation kinetics and the chemistry of anoxic ferruginous seawater. Earth Planet. Sci. Lett.506, 231–242 (2019). [Google Scholar]
  • 18.Maliva, R. G., Knoll, A. H. & Simonson, B. M. Secular change in the Precambrian silica cycle: insights from chert petrology. Geol. Soc. Am. Bull.117, 835–845 (2005). [Google Scholar]
  • 19.Siever, R. The silica cycle in the Precambrian. Geochim. Cosmochim. Acta56, 3265–3272 (1992). [Google Scholar]
  • 20.Konhauser, K. O., Lalonde, S. V., Amskold, L. & Holland, H. D. Was there really an Archean phosphate crisis? Science315, 1234 (2007). [DOI] [PubMed] [Google Scholar]
  • 21.Fischer, W. W. et al. Isotopic constraints on the Late Archean carbon cycle from the Transvaal Supergroup along the western margin of the Kaapvaal Craton, South Africa. Precambrian Res.169, 15–27 (2009). [Google Scholar]
  • 22.Percak-Dennett, E. M. et al. Iron isotope fractionation during microbial dissimilatory iron oxide reduction in simulated Archaean seawater. Geobiology9, 205–220 (2011). [DOI] [PubMed] [Google Scholar]
  • 23.Konhauser, K. O. et al. Could bacteria have formed the Precambrian banded iron formations? Geology30, 1079 (2002). [Google Scholar]
  • 24.Kappler, A., Pasquero, C., Konhauser, K. O. & Newman, D. K. Deposition of banded iron formations by anoxygenic phototrophic Fe(II)-oxidizing bacteria. Geology33, 865–868 (2005). [Google Scholar]
  • 25.Cairns-Smith, A. G. Precambrian solution photochemistry, inverse segregation, and banded iron formations. Nature276, 807–808 (1978). [Google Scholar]
  • 26.Braterman, P. S., Cairns-Smith, A. G. & Sloper, R. W. Photo-oxidation of hydrated Fe2+—significance for banded iron formations. Nature303, 163–164 (1983). [Google Scholar]
  • 27.Rasmussen, B., Meier, D. B., Krapež, B. & Muhling, J. R. Iron silicate microgranules as precursor sediments to 2.5-billion-year-old banded iron formations. Geology41, 435–438 (2013). [Google Scholar]
  • 28.Tosca, N. J., Guggenheim, S. & Pufahl, P. K. An authigenic origin for Precambrian greenalite: Implications for iron formation and the chemistry of ancient seawater. Bulletin128, 511–530 (2016). [Google Scholar]
  • 29.Dreher, C. L. et al. Microbial processes during deposition and diagenesis of banded iron formations. Palaontologische Z.95, 593–610 (2021). [Google Scholar]
  • 30.Dismukes, G. C. et al. The origin of atmospheric oxygen on Earth: The innovation of oxygenic photosynthesis. Proc. Natl. Acad. Sci.98, 2170–2175 (2001). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Schirrmeister, B. E., Gugger, M. & Donoghue, P. C. J. Cyanobacteria and the Great Oxidation Event: evidence from genes and fossils. Palaeontology58, 769–785 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.King, D. W., Lounsbury, H. A. & Millero, F. J. Rates and mechanism of Fe(II) oxidation at nanomolar total iron concentrations. Environ. Sci. Technol.29, 818–824 (1995). [DOI] [PubMed] [Google Scholar]
  • 33.Kappler, A. et al. An evolving view on biogeochemical cycling of iron. Nat. Rev. Microbiol.19, 360–374 (2021). [DOI] [PubMed] [Google Scholar]
  • 34.Cabiscol, E., Piulats, E., Echave, P., Herrero, E. & Ros, J. Oxidative stress promotes specific protein damage in saccharomyces cerevisiae. J. Biol. Chem.275, 27393–27398 (2000). [DOI] [PubMed] [Google Scholar]
  • 35.Morris, J. J., Johnson, Z. I., Szul, M. J., Keller, M. & Zinser, E. R. Dependence of the cyanobacterium prochlorococcus on hydrogen peroxide scavenging microbes for growth at the ocean’s surface. PLOS ONE6, e16805 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Dickson, K. B. & Zhou, J. Role of reactive oxygen species and iron in host defense against infection. Front. Biosci. -Landmark25, 1600–1616 (2020). [Google Scholar]
  • 37.Dreher, C. L. et al. Impact of silica on the identity of minerals formed in Archean oceans during Fe cycling by cyanobacteria and iron(III)-reducing bacteria. Geochim. Cosmochim. Acta404, 202–222 (2025). [Google Scholar]
  • 38.Trabucho Alexandre, J. et al. The mid‐Cretaceous North Atlantic nutrient trap: black shales and OAEs. Paleoceanography25 (2010).
  • 39.Benson, B. B. & Krause, D. Jr. The concentration and isotopic fractionation of oxygen dissolved in freshwater and seawater in equilibrium with the atmosphere. Limnol. Oceanogr.29, 620–632 (1984). [Google Scholar]
  • 40.Weiss, R. F. The solubility of nitrogen, oxygen and argon in water and seawater. Deep Sea Res. Oceanogr. Abstr.17, 721–735 (1970). [Google Scholar]
  • 41.Hinz, I. L., Rossi, L., Ma, C. & Johnson, J. E. Simulated diagenesis of the iron-silica precipitates in banded iron formations. Am. Mineral.108, 1732–1753 (2023). [Google Scholar]
  • 42.Hinz, I. L., Nims, C., Theuer, S., Templeton, A. S. & Johnson, J. E. Ferric iron triggers greenalite formation in simulated Archean seawater. Geology49, 905–910 (2021). [Google Scholar]
  • 43.Halevy, I., Aleksker, M., Schuster, E. M., Popovitz-Biro, R. & Feldman, Y. A key role for green rust in the Precambrian oceans and the genesis of iron formations. Nat. Geosci.10, 135–139 (2017). [Google Scholar]
  • 44.Krissansen-Totton, J., Arney, G. N. & Catling, D. C. Constraining the climate and ocean pH of the early Earth with a geological carbon cycle model. Proc. Natl. Acad. Sci. Usa.115, 4105–4110 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Wydrzynski, T. J. & Hillier, W. Molecular Solar Fuels (Royal Society of Chemistry, 2011).
  • 46.Hegler, F., Posth, N. R., Jiang, J. & Kappler, A. Physiology of phototrophic iron(II)-oxidizing bacteria: implications for modern and ancient environments: Physiology of phototrophic iron(II)-oxidizing bacteria. FEMS Microbiol. Ecol.66, 250–260 (2008). [DOI] [PubMed] [Google Scholar]
  • 47.Nikeleit, V., Roth, L., Maisch, M., Kappler, A. & Bryce, C. Phototrophic Fe(II) oxidation benefits from light/dark cycles. Environ. Microbiol. Rep.16, e13239 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Hauruseu, D. & Koblížek, M. Influence of light on carbon utilization in aerobic anoxygenic phototrophs. Appl. Environ. Microbiol.78, 7414–7419 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Zhi, R. et al. Effects of light-dark cycles on photosynthetic bacteria wastewater treatment and valuable substances production. Bioresour. Technol.274, 496–501 (2019). [DOI] [PubMed] [Google Scholar]
  • 50.Stöckel, J. et al. Global transcriptomic analysis of Cyanothece 51142 reveals robust diurnal oscillation of central metabolic processes. Proc. Natl. Acad. Sci.105, 6156–6161 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Saha, R. et al. Diurnal regulation of cellular processes in the cyanobacterium Synechocystis sp. Strain PCC 6803: Insights from transcriptomic, fluxomic, and physiological analyses. mBio7, e00464–16 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Muhetaer, G. et al. Effects of light intensity and exposure period on the growth and stress responses of two cyanobacteria species: pseudanabaena galeata and microcystis aeruginosa. Water12, 407 (2020). [Google Scholar]
  • 53.Vu, T. T. et al. Computational evaluation of Synechococcus sp. PCC 7002 metabolism for chemical production. Biotechnol. J.8, 619–630 (2013). [DOI] [PubMed] [Google Scholar]
  • 54.Moffett, J. W. & Brand, L. E. Production of strong, extracellular Cu chelators by marine cyanobacteria in response to Cu stress. Limnol. Oceanogr.41, 388–395 (1996). [Google Scholar]
  • 55.Hood, R. D., Higgins, S. A., Flamholz, A., Nichols, R. J. & Savage, D. F. The stringent response regulates adaptation to darkness in the cyanobacterium Synechococcus elongatus. Proc. Natl. Acad. Sci.113, E4867–E4876 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.Ungerer, J., Wendt, K. E., Hendry, J. I., Maranas, C. D. & Pakrasi, H. B. Comparative genomics reveals the molecular determinants of rapid growth of the cyanobacterium Synechococcus elongatus UTEX 2973. Proc. Natl. Acad. Sci.115, E11761–E11770 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Kuan, D., Duff, S., Posarac, D. & Bi, X. Growth optimization of Synechococcus elongatus PCC7942 in lab flasks and a 2-D photobioreactor. Can. J. Chem. Eng.93, 640–647 (2015). [Google Scholar]
  • 58.Herrmann, A. J. & Gehringer, M. M. A low-cost automized anaerobic chamber for long-term growth experiments and sample handling. HardwareX, 10, e00237 (2021).
  • 59.Planavsky, N. J. et al. Evidence for oxygenic photosynthesis half a billion years before the Great Oxidation Event. Nat. Geosci.7, 283–286 (2014). [Google Scholar]
  • 60.Mel’nik, Y. P. Precambrian Banded-Iron-Formations. (Elsevier, 1982).
  • 61.Morris, R. C. & Horwitz, R. C. The origin of the iron-formation-rich Hamersley Group of Western Australia — deposition on a platform. Precambrian Res21, 273–297 (1983). [Google Scholar]
  • 62.Canfield, D. E. The early history of atmospheric oxygen: Homage to Robert M. Garrels. Annu. Rev. Earth Planet. Sci.33, 1–36 (2005). [Google Scholar]
  • 63.Garcia-Pichel, F. et al. Timing the evolutionary advent of cyanobacteria and the later Great Oxidation Event using gene phylogenies of a sunscreen. mBio10, e00561-19 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 64.Schirrmeister, B. E., De Vos, J. M., Antonelli, A. & Bagheri, H. C. Evolution of multicellularity coincided with increased diversification of cyanobacteria and the Great Oxidation Event. Proc. Natl. Acad. Sci.110, 1791–1796 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 65.Szeinbaum, N., Toporek, Y. J., Reinhard, C. T. & Glass, J. B. Microbial helpers allow cyanobacteria to thrive in ferruginous waters. Geobiology19, 510–520 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 66.He, H. et al. An abiotic source of Archean hydrogen peroxide and oxygen that pre-dates oxygenic photosynthesis. Nat. Commun.12, 6611 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 67.Wu, X. et al. Geodynamic oxidation of Archean terrestrial surfaces. Commun. Earth Environ.4, 1–9 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68.Kollias, K., Mylona, E., Adam, K., Papassiopi, N. & Xenidis, A. Suppression of pyrite oxidation by surface silica coating. J. Geosci. Environ. Prot.2, 37–43 (2014). [Google Scholar]
  • 69.Kinsela, A. S. et al. Influence of dissolved silicate on rates of Fe(II) oxidation. Environ. Sci. Technol.50, 11663–11671 (2016). [DOI] [PubMed] [Google Scholar]
  • 70.Fanning, K. A. & Pilson, M. E. On the spectrophotometric determination of dissolved silica in natural waters. Anal. Chem.45, 136–140 (1973). [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary information (787.2KB, pdf)
Peer Review file (978.5KB, pdf)
Reporting Summary (1.6MB, pdf)

Data Availability Statement

The raw data generated in this study have been deposited in the Mendeley database under 10.17632/wp96jybytw.2.

Articles from Nature Communications are provided here courtesy of Nature Publishing Group

RESOURCES